OPTIMIZATION OF MULTIGEN ANALYSIS OF TUMOR SAMPLES

Description

BACKGROUND OF THE INVENTION

[0001] The invention relates to optimized methods for analyzing nucleic acids from tumor samples, e.g., methods having integrated optimized nucleic acid selection, read alignment, and mutation calling. WO 2010 / 029498 A1 describes methods of selecting nucleic acids using solution hybridization.

[0002] The present invention relates to a method of analyzing a tumor nucleic acid sample, comprising: (a) acquiring a library comprising a plurality of nucleic acid molecules from the tumor nucleic acid sample; (b) contacting the library with a plurality of bait sets to provide a library catch being enriched for preselected subgenomic intervals of the tumor nucleic acid sample; (c) acquiring reads for the subgenomic intervals from the tumor nucleic acid sample from said library catch by a next generation sequencing method; (d) aligning said reads to a reference sequence by an alignment method; and (e) assigning a nucleotide value from said reads for a preselected nucleotide position, thereby analysing said tumor nucleic acid sample; wherein each bait set is a plurality of nucleic acid molecules which can hybridize to and thereby capture a target nucleic acid; wherein each bait set provides for a level or depth of sequence coverage that is adjusted for selection for its target subgenomic interval; and wherein the plurality of bait sets comprises a first bait set and a second bait set, wherein the first and second bait sets provide for a depth of sequencing that differs by at least 2 fold, and wherein the plurality of bait sets comprise at least two, three, four, or five of the following: (i) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an exon sequence that includes a single nucleotide alteration associated with a cancerous phenotype; (ii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an in-frame deletion of one or more codons from a reference nucleotide sequence; (iii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an intragenic deletion; (iv) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an intragenic insertion; (v) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a deletion of a full gene; (vi) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an inversion; (vii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an interchromosomal translocation; (viii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a tandem duplication; (ix) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a fusion sequence; (x) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a genomic rearrangement that comprises an intron sequence; or (xi) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a genomic rearrangement that includes an intron sequence from a 5' or 3'-UTR.

[0003] In accordance with a preferred embodiment the plurality of nucleic acid molecules of the library from step (a) comprises fragmented DNA.

[0004] In accordance with a more preferred embodiment the fragmented DNA is sheared or enzymatically prepared genomic DNA.

[0005] In accordance with a preferred embodiment the plurality of nucleic acid molecules of the library from step (a) comprises a sequence not derived from a subject, e.g., an adapter sequence, a primer sequence or a barcode sequence.

[0006] In accordance with another preferred embodiment the plurality of bait sets from step (b) each comprise a binding entity that allows for the capture and separation of a hybrid formed by a bait and a nucleic acid hybridized to the bait.

[0007] In accordance with a preferred embodiment the plurality of bait sets from step (b) is suitable for solution phase hybridization.

[0008] In accordance with a preferred embodiment the tumor nucleic acid sample comprises one or more mutations associated with a cancerous phenotype, e.g., cancer risk, cancer progression, cancer treatment, or resistance to cancer treatment.

[0009] In accordance with a preferred embodiment the amount of nucleic acid used to generate library of step (a) comprises less than 5 micrograms, less than 1 microgram, less than 500 ng, less than 100 ng, or less than 50 ng of fragmented DNA.

[0010] In accordance with a more preferred embodiment the amount of nucleic acid sample used to generate the library of step (a) comprises less than 100 ng of fragmented DNA.

[0011] In accordance with another more preferred embodiment the amount of nucleic acid sample used to generate the library of step (a) comprises less than 50 ng of fragmented DNA.

[0012] In accordance with a preferred embodiment the method comprises sequencing a subgenomic interval chosen from: at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53.

[0013] In accordance with a further preferred embodiment the method comprises sequencing a subgenomic interval chosen from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of subgenomic intervals from a mutated or wild type gene or gene product chosen from at least five or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1, TSC1, TSC2, USP9X, VHL, or WT1.

[0014] In accordance with a preferred embodiment the method further comprises amplifying the nucleic acid molecules in the library catch.

[0015] In accordance with a preferred embodiment the alignment method of step (d) includes: selecting a rearrangement reference sequence for alignment with a read, wherein said rearrangement reference sequence is preselected to align with a preselected rearrangement, e.g., wherein the reference sequence is not identical to the genomic rearrangement; and comparing, e.g., aligning, a read with said preselected rearrangement reference sequence. SUMMARY OF THE INVENTION Alignment

[0016] Methods disclosed herein can integrate the use of multiple, individually tuned, alignment methods or algorithms to optimize performance in sequencing methods, particularly in methods that rely on massively parallel sequencing of a large number of diverse genetic events in a large number of diverse genes, e.g., methods of analyzing tumor samples. In embodiments, multiple alignment methods that are individually customized or tuned to each of a number of variants in different genes are used to analyze reads. In embodiments, tuning can be a function of (one or more of) the gene (or other subgenomic inerval) being sequenced, the tumor type in the sample, the variant being sequenced, or a characteristic of the sample or the subject. The selection or use of alignment conditions that are individually tuned to a number of subgenomic intervals to be sequenced allows optimization of speed, sensitivity and specificity. The method is particularly effective when the alignments of reads for a relatively large number of diverse subgenomic intervals are optimized. also described herein but not claimed is

[0017] Accordingly, in one aspect, a method of analyzing a sample, e.g., a tumor sample. The method comprises: (a) acquiring a library comprising a plurality members from a sample, e.g., a plurality of tumor members from a tumor sample; (b) optionally, enriching the library for preselected sequences, e.g., by contacting the library with a bait set (or plurality of bait sets) to provide selected members (sometimes referred to herein as library catch); (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein; and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method) from said read for the preselected nucleotide position, thereby analyzing said tumor sample, wherein a read from each of X unique subgenomic intervals is aligned with a unique alignment method, wherein unique subgenomic interval means different from the other X-1 subgenoime intervals, and wherein unique alignment method means different from the other X-1 alingment methods, and X is at least 2.

[0018] In an implementation, step (b) is present. In an implementation step (b) is absent.

[0019] In an implementation, X is at least 3, 4, 5, 10, 15, 20, 30, 50, 100, 500, or 1,000.

[0020] In an implementation subgenomic intervals from at least X genes, e.g. at least X genes from Tables 1 and 1A, e.g., genes having the priority 1 annotation in Table 1 and 1A, are aligned with unique alignment methods, and X is equal to 2, 3, 4, 5, 10, 15, 20, or 30.

[0021] In an embodiment, a method (e.g., element (d) of the method recited above) comprises selecting or using an alignment method for analyzing, e.g., aligning, a read, wherein said alignment method is a function of, is selected responsive to, or is optimized for, one or more or all of: (i) tumor type, e.g., the tumor type in said sample; (ii) the gene, or type of gene, in which said subgenomic interval being sequenced is located, e.g., a gene or type of gene characterized by a preselected or variant or type of variant, e.g., a mutation, or by a mutation of a preselected frequency; (iii) the site (e.g., nucleotide position) being analyzed; (iv) the type of variant, e.g., a substitution, within the subgenomic interval being evaluated; (v) the type of sample, e.g., an FFPE sample; and (vi) sequence in or near said subgenomic interval being evaluated, e.g., the expected propensity for misalignment for said subgenomic interval, e.g., the presence of repeated sequences in or near said subgenomic interval.

[0022] As referred to elsewhere herein, a method is particularly effective when the alignment of reads for a relatively large number of subgenomic intervals is optimized. Thus, in an implementation, at least X unique alignment methods are used to analyze reads for at least X unique subgenomic intervals, wherein unique means different from the other X-1, and X is equal to 2, 3, 4, 5, 10, 15, 20, 30, 50, 100, 500, or 1,000.

[0023] In an implementation, subgenomic intervals from at least X genes from Tables 1 and 1A, e.g., having the priority 1 annotation in Table 1 and 1A, are analyzed, and X is equal to 2, 3, 4, 5, 10, 15, 20, or 30. implementation

[0024] In an embediment, a unique alignment method is applied to subgenomic intervals in each of at least 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100 different genes.

[0025] In an implementation, a nucleotide position in at least 20, 40, 60, 80, 100, 120, 140. 160 or 180 genes, e.g., genes from Tables 1 and 1A, is assigned a nucleotide value. In an implementation a unique alignment method is applied to subgenomic intervals in each of at least 10, 20, 30, 40, or 50% of said genes analyzed.

[0026] Methods disclosed herein allow for the rapid and efficient alignment of troublesome reads, e.g., a read having a rearrangment. Thus, in implementation where a read for a subgenomic interval comprises a nucleotide position with a rearrangement, e.g., an indel, the method can comprise using an alignment method that is appropriately tuned and that includes: selecting a rearrangement reference sequence for alignment with a read, wherein said rearrangement reference sequence is preselected to align with a preselected rearrangement (in embodiments the reference sequence is not identical to the genomic rearrangement); comparing, e.g., aligning, a read with said preselected rearrangement reference sequence.

[0027] In implementations, other methods are used to align troublesome reads. These methods are particularly effective when the alignment of reads for a relatively large number of diverse subgenomic intervals is optimized. By way of example, a method of analyzing a tumor sample can comprise: performing a comparison, e.g., an alignment comparison, of a read under a first set of parameters (e.g., a first mapping algorithm or with a first reference sequence), and determining if said read meets a first predetermined alignment criterion (e.g., the read can be aligned with said first reference sequence, e.g., with less than a preselected number of mismatches); if said read fails to meet the first predetermined alignment criterion, performing a second alignment comparison under a second set of parameters, (e.g., a second mapping algorithm or with a second reference sequence); and, optionally, determining if said read meets said second predetermined criterion (e.g., the read can be aligned with said second reference sequence with less than a preselected number of mismatches), wherein said second set of parameters comprises use of a set of parameters, e.g., said second reference sequence, which, compared with said first set of parameters, is more likely to result in an alignment with a read for a preselected variant, e.g., a rearrangement, e.g., an insertion, deletion, or translocation.

[0028] These and other alignment methods are discussed in more detail elsewhere herein, e.g., in the section entitled "Alignment Module." Elements of that module can be included in methods of analyzing a tumor. In implementations, alignment methods from the "Alignment Module" are combined with mutation calling methods from the "Mutation Calling Module" and / or a bait set from the "Bait Module." The method can be applied to set of subgenomic intervals from the "Gene Selection Module."Mutation Calling

[0029] Methods disclosed herein can integrate the use of customized or tuned mutation calling parameters to optimize performance in sequencing methods, particularly in methods that rely on massively parallel sequencing of a large number of diverse genetic events in a large number of diverse genes, e.g., from tumor samples. In implementations of the method mutation calling for each of a number of preselected subgenomic intervals is, individually, customized or fine tuned. The customization or tuning can be based on one or more of the factors described herein, e.g., the type of cancer in a sample, the gene in which subgenomic interval to be sequenced is located, or the variant to be sequenced. This selection or use of alignment conditions finely tuned to a number of subgenomic intervals to be sequenced allows optimization of speed, sensitivity and specificity. The method is particularly effective when the alignment of reads for a relatively large number of diverse subgenomic intervals is optimized.

[0030] Accordingly, in one aspect, also described herein but not claimed is, a method of analyzing a sample, e.g., a tumor sample. The method comprises: (a) acquiring a library comprising a plurality members from a sample, e.g., a plurality of tumor members from the sample, e.g., the tumor sample; (b) optionally, enriching the library for preselected sequences, e.g., by contacting the library with a bait set (or plurality of bait sets) to provide selected members, e.g., a library catch; (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein; and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method or a calling method described herein) from said read for the preselected nucleotide position, thereby analyzing said tumor sample. wherein a nucleotide value is assigned for a nucleotide position in each of X unique subgenomic intervals is assigned by a unique calling method, wherein unique subgenomic interval means different from the other X-1 subgenoimc intervals, and wherein unique calling method means different from the other X-1 calling methods, and X is at least 2. The calling methods can differ, and thereby be unique, e.g., by relying on different Bavesian prior values.

[0031] In an implementation, step (b) is present. In an implementation, step (b) is absent.

[0032] In an implementation, assigning said nucleotide value is a function of a value which is or represents the prior (e.g., literature) expectation of observing a read showing a preselected variant, e.g., a mutation. at said preselected nucleotide position in a tumor of type.

[0033] In an implementation, them method comprises assigning a nucleotide value (e.g., calling a mutation) for at least 10, 20, 40, 50, 60, 70, 80, 90, or 100 preselected nucleotide positions, wherein each assignment is a function of a unque (as opposed to the value for the other assignements) value which is or represents the prior (e.g., literature) expectation of observing a read showing a preselected variant, e.g., a mutation, at said preselected nucleotide position in a tumor of type.

[0034] In an implementation, assigning said nucleotide value is a function of a set of values which represent the probabilities of observing a read showing said preselected variant at said preselected nucleotide position if the variant is present in the sample at a frequency (e.g., 1%, 5%, 10%, etc.) and / or if the variant is absent (e.g., observed in the reads due to base-calling error alone).

[0035] In an implementation, a method (e.g., element (e) of the method recited above) comprises a mutation calling method. The mutation calling methods described herein can include the following: acquiring, for a preselected nucleotide position in each of said X subgenomic intervals: (i) a first value which is or represents the prior (e.g., literature) expectation of observing a read showing a preselected variant, e.g., a mutation, at said preselected nucleotide position in a tumor of type X; and (ii) a second set of values which represent the probabilities of observing a read showing said preselected variant at said preselected nucleotide position if the variant is present in the sample at a frequency (e.g., 1%, 5%, 10%, etc.) and / or if the variant is absent (e.g., observed in the reads due to base-calling error alone); responsive to said values, assigning a nucleotide value (e.g., calling a mutation) from said reads for each of said preselected nucleotide positions by weighing, e.g., by a Bayesian method described herein, the comparison among the values in the second set using the first value (e.g., computing the posterior probability of the presence of a mutation), thereby analyzing said sample.

[0036] In an implementation, the method comprises one or more or all of: (i) assigning a nucleotide value (e.g., calling a mutation) for at least 10, 20, 40, 50, 60, 70, 80, 90, or 100 preselected nucleotide positions, wherein each assignment is based on a unique (as opposed to the other assignments) first and / or second values; (ii) the assignment of method of (i), wherein at least 10, 20, 30 or 40 of the assignments are made with first values which are a function of a probability of a preselected variant being present of less than 5, 10, or 20%, e.g., of the cells in a preselected tumor type; (iii) assigning a nucleotide value (e.g., calling a mutation) for at least X preselected nucleotide positions, each of which of which being associated with a preselected variant having a unique (as opposed to the other X-1 assignments) probability of being present in a tumor of preselected type, e.g., the tumor type of said sample, wherein, optionally, each said of X assignments is based on a unique (as opposed to the other X-1. assignments) first and / or second value (wherein X= 2, 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100); (iv) assigning a nucleotide value (e.g., calling a mutation) at a first and a second nucleotide position, wherein the likelihood of a first preselected variant at said first nucleotide position being present in a tumor of preselected type (e.g., the tumor type of said sample) is at least 2, 5, 10, 20, 30, or 40 times greater than the likelihood of a second preselected variant at said second nucleotide position being present, wherein, optionally, each assignment is based on a unique (as opposed to the other assignments) first and / or second value; (v) assigning a nucleotide value to a plurality of preselected nucleotide positions (e.g., calling mutations), wherein said plurality comprises an assignment for variants falling into one or more, e.g., at least 3, 4, 5, 6, 7, or all, of the following probability ranges: less than .01; .01-.02; greater than 0.02 and less than or equal to 0.03; greater than 0.03 and less than or equal to 0.04; greater than 0.04 and less than or equal to 0.05; greater than 0.05 and less than or equal to 0.1; greater than 0.1 and less than or equal to 0.2; greater than 0.2 and less than or equal to 0.5; greater than 0.5 and less than or equal to 1.0; greater than 1.0 and less than or equal to 2.0; greater than 2.0 and less than or equal to 5.0; greater than 5.0 and less than or equal to 10.0; greater than 10.0 and less than or equal to 20.0; greater than 20.0 and less than or equal to 50.0; and greater than 50 and less than or equal to 100.0 %; wherein, a probability range is the range of probabilities that a preselected variant at a preselected nucleotide position will be present in a tumor of preselected type (e.g., the tumor type of said sample) or the probability that a preselected variant at a preselected nucleotide position will be present in the recited % of the cells in a tumor sample, a library from the tumor sample, or library catch from that library, for a preselected type (e.g., the tumor type of said sample), and wherein, optionally, each assignment is based on a unique first and / or second value (e.g., unique as opposed to the other assignments in a recited probability range or unique as opposed to the first and / or second values for one or more or all of the other listed probability ranges). (vi) assigning a nucleotide value (e.g., calling a mutation) for at least 1, 2 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100 preselected nucleotide positions each, independently, having a preselected variant present in less than 50, 40, 25, 20, 15, 10, 5, 4, 3, 2, 1, 0.5, 0.4, 0.3, 0.2, or 0.1 % of the DNA in said sample, wherein, optionally, each assignment is based on a unique (as opposed to the other assignments) first and / or second value; (vii) assigning a nucleotide value (e.g., calling a mutation) at a first and a second nucleotide position, wherein the likelihood of a preselected variant at the first position in the DNA of said sample is at least 2, 5, 10, 20, 30, or 40 times greater than a the likelihood of a preselected variant at said second nucleotide position in the DNA of said sample, wherein, optionally, each assignment is based on a unique (as opposed to the other assignments) first and / or second value; (viii) assigning a nucleotide value (e.g., calling a mutation) in one or more or all of the following: (1) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in less than 1.0 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (2) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in 1.0- 2.0 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library ; (3) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 2.0 % and less than or equal to 3 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library (4) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 3.0 % and less than or equal to 4 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (5) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 4.0 % and less than or equal to 5 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (6) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 5.0 % and less than or equal to 10 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (7) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 10.0 % and less than or equal to 20 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (8) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 20.0 % and less than or equal to 40 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; (9) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present at greater than 40.0 % and less than or equal to 50 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; or (10) at least 1, 2 3, 4 or 5 preselected nucleotide positions having a preselected variant present in greater than 50.0 % and less than or equal to 100 % of the cells in said sample, of the nucleic acid in a library from said sample, or the nucleic acid in a library catch from that library; wherein, optionally, each assignment is based on a unique first and / or second value (e.g,, unique as opposed to the other assignments in the recited range (e.g., the range in (i) of less than 1%) or unque as opposed to a first and / or second values for a determination in one or more or all of the other listed ranges); or (ix) assigning a nucleotide value (e.g., calling a mutation) at each of X nucleotide positions, each nucleotide position, independently, having a likelihood (of a preselected variant being present in the DNA of said sample) that is unique as compared with the likelihood for a preselected variant at the other X-1 nucleotide positions, wherein X is equal to or greater than 1, 2 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100, and wherein each assignment is based on a unique (as opposed to the other assignments) first and / or second value.

[0037] In implementations of the method, a "threshold value" is used to evaluate reads, and select from the reads a value for a nucleotide position, e.g., calling a mutation at a specific position in a gene. In implementations of the method, a threshold value for each of a number of preselected subgenomic intervals is customized or fine tuned. The customization or tuning can be based on one or more of the factors described herein, e.g., the type of cancer in a sample, the gene in which subgenomic interval to be sequenced is located, or the variant to be sequenced. This provides for calling that is finely tuned to each of a number of subgenomic intervals to be sequenced. The method is particularly effective when a relatively large number of diverse subgenomic intervals are analyzed.

[0038] Thus, in another implementation the method of analyzing a tumor comprises the following mutation calling method: acquiring, for each of said X subgenomic intervals, a threshold value, wherein each of said acquired X threshold values is unique as compared with the other X-1 threshold values, thereby providing X unique threshold values; for each of said X subgenomic intervals, comparing an observed value which is a function of the number of reads having a preselected nucleotide value at a preselected nucleotide position with its unique threshold value, thereby applying to each of said X subgenomic intervals, its unique threshold value: and optionally, responsive to the result of said comparison, assigning a nucleotide value to a preselected nucleotide position, wherein X is equal to or greater than 2.

[0039] In an implementation, the method includes assigning a nucleotide value at at least 2, 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100 preselected nucleotide positions, each having, independently, a first value that is a function of a probability that is less than 0.5, 0.4, 0.25, 0.15, 0.10, 0.05, 0.04, 0.03, 0.02, or 0.01.

[0040] In an implementation, the method includes assigning a nucleotide value at at each of at least X nucleotide positions, each independently having a first value that is unique as compared with the other X-1 first values, and wherein each of said X first values is a function of a probability that is less than 0.5, 0.4, 0.25, 0.15, 0.10, 0.05, 0.04, 0.03, 0.02, or 0.01, wherein X is equal to or greater than 1, 2 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100.

