Therapeutic predictive response signatures and uses thereof in cancer patients

Abstract

Provided herein are pharmacological response activation signatures for use in methods and compositions for predicting the response of a subject suffering from cancer to treatment with select inhibitors.

Description

Attorney Docket No. GNCN-024 / 03WO 320289-2158 THERAPEUTIC PREDICTIVE RESPONSE SIGNATURES AND USES THEREOF IN CANCER PATIENTS CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority from U.S. Provisional Application No. 63 / 658,129 filed June 10, 2024, which is incorporated by reference herein in its entirety for all purposes. FIELD

[0002] The present invention relates to methods, systems and compositions for determining gene expression based therapeutic predictive response signatures on a sample obtained from a subject suffering from cancer. The present invention also relates to methods, systems and compositions for predicting response of subject suffering from cancer to specific cancer pharmacotherapies by determining gene expression signatures from a sample obtained from the subject suffering from cancer. BACKGROUND

[0003] Current patient selection criteria for treating cancer with certain therapeutics rely on mutation panel testing (e.g., Foundation One® or equivalent) or the identification of one or more specific genetic variants, but not all patients respond to the specific therapies selected by said mutational panel testing or genetic variant testing. Moreover, mutational panel testing often does not detect less prevalent mutations that, if identified, and treated with a specific therapy, could result in patient benefit. For example, current patient selection criteria for selecting patients for EGFR inhibitor therapy relies on mutation panel testing (e.g., Foundation One® or equivalent), MET inhibitor therapy patient selection relies on detection of MET exon 14 skipping (METex14) (e.g., Foundation One® or equivalent), KRAS inhibitor therapy patient selection relies on detection of KRAS G12C (e.g., Foundation One® or equivalent), while BRAF / MEK inhibitor therapy and PIK3CA inhibitor therapy patient selection relies on mutation panel testing for BRAF-V600E mutations or PIK3CA mutations (e.g., Foundation One® or equivalent), respectively, but not all patients selected for these various therapies respond to them and often any panels used do not detect less prevalent mutations that, if identified, could be used to select patients for various therapies and, thus, could result in patient benefit. Additionally, current patient selection criteria have not been fully established for certain types of therapies (e.g.,Attorney Docket No. GNCN-024 / 03WO 320289-2158 ERBB2 inhibitor therapy or ERBB3 inhibitor therapy) and, thus, one or more biomarkers to select for patients with said therapies (e.g., active ERBB2 or ERBB3 signaling) could result in patient benefit.

[0004] To address these challenges, provided herein are myriad of predictive response gene signatures and their uses for the aforementioned pharmacotherapies. These predictive response gene signatures address the need for efficient methods for improved patient population classification that could inform prognosis, drug response and patient management based on underlying genomic and biologic tumor characteristics. SUMMARY

[0005] In one aspect, provided herein is a method of determining whether a patient suffering from cancer is likely to respond to treatment with a first type of therapy and / or whether the patient suffering from cancer is unlikely to respond to treatment with a second type of therapy, the method comprising, determining a predictive response signature of a sample obtained from a patient suffering from cancer; and based on the predictive response signature, assessing whether the patient is likely to respond to treatment with the first type of therapy and / or the patient is not likely to respond to treatment with the second type of therapy, wherein a positive predictive response signature predicts that the patient is likely to respond to the treatment with the first type of therapy and / or the patient is not likely to respond to treatment a second type of therapy.

[0006] In another aspect, provided herein is a method for selecting a patient suffering from cancer for treatment with a first type of therapy and / or not selecting the patient suffering from cancer for treatment with a second type of therapy, the method comprising, determining a predictive response signature of a sample obtained from a patient suffering from cancer; and selecting the patient for treatment with the first type of therapy and / or not selecting the patient for treatment with the second type of therapy if the predictive response signature is positive.

[0007] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining the expression levels of a plurality of classifier biomarkers selected from Table 2 and the predictive response signature is an epidermal growth factor receptor (EGFR) predictive response signature. In some cases, the positive EGFR predictive response signature indicates that the patient possesses one or more genetic variants in the epidermal growth factor receptor (egfr) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selectedAttorney Docket No. GNCN-024 / 03WO 320289-2158 from Table 2 to an expression level of the plurality of classifier biomarkers selected from Table 2 in at least one sample training set, wherein the at least one sample training set is from a reference EGFR mutation-containing cancer sample, or is from a reference EGFR mutation-free cancer sample; and classifying the sample as having a positive EGFR predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 2 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 2 from the at least one training set; and classifying the sample as possessing a positive EGFR predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference EGFR mutation-containing cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample. In some cases, the at least one training set is from a reference EGFR mutation-containing cancer sample and from a reference EGFR mutation-free cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward EGFR. In some cases, the EGFR inhibitor is a tyrosine kinase inhibitor. In some cases, theAttorney Docket No. GNCN-024 / 03WO 320289-2158 EGFR inhibitor is a selective tyrosine kinase inhibitor. In some cases, the EGFR inhibitor is a non- selective tyrosine kinase inhibitor. In some cases, the EGFR inhibitor is selected from the group consisting of erlotinib (OSI-744), poziotinib (HM781-36B), osimertinib (AZD9291), AG-490 (Tyrphostin B42), afatinib (BIBW2992), gefitinib (ZD1839), lapatinib (GW-572016), rociletinib (CO-1686), neratinib, lucitanib (E3810), dacomitinib, mobocertinib, vandetanib, canertinib (CI- 1033), BDTX-189, epertinib, AEE788, CUDC-101, pelitinib, sapitinib, varlitinib, pyrotinib, TAK- 285, AC480, tyrophostin AG-528 and any combination thereof. In some cases, the EGFR inhibitor is canertinib (CI-1033). In some cases, the EGFR inhibitor is an antibody or antibody-conjugate. In some cases, the EGFR inhibitor is cetuximab, panitumumab or necitumumab. In some cases, the second type of therapy is a MEK inhibitor or an ERK inhibitor.

[0008] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 4 and the predictive response signature is a MET proto- oncogene, receptor tyrosine kinase (MET) predictive response signature. In some cases, the positive MET predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the MET proto-oncogene, receptor tyrosine kinase (met) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 4 to an expression level of the plurality of classifier biomarkers selected from Table 4 in at least one sample training set, wherein the at least one sample training set is from a reference MET mutation-containing or amplification-containing cancer sample, or is from a reference MET mutation-free cancer sample; and classifying the sample as having a positive MET predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 4 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 4 from the at least one training set; and classifying the sample as possessing a positive MET predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of theAttorney Docket No. GNCN-024 / 03WO 320289-2158 plurality of classifier biomarkers selected from Table 4 from the reference MET mutation- containing or amplification-containing cancer sample. In some cases, the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and from a reference MET mutation-free cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation-containing or amplification- containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization- based analyses. In some cases, the first type of therapy shows inhibitory activity toward MET. In some cases, the MET inhibitor is a tyrosine kinase inhibitor. In some cases, the MET inhibitor is a selective tyrosine kinase inhibitor. In some cases, the MET inhibitor is a non-selective tyrosine kinase inhibitor. In some cases, the MET inhibitor is selected from the group consisting of crizotinib, capmatinib, tepotinib, savolitinib, cabozantinib, glesatinib merestinib and any combination thereof. In some cases, the MET inhibitor is an antibody or antibody-conjugate. In some cases, the MET inhibitor is emibetuzumab, Rilotumumab, Ficlatuzumab, TAK-701, Onartuzumab, ARGX-111, or EM1-mAb. In some cases, the second type of therapy is an mTOR inhibitor, a PI3K inhibitor, an EGFR inhibitor or an AKT inhibitor.

[0009] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 6 and the predictive response signature is a Kirsten Rat Sarcoma viral oncogene homolog (KRAS) predictive response signature. In some cases, theAttorney Docket No. GNCN-024 / 03WO 320289-2158 positive KRAS predictive response signature indicates that the patient possesses one or more genetic variants in the Kirsten Rat Sarcoma viral oncogene homolog (kras) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 6 to an expression level of the plurality of classifier biomarkers selected from Table 6 in at least one sample training set, wherein the at least one sample training set is from a reference KRAS mutation-containing cancer sample, or is from a reference KRAS mutation-free cancer sample; and classifying the sample as having a positive KRAS predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 6 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 6 from the at least one training set; and classifying the sample as possessing a positive KRAS predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference KRAS mutation-containing cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation- containing cancer sample. In some cases, the at least one training set is from a reference KRAS mutation-containing cancer sample and from a reference KRAS mutation-free cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 6. In some cases, the determining the expression levels of the plurality ofAttorney Docket No. GNCN-024 / 03WO 320289-2158 classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward KRAS, MEK1, MEK2 or extracellular signal-regulated kinase (ERK). In some cases, the KRAS inhibitor is selected from the group consisting of Adagrasib, JQ443, GDC-6036, BI 1823911, JNJ-74699157, MK-1084, SCH-53239,SHP099, JAB-3068, RMC-4550, TNO155 and any combination thereof. In some cases, the MEK inhibitor is selected from the group consisting of binimetinib, cobimetinib, selumetinib, trametinib, CI-1040, TAK-733, pimasertib, and any combination thereof. In some cases, the ERK inhibitor is selected from the group consisting of PD0325901, ASN007, ulixertinib (BVD-523), CC-9003,ERK5-IN-2, XMD8-92, DEL-22379 and any combination thereof. In some cases, the second type of therapy is an ERBB2 inhibitor, ERBB4 inhibitor or EGFR inhibitor.

[0010] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 8 and the predictive response signature is a B-Raf protooncogene (BRAF) predictive response signature. In some cases, the positive BRAF predictive response signature indicates that the patient possesses one or more genetic variants in the B-Raf protooncogene (BRAF) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 8 to an expression level of the plurality of classifier biomarkers selected from Table 8 in at least one sample training set, wherein the at least one sample training set is from a reference BRAF mutation-containing cancer sample, or is from a reference BRAF mutation-free cancer sample; and classifying the sample as having a positive BRAF predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 8 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 8 from the at least one training set; and classifying the sample as possessing a positive BRAF predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a referenceAttorney Docket No. GNCN-024 / 03WO 320289-2158 BRAF mutation-containing cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample. In some cases, the at least one training set is from a reference BRAF mutation-containing cancer sample and from a reference BRAF mutation-free cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward BRAF. In some cases, the BRAF inhibitor is selected from the group consisting of vemurafenib, dabrafenib, sorafenib, encorafenib, PLX4032 and any combination thereof.

[0011] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 10 and the predictive response signature is a Erb-B2 receptor tyrosine kinase 2 (ERBB2) predictive response signature. In some cases, the positive ERBB2 predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifierAttorney Docket No. GNCN-024 / 03WO 320289-2158 biomarkers selected from Table 10 to an expression level of the plurality of classifier biomarkers selected from Table 10 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample, or is from a reference ERBB2 mutation-free cancer sample; and classifying the sample as having a positive ERBB2 predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 10 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 10 from the at least one training set; and classifying the sample as possessing a positive ERBB2 predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference ERBB2 mutation- containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification-containing cancer sample. In some cases, the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and from a reference ERBB2 mutation-free cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkersAttorney Docket No. GNCN-024 / 03WO 320289-2158 from Table 10. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward ERBB2. In some cases, the ERBB2 inhibitor is selected from the group consisting of lapatinib, afatinib, AST1306, AEE788, CP724, CP714, CUDC101, TAK285, dacomitinib, pelitinib, AC480, canertinib, tucatinib (irbinitinib, ONT-380),pyrotinib, neratinib, tyrophostin AG-258, (-)-epigallocatechin gallate, mobocertinib, trastuzumab, zenocutuzumab and any combination thereof.

[0012] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 12 and the predictive response signature is a Erb-B3 receptor tyrosine kinase 3 (ERBB3) predictive response signature. In some cases, the positive ERBB3 predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 12 to an expression level of the plurality of classifier biomarkers selected from Table 12 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample, or is from a reference ERBB3 mutation-free cancer sample; and classifying the sample as having a positive ERBB3 predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 12 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 12 from the at least one training set; and classifying the sample as possessing a positive ERBB3 predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference ERBB3 mutation- containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containingAttorney Docket No. GNCN-024 / 03WO 320289-2158 or amplification-containing cancer sample. In some cases, the at least one training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample and from a reference ERBB3 mutation-free cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward ERBB3. In some cases, the ERBB3 inhibitor is selected from the group consisting of sapitinib (AZD8931), elgemtumab (LJM716),antibodies BCD090-P1, BCD090-M2, and BCD090-M456, and any combination thereof.

[0013] In some cases, the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 14 and the predictive response signature is a Phosphoinositide 3-Kinase CA (PIK3CA) predictive response signature. In some cases, the positive ERBB3 predictive response signature indicates that the patient possesses one or more genetic variants in the Phosphoinositide 3-Kinase CA (PIK3CA) gene. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 14 to an expression level of the plurality of classifier biomarkers selected from Table 14 in at least one sample training set, wherein the at least one sample training set is from a reference PIK3CA mutation-containing cancer sample, or is from a reference PIK3CA mutation-free cancerAttorney Docket No. GNCN-024 / 03WO 320289-2158 sample; and classifying the sample as having a positive PIK3CA predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 14 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 14 from the at least one training set; and classifying the sample as possessing a positive PIK3CA predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference PIK3CA mutation-containing cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation- containing cancer sample. In some cases, the at least one training set is from a reference PIK3CA mutation-containing cancer sample and from a reference PIK3CA mutation-free cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward PIK3CA. In some cases, the PIK3CA inhibitor is selected from the group consisting of alpelisib, idelalisib, duvelisib, copanlisib,Attorney Docket No. GNCN-024 / 03WO 320289-2158 apitolisib, bimiralisib, eganelisib, fimepinostat, gedatolisib, linperlisib, nemiralisib, pictilisib, pilaralisib, samotolisib, seletalisib, serabelisib, sonolisib, tenalisib, voxtalisib, AMG 319, AZD8186, GSK2636771, SF1126, acalisib, omipalisib, AZD8835, CAL263, GSK1059615, MEN1611, PWT33597, TG100-115, ZSTK474, GDC0077 and any combination thereof.

[0014] In another aspect, provided herein is a method of treating cancer in a patient, the method comprising: measuring the expression level of a plurality of classifier biomarkers in a sample obtained from a patient suffering from cancer, wherein the plurality of classifier biomarkers are selected from classifier biomarkers listed in Table 2, Table 4, Table 6, Table 8, Table 10, Table 12, Table 14 or any combination thereof, wherein the measured expression levels of the plurality of classifier biomarkers provide a target gene specific activation signature for the sample; and administering a specific therapy to the patient based on presence of a positive target gene specific activation signature, wherein the positive target gene specific activation signature is indicative of presence of one or more genetic variants in the target gene.

[0015] In some cases, the plurality of classifier biomarkers are selected from Table 2, the target gene is epidermal growth factor receptor (egfr) gene and the target gene specific activation signature is an epidermal growth factor receptor (EGFR) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 2 to an expression level of the plurality of classifier biomarkers selected from Table 2 in at least one sample training set, wherein the at least one sample training set is from a reference EGFR mutation-containing cancer sample, or is from a reference EGFR mutation-free cancer sample; and classifying the sample as having a positive EGFR predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 2 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 2 from the at least one training set; and classifying the sample as possessing a positive EGFR predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference EGFR mutation-containing cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-Attorney Docket No. GNCN-024 / 03WO 320289-2158 containing cancer sample. In some cases, the at least one training set is from a reference EGFR mutation-containing cancer sample and from a reference EGFR mutation-free cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the specific therapy shows inhibitory activity toward EGFR. In some cases, the EGFR inhibitor is a tyrosine kinase inhibitor. In some cases, the EGFR inhibitor is a selective tyrosine kinase inhibitor. In some cases, the EGFR inhibitor is a non-selective tyrosine kinase inhibitor. In some cases, the EGFR inhibitor is selected from the group consisting of erlotinib (OSI-744), poziotinibgefitinib (ZD1839), lapatinib (GW-572016), rociletinib (CO-1686), neratinib, lucitanib (E3810), dacomitinib, mobocertinib, vandetanib, canertinib (CI-1033), BDTX-189, epertinib, AEE788, CUDC-101, pelitinib, sapitinib, varlitinib, pyrotinib, TAK-285, AC480, tyrophostin AG-528 and any combination thereof. In some cases, the EGFR inhibitor is canertinib (CI-1033). In some cases, the EGFR inhibitor is an antibody or antibody-conjugate. In some cases, the EGFR inhibitor is cetuximab, panitumumab or necitumumab.

[0016] In some cases, the plurality of classifier biomarkers are selected from Table 4, the target gene is MET proto-oncogene, receptor tyrosine kinase (met) gene and the target gene specific activation signature is a MET proto-oncogene, receptor tyrosine kinase (MET) predictive response signature. In some cases, the method further comprises comparing the expression levelsAttorney Docket No. GNCN-024 / 03WO 320289-2158 of the plurality of classifier biomarkers selected from Table 4 to an expression level of the plurality of classifier biomarkers selected from Table 4 in at least one sample training set, wherein the at least one sample training set is from a reference MET mutation-containing or amplification- containing cancer sample, or is from a reference MET mutation-free cancer sample; and classifying the sample as having a positive MET predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 4 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 4 from the at least one training set; and classifying the sample as possessing a positive MET predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation- containing or amplification-containing cancer sample. In some cases, the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and from a reference MET mutation-free cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation-containing or amplification- containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performingAttorney Docket No. GNCN-024 / 03WO 320289-2158 RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization- based analyses. In some cases, the specific therapy shows inhibitory activity toward MET. In some cases, the MET inhibitor is a tyrosine kinase inhibitor. In some cases, the MET inhibitor is a selective tyrosine kinase inhibitor. In some cases, the MET inhibitor is a non-selective tyrosine kinase inhibitor. In some cases, the MET inhibitor is selected from the group consisting of crizotinib, capmatinib, tepotinib, savolitinib, cabozantinib, glesatinib merestinib and any combination thereof. In some cases, the MET inhibitor is an antibody or antibody-conjugate. In some cases, the MET inhibitor is emibetuzumab, Rilotumumab, Ficlatuzumab, TAK-701, Onartuzumab, ARGX-111, or EM1-mAb.

[0017] In some cases, the plurality of classifier biomarkers are selected from Table 6, the target gene is Kirsten Rat Sarcoma viral oncogene homolog (kras) gene and the target gene specific activation signature is a Kirsten Rat Sarcoma viral oncogene homolog (KRAS) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 6 to an expression level of the plurality of classifier biomarkers selected from Table 6 in at least one sample training set, wherein the at least one sample training set is from a reference KRAS mutation-containing cancer sample, or is from a reference KRAS mutation-free cancer sample; and classifying the sample as having a positive KRAS predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 6 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 6 from the at least one training set; and classifying the sample as possessing a positive KRAS predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference KRAS mutation-containing cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample. In some cases, the at least one training set is from a reference KRAS mutation-containing cancer sample and from a reference KRAS mutation-free cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6Attorney Docket No. GNCN-024 / 03WO 320289-2158 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 6. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the specific therapy shows inhibitory activity toward KRAS, MEK1, MEK2 or extracellular signal-regulated kinase (ERK). In some cases, the KRAS inhibitor is selected from the group consisting of Adagrasib,tipifarnib, SHP099, JAB-3068, RMC-4550, TNO155 and any combination thereof. In some cases, the MEK inhibitor is selected from the group consisting of binimetinib, cobimetinib, selumetinib, trametinib, CI-1040, TAK-733, pimasertib, and any combination thereof. In some cases, the ERK inhibitor is selected from the group consisting of PD0325901, ASN007, ulixertinib (BVD-523),ERK5-IN-1, ERK5-IN-2, XMD8-92, DEL-22379 and any combination thereof.

[0018] In some cases, the plurality of classifier biomarkers are selected from Table 8, the target gene is B-Raf protooncogene (BRAF) gene and the target gene specific activation signature is a B-Raf protooncogene (BRAF) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 8 to an expression level of the plurality of classifier biomarkers selected from Table 8 in at least one sample training set, wherein the at least one sample training set is from a reference BRAF mutation-containing cancer sample, or is from a reference BRAF mutation-free cancer sample; and classifying the sample as having a positive BRAF predictive response signature based on theAttorney Docket No. GNCN-024 / 03WO 320289-2158 results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 8 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 8 from the at least one training set; and classifying the sample as possessing a positive BRAF predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference BRAF mutation-containing cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample. In some cases, the at least one training set is from a reference BRAF mutation-containing cancer sample and from a reference BRAF mutation-free cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the specific therapy shows inhibitory activity toward BRAF. In some cases, the BRAF inhibitor is selected from the group consisting of vemurafenib, dabrafenib, sorafenib, encorafenib, PLX4032 and any combination thereof.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0019] In some cases, the plurality of classifier biomarkers are selected from Table 10, the target gene is Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene and the target gene specific activation signature is a Erb-B2 receptor tyrosine kinase 2 (ERBB2) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 10 to an expression level of the plurality of classifier biomarkers selected from Table 10 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample, or is from a reference ERBB2 mutation-free cancer sample; and classifying the sample as having a positive ERBB2 predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 10 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 10 from the at least one training set; and classifying the sample as possessing a positive ERBB2 predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification-containing cancer sample. In some cases, the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and from a reference ERBB2 mutation-free cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation- containing or amplification-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at leastAttorney Docket No. GNCN-024 / 03WO 320289-2158 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 10. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the first type of therapy shows inhibitory activity toward ERBB2. In some cases, the ERBB2 inhibitor is selected from the group consisting of lapatinib, afatinib, AST1306, AEE788, CP724, CP714,CP-724714, HER@-Inhibitor-1, BDTX-189, varlitinib, sapitinib (AZD8931), allitinib, poziotinib, pyrotinib, neratinib, tyrophostin AG-258, (-)-epigallocatechin gallate, mobocertinib, trastuzumab, zenocutuzumab and any combination thereof.

[0020] In some cases, the plurality of classifier biomarkers are selected from Table 12, the target gene is Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene and the target gene specific activation signature is a Erb-B3 receptor tyrosine kinase 3 (ERBB3) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 12 to an expression level of the plurality of classifier biomarkers selected from Table 12 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample, or is from a reference ERBB3 mutation-free cancer sample; and classifying the sample as having a positive ERBB3 predictive response signature based on the results of the comparing step. In some cases, the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 12 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 12 from the at least one training set; and classifying the sample as possessing a positive ERBB3 predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levelsAttorney Docket No. GNCN-024 / 03WO 320289-2158 of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification-containing cancer sample. In some cases, the at least one training set is from a reference ERBB3 mutation-containing cancer or amplification-containing sample and from a reference ERBB3 mutation-free cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation- containing or amplification-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses. In some cases, the specific therapy shows inhibitory activity toward ERBB3. In some cases, the ERBB3 inhibitor is selected from the group consisting of sapitinib (AZD8931), elgemtumab (LJM716), lumretuzumab (RG7116), KTN3379, patritumab, MM-121, MM-111, MM-141, single domain antibodies BCD090-P1, BCD090-M2, and BCD090-M456, and any combination thereof.

[0021] In some cases, the plurality of classifier biomarkers are selected from Table 14, the target gene is Phosphoinositide 3-Kinase CA (PIK3CA) gene and the target gene specific activation signature is a Phosphoinositide 3-Kinase CA (PIK3CA) predictive response signature. In some cases, the method further comprises comparing the expression levels of the plurality of classifier biomarkers selected from Table 14 to an expression level of the plurality of classifier biomarkers selected from Table 14 in at least one sample training set, wherein the at least one sample training set is from a reference PIK3CA mutation-containing cancer sample, or is from a reference PIK3CA mutation-free cancer sample; and classifying the sample as having a positive PIK3CA predictive response signature based on the results of the comparing step. In some cases, the comparingAttorney Docket No. GNCN-024 / 03WO 320289-2158 comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 14 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 14 from the at least one training set; and classifying the sample as possessing a positive PIK3CA predictive response signature based on the results of the statistical algorithm. In some cases, the at least one training set is from a reference PIK3CA mutation-containing cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample. In some cases, the at least one training set is from a reference PIK3CA mutation-containing cancer sample and from a reference PIK3CA mutation-free cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14. In some cases, the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization- based analyses. In some cases, the specific therapy shows inhibitory activity toward PIK3CA. In some cases, the PIK3CA inhibitor is selected from the group consisting of alpelisib, idelalisib,Attorney Docket No. GNCN-024 / 03WO 320289-2158combination thereof.

[0022] In some cases, the cancer the patient is suffering from is selected from the group consisting of adrenocortical carcinoma (ACC), bladder urothelial carcinoma (BLCA), breast cancer (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), diffuse large B-cell lymphoma (DLBC), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), low grade glioma (LGG), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), mesothelioma (MESO), pancreatic adenocarcinoma (PAAD), pheochromocytoma and paraganglioma (PCPG), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), sarcoma (SARC), skin cutaneous melanoma (SKCM), testicular germ cell tumors (TGCT), thyroid cancer (THCA), thymus cancer (THYM), uterine corpus endometrial carcinoma (UCEC), uterine carcinosarcoma (UCS) and uveal melanoma (UVM). In some cases, the cancer is LUAD. In some cases, the sample is a formalin- fixed, paraffin-embedded (FFPE) tissue sample, fresh or a frozen tissue sample, an exosome, or a bodily fluid obtained from the patient. In some cases, the bodily fluid is blood or fractions thereof, urine, saliva, or sputum.

