Pan-cancer classification based on FMRP pathway activity informs differences in prognosis and treatment response

Abstract

The present invention relates to methods and compositions that provide a companion diagnostic for cancer therapy. Methods for identifying and stratifying patients or groups of patients with cancer as (i) having high or low FMRP activity, (ii) having a high or low risk prognosis, and / or (iii) being responders or non-responders to cancer therapy, and / or (iv) having high or low immune cell infiltration tumors.

Description

[Technical field]

[0001] The present invention relates to methods and compositions that provide a companion diagnostic for cancer therapy. [Background technology]

[0002] The role of upregulated fragile X mental retardation protein (FMRP) protein in cancer cells has been previously shown (see, e.g., U.S. Patent Publication No. 2020-0354718), with upregulated FMRP activity suppressing immune responses to tumors. Genetic disruption of the FMR1 gene, which encodes FMRP, relieves immune suppression in mouse models, activates CD8 T cell-mediated tumor immunity, and results in tumor shrinkage and extended survival compared to isogenic FMRP-expressing tumors lacking genetic disruption of the FMR1 gene.

[0003] Despite the demonstrable role of FMRP in tumor progression and shaping the immunosuppressive tumor microenvironment, the assessment of its functional activity in tumor samples has proven difficult. Due to multiple post-transcriptional and post-translational modifications of FMR1 mRNA and FMRP protein, respectively, the levels of FMR1 mRNA expression and FMRP protein expression are not good biomarkers of the intrinsic immunosuppressive activity of this protein.

[0004] Thus, there is a need for improved methods for assessing FMRP activity in tumors and determining the likelihood that cancers can be successfully treated with various cancer therapies whose effectiveness depends on or is limited by FMRP pathway activity. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] US Patent Application Publication No. 2020-0354718 Summary of the Invention [Means for solving the problem]

[0006] The present invention relates to methods and compositions that provide companion diagnostics for cancer therapy. In particular, the present invention relates to methods and reagents for determining the likelihood that a cancer can be successfully treated with a cancer therapy whose effectiveness depends on or is limited by FMRP pathway activity. The methods and compositions of the present invention are useful for separating cancer patients as potential responders from non-responders to cancer therapy. The present invention is based, at least in part, on the discovery that treatment with a cancer therapy is more likely to be effective when the patient's FMRP activity score is taken into account. [Brief description of the drawings]

[0007] [Figure 1 (1)] Figure 1A-1H show patient classification across 31 different cancer types. In contrast to the case based on FMR1 mRNA expression (panels A and B), which is uninformative, the newly invented FMRP pathway activity signature scoring system (FMRP activity pan-signature: panels C and D; sub-signature 1: panels E and F; sub-signature 3: panels G and H) is informative and statistically significant for all. Each panel shows the association (or non-association) with patient prognosis (A, C, E, and G: overall survival; B, D, F, and H: progression-free survival). To estimate the significance of the correlation, a COX model was used considering tumor type as a covariate. The data used in this figure was downloaded from the latest TCGA PanCan Atlas. [Figure 1 (2)]Figure 1A-H show patient classification across 31 different cancer types. The case based on FMR1 mRNA expression (panels A and B) is uninformative, but in contrast, the newly invented FMRP pathway activity signature scoring system (FMRP activity pan-signature: panels C and D; subsignature 1: panels E and F; subsignature 3: panels G and H) is informative and statistically significant for all. Each panel shows association (or non-association) with patient prognosis (A, C, E, and G: overall survival; B, D, F, and H: progression-free survival). To estimate the significance of the correlation, a COX model was used considering tumor type as a covariate. The data used in this figure was downloaded from the latest TCGA PanCan Atlas. [Figure 2 (1)] Figure 2A-C show FMRP activity scores in breast cancer. A. FMRP activity scores show the highest levels in the basal-like subtype, the most aggressive (invasive) subtype of breast cancer. Signature scores for this panel were derived using only upregulated genes in pan-signature 1. B. FMRP activity scores (pan-signature) correlate with overall survival of all breast cancer patients. Data used in this figure was downloaded from the most recent breast cancer cohort of the TCGA PanCan Atlas. [Figure 2 (2)] Figure 2A-2C show FMRP activity scores in breast cancer. C. FMRP activity scores (pan-signatures) specifically correlate with overall survival in luminal A subtype breast cancer patients. Data used in this figure were downloaded from the most recent breast cancer cohort of the TCGA PanCan Atlas. [Figure 3(1)] Figure 3A-C show FMRP activity scores (pan-signatures) in colorectal cancer. A. FMRP activity scores correlate with overall survival in all colorectal cancer patients. Data used in this figure were downloaded from the most recent colorectal cancer cohort of the TCGA PanCan Atlas. [Figure 3 (2)]Figure 3A-C show FMRP activity scores (pan-signatures) in colorectal cancer. A. FMRP activity score correlates with overall survival in patients with microsatellite stable (MSS) colorectal cancer. B. FMRP activity score correlates with overall survival in patients with microsatellite instability (MSI) colorectal cancer. C. FMRP activity score does not correlate with overall survival in patients with microsatellite instability (MSI) colorectal cancer. Data used in this figure was downloaded from the most recent colorectal cancer cohort of the TCGA PanCan Atlas. [Figure 4(1)] Figures 4A-4D show FMRP activity score correlation with immune checkpoint inhibitor therapy response in cancer patients. Figure 4A. FMRP activity score correlation with overall survival of melanoma patients receiving anti-PD1 therapy (left panel); non-responders to anti-PD1 therapy show higher levels of FMRP activity score (right panel). Figure 4B. FMRP activity score correlation with overall survival of lung cancer patients receiving anti-PD1 or anti-PD-L1 therapy (left panel); non-responders to anti-PD1 or anti-PD- therapy show higher levels of FMRP activity score (right panel). Only upregulated genes in sub-signature 1 were used to derive the signature scores in panels A-B. [Figure 4 (2)] Figure 4A-4D show FMRP activity score correlation with immune checkpoint inhibitor therapy response in cancer patients. Figure 4C. FMRP activity score correlation with overall survival of urothelial carcinoma patients receiving anti-PD-L1 therapy (left panel); non-responders to anti-PD-L1 therapy show higher levels of FMRP activity score (right panel). Only upregulated genes in sub-signature 1 were used to derive the signature score in panel C. Figure 4D. FMRP activity score (pan-signature) correlation with overall survival of melanoma patients receiving anti-CTLA4 therapy (left panel); non-responders to anti-CTLA4 therapy show higher levels of FMRP activity score (right panel). [Diagram 5]Figure 5A and Figure 5B show FMRP activity score correlation with chemotherapy response in cancer patients. Figure 5A. FMRP activity score correlation with disease-free survival in breast cancer patients receiving taxanes (left panel); notably, the signature score is independent of tumor aggressiveness (T stage, right panel) and is also shown using a COX model in survival analysis considering T stage as a covariate, thus revealing that the FMRP activity signature constitutes an independent prognostic marker. Figure 5B. FMRP activity score correlation with progression-free survival in lung cancer patients receiving paclitaxel, cisplatin, or carboplatin (left panel); the signature score is also independent of tumor aggressiveness (T stage, right panel) and constitutes an independent prognostic factor. To estimate the significance of the correlation for survival analysis, a COX model was used, considering T stage as a covariate. To derive the signature scores for all panels of Figure 5, only upregulated genes from sub-signature 1 were used. [Figure 6(1)]Figures 6A-6N show the irreproducibility and lack of correlation between previously published FMRP signatures and the signature described in this invention. FMR1 mRNA expression (Figures 6A and 6B) and FMRP network signature (Luca et al. (2013). The fragile X protein binds mRNAs involved in cancer progression and modulates metastasis formation. EMBO Mol.Med. 5, 1523-1536, Figures 6C and 6D) correlations with breast cancer patient survival are uninformative or not statistically significant. Each panel shows an association (or no association) with patient prognosis (Figures 6A, 6C: overall survival; Figures 6B, 6D: progression-free survival). Figure 6E. The genes constituting the FMRP network signature proposed by Rossella Luca et al., 2013, do not show significant overlap with the pan-signature described in this invention. Correlations of FMR1 mRNA expression (Fig. 6F and 6G) ​​and FMRP network signature (F. Zalfa et al., (2017). The fragile X mental retardation protein regulates tumor invasiveness-related pathways in melanoma cells. Cell Death Dis. 8, e3169, Fig. 6H and 6I) with melanoma patient survival were again uninformative or not statistically significant. Each panel shows the patient prognosis (Fig. 6F, 6H: overall survival; Fig. 6G, 6I: progression-free survival). [Figure 6 (2)] Figures 6A-6N show the non-reproducibility and lack of correlation between previously published FMRP signatures and the signatures described in this invention. Figure 6J. Genes comprising the FMRP network signature proposed by F. Zalfa et al., 2017, show no significant overlap with the pan-signature provided in this invention. FMR1 mRNA expression (Figures 6K and 6L) and RIPK1 mRNA expression (Figures 6M and 6N) correlations with colorectal cancer patient survival are again uninformative or not statistically significant. Each panel shows the patient prognosis (Figures 6K, 6M: overall survival; Figures 6L, 6N: progression-free survival). [Figure 7(1)] Figures 7A-7E show FMRP activity scores (pan-signature) in adrenocortical carcinoma, endometrial carcinoma, esophageal adenocarcinoma, pancreatic adenocarcinoma, and hepatic hepatocellular carcinoma. Figures 7A-7B show the correlation between the pan-signature score and overall survival (OS, left panel) and progression-free survival (PFS, right panel) in adrenocortical carcinoma (A) and endometrial carcinoma (B). [Figure 7(2)] Figure 7A-7E show FMRP activity scores (pan-signature) in adrenocortical carcinoma, endometrial carcinoma, esophageal adenocarcinoma, pancreatic adenocarcinoma, and hepatic hepatocellular carcinoma. Figure 7C shows the correlation between the pan-signature score and overall survival (OS, left panel) and progression-free survival (PFS, right panel) in esophageal adenocarcinoma (C). Figure 7D shows the correlation between the pan-signature score and overall survival in pancreatic adenocarcinoma (D). [Figure 7(3)] Figures 7A to 7E show FMRP activity scores (pan-signature) in adrenocortical carcinoma, endometrial carcinoma, esophageal adenocarcinoma, pancreatic adenocarcinoma, and hepatic hepatocellular carcinoma. Figure 7E shows the correlation between the pan-signature score and overall survival in hepatocellular carcinoma (E). [Figure 8]Figures 8A-8E. Figures 8A and 8B show that FMRP activity scores are negatively associated with both CD8 T infiltration in multiple human tumors. Figure 8A shows the anti-correlation between FMRP activity scores (pan-signature) and CD8 T cell infiltration signatures estimated by the xCell package in human pan-cancer. A linear regression model with tumor type as a covariate was used to estimate the significance of the correlation. Figure 8B shows the anti-correlation between FMRP activity scores (sub-signature 1) and CD8 T cell infiltration signatures, similar to Figure 8A. Only upregulated genes in FMRP activity sub-signature 1 were used to derive the signature scores in this analysis. Figures 8C-8E show no correlation between FMRP activity scores and tumor grade. Figure 8C shows the distribution of FMRP pan-signature scores across different tumor grades in the TCGA human pan-cancer dataset. Figure 8D shows the distribution of FMRP sub-signature 1 scores across different tumor grades in the TCGA human pan-cancer dataset. Figure 8E shows the distribution of FMRP sub-signature 1 scores using only up-regulated genes in the FMRP activity signature list across different tumor grades in the TCGA human pan-cancer dataset. [Figure 9(1)] Figures 9A-9L. Figures 9A-9C show anti-correlation of FMRP activity score (pan-signature) with progression-free survival (PFS, left panel) and CD8 T cell infiltration signature (right panel) in endometrial cancer (A), melanoma (B), and head and neck squamous cell carcinoma (C). Log-rank test was used for survival analysis, and Wilcoxon two-tailed test was used for CD8 T cell association analysis. [Figure 9 (2)]Figures 9A-9L. Figures 9D-9F show box plot comparisons of CD8 T cell infiltration scores in endometrial cancer (D), melanoma (E), and head and neck squamous cell carcinoma (F) tumor samples with high versus low FMRP subsignature 1 scores. Only upregulated genes in FMRP subsignature 1 were used to derive the signature scores in this analysis. Figures 9G-9I show the distribution of FMRP pan-signature scores across different tumor grades in endometrial cancer (G), melanoma (H), and head and neck squamous cell carcinoma (I). [Figure 9 (3)] Figure 9A-L. Figure 9J shows FMRP activity scores (pan-signature) in human breast cancer. i: Box plot comparison of CD8 T cell infiltration scores in tumor samples with high vs. low FMRP signature scores. ii: Box-pot comparison of FMRP signature scores in immune-excluded vs. inflammatory breast cancer tumors (cohort: GSE177043). iii: Box-plot comparison of FMRP signature scores in low vs. high TCR diversity breast cancer tumors (cohort: GSE177043). Wilcoxon two-tailed test. Figure 9K shows the distribution of FMRP pan-signature scores across different tumor grades in the TCGA breast cancer cohort. Figure 9L shows box-pot comparison of FMRP sub-signature 1 scores in immune-excluded vs. inflammatory breast cancer tumors (cohort: GSE177043). [Figure 10(1)] Figures 10A-H show the level of tumor inflammation by CD8 T cells based on pan-signatures in specific cancer cells. Figure 10A shows the inverse correlation between FMRP pan-immunosuppressive signature score and CD8 T cell infiltration signature estimated by the xCell package in human pan-cancers. A linear regression model with tumor type as a covariate was used to estimate the significance of the correlation. Figures 10B-C show the inverse correlation between FMRP pan-immunosuppressive signature score and CD8 T cell infiltration signature in cancer-specific analyses; bladder cancer (B), colorectal adenocarcinoma (C). Wilcoxon two-tailed test was used to estimate the significance. [Figure 10(2)]Figures 10A-H show the level of tumor inflammation by CD8 T cells based on pan-signatures in specific cancer cells. Figures 10D-E show the inverse correlation between FMRP pan-immunosuppressive signature score and CD8 T cell infiltration signature in cancer-specific analyses; glioma (D), liver cancer (E). Wilcoxon two-tailed test was used to estimate significance. [Figure 10(3)] Figures 10A-H show the level of tumor inflammation by CD8 T cells based on pan-signatures in specific cancer cells. Figures 10F-H show the inverse correlation between FMRP pan-immunosuppressive signature score and CD8 T cell infiltration signature in cancer-specific analyses; non-small cell line cancer (F), pancreatic adenocarcinoma (G), and thymic epithelial tumor (H). Wilcoxon two-tailed tests were used to estimate significance. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] The present invention is based on the analysis of gene expression signatures induced by fragile X mental retardation protein (FMRP) protein activity in tumors. FMRP protein is widely upregulated across different types of human cancers, and as shown herein, its functional activity mediates an immunosuppressive effect in the tumor microenvironment, reflected in pathway activity signatures. The present invention relates to a method for evaluating the downstream signaling activity of FMRP protein in tumors, thereby predicting the prognosis of cancer patients, i.e., overall survival and progression-free survival, as well as methods for classifying and stratifying such patients. Furthermore, the present invention relates to a companion diagnostic that can be used in the clinic to stratify and prioritize cancer patients for cancer therapy. The concordant differential expression (expression difference) of genes in the signature list conveys the FMRP pathway activity score disclosed herein, which can be used to stratify cancer patients into groups that may differentially benefit from the above-mentioned and potentially other therapeutic modalities for cancer patients, including drugs that inhibit the functional activity of FMRP.