[0041] In an implementation, a nucleotide position in at least 20, 40, 60, 80, 100, 120, 140, 160 or 180 genes, e.g., genes from Table 1, is assigned a nucleotide value. In an implementation unique first and / or second values are applied to subgenomic intervals in each of at least 10, 20, 30, 40, or 50% of said genes analyzed.

[0042] Implementations of the method can be applied where threshold values for a relatively large number of subgenomic intervals are optimized, as is seen, e.g., from the following implementations. implementation

[0043] In an embodiment, a unique threshold value is applied to subgenomic intervals in each of at least 3, 5, 10, 20, 40, 50, 60, 70, 80, 90, or 100 different genes.

[0044] In an implementation, a nucleotide position in at least 20, 40, 60, 80, 100, 120, 140, 160 or implementation 180 genes, e.g., genes from Table 1, is assigned a nucleotide value. In an implementation a unique threshold value is applied to a subgenomic interval in each of at least 10, 20, 30, 40, or 50% of said genes analyzed.

[0045] In an implementation, a a nucleotide position in at least 5, 10, 20, 30, or 40 genes from Table 1 having the priority 1 annotation is assigned a nucleotide value. In an implementation a unique threshold value is applied to a subgenomic interval in each of at least 10, 20, 30, 40, or 50% of said genes analyzed.

[0046] These and other mutation calling methods are discussed in more detail elsewhere herein, e.g., in the section entitled "Mutation Calling Module." Elements of that module can be included in methods of analyzing a tumor. In implementations, alignment methods from the "Mutation Calling Module" are combined with alignment methods from the "Alignment Module" and / or a bait set from the "Bait Module." The method can be applied to set of subgenomic intervals from the "Gene Selection Module."Bait

[0047] Methods described herein provide for optimized sequencing of a large number of genes and gene products from samples, e.g., tumor samples, from one or more subjects by the appropriate selection of baits, e.g., baits for use in solution hybridization, for the selection of target nucleic acids to be sequenced. The efficiency of selection for various subgenomic intervals, or classes thereof, are matched according to bait sets having preselected efficiency of selection. As used in this section, "efficiency of selection" refers to the level or depth of sequence coverage as it is adjusted according to a target subgenomic interval(s).

[0048] Thus a method (e.g., element (b) of the method recited above) comprises contacting the library with a plurality of baits to provide selected members (e.g., a library catch).

[0049] Accordingly, in one aspect, also described herein but not claimed is, a method of analyzing a sample, e.g., a tumor sample. The method comprises: (a) acquiring a library comprising a plurality of members (e.g., target members) from a sample, e.g., a plurality of tumor members from a tumor sample; (b) contacting the library with a bait set to provide selected members (e.g., a library catch); (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method or a method described herein) from said read for the preselected nucleotide position, thereby analyzing said tumor sample, wherein the method comprises contacting the library with a plurality, e.g., at least two, three, four, or five, of baits or bait sets, wherein each bait or bait set of said plurality has a unique (as opposed to the other bait sets in the plurality), preselected efficiency for selection. E.g., each unique bait or bait set provides for a unique depth of sequencing. The term "bait set", as used herein, collectively refers to one bait or a plurality of bait molecules.

[0050] implementation In an embediment, the efficiency of selection of a first bait set in the plurality differs from the efficiency of a second bait set in the plurality by at least 2 fold. In an implementation, the first and second bait sets provide for a depth of sequencing that differs by at least 2 fold.

[0051] In an implementation, the method comprises contacting one, or a plurality of the following bait sets with the library: a) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 500X or higher sequencing depth, e.g., to sequence a mutation present in no more than 5% of the cells from the sample; b) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 200X or higher, e.g., about 200X-about 500X, sequencing depth, e.g., to sequence a mutation present in no more than 10% of the cells from the sample; c) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 10-100X sequencing depth, e.g., to sequence one or more subgenomic intervals (e.g., exons) that are chosen from: a) a pharmacogenomic (PGx) single nucleotide polymorphism (SNP) that may explain the ability of patient to metabolize different drugs, or b) a genomic SNPs that may be used to uniquely identify (e.g., fingerprint) a patient; d) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 5-50 X sequencing depth, e.g., to detect a structural breakpoint, such as a genomic translocation or an indel. For example, detection of an intronic breakpoint requires 5-50X sequence-pair spanning depth to ensure high detection reliability. Such bait sets can be used to detect, for example, translocation / indel-prone cancer genes; or e) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 0.1-300X sequencing depth, e.g., to detect copy number changes. In one implementation, the sequencing depth ranges from about 0.1-10X sequencing depth to detect copy number changes. In other implementations, the sequencing depth ranges from about 100-300X to detect a genomic SNPs / loci that is used to assess copy number gains / losses of genomic DNA or loss-of-heterozygosity (LOH). Such bait sets can be used to detect, for example, amplification / deletion-prone cancer genes.

[0052] The level of sequencing depth as used herein (e.g., X-fold level of sequencing depth) refers to the level of coverage of reads (e.g., unique reads), after detection and removal of duplicate reads. e.g., PCR duplicate reads.

[0053] In one implementation, the bait set selects a subgenomic interval containing one or more rearrangements, e.g., an intron containing a genomic rearrangement. In such implementations, the bait set is designed such that repetive sequences are masked to increase the selection efficiency. In those implementations where the rearrangement has a known juncture sequence, complementary bait sets can be designed to the juncture sequence to increase the selection efficiency.

[0054] In implementations, the method comprises the use of baits designed to capture two or more different target categories, each category having a different bait design strategies. In implementations, the hybrid capture methods and compositions disclosed herein capture a defined subset of target sequences (e.g., target members) and provide homogenous coverage of the target sequence, while minimizing coverage outside of that subset. In one implementation, the target sequences include the entire exome out of genomic DNA, or a selected subset thereof. The methods and compositions disclosed herein provide different bait sets for achieving different depths and patterns of coverage for complex target nucleic acid sequences (e.g., nucleic acid libraries).

[0055] In an implementation, the method comprises providing selected members of a nucleic acid library (e.g., a library catch). The method includes: providing a library (e.g., a nucleic acid library) comprising a plurality of members, e.g., target nucleic acid members (e.g., including a plurality of tumor members, reference members, and / or PGx members); contacting the library, e.g., in a solution-based reaction, with a plurality of baits (e.g., oligonucleotide baits) to form a hybridization mixture comprising a plurality of bait / member hybrids; separating the plurality of bait / member hybrids from said hybridization mixture, e.g., by contacting said hybridization mixture with a binding entity that allows for separation of said plurality of bait / member hybrid, thereby providing a library-catch (e.g., a selected or enriched subgroup of nucleic acid molecules from the library), wherein the plurality of baits includes two or more of the following: a) a first bait set that selects a high-level target (e.g., one or more tumor members that include a subgenomic interval, such a gene, an exon, or a base) for which the deepest coverage is required to enable a high level of sensitivity for an alteration (e.g., one or more mutations) that appears at a low frequency, e.g., about 5% or less (i.e., 5% of the cells from the sample harbor the alteration in their genome). In one implementation; the first bait set selects (e.g., is complementary to) a tumor member that includes an alteration (e.g., a point mutation) that requires about 500X or higher sequencing depth; b) a second bait set that selects a mid-level target (e.g., one or more tumor members that include a subgenomic interval, such as a gene, an exon, or a base) for which high coverage is required to enable high level of sensitivity for an alteration (e.g., one or more mutations) that appears at a higher frequency than the high-level target in a), e.g., a frequency of about 1 0% (i.e., 10% of the cells from the sample harbor the alteration in their genome). In one implementation; the second bait set selects (e.g., is complementary to) a tumor member that includes an alteration (e.g., a point mutation) that requires about 200X or higher sequencing depth; c) a third bait set that selects a low-level target (e.g., one or more PGx members that includes a subgenomic interval, such as a gene, an exon, or a base) for which low-medium coverage is required to enable high level of sensitivity, e.g., to detect heterozygous alleles. For example, detection of heterozygous alleles requires 10-100X sequencing depth to ensure high detection reliability. In one implementation, third bait set selects one or more subgenomic intervals (e.g., exons) that are chosen from: a) a pharmacogenomic (PGx) single nucleotide polymorphism (SNP) that may explain the ability of patient to metabolize different drugs, or b) a genomic SNPs that may be used to uniquely identify (e.g., fingerprint) a patient; d) a fourth bait set that selects a first intron target (e.g., a member that includes an intron sequence) for which low-medium coverage is required, e.g., to detect a structural breakpoint, such as a genomic translocation or an indel. For example, detection of an intronic breakpoint requires 5-50X sequence-pair spanning depth to ensure high detection reliability. Said fourth bait sets can be used to detect, for example, translocation / indel-prone cancer genes; or e) a fifth bait set that selects a second intron target (e.g., an intron member) for which sparse coverage is required to improve the ability to detect copy number changes. For example, detection of a one-copy deletion of several terminal exons requires 0.1-300X coverage to ensure high detection reliability. In one implementation. the coverage depth ranges from about 0.1-10X to detect copy number changes. In other implementations, the coverage depth ranges from about 100-300X to detect a genomic SNPs / loci that is used to assess copy number gains / losses of genomic DNA or loss-of-heterozygosity (LOH). Said fifth bait sets can be used to detect, for example, amplification / deletion-prone cancer genes.

[0056] Any combination of two, three, four or more of the aforesaid bait sets can be used, for example, a combination of the first and the second bait sets; first and third bait sets; first and fourth bait sets; first and fifth bait sets; second and third bait sets; second and fourth bait sets; second and fifth bait sets; third and fourth bait sets; third and fifth bait sets; fourth and fifth bait sets; first, second and third bait sets; first, second and fourth bait sets; first, second and fifth bait sets; first, second, third, fourth bait sets; first, second, third, fourth and fifth bait sets, and so on.

[0057] In one implementation, each of the first, second, third, fourth, or fifth bait set has a preselected efficiency for selection (e.g., capture). In one implementation, the value for efficiency of selection is the same for at least two, three, four of all five baits according to a)-e). In other implementations, the value for efficiency of selection is different for at least two, three, four of all five baits according to a)-e).

[0058] In some implementations, at least two, three, four, or all five bait sets have a preselected efficiency value that differ. For example, a value for efficiency of selection chosen from one of more of: (i) the first preselected efficiency has a value for first efficiency of selection that is at least about 500X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the second, third, fourth or fifth preselected efficiency of selection (e.g., about 2-3 fold greater than the value for the second efficiency of selection; about 5-6 fold greater than the value for the third efficiency of selection; about 10 fold greater than the value for the fourth efficiency of selection; about 50 to 5000-fold greater than the value for the fifth efficiency of selection); (ii) the second preselected efficiency has a value for second efficiency of selection that is at least about 200X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the third, fourth or fifth preselected efficiency of selection (e.g., about 2 fold greater than the value for the third efficiency of selection; about 4 fold greater than the value for the fourth efficiency of selection; about 20 to 2000-fold greater than the value for the fifth efficiency of selection); (iii) the third preselected efficiency has a value for third efficiency of selection that is at least about 100X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the fourth or fifth preselected efficiency of selection (e.g., about 2 fold greater than the value for the fourth efficiency of selection; about 10 to 1000-fold greater than the value for the fifth efficiency of selection); (iv) the fourth preselected efficiency has a value for fourth efficiency of selection that is at least about 50X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the fifth preselected efficiency of selection (e.g., about 50 to 500-fold greater than the value for the fifth efficiency of selection); or (v) the fifth preselected efficiency has a value for fifth efficiency of selection that is at least about 10X to 0.1 X seguencing depth.

[0059] In certain implementations, the value for efficiency of selection is modified by one or more of: differential representation of different bait sets, differential overlap of bait subsets, differential bait parameters, mixing of different bait sets, and / or using different types of bait sets. For example, a variation in efficiency of selection (e.g., relative sequence coverage of each bait set / target category) can be adjusted by altering one or more of: (i) Differential representation of different bait sets - The bait set design to capture a given target (e.g., a target member) can be included in more / fewer number of copies to enhance / reduce relative target coverage depths; (ii) Differential overlap of bait subsets - The bait set design to capture a given target (e.g., a target member) can include a longer or shorter overlap between neighboring baits to enhance / reduce relative target coverage depths; (iii) Differential bait parameters - The bait set design to capture a given target (e.g., a target member) can include sequence modifications / shorter length to reduce capture efficiency and lower the relative target coverage depths; (iv) Mixing of different bait sets - Bait sets that are designed to capture different target sets can be mixed at different molar ratios to enhance / reduce relative target coverage depths; (v) Using different types of oligonucleotide bait sets -In certain implementations the bait set can include: (a) one or more chemically (e.g., non-enzymatically) synthesized (e.g., individually synthesized) baits, (b) one or more baits synthesized in an array, (c) one or more enzymatically prepared, e.g., in vitro transcribed, baits; (d) any combination of (a), (b) and / or (c), (e) one or more DNA oligonucleotides (e.g., a naturally or non-naturally occurring DNA oligonucleotide), (f) one or more RNA oligonucleotides (e.g., a naturally or non-naturally occurring RNA oligonucleotide), (g) a combination of (e) and (f), or (h) a combination of any of the above.

[0060] The different oligonucleotide combinations can be mixed at different ratios, e.g., a ratio chosen from 1:1, 1:2, 1:3, 1:4, 1:5, 1:10, 1:20, 1:50; 1:100, 1:1000, or the like. In one implementation, the ratio of chemically-synthesized bait to array-generated bait is chosen from 1:5, 1:10, or 1:20. The DNA or RNA oligonucleotides can be naturally- or non-naturally-occurring. In certain implementations, the baits include one or more non-naturally-occurring nucleotide to, e.g., increase melting temperature. Exemplary non-naturally occurring oligonucleotides include modified DNA or RNA nucleotides. Exemplary modified nucleotides (e.g., modified RNA or DNA nucleotides) include, but are not limited to, a locked nucleic acid (LNA), wherein the ribose moiety of an LNA nucleotide is modified with an extra bridge connecting the 2' oxygen and 4' carbon; peptide nucleic acid (PNA), e.g., a PNA composed of repeating N-(2-aminoethyl)-glycine units linked by peptide bonds; a DNA or RNA oligonucleotide modified to capture low GC regions; a bicyclic nucleic acid (BNA); a crosslinked oligonucleotide; a modified 5-methyl deoxycytidine; and 2,6-diaminopurine. Other modified DNA and RNA nucleotides are known in the art.

[0061] In certain implementations, a substantially uniform or homogeneous coverage of a target sequence (e.g., a target member) is obtained. For example, within each bait set / target category, uniformity of coverage can be optimized by modifying bait parameters, for example, by one or more of: (i) Increasing / decreasing bait representation or overlap can be used to enhance / reduce coverage of targets (e.g., target members), which are under / over-covered relative to other targets in the same category; (ii) For low coverage, hard to capture target sequences (e.g., high GC content sequences), expand the region being targeted with the bait sets to cover, e.g., adjacent sequences (e.g., less GC-rich adjancent sequences); (iii) Modifying a bait sequence can be made to reduce secondary structure of the bait and enhance its efficiency of selection; (iv) Modifying a bait length can be used to equalize melting hybridization kinetics of different baits within the same category. Bait length can be modified directly (by producing baits with varying lengths) or indirectly (by producing baits of consistent length, and replacing the bait ends with arbitrary sequence); (v) Modifying baits of different orientation for the same target region (i.e. forward and reverse strand) may have different binding efficiencies. The bait set with either orientation providing optimal coverage for each target may be selected; (vi) Modifying the amount of a binding entity, e.g., a capture tag (e.g. biotin), present on each bait may affect its binding efficiency. Increasing / decreasing the tag level of baits targeting a specific target may be used to enhance / reduce the relative target coverage; (vii) Modifying the type of nucleotide used for different baits can be altered to affect binding affinity to the target, and enhance / reduce the relative target coverage; or (viii) Using modified oligonucleotide baits, e.g., having more stable base pairing, can be used to equalize melting hybridization kinetics between areas of low or normal GC content relative to high GC content.

[0062] For example, different types of oligonucleotide bait sets can be used.

[0063] In one implementation, the value for efficiency of selection is modified by using different types of bait oligonucleotides to encompass pre-selected target regions. For example, a first bait set (e.g., an array-based bait set comprising 10,000-50,000 RNA or DNA baits) can be used to cover a large target area (e.g., 1-2MB total target area). The first bait set can be spiked with a second bait set (e.g., individually synthesized RNA or DNA bait set comprising less than 5,000 baits) to cover a pre-selected target region (e.g., selected subgenomic intervals of interest spanning, e.g., 250kb or less, of a target area) and / or regions of higher secondary structure, e.g., higher GC content. Selected subgenomic intervals of interest may correspond to one or more of the genes or gene products described herein, or a fragment thereof. The second bait set may include about 1-5,000, 2-5,000, 3-5,000, 10-5,000, 100-5,000, 500-5,000, 1100-5,00, 1000-5,000, 2,000-5,000 baits depending on the bait overlap desired. In other implementations, the second bait set can include selected oligo baits (e.g., less than 400, 200, 100, 50, 40, 30, 20, 10, 5, 4, 3, 2 or 1 baits) spiked into the first bait set. The second bait set can be mixed at any ratio of individual oligo baits. For example, the second bait set can include individual baits present as a 1:1 equimolar ratio. Alternatively, the second bait set can include individual baits present at different ratio (e.g., 1:5, 1:10, 1:20), for example, to optimize capture of certain targets (e.g., certain targets can have a 5-10X of the second bait compared to other targets).

[0064] In other implementations, the efficiency of selection is adjusted by leveling the efficiency of individual baits within a group (e.g., a first, second or third plurality of baits) by adjusting the relative abundance of the baits, or the density of the binding entity (e.g., the hapten or affinity tag density) in reference to differential sequence capture efficiency observed when using an equimolar mix of baits, and then introducing a differential excess of internally-leveled group 1 to the overall bait mix relative to internally-leveled group 2.

[0065] In an implementation, the method comprises the use of a plurality of bait sets that includes a bait set that selects a tumor member, e.g., a nucleic acid molecule comprising a subgenomic interval from a tumor cell (also referred to herein as "a tumor bait set"). The tumor member can be any nucleotide sequence present in a tumor cell, e.g., a mutated, a wild-type, a PGx, a reference or an intron nucleotide sequence, as described herein, that is present in a tumor or cancer cell. In one implementation, the tumor member includes an alteration (e.g., one or more mutations) that appears at a low frequency, e.g., about 5% or less of the cells from the tumor sample harbor the alteration in their genome. In other implementations, the tumor member includes an alteration (e.g., one or more mutations) that appears at a frequency of about 10% of the cells from the tumor sample. In other implementations, the tumor member includes a subgenomic interval from a PGx gene or gene product, an intron sequence, e.g., an intron sequence as described herein, a reference sequence that is present in a tumor cell.

[0066] In another aspect, also described herein but not claimed is, a bait set described herein, combinations of individual bait sets described herein, e.g., combinations described herein. The bait set(s) can be part of a kit which can optionally comprise instructions, standards, buffers or enzymes or other reagenats.Gene Selection

[0067] Preselected subgenomic intervals for analysis, e.g., a group or set of subgenomic intervals for sets or groups of genes and other regions, are described herein.

[0068] Thus, in implementations a method comprises sequencing, e.g., by a next generation sequencing method, a subgenomic interval from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the acquired nucleic acid sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1 FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1 MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53, thereby analyzing the tumor sample.

[0069] Accordingly, in one aspect, also described herein but not claimed is, a method of analyzing a sample, e.g., a tumor sample. The method comprises: (a) acquiring a library comprising a plurality members from a sample, e.g., a plurality of tumor members from a tumor sample; (b) optionally, enriching the library for preselected sequences, e.g., by contacting the library with a bait set (or plurality of bait sets) to provide selected members (e.g., a library catch); (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein; and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method or a method described herein) from said read for the preselected nucleotide position, thereby analyzing said tumor sample, wherein the method comprises sequencing, e.g., by a next generation sequencing method, a subgenomic interval from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT,

[0070] KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53. In an implementation, step (b) is present. In an implementation, step (b) is absent.

[0071] In another implementation, subgenomic intervals of one of the following sets or groups are analyzed. E.g., subgenomic intervals associated with a tumor or cancer gene or gene product, a reference (e.g., a wild type) gene or gene product, and a PGx gene or gene product, can provide a group or set of subgenomic intervals from the tumor sample.