[0023] In still another aspect, provided herein is method of detecting a biomarker in a sample obtained from a patient suffering from cancer, the method comprising, consisting essentially of or consisting of measuring the expression level of a plurality of biomarker nucleic acids selected from Table 2, Table 4, Table 6, Table 8, Table 10, Table 12 or Table 14 using an amplification, hybridization and / or sequencing assay. In some cases, the sample was previously diagnosed as being a cancer selected from ACC, BLCA, BRCA, CESC, CHOL, COAD, DLBC, GBM, HNSC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, PAAD, PCPG, PRAD, READ, SARC, SKCM, TGCT, THCA, THYM, UCEC, UCS and UVM. In some cases, the sample was previously diagnosed as being LUAD. In some cases, the amplification, hybridization and / or sequencing assay comprises performing quantitative real time reverse transcriptase polymerase chain reaction (qRT-PCR), RNAseq, microarrays, gene chips, nCounter Gene Expression Assay, Serial Analysis of Gene Expression (SAGE), Rapid Analysis of Gene Expression (RAGE), nuclease protectionAttorney Docket No. GNCN-024 / 03WO 320289-2158 assays, Northern blotting, or any other equivalent gene expression detection techniques. In some cases, the expression level is detected by performing qRT-PCR. In some cases, the detection of the expression level comprises using at least one pair of oligonucleotide primers per each biomarker nucleic acid from the plurality of biomarker nucleic acids selected from Table 2, Table 4, Table 6, Table 8, Table 10, Table 12 or Table 14. In some cases, the sample is a formalin- fixed, paraffin-embedded (FFPE) lung tissue sample, fresh or a frozen tissue sample, an exosome, wash fluids, cell pellets, or a bodily fluid obtained from the patient. In some cases, the bodily fluid is blood or fractions thereof, urine, saliva, or sputum. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consistsAttorney Docket No. GNCN-024 / 03WO 320289-2158 of all the classifier biomarkers from Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14. In some cases, the pluralityAttorney Docket No. GNCN-024 / 03WO 320289-2158 of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG. 1 illustrates three-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per EGFR activation signature status (i.e., positive or negative) to include in the signature of Table 2 for ascertaining EGFR alteration / activation status.

[0025] FIG. 2 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the EGFR classifier of Table 2.

[0026] FIG. 3 illustrates agreement between the calls and EGFR alteration status and the proportion of EGFR-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG.3) as well as an evaluation of the EGFR-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.3).

[0027] FIG. 4 illustrates the association (i.e., Pearson correlation) between the IC50 for specific EGFR inhibitors and EGFR activation signature scores from lung cancer cell lines with expression data and drug sensitivity (IC50) data.

[0028] FIG. 5 illustrates boxplots of EGFR-PRS scores for tumors across multiple cancer types either possessing putative EGFR mutation drivers (M-groups) or not (wildtype or WT group).

[0029] FIG.6 illustrates the overall survival (OS) of LUAD patients as a function of EGFR mutational status (left side graph) or EGFR PRS score (right side graph).

[0030] FIG. 7 illustrates two-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per MET activation signature status (i.e., positive or negative) to include in the signature of Table 4 for ascertaining MET alteration / activation status.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0031] FIG. 8 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the MET classifier of Table 4.

[0032] FIG. 9 illustrates agreement between the calls and MET activation status and the proportion of MET-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG. 9) as well as an evaluation of the MET-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.9).

[0033] FIG. 10 illustrates the association (i.e., Pearson correlation) between the IC50 for specific MET inhibitors and MET-PRS scores from lung cancer cell lines with expression data and drug sensitivity (IC50) data.

[0034] FIG. 11 illustrates boxplots of MET-PRS scores for tumors across multiple cancer types either possessing putative MET mutation drivers (M-groups) or not (wildtype or WT group).

[0035] FIG.12 illustrates the overall survival (OS) of LUAD patients as a function of MET alteration status (left side graph) or MET-PRS score (right side graph).

[0036] FIG.13 illustrates the overall survival (OS) of LUAD patients as a function of MET- PRS score (left side graph) or MET-PRS score and EGFR mutational status (right side graph).

[0037] FIG. 14 illustrates three-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per KRAS activation signature status (i.e., positive or negative) to include in the signature of Table 6 for ascertaining KRAS alteration / activation status.

[0038] FIG.15 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the KRAS classifier of Table 6.

[0039] FIG. 16 illustrates agreement between the calls and KRAS alteration status and the proportion of KRAS-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG.16) as well as an evaluation of the KRAS-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.16).

[0040] FIG. 17 illustrates the association (i.e., Pearson correlation) between the IC50 for specific KRAS inhibitors and KRAS-PRS scores from lung cancer cell lines with expression data and drug sensitivity (IC50) data.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0041] FIG. 18 illustrates boxplots of KRAS-PRS scores for tumors across multiple cancer types either possessing putative KRAS mutation drivers (M-groups) or not (wildtype or WT group).

[0042] FIG.19 illustrates the overall survival (OS) of LUAD patients as a function of KRAS alteration status (left side graph) or KRAS-PRS score (right side graph).

[0043] FIG. 20 illustrates two-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per BRAF activation signature status (i.e., positive or negative) to include in the signature of Table 8 for ascertaining BRAF alteration / activation status.

[0044] FIG.21 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the BRAF classifier of Table 8.

[0045] FIG. 22 illustrates agreement between the calls and BRAF alteration status and the proportion of BRAF-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG. 22) as well as an evaluation of the BRAF-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.22).

[0046] FIG. 23 illustrates two-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per ERBB2 activation signature status (i.e., positive or negative) to include in the signature of Table 10 for ascertaining ERBB2 alteration / activation status.

[0047] FIG.24 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the ERBB2 classifier of Table 10.

[0048] FIG. 25 illustrates agreement between the calls and ERBB2 alteration status and the proportion of ERBB2-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG.25) as well as an evaluation of the ERBB2-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.25).

[0049] FIG. 26 illustrates two-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) lung adenocarcinoma (LUAD) cancer dataset (n=510) to guide the selection of the number of genes per ERBB3 activation signature status (i.e., positive or negative) to include in the signature of Table 12 for ascertaining ERBB2 alteration / activation status.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0050] FIG.27 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the ERBB3 classifier of Table 12.

[0051] FIG. 28 illustrates agreement between the calls and ERBB3 alteration status and the proportion of ERBB3-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG.28) as well as an evaluation of the ERBB3-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.28).

[0052] FIG.29 illustrates samples ranked by their EGFR and MET activation scores (left side) as well as the overall survival (OS) of EGFR-PRS (+) patients as a function of MET-PRS score (right side graph).

[0053] FIG. 30 illustrates two-fold cross validation curves using ClaNC software on The Cancer Genome Atlas (TCGA) prostate adenocarcinoma (PRAD) cancer dataset (n=493) to guide the selection of the number of genes per PIK3CA_PRAD activation signature status (i.e., positive or negative) to include in the signature of Table 14 for ascertaining PIK3CA_PRAD alteration / activation status.

[0054] FIG.31 illustrates ClaNC mean and variance data from the entire training set used to determine the set of genes for the PIK3CA_PRAD classifier of Table 14.

[0055] FIG.32 illustrates agreement between the calls and PIK3CA_PRAD alteration status and the proportion of PIK3CA_PRAD-PRS-activate (yes) calls in the non-altered tumors for the training set (left half of plot in FIG. 32) as well as an evaluation of the PIK3CA_PRAD-PRS model external performance (i.e., outside of the training set data) done using the test set data and comparing the predictive performance to the training set (right half of FIG.32). DETAILED DESCRIPTION Definitions

[0056] While the following terms are believed to be well understood by one of ordinary skill in the art, the following definitions are set forth to facilitate explanation of the presently disclosed subject matter.

[0057] As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Additionally, the use of “or” is intended to include “and / or” unless the context clearly indicates otherwise. Furthermore, to theAttorney Docket No. GNCN-024 / 03WO 320289-2158 extent that the terms "including", "includes", "having", "has", "with", or variants thereof are used in either the detailed description and / or the claims, such terms are intended to be inclusive in a manner similar to the term "comprising". The term "about" as used herein can refer to a range that is 15%, 10%, 8%, 6%, 4%, or 2% plus or minus from a stated numerical value.

[0058] Unless the context requires otherwise, throughout the present specification and claims, the word “comprise” and variations thereof, such as, “comprises” and “comprising” are to be construed in an open, inclusive sense that is as “including, but not limited to”. The use of the alternative (e.g., "or") should be understood to mean either one, both, or any combination thereof of the alternatives. As used herein, the terms "about" and "consisting essentially of" mean + / - 20% of the indicated range, value, or structure, unless otherwise indicated.

[0059] Reference throughout this specification to “one embodiment” or “an embodiment” means that a particular feature, structure or characteristic described in connection with the embodiment may be included in at least one embodiment of the present disclosure. Thus, the appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification may not necessarily all be referring to the same embodiment. It is appreciated that certain features of the disclosure, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the disclosure, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination.

[0060] Throughout this disclosure, various aspects of the methods and compositions provided herein can be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[0061] Unless otherwise indicated, the methods and compositions provided herein can utilize conventional techniques and descriptions of organic chemistry, polymer technology, molecular biology (including recombinant techniques), cell biology, biochemistry, and immunology, whichAttorney Docket No. GNCN-024 / 03WO 320289-2158 are within the skill of the art. Such conventional techniques include polymer array synthesis, hybridization, ligation, and detection of hybridization using a label. Specific illustrations of suitable techniques can be had by reference to the example herein below. However, other equivalent conventional procedures can, of course, also be used. Such conventional techniques and descriptions can be found in standard laboratory manuals such as Genome Analysis: A Laboratory Manual Series (Vols. I-IV), Using Antibodies: A Laboratory Manual, Cells: A Laboratory Manual, PCR Primer: A Laboratory Manual, and Molecular Cloning: A Laboratory Manual (all from Cold Spring Harbor Laboratory Press), Gait, "Oligonucleotide Synthesis: A Practical Approach" 1984, IRL Press, London, Nelson and Cox (2000), Lehninger et al., (2008) Principles of Biochemistry 5th Ed., W.H. Freeman Pub., New York, N.Y. and Berg et al. (2006) Biochemistry, 6.sup.th Ed., W.H. Freeman Pub., New York, N.Y., all of which are herein incorporated in their entirety by reference for all purposes.

[0062] Conventional software and systems may also be used in the methods and compositions provided herein. Computer software products of the invention typically include computer readable medium having computer-executable instructions for performing the logic steps of the method of the invention. Suitable computer readable medium include floppy disk, CD-ROM / DVD / DVD- ROM, hard-disk drive, flash memory, ROM / RAM, magnetic tapes, etc. The computer-executable instructions may be written in a suitable computer language or combination of several languages. Basic computational biology methods are described in, for example, Setubal and Meidanis et al., Introduction to Computational Biology Methods (PWS Publishing Company, Boston, 1997); Salzberg, Searles, Kasif, (Ed.), Computational Methods in Molecular Biology, (Elsevier, Amsterdam, 1998); Rashidi and Buehler, Bioinformatics Basics: Application in Biological Science and Medicine (CRC Press, London, 2000) and Ouelette and Bzevanis Bioinformatics: A Practical Guide for Analysis of Gene and Proteins (Wiley & Sons, Inc., 2.sup.nd ed., 2001). See U.S. Pat. No.6,420,108.

[0063] The methods and compositions provided herein may also make use of various computer program products and software for a variety of purposes, such as probe design, management of data, analysis, and instrument operation. See, U.S. Pat. Nos. 5,593,839, 5,795,716, 5,733,729, 5,974,164, 6,066,454, 6,090,555, 6,185,561, 6,188,783, 6,223,127, 6,229,911 and 6,308,170. Computer methods related to genotyping using high-density microarray analysis may also be usedAttorney Docket No. GNCN-024 / 03WO 320289-2158 in the present methods, see, for example, US Patent Pub. Nos. 20050250151, 20050244883, 20050108197, 20050079536 and 20050042654.

[0064] Additionally, the present disclosure may have preferred embodiments that include methods for providing genetic information over networks such as the Internet as shown in U.S. Patent Pub. Nos. 20030097222, 20020183936, 20030100995, 20030120432, 20040002818, 20040126840, and 20040049354.

[0065] As used herein, the term “individual”, “patient”, or “subject”, can be used interchangeably and can refer to an individual regardless of health and / or disease status. A subject can be a subject, a study participant, a control subject, a screening subject, or any other class of individual from whom a sample can be obtained and assessed in the context of the invention. Accordingly, a subject can be diagnosed with a cancer (including subtypes, or grades thereof), can present with one or more symptoms of a cancer or a predisposing factor, such as a family (genetic) or medical history (medical) factor, for a cancer, can be undergoing treatment or therapy for a cancer, or the like. Alternatively, a subject can be healthy with respect to any of the aforementioned factors or criteria.

[0066] It will be appreciated that the term "healthy" as used herein, can be relative to a cancer status, as the term "healthy" cannot be defined to correspond to any absolute evaluation or status. Thus, an individual defined as healthy with reference to any specified disease or disease criterion can in fact be diagnosed with any other one or more diseases or exhibit any other one or more disease criterion including one or more other cancer types.

[0067] As used herein, the terms “individual,” “patient,” and “subject” can refer to any single animal, more preferably a mammal (including such non-human animals as, for example, dogs, cats, horses, rabbits, zoo animals, cows, pigs, sheep, and non-human primates) for which treatment is desired. In particular, embodiments, the individual or patient herein is a human.

[0068] Further to any of the embodiments provided herein, the cancer can include, but are not limited to, carcinoma, lymphoma, blastoma (including medulloblastoma and retinoblastoma), sarcoma (including liposarcoma and synovial cell sarcoma), neuroendocrine tumors (including carcinoid tumors, gastrinoma, and islet cell cancer), mesothelioma, schwannoma (including acoustic neuroma), meningioma, adenocarcinoma, melanoma, and leukemia or lymphoid malignancies. Examples of a cancer also include, but are not limited to, a lung cancer (e.g., a non- small cell lung cancer (NSCLC)), a kidney cancer (e.g., a kidney urothelial carcinoma or RCC), aAttorney Docket No. GNCN-024 / 03WO 320289-2158 bladder cancer (e.g., a bladder urothelial (transitional cell) carcinoma (e.g., locally advanced or metastatic urothelial cancer, including 1L or 2L+ locally advanced or metastatic urothelial carcinoma), a breast cancer, a colorectal cancer (e.g., a colon adenocarcinoma), an ovarian cancer, a pancreatic cancer (e.g., pancreatic adenocarcinoma or PAAD), a gastric carcinoma, an esophageal cancer, a mesothelioma, a melanoma (e.g., a skin melanoma), a head and neck cancer (e.g., a head and neck squamous cell carcinoma (HNSCC)), a thyroid cancer, a sarcoma (e.g., a soft-tissue sarcoma, a fibrosarcoma, a myxosarcoma, a liposarcoma, an osteogenic sarcoma, an osteosarcoma, a chondrosarcoma, an angiosarcoma, an endotheliosarcoma, a lymphangiosarcoma, a lymphangioendotheliosarcoma, a leiomyosarcoma, or a rhabdomyosarcoma), a prostate cancer, a glioblastoma, a cervical cancer, a thymic carcinoma, a leukemia (e.g., an acute lymphocytic leukemia (ALL), an acute myelocytic leukemia (AML), a chronic myelocytic leukemia (CML), a chronic eosinophilic leukemia, or a chronic lymphocytic leukemia (CLL)), a lymphoma (e.g., a Hodgkin lymphoma or a non-Hodgkin lymphoma (NHL)), a myeloma (e.g., a multiple myeloma (MM)), a mycosis fungoides, a Merkel cell cancer, a hematologic malignancy, a cancer of hematological tissues, a B cell cancer, a bronchus cancer, a stomach cancer, a brain or central nervous system cancer, a peripheral nervous system cancer, a uterine or endometrial cancer, a cancer of the oral cavity or pharynx, a liver cancer, a testicular cancer, a biliary tract cancer, a small bowel or appendix cancer, a salivary gland cancer, an adrenal gland cancer, an adenocarcinoma, an inflammatory myofibroblastic tumor, a gastrointestinal stromal tumor (GIST), a colon cancer, a myelodysplastic syndrome (MDS), a myeloproliferative disorder (MPD), a polycythemia Vera, a chordoma, a synovioma, an Ewing’s tumor, a squamous cell carcinoma, a basal cell carcinoma, an adenocarcinoma, a sweat gland carcinoma, a sebaceous gland carcinoma, a papillary carcinoma, a papillary adenocarcinoma, a medullary carcinoma, a bronchogenic carcinoma, a renal cell carcinoma, a hepatoma, a bile duct carcinoma, a choriocarcinoma, a seminoma, an embryonal carcinoma, a Wilms' tumor, a bladder carcinoma, an epithelial carcinoma, a glioma, an astrocytoma, a medulloblastoma, a craniopharyngioma, an ependymoma, a pinealoma, a hemangioblastoma, an acoustic neuroma, an oligodendroglioma, a meningioma, a neuroblastoma, a retinoblastoma, a follicular lymphoma, a diffuse large B-cell lymphoma, a mantle cell lymphoma, a hepatocellular carcinoma, a thyroid cancer, a small cell cancer, an essential thrombocythemia, an agnogenic myeloid metaplasia, a hypereosinophilic syndrome, aAttorney Docket No. GNCN-024 / 03WO 320289-2158 systemic mastocytosis, a familiar hypereosinophilia, a neuroendocrine cancer, or a carcinoid tumor.

[0069] In some cases, the cancer is selected from an adrenocortical carcinoma (ACC), a cervical kidney renal papillary cell carcinoma (KIRP); breast invasive carcinoma (BRCA); thyroid cancer (THCA); bladder carcinoma (BLCA); a muscle invasive bladder cancer (MIBC); prostate adenocarcinoma (PRAD); kidney chromophobe (KICH); cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC); kidney renal clear cell carcinoma (KIRC); liver hepatocellular carcinoma (LIHC); low grade glioma (LGG); sarcoma (SARC); lung adenocarcinoma (LUAD); colon adenocarcinoma (COAD); head-neck squamous cell carcinomaparaganglioma (PCPG), an esophageal cancer, a mesothelioma, a melanoma, a head and neck cancer, a thyroid cancer, a sarcoma, a prostate cancer, a glioblastoma, a cervical cancer, a thymic carcinoma, a leukemia, a lymphoma, a myeloma, a mycosis fungoides, a merkel cell cancer, an endometrial cancer. In some cases, the cancer is adrenocortical carcinoma (ACC), lung adenocarcinoma (LUAD), colon adenocarcinoma (COAD), breast invasive carcinoma (BRCA),

[0070] The term “nucleic acid” as used herein can refer to a polymeric form of nucleotides of any length, either ribonucleotides, deoxyribonucleotides or peptide nucleic acids (PNAs), that comprise purine and pyrimidine bases, or other natural, chemically or biochemically modified, non-natural, or derivatized nucleotide bases. The backbone of the polynucleotide can comprise sugars and phosphate groups, as may typically be found in RNA or DNA, or modified or substituted sugar or phosphate groups. A polynucleotide may comprise modified nucleotides, suchAttorney Docket No. GNCN-024 / 03WO 320289-2158 as methylated nucleotides and nucleotide analogs. The sequence of nucleotides may be interrupted by non-nucleotide components. Thus, the terms nucleoside, nucleotide, deoxynucleoside and deoxynucleotide generally include analogs such as those described herein. These analogs can be those molecules having some structural features in common with a naturally occurring nucleoside or nucleotide such that when incorporated into a nucleic acid or oligonucleotide sequence, they allow hybridization with a naturally occurring nucleic acid sequence in solution. Typically, these analogs can be derived from naturally occurring nucleosides and nucleotides by replacing and / or modifying the base, the ribose or the phosphodiester moiety. The changes can be tailor made to stabilize or destabilize hybrid formation or enhance the specificity of hybridization with a complementary nucleic acid sequence as desired.

[0071] The term "complementary" as used herein can refer to the hybridization or base pairing between nucleotides or nucleic acids, such as, for instance, between the two strands of a double stranded DNA molecule or between an oligonucleotide primer and a primer binding site on a single stranded nucleic acid to be sequenced or amplified. See, M. Kanehisa Nucleic Acids Res.12:203 (1984), incorporated herein by reference.

[0072] An analyte assay can be a detection or diagnostic method as provided herein. In some cases, the sample can comprise or contain the analyte. The analyte can be derived, removed or extracted from a cell or cells within the sample. The analyte can be a protein or a nucleic acid. The analyte can be a cell-free or extracellular nucleic acid. In some cases, the analyte is a circulating tumor nucleic acid. The nucleic acid can be such DNA or RNA. In some cases, the nucleic acid is cell- free DNA (cfDNA). The cfDNA can be circulating tumor DNA (ctDNA).

[0073] The term “sample” as used herein can refer to a biological sample, such as a liquid biological sample or bodily fluid or a biological tissue. Examples of liquid biological samples or bodily fluids for use in the methods provided herein can include urine, blood, plasma, serum, saliva, ejaculate, stool, sputum, cerebrospinal fluid (CSF), tears, mucus, amniotic fluid or the like. Biological tissues as used herein can be aggregates of cells, usually of a particular kind together with their intercellular substance that form one of the structural materials of a human, animal, plant, bacterial, fungal or viral structure, including connective, epithelium, muscle and nerve tissues. Examples of biological tissues also include organs, tumors, lymph nodes, arteries and individual cell(s). A biological tissue sample can be a biopsy. In one embodiment, the sample is a biopsy of a tumor, which can be referred to as a tumor sample. In one embodiment, the analysesAttorney Docket No. GNCN-024 / 03WO 320289-2158 described herein are performed on biopsies that are freshly obtained or derived. In one embodiment, the analyses described herein are performed on biopsies that are frozen. In one embodiment, the analyses described herein are performed on biopsies that are embedded in paraffin wax. Accordingly, the methods provided herein, including the RT-PCR methods, are sensitive, precise and have multianalyte capability for use with paraffin embedded samples. See, for example, Cronin et al. (2004) Am. J Pathol.164(1):35-42, herein incorporated by reference.

[0074] Formalin fixation and tissue embedding in paraffin wax is a universal approach for tissue processing prior to light microscopic evaluation. A major advantage afforded by formalin- fixed paraffin-embedded (FFPE) specimens is the preservation of cellular and architectural morphologic detail in tissue sections. (Fox et al. (1985) J Histochem Cytochem 33:845-853). The standard buffered formalin fixative in which biopsy specimens are processed is typically an aqueous solution containing 37% formaldehyde and 10-15% methyl alcohol. Formaldehyde is a highly reactive dipolar compound that results in the formation of protein-nucleic acid and protein- protein crosslinks in vitro (Clark et al. (1986) J Histochem Cytochem 34:1509-1512; McGhee and von Hippel (1975) Biochemistry 14:1281-1296, each incorporated by reference herein).

[0075] In one embodiment, the sample used herein is obtained from an individual, and comprises fresh-frozen paraffin embedded (FFPE) tissue.

[0076] The term “tumor,” as used herein, can refer to all neoplastic cell growth and proliferation, whether malignant or benign, and all pre-cancerous and cancerous cells and tissues. The terms “cancer,” “cancerous,” and “tumor” are not mutually exclusive and can be used interchangeably.

[0077] The term “detection” can include any means of detecting, including direct and indirect detection.

[0078] A sample as provided herein can be processed to render it competent for fragmentation, ligation, denaturation, and / or amplification. Exemplary sample processing can include lysing cells of the sample to release nucleic acid, purifying the sample (e.g., to isolate nucleic acid from other sample components, which can inhibit enzymatic reactions), diluting / concentrating the sample, and / or combining the sample with reagents for further nucleic acid processing such as nucleic acid extension, amplification and / or sequencing. In some examples, the sample can be combined with a restriction enzyme, reverse transcriptase, or any other enzyme of nucleic acid processing.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0079] The term “biomarkers” or “classifier biomarkers” or “classifier” can include nucleic acids (e.g., genes) and proteins, and variants and fragments thereof. Such biomarkers can include RNA or DNA, including cDNA, comprising the entire or partial sequence of the nucleic acid sequence encoding the biomarker, or the complement of such a sequence. The biomarker nucleic acids can also include any expression product or portion thereof of the nucleic acid sequences of interest. A biomarker protein is a protein encoded by or corresponding to a DNA or RNA biomarker of the invention. A biomarker protein comprises the entire or partial amino acid sequence of any of the biomarker proteins or polypeptides. The biomarker nucleic acid can be extracted from a cell or can be cell free or extracted from an extracellular vesicular entity such as an exosome or microvesicle.

[0080] A "biomarker" or “classifier biomarker” or “classifier” can be any nucleic acid (e.g., gene) or protein whose level of expression in a tissue or cell is altered compared to that of a normal or healthy cell or tissue. The detection, and in some cases the level, of the biomarkers can permit the differentiation of samples. The “classifier biomarker” or “biomarker” or “classifier” may be one that is up-regulated (e.g., expression is increased) or down-regulated (e.g., expression is decreased) relative to a reference or control as provided herein. The overall expression levelo f e a c h g e n e t e s t e d f r o m a s a m p l e can be referred to herein as the "'expressionprofile" and can be used to classify a training set or a test sample as provided herein. However, it is understood that independent evaluation of expression for each of the genes disclosed herein can be used to classify a training set or a test sample (e.g., as being a therapeutic agent responsive group or not) without the need to group up-regulated and down-regulated genes into one or more gene cassettes.

[0081] As used herein, an “expression profile” or a “biomarker profile” or “gene signature” comprises one or more values corresponding to a measurement of the relative abundance, level, presence, or absence of expression of a discriminative or classifier gene or biomarker. An expression profile can be derived from a subject prior to or subsequent to a diagnosis of a cancer, can be derived from a biological sample collected from a subject at one or more time points prior to or following treatment or therapy, can be derived from a biological sample collected from a subject at one or more time points during which there is no treatment or therapy, or can be collected from a healthy subject. The subject can be a human patient. The one or more biomarkers of the biomarker profiles provided herein can be selected from one or more biomarkers of only Table 2, Table 4, Table 6, Table 8, Table 10, Table 12 or Table 14 or selected from combinations ofAttorney Docket No. GNCN-024 / 03WO 320289-2158 classifier biomarkers selected from Table 2, Table 4, Table 6, Table 8, Table 10, Table 12, or Table 14.