[0009] As used herein, the term "FMRP pathway activity" is also referred to as "FMRP downstream transcriptional network in cancer," "FMRP cancer network signature score," "FMRP cancer signature score," or "FMRP network activity."

[0010] The present invention identifies molecular gene expression biomarkers that reveal FMRP functional activity in tumors and therefore can be used to stratify cancer patients into groups with high, intermediate, and low FMRP pathway activity. These biomarkers can correlate FMRP pathway activity with overall survival (OS) and progression-free survival (PFS), as well as with response to different forms of therapy.

[0011] The present invention allows for the stratification of cancer patients based on the level of tumor inflammation and immune cell infiltration.

[0012] Pan-signature In one embodiment, the present invention provides a "pan-cancer" gene signature, referred to herein as a pan-signature, which can be used to develop a gene expression signature score that can be used to assess the level of FMRP activity in a tumor.

[0013] The pan-signature is a global signature list that includes a complete panel of biomarker genes (a total of 156 genes) discovered by comparing FMRP pathway active tumors versus FMRP pathway inactive tumors and cultured cancer cells. This signature reveals the combined effect of FMRP activity in cancer cells and the tumor microenvironment. The pan-signature is disclosed in Table 1.

[0014] [Table 1(1)] [Table 1(2)] [Table 1(3)] [Table 1(4)] * Secretome refers to the set of proteins differentially secreted by cancer cells with high or low FMRP pathway activity, which can be used, for example, as biomarkers in liquid cytology assays and other diagnostic bioassays. "Upregulation indicates that the gene is positively regulated by FMRP activity, and downregulation, conversely, indicates that the gene is negatively regulated by FMRP activity."