[0072] In an implementation, the method acquires a read, e.g., sequences, a set of subgenomic intervals from the tumor sample, wherein the subgenomic intervals are chosen from at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 or all of the following: A) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more subgenomic intervals from a mutated or wild-type gene or gene product chosen from at least five or more of: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53; B) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of subgenomic intervals from a mutated or wild type gene or gene product chosen from at least five or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1, TSC1, TSC2, USP9X, VHL, or WT1; C) at least five, six, seven, eight, nine, ten, fifteen, twenty, or more subgenomic intervals from a gene or gene product according to Table 1, 1A, 2, 3 or 4; D) at least five, six, seven, eight, nine, ten, fifteen, twenty, or more subgenomic intervals from a gene or gene product that is associated with a tumor or cancer (e.g., is a positive or negative treatment response predictor, is a positive or negative prognostic factor for, or enables differential diagnosis of a tumor or cancer, e.g., a gene or gene product chosen from one or more of: ABL1, AKT1, ALK, AR, BRAF, BRCA1, BRCA2, CEBPA, EGFR, ERBB2, FLT3, JAK2, KIT, KRAS, MET, NPM1, PDGFRA, PIK3CA, RARA, AKT2, AKT3, MAP2K4, NOTCH 1, and TP53; E) at least five, six, seven, eight, nine, ten, or more subgenomic intervals including a mutated or a wild type codon chosen from one or more of: codon 315 of the ABL1 gene; codon 1114, 1338, 1450 or 1556 of APC; codon 600 of BRAF; codon 32, 33, 34, 37, 41 or 45 of CTNNB1; codon 719, 746-750, 768, 790, 858 or 861 of EGFR; codon 835 of FLT3; codon 12, 13, or 61 of HRAS; codon 617 of JAK2; codon 816 of KIT; codon 12, 13, or 61 of KRAS; codon 88, 542, 545, 546, 1047, or 1049 of PIK3CA; codon 130, 173, 233, or 267 of PTEN; codon 918 of RET; codon 175, 245, 248, 273, or 306 of TP53 (e.g., at least five, ten, fifteen, twenty or more subgenomic intervals that include one or more of the codons shown in Table 1). F) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, or more of subgenomic intervals from a mutated or wild type gene or gene product (e.g., single nucleotide polymorphism (SNP)) of a subgenomic interval that is present in a gene or gene product associated with one or more of drug metabolism, drug responsiveness, or toxicity (also referred to therein as "PGx" genes) chosen from: ABCB1, BCC2, ABCC4, ABCG2, C1orf144, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DPYD, ERCC2, ESR2, FCGR3A, GSTP1, ITPA, LRP2, MAN1B1, MTHFR, NQO1, NRP2, SLC19A1, SLC22A2, SLCO1B3, SOD2, SULT1A1, TPMT, TYMS, UGT1A1, or UMPS; G) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, or more of subgenomic intervals from a mutated or wild type PGx gene or gene product (e.g., single nucleotide polymorphism (SNP)) of a subgenomic interval that is present in a gene or gene product associated with one or more of: (i) better survival of a cancer patient treated with a drug (e.g., better survival of a breast cancer patient treated with paclitaxel (e.g., an ABCB1 gene)); (ii) paclitaxel metabolism (e.g., CYP2C8 genes at different loci and mutations shown in Table 2; CYP3A4 gene); (iii) toxicity to a drug (e.g., 6-MP toxicity as seen with ABCC4 gene (Table 2); 5-FU toxicity as seen with DPYD gene, TYMS gene, or UMPS gene (Table 2); purine toxicity as seen with a TMPT gene (Table 2); daunorubicin toxicity as seen with NRP2 gene; Clorf144 gene, CYP1B1 gene (Table 2); or (iv) a side effect to a drug (e.g., ABCG2, TYMS, UGT1A1, ESR1 and ESR2 genes (Table 2)); H) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3; J) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3 in a solid tumor sample from the cancer types specified therein; K) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products according to Table 4; L) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products according to Table 4 in a heme tumor sample from the cancer types specified therein; M) at least five genes or gene products selected from Table 1-4, wherein an allelic variation, e.g., at the preselected position, is associated with a preselected type of tumor and wherein said allelic variation is present in less than 5% of the cells in said tumor type; N) at least five genes or gene products selected from Table 1, 1A-4, which are embedded in a GC-rich region; or O) at least five genes or gene products indicative of a genetic (e.g., a germline risk) factor for developing cancer (e.g., the gene or gene product is chosen from one or more of BRCA1, BRCA2, EGFR, HRAS, KIT, MPL, ALK, PTEN, RET, APC, CDKN2A, MLH1, MSH2, MSH6, NF1, NF2, RB1, TP53, VHL or WT1).

[0073] In yet another implementation, the method acquires a read, e.g., sequences, a set of subgenomic intervals from the tumor sample, wherein the subgenomic intervals are chosen from one, two, three, four, five, ten, fifteen or all of the alterations described in Table 1B. In one implementation, the subgenomic interval includes an alteration classified in one or more of Category A, B, C, D or E. In other implementations, the subgenomic interval includes an alteration in KRAS G13D in a tumor sample, e.g., a colon, lung or breast tumor sample. In other implementations, the subgenomic interval includes an alteration in NRAS Q61K in a tumor sample, e.g., a melanoma or colon tumor sample. In yet other implementations, the subgenomic interval includes an alteration in BRAF V600E in a tumor sample, e.g., a melanoma, colon, or lung tumor sample. In other implementations, the subgenomic interval includes an alteration in BRAF D594G in a tumor sample, e.g., a lung tumor sample. In other implementations, the subgenomic interval includes an alteration in PIK3CA H1047R in a tumor sample, e.g., a breast or colon tumor sample. In yet other implementations, the subgenomic interval includes an alteration in EGFR L858R or T790M in a tumor sample, e.g., a lung tumor sample. In other implementations, the subgenomic interval includes an alteration in ERBB2 in a tumor sample, e.g., an ERBB2 amplification in a breast tumor sample. In other implementations, the subgenomic interval includes an alteration in BRCA1 in a tumor sample, e.g., a BRCA1 biallelic inactivation in a breast tumor sample. In other implementations, the subgenomic interval includes an alteration in BRCA2 in a tumor sample, e.g., a BRCA2 biallelic inactivation in a pancreatic tumor sample. In other implementations, the subgenomic interval includes an alteration in ATM in a tumor sample, e.g., an ATM biallelic inactivation in a breast tumor sample. In other implementations, the subgenomic interval includes an alteration in TSC in a tumor sample, e.g., a TSC biallelic inactivation in a colon tumor sample. In other implementations, the subgenomic interval includes an alteration in PTEN in a tumor sample, e.g., a PTEN biallelic inactivation in a breast or colon tumor sample. In yet other implementations, the subgenomic interval includes an alteration in VHL in a tumor sample, e.g., a VHL biallelic inactivation in a kidney tumor sample. In other implementation, the subgenomic interval includes an alteration in ATR in a tumor sample, e.g., an ATR biallelic inactivation in a breast tumor sample. In other implementations, the subgenomic interval includes an alteration in MYC in a tumor sample, e.g., a MYC biallelic inactivation in a breast tumor sample.

[0074] These and other sets and groups of subgenomic intervals are discussed in more detail elsewhere herein, e.g., in the section entitled "Gene Selection Module."

[0075] Anv of the methods described herein can be combined with one or more of the implementations below.

[0076] In other implementations, the sample is a tumor sample, e.g., includes one or more premalignant or malignant cells. In certain, implementations, the sample, e.g., the tumor sample, is acquired from a solid tumor, a soft tissue tumor or a metastatic lesion. In other implementations, the sample, e.g., the tumor sample, includes tissue or cells from a surgical margin. The sample can be histologically normal tissue. In another implementation, the sample, e.g., tumor sample, includes one or more circulating tumor cells (CTC) (e.g., a CTC acquired from a blood sample).

[0077] In one implementation, the method further includes acquiring a sample, e.g., a tumor sample as described herein. The sample can be acquired directly or indirectly.

[0078] In other implementations, the method includes evaluating a sample, e.g., a histologically normal sample, e.g., from a surgical margin, using the methods described herein. Applicants have discovered that samples obtained from histologically normal tissues (e.g., otherwise histologically normal tissue margins) may still have an alteration as described herein. The methods may thus further include re-classifying a tissue sample based on the presence of the detected alteration.

[0079] In another implementation, at least 10, 20, 30, 40, 50, 60, 70, 80, or 90 % of the reads acquired or analyzed are for subgenomic intervals from genes described herein, e.g., genes from Table 1-1A, or priority 1 genes from Table 1.

[0080] In an implementation, at least 10, 20, 30, 40, 50, 60, 70, 80, or 90 % of the mutation calls made in the method are for subgenomic intervals from genes described herein, e.g., genes from Table 1-1A, or priority 1 genes from Table 1.

[0081] In an implementation, at least 10, 20, 30, 40, 50, 60, 70, 80, or 90 % of the unique threshold values used the method are for subgenomic intervals from genes described herein, e.g., genes from Table 1-1A. or priority 1 genes from Table 1.

[0082] In an implementation, at least 10, 20, 30, 40, 50, 60, 70, 80, or 90 % of the mutation calls annotated, or reported to a third party, are for subgenomic intervals from genes described herein, e.g., genes from Table 1-1A, or priority 1 genes from Table 1.

[0083] In an implementation, the method comprises acquiring a nucleotide sequence read obtained from a tumor and / or control nucleic acid sample (e.g., an FFPE-derived nucleic acid sample).

[0084] In an implementation, the reads are provided by a NGS sequencing method.

[0085] In an implementation, the method includes providing a library of nucleic acid members and sequencing preselected subgenomic intervals from a pluality of members of said library. In implementations the method can include a step of selecting a subset of said library for sequencing, e.g., a solution-based selection or a solid support- (e.g., array-) based selection.

[0086] In an implementation, the method includes the step of contacting a library with a plurality of baits to provide a selected subgroup of nucleic acids, e.g., a library catch. In one implementation, the contacting step is effected in solution hybridization. In another implementation, the contacting step is effected in a solid support, e.g., an array. In certain implementations, the method includes repeating the hybridization step by one or more additional rounds of hybridization. In some implementations, the methods further include subjecting the library catch to one or more additional rounds of hybridization with the same or different collection of baits.

[0087] In vet other implementations, the methods further include analyzing the library catch. In one implementation, the library catch is analyzed by a sequencing method, e.g., a next-generation sequencing method as described herein. The methods include isolating a library catch by, e.g., solution hybridization, and subjecting the library catch by nucleic acid sequencing. In certain implementations, the library catch can be re-sequenced. Next generation sequencing methods are known in the art, and are described, e.g., in Metzker, M. (2010) Nature Biotechnology Reviews 11:31-46.

[0088] In an implementation, the assigned value for a nucleotide position is transmitted to a third party, optionally, with explanatory annotation.

[0089] In an implementation, the assigned value for a nucleotide position is not transmitted to a third party.

[0090] In an implementation, the assigned value for a plurality of nucleotide position is transmitted to a third party, optionally, with explanatory annotations, and the assigned value for a second plurality of nucleotide position is not transmitted to a third party.

[0091] In an implementation, at least 0.01, 0.02, 0.03, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 15, or 30 megabases bases, e.g., genomic bases, are sequenced.

[0092] In an implementation, the method comprises evaluating a plurality of reads that include at least one SNP

[0093] In an implementation, the method comprises determining an SNP allele ratio in the sample and / or control read.

[0094] In an implementation, the method comprises assigning one or more reads to a subject, e.g., by barcode deconvolution.

[0095] In an implementation, the method comprises assigning one or more reads as a tumor read or a control read, e.g., by barcode deconvolution.

[0096] In an implementation, the method comprises mapping, e.g., by alignment with a reference sequence, each of said one or more reads.

[0097] In an implementation, the method comprises memorializing a called mutation.

[0098] In an implementation, the method comprises annotating a called mutation, e.g., annotating a called mutation with an indication of mutation structure, e.g., a mis-sense mutation, or function, e.g., a disease phenotype.

[0099] In an implementation, the method comprises acquiring nucleotide sequence reads for tumor and control nucleic acid.

[0100] In an implementation, the method comprises calling a nucleotide value, e.g., a variant, e.g., a mutation, for each of X subgenomic intervals, e.g., with a Bayesian calling method or a non-Bayesian calling method.

[0101] In an implementation, multiple samples, e.g., from different subjects, are processed simultaneously.

[0102] The methods disclosed herein can be used to detect alterations present in the genome or transcriptome of a subject, and can be applied to DNA and RNA sequencing, e.g., targeted RNA and / or DNA sequencing. Thus, another aspect featured herein includes methods for targeted RNA sequencing, e.g., sequencing of a cDNA derived from an RNA acquired from a sample, e.g., an FFPE-sample, to detect an alteration described herein. The alteration can be rearrangement, e.g., a rearrangement encoding a gene fusion. In other implementations, the method includes detection of a change (e.g., an increase or decrease) in the level of a gene or gene product, e.g., a change in expression of a gene or gene product described herein. The methods can, optionally, include a step of enriching a sample for a target RNA. In other implementations, the methods include the step of depleting the sample of certain high abundance RNAs, e.g., ribosomal or globin RNAs. The RNA sequencing methods can be used, alone or in combination with the DNA sequencing methods described herein. In one implementation, the method includes performing a DNA sequencing step and an RNA sequencing step. The methods can be performed in any order. For example, the method can include confirming by RNA sequencing the expression of an alteration described herein, e.g., confirming expression of mutation or a fusion detected by the DNA sequencing methods of the invention. In other implementations, the method includes performing an RNA sequencing step, followed by a DNA sequencing step.

[0103] In another aspect, also described herein but not claimed is a method comprising building a database of sequencing / alignment artifacts for the targeted subgenomic regions. In an implementation the database can be used to filter out spurious mutation calls and improve specificity. In an implementation the database is built by sequencing unrelated non-tumor (e.g., FFPE) samples or cell-lines and recording non-reference allele events that appear more frequently than expected due to random sequencing error alone in 1 or more of these normal samples. This approach may classify germ-line variation as artifact, but that is acceptable in method concerned with somatic mutations. This mis-classification of germ-line variation as artifact may be ameliorated if desired by filtering this database for known germ-line variation (removing common variants) and for artifacts that appear in only 1 individual (removing rarer variation).

[0104] Methods disclosed herein allow integration of a number of optimized elements including optimized bait-based selection, optimized alignment, and optimized mutation calling, as applied, e.g., to cancer related segments of the genome. Methods described herein provide for NGS-based analysis of tumors that can be optimized on a cancer-by-cancer, gene-by-gene and site-by-site basis. This can be applied e.g., to the genes / sites and tumor types described herein. The methods optimize levels of sensitivity and specificity for mutation detection with a given sequencing technology. Cancer by cancer, gene by gene, and site by site optimization provides very high levels sensitivity / specificity (e.g., >99% for both) that are essential for a clinical product.

[0105] Methods described herein provide for clinical and regulatory grade comprehensive analysis and interpretation of genomic aberrations for a comprehensive set of plausibly actionable genes (which may typically range from 50 to 500 genes) using next generation sequencing technologies from routine, real-world samples in order to inform optimal treatment and disease management decisions.

[0106] Methods described herein provide one-stop-shopping for oncologists / pathologists to send a tumor sample and receive a comprehensive analysis and description of the genomic and other molecular changes for that tumor, in order to inform optimal treatment and disease management decisions.

[0107] Methods described herein provide a robust, real-world clinical oncology diagnostic tool that takes standard available tumor samples and in one test provides a comprehensive genomic and other molecular aberration analysis to provide the oncologist with a comprehensive description of what aberrations may be driving the tumor and could be useful for informing the oncologists treatment decisions.

[0108] Methods described herein provide for a comprehensive analysis of a patient's cancer genome, with clinical grade quality. Methods include the most relevant genes and potential alterations and include one or more of the analysis of mutations, copy number, rearrangments, e.g., translocations, expression, and epigenetic markers. The out put of the genetic analysis can be contextualized with descriptive reporting of actionable results. Methods connect the use with an up to date set of relevant scientific and medical knowledge.

[0109] Methods described herein provide for increasing both the quality and efficiency of care. This includes applications where a tumor is of a rare or poorly studied type such that there is no standard of care or the patient is refractory to established lines of therapy and a rational basis for selection of further therapy or for clinical trial participation could be useful. E.g., methods allow, at any point of therapy, selection where the oncologist would benefit by having the full "molecular image" and / or "molecular sub-diagnosis" available to inform decision making.

[0110] Methods described herein can comprise providing a report, e.g., in electronic, web-based, or paper form, to the patient or to another person or entity, e.g., a caregiver, e.g., a physician, e.g., an oncologist, a hospital, clinic, third-party payor, insurance company or government office. The report can comprise output from the method, e.g., the identification of nucleotide values, the indication of presence or absence of an alteration, mutation, or wildtype sequence, e.g., for sugenomic intervals associated with a tumor of the type of the sample. The report can also comprise information on the role of a sequence, e.g., an alteration, mutation, or wildtype sequence, in disease. Such information can include information on prognosis, resistance, or potential or suggested therapeutic options. The report can comprise information on the likely effectiveness of a therapeutic option, the acceptability of a therapeutic option, or the advisability of applying the therapeutic option to a patient, e.g., a patient having a sequence, alteration or mutation identified in the test, and in implementations, identified in the report. E.g., the report can include information, or a recommendation on, the administration of a drug, e.g., the administration at a preselected dosage or in a preselected treatment regimen, e.g., in combination with other drugs, to the patient. In an implementation, not all mutations identified in the method are identified in the report. E.g., the report can be limited to mutations in genes having a preselected level of correlation with the occurrence, prognosis, stage, or susceptibility of the cancer to treatment, e.g., with a preselected therapeutic option. Methods featured herein allow for delivery of the report, e.g., to an entity described herein, within 7, 14, or 21 days from receipt of the sample by the entity practicing the method.

[0111] Thus, methods featured herein allow a quick turn around time, e.g., within 7, 14 or 21 days of receipt of sample.

[0112] Methods described herein can also be used to evaluate a histologically normal sample, e.g., samples from surgical margins. If one or more alterations as described herein is detected, the tissue can be re-classified, e.g., as malignant or pre-maligant, and / or the course of treatment can be modified.

[0113] In certain aspects, the sequencing methods described herein are useful in non-cancer applications, e.g., in forensic applications (e.g., identification as alternative to, or in addition to, use of dental records), paternity testing, and disease diagnosis and prognosis, e.g., for cystic fibrosis, Huntington's Disease, Alzheimer's Disease, among others. For example, identification of genetic alterations by the methods described herein can indicate the presence or risk of an individual for developing a particular disorder.

[0114] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. . In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

[0115] Other features and advantages of the invention will be apparent from the detailed description, drawings, and from the claims.DESCRIPTION OF THE FIGURES

[0116] The drawings are first described. FIG. 1A-Fis a flowchart depiction of an implementation of a method for multigene analysis of a tumor sample. FIG. 2 depicts the impact of prior expectation and read depth on mutation detection. FIG. 3 depicts the mutation frequencies in more than 100 clinical cancer samples. FIG. 4 is a linear representation of a coverage histogram. The number of targets (y-axis) are depicted as a function of coverage (x-axis). Line #1 represents the coverage using a bait set that includes biotinylated, array-derived RNA oligonucleotide baits spiked with biotinylated, individually synthesized DNA oligonucleotide baits (referred to herein as "Bait set #1"). Line #2 represents the coverage obtained using a bait set that includes biotinylated, array-derived RNA oligonucleotide baits only (referred to herein as "Bait set #2"). The overall average coverage using Bait set #2 was 924, whereas the coverage in areas of high GC content (about 68%) using Bait set #2 was 73. In contrast, when Bait set #1 was used, the overall coverage was about 918, but the coverage was improved to 183 in areas of high GC content. FIG. 5 is a coverage histogram comparing the uniformity in coverage detected with a bait set consisting of biotinylated, individually synthesized DNA oligonucleotide baits only (Bait set #1) and a bait set that includes biotinylated, array-derived RNA oligonucleotide baits spiked with biotinylated, individually synthesized DNA oligonucleotide baits ("Bait set #2"), compared to a bait set that includes biotinylated, array-derived RNA oligonucleotide baits only ("Bait set #3"). The bait sets are shown as #1, 2, and 3 in FIG. 5. Several gaps in coverage were detected using Bait set #3, but were not detected using Bait sets #1-2, as depicted in FIG. 5. DETAILED DESCRIPTION

[0117] Optimized methods and assays for sequencing large numbers of genes and gene products from samples, e.g., tumor samples, from one or more subjects by evaluating a selected group of genes and gene products are disclosed. In one implementation, the methods and assays featured herein are used in a multiplex assay format, e.g., assays incorporated multiple signals from a large number of diverse genetic events in a large number of genes. Disclosed herein are methods and assays that are based, at least in part, on a selected group of genes or gene products that are associated (e.g., positively or negatively) with a cancerous phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment or resistance to treatment). Such preselected genes or gene products enable the application of sequencing methods, particularly methods that rely on massively parallel sequencing of a large number of diverse genes, e.g., from tumor or control samples.

[0118] Certain terms are first defined. Additional terms are defined throughout the specification.

[0119] As used herein, the articles "a" and "an" refer to one or to more than one (e.g., to at least one) of the grammatical object of the article.