[0082] As used herein, the term “oncogene” can refer to a gene that is a mutated (changed or altered) form of a gene that causes the transformation of normal cells into cancerous tumor cells and / or a gene whose aberrant expression or activation at an abnormal point in development for expression or activation of said gene causes the transformation of normal cells into cancerous tumor cells. Oncogenes may cause the growth of cancer cells. Mutations in genes that become oncogenes can be inherited or caused by being exposed to substances in the environment that cause cancer. Oncogenes can also be viral genes that transform a host cell into a tumor cell. An “oncogenic mutation” can refer to a mutation in a gene that causes the transformation of a host cell into a cancerous tumor cell. A mutation as referred to herein should be construed broadly, and include single nucleotide polymorphisms (SNPs), sequence insertions, deletions, inversions, gene amplifications and other sequence replacements. As used herein, the term “non-synonymous” or non-synonymous SNPs” refers to mutations that lead to coding changes in host cell proteins.

[0083] As used herein, the term “EGFR mutation” or “EGFR mutations” can refer to any mutation known in the art in an egfr gene and / or the protein encoded thereby.

[0084] As used herein, the term “MET mutation” or “MET mutations” can refer to any mutation or amplification known in the art in an met gene and / or the protein encoded thereby.

[0085] As used herein, the term “KRAS mutation” or “KRAS mutations” can refer to any mutation known in the art in a Kirsten Rat Sarcoma viral oncogene homolog (kras) gene and / or the protein encoded thereby.

[0086] As used herein, the term “BRAF mutation” or “BRAF mutations” can refer to any mutation known in the art in a B-Raf protooncogene (BRAF) gene and / or the protein encoded thereby.

[0087] As used herein, the term “ERBB2 mutation” or “ERBB2 mutations” can refer to any mutation or amplification known in the art in an Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene and / or the protein encoded thereby.

[0088] As used herein, the term “ERBB3 mutation” or “ERBB3 mutations” can refer to any mutation or amplification known in the art in an Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene and / or the protein encoded thereby.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0089] As used herein, the term “PI3KCA mutation” or “PI3KCA mutations” can refer to any mutation known in the art in a Phosphoinositide 3-Kinase CA (PIK3CA) gene and / or the protein encoded thereby.

[0090] As used herein, the term "determining an expression level" or "determining an expression profile" or “detecting an expression level” or “detecting an expression profile” as used in reference to a biomarker or classifier means the application of a biomarker specific reagent such as a probe, primer or antibody and / or a method to a sample, for example a sample of the subject or patient and / or a control sample, for ascertaining or measuring quantitatively, semi-quantitatively or qualitatively the amount of a biomarker or biomarkers, for example the amount of biomarker polypeptide or mRNA (or cDNA derived therefrom). For example, a level of a biomarker can be determined by a number of methods including for example immunoassays including, for example, immunohistochemistry, ELISA, Western blot, immunoprecipitation and the like, where a biomarker detection agent such as an antibody for example, a labeled antibody, specifically binds the biomarker and permits for example relative or absolute ascertaining of the amount of polypeptide biomarker, hybridization and PCR protocols where a probe or primer or primer set are used to ascertain the amount of nucleic acid biomarker, including for example probe based and amplification based methods including for example microarray analysis, RT-PCR such as quantitative RT-PCR (qRT-PCR), serial analysis of gene expression (SAGE), Northern Blot, digital molecular barcoding technology, for example Nanostring Counter Analysis, and TaqMan quantitative PCR assays. Other methods of mRNA detection and quantification can be applied, such as mRNA in situ hybridization in formalin-fixed, paraffin-embedded (FFPE) tissue samples or cells. This technology is currently offered by the QuantiGene ViewRNA (Affymetrix), which uses probe sets for each mRNA that bind specifically to an amplification system to amplify the hybridization signals; these amplified signals can be visualized using a standard fluorescence microscope or imaging system. This system for example can detect and measure transcript levels in heterogeneous samples; for example, if a sample has normal and tumor cells present in the same tissue section. As mentioned, TaqMan probe-based gene expression analysis (PCR-based) can also be used for measuring gene expression levels in tissue samples, and this technology has been shown to be useful for measuring mRNA levels in FFPE samples. In brief, TaqMan probe-based assays utilize a probe that hybridizes specifically to the mRNA target. This probe contains a quencher dye and a reporter dye (fluorescent molecule) attached to each end, and fluorescence isAttorney Docket No. GNCN-024 / 03WO 320289-2158 emitted only when specific hybridization to the mRNA target occurs. During the amplification step, the exonuclease activity of the polymerase enzyme causes the quencher and the reporter dyes to be detached from the probe, and fluorescence emission can occur. This fluorescence emission is recorded and signals are measured by a detection system; these signal intensities are used to calculate the abundance of a given transcript (gene expression) in a sample.

[0091] The present invention also encompasses a system capable of distinguishing various subtypes of cancer that may or may not be amendable to treatment with an anti-FGFR agent or anti-FGFR3 agent in a sample obtained from a subject suspected of suffering from cancer. Thissystem c a n b e capable of processing a large number of subjects and subject variables suchas expression profiles and other diagnostic criteria. The methods a n d s y s t e m si n c o r p o r a t i n g s a i d m e t h o d s described herein can be used for"pharmacometabonomics," in analogy to pharmacogenomics, e.g., predictive of response to therapy. In this embodiment, subjects could be divided into "responders" and "nonresponders" using the expression profile as evidence of "response," and features of the expression profile could then be used to target future subjects who would likely respond to a particular therapeutic course.

[0092] The expression profile can be used in combination with other diagnostic methods including histochemical, immunohistochemical, cytologic, immunocytologic, and visual diagnostic methods including histologic or morphometric evaluation of samples (e.g., tissue samples).

[0093] In various embodiments of the present invention, the expression profile or signature derived from a subject is compared to a reference expression profile or signature. A “reference expression profile” can be a profile derived from the subject prior to treatment or therapy; can be a profile produced from the subject’s sample at a particular time point (usually prior to or following treatment or therapy but can also include a particular time point prior to or following diagnosis of a type of cancer); or can be derived from a healthy individual or a pooled reference from healthy individuals. A reference expression profile can be specific to cancer types or subtypes known to be responders to a specific inhibitor therapy (e.g., EGFR inhibitors, MET inhibitors, KRAS inhibitors, BRAF inhibitors, ERBB2 inhibitors, ERBB3 inhibitors, PI3KCA inhibitors, etc.) or non-responders to a specific inhibitor therapy (e.g., EGFR inhibitors, MET inhibitors, KRAS inhibitors, BRAF inhibitors, ERBB2 inhibitors, ERBB3 inhibitors, PI3KCAAttorney Docket No. GNCN-024 / 03WO 320289-2158 inhibitors, etc.). A reference expression profile can be specific to cancer types or subtypes known to be proliferative or non-proliferative.

[0094] The reference expression profile or signature can be compared to a test expression profile or signature. A "test expression profile" can be derived from the same subject as the reference expression profile except at a subsequent time point (e.g., one or more days, weeks or months following collection of the reference expression profile) or can be derived from a different subject. In summary, any test expression profile of a subject can be compared to a previously collected profile from a subject whose cancer type or subtype is known to be responsive to a specific inhibitor therapy (e.g., EGFR inhibitors, MET inhibitors, KRAS inhibitors, BRAF inhibitors, ERBB2 inhibitors, ERBB3 inhibitors, PI3KCA inhibitors, etc.) or non-responsive to a specific inhibitor therapy (e.g., EGFR inhibitors, MET inhibitors, KRAS inhibitors, BRAF inhibitors, ERBB2 inhibitors, ERBB3 inhibitors, PI3KCA inhibitors, etc.). Overview

[0095] The p r e s en t i n ve n t io n p r ov i d es methods, compositions or kits that can be usedto provide a n assessment or determination of a target gene mutational or al terationstatus (also referred to as an activation signature or AS) of a sample obtainedfrom a subject suffering from or suspected of suffering from a cancer. The target gene canbe selected from egfr, met, kras, braf, erbb2, erbb3 and pi3kca . In oneembodiment, the assessment or determination of the target gene mutational status comprises measuring an expression level of a defined set of biomarkers in the sample obtained from the subject. The measurement of the expression level can be at the nucleic acid or protein level or any combination thereof. The measurement of the expression level can be performed using of any of the methods provided herein for measuring expression levels atthe nucleic acid or protein level. In one embodiment, the target gene mutational statusis used to determine the likelihood of the subject suffering from or suspected of suffering from a cancer being responsive to treatment with a therapeutic agent or a defined set of therapeutic agents. In this way, the methods provided herein can provide a target gene predictive response signature (PRS). In other words, any activation signature determined for a specific target gene using the methods provided can be also be considered to be a predictive response signature for saidAttorney Docket No. GNCN-024 / 03WO 320289-2158target gene. In another embodiment, the target gene mutational status of the sampleobtained from the subject is predictive of said subject being responsive or non-responsive toa defined set of therapeutic agents. In yet another embodiment, the target genemutational status of the sample obtained from the subject is used in a method to treat the cancer that the subject is suffering from or suspected of suffering from such that a defined set of therapeutic agents is administered to the subject based on the target gene mutational status determined for the sample. The sample can be any type of sample provided herein such as, for example, a tumor sample or biopsy. The cancer can be any cancer known in the art and / or provided herein. The defined set of therapeutic agents can be any agent known in the art and / or provided herein that exhibits inhibitory activity toward the target gene generally or broadly or the target gene specifically.

[0096] In one embodiment, the target gene is the epidermal growth factor receptor (egfr) gene. Further to this embodiment, the methods, compositions or kits provided herein can beused to provide a n assessment or determination of the EGFR mutational or alterationstatus (also referred to as an EGFR activation signature or EGFR-AS) of a sampleobtained from a subject suffering from or suspected of suffering from a cancer. Themeasuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the EGFR activation signature (EGFR-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as an EGFR activation signature (EGFR-AS) or EGFR activation classifier. As alluded to herein, the EGFR-AS can reflect or represent a presence or absence of one or more EGFR mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose EGFR-AS indicates that the subject possesses an EGFR alteration or mutation is said to have a positive EGFR-AS or be EGFR-AS (+). Conversely, samples whose EGFR-AS indicates that the subject does not possess an EGFR alteration or mutation is said to have a negative EGFR-AS or be EGFR-AS (-).

[0097] Whether or not an EGFR-AS of a sample is positive or negative can be determined by comparing the EGFR-AS determined for the sample to the EGFR-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the egfr gene. In one embodiment, the reference or controlAttorney Docket No. GNCN-024 / 03WO 320289-2158 sample is a sample known to not possess or harbor one or more mutations and / or fusions in the egfr gene. In one embodiment, the EGFR-AS of the sample obtained from the subject is compared to the EGFR-AS of a sample known to possess one or more mutations and / or fusions in the egfr gene. In another embodiment, the EGFR-AS of the sample obtained from the subject is compared to the EGFR-AS of a sample known to not possess one or more mutations and / or fusions in the egfr gene. In yet another embodiment, the EGFR-AS of the sample obtained from the subject is compared to the EGFR-AS of a sample known to possess one or more mutations and / or fusions in the egfr gene and the EGFR-AS of a sample known to not possess one or more mutations and / or fusions in the egfr gene. The one or more alterations or mutations in the egfr gene can be any mutation and / or fusion in the egfr gene known in the art.

[0098] In one embodiment, a positive EGFR activation signature (EGFR-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward an epithelial growth factor receptor (EGFR) broadly and / or an EGFR, specifically. In this way, the EGFR-AS is acting as or can be referred to as an EGFR predictive response signature (EGFR-PRS). A positive EGFR-PRS (i.e., EGFR-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., an EGFR inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward an EGFR broadly and / or specifically. The first type of therapy (e.g., an EGFR inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward an EGFR broadly and / or specifically can be a tyrosine kinase inhibitor, an antibody, an antibody-conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward an EGFR can any therapeutic agent known in the art and / or provided herein.

[0099] In some cases, the subject with the positive EGFR-PRS can be non-responsive or even resistant to the second type of therapy. In some cases, the second type of therapy can be a therapeutic agent known in the art and / or provided herein that has inhibitory activity toward MEK (e.g., an MEK inhibitor) or ERK (e.g., an ERK inhibitor).Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0100] In one embodiment, a negative EAS or EGFR-PRS (i.e., EGFR-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward an EGFR generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0101] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine an EGFR activation signature (EAS) or EGFR-PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 2. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers for use in the methods for determining an EGFR- AS or EGFR-PRS consists of only the classifier biomarkers found in Table 2. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0102] In one embodiment, the target gene is the MET proto-oncogene, receptor tyrosine kinase (met) gene. Further to this embodiment, the methods, compositions or kits providedherein can be used to provide a n assessment or determination of the MET mutationalor al teration status (also referred to as a MET activation signature or MET-AS)of a sample obtained from a subject suffering from or suspected of suffering from a cancer.The measuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the MET activation signature (MET- AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as a MET activation signature (MET-AS) or MET activation classifier. As alluded to herein, the MET-AS can reflect or represent a presence or absence of one or more MET mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose MET-AS indicates that the subject possesses a MET alteration orAttorney Docket No. GNCN-024 / 03WO 320289-2158 mutation is said to have a positive MET-AS or be MET-AS (+). Conversely, samples whose MET- AS indicates that the subject does not possess a MET alteration or mutation is said to have a negative MET-AS or be MET-AS (-).

[0103] Whether or not a MET-AS of a sample is positive or negative can be determined by comparing the MET determined for the sample to the MET-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the met gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the met gene. In one embodiment, the MET-AS of the sample obtained from the subject is compared to the MET- AS of a sample known to possess one or more mutations and / or fusions in the met gene. In another embodiment, the MET-AS of the sample obtained from the subject is compared to the MET-AS of a sample known to not possess one or more mutations and / or fusions in the met gene. In yet another embodiment, the MET-AS of the sample obtained from the subject is compared to the MET-AS of a sample known to possess one or more mutations and / or fusions in the met gene and the MET-AS of a sample known to not possess one or more mutations and / or fusions in the met gene. The one or more alterations or mutations in the met gene can be any mutation and / or fusion in the met gene known in the art.

[0104] In one embodiment, a positive MET activation signature (MET-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward a MET broadly and / or a MET, specifically. In this way, the MET-AS is acting as or can be referred to as a MET predictive response signature (MET-PRS). A positive MET-PRS (i.e., MET-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., a MET inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward MET broadly and / or specifically. The first type of therapy (e.g., MET inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward an MET broadly and / or specifically can be a tyrosine kinase inhibitor, an antibody, anAttorney Docket No. GNCN-024 / 03WO 320289-2158 antibody-conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward a MET can any therapeutic agent known in the art and / or provided herein.

[0105] In some cases, the subject with the positive MET-PRS can be non-responsive or even resistant to the second type of therapy. In some cases, the second type of therapy can be a therapeutic agent known in the art and / or provided herein that has inhibitory activity toward mTOR (e.g., an mTOR inhibitor), PI3K (e.g., PI3K inhibitor), an EGFR (e.g., an EGFR inhibitor) or AKT (e.g., an AKT inhibitor).

[0106] In one embodiment, a negative MET-AS or MET-PRS (i.e., MET-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward MET generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0107] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine a MET activation signature (MET-AS) or MET-PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 4. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 or 28 classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers for use in the methods for determining a MET-AS or MET-PRS consists of only the classifier biomarkers found in Table 4. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0108] In one embodiment, the target gene is the Kirsten Rat Sarcoma viral oncogene homolog (kras) gene. Further to this embodiment, the methods, compositions or kits providedherein can be used to provide a n assessment or determination of the KRAS mutationalAttorney Docket No. GNCN-024 / 03WO 320289-2158 or alteration status (also referred to as a KRAS activation signature or KRAS-AS)of a sample obtained from a subject suffering from or suspected of suffering from a cancer.The measuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the KRAS activation signature (KRAS-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as a KRAS activation signature (KRAS-AS) or KRAS activation classifier. As alluded to herein, the KRAS-AS can reflect or represent a presence or absence of one or more KRAS mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose KRAS-AS indicates that the subject possesses a KRAS alteration or mutation is said to have a positive KRAS-AS or be KRAS-AS (+). Conversely, samples whose KRAS-AS indicates that the subject does not possess a KRAS alteration or mutation is said to have a negative KRAS-AS or be KRAS-AS (-).

[0109] Whether or not a KRAS-AS of a sample is positive or negative can be determined by comparing the KRAS determined for the sample to the KRAS-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the kras gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the kras gene. In one embodiment, the KRAS-AS of the sample obtained from the subject is compared to the KRAS-AS of a sample known to possess one or more mutations and / or fusions in the kras gene. In another embodiment, the KRAS-AS of the sample obtained from the subject is compared to the KRAS-AS of a sample known to not possess one or more mutations and / or fusions in the kras gene. In yet another embodiment, the KRAS-AS of the sample obtained from the subject is compared to the KRAS-AS of a sample known to possess one or more mutations and / or fusions in the kras gene and the KRAS-AS of a sample known to not possess one or more mutations and / or fusions in the kras gene. The one or more alterations or mutations in the kras gene can be any mutation and / or fusion in the kras gene known in the art such as, for example, the KRAS non G12C mutations.

[0110] In one embodiment, a positive KRAS activation signature (KRAS-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward a KRAS, MEK1, MEK2, or ERK broadlyAttorney Docket No. GNCN-024 / 03WO 320289-2158 and / or specifically. In this way, the KRAS-AS is acting as or can be referred to as a KRAS predictive response signature (KRAS-PRS). A positive KRAS-PRS (i.e., KRAS-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., a KRAS, MEK1, MEK2,and / or ERK inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward KRAS, MEK1, MEK2 or ERK broadly and / or specifically. The first type of therapy (e.g., KRAS, MEK1, MEK2 or ERK inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward KRAS, MEK1, MEK2 or ERK broadly and / or specifically can be a small molecule, an antibody, an antibody-conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward KRAS, MEK1, MEK2 or ERK can any therapeutic agent known in the art and / or provided herein.

[0111] In some cases, the subject with the positive KRAS-PRS can be non-responsive or even resistant to the second type of therapy. In some cases, the second type of therapy can be a therapeutic agent known in the art and / or provided herein that has inhibitory activity toward ERBB2 (e.g., an ERBB2 inhibitor), an ERBB4 (e.g., ERBB4 inhibitor) or an EGFR (e.g., an EGFR inhibitor).

[0112] In one embodiment, a negative KRAS-AS or KRAS-PRS (i.e., KRAS-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward KRAS, ME1, MEK2, or ERK generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0113] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine a KRAS activation signature (KRAS-AS) or KRAS-PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 6. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers consists of orAttorney Docket No. GNCN-024 / 03WO 320289-2158 comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers for use in the methods for determining a KRAS- AS or KRAS-PRS consists of only the classifier biomarkers found in Table 6. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0114] In one embodiment, the target gene is the B-Raf protooncogene (BRAF) gene. Further to this embodiment, the methods, compositions or kits provided herein can be used toprovide a n assessment or determination of the BRAF mutational or alteration status(also referred to as a BRAF activation signature or BRAF-AS) of a sampleobtained from a subject suffering from or suspected of suffering from a cancer. Themeasuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the BRAF activation signature (BRAF-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as a BRAF activation signature (BRAF-AS) or BRAF activation classifier. As alluded to herein, the BRAF-AS can reflect or represent a presence or absence of one or more BRAF mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose BRAF-AS indicates that the subject possesses a BRAF alteration or mutation is said to have a positive BRAF-AS or be BRAF-AS (+). Conversely, samples whose BRAF-AS indicates that the subject does not possess a BRAF alteration or mutation is said to have a negative BRAF-AS or be BRAF-AS (-).

[0115] Whether or not a BRAF-AS of a sample is positive or negative can be determined by comparing the BRAF determined for the sample to the BRAF-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the braf gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the braf gene. In one embodiment, the BRAF-AS of the sample obtained from the subject is compared to the BRAF-AS of a sample known to possess one or more mutations and / or fusions in the braf gene. In another embodiment, the BRAF-AS of the sample obtained from the subject is compared to the BRAF-AS of a sample known to not possess one or more mutations and / or fusions in theAttorney Docket No. GNCN-024 / 03WO 320289-2158 braf gene. In yet another embodiment, the BRAF-AS of the sample obtained from the subject is compared to the BRAF-AS of a sample known to possess one or more mutations and / or fusions in the braf gene and the BRAF-AS of a sample known to not possess one or more mutations and / or fusions in the braf gene. The one or more alterations or mutations in the braf gene can be any mutation and / or fusion in the braf gene known in the art such as, for example, the braf gene mutations that result in the BRAFV600E mutation.

[0116] In one embodiment, a positive BRAF activation signature (BRAF-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward BRAF broadly and / or specifically. In this way, the BRAF-AS is acting as or can be referred to as a BRAF predictive response signature (BRAF- PRS). A positive BRAF-PRS (i.e., BRAF-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., a BRAF inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward BRAF broadly and / or specifically. The first type of therapy (e.g., a BRAF inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward BRAF broadly and / or specifically can be a small, an antibody, an antibody-conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward BRAF can be any therapeutic agent known in the art and / or provided herein.

[0117] In one embodiment, a negative BRAF-AS or BRAF-PRS (i.e., BRAF-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward BRAF generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0118] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine a BRAF activation signature (BRAF-AS) or BRAF-PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifierAttorney Docket No. GNCN-024 / 03WO 320289-2158 biomarkers listed in Table 8. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199 or 200 classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers for use in the methods for determining a BRAF-AS or BRAF-PRS consists of only the classifier biomarkers found in Table 8. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0119] In one embodiment, the target gene is the Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene. Further to this embodiment, the methods, compositions or kits provided herein canbe used to provide a n assessment or determination of the ERBB2 mutational oralteration status (also referred to as an ERBB2 activation signature or ERBB2-AS) of a sample obtained from a subject suffering from or suspected of suffering from acancer. The measuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the ERBB2 activation signature (ERBB2-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as an ERBB2 activation signature (ERBB2-AS) or ERBB2 activation classifier. As alluded to herein, the ERBB2-AS can reflect or represent a presence or absence of one or more ERBB2 mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose ERBB2-AS indicates that theAttorney Docket No. GNCN-024 / 03WO 320289-2158 subject possesses an ERBB2 alteration or mutation is said to have a positive ERBB2-AS or be ERBB2-AS (+). Conversely, samples whose ERBB2-AS indicates that the subject does not possess an ERBB2 alteration or mutation is said to have a negative ERBB2-AS or be ERBB2-AS (-).

[0120] Whether or not an ERBB2-AS of a sample is positive or negative can be determined by comparing the ERBB2 determined for the sample to the ERBB2-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the erbb2 gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the erbb2 gene. In one embodiment, the ERBB2-AS of the sample obtained from the subject is compared to the ERBB2-AS of a sample known to possess one or more mutations and / or fusions in the erbb2 gene. In another embodiment, the ERBB2-AS of the sample obtained from the subject is compared to the ERBB2-AS of a sample known to not possess one or more mutations and / or fusions in the erbb2 gene. In yet another embodiment, the ERBB2-AS of the sample obtained from the subject is compared to the ERBB2-AS of a sample known to possess one or more mutations and / or fusions in the erbb2 gene and the ERBB2-AS of a sample known to not possess one or more mutations and / or fusions in the erbb2 gene. The one or more alterations or mutations in the erbb2 gene can be any mutation and / or fusion in the erbb2 gene known in the art.

[0121] In one embodiment, a positive ERBB2 activation signature (ERBB2-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward ERBB2 broadly and / or specifically. In this way, the ERBB2-AS is acting as or can be referred to as an ERBB2 predictive response signature (ERBB2-PRS). A positive ERBB2-PRS (i.e., ERBB2-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., an ERBB2 inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward ERBB2 broadly and / or specifically. The first type of therapy (e.g., an ERBB2 inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward ERBB2 broadly and / or specifically can be a small, an antibody, an antibody-Attorney Docket No. GNCN-024 / 03WO 320289-2158 conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward ERBB2 can be any therapeutic agent known in the art and / or provided herein.

[0122] In one embodiment, a negative ERBB2-AS or ERBB2-PRS (i.e., ERBB2-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward ERBB2 generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0123] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine an ERBB2 activation signature (ERBB2-AS) or ERBB2- PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 10. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223 or 224 classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers for use in the methods for determining an ERBB2-AS or ERBB2-PRSAttorney Docket No. GNCN-024 / 03WO 320289-2158 consists of only the classifier biomarkers found in Table 10. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0124] In one embodiment, the target gene is the Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene. Further to this embodiment, the methods, compositions or kits provided herein canbe used to provide a n assessment or determination of the ERBB3 mutational oralteration status (also referred to as an ERBB3 activation signature or ERBB3-AS) of a sample obtained from a subject suffering from or suspected of suffering from acancer. The measuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the ERBB3 activation signature (ERBB3-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as an ERBB3 activation signature (ERBB3-AS) or ERBB3 activation classifier. As alluded to herein, the ERBB3-AS can reflect or represent a presence or absence of one or more ERBB3 mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose ERBB3-AS indicates that the subject possesses an ERBB3 alteration or mutation is said to have a positive ERBB3-AS or be ERBB3-AS (+). Conversely, samples whose ERBB3-AS indicates that the subject does not possess an ERBB3 alteration or mutation is said to have a negative ERBB3-AS or be ERBB3-AS (-).