[0015] As used herein, EIF4G3: eukaryotic translation initiation factor 4 gamma 3; SMPDL3B: sphingomyelin phosphodiesterase acid like 3B; VANGL2: VANGL planar cell polarity protein 2; GBP2: guanylate binding protein 2; POGK: pogo transposable element derived with KRAB domain; IFITM2: interferon induced transmembrane protein 2; IFITM1: interferon induced transmembrane protein 1; IFITM3: interferon induced transmembrane protein 3; PDLIM1: PDZ and LIM domain 1 (PDZ and LIM domain 1); PRDX5: peroxiredoxin 5; PFKP: phosphofructokinase, platelet; SIPA1L2: signal induced proliferation associated 1 like 2; ACSL5: acyl-CoA synthetase long chain family member 5; RBP4: retinol binding protein 4; BNC1: basonuclin 1; PSME2: proteasome activator subunit 2;B2M: beta-2-microglobulin; GAS6: growth arrest specific 6; PSME1: proteasome activator subunit 1; CKMT1B: creatine kinase, mitochondrial 1B; CKMT1A: creatine kinase, mitochondrial 1A; WDR89: WD repeat domain 89; USP50: ubiquitin specific peptidase 50; CRIP1: cysteine ​​rich protein 1; CHCHD10: coiled-coil-helix-coiled-coil-helix domain containing 10; ZNF23: zinc finger protein 23 (zinc finger protein 23); APOB: apolipoprotein B; UBA52: ubiquitin A-52 residue ribosomal protein fusion product 1; POGLUT1: protein O-glucosyltransferase 1; PLAC8: placenta associated 8; STAT1: signal transducer and activator of transcription 1; PDE5A: phosphodiesterase 5A; CPEB2: cytoplasmic polyadenylation element binding protein 2;PCDHB11: protocadherin beta 11; PCDHB12: protocadherin beta 12; PCDHB15: protocadherin beta 15; ATP13A4: ATPase 13A4; HMGB2: high mobility group box 2; RPL29: ribosomal protein L29; PPARGC1A: PPARG coactivator 1 alpha; CHN1: chimerin 1; CCL8: CC motif chemokine ligand 8; SLC4A4: solute carrier family 4 member 4; LSM4: LSM4 homolog, U6 small nuclear RNA and mRNA degradation associated(LSM4 homolog, U6 small nuclear RNA and mRNA degradation associated);KIAA0513:KIAA0513;NME1:NME / NM23 nucleoside diphosphate kinase 1;BST2:bone marrow stromal cell antigen 2;TMEM144:transmembrane protein 144;COL3A1:collagen type III alpha 1 chain;PSMB10:proteasome 20S subunit beta 10;MB21D2:Mab-21 domain containing 2;ZDHHC23:zinc finger DHHC-type palmitoyltransferase 23;MT2A: metallothionein 2A; TFAP2A: transcription factor AP-2 alpha; PARP12: poly(ADP-ribose) polymerase family member 12; HSPB1: heat shock protein family B (small) member 1; HNRNPA2B1: heterogeneous nuclear ribonucleoprotein A2 / B1; ENTPD2: ectonucleoside triphosphate diphosphohydrolase 2; MYLIP: myosin regulatory light chain interacting protein; MTMR7: myotubularin related protein 7; PSMB8: proteasome 20S subunit beta 8 (proteasome 20S subunit beta 8);AUTS2:activator of transcription and developmental regulator AUTS2;UPP1:uridine phosphorylase 1;TAPBP:TAP binding protein;KLRG2:killer cell lectin like receptor G2;PSMB9:proteasome 20S subunit beta 9;MARCKSL1:MARCKS like 1;ID3:inhibitor of DNA binding 3,HLH protein;S100A16: S100 calcium binding protein A16; PLPP3: phospholipid phosphatase 3; GADD45A: growth arrest and DNA damage inducible alpha; S100A4: S100 calcium binding protein A4; DDAH1: dimethylarginine dimethylaminohydrolase 1; MYCL: MYCL proto-oncogene, bHLH transcription factor; CD81: CD81 molecule; SHANK2: SH3 and multiple ankyrin repeat domains; ITIH2: inter-alpha-trypsin inhibitor heavy chain 2;PIK3AP1:phosphoinositide-3-kinase adaptor protein 1;LHFPL6:LHFPL tetraspan subfamily member 6;LGALS3:galectin 3;Multi-Frame-5:FERM domain containing 5;CLDN6:claudin 6;TNFRSF12A:TNF receptor superfamily member 12A;NPC2:NPC intracellular cholesterol transporter 2;CD9:CD9 molecule;ATP11A: ATPase phospholipid transporting 11A; SLC25A21: solute carrier family 25 member 21; CD63: CD63 molecule; B4GALNT3: beta-1,4-N-acetyl-galactosaminyltransferase 3; EMP1: epithelial membrane protein; n 1 (epithelial membrane protein 1);CSTB: cystatin B;WNT10A: Wnt family member 10A;H3-3B: H3.3 histone B;RABAC1: Rab acceptor 1;KCTD17: potassium channel tetramerization domain containing 17;BCAM: basal cell adhesion molecule (Lutheran blood group);CCL15-CCL14: CCL15-CCL14 readthrough (NMD candidate);CCL15: CC motif chemokine ligand 15;CCL23: CC motif chemokine ligand 23 (CC motif chemokine ligand 23); DLG4: discs large MAGUK scaffold protein 4; SPTSSB: serine palmitoyltransferase small subunit B; ANXA5: annexin A5; VAPA: VAMP associated protein A; SOGA1: suppressor of glucose, autophagy associated 1; CST3: cystatin C; MAP1LC3A: microtubule associated protein 1 light chain 3 alpha; MAP9: microtubule associated protein 9; LGALS1: galectin 1;CCDC149: coiled-coil domain containing 149;GNAS: GNAS complex locus;CMBL: carboxymethylenebutenolidase homolog;PTPRN: protein tyrosine phosphatase receptor type N;WTIP: WT1 interacting protein;SPP1: secreted phosphoprotein 1;FXR1: FMR1 autosomal homolog 1;ARHGEF26: Rho guanine nucleotide exchange factor 26;PROS1: protein S;PARP8: poly(ADP-ribose) polymerase family member 8;EIF4A2: eukaryotic translation initiation factor 4A2;OSR1: odd-skipped related transcription factor 1;TFF2: trefoil factor 2;ATF4: activating transcription factor 4;CTSZ: cathepsin Z;UCHL1: ubiquitin C-terminal hydrolase L1;ONECUT2: one cut homeobox 2;EIF1: eukaryotic translation initiation factor 1;LAMP2: lysosomal associated membrane protein 2; CALD1: caldesmon 1; ATP6V1G1: ATPase H+ transporting V1 subunit G1; PRSS35: serine protease 35; KCNK5: potassium two pore domain channel subfamily K member 5; CDKN2B: cyclin dependent kinase inhibitor 2B; AEBP1: AE binding protein 1; SP8: Sp8 transcription factor; CFTR: CF transmembrane conductance regulator; TSPAN7: tetraspanin 7; MPP6: protein associated with LIN7 2, MAGUK family member (LIN7 2-related protein, MAGUK family member);CYSLTR1: cysteinyl leukotriene receptor 1;FSCN1: fascin actin-bundling protein 1;IL33: interleukin 33;PLP2: proteolipid protein 22;ELFN1: extracellular leucine rich repeat and fibronectin type III domain containing 1;IGFBP3: insulin like growth factor binding protein 3;SAT1: spermidine / spermine N1-acetyltransferase 1;AFAP1L1: actin filament associated protein 1 like 1;LPAR4: lysophosphatidic acid receptor 4;ATP6V1F: ATPase H+ transporting V1 subunit F;GRINA: glutamate ionotropic receptor NMDA type subunit associated protein 1;CASD1: CAS1 domain containing 1;HS6ST2: heparan sulfate 6-O-sulfotransferase 2;CD109: CD109 molecule;PGRMC1: progesterone receptor membrane component 1 (progesterone receptor membrane component 1); MAL2: mal, T cell differentiation protein 2; PHF19: PHD finger protein 19; TIMP1: TIMP metallopeptidase inhibitor 1; ASAP1: ArfGAP with SH3 domain, ankyrin repeat and PH domain 1;

[0016] In striking contrast to the lack of association of FMR1 mRNA itself, the FMRP activity signature reveals a statistically significant association with overall and progression-free survival, such that patients with higher FMRP activity have worse overall and progression-free survival. Furthermore, high FMRP activity scores reveal a statistically significant anti-correlation with a CD8 T cell signature that is diagnostic of CTL abundance in human tumors.

[0017] The pan-signature alone is generally sufficient for predicting the prognosis of cancer patients, i.e., overall survival and progression-free survival, and for use in methods for classifying and stratifying such patients, for example, as responders or non-responders to specific cancer therapy.However, when the diagnostic assay based on the pan-signature produces inconclusive results, or additionally / alternatively, further optimization / more accurate results are desired, the present invention further provides three sub-signatures and 28 cancer-specific signatures, as described below.All of these subset signatures contain genes that are up-regulated or down-regulated by FMRP activity, as disclosed in the pan-signature.In particular, using only up-regulated or down-regulated genes as the secondary sub-signature of a particular signature of sub-signature can have utility in itself, as further discussed herein.

[0018] Sub-signature 1 In one embodiment, the present invention provides a "pan-cancer" gene expression signature, referred to herein as sub-signature 1. Sub-signature 1 is a subset of the pan-signature and is based on genes whose expression defines FMRP pathway activity versus inactivity in cancer cells without the influence of stromal and immune cells of the tumor microenvironment (TME). As a result, this signature evaluates the activity of FMRP in cancer cells alone, without the influence of the TME. Sub-signature 1 is disclosed in Table 2.

[0019] [Table 2(1)] [Table 2(2)] "Upregulation indicates that the gene is positively regulated by FMRP activity, and downregulation, conversely, indicates that the gene is negatively regulated by FMRP activity."

[0020] Subsignature 2 In another embodiment, the present invention provides a "pan-cancer" gene expression signature, herein referred to as sub-signature 2. Sub-signature 2 is a subset of the pan-signature and is based only on genes whose expression defines FMRP pathway activity versus inactivity in the tumor. Therefore, this signature evaluates tumor-wide changes, including constitutive accessory (stromal) and immune cells, dictated by FMRP activity in cancer cells, resulting from the effect of intercellular communication between cancer cells and other cell types in the tumor microenvironment (TME). Sub-signature 2 is disclosed in Table 3.

[0021] [Table 3(1)] [Table 3(2)] "Upregulation indicates that the gene is positively regulated by FMRP activity, and downregulation, conversely, indicates that the gene is negatively regulated by FMRP activity."

[0022] Subsignature 3 In another embodiment, the present invention provides a "pan-cancer" gene expression signature, herein referred to as sub-signature 3. Sub-signature 3 is a subset of the pan-signature in which genes corresponding to immune response are excluded. Therefore, this subset can be applied to evaluate FMRP pathway activity without the indirect influence of immune cells in the tumor microenvironment (TME). Sub-signature 3 is disclosed in Table 4.

[0023] [Table 4(1)] [Table 4(2)] "Up-regulated" indicates that the gene is positively regulated by FMRP activity, and "down-regulated," conversely, indicates that the gene is negatively regulated by FMRP activity.

[0024] In addition to the four pan-cancer signatures, the present invention provides exemplary cancer-specific signatures optimized to selectively score FMRP pathway activity in 28 individual cancer types, which are shown in Tables 5-32 below.

[0025] [Table 5(1)] [Table 5(2)] [Table 5(3)] [Table 5(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0026] [Table 6(1)] [Table 6(2)] [Table 6(3)] [Table 6(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0027] [Table 7(1)] [Table 7(2)] [Table 7(3)] [Table 7(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0028] [Table 8(1)] [Table 8(2)] [Table 8(3)] [Table 8(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0029] [Table 9(1)] [Table 9(2)] [Table 9(3)] [Table 9(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0030] [Table 10(1)] [Table 10(2)] [Table 10(3)] [Table 10(4)] "Upregulated" indicates that the gene is positively regulated by FMRP activity, "downregulated" conversely, that the gene is negatively regulated by FMRP activity, and 0 indicates a gene that is not correlated with or not regulated by FMRP activity for this particular cancer type.

[0031] Pan-immunosuppressive signature In one embodiment, the invention provides an independent pan-cancer "FMRP immunosuppression" gene signature, referred to herein as the pan-immunosuppression signature. The pan-immunosuppression signature is based on short-term FMRP knockout in cultured cells and can be used to assess the level of immunosuppression induced by FMRP activity and develop a gene expression signature score that represents the level of CD8 infiltration in tumors at a pan-cancer level, as well as the verity of a particular cancer type.

[0032] The pan-immunosuppressive signature is a global signature list that includes the complete panel of biomarker genes (195 genes total) discovered by comparing FMRP-active cultured cancer cells versus FMRP knockout (by siRNA, and thus inactive) cultured cancer cells. The pan-immunosuppressive signature is disclosed in Table 33.

[0033] [Table 11(1)] [Table 11(2)] [Table 11(3)] [Table 11(4)] [Table 11(5)] * Secretome refers to the set of proteins differentially secreted by cancer cells with high or low FMRP pathway activity, which can be used, for example, as biomarkers in liquid cytology assays and other diagnostic bioassays. "Upregulation indicates that the gene is positively regulated by FMRP activity, and downregulation, conversely, indicates that the gene is negatively regulated by FMRP activity."