[0120] "About" and "approximately" shall generally mean an acceptable degree of error for the quantit: measured given the nature or precision of the measurements. Exemplary degrees of error are within 20 percent (%), typically, within 10%, and more typically, within 5% of a given value or range of values.

[0121] "Acquire" or "acquiring" as the terms are used herein, refer to obtaining possession of a physical entity, or a value, e.g., a numerical value, by "directly acquiring" or "indirectly acquiring" the physical entity or value. "Directly acquiring" means performing a process (e.g., performing a synthetic or analytical method) to obtain the physical entity or value. "Indirectly acquiring" refers to receiving the physical entity or value from another party or source (e.g., a third party laboratory that directly acquired the physical entity or value). Directly acquiring a physical entity includes performing a process that includes a physical change in a physical substance, e.g., a starting material. Exemplary changes include making a physical entity from two or ore starting materials, shearing or fragmenting a substance, separating or purifying a substance, .combining two or more separate entities into a mixture, performing a chemical reaction that includes breaking or forming a covalent or non covalent bond. Directly acquiring a value includes performing a process that includes a physical change in a sample or another substance, e.g., performing an analytical process which includes a physical change in a substance, e.g., a sample, analyte, or reagent (sometimes referred to herein as "physical analysis"), performing an analytical method, e.g., a method which includes one or more of the following: separating or purifying a substance, e.g., an analyte, or a fragment or other derivative thereof, from another substance; combining an analyte, or fragment or other derivative thereof, with another substance, e.g., a buffer, solvent, or reactant; or changing the structure of an analyte, or a fragment or other derivative thereof, e.g., by breaking or forming a covalent or non covalent bond, between a first and a second atom of the analyte; or by changing the structure of a reagent, or a fragment or other derivative thereof, e.g., by breaking or forming a covalent or non covalent bond, between a first and a second atom of the reagent.

[0122] "Acquiring a sequence" or "acquiring a read" as the term is used herein, refers to obtaining possession of a nucleotide sequence or amino acid sequence, by "directly acquiring" or "indirectly acquiring" the sequence or read. "Directly acquiring" a sequence or read means performing a process (e.g., performing a synthetic or analytical method) to obtain the sequence, such as performing a sequencing method (e.g., a Next Generation Sequencing (NGS) method). "Indirectly acquiring" a sequence or read refers to receiving information or knowledge of, or receiving, the sequence from another party or source (e.g., a third party laboratory that directly acquired the sequence). The sequence or read acquired need not be a full sequence, e.g., sequencing of at least one nucleotide, or obtaining information or knowledge, that identifies one or more of the alterations disclosed herein as being present in a subject constitutes acquiring a sequence.

[0123] Directly acquiring a sequence or read includes performing a process that includes a physical change in a physical substance, e.g., a starting material, such as a tissue or cellular sample, e.g., a biopsy, or an isolated nucleic acid (e.g., DNA or RNA) sample. Exemplary changes include making a physical entity from two or more starting materials, shearing or fragmenting a substance, such as a genomic DNA fragment; separating or purifying a substance (e.g., isolating a nucleic acid sample from a tissue); combining two or more separate entities into a mixture, performing a chemical reaction that includes breaking or forming a covalent or non-covalent bond. Directly acquiring a value includes performing a process that includes a physical change in a sample or another substance as described above.

[0124] "Acquiring a sample" as the term is used herein, refers to obtaining possession of a sample, e.g., a tissue sample or nucleic acid sample, by "directly acquiring" or "indirectly acquiring" the sample. "Directly acquiring a sample" means performing a process (e.g., performing a physical method such as a surgery or extraction) to obtain the sample. "Indirectly acquiring a sample" refers to receiving the sample from another party or source (e.g., a third party laboratory that directly acquired the sample). Directly acquiring a sample includes performing a process that includes a physical change in a physical substance, e.g., a starting material, such as a tissue, e.g., a tissue in a human patient or a tissue that has was previously isolated from a patient. Exemplary changes include making a physical entity from a starting material, dissecting or scraping a tissue; separating or purifying a substance (e.g., a sample tissue or a nucleic acid sample); combining two or more separate entities into a mixture; performing a chemical reaction that includes breaking or forming a covalent or non-covalent bond. Directly acquiring a sample includes performing a process that includes a physical change in a sample or another substance, e.g., as described above.

[0125] "Alignment selector," as used herein, refers to a parameter that allows or directs the selection of an alignment method, e.g., an alignment algorithm or parameter, that can optimize the sequencing of a preselected subgenomic interval. An alignment selector can be specific to, or selected as a function, e.g., of one or more of the following: 1. The sequence context, e.g., sequence context, of a subgenomic interval (e.g., the preselected nucleotide position to be evaluated) that is associated with a propensity for misalignment of reads for said subgenomic interval. E.g., the existence of a sequence element in or near the subgenomic interval to be evaluated that is repeated elsewhere in the genome can cause misalignment and thereby reduce performance. Performance can be enhanced by selecting an algorithm or an algorithm parameter that minimizes misalignment. In this case the value for the alignment selector can be a function of the sequence context, e.g., the presence or absence of a sequence of preselected length that is repeated at least a preselected number of times in the genome (or in the portion of the genome being analyzed). 2. The tumor type being analyzed. E.g., a specific tumor type can be characterized by increased rate of deletions. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter that is more sensitive to indels. In this case the value for the alignment selector can be a function of the tumor type, e.g., an identifier for the tumor type. In an implementation the value is the identity of the tumor type, e.g., breast cancer. 3. The gene, or type of gene, being analyzed, e.g., a gene, or type of gene, can be analyzed. Oncogenes, by way of example, are often characterized by substitutions or in-frame indels. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter that is particularly sensitive to these variants and specific against others. Tumor suppressors are often characterized by frame-shift indels. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter that is particularly sensitive to these variants. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter matched with the subgenomic interval. In this case the value for the alignment selector can be a function of the gene or gene type, e.g., an identifier for gene or gene type. In an implementation the value is the identity of the gene. 4. The site (e.g., nucleotide position) being analyzed. In this case the value for the alignment selector can be a function of the site or the type of site, e.g., an identifier for the site or site type. In an implementation the value is the identity of the site. (E.g., if the gene containing the site is highly homologous with another gene, normal / fast short read alignment algorithms (e.g., BWA) may have difficulty distinguishing between the two genes, potentially necessitating more intensive alignment methods (Smith-Waterman) or even assembly (ARACHNE). Similarly, if the gene sequence contains low-complexity regions (e.g., AAAAAA), more intensive alignment methods may be necessary. 5. The variant, or type of variant, associated with the subgenomic interval being evaluated. E.g., a substitution, insertion, deletion, translocation or other rearrangement. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter that is more sensitive to the specific variant type. In this case the value for the alignment selector can he a function of the type of variant, e.g., an identifier for the type of variant. In an implementation the value is the identity of the type of variant, e.g., a substitution. 6. The type of sample, a FFPE or other fixed sample. Sample type / quality can affect error (spurious observation of non-reference sequence) rate. Thus, performance can be enhanced by selecting an algorithm or algorithm parameter that accurately model the true error rate in the sample. In this case the value for the alignment selector can be a function of the type of sample, e.g., an identifier for the sample type. In an implementation, the value is the identity of the sampe type, e.g., a fixed sample.

[0126] "Alteration" or "altered structure" as used herein, of a gene or gene product (e.g., a marker gene or gene product) refers to the presence of a mutation or mutations within the gene or gene product, e.g., a mutation, which affects amount or activity of the gene or gene product, as compared to the normal or wild-type gene. The alteration can be in amount, structure, and / or activity in a cancer tissue or cancer cell, as compared to its amount, structure, and / or activity, in a normal or healthy tissue or cell (e.g., a control), and is associated with a disease state, such as cancer. For example, an alteration which is associated with cancer, or predictive of responsiveness to anti-cancer therapeutics, can have an altered nucleotide sequence (e.g., a mutation), amino acid sequence, chromosomal translocation, intra-chromosomal inversion, copy number, expression level, protein level, protein activity, or methylation status, in a cancer tissue or cancer cell, as compared to a normal, healthy tissue or cell. Exemplary mutations include, but are not limited to, point mutations (e.g., silent, missense, or nonsense), deletions, insertions, inversions, linking mutations, duplications, translocations, inter- and intra-chromosomal rearrangements. Mutations can be present in the coding or non-coding region of the gene. In certain implementations, the alteration(s) is detected as a rearrangement, e.g., a genomic rearrangement comprising one or more introns or fragments thereof (e.g., one or more rearrangements in the 5'- and / or 3'-UTR). In certain implementations, the alterations are associated (or not associated) with a phenotype, e.g., a cancerous phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment or resistance to cancer treatment). In one implementation, the alteration is associated with one or more of: a genetic risk factor for cancer, a positive treatment response predictor, a negative treatment response predictor, a positive prognostic factor, a negative prognostic factor, or a diagnostic factor.

[0127] "Bait", as used herein, is type of hybrid capture reagent. A bait can be a nucleic acid molecule, e.g., a DNA or RNA molecule, which can hybridize to (e.g., be complementary to), and thereby allow capture of a target nucleic acid. In one implementation, a bait is an RNA molecule (e.g., a naturally-occurring or modified RNA molecule); a DNA molecule (e.g., a naturally-occurring or modified DNA molecule), or a combination thereof. In other implementations, a bait includes a binding entity, e.g., an affinity tag, that allows capture and separation, e.g., by binding to a binding entity, of a hybrid formed by a bait and a nucleic acid hybridized to the bait. In one implementation, a bait is suitable for solution phase hybridization.

[0128] "Bait set," as used herein, refers to one or a plurality of bait molecules.

[0129] "Binding entity" means any molecule to which molecular tags can be directly or indirectly attached that is capable of specifically binding to an analyte. The binding entity can be an affinity tag on each bait sequence. In certain implementations, the binding entity allows for separation of the bait / member hybrids from the hybridization mixture by binding to a partner, such as an avidin molecule, or an antibody that binds to the hapten or an antigen-binding fragment thereof. Exemplary binding entities include, but are not limited to, a biotin molecule, a hapten, an antibody, an antibody binding fragment, a peptide, and a protein.

[0130] "Complementary" refers to sequence complementarity between regions of two nucleic acid strands or between two regions of the same nucleic acid strand. It is known that an adenine residue of a first nucleic acid region is capable of forming specific hydrogen bonds ("base pairing") with a residue of a second nucleic acid region which is antiparallel to the first region if the residue is thymine or uracil. Similarly, it is known that a cytosine residue of a first nucleic acid strand is capable of base pairing with a residue of a second nucleic acid strand which is antiparallel to the first strand if the residue is guanine. A first region of a nucleic acid is complementary to a second region of the same or a different nucleic acid if, when the two regions are arranged in an antiparallel fashion, at least one nucleotide residue of the first region is capable of base pairing with a residue of the second region. In certain implementations, the first region comprises a first portion and the second region comprises a second portion, whereby, when the first and second portions are arranged in an antiparallel fashion, at least about 50%, at least about 75%, at least about 90%, or at least about 95% of the nucleotide residues of the first portion are capable of base pairing with nucleotide residues in the second portion. In other implementations, all nucleotide residues of the first portion are capable of base pairing with nucleotide residues in the second portion.

[0131] The term "cancer" or "tumor" is used interchangeably herein. These terms refer to the presence of cells possessing characteristics typical of cancer-causing cells, such as uncontrolled proliferation, immortality, metastatic potential, rapid growth and proliferation rate, and certain characteristic morphological features. Cancer cells are often in the form of a tumor, but such cells can exist alone within an animal, or can be a non-tumorigenic cancer cell, such as a leukemia cell. These terms include a solid tumor, a soft tissue tumor, or a metastatic lesion. As used herein, the term "cancer" includes premalignant, as well as malignant cancers.

[0132] "Likely to" or "increased likelihood," as used herein, refers to an increased probability that an item, object, thing or person will occur. Thus, in one example, a subject that is likely to respond to treatment has an increased probability of responding to treatment relative to a reference subject or group of subjects.

[0133] "Unlikely to" refers to a decreased probability that an event, item, object, thing or person will occur with respect to a reference. Thus, a subject that is unlikely to respond to treatment has a decreased probability of responding to treatment relative to a reference subject or group of subjects.

[0134] "Control member" refers to a member having sequence from a non-tumor cell.

[0135] "Indel alignment sequence selector," as used herein, refers to a parameter that allows or directs the selection of a sequence to which a read is to be aligned with in the case of a preselcted indel. Use of such a sequence can optimize the sequencing of a preselected subgenomic interval comprising an indel. The value for an indel alignment sequence selector is a function of a preselected indel, e.g., an identifier for the indel. In an implementation the value is the identity of the indel.

[0136] As used herein, the term "library" refers to a collection of members. In one implementation, the library includes a collection of nucleic acid members, e.g., a collection of whole genomic, subgenomic fragments, cDNA, cDNA fragments, RNA, RNA fragments, or a combination thereof. In one implementation, a portion or all of the library members comprises an adapter sequence. The adapter sequence can be located at one or both ends. The adapter sequence can be useful, e.g., for a sequencing method (e.g., an NGS method), for amplification, for reverse transcription, or for cloning into a vector.

[0137] The libary can comprise a collection of members, e.g., a target member (e.g., a tumor member, a reference member, a PGx member or a combination thereof). The members of the library can be from a single individual. In implementations, a library can comprise members from more than one subject (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30 or more subjects), e.g., two or more libraries from different subjects can be combined to from a library having members from more than one subject. In one implementation, the subject is human having, or at risk of having, a cancer or tumor.

[0138] "Library-catch" refers to a subset of a library, e.g., a subset enriched for preselected subgenomic intervals, e.g., product captured by hybridization with preselected baits.

[0139] "Member" or "library member" or other similar term, as used herein, refers to a nucleic acid molecule, e.g., a DNA, RNA, or a combination thereof, that is the member of a library. Typically, a member is a DNA molecule, e.g., genomic DNA or cDNA. A member can be fragmented, e.g., sheared or enzymatically prepared, genomic DNA. Members comprise sequence from a subject and can also comprise sequence not derived from the subject, e.g., adapters sequence, a primer sequence, or other sequences that allow for identification, e.g., "barcode" sequences.

[0140] "Next-generation sequencing or NGS or NG sequencing" as used herein, refers to any sequencing method that determines the nucleotide sequence of either individual nucleic acid molecules (e.g., in single molecule sequencing) or clonally expanded proxies for individual nucleic acid molecules in a high through-putfashion (e.g., greater than 10 3< , 10 4< , 10 5< or more molecules are sequenced simultaneously). In one implementation, the relative abundance of the nucleic acid species in the library can be estimated by counting the relative number of occurrences of their cognate sequences in the data generated by the sequencing experiment. Next generation sequencing methods are known in the art, and are described, e.g., in Metzker, M. (2010) Nature Biotechnology Reviews 11:31-46. Next generation sequencing can detect a variant present in less than 5% of the nucleic acids in a sample.

[0141] "Nucleotide value" as referred herein, represents the identity of the nucleotide(s) occupying or assigned to a preselected nucleotide position. Typical nucleotide values include: missing (e.g., deleted); addtional (e.g., an insertion of one or more nucleotides, the identity of which may or may not be included); or present (occupied); A; T; C; or G. Other values can be, e.g., not Y, wherein Y is A, T, G, or C; A or X, wherein X is one or two of T, G, or C; T or X, wherein X is one or two of A, G, or C; G or X, wherein X is one or two of T, A, or C; C or X, wherein X is one or two of T, G, or A; a pyrimidine nucleotide; or a purine nucleotide. A nucleotide value can be a frequency for 1 or more, e.g., 2, 3, or 4, bases (or other value described herein, e.g., missing or additional) at a nucleotide position. E.g., a nucleotide value can comprise a frequency for A, and a frequency for G, at a nucleotide position.

[0142] "Or" is used herein to mean, and is used interchangeably with, the term "and / or", unless context clearly indicates otherwise. The use of the term "and / or" in some places herein does not mean that uses of the term "or" are not interchangeable with the term "and / or" unless the context clearly indicates otherwise.

[0143] "Primary control" refers to a non tumor tissue other than NAT tissue in a tumor sample. Blood is a typical primary control.

[0144] "Rearrangement alignment sequence selector," as used herein, refers to a parameter that allows or directs the selection of a sequence to which a read is to be aligned with in the case of a preselected rearrangement. Use of such a sequence can optimize the sequencing of a preselected subgenomic interval comprising a rearrangement. The value for a rearrangement alignment sequence selector is a function of a preselected rearrangement, e.g., an identifier for the rearrangement. In an implementation the value is the identity of the rearrangement. An "indel alignment sequence selector" (also defined elsewhere herein) is an example of a rearrangement alignment sequence selector.

[0145] "Sample," "tissue sample," "patient sample," "patient cell or tissue sample" or "specimen" each refers to a collection of similar cells obtained from a tissue, or circulating cells, of a subject or patient. The source of the tissue sample can be solid tissue as from a fresh, frozen and / or preserved organ, tissue sample, biopsy, or aspirate; blood or any blood constituents; bodily fluids such as cerebral spinal fluid, amniotic fluid, peritoneal fluid or interstitial fluid; or cells from any time in gestation or development of the subject. The tissue sample can contain compounds that are not naturally intermixed with the tissue in nature such as preservatives, anticoagulants, buffers, fixatives, nutrients, antibiotics or the like. In one implementation, the sample is preserved as a frozen sample or as formaldehyde- or paraformaldehyde-fixed paraffin-embedded (FFPE) tissue preparation. For example, the sample can be embedded in a matrix, e.g., an FFPE block or a frozen sample.

[0146] In one implementation, the sample is a tumor sample, e.g., includes one or more premalignant or malignant cells. In certain, implementations, the sample, e.g., the tumor sample, is acquired from a solid tumor, a soft tissue tumor or a metastatic lesion. In other implementations, the sample, e.g., the tumor sample, includes tissue or cells from a surgical margin. In another implementation embediment, the sample, e.g., tumor sample, includes one or more circulating tumor cells (CTC) (e.g., a CTC acquired from a blood sample).

[0147] "Sensitivity," as used herein, is a measure of the ability of a method to detect a preselected sequence variant in a heterogeneous population of sequences. A method has a sensitivity of S% for variants of F% if, given a sample in which the preselected sequence variant is present as at least F% of the sequences in the sample, the method can detect the preselected sequence at a preselected confidence of C%, S% of the time. By way of example, a method has a sensitivity of 90% for variants of 5% if, given a sample in which the preselected variant sequence is present as at least 5% of the sequences in the sample, the method can detect the preselected sequence at a preselected confidence of 99%, 9 out of 10 times (F=5%; C=99%; S=90%). Exemplary sensitivities include those of S=90%, 95%, 99% for sequence variants at F=1%, 5%, 10%, 20%, 50%, 100% at confidence levels of C= 90%, 95%, 99%, and 99.9%.

[0148] "Specificity," as used herein, is a measure of the ability of a method to distinguish a truly occurring preselected sequence variant from sequencing artifacts or other closely related sequences. It is the ability to avoid false positive detections. False positive detections can arise from errors introduced into the sequence of interest during sample preparation, sequencing error, or inadvertent sequencing of closely related sequences like pseudo-genes or members of a gene family. A method has a specificity of X % if, when applied to a sample set of N Total sequences, in which X True sequences are truly variant and X Not true are not truly variant, the method selects at least X % of the not truly variant as not variant. E.g., a method has a specificity of 90 % if, when applied to a sample set of 1,000 sequences, in which 500 sequences are truly variant and 500 are not truly variant, the method selects 90 % of the 500 not truly variant sequences as not variant. Exemplary specificities include 90, 95, 98, and 99 %.

[0149] A "tumor nucleic acid sample" as used herein, refers to nucleic acid molecules from a tumor or cancer sample. Typically, it is DNA, e.g., genomic DNA, or cDNA derived from RNA, from a tumor or cancer sample. In certain implementations, the tumor nucleic acid sample is purified or isolated (e.g., it is removed from its natural state).

[0150] A "control" or "reference" "nucleic acid sample" as used herein, refers to nucleic acid molecules from a control or reference sample. Typically, it is DNA, e.g., genomic DNA, or cDNA derived from RNA, not containing the alteration or variation in the gene or gene product. In certain implementations, the reference or control nucleic acid sample is a wild type or a non-mutated sequence. In certain implementations, the reference nucleic acid sample is purified or isolated (e.g., it is removed from its natural state). In other implementations, the reference nucleic acid sample is from a non-tumor sample, e.g., a blood control, a normal adjacent tumor (NAT), or any other non-cancerous sample from the same or a different subject.

[0151] "Sequencing" a nucleic acid molecule requires determining the identity of at least 1 nucleotide in the molecule. In implementations the identity of less than all of the nucleotides in a molecule are determined. In other implementations, the identity of a majority or all of the nucleotides in the molecule is determined.