[0125] Whether or not an ERBB3-AS of a sample is positive or negative can be determined by comparing the ERBB3 determined for the sample to the ERBB3-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the erbb3 gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the erbb3 gene. In one embodiment, the ERBB3-AS of the sample obtained from the subject is compared to the ERBB3-AS of a sample known to possess one or more mutations and / or fusions in the erbb3 gene. In another embodiment, the ERBB3-AS of the sample obtained from the subject is compared to the ERBB3-AS of a sample known to not possess one or more mutations and / or fusions in the erbb3 gene. In yet another embodiment, the ERBB3-AS of the sample obtained from the subject is compared to the ERBB3-AS of a sample known to possess one or more mutations and / or fusions in the erbb3 gene and the ERBB3-AS of a sample known to not possess one or more mutations and / or fusions in the erbb3 gene. The one or more alterations or mutations in the erbb3 gene can be any mutation and / or fusion in the erbb3 gene known in the art.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0126] In one embodiment, a positive ERBB3 activation signature (ERBB3-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward ERBB3 broadly and / or specifically. In this way, the ERBB3-AS is acting as or can be referred to as an ERBB3 predictive response signature (ERBB3-PRS). A positive ERBB3-PRS (i.e., ERBB3-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., an ERBB3 inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward ERBB3 broadly and / or specifically. The first type of therapy (e.g., an ERBB3 inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward ERBB3 broadly and / or specifically can be a small, an antibody, an antibody- conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward ERBB3 can be any therapeutic agent known in the art and / or provided herein.

[0127] In one embodiment, a negative ERBB3-AS or ERBB3-PRS (i.e., ERBB3-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward ERBB3 generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0128] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine an ERBB3 activation signature (ERBB3-AS) or ERBB3- PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 12. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%Attorney Docket No. GNCN-024 / 03WO 320289-2158 or 99% of the classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers for use in the methods for determining an ERBB3-AS or ERBB3-PRS consists of only the classifier biomarkers found in Table 12. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0129] In one embodiment, the target gene is the Phosphoinositide 3-Kinase CA (PIK3CA) gene. Further to this embodiment, the methods, compositions or kits provided hereincan be used to provide a n assessment or determination of the PIK3CA mutational oralteration status (also referred to as a PIK3CA activation signature or PIK3CA-AS) of a sample obtained from a subject suffering from or suspected of suffering from acancer. The measuring of the expression level of the defined set of biomarkers generates or produces an expression profile that represents the PIK3CA activation signature (PIK3CA-AS) of the sample. In this way, a set of biomarkers as provided herein can each be referred to as a PIK3CA activation signature (PIK3CA-AS) or PIK3CA activation classifier. As alluded to herein, the PIK3CA-AS can reflect or represent a presence or absence of one or more PIK3CA mutation(s) or alteration(s) in the sample obtained from the subject. Samples whose PIK3CA-AS indicates that the subject possesses a PIK3CA alteration or mutation is said to have a positive PIK3CA-AS or be PIK3CA-AS (+). Conversely, samples whose PIK3CA-AS indicates that the subject does not possess a PIK3CA alteration or mutation is said to have a negative PIK3CA-AS or be PIK3CA- AS (-).

[0130] Whether or not a PIK3CA-AS of a sample is positive or negative can be determined by comparing the PIK3CA determined for the sample to the PIK3CA-AS for one or more reference or control samples. In one embodiment, the reference or control sample is a sample known to possess one or more mutations and / or fusions in the pik3ca gene. In one embodiment, the reference or control sample is a sample known to not possess or harbor one or more mutations and / or fusions in the pik3ca gene. In one embodiment, the PIK3CA-AS of the sample obtained from the subject is compared to the PIK3CA-AS of a sample known to possess one or more mutations and / or fusions in the pik3ca gene. In another embodiment, the PIK3CA-AS of the sample obtained from the subject is compared to the PIK3CA-AS of a sample known to notAttorney Docket No. GNCN-024 / 03WO 320289-2158 possess one or more mutations and / or fusions in the pik3ca gene. In yet another embodiment, the PIK3CA-AS of the sample obtained from the subject is compared to the PIK3CA-AS of a sample known to possess one or more mutations and / or fusions in the pik3ca gene and the PIK3CA-AS of a sample known to not possess one or more mutations and / or fusions in the pik3ca gene. The one or more alterations or mutations in the pik3ca gene can be any mutation and / or fusion in the pik3ca gene known in the art.

[0131] In one embodiment, a positive PIK3CA activation signature (PIK3CA-AS) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward PIK3CA broadly and / or specifically. In this way, the PIK3CA-AS is acting as or can be referred to as a PIK3CA predictive response signature (PIK3CA-PRS). A positive PIK3CA-PRS (i.e., PIK3CA-PRS (+)) can indicate that the subject can be responsive to a first type of therapy (e.g., a PIK3CA inhibitory therapeutic agent) and / or not responsive to a second type of therapy. The first type of therapy can be a therapeutic agent or defined set of therapeutic agents that exhibit(s) inhibitory activity toward PIK3CA broadly and / or specifically. The first type of therapy (e.g., an PIK3CA inhibitory therapeutic agent) can be administered to the subject in a therapeutically effective dose or doses alone or in combination with one or more additional therapeutic agents or modalities as described herein. The therapeutic agent that exhibit(s) inhibitory activity toward PIK3CA broadly and / or specifically can be a small, an antibody, an antibody-conjugate or any combination thereof. The therapeutic agent that exhibit(s) inhibitory activity toward PIK3CA can be any therapeutic agent known in the art and / or provided herein.

[0132] In one embodiment, a negative PIK3CA-AS or PIK3CA-PRS (i.e., PIK3CA-PRS (-)) of a sample obtained from a subject suffering from or suspected of suffering from a cancer indicates that the subject may be responsive to a therapeutic agent or defined set of therapeutic agents other than those that exhibit(s) inhibitory activity toward PIK3CA generally and / or specifically such as one or more therapeutic agents or modalities known in the art and / or as described herein.

[0133] In one embodiment, the set of biomarkers for use in the compositions, methods and kits provided herein in order to determine a PIK3CA activation signature (PIK3CA-AS) orAttorney Docket No. GNCN-024 / 03WO 320289-2158 PIK3CA-PRS of a sample obtained from a subject is a plurality of classifier biomarkers selected from the classifier biomarkers listed in Table 14. In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 or 36 classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers for use in the methods for determining an PIK3CA-AS or PIK3CA-PRS consists of only the classifier biomarkers found in Table 14. The detection can be ascertained by using any amplification, hybridization and / or sequencing assay disclosed herein.

[0134] The expression level of any and all genes utilized in a specific target activation signature or predictive response signature or combination of specific target activation signature or predictive response signature as provided herein can be normalized as provided herein, such as, for example, normalizing expression of the classifier genes or classifier gene pairs by using expression levels from one or more reference or housekeeping genes. The housekeeping genes can be any housekeeping genes known in the art and / or provided herein such as, for example, GAPDH and / or beta-actin. Measuring Biomarkers Expression Levels

[0135] In one embodiment, the detecting, determining or measuring the expression level of any classifier biomarker in any sample in any of the methods provided herein is performed at the nucleic acid level. The nucleic acid can be DNA, cDNA or RNA. Measuring the nucleic acid level can be performed by any suitable technique including, but not limited to, RNA-seq, a reverse transcriptase polymerase chain reaction (RT-PCR), a microarray hybridization assay, or another hybridization assay, e.g., a NanoString assay for example, with primers and / or probes specific to the classifier biomarkers, and / or the like. In some cases, the primers useful for the amplification methods (e.g., RT-PCR or qRT-PCR) are any forward and reverse primers suitable for binding to a classifier biomarker gene provided herein, such as the classifier biomarkers listed in Tables 2, 4, 6, 8, 10, 12 and / or 14.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0136] In one embodiment, the measuring or detecting step for methods provided herein that comprise determining a specific target gene activation signature (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) or specific target gene predictive response signature (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) of a sample obtained from a subject suffering from or suspected of suffering from a cancer as provided herein is at the nucleic acid level. The measuring or detecting step can entail performing RNA-seq, a reverse transcriptase polymerase chain reaction (RT-PCR) or a hybridization assay with oligonucleotides that are substantially complementary to portions of cDNA molecules of the at least one or plurality of classifier biomarker(s) of Tables 2, 4, 6, 8, 10, 12 and / or 14 under conditions suitable for RNA- seq, RT-PCR or hybridization and obtaining expression levels of the at least one or plurality of classifier biomarkers based on the detecting step. In one embodiment, the measuring or detecting step for methods of determining an AS or PRS as provided herein comprises mixing the sample with one or more oligonucleotides that are complementary or substantially complementary to portions of cDNA molecules of the at least one or plurality of classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 under conditions suitable for hybridization of the one or more oligonucleotides to their complements or substantial complements; detecting whether hybridization occurs between the one or more oligonucleotides to their complements or substantial complements; and obtaining hybridization values of the at least one or plurality of classifier biomarkers based on the detecting step such that the hybridization values represent expression levels. In one embodiment, the measuring or detecting step for methods of determining a specific target gene activation signature (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) or specific target gene predictive response signature (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) as provided herein comprises mixing the sample with oligonucleotides that are complementary or substantially complementary to portions of DNA (e.g., cDNA) molecules of the at least one or plurality of classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 under conditions suitable for hybridization of the oligonucleotides to their complements or substantial complements and subsequent amplification of said DNA (e.g. cDNA); detecting whether amplification occurred between the oligonucleotides and their complements or substantial complements; and obtaining expression levels of the amplicons of the at least one or plurality of classifier biomarkers based on the detecting step.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0137] In one embodiment, the expression levels of the at least one or plurality of the classifier biomarkers of the sample obtained from the subject suffering from or suspected of suffering from a cancer are then compared to reference expression levels of the at least one or plurality of the classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 from at least one sample training set. The at least one sample training set can comprise, (i) expression levels from a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (+) sample and / or (ii) expression levels from a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (-) sample. The sample can then be classified as a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (+) or a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (-) subtype or sample based on the results of the comparing step.

[0138] In one embodiment, the comparing step can comprise applying a statistical algorithm which comprises determining a correlation between the expression data obtained by measuring one or a plurality of biomarkers from Tables 2, 4, 6, 8, 10, 12 and / or 14 on the sample obtained from the subject and the expression data from the at least one training set(s); and classifying the sample as a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (+) or a target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS (-) subtype or sample based on the results of the statistical algorithm. The statistical algorithm can entail finding the centroid to which the target gene specific (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 and / or PIK3CA) AS of the sample obtained from the subject is nearest from the centroids constructed from the expression data from the at least one training set, using any distance measure e.g., Euclidean distance or correlation. The centroids can be constructed using any method known in the art for generating centroids such as, for example, those found in Mullins et al. (2007) Clin Chem. 53(7):1273-9 or Dabney (2005) Bioinformatics 21(22):4148-4154 The FAS of the sample obtained from subject can then be assigned based on the use of a classification to the nearest centroid (CLaNC) algorithm as applied to the expression data generated from the sample obtained from the subject and the centroid(s) constructed for the at least one training set. The CLaNC algorithm for use in the methods, compositions and kits provided herein can be the CLaNC algorithm implemented by the CLaNC software found in Dabney AR. ClaNC: Point-and-click software for classifying microarrays toAttorney Docket No. GNCN-024 / 03WO 320289-2158 nearest centroids. Bioinformatics.2006;22: 122–123, which is herein incorporated by reference in its entirety or equivalents or derivatives thereof.

[0139] The classifier biomarkers described herein can include RNA comprising the entire or partial sequence of any of the nucleic acid sequences of interest, or their non-natural cDNA product, obtained synthetically in vitro in a reverse transcription reaction. The term “fragment” is intended to refer to a portion of the polynucleotide that generally comprise at least 10, 15, 20, 50, 75, 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 800, 900, 1,000, 1,200, or 1,500 contiguous nucleotides, or up to the number of nucleotides present in a full-length biomarker polynucleotide disclosed herein. A fragment of a biomarker polynucleotide will generally encode at least 15, 25, 30, 50, 100, 150, 200, or 250 contiguous amino acids, or up to the total number of amino acids present in a full-length biomarker protein of the invention.

[0140] In some embodiments, overexpression, such as of an RNA transcript or its expression product, is determined by normalization to the level of reference RNA transcripts or their expression products, which can be all measured transcripts (or their products) in the sample or a particular reference set of RNA transcripts (or their non-natural cDNA products). Normalization is performed to correct for or normalize away both differences in the amount of RNA or cDNA assayed and variability in the quality of the RNA or cDNA used. Therefore, an assay typically measures and incorporates the expression of certain normalizing genes, including -Actin. Alternatively, normalization can be based on the mean or median signal of all of the assayed biomarkers or a large subset thereof (global normalization approach).

[0141] Isolated mRNA from samples obtained from a patient or subject can be used in the methods, compositions and kits provided herein. As necessitated, the isolated mRNA can be used in hybridization or amplification assays that include, but are not limited to, Southern or Northern analyses, PCR analyses and probe arrays, NanoString Assays. One method for the detection of mRNA levels involves contacting the isolated mRNA or synthesized cDNA with a nucleic acid molecule (probe) that can hybridize to the mRNA encoded by the gene being detected. The nucleic acid probe can be, for example, a cDNA, or a portion thereof, such as an oligonucleotide of at least 7, 15, 30, 50, 100, 250, or 500 nucleotides in length and sufficient to specifically hybridize under stringent conditions to the non-natural cDNA or mRNA biomarker of the present invention.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0142] As explained above, in one embodiment, once the mRNA is obtained from a sample, it is converted to complementary DNA (cDNA) in a hybridization reaction. Conversion of the mRNA to cDNA can be performed with oligonucleotides or primers comprising sequence that is complementary to a portion of a specific mRNA. Conversion of the mRNA to cDNA can be performed with oligonucleotides or primers comprising random sequence. Conversion of the mRNA to cDNA can be performed with oligonucleotides or primers comprising sequence that is complementary to the poly(A) tail of an mRNA. cDNA does not exist in vivo and therefore is a non-natural molecule. In a further embodiment, the cDNA is then amplified, for example, by the polymerase chain reaction (PCR) or other amplification method known to those of ordinary skill in the art. PCR can be performed with the forward and / or reverse primers comprising sequence complementary to at least a portion of a classifier gene provided herein, such as the classifier biomarkers in Tables 2, 4, 6, 8, 10, 12 and / or 14. The product of this amplification reaction, i.e., amplified cDNA is necessarily a non-natural product. As mentioned above, cDNA is a non-natural molecule. Second, in the case of PCR, the amplification process serves to create hundreds of millions of cDNA copies for every individual cDNA molecule of starting material. The number of copies generated is far removed from the number of copies of mRNA that are present in vivo.

[0143] In one embodiment, cDNA is amplified with primers that introduce an additional DNA sequence (adapter sequence) onto the fragments (with the use of adapter-specific primers). The adaptor sequence can be a tail, wherein the tail sequence is not complementary to the cDNA. For example, the forward and / or reverse primers comprising sequence complementary to at least a portion of a classifier gene provided herein, such as the classifier biomarkers from Tables 2, 4, 6, 8, 10, 12 and / or 14 can comprise tail sequence. Amplification therefore serves to create non- natural double stranded molecules from the non-natural single stranded cDNA, by introducing barcode, adapter and / or reporter sequences onto the already non-natural cDNA. In one embodiment, during amplification with the adapter-specific primers, a detectable label, e.g., a fluorophore, is added to single strand cDNA molecules. Amplification therefore also serves to create DNA complexes that do not occur in nature, at least because (i) cDNA does not exist in vivo, (ii) adapter sequences are added to the ends of cDNA molecules to make DNA sequences that do not exist in vivo, (iii) the error rate associated with amplification further creates DNA sequences that do not exist in vivo, (iv) the disparate structure of the cDNA molecules as compared to what exists in nature, and (v) the chemical addition of a detectable label to the cDNA molecules.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0144] In one embodiment, the synthesized cDNA (for example, amplified cDNA) is immobilized on a solid surface via hybridization with a probe, e.g., via a microarray. In another embodiment, cDNA products are detected via real-time polymerase chain reaction (PCR) via the introduction of fluorescent probes that hybridize with the cDNA products. For example, in one embodiment, biomarker detection is assessed by quantitative fluorogenic RT-PCR (e.g., with TaqMan® probes). For PCR analysis, well known methods are available in the art for the determination of primer sequences for use in the analysis.

[0145] Classifier biomarkers provided herein in one embodiment, are detected via a hybridization reaction that employs a capture probe and / or a reporter probe. For example, the hybridization probe is a probe derivatized to a solid surface such as a bead, glass or silicon substrate. In another embodiment, the capture probe is present in solution and mixed with the patient’s sample, followed by attachment of the hybridization product to a surface, e.g., via a biotin-avidin interaction (e.g., where biotin is a part of the capture probe and avidin is on the surface). The hybridization assay, in one embodiment, employs both a capture probe and a reporter probe. The reporter probe can hybridize to either the capture probe or the biomarker nucleic acid. Reporter probes e.g., are then counted and detected to determine the level of biomarker(s) in the sample. The capture and / or reporter probe, in one embodiment contain a detectable label, and / or a group that allows functionalization to a surface.

[0146] For example, the nCounter gene analysis system (see, e.g., Geiss et al. (2008) Nat. Biotechnol.26, pp.317-325, incorporated by reference in its entirety for all purposes, is amenable for use with the methods provided herein.

[0147] Hybridization assays described in U.S. Patent Nos. 7,473,767 and 8,492,094, the disclosures of which are incorporated by reference in their entireties for all purposes, are amenable for use with the methods provided herein, i.e., to detect the biomarkers and biomarker combinations described herein.

[0148] Classifier biomarker levels may be monitored using a membrane blot (such as used in hybridization analysis such as Northern, Southern, dot, and the like), or microwells, sample tubes, gels, beads, or fibers (or any solid support comprising bound nucleic acids). See, for example, U.S. Pat. Nos. 5,770,722, 5,874,219, 5,744,305, 5,677,195 and 5,445,934, each incorporated by reference in their entireties.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0149] In one embodiment, microarrays are used to detect classifier biomarker levels. Microarrays are particularly well suited for this purpose because of the reproducibility between different experiments. DNA microarrays provide one method for the simultaneous measurement of the expression levels of large numbers of genes. Each array consists of a reproducible pattern of capture probes attached to a solid support. Labeled RNA or DNA is hybridized to complementary probes on the array and then detected by laser scanning hybridization intensities for each probe on the array are determined and converted to a quantitative value representing relative gene expression levels. See, for example, U.S. Pat. Nos. 6,040,138, 5,800,992 and 6,020,135, 6,033,860, and 6,344,316, each incorporated by reference in their entireties. High- density oligonucleotide arrays are particularly useful for determining the gene expression profile for a large number of RNAs in a sample.

[0150] Techniques for the synthesis of these arrays using mechanical synthesis methods are described in, for example, U.S. Pat. No.5,384,261. Although a planar array surface is generally used, the array can be fabricated on a surface of virtually any shape or even a multiplicity of surfaces. Arrays can be nucleic acids (or peptides) on beads, gels, polymeric surfaces, fibers (such as fiber optics), glass, or any other appropriate substrate. See, for example, U.S. Pat. Nos. 5,770,358, 5,789,162, 5,708,153, 6,040,193 and 5,800,992, each incorporated by reference in their entireties. Arrays can be packaged in such a manner as to allow for diagnostics or other manipulation of an all-inclusive device. See, for example, U.S. Pat. Nos.5,856,174 and 5,922,591, each incorporated by reference in their entireties.

[0151] Serial analysis of gene expression (SAGE) in one embodiment is employed in the methods described herein. SAGE is a method that allows the simultaneous and quantitative analysis of a large number of gene transcripts, without the need of providing an individual hybridization probe for each transcript. First, a short sequence tag (about 10-14 bp) is generated that contains sufficient information to uniquely identify a transcript, provided that the tag is obtained from a unique position within each transcript. Then, many transcripts are linked together to form long serial molecules, that can be sequenced, revealing the identity of the multiple tags simultaneously. The expression pattern of any population of transcripts can be quantitatively evaluated by determining the abundance of individual tags, and identifying the gene corresponding to each tag. See, Velculescu et al. Science 270:484-87, 1995; Cell 88:243-51, 1997, incorporated by reference in its entirety.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0152] An additional method of classifier biomarker level analysis at the nucleic acid level is the use of a sequencing method, for example, RNAseq, next generation sequencing, and massively parallel signature sequencing (MPSS), as described by Brenner et al. (Nat. Biotech. 18:630-34, 2000, incorporated by reference in its entirety). This is a sequencing approach that combines non-gel-based signature sequencing with in vitro cloning of millions of templates on separate 5 μm diameter microbeads. First, a microbead library of DNA templates is constructed by in vitro cloning. This is followed by the assembly of a planar array of the template-containing microbeads in a flow cell at a high density (typically greater than 3.0 X 106microbeads / cm2). The free ends of the cloned templates on each microbead are analyzed simultaneously, using a fluorescence-based signature sequencing method that does not require DNA fragment separation. This method has been shown to simultaneously and accurately provide, in a single operation, hundreds of thousands of gene signature sequences from a yeast cDNA library.

[0153] Another method of classifier biomarker level analysis at the nucleic acid level is the use of an amplification method such as, for example, RT-PCR or quantitative RT-PCR (qRT- PCR). Methods for determining the level of biomarker mRNA in a sample may involve the process of nucleic acid amplification, e.g., by RT-PCR (the experimental embodiment set forth in Mullis, 1987, U.S. Pat. No.4,683,202), ligase chain reaction (Barany (1991) Proc. Natl. Acad. Sci. USA 88:189-193), self-sustained sequence replication (Guatelli et al. (1990) Proc. Natl. Acad. Sci. USA 87:1874-1878), transcriptional amplification system (Kwoh et al. (1989) Proc. Natl. Acad. Sci. USA 86:1173-1177), Q-Beta Replicase (Lizardi et al. (1988) Bio / Technology 6:1197), rolling circle replication (Lizardi et al., U.S. Pat. No.5,854,033) or any other nucleic acid amplification method, followed by the detection of the amplified molecules using techniques well known to those of skill in the art. Numerous different PCR or qRT-PCR protocols are known in the art and can be directly applied or adapted for use using the presently described compositions for the detection and / or quantification of expression of discriminative genes in a sample. See, for example, Fan et al. (2004) Genome Res.14:878-885, herein incorporated by reference. Generally, in PCR, a target polynucleotide sequence is amplified by reaction with at least one oligonucleotide primer or pair of oligonucleotide primers. The primer(s) hybridize to a complementary region of the target nucleic acid and a DNA polymerase extends the primer(s) to amplify the target sequence. Under conditions sufficient to provide polymerase-based nucleic acid amplification products, a nucleic acid fragment of one size dominates the reaction products (the target polynucleotide sequenceAttorney Docket No. GNCN-024 / 03WO 320289-2158 which is the amplification product). The amplification cycle is repeated to increase the concentration of the single target polynucleotide sequence. The reaction can be performed in any thermocycler commonly used for PCR.

[0154] Quantitative RT-PCR (qRT-PCR) (also referred to as real-time RT-PCR) is preferred under some circumstances because it provides not only a quantitative measurement, but also reduced time and contamination. As used herein, "quantitative PCR” (or "real time qRT- PCR") refers to the direct monitoring of the progress of a PCR amplification as it is occurring without the need for repeated sampling of the reaction products. In quantitative PCR, the reaction products may be monitored via a signaling mechanism (e.g., fluorescence) as they are generated and are tracked after the signal rises above a background level but before the reaction reaches a plateau. The number of cycles required to achieve a detectable or "threshold" level of fluorescence varies directly with the concentration of amplifiable targets at the beginning of the PCR process, enabling a measure of signal intensity to provide a measure of the amount of target nucleic acid in a sample in real time. A DNA binding dye (e.g., SYBR green) or a labeled probe can be used to detect the extension product generated by PCR amplification. Any probe format utilizing a labeled probe comprising the sequences of the invention may be used.

[0155] Immunohistochemistry methods are also suitable for detecting the levels of the classifier biomarkers of the present invention. Samples can be frozen for later preparation or immediately placed in a fixative solution. Tissue samples can be fixed by treatment with a reagent, such as formalin, glutaraldehyde, methanol, or the like and embedded in paraffin. Methods for preparing slides for immunohistochemical analysis from formalin-fixed, paraffin-embedded tissue samples are well known in the art.

[0156] In one embodiment, the levels of the classifier biomarkers provided herein, such as the classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 (or subsets thereof as provided herein), are normalized against the expression levels of all RNA transcripts or their non-natural cDNA expression products, or protein products in the sample, or of a reference set of RNA transcripts or a reference set of their non-natural cDNA expression products, or a reference set of their protein products in the sample.

[0157] In one embodiment, the detecting, determining or measuring the expression level of any classifier biomarker in any of the methods provided herein is performed at the protein level. In one embodiment, an activation signature or predictive response signature provided hereinAttorney Docket No. GNCN-024 / 03WO 320289-2158 be evaluated using levels of protein expression of one or more of the classifier genes provided herein, such as the classifier biomarkers listed in Tables 2, 4, 6, 8, 10, 12 and / or 14. The level of protein expression can be measured using an immunological detection method. Immunological detection methods which can be used herein include, but are not limited to, competitive and non-competitive assay systems using techniques such as Western blots, radioimmunoassays, ELISA (enzyme linked immunosorbent assay), "sandwich" immunoassays, immunoprecipitation assays, precipitin reactions, gel diffusion precipitin reactions, immunodiffusion assays, agglutination assays, complement-fixation assays, immunoradiometric assays, fluorescent immunoassays, protein A immunoassays, and the like. Such assays are routine and well known in the art (see, e.g., Ausubel e t a l , eds, 1994,Current Protocols in Molecular Biology, Vol. I, John Wiley & Sons, Inc., New York, whichis incorporated by reference herein in its entirety).

[0158] In one embodiment, antibodies specific for c l a s s i f i e r biomarker proteins areutilized to detect the expression of a classifier biomarker protein in a body sample. The method comprises obtaining a body sample from a patient or a subject, contacting the body sample with at least one antibody directed to a classifier biomarker selected from Tables 2, 4, 6, 8, 10, 12 and / or 14, and detecting antibody binding to determine if the classifier biomarker is expressed in the patient sample. One of skill in the art will recognize that the immunocytochemistry method described herein below may be performed manually or in an automated fashion.