[0034] As used herein, MRC1 is mannose receptor C-type 1, KDELR3 is KDEL endoplasmic reticulum protein retention receptor 3, SLC7A1 is solute carrier family 7 member 1, PIK3CD is phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit delta, BCAT1 is branched chain amino acid transaminase 1, JDP2 is Jun dimerization protein 2, ADGRA2 is adhesion G protein-coupled receptor A2, and HMOX1 is heme oxygenase. 1, COBL is cordon-bleu WH2 repeat protein, PSAT1 is phosphoserine aminotransferase 1, CHD5 is chromodomain helicase DNA binding protein 5, CHAC1 is ChaC glutathione specific gamma-glutamylcyclotransferase 1, ATP2A3 is ATPase sarcoplasmic / endoplasmic reticulum Ca2+ transporting 3, and EIF4EBP1 is eukaryotic translation initiationfactor 4E binding protein 1, CA6 is carbonic anhydrase 6, AVIL is advillin, PSPH is phosphoserine phosphatase, HMGA1 is high mobility group AT-hook 1, ATF4 is activating transcription factor 4, SLC1A4 is solute carrier family 1 member 4, CIART is circadian associated repressor of transcription, TRIB3 is tribbles pseudokinase 3, LIMS4 is LIM zinc finger domain containing 4, AREG is amphiregulin, IFRD1 is interferon related developmental regulator 1 (interferon-related developmental regulator 1), SLC7A11 is solute carrier family 7 member 11, ASNS is asparagine synthetase (glutamine-hydrolyzing), ACAT2 is acetyl-CoA acetyltransferase 2, LHFPL2 is LHFPL tetraspan subfamily member 2, EXTL1 is exostosin like glycosyltransferase 1, and FOSL1 is FOS like 1,AP-1CDSN is corneodesmosin, SNAI2 is snail family transcriptional repressor 2, ALDH1L2 is aldehyde dehydrogenase 1 family member L2, SLC7A5 is solute carrier family 7 member 5, TMEM266 is transmembrane protein 266, PCK2 is phosphoenolpyruvate carboxykinase 2, mitochondrial, PHF19 is PHD finger protein 19, FTL is ferritin light chain, and GRAMD2A is GRAM domain containing 2A (GRAM domain containing 2A), CPS1 is carbamoyl-phosphate synthase 1, CAV1 is caveolin 1, UNC13C is unc-13 homolog C, BEND6 is BEN domain containing 6, TIGIT is T cell immunoreceptor with Ig and ITIM domains, YARS1 is tyrosyl-tRNA synthetase 1, LIMS3 is LIM zinc finger domain containing 3, STBD1 is starch binding domain 1, and ZEB2 is zincfinger E-box binding homeobox 2, RAB7B is RAB7B, member RAS oncogene family, DDIT3 is DNA damage inducible transcript 3, CTH is cystathionine gamma-lyase, CARS1 is cysteinyl-tRNA synthetase 1, ILDR2 is immunoglobulin like domain containing receptor 2, ANGPTL6 is angiopoietin like 6, ABHD14A is abhydrolase domain containing 14A, and MTHFD2 is methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 2, methylenetetrahydrofolate cyclohydrolase (methylenetetrahydrofolate dehydrogenase (NADP+-dependent) 2, methenyltetrahydrofolate cyclohydrolase), P2RX3 is purinergic receptor P2X 3, GPR141 is G protein-coupled receptor 141, ATF5 is activating transcription factor 5, ALDH18A1 is aldehyde dehydrogenase 18 family member A1, PYCR1 is pyrroline-5-carboxylate reductase 1, and SNHG12 is small nuclear RNA host gene12, CD68 is CD68 molecule, TMEM50B is transmembrane protein 50B, URAD is ureidoimidazoline (2-oxo-4-hydroxy-4-carboxy-5-) decarboxylase, CST9L is cystatin 9 like, FLRT3 is fibronectin leucine rich transmembrane protein 3, MCF2L is MCF.2 cell line derived transforming sequence like, and FAM3B is FAM3 metabolism regulating signaling molecule. B (FAM3 metabolic regulatory signaling molecule B), SLC2A10 is solute carrier family 2 member 10, OLFM4 is olfactomedin 4, HAO1 is hydroxyacid oxidase 1, IFNGR2 is interferon gamma receptor 2, CYP2C18 is cytochrome P450 family 2 subfamily C member 18, GPD1 is glycerol-3-phosphate dehydrogenase 1, DEPP1 is DEPP1 autophagy regulator, DDC is dopa decarboxylase, and SLC39A9 is solute carrier family 39 member9 (solute carrier family 39 member 9), CYP2D7 is cytochrome P450 family 2 subfamily D member 7 (gene / pseudogene), and MX1 is MX dynamin like GTPa se 1 (MX dynamin-like GTPase 1), AMBP is alpha-1-microglobulin / bikunin precursor, SMIM24 is small integral membrane protein 24, IL13RA2 is interleukin 13 receptor subunit alpha 2, DMKN is dermokine, CLU is clusterin, TFF3 is trefoil factor 3, SLC18A1 is solute carrier family 18 member A1, WDR1 is WD repeat domain 1, TMPRSS6 is transmembrane serine protease 6, and DHRS3 is dehydrogenase / reductase 3, BCL2L14 is BCL2 like 14, LDLRAD3 is low density lipoprotein receptor class A domain containing 3, IGFBP5 is insulin like growth factor binding protein 5, ALDOB is aldolase, fructose-bisphosphate B, FABP1 is fatty acid binding protein 1, SCAMP1 is secretory carrier membrane protein 1, HADHB is hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunitbeta, FAM3D is FAM3 metabolism regulating signaling molecule D, CLCA1 is chloride channel accessory 1, UQCRC2 is ubiquinol-cytochrome c reductase core protein 2, TLR3 is toll like receptor 3, PSCA is prostate stem cell antigen, CLDN2 is claudin 2, PIWIL4 is piwi like RNA-mediated gene silencing 4, ACE2 is angiotensin converting enzyme 2, MUC20 is mucin 20, cell surface SLC44A3 is solute carrier family 44 member 3, FRK is fyn related Src family tyrosine kinase, SPP2 is secreted phosphoprotein 2, DMBT1 is deleted in malignant brain tumors 1, PLA2G10 is phospholipase A2 group X, ATP7A is ATPase copper transporting alpha, and GALNT17 is polypeptide N-acetylgalactosaminyltransferaseASB13 is ankyrin repeat and SOCS box containing 13; KRT7 is keratin 7; ANXA13 is annexin A13; CKMT1B is creatine kinase, mitochondrial 1B; CKMT1A is creatine kinase, mitochondrial 1A; FMR1 is FMRP translational regulator 1; ATP1A3 is ATPase Na+ / K+ transporting subunit alpha 3; SOBP is sine oculis binding protein homolog; NAALADL2 is N-acetylglucosamine transporter; alpha-linked acidic dipeptidase like 2, KCNK16 is potassium two pore domain channel subfamily K member 16, CYP2D6 is cytochrome P450 family 2 subfamily D member 6, EPS8L1 is EPS8 like 1, F5 is coagulation factor V, UGT1A6 is UDP glucuronosyltransferase family 1 member A6, KRT20 is keratin 20, and CDH16 is cadherincadherin 16, PGC is progastricsin, ANO7 is anoctamin 7, USH1C is USH1 protein network component harmonin, TMPRSS4 is transmembrane serine protease 4, UGT1A10 is UDP glucuronosyltransferase family 1 member A10, UGT1A9 is UDP glucuronosyltransferase family 1 member A9, UGT1A8 is UDP glucuronosyltransferase family 1 member A8, and UGT1A7 is UDP glucuronosyltransferase family 1 member A7, CD55 is CD55 molecule (Cromer blood group), IL5RA is interleukin 5 receptor subunit alpha, CXCL17 is CXC motif chemokine ligand 17, GKN2 is gastrokine 2, TMC4 is transmembrane channel like 4, CTSE is cathepsin E, ABCB9 is ATP binding cassette subfamily B member 9, and CYP4B1 is cytochrome P450 family 4 subfamily B member1, SLC9A4 is solute carrier family 9 member A4, CHST4 is carbohydrate sulfotransferase 4, OTOP3 is otopetrin 3, LIPA is lipase A, lysosomal acid type, MUC1 is mucin 1, cell surface associated, CD38 is CD38 molecul, HMGCS2 is 3-hydroxy-3-methylglutaryl-CoA synthase 2, ABCC8 is ATP binding cassette subfamily C member 8, and RBP2 is retinol binding protein 2, GIMAP8 is GTPase, IMAP family member 8, EHF is ETS homologous factor, STAB2 is stabilin 2, TMEM236 is transmembrane protein 236, C2orf72 is chromosome 2 open reading frame 72, ACSM3 is acyl-CoA synthetase medium chain family member 3, SGK1 is serum / glucocorticoid regulated kinase 1, and FXYD3 is FXYD domain containing ion transport regulator3 (FXYD domain-containing ion transport regulator 3), VIL1 is villin 1, ADGRG7 is adhesion G protein-coupled receptor G7, ABCG8 is ATP binding cassette subfamily G member 8, and MUC3A is mucin 3A, cell surface associated s100A14 is S100 calcium binding protein A14; PYURF is PIGY upstream open reading frame; HP is haptoglobin; HPR is haptoglobin-related protein; GPA33 is glycoprotein A33; FOXJ1 is forkhead box J1; AQP1 is aquaporin 1 (Colton blood group); SPTBN2 is spectrin beta,non-erythrocytic 2 (spectrin beta, non-erythrocytic 2), TM4SF20 is transmembrane 4 L six family member 20, CES3 is carboxylesterase 3, KRT23 is keratin 23, PIGR is polymeric immunoglobulin receptor, APOA1 is apolipoprotein A1, SLFN12 is schlafen family member 12, TRPM8 is transient receptor potential cation channel subfamily M member 8, CLCN2 is chloride voltage-gated channel 2, EPHA1 is EPH receptor A1, KIF12 is kinesin family member 12, PDZK1IP1 is PDZK1 interacting protein 1, and PHGR1 is proline,histidine and glycine rich 1, PILRA is paired immunoglobin like type 2 receptor alpha, PZP is PZP alpha-2-macroglobulin like, TTYH1 is tweety family member 1, SYCN is syncollin, SULT1A1 is sulfotransferase family 1A member 1, H19 is H19 imprinted maternally expressed transcript, and MUC4 is mucin 4, cell surface associated.

[0035] As used herein, the pan-signature list, the sub-signature lists (sub-signatures 1, 2, and / or 3), the cancer type-specific lists, and the pan-immunosuppression signature list are individually and collectively referred to herein as "signatures of the invention."

[0036] The present invention relates to the identification and use of gene expression patterns (or profiles or signatures) that are clinically relevant for cancer therapy. In particular, the present invention identifies genes that correlate with the assessment, treatment and monitoring of patients for cancer therapy.

[0037] The identified gene biomarkers embodied in the pan-signature list, sub-signature list, cancer type-specific list, and pan-immunosuppression list constituting the present invention do not involve or require evaluation of FMR1 mRNA or FMRP protein expression, but rather independently predict the level of signaling activity downstream of FMRP expression, with high levels of pathway activity in tumors predicting the ability to suppress tumor immunity and / or stimulate invasion and metastasis. The above signatures can be the basis of a multi-biomarker assay to stratify cancer patients based on FMRP activity to predict prognosis and inform treatment selection, and thus can serve as a "companion diagnostic" for cancer therapy.