[0152] "Subgenomic interval" as referred to herein, refers to a portion of genomic sequence. In an implementation a subgenomic interval can be a single nucleotide position, e.g., a nucleotide position variants of which are associated (positively or negatively) with a tumor phenotype. In an implementation a subgenomic interval comprises more than one nucleotide position. Such implementations include sequences of at least 2, 5, 10, 50, 100, 150, or 250 nucleotide positions in length. Subgenomic intervals can comprise an entire gene, or a preselected portion thereof, e.g., the coding region (or portions there of), a preselected intron (or portion thereof) or exon (or portion thereof). A subgenomic interval can comprise all or a part of a fragment of a naturally occurring, e.g., genomic, nucleic acid. E.g., a subgenomic interval can correspond to a fragment of genomic DNA which is subjected to a sequencing reaction. In implementations a subgenomic interval is continuous sequence from a genomic source. In implementations a subgenomic interval includes sequences that are not contiguous in the genome, e.g., it can include junctions formed found at exon-exon junctions in cDNA.

[0153] In an implementation, a subgenomic interval comprises or consists of: a single nucleotide position; an intragenic region or an intergenic region; an exon or an intron, or a fragment thereof, typically an exon sequence or a fragment thereof; a coding region or a non-coding region, e.g., a promoter, an enhancer, a 5' untranslated region (5' UTR), or a 3' untranslated region (3' UTR), or a fragment thereof; a cDNA or a fragment thereof; an SNP; a somatic mutation, a germ line mutation or both; an alteration, e.g., a point or a single mutation; a deletion mutation (e.g., an in-frame deletion, an intragenic deletion, a full gene deletion); an insertion mutation (e.g., intragenic insertion); an inversion mutation (e.g., an intra-chromosomal inversion); a linking mutation; a linked insertion mutation; an inverted duplication mutation; a tandem duplication (e.g., an intrachromosomal tandem duplication); a translocation (e.g., a chromosomal translocation, a non-reciprocal translocation); a rearrangement (e.g., a genomic rearrangement (e.g., a rearrangement of one or more introns, or a fragment thereof; a rearranged intron can include a a 5'- and / or 3'- UTR); a change in gene copy number; a change in gene expression; a change in RNA levels, or a combination thereof. The "copy number of a gene" refers to the number of DNA sequences in a cell encoding a particular gene product. Generally, for a given gene, a mammal has two copies of each gene. The copy number can be increased, e.g., by gene amplification or duplication, or reduced by deletion.

[0154] "Threshold value," as used herein, is a value that is a function of the number of reads required to be present to assign a nucleotide value to a subgenomic interval. E.g., it is a function of the number of reads having a specific nucleotide value, e.g., A, at a nucleotide position, required to assign that nucleotide value to that nucleotide position in the subgenomic interval. The threshold value can, e.g., be expressed as (or as a function of) a number of reads, e.g., an integer, or as a proportion of reads having the preselected value. By way of example, if the threshold value is X, and X+1 reads having the nucleotide value of "A" are present, then the value of "A" is assigned to the preselected position in the subgenomic interval. The threshold value can also be expressed as a function of a mutation or variant expectation, mutation frequency, or of Bayesian prior. In an implementation, a preselected mutation frequency would require a preselected number or proportion of reads having a nucleotide value, e.g., A or G, at a preselected position, to call that that nucleotide value. In implementations the threshold value can be a function of mutation expectation, e.g., mutation frequency, and tumor type. E.g., a preslected variant at a preselected nucleotide position could have a first threshold value if the patient has a first tumor type and a second threshold value if the patient has a second tumor type.

[0155] As used herein, "target member" refers to a nucleic acid molecule that one desires to isolate from the nucleic acid library. In one implementation, the target members can be a tumor member, a reference member, a control member, or a PGx member as described herein.

[0156] "Tumor member," or other similar term (e.g., a "tumor or cancer-associated member"), as used herein refers to a member having sequence from a tumor cell. In one implementation, the tumor member includes a subgenomic interval having a sequence (e.g., a nucleotide sequence) that has an alteration (e.g., a mutation) associated with a cancerous phenotype. In other implementations, the tumor member includes a subgenomic interval having a wild type sequence (e.g., a wild type nucleotide sequence). For example, a subgenomic interval from a heterozygous or homozygous wild type allele present in a cancer cell. A tumor member can include a reference member or a PGx member.

[0157] "Reference member," or other similar term (e.g., a "control member"), as used herein, refers to a member that comprises a subgenomic interval having a sequence (e.g., a nucleotide sequence) that is not associated with the cancerous phenotype. In one implementation, the reference member includes a wild-type or a non-mutated nucleotide sequence of a gene or gene product that when mutated is associated with the cancerous phenotype. The reference member can be present in a cancer cell or non-cancer cell.

[0158] "PGx member" or other similar term, as used herein, refers to a member that comprises a subgenomic interval that is associated with the pharmacogenetic or pharmacogenomic profile of a gene. In one implementation, the PGx member includes an SNP (e.g., an SNP as described herein). In other implementations, the PGx member includes a subgenomic interval according to Table 1 or Table 2.

[0159] "Variant," as used herein, refers to a structure that can be present at a subgenomic interval that can have more than one structure, e.g., an allele at a polymorphic locus.

[0160] Headings, e.g., (a), (b), (i) etc, are presented merely for ease of reading the specification and claims. The use of headings in the specification or claims does not require the steps or elements be performed in alphabetical or numerical order or the order in which they are presented.Selection of Gene or Gene Products

[0161] The selected genes or gene products (also referred to herein as the "target genes or gene products") can include subgenomic intervals comprising intragenic regions or intergenic regions. For example, the subgenomic interval can include an exon or an intron, or a fragment thereof, typically an exon sequence or a fragment thereof. The subgenomic interval can include a coding region or a non-coding region, e.g., a promoter, an enhancer, a 5' untranslated region (5' UTR), or a 3' untranslated region (3' UTR), or a fragment thereof. In other implementations, the subgenomic interval includes a cDNA or a fragment thereof. In other implementations, the subgenomic interval includes an SNP, e.g., as described herein.

[0162] In other implementations, the subgenomic intervals include substantially all exons in a genome, e.g., one or more of the subgenomic intervals as described herein (e.g., exons from selected genes or gene products of interest (e.g., genes or gene products associated with a cancerous phenotype as described herein)). In one implementation, the subgenomic interval includes a somatic mutation, a germ line mutation or both. In one implementation, the subgenomic interval includes an alteration, e.g., a point or a single mutation, a deletion mutation (e.g., an in-frame deletion, an intragenic deletion, a full gene deletion), an insertion mutation (e.g., intragenic insertion), an inversion mutation (e.g., an intra-chromosomal inversion), a linking mutation, a linked insertion mutation, an inverted duplication mutation, a tandem duplication (e.g., an intrachromosomal tandem duplication), a translocation (e.g., a chromosomal translocation, a non-reciprocal translocation). a rearrangement, a change in gene copy number, or a combination thereof. In certain implementations, the subgenomic interval constitutes less than 5, 1, 0.5, 0.1%. 0.01%. 0.001% of the coding region of the genome of the tumor cells in a sample. implementations In other embodiments, the subgenomic intervals are not involved in a disease, e.g., are not associated with a cancerous phenotype as described herein.

[0163] In one implementation, the target gene or gene product is a biomarker. As used herein, a "biomarker" or "marker" is a gene, mRNA, or protein which can be altered, wherein said alteration is associated with cancer. The alteration can be in amount, structure, and / or activity in a cancer tissue or cancer cell, as compared to its amount, structure, and / or activity, in a normal or healthy tissue or cell (e.g., a control), and is associated with a disease state, such as cancer. For example, a marker associated with cancer, or predictive of responsiveness to anti-cancer therapeutics, can have an altered nucleotide sequence, amino acid sequence, chromosomal translocation, intra-chromosomal inversion, copy number, expression level, protein level, protein activity, or methylation status, in a cancer tissue or cancer cell as compared to a normal, healthy tissue or cell. Furthermore, a "marker" includes a molecule whose structure is altered, e.g., mutated (contains an mutation), e.g., differs from the wild type sequence at the nucleotide or amino acid level, e.g., by substitution, deletion, or insertion, when present in a tissue or cell associated with a disease state, such as cancer.

[0164] In one implementation, the target gene or gene product includes a single-nucleotide polymorphism (SNP). In another implementation, the gene or gene product has a small deletion, e.g., a small intragenic deletion (e.g., an in-frame or frame-shift deletion). In yet another implementation, the target sequence results from the deletion of an entire gene. In still another implementation, the target sequence has a small insertion, e.g., a small intragenic insertion. In one implementation, the target sequence results from an inversion, e.g., an intrachromosal inversion. In another implementation, the target sequence results from an interchromosal translocation. In yet another implementation, the target sequence has a tandem duplication. In one implementation, the target sequence has an undesirable feature (e.g., high GC content or repeat element). In another implementation, the target sequence has a portion of nucleotide sequence that cannot itself be successfully targeted, e.g., because of its repetitive nature. In one implementation, the target sequence results from alternative splicing. In another implementation, the target sequence is chosen from a gene or gene product, or a fragment thereof according to Table 1, 1A, 2, 3, or 4.

[0165] Cancers include, but are not limited to, B cell cancer, e.g., multiple myeloma, melanomas, breast cancer, lung cancer (such as non-small cell lung carcinoma or NSCLC), bronchus cancer, colorectal cancer, prostate cancer, pancreatic cancer, stomach cancer, ovarian cancer, urinary bladder cancer, brain or central nervous system cancer, peripheral nervous system cancer, esophageal cancer, cervical cancer, uterine or endometrial cancer, cancer of the oral cavity or pharynx, liver cancer, kidney cancer, testicular cancer, biliary tract cancer, small bowel or appendix cancer, salivary gland cancer, thyroid gland cancer, adrenal gland cancer, osteosarcoma, chondrosarcoma, cancer of hematological tissues, adenocarcinomas, inflammatory myofibroblastic tumors, gastrointestinal stromal tumor (GIST), colon cancer, multiple myeloma (MM), myelodysplastic syndrome (MDS), myeloproliferative disorder (MPD), acute lymphocytic leukemia (ALL), acute myelocytic leukemia (AML), chronic myelocytic leukemia (CML), chronic lymphocytic leukemia (CLL), polycythemia Vera, Hodgkin lymphoma, non-Hodgkin lymphoma (NHL), soft-tissue sarcoma, fibrosarcoma, myxosarcoma, liposarcoma, osteogenic sarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, mesothelioma, Ewing's tumor, leiomyosarcoma, rhabdomyosarcoma, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinomas, medullary carcinoma, bronchogenic carcinoma, renal cell carcinoma, hepatoma, bile duct carcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilms' tumor, bladder carcinoma, epithelial carcinoma, glioma, astrocytoma, medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, neuroblastoma, retinoblastoma, follicular lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, hepatocellular carcinoma, thyroid cancer, gastric cancer, head and neck cancer, small cell cancers, essential thrombocythemia, agnogenic myeloid metaplasia, hypereosinophilic syndrome, systemic mastocytosis, familiar hypereosinophilia, chronic eosinophilic leukemia, neuroendocrine cancers, carcinoid tumors, and the like.

[0166] In one implementation, the target gene or gene product is chosen a full length, or a fragment thereof, selected from the group consisting of ABCB1, ABCC2, ABCC4, ABCG2, ABL1, ABL2, AKT1, AKT2, AKT3, ALK, APC, AR, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRAF, BRCA1, BRCA2, C1orf144, CARD11, CBL, CCND1, CCND2, CCND3, CCNE1, CDH1, CDH2, CDH20, CDH5, CDK4, CDK6, CDK8, CDKN2A, CDKN2B, CDKN2C, CEBPA, CHEK1, CHEK2, CRKL, CRLF2, CTNNB1, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DNMT3A, DOT1L, DPYD, EGFR, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB2, ERBB3, ERBB4, ERCC2, ERG, ESR1, ESR2, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FCGR3A, FGFR1, FGFR2, FGFR3, FGFR4, FLT1, FLT3, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GSTP1, GUCY1A2, HOXA3, HRAS, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, ITPA, JAK1, JAK2, JAK3, JUN, KDR, KIT, KRAS, LRP1B, LRP2, LTK, MAN1B1, MAP2K1, MAP2K2, MAP2K4, MCL1, MDM2, MDM4, MEN1, MET, MITF, MLH1, MLL, MPL, MRE11A, MSH2, MSH6, MTHFR, MTOR, MUTYH, MYC, MYCL1, MYCN, NF1, NF2, NKX2-1, NOTCH1, NPM1, NQO1, NRAS, NRP2, NTRK1, NTRK3, PAK3, PAX5, PDGFRA, PDGFRB, PIK3CA, PIK3R1, PKHD1, PLCG1, PRKDC, PTCH1, PTEN, PTPN11, PTPRD, RAF1, RARA, RB1, RET, RICTOR, RPTOR, RUNX1, SLC19A1, SLC22A2, SLCO1B3, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SMO, SOD2, SOX10, SOX2, SRC, STK11, SULT1A1, TBX22, TET2, TGFBR2, TMPRSS2, TOP1, TP53, TPMT, TSC1, TSC2, TYMS, UGT1A1, UMPS, USP9X, VHL, and WT1.

[0167] In one implementation, the target gene or gene product, or a fragment thereof, has one or more SNPs that are relevant to pharmacogenetics and pharmacogenomics (PGx), e.g., drug metabolism and toxicity. Exemplary genes or gene products include, but not limited to, ABCB1, ABCC2, ABCC4, ABCG2, C1orf144, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DPYD, ERCC2, ESR2, FCGR3A, GSTP1, ITPA, LRP2, MAN1B1, MTHFR, NQO1, NRP2, SLC19A1, SLC22A2, SLCO1B3, SOD2, SULT1A1, TPMT, TYMS, UGT1A1, and UMPS.

[0168] In another implementation, the target gene or gene product, or a fragment thereof, has one or more codons that are associated with cancer. Exemplary genes or gene products include, but not limited to, ABL1 (e.g., codon 315), AKT1, ALK, APC (e.g., codon 1114, 1338, 1450, and 1556), AR, BRAF (e.g., codon 600), CDKN2A, CEBPA, CTNNB1 (e.g., codon 32, 33, 34, 37, 41, and 45), EGFR (e.g., 719, 746-750, 768, 790, 858, and 861), ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3 (e.g., codon 835), HRAS (e.g., codon 12, 13, and 61), JAK2 (e.g., codon 617), KIT (e.g., codon 816), KRAS (e.g., codon 12, 13, and 61), MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, PDGFRA, PIK3CA (e.g., codon 88, 542, 545, 546, 1047, and 1049), PTEN (e.g., codon 130, 173, 233, and 267), RB1, RET (e.g., codon 918), TP53 (e.g., 175 245, 248, 273, and 306).

[0169] In yet another implementation, the target gene or gene product, or a fragment thereof, are associated with cancer. Exemplary genes or gene products include, but not limited to, ABL2, AKT2, AKT3, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CARD11, CBL, CCND1, CCND2, CCND3, CCNE1, CDH1, CDH2, CDH20, CDH5, CDK4, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDR, LRP1B, LTK, MAP2K1, MAP2K2, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK3, PAK3, PAX5, PDGFRB, PIK3R1, PKHD1, PLCG1, PRKDC, PTCH1, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SMO, SOX10, SOX2, SRC, STK11, TBX22, TET2, TGFBR2, TMPRSS2, TOP1, TSC1, TSC2, USP9X, VHL, and WT1.

[0170] Applications of the foregoing methods include using a library of oligonucleotides containing all known sequence variants (or a subset thereof) of a particular gene or genes for sequencing in medical specimens.Gene Selection Module

[0171] This module discloses sets of subgenomic intervals for use in methods featured herein, e.g., subgenomic intervals for sets or groups of genes and other regions described herein.

[0172] Optimized methods and assays for sequencing large numbers of genes and gene products from samples, e.g., tumor samples, from one or more subjects are disclosed. In one implementation, the methods and assays featured herein are used in a multiplex, multi-gene assay format, e.g., assays that incorporate multiple signals from a large number of diverse genetic events in a large number of genes. Disclosed herein are methods and assays that are based, at least in part, on a pre-selected set of genes or gene products that are associated (e.g., positively or negatively) with a cancerous phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment response or resistance to cancer treatment). Such pre-selected genes or gene products enable the application of sequencing methods, particularly methods that rely on massively parallel sequencing of a large number of diverse genes, e.g., from tumor or control samples.

[0173] Accordingly, also described herein but not claimed is a method of analyzing a sample, e.g., a tumor sample. The method comprises: (a) acquiring a library comprising a plurality members from a sample, e.g., a plurality of tumor members from a tumor sample; (b) optionally, enriching the library for preselected sequences, e.g., by contacting the library with a bait set (or plurality of bait sets) to provide selected members (sometimes referred to herein as library catch); (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein; and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method or a method described herein) from said read for the preselected nucleotide position, thereby analyzing said tumor sample, wherein the method comprises sequencing, e.g., by a next generation sequencing method, a subgenomic interval from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53.

[0174] In an implementation, step (b) is present. In an implementation, step (b) is absent.

[0175] Thus, in implementations a method comprises sequencing, e.g., by a next generation sequencing method, a subgenomic interval from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the acquired nucleic acid sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53, thereby analyzing the tumor sample.

[0176] In certain implementations, the method, or the assay, further includes sequencing a subgenomic interval from a gene or gene product chosen from one, two, three, four, five, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1. TSC1, TSC2, USP9X, VHL, or WT1.

[0177] In other implementations, the method, or the assay, further includes sequencing a subgenomic interval that is present in a gene or gene product associated with one or more of drug metabolism, drug responsiveness, or toxicity (also referred to therein as "PGx" genes). In certain implementations, the subgenomic interval sequenced includes an alteration (e.g., single nucleotide polymorphism (SNP)). In one implementation, the subgenomic interval sequenced is from a gene or gene product chosen from one, two, three, four, five, ten, fifteen, twenty, twenty-five, thirty or more of: ABCB1, BCC2, ABCC4, ABCG2, C1orf144, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DPYD, ERCC2, ESR2, FCGR3A, GSTP1, ITPA, LRP2, MAN1B1, MTHFR, NQO1, NRP2, SLC19A1, SLC22A2, SLCO1B3, SOD2, SULT1A1, TPMT, TYMS, UGT1A1, or UMPS.

[0178] In other implementations, the method, or the assay, further includes sequencing a subgenomic interval that is present in a gene or gene product chosen from one, two, three, four, five, ten, fifteen, twenty or more of ARFRP1, BCL2A1, CARD11, CDH20, CDH5, DDR2, EPHA3, EPHA5, EPHA7, EPHB1, FOXP4, GPR124, GUCY1A2, INSR, LRP1B, LTK, PAK3, PHLPP2, PLCG1, PTPRD, STAT3, TBX22 or USP9X.

[0179] In certain implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence from at least 50. 75. 100, 150, 200 or more genes or gene products from Table 1 or 1A. In other implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence from at least 50, 75, 100, 150, 200 or more genes or gene products from Table 1 or 1A acquired from a tumor sample from the cancer types specified therein. In yet other implementations, the sequenced subgenomic interval includes a combination of the Priority 1 genes and the PGx genes according to Table 1 or 1A (e.g., at least 5, 10, 20 or 30 Priority 1 genes; and at least 5, 10, 20 or 30 PGX genes according to Table 1 or 1A). In other implementations, the sequenced subgenomic interval includes a combination of the Priority 1 genes, Cancer genes and PGx genes according to Table 1 or 1A (e.g., at least 5, 10, 20 or 30 Priority 1 genes; at least 5, 10, 20 or 30 Cancer genes; and at least 5, 10, 20 or 30 PGX genes according to Table 1 or 1A).

[0180] In certain implementations, the sequenced subgenomic interval of the nucleic acid sample includes a codon chosen from one or more of: codon 315 of the ABL1 gene; codon 1114, 1338, 1450 or 1556 of APC; codon 600 of BRAF; codon 32, 33, 34, 37, 41 or 45 of CTNNB1; codon 719, 746-750, 768, 790, 858 or 861 of EGFR; codon 835 of FLT3; codon 12, 13, or 61 of HRAS; codon 617 of JAK2; codon 816 of KIT; codon 12, 13, or 61 of KRAS; codon 88, 542, 545, 546, 1047, or 1049 of PIK3CA; codon 130, 173. 233. or 267 of PTEN; codon 918 of RET; codon 175, 245, 248, 273, or 306 of TP53. In certain implementations, two, three, four, five, ten, fifteen, twenty or more of the aforesaid codons are sequenced. In other implementations, the sequenced subgenomic interval includes one or more of the codons shown in Table 1 or 1A.