[0159] As provided throughout, the methods set forth herein provide a method for determining an activation signature or predictive response signature provided herein of a subject. Once the classifier biomarker levels are determined, for example by measuring non-natural cDNA biomarker levels or non-natural mRNA-cDNA biomarker complexes, the classifier biomarker levels can be compared to reference values or a reference sample, for example with the use of statistical methods or direct comparison of detected levels, to make a determination of the activation signature or predictive response signature. Based on the comparison, the patient’s sample is classified as being AS (+) or (-) (or PRS (+) or PRS(-)).

[0160] In one embodiment, expression level values of the at least one classifier biomarkers provided herein, such as the classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 areAttorney Docket No. GNCN-024 / 03WO 320289-2158 compared to reference expression level value(s) from at least one sample training set, wherein the at least one sample training set comprises expression level values from a reference sample(s).

[0161] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 2 from a sample known to possess one or more alterations or mutations in an egfr gene alone, a sample known not to possesses one or more alterations or mutations in an egfr gene alone or a combination thereof. The one or more alterations or mutations in the egfr gene can be any mutation and / or fusion in the egfr gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the EGFR-AS is then made.

[0162] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 2 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 2 from a sample known to possess one or more alterations or mutations in an egfr gene alone, a sample known not to possesses one or more alterations or mutations in an egfr gene alone or a combination thereof. The one or more alterations or mutations in the egfr gene can be any mutation and / or fusion in the egfr gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the EGFR-AS is then made.

[0163] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 4 from a sample known to possess one or more alterations or mutations in a met gene alone, a sample known not to possesses one or more alterations or mutations in a met gene alone or a combination thereof. The one or more alterations or mutations in the met gene can be any mutation and / or fusion in the met gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlationAttorney Docket No. GNCN-024 / 03WO 320289-2158 between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the MET-AS is then made.

[0164] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 4 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 4 from a sample known to possess one or more alterations or mutations in a met gene alone, a sample known not to possesses one or more alterations or mutations in a met gene alone or a combination thereof. The one or more alterations or mutations in the met gene can be any mutation and / or fusion in the met gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the MET-AS is then made.

[0165] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 6 from a sample known to possess one or more alterations or mutations in a kras gene alone, a sample known not to possesses one or more alterations or mutations in a kras gene alone or a combination thereof. The one or more alterations or mutations in the kras gene can be any mutation and / or fusion in the kras gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the KRAS-AS is then made.

[0166] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 6 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 6 from a sample known to possess one or more alterations or mutations in a kras gene alone, a sample known not to possesses one or more alterations or mutations in a kras gene alone or a combination thereof. The one or more alterations or mutations in the kras gene can be any mutation and / or fusion in theAttorney Docket No. GNCN-024 / 03WO 320289-2158 kras gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the KRAS-AS is then made.

[0167] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 8 from a sample known to possess one or more alterations or mutations in a braf gene alone, a sample known not to possesses one or more alterations or mutations in a braf gene alone or a combination thereof. The one or more alterations or mutations in the braf gene can be any mutation and / or fusion in the braf gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the BRAF-AS is then made.

[0168] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 8 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 8 from a sample known to possess one or more alterations or mutations in a braf gene alone, a sample known not to possesses one or more alterations or mutations in a braf gene alone or a combination thereof. The one or more alterations or mutations in the braf gene can be any mutation and / or fusion in the braf gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the BRAF-AS is then made.

[0169] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 10 from a sample known to possess one or more alterations or mutations in an erbb2 gene alone, a sample known not to possesses one or more alterations or mutations in an erbb2 gene alone or a combination thereof. The one or more alterations or mutations in the erbb2 gene can be any mutation and / or fusion in the erbb2 gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / orAttorney Docket No. GNCN-024 / 03WO 320289-2158 reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the ERBB2-AS is then made.

[0170] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 10 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 10 from a sample known to possess one or more alterations or mutations in an erbb2 gene alone, a sample known not to possesses one or more alterations or mutations in an erbb2 gene alone or a combination thereof. The one or more alterations or mutations in the erbb2 gene can be any mutation and / or fusion in the erbb2 gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the ERBB2-AS is then made.

[0171] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 12 from a sample known to possess one or more alterations or mutations in an erbb3 gene alone, a sample known not to possesses one or more alterations or mutations in an erbb3 gene alone or a combination thereof. The one or more alterations or mutations in the erbb3 gene can be any mutation and / or fusion in the erbb3 gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the ERBB3-AS is then made.

[0172] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 12 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 12 from a sample known to possess one or more alterations or mutations in an erbb3 gene alone, a sample known not to possesses one or more alterations or mutations in an erbb3 gene alone or a combinationAttorney Docket No. GNCN-024 / 03WO 320289-2158 thereof. The one or more alterations or mutations in the erbb3 gene can be any mutation and / or fusion in the erbb3 gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the ERBB3-AS is then made.

[0173] In one embodiment, the at least one sample training set comprises expression level values of the plurality of classifier biomarkers found in Table 14 from a sample known to possess one or more alterations or mutations in a pik3ca gene alone, a sample known not to possesses one or more alterations or mutations in a pik3ca gene alone or a combination thereof. The one or more alterations or mutations in the pik3ca gene can be any mutation and / or fusion in the pik3ca gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the PIK3CA-AS is then made.

[0174] In a separate embodiment, hybridization values of the plurality of classifier biomarkers found in Table 14 are compared to reference hybridization value(s) from at least one sample training set, wherein the at least one sample training set comprises hybridization values from a reference sample(s). In a further embodiment, the at least one sample training set comprises hybridization values of the plurality of classifier biomarkers found in Table 14 from a sample known to possess one or more alterations or mutations in a pik3ca gene alone, a sample known not to possesses one or more alterations or mutations in a pik3ca gene alone or a combination thereof. The one or more alterations or mutations in the pik3ca gene can be any mutation and / or fusion in the pik3ca gene known in the art. Methods for comparing detected levels of biomarkers to reference values and / or reference samples are provided herein. Based on this comparison, in one embodiment a correlation between the biomarker levels obtained from the subject’s sample and the reference values is obtained. An assessment of the PIK3CA-AS is then made. Sample Types

[0175] In one embodiment, the sample used in any method provided herein is obtained from an individual and comprises formalin-fixed paraffin-embedded (FFPE) tissue. However, other tissue and sample types are amenable for use in any of the methods provided herein. In oneAttorney Docket No. GNCN-024 / 03WO 320289-2158 embodiment, the other tissue and sample types can be fresh frozen tissue, wash fluids or cell pellets, or the like. In one embodiment, the sample can be a bodily fluid obtained from the individual. The bodily fluid can be blood or fractions thereof (e.g., serum, plasma), urine, sputum, saliva or cerebrospinal fluid (CSF). A biomarker or each biomarker in a pair of biomarkers for use in any method or composition provided herein can be a nucleic acid. A biomarker nucleic acid (e.g., DNA or RNA) as provided herein can be extracted from a cell or can be cell free or extracted from an extracellular vesicular entity such as an exosome or microvesicle. The sample can contain cellular as well as extracellular sources of nucleic acid for use in the methods provided herein. The methods provided herein, including the RT-PCR methods, can be sensitive, precise and have multi- analyte capability for use with paraffin embedded samples. See, for example, Cronin et al. (2004) Am. J Pathol.164(1):35-42, herein incorporated by reference.

[0176] Formalin fixation and tissue embedding in paraffin wax is a universal approach for tissue processing prior to light microscopic evaluation. A major advantage afforded by formalin- fixed paraffin-embedded (FFPE) specimens is the preservation of cellular and architectural morphologic detail in tissue sections. (Fox et al. (1985) J Histochem Cytochem 33:845-853). The standard buffered formalin fixative in which biopsy specimens are processed is typically an aqueous solution containing 37% formaldehyde and 10-15% methyl alcohol. Formaldehyde is a highly reactive dipolar compound that results in the formation of protein-nucleic acid and protein- protein crosslinks in vitro (Clark et al. (1986) J Histochem Cytochem 34:1509-1512; McGhee and von Hippel (1975) Biochemistry 14:1281-1296, each incorporated by reference herein).

[0177] Methods are known in the art for the isolation of RNA from FFPE tissue. In one embodiment, total RNA can be isolated from FFPE tissues as described by Bibikova et al. (2004) American Journal of Pathology 165:1799-1807, herein incorporated by reference. Likewise, the High Pure RNA Paraffin Kit (Roche) can be used. Paraffin is removed by xylene extraction followed by ethanol wash. RNA can be isolated from sectioned tissue blocks using the MasterPure® Purification kit (Epicenter, Madison, Wis.); a DNase I treatment step is included. RNA can be extracted from frozen samples using Trizol reagent according to the supplier's instructions (Invitrogen Life Technologies, Carlsbad, Calif.). Samples with measurable residual genomic DNA can be re-subjected to DNaseI treatment and assayed for DNA contamination. All purification, DNase treatment, and other steps can be performed according to the manufacturer's protocol. After total RNA isolation, samples can be stored at -80 ºC until use.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0178] General methods for mRNA extraction are well known in the art and are disclosed in standard textbooks of molecular biology, including Ausubel et al., ed., Current Protocols in Molecular Biology, John Wiley & Sons, New York 1987-1999. Methods for RNA extraction from paraffin embedded tissues are disclosed, for example, in Rupp and Locker (Lab Invest. 56:A67, 1987) and De Andres et al. (Biotechniques 18:42-44, 1995). In particular, RNA isolation can be performed using a purification kit, a buffer set and protease from commercial manufacturers, such as Qiagen (Valencia, Calif.), according to the manufacturer's instructions. For example, total RNA from cells in culture can be isolated using Qiagen RNeasy mini-columns. Other commercially available RNA isolation kits include MasterPure®. Complete DNA and RNA Purification Kit (Epicentre, Madison, Wis.) and Paraffin Block RNA Isolation Kit (Ambion, Austin, Tex.). Total RNA from tissue samples can be isolated, for example, using RNA Stat-60 (Tel-Test, Friendswood, Tex.). RNA prepared from a tumor can be isolated, for example, by cesium chloride density gradient centrifugation. Additionally, large numbers of tissue samples can readily be processed using techniques well known to those of skill in the art, such as, for example, the single- step RNA isolation process of Chomczynski (U.S. Pat. No.4,843,155, incorporated by reference in its entirety for all purposes).

[0179] In one embodiment, a sample for use in any of the methods provided herein comprises cells harvested from a tissue sample, for example, a tumor sample. The tumor sample can be a cancerous tumor. The cancerous tumor can be any type of cancer known in the art and / or provided herein. Cells can be harvested from a biological sample using standard techniques known in the art. For example, in one embodiment, cells are harvested by centrifuging a cell sample and resuspending the pelleted cells. The cells can be resuspended in a buffered solution such as phosphate-buffered saline (PBS). After centrifuging the cell suspension to obtain a cell pellet, the cells can be lysed to extract nucleic acid, e.g., messenger RNA. All samples obtained from a subject, including those subjected to any sort of further processing, are considered to be obtained from the subject.

[0180] The sample, in one embodiment, is further processed before the detection of the biomarker levels of the combination of biomarkers set forth herein. For example, mRNA in a cell or tissue sample can be separated from other components of the sample. The sample can be concentrated and / or purified to isolate mRNA in its non-natural state, as the mRNA is not in its natural environment. For example, studies have indicated that the higher order structure of mRNAAttorney Docket No. GNCN-024 / 03WO 320289-2158 in vivo differs from the in vitro structure of the same sequence (see, e.g., Rouskin et al. (2014). Nature 505, pp.701-705, incorporated herein in its entirety for all purposes).

[0181] mRNA from the sample in one embodiment, is hybridized to a synthetic DNA probe, which in some embodiments, includes a detection moiety (e.g., detectable label, capture sequence, barcode reporting sequence). Accordingly, in these embodiments, a non-natural mRNA- cDNA complex is ultimately made and used for detection of the biomarker. In another embodiment, mRNA from the sample is directly labeled with a detectable label, e.g., a fluorophore. In a further embodiment, the non-natural labeled-mRNA molecule is hybridized to a cDNA probe and the complex is detected.

[0182] In one embodiment, once the mRNA is obtained from a sample, it is converted to complementary DNA (cDNA) in a hybridization reaction or is used in a hybridization reaction together with one or more cDNA probes. cDNA does not exist in vivo and therefore is a non- natural molecule. Furthermore, cDNA-mRNA hybrids are synthetic and do not exist in vivo. Besides cDNA not existing in vivo, cDNA is necessarily different than mRNA, as it includes deoxyribonucleic acid and not ribonucleic acid. The cDNA is then amplified, for example, by the polymerase chain reaction (PCR) or other amplification method known to those of ordinary skill in the art. For example, other amplification methods that may be employed include the ligase chain reaction (LCR) (Wu and Wallace, Genomics, 4:560 (1989), Landegren et al., Science, 241:1077 (1988), incorporated by reference in its entirety for all purposes, transcription amplification (Kwoh et al., Proc. Natl. Acad. Sci. USA, 86:1173 (1989), incorporated by reference in its entirety for all purposes), self-sustained sequence replication (Guatelli et al., Proc. Nat. Acad. Sci. USA, 87:1874 (1990), incorporated by reference in its entirety for all purposes), incorporated by reference in its entirety for all purposes, and nucleic acid based sequence amplification (NASBA). Guidelines for selecting primers for PCR amplification are known to those of ordinary skill in the art. See, e.g., McPherson et al., PCR Basics: From Background to Bench, Springer- Verlag, 2000, incorporated by reference in its entirety for all purposes. The product of this amplification reaction, i.e., amplified cDNA is also necessarily a non-natural product. First, as mentioned above, cDNA is a non-natural molecule. Second, in the case of PCR, the amplification process serves to create hundreds of millions of cDNA copies for every individual cDNA molecule of starting material. The numbers of copies generated are far removed from the number of copies of mRNA that are present in vivo.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0183] In one embodiment, cDNA is amplified with primers that introduce an additional DNA sequence (e.g., adapter, reporter, capture sequence or moiety, barcode) onto the fragments (e.g., with the use of adapter-specific primers), or mRNA or cDNA biomarker sequences are hybridized directly to a cDNA probe comprising the additional sequence (e.g., adapter, reporter, capture sequence or moiety, barcode). Amplification and / or hybridization of mRNA to a cDNA probe therefore serves to create non-natural double stranded molecules from the non-natural single stranded cDNA, or the mRNA, by introducing additional sequences and forming non-natural hybrids. Further, as known to those of ordinary skill in the art, amplification procedures have error rates associated with them. Therefore, amplification introduces further modifications into the cDNA molecules. In one embodiment, during amplification with the adapter-specific primers, a detectable label, e.g., a fluorophore, is added to single strand cDNA molecules. Amplification therefore also serves to create DNA complexes that do not occur in nature, at least because (i) cDNA does not exist in vivo, (i) adapter sequences are added to the ends of cDNA molecules to make DNA sequences that do not exist in vivo, (ii) the error rate associated with amplification further creates DNA sequences that do not exist in vivo, (iii) the disparate structure of the cDNA molecules as compared to what exists in nature, and (iv) the chemical addition of a detectable label to the cDNA molecules.

[0184] In some embodiments, the expression of a biomarker of interest (e.g., one or a plurality of biomarkers from Tables 2, 4, 6, 8, 10, 12 and / or 14 ) is detected at the nucleic acid level via detection of non-natural cDNA molecules. Types of Cancer

[0185] Further to any of the embodiments provided herein, the sample obtained from a subject subjected any of the methods provided herein can be a tumor sample. The tumor sample can be a cancerous tumor. The cancer can include, but is not limited to, carcinoma, lymphoma, blastoma (including medulloblastoma and retinoblastoma), sarcoma (including liposarcoma and synovial cell sarcoma), neuroendocrine tumors (including carcinoid tumors, gastrinoma, and islet cell cancer), mesothelioma, schwannoma (including acoustic neuroma), meningioma, adenocarcinoma, melanoma, and leukemia or lymphoid malignancies. Examples of a cancer also include, but are not limited to, a lung cancer (e.g., a non-small cell lung cancer (NSCLC) such as lung adenocarcinoma (LUAD) or lung squamous cell carcinoma (LUSC)), a kidney cancer (e.g.,Attorney Docket No. GNCN-024 / 03WO 320289-2158 a kidney urothelial carcinoma or RCC), a bladder cancer (e.g., a bladder urothelial (transitional cell) carcinoma (e.g., locally advanced or metastatic urothelial cancer, including 1L or 2L+ locally advanced or metastatic urothelial carcinoma), a muscle invasive bladder cancer (MIBC), a breast cancer, a colorectal cancer (e.g., a colon adenocarcinoma), an ovarian cancer, a pancreatic cancer (e.g., pancreatic adenocarcinoma or PAAD), a gastric carcinoma, an esophageal cancer, a mesothelioma, a melanoma (e.g., a skin melanoma), a head and neck cancer (e.g., a head and neck squamous cell carcinoma (HNSCC)), a thyroid cancer, a sarcoma (e.g., a soft-tissue sarcoma, a fibrosarcoma, a myxosarcoma, a liposarcoma, an osteogenic sarcoma, an osteosarcoma, a chondrosarcoma, an angiosarcoma, an endotheliosarcoma, a lymphangiosarcoma, a lymphangioendotheliosarcoma, a leiomyosarcoma, or a rhabdomyosarcoma), a prostate cancer, a glioblastoma, a cervical cancer, a thymic carcinoma, a leukemia (e.g., an acute lymphocytic leukemia (ALL), an acute myelocytic leukemia (AML), a chronic myelocytic leukemia (CML), a chronic eosinophilic leukemia, or a chronic lymphocytic leukemia (CLL)), a lymphoma (e.g., a Hodgkin lymphoma or a non-Hodgkin lymphoma (NHL)), a myeloma (e.g., a multiple myeloma (MM)), a mycosis fungoides, a Merkel cell cancer, a hematologic malignancy, a cancer of hematological tissues, a B cell cancer, a bronchus cancer, a stomach cancer, a brain or central nervous system cancer, a peripheral nervous system cancer, a uterine or endometrial cancer, a cancer of the oral cavity or pharynx, a liver cancer, a testicular cancer, a biliary tract cancer, a small bowel or appendix cancer, a salivary gland cancer, an adrenal gland cancer, an adenocarcinoma, an inflammatory myofibroblastic tumor, a gastrointestinal stromal tumor (GIST), a colon cancer, a myelodysplastic syndrome (MDS), a myeloproliferative disorder (MPD), a polycythemia Vera, a chordoma, a synovioma, an Ewing’s tumor, a squamous cell carcinoma, a basal cell carcinoma, an adenocarcinoma, a sweat gland carcinoma, a sebaceous gland carcinoma, a papillary carcinoma, a papillary adenocarcinoma, a medullary carcinoma, a bronchogenic carcinoma, a renal cell carcinoma, a hepatoma, a bile duct carcinoma, a choriocarcinoma, a seminoma, an embryonal carcinoma, a Wilms' tumor, a bladder carcinoma, an epithelial carcinoma, a glioma, an astrocytoma, a medulloblastoma, a craniopharyngioma, an ependymoma, a pinealoma, a hemangioblastoma, an acoustic neuroma, an oligodendroglioma, a meningioma, a neuroblastoma, a retinoblastoma, a follicular lymphoma, a diffuse large B-cell lymphoma, a mantle cell lymphoma, a hepatocellular carcinoma, a thyroid cancer, a small cell cancer, an essential thrombocythemia, an agnogenic myeloid metaplasia, a hypereosinophilic syndrome, aAttorney Docket No. GNCN-024 / 03WO 320289-2158 systemic mastocytosis, a familiar hypereosinophilia, a neuroendocrine cancer, or a carcinoid tumor.

[0186] In some cases, the cancer that the subject from which a sample is obtained is suffering or suspected of suffering from is selected from a cervical kidney renal papillary cell carcinoma (KIRP); breast invasive carcinoma (BRCA); thyroid cancer (THCA); bladder carcinoma (BLCA); prostate adenocarcinoma (PRAD); kidney chromophobe (KICH); cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC); kidney renal clear cell carcinoma (KIRC); liver hepatocellular carcinoma (LIHC); low grade glioma (LGG); sarcoma (SARC); lung adenocarcinoma (LUAD); colon adenocarcinoma (COAD); head-neck squamous cell carcinomaesophageal cancer, a mesothelioma, a melanoma, a head and neck cancer, a thyroid cancer, a sarcoma, a prostate cancer, a glioblastoma, a cervical cancer, a thymic carcinoma, a leukemia, a lymphoma, a myeloma, a mycosis fungoides, a merkel cell cancer, an endometrial cancer. In some cases, the cancer is LUAD, LGG, LIHC, KIRC, KICH, MESO, ACC or KIRP. In some cases, cancer can be selected from the group consisting of ACC, BLCA, BRCA, CESC, CHOL, COAD, DLBC, GBM, HNSC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, PAAD, PCPG, PRAD, READ, SARC, SKCM, TGCT, THCA, THYM, UCEC, UCS and UVM. In embodiment, the cancer that the subject from which a sample is obtained is suffering or suspected of suffering from is lung cancer. In one embodiment, the lung cancer is LUAD. Statistical Methods

[0187] Various statistical methods can be used to aid in the comparison of the biomarker levels obtained from the patient and reference biomarker levels, for example, from at least one sample training set.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0188] In one embodiment, a supervised pattern recognition method is employed. Examples of supervised pattern recognition methods can include, but are not limited to, the nearest centroid methods (Dabney (2005) Bioinformatics 21(22):4148-4154 and Tibshirani et al. (2002) Proc. Natl. Acad. Sci. USA 99(10):6576-6572); soft independent modeling of class analysis (SIMCA) (see, for example, Wold, 1976); partial least squares analysis (PLS) (see, for example, Wold, 1966; Joreskog, 1982; Frank, 1984; Bro, R., 1997); linear discriminant analysis (LDA) (see, for example, Nillson, 1965); K-nearest neighbor analysis (KNN) (sec, for example, Brown et al., 1996); artificial neural networks (ANN) (see, for example, Wasserman, 1989; Anker et al., 1992; Hare, 1994); probabilistic neural networks (PNNs) (see, for example, Parzen, 1962; Bishop, 1995; Speckt, 1990; Broomhead et al., 1988; Patterson, 1996); rule induction (RI) (see, for example, Quinlan, 1986); and, Bayesian methods (see, for example, Bretthorst, 1990a, 1990b, 1988). In one embodiment, the classifier for identifying tumor subtypes based on gene expression data is the centroid based method described in Mullins et al. (2007) Clin Chem.53(7):1273-9, each of which is herein incorporated by reference in its entirety. In another embodiment, the classifier for identifying an FAS based on gene expression data is used in a nearest centroid based method as described in Dabney (2005) Bioinformatics 21(22):4148-4154, which is incorporated herein by reference in its entirety. The nearest centroid based method can be performed using CLaNC software as described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics.2006;22: 122–123 or equivalents or derivatives thereof.

[0189] In other embodiments, an unsupervised training approach is employed, and therefore, no training set is used.

[0190] In one embodiment, a rank-based classifier such as the Top Scoring Pair (TSP; Leek, 2009) and kTSP (Afsari et al., 2014) is employed. Rank-based classifiers such as the Top Scoring Pair (TSP; Leek, 2009) and kTSP (Afsari et al., 2014) depend only on the relative ranks of the expression of genes within a sample, allowing such classifiers to be robust against platform- specific effects and study-to-study variations due to data normalization and preprocessing (Patil et al., 2015)

[0191] The kTSP approach can select k pairs of genes A and B such that gene A expression>gene B expression implies sample membership to class 1 (e.g., FAS (+)), otherwise implying membership to class 2 (e.g., (FAS (-)). The default decision rule in Afsari et al., 2015 following feature selection weights each TSP equally in their class prediction ("voting"), despiteAttorney Docket No. GNCN-024 / 03WO 320289-2158 the fact that some TSPs may better discriminate between classes than others. The kTSP approach of Afsari et al., 2015 can be utilized to generate a rank-based classifier for use in the methods described herein by implementing a custom decision rule that inputs the selected k gene pairs into a penalized logistic regression classifier to estimate the relative contribution each of the k selected TSPs in predicting class membership (defined here as FAS (+) versus otherwise), similar to (Shi et al., 2011). In fitting the model, class membership can be the binary outcome variable, and each covariate can correspond to a TSP, consisting of a binary integer vector which can take on the value of 1 for a sample if gene A>gene B in expression for that TSP, and 0 otherwise for each sample.

[0192] Referring to sample training sets for supervised learning approaches again, in some embodiments, a sample training set(s) can include expression data of a plurality or all of the classifier biomarkers (e.g., all the classifier biomarkers of Tables 2, 4, 6, 8, 10, 12 and / or 14 ) from a sample (e.g., tumor sample) obtained from subject suffering from cancer.

[0193] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 2. In some cases, the plurality of classifier biomarkers for use in the methods for determining an EGFR-AS or EGFR-PRS consists of only the classifier biomarkers found in Table 2.

[0194] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 or 28 classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 4. In some cases, the plurality of classifier biomarkers for use in the methods for determining a MET-AS or MET-PRS consists of only the classifier biomarkers found in Table 4.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0195] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 6. In some cases, the plurality of classifier biomarkers for use in the methods for determining a KRAS-AS or KRAS- PRS consists of only the classifier biomarkers found in Table 6.

[0196] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199 or 200 classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 8. In some cases, the plurality of classifier biomarkers for use in the methods for determining a BRAF-AS or BRAF-PRS consists of only the classifier biomarkers found in Table 8.

[0197] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104,Attorney Docket No. GNCN-024 / 03WO 320289-2158 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223 or 224 classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 10. In some cases, the plurality of classifier biomarkers for use in the methods for determining an ERBB2-AS or ERBB2-PRS consists of only the classifier biomarkers found in Table 10.