[0038] As used herein, a companion diagnostic refers to a diagnostic method and / or reagent used to identify patients susceptible to treatment with a particular treatment, or to monitor treatment, and / or to identify effective dosages for patients or subgroups or other groups of patients. Companion diagnostic refers to the reagent and also to tests performed with the reagent.

[0039] As used herein, "patient," "subject," and "individual" are used interchangeably and refer to a human subject having or exhibiting symptoms of cancer.

[0040] In an embodiment, the invention provides a method of identifying a patient having cancer as having high or low FMRP activity, having a high or low risk prognosis, and / or being a responder or non-responder to a cancer therapy. In an embodiment, the method comprises obtaining a sample from the patient, determining in the sample the expression levels of genes in one or more of the signatures set forth in Tables 1-33, comparing the expression levels in the sample to the levels of the genes expressed in a control, identifying gene(s) that are differentially expressed between the sample and the control, and classifying the patient as having (i) high or low FMRP activity, (ii) having a high or low risk prognosis, and / or (iii) being a responder or non-responder to a cancer therapy, and / or (iv) having a high or low immune cell infiltration tumor based on the match of differential expression with the one or more signatures.

[0041] As used in any of the embodiments herein, the term "control" or the like refers to one or more samples with known FMRP activity status and / or clinical information. Thus, FMRP activity in a patient sample (query sample) is determined in comparison to this control, and thus the clinical outcome (prognosis or response to cancer therapy) is predicted. The control may be of the same or different composition as the patient sample, including, but not limited to, one or more tumor samples from the same cancer type with known prognosis and / or response to some form of therapy; or a cohort of samples from a publicly available dataset (e.g., TCGA) that profiles tumor samples with various FMRP activities; in addition, a cohort of related normal samples can serve as a control cohort, depending on the tissue and the activity of FMRP in normal cells, in some cases. For example, if the patient has breast cancer, the control may be a set of previously analyzed tumor samples from a cohort of breast cancer patients, where some breast cancer patients have high FMRP activity scores and others have low FMRP activity scores, potentially supplemented with additional clinical or pathological information. This cohort can be used as a reference set to establish high vs. low FMRP activity scores for new tumors referred and specific prognostic / therapeutic questions addressed. Alternatively, for example, the TCGA cohort of breast cancer tumors can be separated into groups with high, neutral, or low FMRP activity scores and used as a reference to classify tumors referred for FMRP activity.

[0042] For FMRP activity signature to have predictive power, at least one, or at least two, or at least ten genes from the general signature list and / or from sub-signature, or from its cancer type-specific signature list, must be differentially expressed between patient samples and controls.If this criterion is met, the query sample is classified as follows: if the majority of differentially expressed genes follow the expected up / down-regulated call in the signature list, that is, if the differentially up-regulated genes in the sample are in the signature list of up-regulated genes, and the differentially down-regulated genes in the sample are also in the signature list of down-regulated genes, then the query sample has higher FMRP activity compared to the control. Conversely, if the majority of differentially expressed genes show the opposite pattern in the signature list, i.e., the differentially upregulated genes in the sample are ostensibly part of the signature list of downregulated genes, and the differentially downregulated genes in the sample are derived from the signature list of upregulated genes, then the query sample is determined to have lower FMRP activity compared to the control. As used herein, the phrase "majority of differentially expressed genes" generally means that 2 / 3 of the differentially expressed genes in the sample either follow or do not follow the regulated call (i.e., upregulated / downregulated) in the signature list.

[0043] As shown in the figures accompanying this specification, for other patients with the same cancer type or subtype, A lower FMRP activity signature score was associated with a better prognosis. a low FMRP activity signature score is associated with the patient being a relatively good responder to treatment with a checkpoint inhibitor, targeted cancer therapy, chemotherapy, or radiation, but a non-responder or a poor responder to treatment with an FMRP inhibitor; a high FMRP activity signature score is associated with the patient being a relatively good responder to treatment with an FMRP inhibitor, but a non-responder or a relatively poor responder to treatment with a checkpoint inhibitor, targeted cancer therapy, chemotherapy, or radiation unless combined with an FMRP inhibitor; A low pan-immunosuppression signature score is associated with higher tumor inflammation driven by T cells.

[0044] As used in any of the methods described herein, the term "differentially expressed" or "altered expression" is used interchangeably to refer to the difference in the expression level of an RNA of a biomarker of the invention as measured by the amount or level of the mRNA of the biomarker and / or one or more splice variants of the mRNA of the biomarker in one sample compared to the expression level of the same biomarker of the invention in a second sample. "Differentially expressed" or "altered expression" or "change in expression" can also include the measurement of the protein encoded by the biomarker of the invention in one sample or sample population compared to the amount or level of protein expression in a second sample or sample population. Differential expression can be determined as described herein and as understood by one of skill in the art. A gene or protein is upregulated or downregulated in a cancer patient compared to a control. A gene is considered to be either upregulated or downregulated if its expression in a patient sample is increased or decreased by at least 1.5-fold compared to its expression level in the corresponding control. For purposes herein, the change in expression of a gene is the result of FMRP functional activity in the tumor.

[0045] As used in any of the embodiments herein, phrases such as "relative to the level of a gene expressed in a control" refer to the expression level of a gene of the present invention in a control sample, depending on each particular study described herein.

[0046] In an embodiment, the present invention provides a method of identifying a patient having cancer as eligible for cancer therapy, which comprises obtaining a sample from a patient, determining the expression levels in said sample of genes in one or more of the signatures set forth in Tables 1-33, comparing the expression levels to the levels of said genes expressed in a matched control, and identifying said patient as eligible to receive cancer therapy based on a match of differential expression with said signature.

[0047] In an embodiment, the invention provides a method of identifying a patient having cancer as a responder to a cancer therapy, the method comprising obtaining a sample from the patient, determining the expression levels of genes in one or more of the signatures set forth in Tables 1-33 in the sample, comparing the expression levels to the levels of the genes expressed in a matched control, and identifying the patient as a responder to a cancer therapy based on a match of differential expression with the signature.

[0048] In an embodiment, the invention provides a method of treating a patient with cancer, comprising obtaining a sample from the patient, determining the expression levels of genes in one or more of the signatures set forth in Tables 1-33 in the sample, comparing the expression levels in the sample to the levels of the genes expressed in a control, identifying genes that are differentially expressed between the sample and the control, classifying the patient as having (i) high or low FMRP activity, (ii) a high or low risk prognosis, and / or (iii) a responder or non-responder to a cancer therapy, and / or (iv) a high or low immune cell infiltration tumor based on the match of differential expression with the signature, and administering a cancer therapy to the patient.

[0049] In any of the embodiments herein, the method includes determining the expression levels of genes in one signature set forth in Tables 1-33. In any of the embodiments herein, the method includes determining the expression levels of genes in two or more signatures set forth in Tables 1-33.

[0050] In any of the embodiments herein, the method includes determining the expression level of each gene in a global signature set forth in Table 1. In any of the embodiments herein, the method includes determining the expression level of genes in one or more of the signatures set forth in Tables 1-33.

[0051] In any of the embodiments herein, the method comprises determining the expression level of each gene in one or more sub-signatures and / or cancer type-specific signatures set forth in Tables 2-33 in a tissue sample and comparing these expression levels to the levels of said genes expressed in a control. In any of the embodiments herein, the method comprises determining the expression level of genes in one or more sub-signatures set forth in Tables 2-4. In any of the embodiments herein, the method comprises determining the expression level of genes in one or more cancer specific signatures set forth in Tables 5-32.

[0052] In any of the embodiments herein, the method comprises determining the expression levels of genes in a pan-immunosuppression signature set forth in Table 33.

[0053] The present invention also provides a method for developing a signature score as a biomarker of FMRP activity in a group of patients with cancer. In some embodiments, when there is a particular set of samples from cancer patients analyzed without a separate reference set, for example, from a group that contains a unique histological or molecular subtype of a particular cancer type or has a variety of responses to a particular therapy (tumor size, PSF, OS), a signature score can be derived for each sample relative to all other samples in the group.

[0054] In an embodiment, the present invention provides a method for stratifying a group of patients having cancer as having (i) high or low FMRP activity, (ii) high or low risk prognosis, and / or (iii) responder or non-responder to cancer therapy, and / or (iv) high or low immune cell infiltration tumors. In an embodiment, the method comprises obtaining a sample from each patient of the group, determining for each sample the expression level of genes in one or more of the signatures set forth in Tables 1-33, establishing an FMRP activity score for each sample, and identifying each patient as having (i) high or low FMRP activity, (ii) high or low risk prognosis, and / or (iii) responder or non-responder to cancer therapy, and / or (iv) high or low immune cell infiltration tumors based on said FMRP activity score.

[0055] In an embodiment, the invention provides a method for stratifying a group of patients having cancer as eligible for cancer therapy, the method comprising obtaining a sample from each patient of the group, determining for each sample the expression level of genes in one or more of the signatures set forth in Tables 1-33, establishing an FMRP activity score for each sample, and identifying the patients as eligible to receive cancer therapy based on the FMRP activity score.

[0056] In an embodiment, the invention provides a method for stratifying a group of patients having cancer into responders to cancer therapy, the method comprising obtaining a sample from each patient of the group, determining for each sample the expression level of genes in one or more of the signatures set forth in Tables 1-33, establishing an FMRP activity score for each sample, and identifying the patients as eligible to receive cancer therapy based on the FMRP activity score.

[0057] In an embodiment, the invention provides a method of treating a group of patients with cancer, comprising obtaining a sample from each patient of said group, determining for each sample the expression level of genes in one or more of the signatures set forth in Tables 1-33, establishing an FMRP activity score for each sample, identifying each patient as having (i) high or low FMRP activity, (ii) a high or low risk prognosis, and / or (iii) a responder or non-responder to a cancer therapy, and / or (iv) a high or low immune cell infiltration tumor based on said FMRP activity score, and administering a cancer therapy to each patient.

[0058] In an embodiment, the present invention provides a method for predicting T cell infiltration in a cancer patient, comprising obtaining a sample from the patient, determining the expression levels of the genes listed in Table 33 in the sample, comparing the expression levels in step (b) to the levels of the genes expressed in a control, identifying differentially expressed gene(s) between the sample and the control, and classifying the patient as having (i) high or low FMRP immunosuppressive activity, (ii) high or low immune cell infiltration based on the match of differential expression with the signature.