[0181] In other implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence from at least one, five, ten fifteen, twenty, twenty-five or more PGx genes or gene products from Table 1. In other implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence from at least 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, or more PGx genes or gene products from Table 2. In yet other implementations, the sequenced subgenomic interval includes a nucleotide sequence from at least one PGx gene (and / or at least one PGx gene mutation) according to Table 2 that is associated with one or more of: drug metabolism, drug responsiveness, drug toxicity or side effects. For example, the sequenced subgenomic interval can include a nucleotide sequence from at least one PGx gene associated with better survival of a cancer patient treated with a drug (e.g., better survival of a breast cancer patient treated with paclitaxel (e.g., an ABCB1 gene)). In other implementations, the sequenced subgenomic interval is associated with paclitaxel metabolism (e.g., CYP2C8 genes at different loci and mutations shown in Table 2; a CYP3A4 gene). In yet other implementations, the sequenced subgenomic interval is associated with toxicity to a drug (e.g., 6-MP toxicity as seen with ABCC4 gene (Table 2); 5-FU toxicity as seen with DPYD gene, TYMS gene, and UMPS gene (Table 2); purine toxicity as seen with TMPT gene (Table 2); daunorubicin toxicity as seen with NRP2 gene; C1orf144 gene, CYP1B1 gene (Table 2)). In other implementations, the sequenced subgenomic interval is associated with a side effect to a drug (e.g., ABCG2, TYMS, UGT1A1, ESR1 and ESR2 genes (Table 2)).

[0182] In another implementation, subgenomic intervals from one of the following sets or groups are analyzed. E.g., subgenomic intervals associated with a tumor or cancer gene or gene product, a reference (e.g., a wild type) gene or gene product, or a PGx gene or gene product, thereby obtaining a selected subset of subgenomic intervals from the tumor sample.

[0183] In an implementation, the method sequences a subset of subgenomic intervals from the tumor sample, wherein the subgenomic intervals are chosen from at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 or all of the following: A) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more subgenomic intervals from a mutated or wild-type gene or gene product chosen from at least five or more of: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53; B) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of subgenomic intervals from a mutated or wild type gene or gene product chosen from at least five or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1, TSC1, TSC2, USP9X, VHL, or WT1; C) at least five, six, seven, eight, nine, ten, fifteen, twenty, or more subgenomic intervals from a gene or gene product according to Table 1, 1A, 2, 3 or 4; D) at least five, six, seven, eight, nine, ten, fifteen, twenty, or more subgenomic intervals from a gene or gene product that is associated with a tumor or cancer (e.g., is a positive or negative treatment response predictor, is a positive or negative prognostic factor for, or enables differential diagnosis of a tumor or cancer, e.g., a gene or gene product chosen from one or more of: ABL1, AKT1, ALK, AR, BRAF, BRCA1, BRCA2, CEBPA, EGFR, ERBB2, FLT3, JAK2, KIT, KRAS, MET, NPM1, PDGFRA, PIK3CA, RARA, AKT2, AKT3, MAP2K4, NOTCH1, and TP53; E) at least five, six, seven, eight, nine, ten, or more subgenomic intervals including a mutated or a wild type codon chosen from one or more of: codon 315 of the ABL1 gene; codon 1114, 1338, 1450 or 1556 of APC; codon 600 of BRAF; codon 32, 33, 34, 37, 41 or 45 of CTNNB1; codon 719, 746-750, 768, 790, 858 or 861 of EGFR; codon 835 of FLT3; codon 12, 13, or 61 of HRAS; codon 617 of JAK2; codon 816 of KIT; codon 12, 13, or 61 of KRAS; codon 88, 542, 545, 546, 1047, or 1049 of PIK3CA; codon 130, 173, 233, or 267 of PTEN; codon 918 of RET; codon 175, 245, 248, 273, or 306 of TP53 (e.g., at least five, ten, fifteen, twenty or more subgenomic intervals that include one or more of the codons shown in Table 1 or 1A). F) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, or more of subgenomic intervals from a mutated or wild type gene or gene product (e.g., single nucleotide polymorphism (SNP)) of a subgenomic interval that is present in a gene or gene product associated with one or more of drug metabolism, drug responsiveness, or toxicity (also referred to therein as "PGx" genes) chosen from: ABCB1, BCC2, ABCC4, ABCG2, C1orf144, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DPYD, ERCC2, ESR2, FCGR3A, GSTP1, ITPA, LRP2, MAN1B1, MTHFR, NQO1, NRP2, SLC19A1, SLC22A2, SLCO1B3, SOD2, SULT1A1, TPMT, TYMS, UGT1A1, or UMPS; G) at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, or more of subgenomic intervals from a mutated or wild type PGx gene or gene product (e.g., single nucleotide polymorphism (SNP)) of a subgenomic interval that is present in a gene or gene product associated with one or more of: (i) better survival of a cancer patient treated with a drug (e.g., better survival of a breast cancer patient treated with paclitaxel (e.g., an ABCB1 gene)); (ii) paclitaxel metabolism (e.g., CYP2C8 genes at different loci and mutations shown in Table 2; CYP3A4 gene); (iii) toxicity to a drug (e.g., 6-MP toxicity as seen with ABCC4 gene (Table 2); 5-FU toxicity as seen with DPYD gene, TYMS gene, or UMPS gene (Table 2); purine toxicity as seen with a TMPT gene (Table 2); daunorubicin toxicity as seen with NRP2 gene; C1orf144 gene, CYP1B1 gene (Table 2); or (iv) a side effect to a drug (e.g., ABCG2, TYMS, UGT1A1, ESR1 and ESR2 genes (Table 2)); H) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3; J) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3 in a solid tumor sample from the cancer types specified therein; K) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products according to Table 4; L) a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products according to Table 4 in a heme tumor sample from the cancer types specified therein; M) at least five genes or gene products selected from Table 1, 1A-4, wherein an allelic variation, e.g., at the preselected position, is associated with a preselected type of tumor and wherein said allelic variation is present in less than 5% of the cells in said tumor type; N) at least five genes or gene products selected from Table 1, 1A-4, which are embedded in a GC-rich region; or O) at least five genes or gene products indicative of a genetic (e.g., a germline risk) factor for developing cancer (e.g., the gene or gene product is chosen from one or more of BRCA1, BRCA2, EGFR, HRAS, KIT, MPL, ALK, PTEN, RET, APC, CDKN2A, MLH1, MSH2, MSH6, NF1, NF2, RB1, TP53, VHL or WT1).

[0184] In certain implementations, the acquiring step of the method or assay includes acquiring a library that includes a plurality of tumor or cancer-associated members, reference members and / or PGx members as described herein from said tumor sample. In certain implementations, the selecting step includes solution based hybridization (e.g., to select or enrich for the tumor or cancer-associated member, the reference member (e.g., the wild type member), or the PGx member, each comprising a subgenomic interval from a gene or gene product as described herein.

[0185] Additional implementations or features herein are as follows: In one implementation, the subgenomic interval of the nucleic acid sample includes an intragenic region or an intergenic region. In one implementation, the subgenomic interval includes a gene or fragment thereof, an exon or a fragment thereof, or a preselected nucleotide position. In another implementation, the subgenomic interval includes an exon or an intron, or a fragment thereof, typically an exon or a fragment thereof. In one implementation, the subgenomic interval includes a coding region or a non-coding region, e.g., a promoter, an enhancer, a 5' untranslated region (5' UTR), or a 3' untranslated region (3' UTR), or a fragment thereof.

[0186] In other implementations, the subgenomic interval of the nucleic acid sample includes an alteration (e.g., one or more mutations) associated, e.g., positively or negatively, with a cancerous phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment or resistance to treatment). In yet another implementation, the subgenomic interval includes an alteration, e.g., a point or a single mutation, a deletion mutation (e.g., an in-frame deletion, an intragenic deletion, a full gene deletion), an insertion mutation (e.g., intragenic insertion), an inversion mutation (e.g., an intra-chromosomal inversion), a linking mutation, a linked insertion mutation, an inverted duplication mutation, a tandem duplication (e.g., an intrachromosomal tandem duplication), a translocation (e.g., a chromosomal translocation, a non-reciprocal translocation), a rearrangement, a change in gene copy number, or a combination thereof.

[0187] In other implementations, the subgenomic interval of the nucleic acid sample includes a nucleic acid molecule (in the same or a different subgenomic interval) not associated with the cancerous phenotype for the tumor of the type from the sample. In one implementation, the sequenced subgenomic interval includes a wild-type or a non-mutated nucleotide sequence of a gene or gene product (e.g., an exon sequence or a fragment thereof) that when mutated is associated with a cancerous phenotype (e.g., a wild type or a non-mutated sequence of a gene or gene product as described herein). For example, the sequenced subgenomic interval is from a normal (e.g., non-cancerous) reference sample (e.g., form the same subject from whom the tumor sample was obtained); a normal adjacent tissue (NAT) or a blood sample from the same subject having or at risk of having the tumor. In other implementations, the sequenced subgenomic interval is from a different subject as the tumor or cancer-associated member (e.g., is from one or more of the same or a different tumor sample from a different subject; a normal (e.g., non-cancerous) reference sample; a normal adjacent tissue (NAT); or a blood sample), from one or more different subjects (e.g., healthy subjects or other subjects having or at risk of having the tumor).

[0188] In other implementations, the subgenomic interval of the nucleic acid sample includes one or more translocation alterations as shown in Table 3, Table 4, or a combination thereof. In certain implementations, the sequenced subgenomic interval includes a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3. In other implementations, the sequenced subgenomic interval includes a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 110 or more genes or gene products according to Table 3 in a tumor sample from the cancer types specified therein. In other implementations, the sequenced subgenomic interval includes a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products according to Table 4. In other implementations, the sequenced subgenomic interval includes a translocation alteration of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 150, 200 or more genes or gene products from Table 4 in a tumor sample from the cancer types specified therein.

[0189] In one implementation, the subgenomic interval of the nucleic acid sample includes an exon sequence that includes a single nucleotide alteration associated with a cancerous phenotype. For example, the subgenomic interval includes nucleotides 25,398,215-25,398,334 of chromosome 12. In other implementations, the subgenomic interval includes a C-T substitution at position 25,398,286, which represents a G12S mutation in the KRAS gene.

[0190] In another implementation, the subgenomic interval of the nucleic acid sample includes an in-frame deletion of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more codons from a reference nucleotide (e.g., chromosome) sequence. In one implementation, the subgenomic interval includes an in-frame deletion of codons 746-750 of the EGFR gene (e.g., the subgenomic interval includes nucleotides 55,242,400 to 55,242,535 of chromosome 7, but lacks nucleotides 55,242,464 to 55,242,479).

[0191] In yet another implementation, the subgenomic interval of the nucleic acid sample includes a deletion of the dinucleotide sequence "CA" from codon 64 of the PTEN gene (e.g., the subgenomic interval includes nucleotides 9,675,214 to 89,675,274 of chromosome 10, followed by bases 89,675,277 to 89,675,337 of chromosome 10.

[0192] In yet another implementation, the subgenomic interval of the nucleic acid sample includes an insertion of amino acid residues "Gly-Met" following codon 136 of the PTEN (e.g., the subgenomic interval includes nucleotides 89,692,864 to 89,692,924 of chromosome 10, followed by a nucleotide sequence "GGNATG", followed by nucleotides 89,692,925 to 89,692,980 of chromosome 10).

[0193] In yet another implementation, the subgenomic interval of the nucleic acid sample includes a deletion of the CDKN2A gene (e.g., the subgenomic interval includes nucleotides 21,961,007 to 21,961,067 of chromosome 9 adjacent to bases 22,001,175 to 22,001,235 of chromosome 9).

[0194] In another implementation, the sequenced subgenomic interval of the nucleic acid sample includes an inversion producing an EML4:ALK fusion (e.g., the subgenomic interval includes nucleotides 42,522,893 to 42,522,953 of chromosome 2, juxtaposed with nucleotides 29,449,993 to 29,449,933 of chromosome 2).

[0195] In another implementation, the subgenomic interval of the nucleic acid sample includes an interchromosal translocation resulting in a BCR-ABL fusion (e.g., the subgenomic interval includes nucleotides 23,632,552 to 23,632,612 of chromosome 22, juxtaposed with nucleotides 133,681,793 to 133.681.853 of chromosome 9).

[0196] In another implementation, the subgenomic interval of the nucleic acid sample includes an internal tandem duplication (ITD) mutation in the FLT3 gene (e.g., the subgenomic interval includes nucleotides 28,608,259 to 28,608,285 of chromosome 13 repeated twice in the same orientation.

[0197] In another implementation, the subgenomic interval of the nucleic acid sample includes a microsatellite marker sequence (e.g., the subgenomic interval includes a microsatellite marker sequence of D2S123, e.g., nucleotides 51,288,380 to 51,288,500 and nucleotides 51,288,560 to 51,288,680 of chromosome 2.

[0198] In another implementation, the subgenomic interval of the nucleic acid sample includes a nucleotide sequence corresponding to a fusion sequence (e.g., a fusion transcript or a cancer associated alternative spliced form of a non-fusion transcript).

[0199] In other implementation, the subgenomic interval of the nucleic acid sample includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is indicative of a cancer-related phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment response or resistance to treatment, tumor staging, metastatic likelihood, etc.). In certain implementation, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is predictive of a positive clinical outcome, and / or responsiveness to therapy. In other implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is predictive of a negative clinical outcome, and / or responsiveness to therapy. In certain implementations, the sequenced subgenomic interval of the nucleic acid sample includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is indicative of a genetic (e.g., a germline risk) factor for developing cancer (e.g., the gene or gene product is chosen from one or more of BRCA1, BRCA2, EGFR, HRAS, KIT, MPL, ALK, PTEN, RET, APC, CDKN2A, MLH1, MSH2, MSH6, NF1, NF2, RB1, TP53, VHL or WT1).

[0200] In other implementations, the subgenomic interval of the nucleic acid sample is from one or more genes or gene products shown in Table 1, 1A, 3 or 4, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of the cancer types described in Tables 1, 1A, 3 or 4.

[0201] In one implementation, the subgenomic interval of the nucleic acid sample is from an ABL-1 gene or gene product, that is associated with a cancerous phenotype, e.g., a soft-tissue malignancy chosen from one or more of CML, ALL or T-ALL. In other implementations, the sequenced subgenomic interval of the nucleic acid sample is from an AKT1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, colorectal, ovarian, or non-small cell lung carcinoma (NSCLC).

[0202] In other implementations, the subgenomic interval of the nucleic acid sample is from an ALK gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of ALCL, NSCLC or neuroblastoma.

[0203] In other implementations, the subgenomic interval of the nucleic acid sample is from an APC gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, pancreatic, desmoid, hepatoblastoma, glioma, or other CNS cancers or tumors.

[0204] In other implementations, embodiments, the subgenomic interval of the nucleic acid sample is from a BRAF gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, colorectal cancer, lung cancer, other epithelial malignancies, or hamatological malignancies including AML or ALL.

[0205] In other implementations, the subgenomic interval of the nucleic acid sample is from a CDKN2A gene or gene product,that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, pancreatic, or other tumor types.

[0206] In other implementations, the sequenced subgenomic interval of the nucleic acid sample is from a CEBPA gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of AML or MDS.

[0207] In other implementations, the subgenomic interval of the nucleic acid sample is from a CTNNB1 gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, ovarian, hepatoblastoma, or pleomorphic salivary adenoma.

[0208] In other implementations embodiments, the subgenomic interval of the nucleic acid sample is from an EGFR gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of glioma, lung cancer, or NSCLC.

[0209] In other implementations, the subgenomic interval of the nucleic acid sample is from an ERBB2 gene or gene product, that is associated, e.g., positively or negatively, with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, ovarian, NSCLC, gastric or other solid tumors.

[0210] In other implementations, the subgenomic interval of the nucleic acid sample is from an ESR1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, ovarian or endometrial tumors.

[0211] In other implementations, the subgenomic interval of the nucleic acid sample is from an FGFR1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of MPD or NHL.

[0212] In other implementations, the subgenomic interval of the nucleic acid sample is from an FGFR2 gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of gastric, NSCLC or endometrial tumors. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of gastric, NSCLC or endometrial tumors.

[0213] In other implementations, the subgenomic interval of the nucleic acid sample is from an FGFR3 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of bladder cancer, multiple myeloma or T-cell lymphoma.

[0214] In other implementations, the subgenomic interval of the nucleic acid sample is from an FLT3 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of AML, melanoma, colorectal, papillary thyroid, ovarian, non small-cell lung cancer (NSCLC), cholangiocarcinoma, or pilocytic astrocytoma.

[0215] In other implementations, the subgenomic interval of the nucleic acid sample is from an HRAS gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of rhadomyosarcoma, ganglioneuroblastoma, bladder, sarcomas, or other cancer types.

[0216] In other implementations, the subgenomic interval of the nucleic acid sample is from a JAK2 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of ALL, AML, MPD or CML.

[0217] In other implementations, the subgenomic interval of the nucleic acid sample is from a KIT gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of gastrointestinal stromal tumors (GIST), AML, TGCT, mastocytosis, mucosal melanoma, or epithelioma.

[0218] In other implementations, the subgenomic interval of the nucleic acid sample is from a KRAS gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of pancreatic, colon, colorectal, lung, thyroid, or AML.

[0219] In other implementations, the subgenomic interval of the nucleic acid sample is from a MET gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of renal or head-neck squamous cell carcinoma.

[0220] In other implementations, the sequenced subgenomic interval of the nucleic acid sample is from an MLL gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of AML or ALL.

[0221] In other implementations, the subgenomic interval of the nucleic acid sample is from an NF1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of neurofibroma or glioma.

[0222] In other implementations, the subgenomic interval of the nucleic acid sample is from a NOTCH1 gene or gene product that is associated with a cancerous phenotype, e.g., a T-ALL cancer.

[0223] In other implementations, the subgenomic interval of the nucleic acid sample is from an NPM1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of NHL. APL or AML.

[0224] In other implementations, the subgenomic interval of the nucleic acid sample is from an NRAS gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, colorectal cancer, multiple myeloma, AML, or thyroid cancer.

[0225] In other implementations, the subgenomic interval of the nucleic acid sample is from a PDGFRA gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of GIST or idiopathic hypereosinophilic syndrome.

[0226] In other implementations, the subgenomic interval of the nucleic acid sample is from a PIK3CA gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, gastric, gliobastoma, or breast cancer.

[0227] In other implementations, the subgenomic interval of the nucleic acid sample is from a PTEN gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, glioma, prostate, or endometrial cancer.

[0228] In other implementations, the subgenomic interval of the nucleic acid sample is from an RB1 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of retinoblastoma, sarcoma, breast, or small cell lung carcinoma.

[0229] In other implementations, the subgenomic interval of the nucleic acid sample is from a RET gene or gene product, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of medullary thyroid, papillary thyroid, or pheochromocytoma.

[0230] In other implementations, the subgenomic interval of the nucleic acid sample is from a TP53 gene or gene product that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, colorectal, lung, sarcoma, adrenocortical, glioma, or other tumor types.

[0231] In one implementation, the subgenomic interval of the nucleic acid sample is a positive predictor of therapeutic response. Examples of a positive predictor of a therapeutic response include, but are not limited to, an activating mutation in the EGFR gene that predicts responsiveness to small molecule EGFR TKIs (e.g., Iressa / gefitinib) in NSCLC patients; presence of an EML4 / ALK fusion gene predicts responsiveness to ALK inhibitors (e.g. PF-02341066) in NSCLC patients; presence of a BRAF V600E mutation predicts responsiveness to BRAF inhibition (e.g. PLX-4032) in melanoma patients.

[0232] In other implementations, the subgenomic interval of the nucleic acid sample is a negative predictor of therapeutic response. Examples of a negative predictor of a therapeutic response include, but are not limited to, an activating mutation in the KRAS gene that predict lack of response to anti-EGFR monoclonal antibodies (cetuximab, panitumumab) in CRC patients; and the presence of an M351T mutation in the BCR / Abl fusion gene predicts resistance to Gleevec / imatinib in CML patients.

[0233] In other implementations, the subgenomic interval of the nucleic acid sample is a prognostic factor. Examples of prognostic factors include, but are not limited to, the presence of an insertion mutation in the FLT3 gene, which is a negative prognostic for relapse in AML patients; the presence of specific RET gene mutations, e.g. M918T, which are negative prognostic factors for survival in medullary thyroid carcinoma patients.

[0234] In other implementations, the subgenomic interval of the nucleic acid sample is a diagnostic factor. Examples of prognostic factors include, but are not limited to, the presence of a BCR / Abl fusion gene, which is diagnostic for CML; and the presence of a SMARCB1 mutation, which is diagnostic of Rhabdoid tumor of the kidney.

[0235] In other implementations, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is present in a minority (e.g., less than 5%) of the cells in the tumor sample. In one implementation, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is associated, e.g., positively or negatively, with a cancer-related phenotype, but which is present in a minority (e.g., less than 5%) of the cells in the tumor sample. In other implementations, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is present in less than 50, 40, 30, 10, 5, or 1% of the cells in a tumor sample. In yet other implementations, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is present in more than 50, 60, 70, 80%, or more of the cells in a tumor sample.