[0198] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 12. In some cases, the plurality of classifier biomarkers for use in the methods for determining an ERBB3-AS or ERBB3-PRS consists of only the classifier biomarkers found in Table 12.

[0199] In some cases, the plurality of biomarker consists of or comprises exactly, at most or at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 or 36 classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers consists of or comprises exactly, at most or at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers found in Table 14. In some cases, the plurality of classifier biomarkers for use in theAttorney Docket No. GNCN-024 / 03WO 320289-2158 methods for determining an PIK3CA-AS or PIK3CA-PRS consists of only the classifier biomarkers found in Table 14.

[0200] In some embodiments, the sample training set(s) are normalized to remove sample-to- sample variation.

[0201] In some embodiments, comparing can include applying a statistical algorithm, such as, for example, any suitable multivariate statistical analysis model, which can be parametric or non-parametric. In some embodiments, applying the statistical algorithm can include determining a correlation between the expression data obtained from the human lung tissue sample and the expression data from the adenocarcinoma training set(s). In some embodiments, cross-validation is performed, such as (for example), leave-one-out cross-validation (LOOCV). In some embodiments, integrative correlation is performed. In some embodiments, a Spearman correlation is performed. In some embodiments, a centroid based method is employed for the statistical algorithm. The centroids can be constructed using any method known in the art for generating centroids such as, for example, those found in Mullins et al. (2007) Clin Chem. 53(7):1273-9 or the nearest centroid method found in Dabney (2005) Bioinformatics 21(22):4148-4154, which is herein incorporated by reference in its entirety. In one embodiment, a correlation analysis is performed on the expression data obtained from the sample obtained from a subject suffering or suspected of suffering from a cancer and the centroid(s) constructed on the expression data from the training set(s). The correlation analysis can be a Spearman correlation or a Pearson correlation. In one embodiment, a distance measure analysis (e.g., Euclidean distance) is performed on the expression data obtained from the sample and the centroid(s) constructed on the expression data from the training set(s).

[0202] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing an EGFR-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive EGFR-AS (in other words to possess one or more known EGFR mutations and / or fusions) or negative EGFR-AS (in other words, free of known EGFR mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 2) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using anyAttorney Docket No. GNCN-024 / 03WO 320289-2158 of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining an EGFR-AS, the EGFR-AS (+) or EGFR-AS (-) centroids can be the centroids found in Table 2.

[0203] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing a MET-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive MET-AS (in other words to possess one or more known MET mutations and / or fusions) or negative MET-AS (in other words, free of known MET mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 4) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining a MET-AS, the MET-AS (+) or MET-AS (-) centroids can be the centroids found in Table 4.

[0204] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing a KRAS-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive KRAS-AS (in other words to possess one or more known KRAS mutations and / or fusions) or negative KRAS-AS (in other words, free of known KRAS mutations and / or fusions). In one embodiment, the gene expressionAttorney Docket No. GNCN-024 / 03WO 320289-2158 levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 6) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining a KRAS-AS, the KRAS-AS (+) or KRAS-AS (-) centroids can be the centroids found in Table 6.

[0205] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing a BRAF-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive BRAF-AS (in other words to possess one or more known BRAF mutations and / or fusions) or negative BRAF-AS (in other words, free of known BRAF mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 8) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining a BRAF-AS, the BRAF-AS (+) or BRAF-AS (-) centroids can be the centroids found in Table 8.

[0206] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing an ERBB2-AS, a reference sample or reference gene expressionAttorney Docket No. GNCN-024 / 03WO 320289-2158 data is obtained or derived from an individual known to have a positive ERBB2-AS (in other words to possess one or more known ERBB2 mutations and / or fusions) or negative ERBB2-AS (in other words, free of known ERBB2 mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 10) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining an ERBB2-AS, the ERBB2-AS (+) or ERBB2-AS (-) centroids can be the centroids found in Table 10.

[0207] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing an ERBB3-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive ERBB3-AS (in other words to possess one or more known ERBB3 mutations and / or fusions) or negative ERBB3-AS (in other words, free of known ERBB3 mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 12) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining an ERBB3-AS, the ERBB3-AS (+) or ERBB3-AS (-) centroids can be the centroids found in Table 12.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0208] Results of the gene expression performed on a sample from a subject (test sample) may be compared to a biological sample(s) or data derived from a reference biological sample(s). In some embodiments for assessing a PIK3CA-AS, a reference sample or reference gene expression data is obtained or derived from an individual known to have a positive PIK3CA-AS (in other words to possess one or more known PIK3CA mutations and / or fusions) or negative PIK3CA-AS (in other words, free of known PIK3CA mutations and / or fusions). In one embodiment, the gene expression levels or profile for the at least one or plurality of classifier biomarker provided herein (e.g., Table 14) measured or detected in the test sample may be compared to centroids constructed from the gene expression performed on the reference sample. The centroids can be constructed using any of the methods provided herein such as, for example, using the ClaNC software described in Dabney AR. ClaNC: Point-and-click software for classifying microarrays to nearest centroids. Bioinformatics. 2006;22: 122–123 or equivalents or derivatives related thereto. Classification or determination of the subtype of the test sample can then be ascertained by determining the nearest centroid from the reference or normal sample to which the expression levels or profile from said test sample is nearest based on a distance measure or correlation. The distance measure can be a Euclidean distance. In embodiments related to determining a PIK3CA- AS, the PIK3CA-AS (+) or PIK3CA-AS (-) centroids can be the centroids found in Table 14.

[0209] The reference sample may be assayed at the same time, or at a different time from the test sample. Alternatively, the biomarker level information from a reference sample may be stored in a database or other means for access at a later date.

[0210] The biomarker level results of an assay on the test sample may be compared to the results of the same assay on a reference sample. In some cases, the results of the assay on the reference sample are from a database, or a reference value(s). In some cases, the results of the assay on the reference sample are a known or generally accepted value or range of values by those skilled in the art. In some cases, the comparison is qualitative. In other cases, the comparison is quantitative. In some cases, qualitative or quantitative comparisons may involve but are not limited to one or more of the following: comparing fluorescence values, spot intensities, absorbance values, chemiluminescent signals, histograms, critical threshold values, statistical significance values, expression levels of the genesdescribed herein, mRNA copy numbers.

[0211] In one embodiment, an odds ratio (OR) is calculated for each biomarker or biomarker pair expression level measurement. Here, the OR is a measure of association betweenAttorney Docket No. GNCN-024 / 03WO 320289-2158 the measured biomarker or biomarker pair values for the patient and an outcome, e.g., FGFR3 activation signature. For example, see, J. Can. Acad. Child Adolesc. Psychiatry 2010; 19(3): 227- 229, which is incorporated by reference in its entirety for all purposes.

[0212] In one embodiment, a specified statistical confidence level may be determined in order to provide a confidence level regarding any one or a combination of the activation signatures or predictive response signatures provided herein. For example, it may be determined that a confidence level of greater than 90% may be a useful predictor of any one or a combination of the activation signatures or predictive response signatures provided herein. In other embodiments, more or less stringent confidence levels may be chosen. For example, a confidence level of about or at least about 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 97.5%, 99%, 99.5%, or 99.9% may be chosen. The confidence level provided may in some cases be related to the quality of the sample, the quality of the data, the quality of the analysis, the specific methods used, and / or the number of gene expression values (i.e., the number of genes) analyzed. The specified confidence level for providing the likelihood of response may be chosen on the basis of the expected number of false positives or false negatives. Methods for choosing parameters for achieving a specified confidence level or for identifying markers with diagnostic power include but are not limited to Receiver Operating Characteristic (ROC) curve analysis, binormal ROC, principal component analysis, odds ratio analysis, partial least squares analysis, singular value decomposition, least absolute shrinkage and selection operator analysis, least angle regression, and the threshold gradient directed regularization method.

[0213] Determining any one or a combination of the activation signatures or predictive response signatures provided herein in some cases can be improved through the application of algorithms designed to normalize and or improve the reliability of the gene expression data. In some embodiments of the present invention, the data analysis utilizes a computer or other device, machine or apparatus for application of the various algorithms described herein due to the large number of individual data points that are processed. A “machine learning algorithm” refers to a computational-based prediction methodology, also known to persons skilled in the art as a “classifier,” employed for characterizing a gene expression profile or profiles, e.g., to determine any one or a combination of the activation signatures or predictive response signatures provided herein. The biomarker levels, determined by, e.g., microarray-based hybridization assays, sequencing assays, NanoString assays, etc., are in one embodiment subjected to the algorithm inAttorney Docket No. GNCN-024 / 03WO 320289-2158 order to classify the profile. In embodiments related to assessing any one or a combination of the activation signatures or predictive response signatures provided herein, supervised learning generally involves “training” a classifier to recognize the distinctions among an activation signature or predictive response signature positive or non-target gene activation signature or predictive response signature, and then “testing” the accuracy of the classifier on an independent test set. Therefore, for new, unknown samples the classifier can be used to predict, for example, the class (e.g., AS or PRS (+) vs. AS or PRS (-)) in which the samples belong.

[0214] In some embodiments, a robust multi-array average (RMA) method may be used to normalize raw data. The RMA method begins by computing background-corrected intensities for each matched cell on a number of microarrays. In one embodiment, the background corrected values are restricted to positive values as described by Irizarry et al. (2003). Biostatistics April 4 (2): 249-64, incorporated by reference in its entirety for all purposes. After background correction, the base-2 logarithm of each background corrected matched-cell intensity is then obtained. The background corrected, log-transformed, matched intensity on each microarray is then normalized using the quantile normalization method in which for each input array and each probe value, the array percentile probe value is replaced with the average of all array percentile points, this method is more completely described by Bolstad et al. Bioinformatics 2003, incorporated by reference in its entirety. Following quantile normalization, the normalized data may then be fit to a linear model to obtain an intensity measure for each probe on each microarray. Tukey’s median polish algorithm (Tukey, J. W., Exploratory Data Analysis.1977, incorporated by reference in its entirety for all purposes) may then be used to determine the log-scale intensity level for the normalized probe set data.

[0215] Various other software programs may be implemented. In certain methods, feature selection and model estimation may be performed by logistic regression with lasso penalty using glmnet (Friedman et al. (2010). Journal of statistical software 33(1): 1-22, incorporated by reference in its entirety). Raw reads may be aligned using TopHat (Trapnell et al. (2009). Bioinformatics 25(9): 1105-11, incorporated by reference in its entirety). In methods, top features (N ranging from 10 to 200) are used to train a linear support vector machine (SVM) (Suykens JAK, Vandewalle J. Least Squares Support Vector Machine Classifiers. Neural Processing Letters 1999; 9(3): 293-300, incorporated by reference in its entirety) using the e1071 library (Meyer D. Support vector machines: the interface to libsvm in package e1071. 2014, incorporated byAttorney Docket No. GNCN-024 / 03WO 320289-2158 reference in its entirety). Confidence intervals, in one embodiment, are computed using the pROC package (Robin X, Turck N, Hainard A, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC bioinformatics 2011; 12: 77, incorporated by reference in its entirety).

[0216] In addition, data may be filtered to remove data that may be considered suspect. In one embodiment, data derived from microarray probes that have fewer than about 4, 5, 6, 7 or 8 guanosine + cytosine nucleotides may be considered to be unreliable due to their aberrant hybridization propensity or secondary structure issues. Similarly, data deriving from microarray probes that have more than about 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, or 22 guanosine + cytosine nucleotides may in one embodiment be considered unreliable due to their aberrant hybridization propensity or secondary structure issues.

[0217] In some embodiments of the present invention, data from probe-sets may be excluded from analysis if they are not identified at a detectable level (above background).

[0218] In some embodiments of the present disclosure, probe-sets that exhibit no, or low variance may be excluded from further analysis. Low-variance probe-sets are excluded from the analysis via a Chi-Square test. In one embodiment, a probe-set is considered to be low-variance if its transformed variance is to the left of the 99 percent confidence interval of the Chi-Squared distribution with (N-l) degrees of freedom. (N-l)*Probe-set Variance / (Gene Probe-set Variance). Chi-Sq(N-l) where N is the number of input CEL files, (N-l) is the degrees of freedom for the Chi- Squared distribution, and the “probe-set variance for the gene” is the average of probe-set variances across the gene. In some embodiments of the present invention, probe-sets for a given mRNA or group of mRNAs may be excluded from further analysis if they contain less than a minimum number of probes that pass through the previously described filter steps for GC content, reliability, variance and the like. For example, in some embodiments, probe-sets for a given gene or transcript cluster may be excluded from further analysis if they contain less than about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or less than about 20 probes.

[0219] Methods of classifier biomarker level data analysis in one embodiment, further include the use of a feature selection algorithm as provided herein. In some embodiments of the present invention, feature selection is provided by use of the LIMMA software package (Smyth, G. K. (2005). Limma: linear models for microarray data. In: Bioinformatics and Computational Biology Solutions using R and Bioconductor, R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, W. HuberAttorney Docket No. GNCN-024 / 03WO 320289-2158 (eds.), Springer, New York, pages 397-420, incorporated by reference in its entirety for all purposes).

[0220] Methods of classifier biomarker level data analysis, in one embodiment, include the use of a pre-classifier algorithm. For example, an algorithm may use a specific molecular fingerprint to pre-classify the samples according to their composition and then apply a correction / normalization factor. This data / information may then be fed into a final classification algorithm which would incorporate that information to aid in the final diagnosis.

[0221] Methods of classifier biomarker level data analysis, in one embodiment, further include the use of a classifier algorithm as provided herein. In one embodiment of the present invention, a diagonal linear discriminant analysis, CLaNC, k-nearest neighbor algorithm, top scoring pair, k- top scoring pair, support vector machine (SVM) algorithm, linear support vector machine, random forest algorithm, or a probabilistic model-based method or a combination thereof is provided for classification of microarray data or RNA-seq data. In some embodiments, identified markers that distinguish samples (e.g., AS or PRS (+), AS or PRS(-)) are selected based on statistical significance of the difference in biomarker levels between classes of interest. In some cases, the statistical significance is adjusted by applying a Benjamin Hochberg or another correction for false discovery rate (FDR).

[0222] In some cases, the classifier algorithm may be supplemented with a meta-analysis approach such as that described by Fishel and Kaufman et al.2007 Bioinformatics 23(13): 1599- 606, incorporated by reference in its entirety for all purposes. In some cases, the classifier algorithm may be supplemented with a meta-analysis approach such as a repeatability analysis.

[0223] Methods for deriving and applying posterior probabilities to the analysis of biomarker level data are known in the art and have been described for example in Smyth, G. K.2004 Stat. Appi. Genet. Mol. Biol.3: Article 3, incorporated by reference in its entirety for all purposes. In some cases, the posterior probabilities may be used in the methods of the present invention to rank the markers provided by the classifier algorithm.

[0224] A statistical evaluation of the results of the classifier biomarker level profiling may provide a quantitative value or values indicative of one or more of the activation signatures or predictive response signatures provided herein; the likelihood of the success of a particular therapeutic intervention, e.g., FGFR inhibitor therapy, angiogenesis inhibitor therapy, chemotherapy, immunotherapy or any combination thereof. In one embodiment, the data isAttorney Docket No. GNCN-024 / 03WO 320289-2158 presented directly to the physician in its most useful form to guide patient care or is used to define patient populations in clinical trials or a patient population for a given medication. The results of the molecular profiling can be statistically evaluated using a number of methods known to the art including, but not limited to: the students T test, the two sided T test, Pearson rank sum analysis, hidden Markov model analysis, analysis of q-q plots, principal component analysis, one way ANOVA, two way ANOVA, LIMMA and the like.

[0225] In some cases, accuracy may be determined by tracking the subject over time to determine the accuracy of the original diagnosis. In other cases, accuracy may be established in a deterministic manner or using statistical methods. For example, receiver operator characteristic (ROC) analysis may be used to determine the optimal assay parameters to achieve a specific level of accuracy, specificity, positive predictive value, negative predictive value, and / or false discovery rate.

[0226] In some cases, the results of the classifier biomarker level profiling assays, are entered into a database for access by representatives or agents of a molecular profiling business, the individual, a medical provider, or insurance provider. In some cases, assay results include sample classification, identification, or diagnosis by a representative, agent or consultant of the business, such as a medical professional. In other cases, a computer or algorithmic analysis of the data is provided automatically. In some cases, the molecular profiling business may bill the individual, insurance provider, medical provider, researcher, or government entity for one or more of the following: molecular profiling assays performed, consulting services, data analysis, reporting of results, or database access.

[0227] In some embodiments of the present invention, the results of the classifier biomarker level profiling assays are presented as a report on a computer screen or as a paper record. In some embodiments, the report may include, but is not limited to, such information as one or more of the following: the levels of classifier biomarkers (e.g., as reported by copy number or fluorescence intensity, etc.) as compared to the reference sample or reference value(s); the likelihood the subject will respond to a particular therapy and / or will be resistant or non-responsive to a particular therapy, based on the classifier biomarker values and the activation / predictive response signature or any combination of activation / predictive response signatures and proposed therapies.

[0228] In one embodiment, the results of the gene expression profiling may be classified into one or more of the following: target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3,Attorney Docket No. GNCN-024 / 03WO 320289-2158 PIK3CA) activation signature positive; possessing one or more target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) alterations or mutations; target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) activation signature negative; free of one or more target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) alterations or mutations); likely to respond to target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy; likely to respond to angiogenesis inhibitor, immunotherapy or chemotherapy; unlikely to respond to target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy; unlikely to respond to angiogenesis inhibitor, immunotherapy or chemotherapy; likely to be resistant to specific type of inhibitor therapy or a combination thereof.

[0229] In some embodiments of the present invention, results are classified using a trained algorithm. Trained algorithms of the present invention include algorithms that have been developed using a reference set of known gene expression values and / or normal samples, for example, samples from individuals diagnosed with a particular target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) mutation. In some cases, a reference set of known gene expression values are obtained from individuals who have been diagnosed with a particular target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) mutation and are also known to respond (or not respond) to a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy. In some cases, a reference set of known gene expression values are obtained from individuals who have been diagnosed without a particular target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) mutation and are also known to respond (or not respond) to a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy. In some cases, a reference set of known gene expression values are obtained from individuals who have been diagnosed with a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) mutation, and are also known to respond (or not respond) to a treatment modality other than a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy (such as, for example, chemotherapy, immunotherapy, angiogenesis inhibitors, radiotherapy, surgical intervention, etc.). In some cases, a reference set of known gene expression values are obtained from individuals who have been diagnosed without a particular target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) mutation and are also known to respond (or not respond) to a treatment modality other than a target gene (e.g., EGFR,Attorney Docket No. GNCN-024 / 03WO 320289-2158 MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy (such as, for example, chemotherapy, immunotherapy, angiogenesis inhibitors, radiotherapy, surgical intervention, etc.).

[0230] Algorithms suitable for categorization of samples include but are not limited to k- nearest neighbor algorithms, k-top scoring pairs (TSPs), top scoring pairs (TSPs), support vector machines, linear discriminant analysis, CLaNC, diagonal linear discriminant analysis, up down, naive Bayesian algorithms, neural network algorithms, hidden Markov model algorithms, genetic algorithms, or any combination thereof.

[0231] When a binary classifier is compared with actual true values (e.g., values from a biological sample), there are typically four possible outcomes. If the outcome from a prediction is p (where “p” is a positive classifier output, such as the presence of a deletion or duplication syndrome) and the actual value is also p, then it is called a true positive (TP); however, if the actual value is n then it is said to be a false positive (FP). Conversely, a true negative has occurred when both the prediction outcome and the actual value are n (where “n” is a negative classifier output, such as no deletion or duplication syndrome), and false negative is when the prediction outcome is n while the actual value is p. In one embodiment, consider a test that seeks to determine whether a person is likely or unlikely to respond to a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) inhibitor therapy. A false positive in this case occurs when the person tests positive, but actually does respond. A false negative, on the other hand, occurs when the person tests negative, suggesting they are unlikely to respond, when they actually are likely to respond. The same holds true for classifying any one of or a combination of the activation / predictive response signature(s) provided herein.

[0232] The positive predictive value (PPV), or precision rate, or post-test probability of disease, is the proportion of subjects with positive test results who are correctly diagnosed as likely or unlikely to respond, or diagnosed with a positive target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3, PIK3CA) activation status, or a combination thereof. It reflects the probability that a positive test reflects the underlying condition being tested for. Its value does however depend on the prevalence of the disease, which may vary. In one example the following characteristics are provided: FP (false positive); TN (true negative); TP (true positive); FN (falsenegative). False positive rate ( )=FP / (FP+TN)-specificity; False negative rate ( )=FN / (TP+FN)-sensitivity; Power = sensitivity = 1- ; Likelihood-ratio positive=sensitivity / (l-specificity);Attorney Docket No. GNCN-024 / 03WO 320289-2158 Likelihood-ratio negative = (1 -sensitivity) / specificity. The negative predictive value (NPV) is the proportion of subjects with negative test results who are correctly diagnosed.

[0233] In some embodiments, the results of the classifier biomarker level analysis of the subject methods provide a statistical confidence level that a given diagnosis is correct. In some embodiments, such statistical confidence level is at least about, or more than about 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% 99.5%, or more.

[0234] In some embodiments, the method further includes classifying the sample as being AS (+) or (-) or PRS (+) or (-) based on the comparison of classifier biomarker levels in the sample and reference biomarker levels, for example present in at least one training set. In some embodiments, the sample is classified as being AS (+) or (-) or PRS (+) or (-) if the results of the comparison meet one or more criterion such as, for example, a minimum percent agreement, a value of a statistic calculated based on the percentage agreement such as (for example) a kappa statistic, a minimum correlation (e.g., Pearson’s correlation) and / or the like.

[0235] It is intended that the methods described herein can be performed by software (stored in memory and / or executed on hardware), hardware, or a combination thereof. Hardware modules may include, for example, a general-purpose processor, a field programmable gate array (FPGA), and / or an application specific integrated circuit (ASIC). Software modules (executed on hardware) can be expressed in a variety of software languages (e.g., computer code), including Unix utilities, C, C++, Java™, Ruby, SQL, SAS®, the R programming language / software environment, Visual Basic™, and other object-oriented, procedural, or other programming language and development tools. Examples of computer code include, but are not limited to, micro-code or micro- instructions, machine instructions, such as produced by a compiler, code used to produce a web service, and files containing higher-level instructions that are executed by a computer using an interpreter. Additional examples of computer code include, but are not limited to, control signals, encrypted code, and compressed code.

[0236] Some embodiments described herein relate to devices with a non-transitory computer- readable medium (also can be referred to as a non-transitory processor-readable medium or memory) having instructions or computer code thereon for performing various computer- implemented operations and / or methods disclosed herein. The computer-readable medium (or processor-readable medium) is non-transitory in the sense that it does not include transitory propagating signals per se (e.g., a propagating electromagnetic wave carrying information on aAttorney Docket No. GNCN-024 / 03WO 320289-2158 transmission medium such as space or a cable). The media and computer code (also can be referred to as code) may be those designed and constructed for the specific purpose or purposes. Examples of non-transitory computer-readable media include, but are not limited to: magnetic storage media such as hard disks, floppy disks, and magnetic tape; optical storage media such as Compact Disc / Digital Video Discs (CD / DVDs), Compact Disc-Read Only Memories (CD-ROMs), and holographic devices; magneto-optical storage media such as optical disks; carrier wave signal processing modules; and hardware devices that are specially configured to store and execute program code, such as Application-Specific Integrated Circuits (ASICs), Programmable Logic Devices (PLDs), Read-Only Memory (ROM) and Random-Access Memory (RAM) devices. Other embodiments described herein relate to a computer program product, which can include, for example, the instructions and / or computer code discussed herein.

[0237] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 2 are capable of classifying an EGFR alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 2 are capable of classifying an EGFR alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0238] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 or 28 classifier biomarkers found in Table 4 are capable of classifying MET alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%,Attorney Docket No. GNCN-024 / 03WO 320289-2158 about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 4 are capable of classifying MET alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0239] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 6 are capable of classifying KRAS alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 6 are capable of classifying KRAS alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0240] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135,Attorney Docket No. GNCN-024 / 03WO 320289-2158 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199 or 200 classifier biomarkers found in Table 8 are capable of classifying BRAF alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 8 are capable of classifying BRAF alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0241] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223 or 224 classifier biomarkers found in Table 10 are capable of classifying an ERBB2 alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%,Attorney Docket No. GNCN-024 / 03WO 320289-2158 about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 10 are capable of classifying an ERBB2 alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0242] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 classifier biomarkers found in Table 12 are capable of classifying an ERBB3 alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 12 are capable of classifying an ERBB3 alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0243] In some embodiments, at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 or 36 classifier biomarkers found in Table 14 are capable of classifying PIK3CA alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, aboutAttorney Docket No. GNCN-024 / 03WO 320289-2158 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between. In some embodiments, at least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% of the classifier biomarkers found in Table 14 are capable of classifying PIK3CA alteration status with a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0244] In some embodiments, any combination of biomarkers disclosed herein (e.g., in Tables 2, 4, 6, 8, 10, 12 and / or 14) can be used to obtain a predictive success of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0245] In some embodiments, any combination of biomarkers disclosed herein (e.g., in Tables 2, 4, 6, 8, 10, 12 and / or 14) can be used to obtain a sensitivity or specificity of at least about 70%, at least about 71%, at least about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, about 99%, up to 100%, and all values in between.