[0059] As used in this embodiment, the term "signature score", also referred to herein as "FMRP activity signature score", generally refers to a quantitative score that predicts whether a patient will benefit from currently available cancer therapies whose effectiveness is limited by or otherwise dependent on FMRP activity, or potentially regulated by FMRP. The signature score is calculated by summing up the z-scores, e.g., the number of standard deviations that expression is above or below the average expression of the gene in all samples, of genes in a particular FMRP activity signature list (e.g., pan-signature and / or sub-signature and / or its cancer-specific signature and / or its pan-immunosuppression signature). For downregulated genes in a signature, the z-scores are multiplied by minus one (-1) and then summed to derive the final signature score.

[0060] In any of the methods described herein, according to the signature score, cancer patients with a low FMRP activity score are predicted to have a better prognosis and a better response to cancer therapy compared to cancer patients with a high FMRP activity score. Cancer patients with a high FMRP activity score are predicted to have a better response to treatment with an FMRP inhibitor.

[0061] In some embodiments, the predictive power of the FMRP activity signature score in such a group can be confirmed if at least one, or at least two, or at least ten genes from the signature list are differentially expressed between the top 50% of samples in terms of the signature score (samples with a signature score higher than the median) and the bottom 50% of samples in terms of the signature score (samples with a signature score lower than the median). If this criterion is met, samples with a low FMRP activity signature score (samples with a signature score lower than the median or the first quartile) have a better prognosis or better response to cancer therapy, while samples with a high FMRP activity signature score (samples with a signature score higher than the median or the first quartile) have a worse prognosis or poor / no response to cancer therapy, or potentially a better response to treatment with an FMRP inhibitor.

[0062] The present invention provides a companion diagnostic assay for the classification of patients for cancer treatment, comprising assessing the expression levels of the genes set forth in Tables 1-33 or combinations thereof in a patient tissue sample. The assays of the present invention include assay methods for identifying patients eligible to receive cancer therapy and for monitoring the patient's response to such therapy. The methods of the present invention include the assessment of expression of the above genes in blood, urine or other body fluid samples by immunoassays, proteomic assays or nucleic acid hybridization or amplification or sequencing assays, and in tissue or other cell body samples by immunohistochemistry or in situ hybridization assays.

[0063] The gene expression patterns of the present invention, also referred to as "gene expression patterns" or "gene expression profiles" or "gene signatures", are identified as described herein. Generally, the gene expression profile of a sample is obtained by quantifying the expression levels of mRNA corresponding to a number of genes identified in the signature list of Tables 1-33. The signatures are then analyzed to identify genes whose expression positively correlates with the identification and monitoring of patients eligible for cancer treatment.

[0064] In any of the embodiments herein, a gene signature represents a combined pattern of the results of an analysis of the expression levels of one or more genes of a signature of the invention. In an embodiment, a gene signature represents a combined pattern of the results of an analysis of the expression levels of one or more genes of a signature of the invention. , 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, or all of the expression levels.

[0065] In any of the embodiments herein, a gene signature represents a combined pattern of the results of an analysis of the expression levels of one or more genes of one or more signatures of the invention. In an embodiment, a gene signature represents a combination of 1, 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, 109, 108, 109, 109, 110 9, 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, or all of the expression levels of these genes.

[0066] In any of the embodiments herein, the gene signature represents a combined pattern of the results of an analysis of the expression levels of 10 or more genes of the signature of the invention. In an embodiment, the gene signature represents a combined pattern of the results of an analysis of the expression levels of 10 or more genes of the signature of the invention. 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, 1 27, 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, or all of the expression levels.

[0067] In any of the embodiments herein, the gene signature represents a combined pattern of the results of the analysis of the expression levels of 10 or more genes of one or more signatures of the invention. In an embodiment, the gene signature represents a combined pattern of the results of the analysis of the expression levels of 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, 109, 108, 110, 111, 121, 131, 141, 14 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, or all of the expression levels.

[0068] In any embodiment of the present invention, the minimum number of genes (biomarkers) required for use of the signature list of the present invention to improve patient identification or stratification is at least one (1) of the biomarkers of the signature of the present invention. In any embodiment of the present invention, the minimum number of genes required for use of the signature list of the present invention is at least two (2) of the biomarkers of the signature of the present invention. In any embodiment of the present invention, the minimum number of genes required for use of the signature list of the present invention is at least ten (10) of the biomarkers of the signature of the present invention. In any embodiment of the present invention, the minimum number of genes (biomarkers) required for use of the signature list of the present invention is at least 1, 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, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 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, 7 8, 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 1, 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 types (pieces), or all.

[0069] In any embodiment of the invention, the minimum number of genes (biomarkers) required for the use of the biomarkers of the signature list of the invention to improve patient identification or stratification is at least 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, 109, 109, 102, 104, 105, 106, 107, 108, 109, 109, 102, , 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, or all of these.

[0070] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least one of the biomarkers in Table 1, Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32, or a combination thereof. In another embodiment, the biomarkers for improved patient identification or stratification include at least one of the biomarkers in Table 1 and / or Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32. In another embodiment, biomarkers for improving patient identification or stratification include at least one of the biomarkers in Table 1, and one or more of Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, and / or Table 32.

[0071] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least one of the biomarkers in Table 33.

[0072] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least two of the biomarkers in Table 1, Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32, or a combination thereof. In another embodiment, the biomarkers for improved patient identification or stratification include at least two of the biomarkers in Table 1 and / or Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 1, and one or more of Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, and / or Table 32.

[0073] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least two of the biomarkers in Table 33.

[0074] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least ten of the biomarkers in Table 1, Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32, or a combination thereof. In another embodiment, the biomarkers for improved patient identification or stratification include at least ten of the biomarkers in Table 1 and / or Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, Table 32. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 1, and one or more of Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, Table 13, Table 14, Table 15, Table 16, Table 17, Table 18, Table 19, Table 20, Table 21, Table 22, Table 23, Table 24, Table 25, Table 26, Table 27, Table 28, Table 29, Table 30, Table 31, and / or Table 32.

[0075] In any embodiment of the invention, the biomarkers for improved patient identification or stratification include at least 10 of the biomarkers in Table 33.

[0076] In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 1. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 2. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 3. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 4. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 5. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 6. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 7. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 8. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 9. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 10. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 11. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 12. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 13. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 14. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 15. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 16.In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 17. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 18. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 19. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 20. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 21. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 22. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 23. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 24. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 25. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 26. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 27. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 28. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 29. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 30. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 31. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 32.

[0077] In another embodiment, the biomarkers for improving patient identification or stratification comprise at least one of the biomarkers in Table 33.

[0078] In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 1. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 2. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 3. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 4. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 5. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 6. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 7. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 8. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 9. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 10. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 11. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 12. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 13. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 14. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 15. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 16.In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 17. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 18. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 19. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 20. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 21. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 22. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 23. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 24. In another embodiment, the biomarkers for improving patient identification or stratification include at least two of the biomarkers in Table 25. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 26. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 27. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 28. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 29. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 30. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 31. In another embodiment, the biomarkers for improving patient identification or stratification comprise at least two of the biomarkers in Table 32.

[0079] In another embodiment, the biomarkers for improved patient identification or stratification include at least two of the biomarkers in Table 33.

[0080] In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 1. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 2. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 3. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 4. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 5. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 6. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 7. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 8. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 9. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 10. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 11. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 12. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 13. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 14. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 15. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 16.In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 17. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 18. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 19. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 20. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 21. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 22. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 23. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 24. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 25. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 26. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 27. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 28. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 29. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 30. In another embodiment, the biomarkers for improving patient identification or stratification include at least ten of the biomarkers in Table 31.In another embodiment, the biomarkers for improving patient identification or stratification include at least 10 of the biomarkers in Table 32.

[0081] In another embodiment, the biomarkers for improving patient identification or stratification include at least 10 of the biomarkers in Table 33.

[0082] A gene signature can result from measuring the expression of RNA and / or protein expressed by genes corresponding to the biomarkers of Table 1 and / or Tables 2-32 of the present invention. In the case of RNA, it refers to the RNA transcripts transcribed from the genes corresponding to the biomarkers of the present invention. In the case of protein, it refers to the protein translated from the genes corresponding to the biomarkers of the present invention. For example, techniques for measuring the expression of RNA products of the biomarkers of the present invention include PCR-based methods (including RT-PCR) and non-PCR-based methods as well as microarray analysis. For measuring the protein products of the biomarkers of the present invention, techniques include Western blotting and ELISA analysis, as well as proteomic profiling (e.g., mass spectrometry, Imaging Mass Cytometry (histo-CyTOF, etc.).

[0083] The assays of the present invention include both assays for selecting patients eligible to receive cancer therapy and assays for monitoring patient response. These assays can be performed by protein assays and nucleic acid assays. Any type of protein or nucleic acid assay can be used. Protein assays useful in the present invention are well known in the art and include (i) immunoassays involving binding of labeled antibodies or proteins to expressed proteins or fragments thereof, (ii) mass spectrometry to determine the expressed proteins or fragments thereof of these biomarkers, and (iii) proteomics-based assays or "protein chip" assays. Useful immunoassays include both liquid phase assays performed using any format known in the art, such as, but not limited to, ELISA format, sandwich format, competitive inhibition format (including both forward or reverse competitive inhibition assays) or fluorescence polarization format, and solid phase assays such as immunohistochemistry (referred to as "IHC").

[0084] IHC is a method to detect the presence of specific proteins in cells or tissues and consists of the following steps: 1) preparing a slide with the tissue to be examined; 2) a primary antibody is added to the slide and binds to the specific antigen; 3) the resulting antibody-antigen complex is bound by a secondary enzyme-linked antibody; 4) in the presence of a substrate and chromogen, the enzyme forms a colored deposit ("stain") at the antibody-antigen binding site; 5) the slide is examined under a microscope to identify the presence and level of stain.

[0085] Nucleic acid assay methods useful in the present invention are also well known in the art and include (i) in situ hybridization assays on intact tissue or cell samples to detect mRNA levels or chromosomal DNA alterations, (ii) microarray hybridization assays to detect mRNA levels or chromosomal DNA alterations, (iii) RT-PCR or other amplification assays to detect mRNA levels, or (iv) PCR or other amplification assays to detect chromosomal DNA alterations. Assays that use synthetic analogs of nucleic acids, such as peptide nucleic acids, in any of these formats can also be used.