[0236] In yet other implementations, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is present in less than 5, 1, 0.5, 0.1%, 0.01%, 0.001% of the coding region of the genome of the tumor cells in the tumor sample.

[0237] In one implementation, the nucleic acid sample includes a subgenomic interval from a gene or gene product that is associated with a tumor or cancer (e.g., is a positive or negative treatment response predictor, is a positive or negative prognostic factor for, or enables differential diagnosis of a tumor or cancer, e.g., a gene or gene product chosen from one or more of: ABL1, AKT1, ALK, AR, BRAF, BRCA1, BRCA2, CEBPA, EGFR, ERBB2, FLT3, JAK2, KIT, KRAS, MET, NPM1, PDGFRA, PIK3CA, RARA, AKT2, AKT3, MAP2K4, NOTCH1, and TP53.

[0238] In one implementation, the cancerous phenotype associated with the gene or gene product is the same tumor type as the tumor sample. In other implementations, the cancerous phenotype associated with the gene or gene product is from a different tumor type as the tumor sample.

[0239] In certain implementations, the method or assay includes sequencing nucleic acid samples from tumor samples from at least X subjects, (wherein X = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, or more subjects). In one implementation, the subject is human having, or at risk of having, a cancer or tumor. The method includes sequencing at least 5, 10, 15, 20, 30, 40, 50, 75 or more genes or gene products described herein (e.g., genes or gene products from Table 1, 1A, 2, 3, or 4) from at least X subjects. In certain implementations, the gene or gene product includes an alteration that is associated with a cancerous phenotype, e.g., one or more of cancer risk, cancer progression, cancer treatment response or resistance to treatment.

[0240] In other implementations or in addition to the aforesaid implementations, the method or assay includes sequencing a control or reference subgenomic interval from a gene or gene product from the same subject as the tumor sample, e.g., a wild-type or a non-mutated nucleotide sequence of a gene or gene product described herein (e.g., genes or gene products from Table 1, 1A, 2, 3, or 4). In one implementation, the control gene or gene product is from the same subject or a different subject as the tumor sample (e.g., is from one or more of the same or a different tumor sample; a normal (e.g., non-cancerous) sample; a normal adjacent tissue (NAT); or a blood sample), from the same subject having or at risk of having the tumor, or from a different subject.

[0241] In other implementations or in addition to the aforesaid embodiments, the method or assay includes sequencing a subgenomic interval that is present in a gene associated with drug metabolism, drug responsiveness, or toxicity (the PGx genes as described herein). In certain implementations, the subgenomic interval sequenced includes an alteration (e.g., single nucleotide polymorphism (SNP)).

[0242] In certain implementations, the method, or assay, includes sequencing (and / or reporting the results of sequencing) a first set of genes or gene products from Table 1, 1A, 2, 3, or 4 from a first subject. In other implementations, the method, or assay, includes sequencing (and / or reporting the results of sequencing) a second set, a third set or more (e.g., an overlapping but different) set of genes or gene products from Table 1, 1A, 2, 3, or 4 from a first or a second subject. In certain implementations, the tumor sample from a first subject includes a tumor of a first type and the tumor sample from a second subject includes a tumor of a second type. In other implementations, the tumor sample from the first subject and the second subject are from the same tumor type.

[0243] In certain implementations, the method or assay further includes one or more of: (i) fingerprinting the nucleic acid sample; (ii) quantifying the abundance of a gene or gene product (e.g., a gene or gene product as described herein) in the nucleic acid sample; (iii) quantifying the relative abundance of a transcript in the sample; (iv) identifying the nucleic acid sample as belonging to a particular subject (e.g., a normal control or a cancer patient); (v) identifying a genetic trait in the nucleic acid sample (e.g., one or more subject's genetic make-up (e.g., ethnicity, race, familial traits)); (vi) determining the ploidy in the nucleic acid sample; determining a loss of heterozygosity in the nucleic acid sample; (vii) determining the presence or absence of a gene duplication event in the nucleic acid sample; (viii) determining the presence or absence of a gene amplification event in the nucleic acid sample; or (ix) determining the level of tumor / normal cellular admixture in the nucleic acid sample.

[0244] In other implementations, the nucleic acid sample includes a library, or a selected library output, that includes a plurality of tumor nucleic acid members, reference or control (e.g., wild type) nucleic acid members, and / or PGx associated nucleic acid members (e.g., a nucleic acid that includes a subgenomic interval as described herein) from the tumor sample. In one implementation, the library (e.g., the nucleic acid library) includes a plurality of members, e.g., target nucleic acid members from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30 or more subjects. In one implementation, the subject is human having, or at risk of having, a cancer or tumor. In certain implementations, the library further comprises tumor or cancer-associated nucleic acid members and control nucleic acid fragments from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30 or more subjects.

[0245] In certain implementations, the selected subset of subgenomic intervals are separated or enriched from the nucleic acid sample by solution- or solid support-based hybridization. In one implementation, the method, or assay, provides selected members of a nucleic acid library (e.g., a library catch). The method includes: providing a library (e.g., a nucleic acid library) comprising a plurality of members, e.g., target members (e.g., including a plurality of tumor or cancer-associated members, reference members, and / or PGx members); contacting the library, e.g., in a solution- or solid support-based reaction, with a plurality of baits (e.g., oligonucleotide baits) to form a hybridization mixture comprising a plurality of bait / member hybrids; separating the plurality of bait / member hybrids from said hybridization mixture, e.g., by contacting said hybridization mixture with a binding entity that allows for separation of said plurality of bait / member hybrid, thereby providing a library-catch (e.g., a selected or enriched subgroup of nucleic acid molecules from the library), wherein the plurality of baits includes at least one, or two of the following: a) a first bait set that selects a tumor or cancer-associated or a reference (e.g., wild type) member comprising a subgenomic interval from a tumor or a reference gene or gene product as described herein, e.g., a tumor or a reference gene or gene product as described in Table 1, 1A, 3 or 4; b) a second bait set that selects a PGx member comprising a subgenomic interval (in the same or a different subgenomic interval as in a) from a gene or gene product as described in Table 1 or 2.

[0246] In certain implementations, the method, or assay, further includes the step of sequencing said members. In certain implementations, tumor members from at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30 or more subjects are sequenced (e.g., at least 50, 75, 100 or 150 subgenomic intervals from the genes or gene products from Table 1 or 1A are sequenced from each subject).

[0247] In certain implementations, the method, or assay, further includes the step of detecting, in the nucleic acid sample, a preselected alteration (e.g., an allelic variation) in at least 10 (e.g., 20, 30, 40) Priority, Cancer, or PGx genes or gene products from Table 1. In certain implementations, the alteration (e.g., the allelic variation) includes a cytogenetic abnormality, a non-reciprocal translocation, a rearrangement, an intra-chromosomal inversion, a mutation, a point mutations, a deletion, a change in gene copy number, an SNP, among others.

[0248] In certain implementations, the method, or assay, further includes the step of comparing the detected presence or absence of the alteration (e.g., the allelic variation) to a reference value (e.g., a literature report or the sequence of a control sample, e.g., blood matched controls or NAT (normal adjacent tumor), from the same subject as the tumor sample, or a different subject).

[0249] In certain implementations, the method, or assay, further includes the step of memorializing the presence or absence of the alteration (e.g., the preselected allelic variation), and, e.g., providing a report comprising the memorialization.

[0250] In certain implementations, the method, or assay, further includes the step of annotating the alteration, and, e.g., annotating a preselected allelic variation with an indication of a mutation structure, e.g., a mis-sense mutation, or function, e.g., an association with a disease phenotype.

[0251] In certain implementations, the method, or assay, further includes the step of providing a data set, wherein each element of the dataset comprises the association of a tumor type, a gene and a preselected alteration (e.g., allelic variation) (a "TGA").

[0252] In certain implementations, the method, or assay, further includes the step of memorializing the presence or absence of a TGA, and optionally an associated annotation, of a subject to form a report.

[0253] In certain implementations, the method, or assay, further includes the step of transmitting the report to a recipient party.

[0254] Assays, e.g., multiplex assays, that include the aforesaid selection methods and reagents are also provided.Nucleic Acid Samples

[0255] A variety of tissue samples can be the source of the nucleic acid samples used in the present methods. Genomic or subgenomic nucleic acid (e.g., DNA or RNA) can be isolated from a subject's sample (e.g., a tumor sample, a normal adjacent tissue (NAT), a blood sample, a sample containing circulating tumor cells (CTC) or any normal control)). In certain implementations, the tissue sample is preserved as a frozen sample or as formaldehyde- or paraformaldehyde-fixed paraffin-embedded (FFPE) tissue preparation. For example, the sample can be embedded in a matrix, e.g., an FFPE block or a frozen sample. The isolating step can include flow-sorting of individual chromosomes; and / or micro-dissecting a subject's sample (e.g., a tumor sample, a NAT, a blood sample).

[0256] An "isolated" nucleic acid molecule is one which is separated from other nucleic acid molecules which are present in the natural source of the nucleic acid molecule. In certain implementations, an "isolated" nucleic acid molecule is free of sequences (such as protein-encoding sequences) which naturally flank the nucleic acid (i.e., sequences located at the 5' and 3' ends of the nucleic acid) in the genomic DNA of the organism from which the nucleic acid is derived. For example, in various implementations, the isolated nucleic acid molecule can contain less than about 5 kB, less than about 4 kB, less than about 3 kB, less than about 2 kB, less than about 1 kB, less than about 0.5 kB or less than about 0.1 kB of nucleotide sequences which naturally flank the nucleic acid molecule in genomic DNA of the cell from which the nucleic acid is derived. Moreover, an "isolated" nucleic acid molecule, such as a cDNA molecule, can be substantially free of other cellular material or culture medium when produced by recombinant techniques, or substantially free of chemical precursors or other chemicals when chemically synthesized.

[0257] The language "substantially free of other cellular material or culture medium" includes preparations of nucleic acid molecule in which the molecule is separated from cellular components of the cells from which it is isolated or recombinantly produced. Thus, nucleic acid molecule that is substantially free of cellular material includes preparations of nucleic acid molecule having less than about 30%, less than about 20%, less than about 10%, or less than about 5% (by dry weight) of other cellular material or culture medium.

[0258] In certain implementations, the nucleic acid is isolated from an aged sample, e.g., an aged FFPE sample. The aged sample, can be, for example, years old, e.g., 1 year, 2 years, 3 years, 4 years, 5 years, 10 years, 15 years, 20 years, 25 years, 50 years, 75 years, or 100 years old or older.

[0259] A nucleic acid sample can be obtained from tissue samples (e.g., a biopsy or FFPE sample) of various sizes. For example, the nucleic acid can be isolated from a tissue sample from 5 to 200 µm, or larger. For example, the tissue sample can measure 5 µm, 10 µm, 20 µm, 30 µm, 40 µm, 50 µm, 70 µm, 100 µm, 110 µm, 120 µm., 150 µm or 200 µm or larger.

[0260] Protocols for DNA isolation from a tissue sample are provided in Example 1. Additional methods to isolate nucleic acids (e.g., DNA) from formaldehyde- or paraformaldehyde-fixed, paraffin-embedded (FFPE) tissues are disclosed, e.g., in Cronin M. et al., (2004) Am J Pathol. 164(1):35-42; Masuda N. et al., (1999) Nucleic Acids Res. 27(22):4436-4443; Specht K. et al., (2001) Am J Pathol. 158(2):419-429, Ambion RecoverAll ™< Total Nucleic Acid Isolation Protocol (Ambion, Cat. No. AM1975, September 2008), Maxwell ®< 16 FFPE Plus LEV DNA Purification Kit Technical Manual (Promega Literature #TM349, February 2011), E.Z.N.A. ®< FFPE DNA Kit Handbook (OMEGA bio-tek, Norcross, GA, product numbers D3399-00, D3399-01, and D3399-02; June 2009), and QIAamp ®< DNA FFPE Tissue Handbook (Qiagen, Cat. No. 37625, October 2007). RecoverAll ™< Total Nucleic Acid Isolation Kit uses xylene at elevated temperatures to solubilize paraffin-embedded samples and a glass-fiber filter to capture nucleic acids. Maxwell ®< 16 FFPE Plus LEV DNA Purification Kit is used with the Maxwell ®< 16 Instrument for purification of genomic DNA from 1 to 10 µm sections of FFPE tissue. DNA is purified using silica-clad paramagnetic particles (PMPs), and eluted in low elution volume. The E.Z.N.A. ®< FFPE DNA Kit uses a spin column and buffer system for isolation of genomic DNA. QIAamp ®< DNA FFPE Tissue Kit uses QIAamp ®< DNA Micro technology for purification of genomic and mitochondrial DNA.Protocols for DNA isolation from blood are disclosed, e.g., in the Maxwell ®< 16 LEV Blood DNA Kit and Maxwell 16 Buccal Swab LEV DNA Purification Kit Technical Manual (Promega Literature #TM333, January 1, 2011).

[0261] Protocols for RNA isolation are disclosed, e.g., in the Maxwell ®< 16 Total RNA Purification Kit Technical Bulletin (Promega Literature #TB351, August 2009).

[0262] The isolated nucleic acid samples (e.g., genomic DNA samples) can be fragmented or sheared by practicing routine techniques. For example, genomic DNA can be fragmented by physical shearing methods, enzymatic cleavage methods, chemical cleavage methods, and other methods well known to those skilled in the art. The nucleic acid library can contain all or substantially all of the complexity of the genome. The term "substantially all" in this context refers to the possibility that there can in practice be some unwanted loss of genome complexity during the initial steps of the procedure. The methods described herein also are useful in cases where the nucleic acid library is a portion of the genome, i.e., where the complexity of the genome is reduced by design. In some implementations, any selected portion of the genome can be used with the methods described herein. In certain implementations, the entire exome or a subset thereof is isolated.

[0263] Methods featured herein can further include isolating a nucleic acid sample to provide a library (e.g., a nucleic acid library as described herein). In certain implementations, the nucleic acid sample includes whole genomic, subgenomic fragments, or both. The isolated nucleic acid samples can be used to prepare nucleic acid libraries. Thus, in one implementation, the methods featured herein further include isolating a nucleic acid sample to provide a library (e.g., a nucleic acid library as described herein). Protocols for isolating and preparing libraries from whole genomic or subgenomic fragments are known in the art (e.g., Illumina's genomic DNA sample preparation kit). In certain implementations, the genomic or subgenomic DNA fragment is isolated from a subject's sample (e.g., a tumor sample, a normal adjacent tissue (NAT), a blood sample or any normal control)). In one implementation, the sample (e.g., the tumor or NAT sample) is a preserved specimen. For example, the sample is embedded in a matrix, e.g., an FFPE block or a frozen sample. In certain implementations, the isolating step includes flow-sorting of individual chromosomes; and / or microdissecting a subject's sample (e.g., a tumor sample, a NAT, a blood sample). In certain implementations, the nucleic acid sample used to generate the nucleic acid library is less than 5 microgram, less than 1 microgram, or less than 500ng, less than 200ng, less than 100ng, less than 50ng, less than 10ng, less than 5 ng, or less than 1 ng.

[0264] In still other implementations, the nucleic acid sample used to generate the library includes RNA or cDNA derived from RNA. In some implementations, the RNA includes total cellular RNA. In other implementations, certain abundant RNA sequences (e.g., ribosomal RNAs) have been depleted. In some implementations, the poly(A)-tailed mRNA fraction in the total RNA preparation has been enriched. In some implementations, the cDNA is produced by random-primed cDNA synthesis methods. In other implementations, the cDNA synthesis is initiated at the poly(A) tail of mature mRNAs by priming by oligo(dT)-containing oligonucleotides. Methods for depletion, poly(A) enrichment, and cDNA synthesis are well known to those skilled in the art.

[0265] The method can further include amplifying the nucleic acid sample by specific or non-specific nucleic acid amplification methods that are well known to those skilled in the art. In some implementations, certain implementations, the nucleic acid sample is amplified, e.g., by whole-genome amplification methods such as random-primed strand-displacement amplification.

[0266] In other implementations, the nucleic acid sample is fragmented or sheared by physical or enzymatic methods and ligated to synthetic adapters, size-selected (e.g., by preparative gel electrophoresis) and amplified (e.g., by PCR). In other implementations, the fragmented and adapter-ligated group of nucleic acids is used without explicit size selection or amplification prior to hybrid selection.

[0267] In other implementations, the isolated DNA (e.g., the genomic DNA) is fragmented or sheared. In some implementations, the library includes less than 50% of genomic DNA, such as a subfraction of genomic DNA that is a reduced representation or a defined portion of a genome, e.g., that has been subfractionated by other means. In other implementations, the library includes all or substantially all genomic DNA.

[0268] In some implementations, the library includes less than 50% of genomic DNA, such as a subfraction of genomic DNA that is a reduced representation or a defined portion of a genome, e.g., that has been subfractionated by other means. In other implementations, the library includes all or substantially all genomic DNA. Protocols for isolating and preparing libraries from whole genomic or subgenomic fragments are known in the art (e.g., Illumina's genomic DNA sample preparation kit), and are described herein as Examples 2A, 2B and 3. Alternative methods for DNA shearing are described herein as Example 2B. For example, alternative DNA shearing methods can be more automatable and / or more efficient (e.g., with degraded FFPE samples). Alternatives to DNA shearing methods can also be used to avoid a ligation step during library preparation.

[0269] The methods described herein can be performed using a small amount of nucleic acids, e.g., when the amount of source DNA is limiting (e.g., even after whole-genome amplification). In one implementation, the nucleic acid comprises less than about 5 µg, 4 µg, 3 µg, 2 µg, 1 µg, 0.8 µg, 0.7 µg, 0.6 µg, 0.5 µg, or 400 ng, 300 ng, 200 ng, 100 ng, 50 ng, 10 ng, 5 ng, 1 ng, or less of nucleic acid sample. For example, one can typically begin with 50-100 ng of genomic DNA. One can start with less, however, if one amplifies the genomic DNA (e.g., using PCR) before the hybridization step, e.g., solution hybridization. Thus it is possible, but not essential, to amplify the genomic DNA before hybridization, e.g., solution hybridization.

[0270] The nucleic acid sample used to generate the library can also include RNA or cDNA derived from RNA. In some embodiments, the RNA includes total cellular RNA. In other implementations, certain abundant RNA sequences (e.g., ribosomal RNAs) have been depleted. In other implementations, the poly(A)-tailed mRNA fraction in the total RNA preparation has been enriched. In some implementations, the cDNA is produced by random-primed cDNA synthesis methods. In other implementations, the cDNA synthesis is initiated at the poly(A) tail of mature mRNAs by priming by oligo(dT)-containing oligonucleotides. Methods for depletion, poly(A) enrichment, and cDNA synthesis are well known to those skilled in the art.

[0271] The method can further include amplifying the nucleic acid sample by specific or non-specific nucleic acid amplification methods that are known to those skilled in the art. The nucleic acid sample can be amplified, e.g., by whole-genome amplification methods such as random-primed strand-displacement amplification.

[0272] The nucleic acid sample can be fragmented or sheared by physical or enzymatic methods as described herein, and ligated to synthetic adapters, size-selected (e.g., by preparative gel electrophoresis) and amplified (e.g., by PCR). The fragmented and adapter-ligated group of nucleic acids is used without explicit size selection or amplification prior to hybrid selection.Library Members

[0273] "Member" or "library member" or other similar term, as used herein, refers to a nucleic acid molecule, e.g., DNA or RNA, that is the member of a library (or "library-catch"). The library member can be one or more of a tumor member, a reference member, or a PGx member as described herein. Typically, a member is a DNA molecule, e.g., a genomic DNA or cDNA, molecule. A member can be fragmented, e.g., enzymatically or by shearing, genomic DNA. Members can comprise a nucleotide sequence from a subject and can also comprise a nucleotide sequence not derived from the subject, e.g., primers or adapters (e.g., for PCR amplification or for sequencing), or sequences that allow for identification of a sample, e.g., "barcode" sequences.

[0274] As used herein, "target member" refers to a nucleic acid molecule that one desires to isolate from the nucleic acid library. In one implementation, the target members can be a tumor member, a reference member, or a PGx member as described herein. The members that are actually selected from the nucleic acid library is referred to herein as the "library catch." In one implementation, the library-catch includes a selection or enrichment of members of the library, e.g., the enriched or selected output of a library after one or more rounds of hybrid capture as described herein.

[0275] The target members may be a subgroup of the library, i.e., that not all of the library members are selected by any particular use of the processes described herein. In other implementations, the target members are within a desired target region. For example, the target members may in some implementations be a percentage of the library members that is as low as 10% or as high as 95%-98% or higher. In one implementation, the library catch includes at least about 20%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 98%, 99%, 99.9% or more of the target members. In another implementation, the library contains 100% of the target members. In one implementation, the purity of the library catch (percentage of reads that align to the targets) is at least about 20%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 98%, 99%, 99.9% or more.