[0246] In one embodiment, use of an activation signature or predictive response signature provided herein for predicting or ascertaining a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any of the methods provided herein can do so with predictive success greater than conducting a conventional mutational analysis (e.g., DNA mutational analysis) of said sample for any known target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) oncogenic mutation. The predictive success of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering fromAttorney Docket No. GNCN-024 / 03WO 320289-2158 cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be at least, at most or about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% 99.5%, or more greater than the predictive success of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any conventional mutational analysis (e.g., DNA mutational analysis). The predictive success of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 99 or 100 times greater than the predictive success of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any conventional mutational analysis (e.g., DNA mutational analysis). The cancer can be any cancer known in the art and / or provided herein. Examples of conventional mutational analysis include but are not limited to whole exome sequencing (WES), whole genome sequencing, RNA-seq, RT-PCR, etc.

[0247] In one embodiment, use of the activation signatures or predictive response signatures provided herein for predicting or ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any of the methods provided herein can do so with a sensitivity greater than conducting a conventional mutational analysis (e.g., DNA mutational analysis) of said sample for any known target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) oncogenic mutation. The sensitivity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be at least, at most or about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% 99.5%, or more greater than the sensitivity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using anyAttorney Docket No. GNCN-024 / 03WO 320289-2158 conventional mutational analysis (e.g., DNA mutational analysis). The sensitivity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 99 or 100 times greater than the sensitivity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any conventional mutational analysis (e.g., DNA mutational analysis). The cancer can be any cancer known in the art and / or provided herein. Examples of conventional mutational analysis include but are not limited to whole exome sequencing (WES), whole genome sequencing, RNA-seq, RT-PCR, etc.

[0248] In one embodiment, use of the activation signatures or predictive response signatures provided herein for predicting or ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any of the methods provided herein can do so with a specificity greater than conducting a conventional mutational analysis (e.g., DNA mutational analysis) of said sample for any known target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) oncogenic mutation. The specificity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be at least, at most or about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% 99.5%, or more greater than the specificity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any conventional mutational analysis (e.g., DNA mutational analysis). The specificity of ascertaining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any one of the activation signatures or predictive response signatures provided herein in a detection or diagnostic method provided herein can be 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 99 or 100 times greater than the specificity of ascertaining the target geneAttorney Docket No. GNCN-024 / 03WO 320289-2158 (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) mutational status of a sample obtained from a patient suffering from cancer using any conventional mutational analysis (e.g., DNA mutational analysis). The cancer can be any cancer known in the art and / or provided herein. Examples of conventional mutational analysis include but are not limited to whole exome sequencing (WES), whole genome sequencing, RNA-seq, RT-PCR, etc. Prognostic Uses

[0249] In one aspect, provided herein is a method for determining a disease outcome in a subject suffering from or suspected of suffering from cancer. The cancer can be any cancer known in the art and / or provided herein. In one embodiment, the subject is suffering from or suspected of suffering from a cancer selected from KIRP, BRCA, THCA, BLCA, PRAD, KICH, CESC, KIRC, LIHC, LGG, SARC, LUAD, COAD, HNSC, UCEC, GBM, ESCA, STAD, OV or READ. The disease outcome can be a prognosis. The prognostic information that can be obtained by the methods provided herein can comprise a number of possible endpoints, which can be selected from time from surgery to distant metastases (distant recurrence-free survival), time of disease-free survival (recurrence free survival), time of progression-free survival (progression free survival) and time of overall survival. In some cases, Kaplan-Meier plots (Kaplan and Meier. J Am Stat Assoc 53: 457-481 (1958)) can be used to display time-to-event curves for any or all of these three endpoints. In some cases, a cox regression (or proportional hazards regression) can be performed in order to determine a hazard ratio for any or all of these three endpoints. In one embodiment, a cox regression (or proportional hazards regression) is used to assess the prognostic performance in terms of overall survival of an AS (+) and / or AS (-) sample as determined using the methods provided herein. The Cox Proportional Hazards analysis is a regression method for survival data that provides an estimate of the hazard ratio and its confidence interval. The Cox model is a well- recognized statistical technique for exploring the relationship between the survival of a subject and particular variables. This statistical method permits estimation of the hazard (i.e., risk) of individuals given their prognostic variables (e.g., target gene (e.g., ) activation status with or without other additional clinical factors, as described herein). The "hazard ratio" is the risk of death at any given time point for patients displaying particular prognostic variables. See generally Spruance et al., Antimicrob. Agents & Chemo.48:2787-92 (2004). The additional clinical factors can include age, sex, tumor diameter, tumor stage and smoking history. A relevant time intervalAttorney Docket No. GNCN-024 / 03WO 320289-2158 or time point can be at least 1 year, at least two years, at least three years, at least five years, or at least ten years.

[0250] In one embodiment, the method for determining a disease outcome for a subject suffering from or suspected of suffering from a cancer can comprise: (a) determining a target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) activation signature of a sample obtained from the subject, wherein the determining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) activation signature comprises determining the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) AS of the sample obtained from the subject using any of the diagnostic or detection methods provided herein on any of the target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) activation signatures (i.e., EGFR-AS; MET-AS; KRAS-AS; BRAF-AS; ERBB2-AS; ERBB3-AS; PIK3CA-AS) provided herein. Further to either of these embodiments, a positive target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) AS in the sample obtained from the subject as compared to a control sample can be indicative of a poor disease outcome for the subject. In one embodiment, a positive target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) AS can be indicative of poor overall survival as compared to a control sample such as a tumor sample with a negative target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) AS obtained from a control subject or a sample obtained from a control subject not suffering from cancer. In still another embodiment, a negative target gene (e.g., EGFR, MET, KRAS, BRAF, ERBB2, ERBB3 or PIK3CA) AS in the sample obtained from the subject as compared to a control sample can be indicative of a poor disease outcome for the subject. The expression level of any and all classifier genes can be normalized as provided herein, such as, for example, normalizing expression of the classifier genes by using expression levels from one or more reference or housekeeping genes. Therapeutic Uses EGFR Inhibitors

[0251] In one embodiment, an agent for use in any of the diagnostic and / or therapeutic methods provided herein is an agent that shows or exhibits inhibitory activity towards an EGFR. In one embodiment, the detection of a positive EGFR-AS (which can also be referred to as anAttorney Docket No. GNCN-024 / 03WO 320289-2158 EGFR-PRS) in a sample obtained from a patient using an EGFR activation signatures as provided herein (e.g., Table 2) indicates that the patient is a responder to an agent that shows or exhibits inhibitory activity towards an EGFR. The agent that shows or exhibits inhibitory activity towards an EGFR can be administered to a responder (patient with a positive EGFR-AS or EGFR-PRS) alone or in combination with an additional therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.

[0252] The agent that shows or exhibits inhibitory activity towards an EGFR can be any agent known in the art that exhibits inhibitory activity toward epidermal growth factor receptors generally or EGFR, specifically. In one embodiment, the agent is a tyrosine kinase inhibitor. The tyrosine kinase inhibitor can be any tyrosine kinase inhibitor known in the art. The tyrosine kinase inhibitor can be a selective or non-selective tyrosine kinase inhibitor. The agent can be selected from the group consisting of erlotinib (OSI-744), poziotinib (HM781-36B), osimertinib (AZD9291), AG-490 (Tyrphostin B42), afatinib (BIBW2992), gefitinib (ZD1839), lapatinib (GW- 572016), rociletinib (CO-1686), neratinib, lucitanib (E3810), dacomitinib, mobocertinib, vandetanib, canertinib (CI-1033), BDTX-189, epertinib, AEE788, CUDC-101, pelitinib, sapitinib, varlitinib, pyrotinib, TAK-285, AC480, tyrophostin AG-528. In one embodiment, the agent is an antibody or antibody-conjugate. The antibody or antibody-conjugate can be selected from cetuximab, panitumumab or necitumumab. In one embodiment, the agent is a combination of agents that exhibit inhibitory activity toward epidermal growth factor receptors generally or EGFR specifically.

[0253] In one embodiment, the detection of a negative EGFR-AS or EGFR-PRS in a sample obtained from a patient using the EGFR activation signature provided herein (e.g., Table 2) indicates that the patient is a non-responder to an agent that shows or exhibits inhibitory activity towards an EGFR. The agent that shows or exhibits inhibitory activity towards an EGFR can thusly, not be administered to a non-responder (patient with a negative EGFR-AS or EGFR-PRS). Instead, a patient determined to be a non-responder using any of the diagnostic or detection methods provided herein is administered a non-EGFR inhibitor therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.Attorney Docket No. GNCN-024 / 03WO 320289-2158

[0254] In one embodiment, the detection of a positive EGFR-AS (which can also be referred to as an EGFR-PRS) in a sample obtained from a patient using an EGFR activation signatures as provided herein (e.g., Table 2) indicates that the patient is resistant to an agent that shows or exhibits inhibitory activity towards an MEK (e.g., MEK inhibitor) and / or ERK (e.g., ERK inhibitor). MET Inhibitors

[0255] In one embodiment, an agent for use in any of the diagnostic and / or therapeutic methods provided herein is an agent that shows or exhibits inhibitory activity towards a MET. In one embodiment, the detection of a positive MET-AS (which can also be referred to as an MET- PRS) in a sample obtained from a patient using a MET activation signatures as provided herein (e.g., Table 4) indicates that the patient is a responder to an agent that shows or exhibits inhibitory activity towards a MET. The agent that shows or exhibits inhibitory activity towards a MET can be administered to a responder (patient with a positive MET-AS or MET-PRS) alone or in combination with an additional therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.

[0256] The agent that shows or exhibits inhibitory activity towards a MET can be any agent known in the art that exhibits inhibitory activity a MET proto-oncogene, receptor tyrosine kinase (MET) generally or specifically. In one embodiment, the agent is a tyrosine kinase inhibitor. The tyrosine kinase inhibitor can be any tyrosine kinase inhibitor known in the art. The tyrosine kinase inhibitor can be a selective or non-selective tyrosine kinase inhibitor. The agent can be selected from the group consisting of crizotinib, capmatinib, tepotinib, savolitinib, cabozantinib, glesatinib merestinib . In one embodiment, the agent is an antibody or antibody-conjugate. The antibody or antibody-conjugate can be selected from emibetuzumab, Rilotumumab, Ficlatuzumab, TAK-701, Onartuzumab, ARGX-111, or EM1-mAb. In one embodiment, the agent is a combination of agents that exhibit inhibitory activity toward a MET generally or specifically.

[0257] In one embodiment, the detection of a negative MET-AS or MET-PRS in a sample obtained from a patient using the MET activation signature provided herein (e.g., Table 4) indicates that the patient is a non-responder to an agent that shows or exhibits inhibitory activity towards a MET. The agent that shows or exhibits inhibitory activity towards a MET can thusly,Attorney Docket No. GNCN-024 / 03WO 320289-2158 not be administered to a non-responder (patient with a negative MET-AS or MET-PRS). Instead, a patient determined to be a non-responder using any of the diagnostic or detection methods provided herein is administered a non-MET inhibitor therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.

[0258] In one embodiment, the detection of a positive MET-AS (which can also be referred to as a MET-PRS) in a sample obtained from a patient using a MET activation signatures as provided herein (e.g., Table 4) indicates that the patient is resistant to an agent that shows or exhibits inhibitory activity towards a PI3K (e.g., a PI3K inhibitor), an AKT (e.g., an AKT inhibitor), an EGFR (e.g., EGFR inhibitor) and / or mTOR (e.g., mTOR inhibitor). KRAS Inhibitors

[0259] In one embodiment, an agent for use in any of the diagnostic and / or therapeutic methods provided herein is an agent that shows or exhibits inhibitory activity towards KRAS. In one embodiment, the detection of a positive KRAS-AS (which can also be referred to as KRAS- PRS) in a sample obtained from a patient using a KRAS activation signatures as provided herein (e.g., Table 6) indicates that the patient is a responder to an agent that shows or exhibits inhibitory activity towards Kirsten Rat Sarcoma viral oncogene homolog (KRAS), mitogen-activated protein kinase kinase 1 (MEK1), mitogen-activated protein kinase kinase 2 (MEK2), extracellular signal- regulated kinase (ERK). The agent that shows or exhibits inhibitory activity towards KRAS, MEK1, MEK2 or ERK can be administered to a responder (patient with a positive KRAS-AS or KRAS-PRS) alone or in combination with an additional therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.

[0260] The agent that shows or exhibits inhibitory activity towards KRAS can be any agent known in the art that exhibits inhibitory activity Kirsten Rat Sarcoma viral oncogene homolog (KRAS), mitogen-activated protein kinase kinase 1 (MEK1), mitogen-activated protein kinase kinase 2 (MEK2), extracellular signal-regulated kinase (ERK) generally or specifically. The agent can be a KRAS inhibitor selected from the group consisting of Adagrasib, JQ443, GDC-6036, BIAttorney Docket No. GNCN-024 / 03WO 320289-21583144, Abd-7, BI02852, BI-3406, tipifarnib, SHP099, JAB-3068, RMC-4550 and TNO155. The agent can be a MEK inhibitor selected from the group consisting of binimetinib, cobimetinib, selumetinib, trametinib, CI-1040, TAK-733 and pimasertib. The agent can be a ERK inhibitor is selected from the group consisting of PD0325901, ASN007, ulixertinib (BVD-523), CC-9003,ERK5-IN-2, XMD8-92 and DEL-22379. In one embodiment, the agent is a combination of agents that exhibit inhibitory activity toward KRAS, MEK1, MEK2 or ERK generally or specifically.

[0261] In one embodiment, the detection of a negative KRAS-AS or KRAS-PRS in a sample obtained from a patient using the KRAS activation signature provided herein (e.g., Table 6) indicates that the patient is a non-responder to an agent that shows or exhibits inhibitory activity towards KRAS, MEK1, MEK2 or ERK. The agent that shows or exhibits inhibitory activity towards KRAS, MEK1, MEK2 or ERK can thusly, not be administered to a non-responder (patient with a negative KRAS-AS or KRAS-PRS). Instead, a patient determined to be a non-responder using any of the diagnostic or detection methods provided herein is administered a non-KRAS, non-MEK1, non-MEK2 or non-ERK inhibitor therapy or therapies. The additional therapy or therapies can be selected from the group consisting of a chemotherapeutic agent, an angiogenesis inhibitor, immunotherapy, radiotherapy, surgical intervention and any combination thereof.

[0262] In one embodiment, the detection of a positive KRAS-AS (which can also be referred to as a KRAS-PRS) in a sample obtained from a patient using a KRAS activation signatures as provided herein (e.g.,...

Claims

Attorney Docket No. GNCN-024 / 03WO 320289-2158 What is claimed:

1. A method of determining whether a patient suffering from cancer is likely to respond to treatment with a first type of therapy and / or whether the patient suffering from cancer is unlikely to respond to treatment with a second type of therapy, the method comprising, determining a predictive response signature of a sample obtained from a patient suffering from cancer; and based on the predictive response signature, assessing whether the patient is likely to respond to treatment with the first type of therapy and / or the patient is not likely to respond to treatment with the second type of therapy, wherein a positive predictive response signature predicts that the patient is likely to respond to the treatment with the first type of therapy and / or the patient is not likely to respond to treatment a second type of therapy.

2. A method for selecting a patient suffering from cancer for treatment with a first type of therapy and / or not selecting the patient suffering from cancer for treatment with a second type of therapy, the method comprising, determining a predictive response signature of a sample obtained from a patient suffering from cancer; and selecting the patient for treatment with the first type of therapy and / or not selecting the patient for treatment with the second type of therapy if the predictive response signature is positive.

3. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining the expression levels of a plurality of classifier biomarkers selected from Table 2 and the predictive response signature is an epidermal growth factor receptor (EGFR) predictive response signature.

4. The method of claim 3, wherein the positive EGFR predictive response signature indicates that the patient possesses one or more genetic variants in the epidermal growth factor receptor (egfr) gene.

5. The method of claim 3, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 2 to an expression level of the plurality of classifier biomarkers selected from Table 2 in at least one sample training set, wherein the at least one sample training set is from a reference EGFR mutation-containing cancer sample, or is from a reference EGFR mutation-free cancer sample; and classifying the sample as having a positive EGFR predictive response signature based on the results of the comparing step.Attorney Docket No. GNCN-024 / 03WO 320289-2158 6. The method of claim 5, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 2 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 2 from the at least one training set; and classifying the sample as possessing a positive EGFR predictive response signature based on the results of the statistical algorithm.

7. The method of claim 5, wherein the at least one training set is from a reference EGFR mutation-containing cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample.

8. The method of claim 5, wherein the at least one training set is from a reference EGFR mutation-containing cancer sample and from a reference EGFR mutation-free cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample.

9. The method of claim 3, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2.

10. The method of claim 3, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2.

11. The method of claim 3, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2.

12. The method of claim 3, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.Attorney Docket No. GNCN-024 / 03WO 320289-2158 13. The method of claim 3, wherein the first type of therapy shows inhibitory activity toward EGFR.

14. The method of claim 13, wherein the EGFR inhibitor is a tyrosine kinase inhibitor.

15. The method of claim 14, wherein the EGFR inhibitor is a selective tyrosine kinase inhibitor.

16. The method of claim 14, wherein the EGFR inhibitor is a non-selective tyrosine kinase inhibitor.

17. The method of claim 13, wherein the EGFR inhibitor is selected from the group consisting of erlotinib (OSI-744), poziotinib (HM781-36B), osimertinib (AZD9291), AG-490 (Tyrphostin B42), afatinib (BIBW2992), gefitinib (ZD1839), lapatinib (GW-572016), rociletinib (CO-1686), neratinib, lucitanib (E3810), dacomitinib, mobocertinib, vandetanib, canertinib (CI- 1033), BDTX-189, epertinib, AEE788, CUDC-101, pelitinib, sapitinib, varlitinib, pyrotinib, TAK- 285, AC480, tyrophostin AG-528 and any combination thereof.

18. The method of claim 13, wherein the EGFR inhibitor is canertinib (CI-1033).

19. The method of claim 13, wherein the EGFR inhibitor is an antibody or antibody- conjugate.

20. The method of claim 19, wherein the EGFR inhibitor is cetuximab, panitumumab or necitumumab.

21. The method of claim 3, wherein the second type of therapy is a MEK inhibitor or an ERK inhibitor.

22. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 4 and the predictive response signature is a MET proto-oncogene, receptor tyrosine kinase (MET) predictive response signature.

23. The method of claim 22, wherein the positive MET predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the MET proto-oncogene, receptor tyrosine kinase (met) gene.

24. The method of claim 22, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 4 to an expression level of the plurality of classifier biomarkers selected from Table 4 in at least one sample training set, wherein the at leastAttorney Docket No. GNCN-024 / 03WO 320289-2158 one sample training set is from a reference MET mutation-containing or amplification-containing cancer sample, or is from a reference MET mutation-free cancer sample; and classifying the sample as having a positive MET predictive response signature based on the results of the comparing step.

25. The method of claim 24, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 4 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 4 from the at least one training set; and classifying the sample as possessing a positive MET predictive response signature based on the results of the statistical algorithm.

26. The method of claim 24, wherein the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation- containing or amplification-containing cancer sample.

27. The method of claim 24, wherein the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and from a reference MET mutation-free cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation-containing or amplification-containing cancer sample.

28. The method of claim 22, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4.

29. The method of claim 22, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4.Attorney Docket No. GNCN-024 / 03WO 320289-2158 30. The method of claim 22, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4.

31. The method of claim 22, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

32. The method of claim 22, wherein the first type of therapy shows inhibitory activity toward MET.

33. The method of claim 32, wherein the MET inhibitor is a tyrosine kinase inhibitor.

34. The method of claim 33, wherein the MET inhibitor is a selective tyrosine kinase inhibitor.

35. The method of claim 33, wherein the MET inhibitor is a non-selective tyrosine kinase inhibitor.

36. The method of claim 32, wherein the MET inhibitor is selected from the group consisting of crizotinib, capmatinib, tepotinib, savolitinib, cabozantinib, glesatinib merestinib and any combination thereof.

37. The method of claim 29, wherein the MET inhibitor is an antibody or antibody- conjugate.

38. The method of claim 35, wherein the MET inhibitor is emibetuzumab, Rilotumumab, Ficlatuzumab, TAK-701, Onartuzumab, ARGX-111, or EM1-mAb.

39. The method of claim 22, wherein the second type of therapy is an mTOR inhibitor, a PI3K inhibitor, an EGFR inhibitor or an AKT inhibitor.

40. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 6 and the predictive response signature is a Kirsten Rat Sarcoma viral oncogene homolog (KRAS) predictive response signature.

41. The method of claim 40, wherein the positive KRAS predictive response signature indicates that the patient possesses one or more genetic variants in the Kirsten Rat Sarcoma viral oncogene homolog (kras) gene.

42. The method of claim 40, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 6 to an expression level of the plurality ofAttorney Docket No. GNCN-024 / 03WO 320289-2158 classifier biomarkers selected from Table 6 in at least one sample training set, wherein the at least one sample training set is from a reference KRAS mutation-containing cancer sample, or is from a reference KRAS mutation-free cancer sample; and classifying the sample as having a positive KRAS predictive response signature based on the results of the comparing step.

43. The method of claim 42, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 6 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 6 from the at least one training set; and classifying the sample as possessing a positive KRAS predictive response signature based on the results of the statistical algorithm.

44. The method of claim 42, wherein the at least one training set is from a reference KRAS mutation-containing cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample.

45. The method of claim 42, wherein the at least one training set is from a reference KRAS mutation-containing cancer sample and from a reference KRAS mutation-free cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample.

46. The method of claim 40, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6.

47. The method of claim 40, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6.

48. The method of claim 40, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 6.Attorney Docket No. GNCN-024 / 03WO 320289-2158 49. The method of claim 40, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

50. The method of claim 40, wherein the first type of therapy shows inhibitory activity toward KRAS, MEK1, MEK2 or extracellular signal-regulated kinase (ERK).

51. The method of claim 50, wherein the KRAS inhibitor is selected from the group consisting of Adagrasib, JQ443, GDC-6036, BI 1823911, JNJ-74699157, MK-1084, SCH-53239, AMG510, MRTX849, ARS-3248, compound 3144, Abd-7, BI02852, BI-3406, tipifarnib, SHP099, JAB-3068, RMC-4550, TNO155 and any combination thereof.

52. The method of claim 50, wherein the MEK inhibitor is selected from the group consisting of binimetinib, cobimetinib, selumetinib, trametinib, CI-1040, TAK-733, pimasertib, and any combination thereof.

53. The method of claim 50, wherein the ERK inhibitor is selected from the group consisting of PD0325901, ASN007, ulixertinib (BVD-523), CC-9003, ravoxertinib (GDC-0994), MK-8353, KO-947, LTT462, temuterib (LY3214996), magnolin, SCH772984, pluripotin, FR 180204, resveratrol, VX-11e, AZD0364. MRTX-1257, ERK5-IN-1, ERK5-IN-2, XMD8-92, DEL-22379 and any combination thereof.

54. The method of claim 40, wherein the second type of therapy is an ERBB2 inhibitor, ERBB4 inhibitor or EGFR inhibitor.

55. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 8 and the predictive response signature is a B-Raf protooncogene (BRAF) predictive response signature.

56. The method of claim 55, wherein the positive BRAF predictive response signature indicates that the patient possesses one or more genetic variants in the B-Raf protooncogene (BRAF) gene.

57. The method of claim 55, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 8 to an expression level of the plurality of classifier biomarkers selected from Table 8 in at least one sample training set, wherein the at least one sample training set is from a reference BRAF mutation-containing cancer sample, or is fromAttorney Docket No. GNCN-024 / 03WO 320289-2158 a reference BRAF mutation-free cancer sample; and classifying the sample as having a positive BRAF predictive response signature based on the results of the comparing step.

58. The method of claim 57, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 8 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 8 from the at least one training set; and classifying the sample as possessing a positive BRAF predictive response signature based on the results of the statistical algorithm.

59. The method of claim 57, wherein the at least one training set is from a reference BRAF mutation-containing cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample.

60. The method of claim 57, wherein the at least one training set is from a reference BRAF mutation-containing cancer sample and from a reference BRAF mutation-free cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample.

61. The method of claim 55, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8.

62. The method of claim 55, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8.

63. The method of claim 55, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8.Attorney Docket No. GNCN-024 / 03WO 320289-2158 64. The method of claim 55, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

65. The method of claim 55, wherein the first type of therapy shows inhibitory activity toward BRAF.

66. The method of claim 65, wherein the BRAF inhibitor is selected from the group consisting of vemurafenib, dabrafenib, sorafenib, encorafenib, PLX4032 and any combination thereof.

67. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 10 and the predictive response signature is a Erb-B2 receptor tyrosine kinase 2 (ERBB2) predictive response signature.

68. The method of claim 67, wherein the positive ERBB2 predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene.

69. The method of claim 67, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 10 to an expression level of the plurality of classifier biomarkers selected from Table 10 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB2 mutation-containing or amplification- containing cancer sample, or is from a reference ERBB2 mutation-free cancer sample; and classifying the sample as having a positive ERBB2 predictive response signature based on the results of the comparing step.

70. The method of claim 69, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 10 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 10 from the at least one training set; and classifying the sample as possessing a positive ERBB2 predictive response signature based on the results of the statistical algorithm.

71. The method of claim 69, wherein the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levelsAttorney Docket No. GNCN-024 / 03WO 320289-2158 of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification-containing cancer sample.

72. The method of claim 69, wherein the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and from a reference ERBB2 mutation-free cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification- containing cancer sample.

73. The method of claim 67, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10.

74. The method of claim 67, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10.

75. The method of claim 67, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 10.

76. The method of claim 67, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

77. The method of claim 67, wherein the first type of therapy shows inhibitory activity toward ERBB2.

78. The method of claim 77, wherein the ERBB2 inhibitor is selected from the group consisting of lapatinib, afatinib, AST1306, AEE788, CP724, CP714, CUDC101, TAK285, dacomitinib, pelitinib, AC480, canertinib, tucatinib (irbinitinib, ONT-380), CP-724714, HER@- Inhibitor-1, BDTX-189, varlitinib, sapitinib (AZD8931), allitinib, poziotinib, pyrotinib, neratinib,Attorney Docket No. GNCN-024 / 03WO 320289-2158 tyrophostin AG-258, (-)-epigallocatechin gallate, mobocertinib, trastuzumab, zenocutuzumab and any combination thereof.

79. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 12 and the predictive response signature is a Erb-B3 receptor tyrosine kinase 3 (ERBB3) predictive response signature.

80. The method of claim 79, wherein the positive ERBB3 predictive response signature indicates that the patient possesses one or more genetic variants or amplifications in the Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene.

81. The method of claim 79, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 12 to an expression level of the plurality of classifier biomarkers selected from Table 12 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB3 mutation-containing or amplification- containing cancer sample, or is from a reference ERBB3 mutation-free cancer sample; and classifying the sample as having a positive ERBB3 predictive response signature based on the results of the comparing step.

82. The method of claim 81, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 12 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 12 from the at least one training set; and classifying the sample as possessing a positive ERBB3 predictive response signature based on the results of the statistical algorithm.

83. The method of claim 81, wherein the at least one training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification-containing cancer sample.

84. The method of claim 81, wherein the at least one training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample and from a reference ERBB3 mutation-free cancer sample and the sample is classified as possessing the positive ERBB3Attorney Docket No. GNCN-024 / 03WO 320289-2158 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification- containing cancer sample.

85. The method of claim 79, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12.

86. The method of claim 79, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12.

87. The method of claim 79, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12.

88. The method of claim 79, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

89. The method of claim 79, wherein the first type of therapy shows inhibitory activity toward ERBB3.

90. The method of claim 89, wherein the ERBB3 inhibitor is selected from the group consisting of sapitinib (AZD8931), elgemtumab (LJM716), lumretuzumab (RG7116), KTN3379, patritumab, MM-121, MM-111, MM-141, single domain antibodies BCD090-P1, BCD090-M2, and BCD090-M456, and any combination thereof.

91. The method of claim 1 or 2, wherein the determining the predictive response signature of the sample obtained from the patient suffering from cancer comprises determining expression levels of a plurality of classifier biomarkers selected from Table 14 and the predictive response signature is a Phosphoinositide 3-Kinase CA (PIK3CA) predictive response signature.

92. The method of claim 91, wherein the positive ERBB3 predictive response signature indicates that the patient possesses one or more genetic variants in the Phosphoinositide 3-Kinase CA (PIK3CA) gene.Attorney Docket No. GNCN-024 / 03WO 320289-2158 93. The method of claim 91, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 14 to an expression level of the plurality of classifier biomarkers selected from Table 14 in at least one sample training set, wherein the at least one sample training set is from a reference PIK3CA mutation-containing cancer sample, or is from a reference PIK3CA mutation-free cancer sample; and classifying the sample as having a positive PIK3CA predictive response signature based on the results of the comparing step.

94. The method of claim 93, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 14 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 14 from the at least one training set; and classifying the sample as possessing a positive PIK3CA predictive response signature based on the results of the statistical algorithm.

95. The method of claim 93, wherein the at least one training set is from a reference PIK3CA mutation-containing cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample.

96. The method of claim 93, wherein the at least one training set is from a reference PIK3CA mutation-containing cancer sample and from a reference PIK3CA mutation-free cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample.

97. The method of claim 91, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14.

98. The method of claim 91, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at leastAttorney Docket No. GNCN-024 / 03WO 320289-2158 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14.

99. The method of claim 91, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14.

100. The method of claim 91, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

101. The method of claim 91, wherein the first type of therapy shows inhibitory activity toward PIK3CA.

102. The method of claim 101, wherein the PIK3CA inhibitor is selected from the group consisting of alpelisib, idelalisib, duvelisib, copanlisib, umbralisib, buparlisib, dactolisib, leniolisib, parsaclisib, paxalisib, taselisib, zandelisib, inacolisib, apitolisib, bimiralisib, eganelisib, fimepinostat, gedatolisib, linperlisib, nemiralisib, pictilisib, pilaralisib, samotolisib, seletalisib, serabelisib, sonolisib, tenalisib, voxtalisib, AMG 319, AZD8186, GSK2636771, SF1126, acalisib, omipalisib, AZD8835, CAL263, GSK1059615, MEN1611, PWT33597, TG100-115, ZSTK474, GDC0077 and any combination thereof.

103. The method of claim 1 or 2, wherein the cancer the patient is suffering from is selected from the group consisting of adrenocortical carcinoma (ACC), bladder urothelial carcinoma (BLCA), breast cancer (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), diffuse large B-cell lymphoma (DLBC), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), low grade glioma (LGG), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), mesothelioma (MESO), pancreatic adenocarcinoma (PAAD), pheochromocytoma and paraganglioma (PCPG), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), sarcoma (SARC), skin cutaneous melanoma (SKCM), testicular germ cell tumors (TGCT), thyroid cancer (THCA), thymus cancer (THYM), uterine corpus endometrial carcinoma (UCEC), uterine carcinsarcoma (UCS) and uveal melanoma (UVM)..

104. The method of claim 103, wherein the cancer is LUAD.Attorney Docket No. GNCN-024 / 03WO 320289-2158 105. The method of claim 1 or 2, wherein the sample is a formalin-fixed, paraffin- embedded (FFPE) tissue sample, fresh or a frozen tissue sample, an exosome, or a bodily fluid obtained from the patient.

106. The method of claim 105, wherein the bodily fluid is blood or fractions thereof, urine, saliva, or sputum.

107. A method of treating cancer in a patient, the method comprising: measuring the expression level of a plurality of classifier biomarkers in a sample obtained from a patient suffering from cancer, wherein the plurality of classifier biomarkers are selected from classifier biomarkers listed in Table 2, Table 4, Table 6, Table 8, Table 10, Table 12, Table 14 or any combination thereof, wherein the measured expression levels of the plurality of classifier biomarkers provide a target gene specific activation signature for the sample; and administering a specific therapy to the patient based on presence of a positive target gene specific activation signature, wherein the positive target gene specific activation signature is indicative of presence of one or more genetic variants in the target gene.

108. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 2, the target gene is epidermal growth factor receptor (egfr) gene and the target gene specific activation signature is an epidermal growth factor receptor (EGFR) predictive response signature.

109. The method of claim 108, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 2 to an expression level of the plurality of classifier biomarkers selected from Table 2 in at least one sample training set, wherein the at least one sample training set is from a reference EGFR mutation-containing cancer sample, or is from a reference EGFR mutation-free cancer sample; and classifying the sample as having a positive EGFR predictive response signature based on the results of the comparing step.

110. The method of claim 109, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 2 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 2 from the at least one training set; and classifying the sample as possessing a positive EGFR predictive response signature based on the results of the statistical algorithm.Attorney Docket No. GNCN-024 / 03WO 320289-2158 111. The method of claim 110, wherein the at least one training set is from a reference EGFR mutation-containing cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample.

112. The method of claim 110, wherein the at least one training set is from a reference EGFR mutation-containing cancer sample and from a reference EGFR mutation-free cancer sample and the sample is classified as possessing the positive EGFR predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 2 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 2 from the reference EGFR mutation-containing cancer sample.

113. The method of claim 108, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2.

114. The method of claim 108, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2.

115. The method of claim 108, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2.

116. The method of claim 108, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

117. The method of claim 108, wherein the specific therapy shows inhibitory activity toward EGFR.

118. The method of claim 117, wherein the EGFR inhibitor is a tyrosine kinase inhibitor.

119. The method of claim 118, wherein the EGFR inhibitor is a selective tyrosine kinase inhibitor.

120. The method of claim 118, wherein the EGFR inhibitor is a non-selective tyrosine kinase inhibitor.Attorney Docket No. GNCN-024 / 03WO 320289-2158 121. The method of claim 117, wherein the EGFR inhibitor is selected from the group consisting of erlotinib (OSI-744), poziotinib (HM781-36B), osimertinib (AZD9291), AG-490 (Tyrphostin B42), afatinib (BIBW2992), gefitinib (ZD1839), lapatinib (GW-572016), rociletinib (CO-1686), neratinib, lucitanib (E3810), dacomitinib, mobocertinib, vandetanib, canertinib (CI- 1033), BDTX-189, epertinib, AEE788, CUDC-101, pelitinib, sapitinib, varlitinib, pyrotinib, TAK- 285, AC480, tyrophostin AG-528 and any combination thereof.

122. The method of claim 117, wherein the EGFR inhibitor is canertinib (CI-1033).

123. The method of claim 117, wherein the EGFR inhibitor is an antibody or antibody- conjugate.

124. The method of claim 123, wherein the EGFR inhibitor is cetuximab, panitumumab or necitumumab.

125. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 4, the target gene is MET proto-oncogene, receptor tyrosine kinase (met) gene and the target gene specific activation signature is a MET proto-oncogene, receptor tyrosine kinase (MET) predictive response signature.

126. The method of claim 125, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 4 to an expression level of the plurality of classifier biomarkers selected from Table 4 in at least one sample training set, wherein the at least one sample training set is from a reference MET mutation-containing or amplification- containing cancer sample, or is from a reference MET mutation-free cancer sample; and classifying the sample as having a positive MET predictive response signature based on the results of the comparing step.

127. The method of claim 126, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 4 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 4 from the at least one training set; and classifying the sample as possessing a positive MET predictive response signature based on the results of the statistical algorithm.

128. The method of claim 126, wherein the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of theAttorney Docket No. GNCN-024 / 03WO 320289-2158 plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation- containing or amplification-containing cancer sample.

129. The method of claim 126, wherein the at least one training set is from a reference MET mutation-containing or amplification-containing cancer sample and from a reference MET mutation-free cancer sample and the sample is classified as possessing the positive MET predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 4 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 4 from the reference MET mutation-containing or amplification-containing cancer sample.

130. The method of claim 125, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4.

131. The method of any one of claims 125-129, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4.

132. The method of claim 125, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4.

133. The method of claim 125, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

134. The method of claim 125, wherein the specific therapy shows inhibitory activity toward MET.

135. The method of claim 134, wherein the MET inhibitor is a tyrosine kinase inhibitor.

136. The method of claim 135, wherein the MET inhibitor is a selective tyrosine kinase inhibitor.

137. The method of claim 135, wherein the MET inhibitor is a non-selective tyrosine kinase inhibitor.Attorney Docket No. GNCN-024 / 03WO 320289-2158 138. The method of claim 134, wherein the MET inhibitor is selected from the group consisting of crizotinib, capmatinib, tepotinib, savolitinib, cabozantinib, glesatinib merestinib and any combination thereof.

139. The method of claim 134, wherein the MET inhibitor is an antibody or antibody- conjugate.

140. The method of claim 139, wherein the MET inhibitor is emibetuzumab, Rilotumumab, Ficlatuzumab, TAK-701, Onartuzumab, ARGX-111, or EM1-mAb.

141. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 6, the target gene is Kirsten Rat Sarcoma viral oncogene homolog (kras) gene and the target gene specific activation signature is a Kirsten Rat Sarcoma viral oncogene homolog (KRAS) predictive response signature.

142. The method of claim 141, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 6 to an expression level of the plurality of classifier biomarkers selected from Table 6 in at least one sample training set, wherein the at least one sample training set is from a reference KRAS mutation-containing cancer sample, or is from a reference KRAS mutation-free cancer sample; and classifying the sample as having a positive KRAS predictive response signature based on the results of the comparing step.

143. The method of claim 142, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 6 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 6 from the at least one training set; and classifying the sample as possessing a positive KRAS predictive response signature based on the results of the statistical algorithm.

144. The method of claim 142, wherein the at least one training set is from a reference KRAS mutation-containing cancer sample and the sample is classified as possessing the positive KRAS predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample.

145. The method of claim 142, wherein the at least one training set is from a reference KRAS mutation-containing cancer sample and from a reference KRAS mutation-free cancer sample and the sample is classified as possessing the positive KRAS predictive response signatureAttorney Docket No. GNCN-024 / 03WO 320289-2158 if the expression levels of the plurality of classifier biomarkers selected from Table 6 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 6 from the reference KRAS mutation-containing cancer sample.

146. The method of claim 141, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6.

147. The method of claim 141, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6.

148. The method of claim 141, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 6.

149. The method of claim 141, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

150. The method of claim 141, wherein the specific therapy shows inhibitory activity toward KRAS, MEK1, MEK2 or extracellular signal-regulated kinase (ERK).

151. The method of claim 150, wherein the KRAS inhibitor is selected from the group consisting of Adagrasib, JQ443, GDC-6036, BI 1823911, JNJ-74699157, MK-1084, SCH-53239, AMG510, MRTX849, ARS-3248, compound 3144, Abd-7, BI02852, BI-3406, tipifarnib, SHP099, JAB-3068, RMC-4550, TNO155 and any combination thereof.

152. The method of claim 150, wherein the MEK inhibitor is selected from the group consisting of binimetinib, cobimetinib, selumetinib, trametinib, CI-1040, TAK-733, pimasertib, and any combination thereof.

153. The method of claim 150, wherein the ERK inhibitor is selected from the group consisting of PD0325901, ASN007, ulixertinib (BVD-523), CC-9003, ravoxertinib (GDC-0994), MK-8353, KO-947, LTT462, temuterib (LY3214996), magnolin, SCH772984, pluripotin, FR 180204, resveratrol, VX-11e, AZD0364. MRTX-1257, ERK5-IN-1, ERK5-IN-2, XMD8-92, DEL-22379 and any combination thereof.Attorney Docket No. GNCN-024 / 03WO 320289-2158 154. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 8, the target gene is B-Raf protooncogene (BRAF) gene and the target gene specific activation signature is a B-Raf protooncogene (BRAF) predictive response signature.

155. The method of claim 154, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 8 to an expression level of the plurality of classifier biomarkers selected from Table 8 in at least one sample training set, wherein the at least one sample training set is from a reference BRAF mutation-containing cancer sample, or is from a reference BRAF mutation-free cancer sample; and classifying the sample as having a positive BRAF predictive response signature based on the results of the comparing step.

156. The method of claim 155, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 8 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 8 from the at least one training set; and classifying the sample as possessing a positive BRAF predictive response signature based on the results of the statistical algorithm.

157. The method of claim 155, wherein the at least one training set is from a reference BRAF mutation-containing cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample.

158. The method of claim 155, wherein the at least one training set is from a reference BRAF mutation-containing cancer sample and from a reference BRAF mutation-free cancer sample and the sample is classified as possessing the positive BRAF predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 8 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 8 from the reference BRAF mutation-containing cancer sample.

159. The method of claim 154, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8.Attorney Docket No. GNCN-024 / 03WO 320289-2158 160. The method of claim 154, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8.

161. The method of claim 154, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8.

162. The method of claim 154, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

163. The method of claim 154, wherein the specific therapy shows inhibitory activity toward BRAF.

164. The method of claim 163, wherein the BRAF inhibitor is selected from the group consisting of vemurafenib, dabrafenib, sorafenib, encorafenib, PLX4032 and any combination thereof.

165. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 10, the target gene is Erb-B2 receptor tyrosine kinase 2 (ERBB2) gene and the target gene specific activation signature is a Erb-B2 receptor tyrosine kinase 2 (ERBB2) predictive response signature.

166. The method of claim 165, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 10 to an expression level of the plurality of classifier biomarkers selected from Table 10 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB2 mutation-containing or amplification- containing cancer sample, or is from a reference ERBB2 mutation-free cancer sample; and classifying the sample as having a positive ERBB2 predictive response signature based on the results of the comparing step.

167. The method of claim 166, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 10 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 10 from the at least one training set;Attorney Docket No. GNCN-024 / 03WO 320289-2158 and classifying the sample as possessing a positive ERBB2 predictive response signature based on the results of the statistical algorithm.

168. The method of claim 166, wherein the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification-containing cancer sample.

169. The method of claim 166, wherein the at least one training set is from a reference ERBB2 mutation-containing or amplification-containing cancer sample and from a reference ERBB2 mutation-free cancer sample and the sample is classified as possessing the positive ERBB2 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 10 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 10 from the reference ERBB2 mutation-containing or amplification- containing cancer sample.

170. The method of claim 165, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10.

171. The method of claim 165, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10.

172. The method of claim 165, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 10.

173. The method of claim 165, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.Attorney Docket No. GNCN-024 / 03WO 320289-2158 174. The method of claim 165, wherein the first type of therapy shows inhibitory activity toward ERBB2.

175. The method of claim 174, wherein the ERBB2 inhibitor is selected from the group consisting of lapatinib, afatinib, AST1306, AEE788, CP724, CP714, CUDC101, TAK285, dacomitinib, pelitinib, AC480, canertinib, tucatinib (irbinitinib, ONT-380), CP-724714, HER@- Inhibitor-1, BDTX-189, varlitinib, sapitinib (AZD8931), allitinib, poziotinib, pyrotinib, neratinib, tyrophostin AG-258, (-)-epigallocatechin gallate, mobocertinib, trastuzumab, zenocutuzumab and any combination thereof.

176. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 12, the target gene is Erb-B3 receptor tyrosine kinase 3 (ERBB3) gene and the target gene specific activation signature is a Erb-B3 receptor tyrosine kinase 3 (ERBB3) predictive response signature.

177. The method of claim 176, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 12 to an expression level of the plurality of classifier biomarkers selected from Table 12 in at least one sample training set, wherein the at least one sample training set is from a reference ERBB3 mutation-containing or amplification- containing cancer sample, or is from a reference ERBB3 mutation-free cancer sample; and classifying the sample as having a positive ERBB3 predictive response signature based on the results of the comparing step.

178. The method of claim 177, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 12 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 12 from the at least one training set; and classifying the sample as possessing a positive ERBB3 predictive response signature based on the results of the statistical algorithm.

179. The method of claim 177, wherein the at least one training set is from a reference ERBB3 mutation-containing or amplification-containing cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification-containing cancer sample.Attorney Docket No. GNCN-024 / 03WO 320289-2158 180. The method of claim 177, wherein the at least one training set is from a reference ERBB3 mutation-containing cancer or amplification-containing sample and from a reference ERBB3 mutation-free cancer sample and the sample is classified as possessing the positive ERBB3 predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 12 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 12 from the reference ERBB3 mutation-containing or amplification- containing cancer sample.

181. The method of claim 176, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12.

182. The method of claim 176, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12.

183. The method of claim 176, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12.

184. The method of claim 176, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

185. The method of claim 176, wherein the specific therapy shows inhibitory activity toward ERBB3.

186. The method of claim 185, wherein the ERBB3 inhibitor is selected from the group consisting of sapitinib (AZD8931), elgemtumab (LJM716), lumretuzumab (RG7116), KTN3379, patritumab, MM-121, MM-111, MM-141, single domain antibodies BCD090-P1, BCD090-M2, and BCD090-M456, and any combination thereof.

187. The method of claim 107, wherein the plurality of classifier biomarkers are selected from Table 14, the target gene is Phosphoinositide 3-Kinase CA (PIK3CA) gene and the target gene specific activation signature is a Phosphoinositide 3-Kinase CA (PIK3CA) predictive response signature.Attorney Docket No. GNCN-024 / 03WO 320289-2158 188. The method of claim 187, further comprising comparing the expression levels of the plurality of classifier biomarkers selected from Table 14 to an expression level of the plurality of classifier biomarkers selected from Table 14 in at least one sample training set, wherein the at least one sample training set is from a reference PIK3CA mutation-containing cancer sample, or is from a reference PIK3CA mutation-free cancer sample; and classifying the sample as having a positive PIK3CA predictive response signature based on the results of the comparing step.

189. The method of claim 188, wherein the comparing comprises applying a statistical algorithm that comprises determining a correlation between the expression levels of the plurality of classifier biomarkers selected from Table 14 obtained from the sample and the expression levels of the plurality of classifier biomarkers selected from Table 14 from the at least one training set; and classifying the sample as possessing a positive PIK3CA predictive response signature based on the results of the statistical algorithm.

190. The method of claim 188, wherein the at least one training set is from a reference PIK3CA mutation-containing cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample.

191. The method of claim 188, wherein the at least one training set is from a reference PIK3CA mutation-containing cancer sample and from a reference PIK3CA mutation-free cancer sample and the sample is classified as possessing the positive PIK3CA predictive response signature if the expression levels of the plurality of classifier biomarkers selected from Table 14 correlate with the expression levels of the plurality of classifier biomarkers selected from Table 14 from the reference PIK3CA mutation-containing cancer sample.

192. The method of claim 187, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14.

193. The method of claim 187, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at leastAttorney Docket No. GNCN-024 / 03WO 320289-2158 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14.

194. The method of claim 187, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14.

195. The method of claim 187, wherein the determining the expression levels of the plurality of classifier biomarkers is at a nucleic acid level by performing RNA sequencing, reverse transcriptase polymerase chain reaction (RT-PCR) or hybridization-based analyses.

196. The method of claim 187, wherein the specific therapy shows inhibitory activity toward PIK3CA.

197. The method of claim 196, wherein the PIK3CA inhibitor is selected from the group consisting of alpelisib, idelalisib, duvelisib, copanlisib, umbralisib, buparlisib, dactolisib, leniolisib, parsaclisib, paxalisib, taselisib, zandelisib, inacolisib, apitolisib, bimiralisib, eganelisib, fimepinostat, gedatolisib, linperlisib, nemiralisib, pictilisib, pilaralisib, samotolisib, seletalisib, serabelisib, sonolisib, tenalisib, voxtalisib, AMG 319, AZD8186, GSK2636771, SF1126, acalisib, omipalisib, AZD8835, CAL263, GSK1059615, MEN1611, PWT33597, TG100-115, ZSTK474, GDC0077 and any combination thereof.

198. The method of claim 107, wherein the cancer the patient is suffering from is selected from the group consisting of adrenocortical carcinoma (ACC), bladder urothelial carcinoma (BLCA), breast cancer (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), diffuse large B-cell lymphoma (DLBC), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), low grade glioma (LGG), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), mesothelioma (MESO), pancreatic adenocarcinoma (PAAD), pheochromocytoma and paraganglioma (PCPG), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), sarcoma (SARC), skin cutaneous melanoma (SKCM), testicular germ cell tumors (TGCT), thyroid cancer (THCA), thymus cancer (THYM), uterine corpus endometrial carcinoma (UCEC), uterine carcinsarcoma (UCS) and uveal melanoma (UVM).

199. The method of claim 198, wherein the cancer is LUAD.Attorney Docket No. GNCN-024 / 03WO 320289-2158 200. The method of claim 107, wherein the sample is a formalin-fixed, paraffin- embedded (FFPE) tissue sample, fresh or a frozen tissue sample, an exosome, or a bodily fluid obtained from the patient.

201. The method of claim 200, wherein the bodily fluid is blood or fractions thereof, urine, saliva, or sputum.

202. A method of detecting a biomarker in a sample obtained from a patient suffering from cancer, the method comprising, consisting essentially of or consisting of measuring the expression level of a plurality of biomarker nucleic acids selected from Table 2, Table 4, Table 6, Table 8, Table 10, Table 12 or Table 14 using an amplification, hybridization and / or sequencing assay.

203. The method of claim 202, wherein the sample was previously diagnosed as being a cancer selected from ACC, BLCA, BRCA, CESC, CHOL, COAD, DLBC, GBM, HNSC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, MESO, PAAD, PCPG, PRAD, READ, SARC, SKCM, TGCT, THCA, THYM, UCEC, UCS and UVM.

204. The method of claim 203, wherein the sample was previously diagnosed as being LUAD.

205. The method of claim 202, wherein the amplification, hybridization and / or sequencing assay comprises performing quantitative real time reverse transcriptase polymerase chain reaction (qRT-PCR), RNAseq, microarrays, gene chips, nCounter Gene Expression Assay, Serial Analysis of Gene Expression (SAGE), Rapid Analysis of Gene Expression (RAGE), nuclease protection assays, Northern blotting, or any other equivalent gene expression detection techniques.

206. The method of claim 205, wherein the expression level is detected by performing qRT-PCR.

207. The method of claim 206, wherein the detection of the expression level comprises using at least one pair of oligonucleotide primers per each biomarker nucleic acid from the plurality of biomarker nucleic acids selected from Table 2, Table 4, Table 6, Table 8, Table 10, Table 12 or Table 14.

208. The method of claim 202, wherein the sample is a formalin-fixed, paraffin- embedded (FFPE) lung tissue sample, fresh or a frozen tissue sample, an exosome, wash fluids, cell pellets, or a bodily fluid obtained from the patient.Attorney Docket No. GNCN-024 / 03WO 320289-2158 209. The method of claim 208, wherein the bodily fluid is blood or fractions thereof, urine, saliva, or sputum.

210. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 2.

211. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 2.

212. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 2.

213. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26 or at least 28 classifier biomarkers selected from Table 4.

214. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 4.

215. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 4.

216. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16 or at least 18 classifier biomarkers selected from Table 6.

217. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 6.Attorney Docket No. GNCN-024 / 03WO 320289-2158 218. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 6.

219. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190 or at least 195 classifier biomarkers selected from Table 8.

220. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 8.

221. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 8.

222. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 10, at least 20, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, at least 100, at least 110, at least 120, at least 130, at least 140, at least 150, at least 160, at least 170, at least 180, at least 190, at least 200, at least 210 or at least 220 classifier biomarkers selected from Table 10.

223. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 10.

224. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 10.

225. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, or at least 18 classifier biomarkers selected from Table 12.

226. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, atAttorney Docket No. GNCN-024 / 03WO 320289-2158 least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 12.

227. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 12.

228. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of at least 2, at least 4, at least 6, at least 8, at least 10, at least 12, at least 14, at least 16, at least 18, at least 20, at least 22, at least 24, at least 26, at least 28, at least 30, at least 32, or at least 34 classifier biomarkers selected from Table 14.

229. The method of claim 202, wherein the plurality of classifier biomarkers comprises at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95% or at least 99% of the classifier biomarkers selected from Table 14.

230. The method of claim 202, wherein the plurality of classifier biomarkers comprises, consists essentially of or consists of all the classifier biomarkers from Table 14.

Images