[0086] The present invention provides methods for identifying altered expression levels of genes in the pan-signature (Table 1) or a subset thereof for both response prediction and monitoring of patient response to cancer therapy. Assays for response prediction are performed prior to therapy selection, and samples determined to have at least one, or at least two, or at least ten differentially expressed genes from the pan-signature and / or sub-signature and / or cancer-specific signature list, as the case may be, compared to a control as defined herein, and classified as having a high or low FMRP activity score, will be eligible to receive the particular cancer therapy determined to be differentially responsive as a function of FMRP activity.

[0087] To monitor the response of a patient to an FMRP inhibitor, the assay can be carried out at the beginning of therapy to establish the FMRP activity score and the baseline level of the gene in tissue samples.The same tissue is then sampled and assayed, and the level of the gene is compared to the baseline.If the level remains the same or decreases, therapy is likely to be effective and can be continued.If a significant increase occurs above the baseline level, the patient may not be responding.

[0088] As used herein, cancer therapy includes, but is not limited to, treatment with one or more inhibitors of FMRP protein expression or activity, treatment with one or more immune checkpoint inhibitors, chemotherapy treatment, radiation, targeted cancer therapy, or a combination thereof. In an embodiment, cancer therapy includes, but is not limited to, treatment with an inhibitor of FMRP protein expression or activity, treatment with an immune checkpoint inhibitor, chemotherapy treatment, or a combination thereof. In an embodiment, cancer therapy is treatment with an inhibitor of FMRP protein expression or activity. In an embodiment, cancer therapy is treatment with an immune checkpoint inhibitor. In an embodiment, cancer therapy is chemotherapy treatment.

[0089] As used herein, when referring to therapeutic treatment, the term "in combination" refers to the use of more than one type of therapy. The use of the term "in combination" does not restrict the order in which the therapies are administered to a subject. Such combinations may also include more than a single administration of the therapy. The administration of the therapies may be by the same or different routes. One or more therapies may be administered simultaneously. The term "co-administered" or "co-administration" generally refers to the administration of at least two different substances sufficiently close in time. Co-administration refers to the administration of at least two different substances at the same time, in any order, either in a single dose or in separate doses, and in a sequence spaced in time by up to several days.

[0090] Checkpoint inhibitors include, but are not limited to, anti-PD1, anti-PDL1 and anti-CTLA inhibitors (antibodies). In embodiments, the checkpoint inhibitor is an anti-CTLA-4 antagonist antibody, such as ipilimumab, tremelimumab, and BMS-986249. In embodiments, the checkpoint inhibitor is an anti-PD-1 or anti-PD-L1 antagonist antibody, such as avelumab, atezolizumab, CX-072, pembrolizumab, nivolumab, cemiplimab, spartalizumab, tislelizumab, JNJ-63723283, genolimzumab, AMP-514, AGEN2034, durvalumab, and JNC-1.

[0091] Chemotherapeutic agents include afatinib, capecitabine, carboplatin, cisplatin, cobimetanib, crizotinib, cyclophosphamide, dabrafenib, dacarbazine, dexamethasone, docetaxel, doxorubicin, daunorubicin, epirubicin, eribulin, erlotinib, etoposide, fludarabine, 5-FU, gemcitabine, gefitinib, irinotecan, ixabepilone, CHOP (C: CYTOXAN® (cyclophosphamide); H: A These include, but are not limited to, DIAMYCIN® (hydroxydoxorubicin; O: vincristine (ONCOVIN®); P: prednisone), methotrexate, mitoxantrone, oxaliplatin, paclitaxel, nanoparticle albumin-bound paclitaxel (nab-paclitaxel), pemetrexed, rapamycin, RITUXIN® (rituximab), temozolomide, trametinib, vemurafenib, vinorelbine, and vincristine.

[0092] Targeted therapies include, but are not limited to, EGFR, ALK, ROS, RAS, BRAF, or BCL2.

[0093] In any of the embodiments herein, where the cancer therapy is an FMRP inhibitor, tumors with high FMRP activity scores can be selected. In any of the embodiments herein, where the cancer therapy is an immune checkpoint inhibitor and / or chemotherapy, patients with low FMRP activity scores can be selected, unless those therapies are combined with an FMRP inhibitor.

[0094] In embodiments, cancers include, but are not limited to, AML (acute myeloid leukemia), BRCA (breast cancer), CCC (cholangiocarcinoma), CLL (chronic lymphocytic leukemia), CRC (colorectal cancer), GBC (gallbladder cancer), GBM (glioblastoma), GC (gastric cancer), GEJC (gastroesophageal junction cancer), HCC (hepatocellular carcinoma), HNSCC (head and neck squamous cell carcinoma), MEL (melanoma), NHL (non-Hodgkin's lymphoma), NSCLC (non-small cell lung cancer), OC (ovarian cancer), OSCAR (esophageal cancer), PACA (pancreatic cancer), PRCA (prostate cancer), RCC (renal cell carcinoma), SCLC (small cell lung cancer), UBC (bladder cancer), and UEC (endometrial cancer). In embodiments, the cancers include, but are not limited to, gastric cancer, breast cancer, optionally triple-negative breast cancer (TNBC), non-small cell lung cancer (NSCLC), melanoma, renal cell carcinoma (RCC), bladder cancer, endometrial cancer, diffuse large B-cell lymphoma (DLBCL), Hodgkin's lymphoma, ovarian cancer, and head and neck squamous cell carcinoma (HNSCC).

[0095] In any of the embodiments herein, the biomarker and signature lists of the present invention are useful for cancer in general, and for adrenocortical carcinoma, bladder cancer, breast cancer, cervical cancer, colon adenocarcinoma, esophageal cancer, glioblastoma, head and neck cancer, chromophobe renal cell carcinoma, clear cell renal carcinoma, papillary renal cell carcinoma, acute myeloid leukemia, glioma, hepatocellular carcinoma, lung adenocarcinoma, lung squamous cell carcinoma, ovarian cancer, pancreatic adenocarcinoma, pheochromocytoma and paraganglioma, prostate adenocarcinoma, rectal adenocarcinoma, sarcoma, melanoma, gastric adenocarcinoma, testicular tumor, thyroid cancer, thymoma, or endometrial cancer in particular.

[0096] The present invention includes diagnostic assays performed on any type of patient sample (also referred to as a "sample", "tissue sample", or "query sample") or derivative thereof, including peripheral blood, tumor or suspected tumor tissue (including fresh frozen and fixed or paraffin embedded tissue), cell isolates such as circulating epithelial cells separated or identified in a blood sample, lymph node tissue, bone marrow and fine needle aspirates. Preferred samples for use herein are peripheral blood, tumor or suspected tumor tissue and bone marrow.

[0097] Furthermore, the present invention provides cell-based assays comprising cancer cells expressing high levels of FMRP protein and that FMRP protein's gene signature of pathway activity for use in identifying and / or validating inhibitors of said FMRP activity. Such activity inhibition assays can be powerful tools when applied in screening efforts aimed at discovering and developing pharmaceuticals targeting FMRP and / or the FMRP immunosuppressive and pro-invasive / pro-metastatic pathways. For diagnostic applications, such cell-based assays could use mRNA or protein representing the signature genes. EXAMPLES

[0098] The present invention was developed using mouse cancer cell lines and tumors that express FMRP differently or lack expression of FMRP due to genetic disruption of the FMR1 gene. Importantly, the identified biomarkers and methods for developing signature scores reporting FMRP pathway activity are demonstrably applicable across multiple human cancer types and can be used to predict the prognosis of cancer patients with a variety of tumor types.

[0099] The present invention demonstrates that the FMRP activity signature score can predict which patients will benefit from FMRP inhibitor therapy. The present invention represents a companion diagnostic for "precision medicine" strategies that reveal the extent of FMRP pathway activity and putative immunosuppressive capacity to more accurately select patients most likely to respond to potential inhibitors of FMRP.

[0100] In addition, the present invention demonstrates that the FMRP activity signature score can predict which patients will benefit from immune checkpoint inhibitor therapy. Thus, the identified biomarkers and corresponding methods can be used together and / or in addition to current biomarkers to classify patients for treatment with or without immunotherapy. The FMRP activity score may also be applicable to clinical decisions for treating cancer patients with other treatment modalities that involve adaptive immune responses, as exemplified for chemotherapy.

[0101] Example 1 – Development of an FMRP activity signature The present invention is based on two separate experiments applying state-of-the-art gene knockout systems performed both in vitro (cell cultures) and in vivo (tumor-bearing mice). Gene expression levels were measured using bulk and single-cell RNA sequencing techniques, as well as advanced bioinformatics analysis and corresponding methods to establish gene lists to develop a signature score representing FMRP pathway activity in cancer cells.

[0102] FMR1 (the gene encoding the FMRP protein) was genetically deleted in mouse pancreatic cancer cell lines by using the CRISPR-Cas9 system to target deletion of the essential first exon in the FMR1 gene. In the first model, cancer cells in culture were subjected to RNA sequencing analysis to identify differentially expressed genes (fold change >1.5) by comparing an isogenic cell line in which the FMR1 gene was intact and its gene product FMRP was expressed (FMRP-WT) with a derivative in which FMR1 was deleted and FMRP was not expressed (FMRP-KO). This list of significantly differentially expressed genes defines a "signature" consisting of genes that FMRP regulates, directly or indirectly, in cancer cells that express it. We term this gene set FMRP activity "subsignature 1". In the second model, FMRP-WT and FMRP-KO cancer cells were inoculated (subcutaneously) into immune-competent mice and allowed to form solid tumors. Tumors were resected and subjected to single-cell RNA sequencing analysis, followed by identifying genes that were differentially expressed (fold change >1.5) between FMRP-WT and FMRP-KO tumors to define a second set of genes, named FMRP activity "sub-signature 2". The union of these two differentially expressed gene lists constitutes and defines the FMRP activity "pan-signature". In addition, genes reflecting an indirect innate immune response in tumors, annotated from the Gene-Ontology signature list, were excluded from the pan-signature. The remaining genes define FMRP activity "sub-signature 3". In addition, derivative cancer type-specific signatures were developed by subselecting genes from the pan-signature, including only genes that collectively showed significant correlation (hazard ratio >1.2) with overall survival and progression-free survival in the TCGA cohort of a particular cancer type, using the COX model.

[0103] Example 2 Tumor samples from TCGA were classified based on the quartiles of the signature score after inferring the signature score: FMRP-low (samples with score < Q1), FMRP-intermediate (samples with score between Q1 and Q3), FMRP-high (samples with score > Q3). Kaplan-Meier survival analysis was used to evaluate the relationship between the signature score and survival rate. The COX model was used to determine the association between the predictor variables and obtain the adjusted hazard ratio. The tumor type was included as a covariate in the COX model.

[0104] Figure 1 shows patient classifications across 31 different cancer types. While classification based on FMR1 mRNA expression (Panels A and B) is uninformative, in contrast, the newly invented FMRP pathway activity signature scores (FMRP activity pan-signature: Panels C and D; sub-signature 1: Panels E and F; sub-signature 3: Panels G and H) are informative and statistically significant for all. Each panel shows the association (or lack thereof) with patient prognosis (A, C, E, and G: overall survival; B, D, F, and H: progression-free survival). To estimate the significance of the correlation, a COX model was used considering the tumor type as a covariate. The data used in this figure were downloaded from the latest TCGA PanCan Atlas.

[0105] Example 3 The application of the classification method according to the present invention is applied to two different cancer types in which FMRP is involved, namely breast cancer and colorectal cancer. The use of the present invention for predicting patient responses to immune checkpoint inhibitors and chemotherapy in several cancer types is demonstrated. For these analyses, the pan-signature was used unless otherwise specifically mentioned in the legend.

[0106] The FMRP signature scores for each tumor sample were developed as described above. For survival analysis, as in Figure 1 above, the samples were classified based on the signature scores (as shown in the figure, for Figures 2 / 3 / 5, low score < Q1 and high score > Q3; for Figure 4, low score < Q2 and high score > Q2). For box plots (correlation analysis), the signature scores in each subtype were compared and tested for significant differences using the Wilcoxon test. The subtypes used in each figure are as follows; Subtypes for Figure 2: Breast cancer PAM50 subtypes; Subtypes for Figure 4: Responders and non-responders; Subtypes for Figure 5: Tumor T stage.

[0107] Figure 2: FMRP activity score in breast cancer. Figure 2A. The FMRP activity score shows the highest levels in the basal-like subtype, which is the most aggressive subtype of breast cancer. The signature score for this panel was derived using only the upregulated genes in the pan-signature. Figure 2B. The FMRP activity score correlates with the overall survival of all breast cancer patients. Figure 2C. The FMRP activity score specifically correlates with the overall survival of the luminal A subtype of breast cancer patients. The data used in this figure were downloaded from the latest breast cancer cohort of the TCGA PanCan Atlas.

[0108] Figure 3 shows the FMRP activity score in colorectal cancer. Figure 3A. The FMRP activity score correlates with the overall survival of all colorectal cancer patients. Figure 3B. The FMRP activity score correlates with the overall survival of microsatellite stable (MSS) colorectal cancer patients. Figure 3C. Shows the lack of correlation between the overall survival of microsatellite instability (MSI) colorectal cancer patients and the FMRP activity score. The data used in this figure were downloaded from the latest colorectal cancer cohort of the TCGA PanCan Atlas.

[0109] Figure 4 shows FMRP activity score correlation with immune checkpoint inhibitor therapy response in cancer patients. Figure 4A. FMRP activity score correlation with overall survival of melanoma patients receiving anti-PD1 therapy (left panel); non-responders to anti-PD1 therapy show higher levels of FMRP activity score (right panel). Figure 4B. FMRP activity score correlation with overall survival of lung cancer patients receiving anti-PD1 or anti-PD-L1 therapy (left panel); non-responders to anti-PD1 or anti-PD- therapy show higher levels of FMRP activity score (right panel). Figure 4C. FMRP activity score correlation with overall survival of urothelial carcinoma patients receiving anti-PD-L1 therapy (left panel); non-responders to anti-PD-L1 therapy show higher levels of FMRP activity score (right panel). Only upregulated genes in sub-signature 1 were used to derive the signature scores in panels A-C. Figure 4D. FMRP activity score (pan-signature) correlation with overall survival in melanoma patients receiving anti-CTLA4 therapy (left panel); non-responders to anti-CTLA4 therapy show higher levels of FMRP activity score (right panel).

[0110] Figure 5 shows FMRP activity score correlation with chemotherapy response in cancer patients. Figure 5A. FMRP activity score correlation with disease-free survival in breast cancer patients receiving taxanes (left panel); notably, the signature score is independent of tumor aggressiveness (T stage, right panel) and is also shown using a COX model in a survival analysis considering T stage as a covariate, thus revealing that the FMRP activity signature constitutes an independent prognostic marker. Figure 5B. FMRP activity score correlation with progression-free survival in lung cancer patients receiving paclitaxel, cisplatin, or carboplatin (left panel); the signature score is again independent of tumor aggressiveness (T stage, right panel) and constitutes an independent prognostic factor. To estimate the significance of the correlation for survival analysis, a COX model was used, considering T stage as a covariate. To derive the signature scores for all panels, only upregulated genes from subsignature 1 were used.

[0111] Example 4 Figure 6 shows the non-reproducibility and lack of correlation between previously published FMRP signatures and the signature described in the present invention. FMR1 mRNA expression (Figures 6A and 6B) and FMRP network signature (Luca et al. (2013), Figures 6C and 6D) correlations with breast cancer patient survival are uninformative or not statistically significant. Each panel shows the association (or not) with patient prognosis (Figures 6A, 6C: overall survival; Figures 6B, 6D: progression-free survival). Figure 6E. The genes constituting the FMRP network signature proposed by Rossella Luca et al., 2013, do not show significant overlap with the pan-signature 1 described in the present invention. FMR1 mRNA expression (Figures 6F and 6G) ​​and FMRP network signature (Zalfa et al., (2017), Figures 6H and 6I) correlations with melanoma patient survival are again uninformative or not statistically significant. Each panel shows the prognosis of the patient (Figure 6F, Figure 6H: overall survival; Figure 6G, Figure 6I: progression-free survival). Figure 6J. Genes comprising the FMRP network signature proposed by F. Zalfa et al., 2017, do not show significant overlap with the pan-signature provided in the present invention. FMR1 mRNA expression (Figure 6K and Figure 6L) and RIPK1 mRNA expression (Figure 6M and Figure 6N) correlation with colorectal cancer patient survival is again uninformative or not statistically significant. Each panel shows the prognosis of the patient (Figure 6K, Figure 6M: overall survival; Figure 6L, Figure 6N: progression-free survival).

[0112] Example 5 Mouse PDAC cancer cell lines were transfected with siRNA targeting FMR1 mRNA, which resulted in significant knockdown of FMRP expression. 24 hours after transfection with siFMRP and siControl (not targeting any mRNA), cells were subjected to RNA-seq analysis, and then a signature was developed based on the upregulated genes in siCTRL vs. siFMRP cancer cells. Figure 10 shows the inverse correlation of the level of tumor inflammation and CD8 T cells for this pan-immunosuppressive signature, reflecting its ability to suppress T cell inflammation.

[0113] While the present invention has been particularly shown and described with reference to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention as encompassed by the appended claims.

Claims

1. 1. A method of identifying a patient having cancer as (i) having high or low FMRP activity, (ii) having a high or low risk prognosis, and / or (iii) being a responder or non-responder to a cancer therapy, comprising: (a) determining in a sample obtained from a patient the expression level of genes in one or more of the signatures set forth in Tables 1-32; (b) comparing the expression level in step (a) to the level of the gene expressed in a control; (c) identifying genes that are differentially expressed between the sample and the control; (d) classifying said patient as (i) having high or low FMRP activity, (ii) having a high or low risk prognosis, and / or (iii) being a responder or non-responder to cancer therapy based on the match of differential expression with one or more of said signatures; A method comprising:

2. 2. The method of claim 1, wherein the expression levels of the genes are determined in i) one signature set forth in Tables 1-32, or ii) two or more signatures set forth in Tables 1-32.

3. The method of claim 1, wherein i) at least one gene in one or more of the signatures is differentially expressed compared to the control, or ii) at least 10 genes in one or more of the signatures are differentially expressed compared to the control.

4. 2. The method of claim 1, wherein the method identifies the patient as: i) having high or low FMRP activity, ii) having a high-risk or low-risk prognosis, or iii) being a responder or non-responder to cancer therapy.

5. The method of claim 1, wherein the patient sample is a blood or other bodily fluid or tissue sample.

6. 10. The method of claim 1, further comprising administering a cancer therapy to the patient in step (d), wherein the cancer therapy is an immune checkpoint inhibitor, an anti-FMRP therapy, chemotherapy, radiation therapy, targeted therapy, or a combination thereof.

7. 7. The method of claim 6, further comprising administering an immune checkpoint inhibitor and / or chemotherapy in combination with the FMRP inhibitor.

8. 1. A method of stratifying a group of patients having cancer as (i) having high or low FMRP activity, (ii) having a high-risk or low-risk prognosis, and / or (iii) being a responder or non-responder to a cancer therapy, comprising: (a) determining for a sample obtained from each patient of said group the expression level of genes in one or more of the signatures set out in Tables 1 to 32; (b) establishing an FMRP activity score for each sample; (c) classifying each patient as (i) having high or low FMRP activity, (ii) having a high-risk or low-risk prognosis, and / or (iii) being a responder or non-responder to cancer therapy based on said FMRP activity score; A method comprising:

9. 9. The method of claim 8, wherein the expression levels of the genes are determined in i) one signature set forth in Tables 1-32, or ii) two or more signatures set forth in Tables 1-32.

10. The method described in claim 8, wherein i) at least one gene in one or more of the signatures is differentially expressed, or ii) at least 10 genes in one or more of the signatures are differentially expressed compared to a control.

11. The method of claim 8, wherein the method i) identifies the patient as having high or low FMRP activity, ii) identifies the patient as having a high-risk or low-risk prognosis, or iii) identifies the patient as a responder or non-responder to cancer therapy.

12. The method of claim 8, wherein the sample is a blood or other body fluid or tissue sample.

13. 10. The method of claim 8, further comprising administering a cancer therapy to the patient in step (c), wherein the cancer therapy is an immune checkpoint inhibitor, an anti-FMRP therapy, chemotherapy, radiation therapy, targeted therapy, or a combination thereof.

14. 14. The method of claim 13, further comprising administering an immune checkpoint inhibitor and / or chemotherapy in combination with the FMRP inhibitor.

15. 1. A method for predicting T cell infiltration, comprising: (a) determining the expression level of the genes listed in Table 33 in a tumor sample (biopsy, resection) obtained from a patient; (b) comparing the expression level in step (a) to the level of the gene expressed in a control; (c) identifying genes that are differentially expressed between the tumor sample and the control; (d) classifying said patients as having (i) high or low FMRP immunosuppressive activity and (ii) high or low immune cell infiltration based on the match of differential expression with the signature; A method comprising:

Images