[0276] The target members (or the library catch) obtained from genomic DNA can include a small fraction of the total genomic DNA, such that it includes less than about 0.0001%, at least about 0.0001%, at least about 0.001%, at least about 0.01 %, or at least about 0.1 % of genomic DNA, or a more significant fraction of the total genomic DNA, such that it includes at least about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, or 10% of genomic DNA, or more than 10% of genomic DNA.

[0277] In one implementation, the target members (or the library catch) are selected from a complex mixture of genome. For example, the selection of the DNA from one cell type (e.g., cancer cells) from a sample containing the DNA from other cell types (e.g., normal cells). In such applications, the target member can include less than 0.0001%, at least 0.0001%, at least about 0.001%, at least about 0.01%, or at least about 0.1 % of the total complexity of the nucleic acid sequences present in the complex sample, or a more significant fraction such that it includes at least about 1%, 2%, 5%, 10% or more than 10% of the total complexity of nucleic acid sequences present in the complex sample.

[0278] In one implementation, the target member (or the library catch) selected by the methods described herein (e.g., solution hybridization selection methods) include all or a portion of exons in a genome, such as greater than about 0.1%, 1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 95% of the genomic exons. In another implementation, the target member (or the library catch) can be a specific group of exons, e.g., at least about 100, 200, 300, 400, 500, 600, 700, 800, 900, or 1000 particular exons, e.g., exons associated with particular diseases such as cancer. In yet another implementation, the target member (or the library catch) contains exons or other parts of selected genes of interest. The use of specific bait sequences allows the practitioner to select target sequences (ideal set of sequences selected) and subgroups of nucleic acids (actual set of sequences selected) containing as many or as few exons (or other sequences) from a group of nucleic acids for a particular selection.

[0279] In one implementation, the target member (or the library catch) includes a set of cDNAs. Capturing cDNAs can be used, for example, to find splice variants, and to identify fusion transcripts (e.g., from genomic DNA translocations). In another implementation, the target member (and the library catch) is used to find single base changes and other sequence changes expressed in the RNA fraction of a cell, tissue, or organ, e.g., in a tumor.

[0280] The target member (or the library catch) (e.g., exons, cDNAs and other sequences) can be related or unrelated as desired. For example, selected target member (and the library catch) can be obtained from a group of nucleic acids that are genes involved in a disease, such as a group of genes implicated in one or more diseases such as cancers, a group of nucleic acids containing specific SNPs.Design and Construction of Baits

[0281] A bait can be a nucleic acid molecule, e.g., a DNA or RNA molecule, which can hybridize to (e.g., be complementary to), and thereby allow capture of a target nucleic acid. In one implementation, a bait is an RNA molecule. In other implementations, a bait includes a binding entity, e.g., an affinity tag, that allows capture and separation, e.g., by binding to a binding entity, of a hybrid formed by a bait and a nucleic acid hybridized to the bait. In one implementation, a bait is suitable for solution phase hybridization.

[0282] Typically, RNA molecules are used as bait sequences. A RNA-DNA duplex is more stable than a DNA-DNA duplex, and therefore provides for potentially better capture of nucleic acids.

[0283] RNA baits can be made as described elsewhere herein, using methods known in the art including, but not limited to, de novo chemical synthesis and transcription of DNA molecules using a DNA-dependent RNA polymerase. In one implementation, the bait sequence is produced using known nucleic acid amplification methods, such as PCR, e.g., using human DNA or pooled human DNA samples as the template. The oligonucleotides can then be converted to RNA baits. In one implementation, in vitro transcription is used, for example, based on adding an RNA polymerase promoter sequence to one end of the oligonucleotide. In one implementation, the RNA polymerase promoter sequence is added at the end of the bait by amplifying or reamplifying the bait sequence, e.g., using PCR or other nucleic acid amplification methods, e.g., by tailing one primer of each target-specific primer pairs with an RNA promoter sequence. In one implementation, the RNA polymerase is a T7 polymerase, a SP6 polymerase, or a T3 polymerase. In one implementation, RNA bait is labeled with a tag, e.g., an affinity tag. In one implementation, RNA bait is made by in vitro transcription, e.g., using biotinylated UTP. In another implementation, RNA bait is produced without biotin and then biotin is crosslinked to the RNA molecule using methods well known in the art, such as psoralen crosslinking. In one implementation, the RNA bait is an RNase-resistant RNA molecule, which can be made, e.g., by using modified nucleotides during transcription to produce RNA molecule that resists RNase degradation. In one implementation, the RNA bait corresponds to only one strand of the double-stranded DNA target. Typically, such RNA baits are not self-complementary and are more effective as hybridization drivers.

[0284] The bait sets can be designed from reference sequences, such that the baits are optimal for selecting targets of the reference sequences. In some implementations, bait sequences are designed using a mixed base (e.g., degeneracy). For example, the mixed base(s) can be included in the bait sequence at the position(s) of a common SNP or mutation, to optimize the bait sequences to catch both alleles (e.g., SNP and non-SNP; mutant and non-mutant). In some implementations, all known sequence variations (or a subset thereof) can be targeted with multiple oligonucleotide baits, rather than by using mixed degenerate oligonucleotides.

[0285] In certain implementations, the bait set includes an oligonucleotide (or a plurality of oligonucleotides) between about 100 nucleotides and 300 nucleotides in length. Typically, the bait set includes an oligonucleotide (or a plurality of oligonucleotides) between about 130 nucleotides and 230 nucleotides, or about 150 and 200 nucleotides, in length. In other implementations, the bait set includes an oligonucleotide (or a plurality of oligonucleotides) between about 300 nucleotides and 1000 nucleotides in length.

[0286] In some implementations, the target member-specific sequences in the oligonucleotide is between about 40 and 1000 nucleotides, about 70 and 300 nucleotides, about 100 and 200 nucleotides in length, typically between about 120 and 170 nucleotides in length.

[0287] In some implementations, the bait set includes a binding entity. The binding entity can be an affinity tag on each bait sequence. In some implementations, the affinity tag is a biotin molecule or a hapten. In certain implementations, the binding entity allows for separation of the bait / member hybrids from the hybridization mixture by binding to a partner, such as an avidin molecule, or an antibody that binds to the hanten or an antigen-binding fragment thereof.

[0288] In other implementations, the oligonucleotides in the bait set contains forward and reverse complemented sequences for the same target member sequence whereby the oligonucleotides with reverse-complemented member-specific sequences also carry reverse complemented universal tails. This can lead to RNA transcripts that are the same strand, i.e., not complementary to each other.

[0289] In other implementations, the bait set includes oligonucleotides that contain degenerate or mixed bases at one or more positions. In still other implementations, the bait set includes multiple or substantially all known sequence variants present in a population of a single species or community of organisms. In one implementation, the bait set includes multiple or substantially all known sequence variants present in a human population.

[0290] In other implementations, the bait set includes cDNA sequences or is derived from cDNAs sequences. In other implementations, the bait set includes amplification products (e.g., PCR products) that are amplified from genomic DNA, cDNA or cloned DNA.

[0291] In other implementations, the bait set includes RNA molecules. In some implementations, the set includes chemically, enzymatically modified, or in vitro transcribed RNA molecules, including but not limited to, those that are more stable and resistant to RNase.

[0292] In yet other implementations, the baits are produced by methods described in US 2010 / 0029498 and Gnirke, A. et al. (2009) Nat Biotechnol. 27(2):182-189. For example, biotinylated RNA baits can be produced by obtaining a pool of synthetic long oligonucleotides, originally synthesized on a microarray, and amplifying the oligonucleotides to produce the bait sequences. In some implementations, the baits are produced by adding an RNA polymerase promoter sequence at one end of the bait sequences, and synthesizing RNA sequences using RNA polymerase. In one implementation, libraries of synthetic oligodeoxynucleotides can be obtained from commercial suppliers, such as Agilent Technologies, Inc., and amplified using known nucleic acid amplification methods.

[0293] Accordingly, a method of making the aforesaid bait set is provided. The method includes selecting one or more target specific bait oligonucleotide sequences (e.g., one or more mutation capturing, reference or control oligonucleotide sequences as described herein); obtaining a pool of target specific bait oligonucleotide sequences (e.g., synthesizing the pool of target specific bait oligonucleotide sequences, e.g., by microarray synthesis); and optionally, amplifying the oligonucleotides to produce the bait set,

[0294] In other implementations, the methods further include amplifying (e.g., bv PCR) the oligonucleotides using one or more biotinylated primers. In some implementations, the oligonucleotides include a universal sequence at the end of each oligonucleotide attached to the microarray. The methods can further include removing the universal sequences from the oligonucleotides. Such methods can also include removing the complementary strand of the oligonucleotides, annealing the oligonucleotides, and extending the oligonucleotides. In some of these implementations, the methods for amplifying (e.g., by PCR) the oligonucleotides use one or more biotinylated primers. In some implementations, the method further includes size selecting the amplified oligonucleotides.

[0295] In one implementation, an RNA bait set is made. The methods include producing a set of bait sequences according to the methods described herein, adding a RNA polymerase promoter sequence at one end of the bait sequences, and synthesizing RNA sequences using RNA polymerase. The RNA polymerase can be chosen from a T7 RNA polymerase, an SP6 RNA polymerase or a T3 RNA polymerase. In other implementations, the RNA polymerase promoter sequence is added at the ends of the bait sequences by amplifying (e.g., by PCR) the bait sequences. In implementations where the bait sequences are amplified by PCR with specific primer pairs out of genomic or cDNA, adding an RNA promoter sequence to the 5' end of one of the two specific primers in each pair will lead to a PCR product that can be transcribed into a RNA bait using standard methods.

[0296] In other implementations, bait sets can be produced using human DNA or pooled human DNA samples as the template. In such implementations, the oligonucleotides are amplified by polymerase chain reaction (PCR). In other implementations, the amplified oligonucleotides are reamplified by rolling circle amplification or hyperbranched rolling circle amplification. The same methods also can be used to produce bait sequences using human DNA or pooled human DNA samples as the template. The same methods can also be used to produce bait sequences using subfractions of a genome obtained by other methods, including but not limited to restriction digestion, pulsed-field gel electrophoresis, flow-sorting, CsCl density gradient centrifugation, selective kinetic reassociation, microdissection of chromosome preparations and other fractionation methods known to those skilled in the art.

[0297] In certain implementations, the number of baits in the bait set is less than 1,000. In other implementations, the number of baits in the bait set is greater than 1,000, greater than 5,000, greater than 10,000, greater than 20,000, greater than 50,000, greater than 100,000, or greater than 500,000.

[0298] In one implementation, the bait sequence selects a base complementary to a SNP, e.g., to increase its binding capacity (e.g., affinity and / or specificity) in a target gene or gene product, or a fragment thereof, which encodes the SNP. Exemplary genes or gene products include, but not limited to, ABCB1, ABCC2, ABCC4, ABCG2, C1orf144, CYP1B1, CYP2C19, CYP2C8, CYP2D6, CYP3A4, CYP3A5, DPYD, ERCC2, ESR2, FCGR3A, GSTP1, ITPA, LRP2, MAN1B1, MTHFR, NQO1, NRP2, SLC19A1, SLC22A2, SLCO1B3, SOD2, SULT1A1, TPMT, TYMS, UGT1A1, and UMPS.

[0299] In another implementation, the bait set selects a codon in a target gene or gene product, or a fragment thereof, which is associated with cancer. Exemplary genes or gene products include, but not limited to, ABL1 (e.g., codon 315), AKT1, ALK, APC (e.g., codon 1114, 1338, 1450, and 1556), AR, BRAF (e.g., codon 600), CDKN2A, CEBPA, CTNNB1 (e.g., codon 32, 33, 34, 37, 41, and 45), EGFR (e.g., 719, 746-750, 768, 790, 858, and 861), ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3 (e.g., codon 835), HRAS (e.g., codon 12, 13, and 61), JAK2 (e.g., codon 617), KIT (e.g., codon 816), KRAS (e.g., codon 12, 13, and 61), MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, PDGFRA, PIK3CA (e.g., codon 88, 542, 545, 546, 1047, and 1049), PTEN (e.g., codon 130, 173, 233, and 267), RB1, RET (e.g., codon 918), TP53 (e.g., 175, 245, 248, 273, and 306)

[0300] In yet another implementation, the bait set selects a target gene or gene product, or a fragment thereof, which is associated with cancer. Exemplary genes or gene products include, but not limited to, ABL2, AKT2, AKT3, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CARD11, CBL, CCND1, CCND2, CCND3, CCNE1, CDH1, CDH2, CDH20, CDH5, CDK4, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDR, LRP1B, LTK, MAP2K1, MAP2K2, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK3, PAK3, PAX5, PDGFRB, PIK3R1, PKHD1, PLCG1, PRKDC, PTCH1, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SMO, SOX10, SOX2, SRC, STK11, TBX22, TET2, TGFBR2, TMPRSS2, TOP1, TSC1, TSC2, USP9X, VHL, and WT1.

[0301] The length of the bait sequence can be between about 70 nucleotides and 1000 nucleotides. In one implementation, the bait length is between about 100 and 300 nucleotides, 110 and 200 nucleotides, or 120 and 170 nucleotides, in length. In addition to those mentioned above, intermediate oligonucleotide lengths of about 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 300, 400, 500, 600, 700, 800, and 900 nucleotides in length can be used in the methods described herein. In some implementations, oligonucleotides of about 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, or 230 bases can be used.

[0302] Each bait sequence can include a target-specific (e.g., a member-specific) bait sequence and universal tails on one or both ends. As used herein, the term "bait sequence" can refer to the target-specific bait sequence or the entire oligonucleotide including the target-specific "bait sequence" and other nucleotides of the oligonucleotide. The target-specific sequences in the baits are between about 40 nucleotides and 1000 nucleotides in length. In one implementation, the target-specific sequence is between about 70 nucleotides and 300 nucleotides in length. In another implementation, the target-specific sequence is between about 100 nucleotides and 200 nucleotides in length. In yet another implementation, the target-specific sequence is between about 120 nucleotides and 170 nucleotides in length, typically 120 nucleotides in length. Intermediate lengths in addition to those mentioned above also can be used in the methods described herein, such as target-specific sequences of about 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 300, 400, 500, 600, 700, 800, and 900 nucleotides in length, as well as target-specific sequences of lengths between the above-mentioned lengths.

[0303] In one implementation, the bait is an oligomer (e.g., comprised of RNA oligomers, DNA oligomers, or a combination thereof) about 50 to 200 nucleotides in length (e.g., about 50, 60, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 190, or 200 nucleotides in length). In one implementation, each bait oligomer includes about 120 to 170, or typically, about 120 nucleotides, which are a target specific bait sequence. The bait can comprise additional non-target specific nucleotide sequences at one or both ends. The additional nucleotide sequences can be used, e.g., for PCT amplification or as a bait identifier. In certain implementations, the bait additionally comprises a binding entity as described herein (e.g., a capture tag such as a biotin molecule). The binding entity, e.g., biotin molecule, can be attached to the bait, e.g., at the 5'-. 3'-end, or internally (e.g., by incorporating a biotinylated nucleotide), of the bait. In one implementation, the biotin molecule is attached at the 5'-end of the bait.

[0304] In one exemplary implementation, the bait is an oligonucleotide about 150 nucleotides in length, of which 120 nucleotides are target-specific "bait sequence". The other 30 nucleotides (e.g., 15 nucleotides on each end) are universal arbitrary tails used for PCR amplification. The tails can be any sequence selected by the user. For example, the pool of synthetic oligonucleotides can include oligonucleotides of the sequence of 5'-ATCGCACCAGCGTGTN 120 CACTGCGGCTCCTCA-3' (SEQ ID NO:1) with N 120 indicating the target-specific bait sequences.

[0305] The bait sequences described herein can be used for selection of exons and short target sequences. In one implementation, the bait is between about 100 nucleotides and 300 nucleotides in length. In another implementation, the bait is between about 130 nucleotides and 230 nucleotides in length. In yet another implementation, the bait is between about 150 nucleotides and 200 nucleotides in length. The target-specific sequences in the baits, e.g., for selection of exons and short target sequences, are between about 40 nucleotides and 1000 nucleotides in length. In one implementation embodiment, the target-specific sequence is between about 70 nucleotides and 300 nucleotides in length. In another implementation, the target-specific sequence is between about 100 nucleotides and 200 nucleotides in length. In yet another implementation, the target-specific sequence is between about 120 nucleotides and 170 nucleotides in length.

[0306] In some implementation, long oligonucleotides can minimize the number of oligonucleotides necessary to capture the target sequences. For example, one oligonucleotide can be used per exon. It is known in the art that the mean and median lengths of the protein-coding exons in the human genome are about 164 and 120 base pairs, respective. Longer baits can be more specific and capture better than shorter ones. As a result, the success rate per oligonucleotide bait sequence is higher than with short oligonucleotides. In one implementation, the minimum bait-covered sequence is the size of one bait (e.g., 120-170 bases), e.g., for capturing exon-sized targets. In determining the length of the bait sequences, one also can take into consideration that unnecessarily long baits catch more unwanted DNA directly adjacent to the target. Longer oligonucleotide baits can also be more tolerant to polymorphisms in the targeted region in the DNA samples than shorter ones. Typically, the bait sequences are derived from a reference genome sequence. If the target sequence in the actual DNA sample deviates from the reference sequence, for example if it contains a single-nucleotide polymorphism (SNP), it can hybridize less efficiently to the bait and may therefore be under-represented or completely absent in the sequences hybridized to the bait sequences. Allelic drop-outs due to SNPs can be less likely with the longer synthetic baits molecules for the reason that a single mispair in, e.g., 120 to 170 bases can have less of an effect on hybrid stability than a single mismatch in, 20 or 70 bases, which are the typical bait or primer lengths in multiplex amplification and microarray capture, respectively.

[0307] For selection of targets that are long compared to the length of the capture baits, such as genomic regions, bait sequence lengths are typically in the same size range as the baits for short targets mentioned above, except that there is no need to limit the maximum size of bait sequences for the sole purpose of minimizing targeting of adjacent sequences. Alternatively, oligonucleotides can be titled across a much wider window (typically 600 bases). This method can be used to capture DNA fragments that are much larger (e.g., about 500 bases) than a typical exon. As a result, much more unwanted flanking non-target sequences are selected.Bait Synthesis

[0308] The baits can be any type of oligonucleotide, e.g., DNA or RNA. The DNA or RNA baits ("oligo baits") can be synthesized individually, or can be synthesized in an array, as a DNA or RNA bait set ("array baits"). An oligo bait, whether provided in an array format, or as an isolated oligo, is typically single stranded. The bait can additionally comprise a binding entity as described herein (e.g., a capture tag such as a biotin molecule). The binding entity, e.g., biotin molecule, can be attached to the bait, e.g., at the 5' or 3'-end of the bait, typically, at the 5'-end of the bait.

[0309] In some implementations, individual oligo baits can be added to an array bait set. In these cases, the oligo baits can be designed to target the same areas as those targeted by the array baits, and additional oligo baits can be designed and added to the standard array baits to achieve enhanced, or more thorough, coverage in certain areas of the genome. For example, additional oligo baits can be designed to target areas of poor sequencing coverage following an initial sequencing round with a standard array bait set. In some implementations, the oligo baits are designed to have a tiled effect over the area of coverage for the array bait set, or a tiled effect over the area of coverage for other oligo baits.

[0310] In one implementation, the individual oligo baits are DNA oligos that are used to supplement an RNA or DNA oligo array bait set, or a combination thereof (e.g., a commercially available array bait set). In other implementations, individual oligo baits are DNA oligos that are used to supplement an RNA or DNA oligo bait set, or a combination thereof, that is a collection of individually designed and synthesized oligos. In one implementation, the individual oligo baits are RNA oligos that are used to supplement an RNA or DNA oligo array bait set, or a combination thereof (e.g., a commercially available array bait set). In other implementations individual oligo baits are RNA oligos that are used to supplement an RNA or DNA oligo bait set, or a combination thereof, that is a collection of individually designed and synthesized oligos.

[0311] In yet another implementation, the individual oligo baits are DNA oligos that are used to supplement a DNA oligo array bait set (e.g., a commercially available array bait set), and in other implementations individual oligo baits are DNA oligos that are used to supplement a DNA oligo bait set that is a collection of individually designed and synthesized oligos.

[0312] In yet another implementation, the individual oligo baits are DNA oligos that are used to supplement a RNA oligo array bait set (e.g., a commercially available array bait set), and in other implementations individual oligo baits are DNA oligos that are used to supplement a RNA oligo bait set that is a collection of individually designed and synthesized oligos.

[0313] In yet another implementation, the individual oligo baits are RNA oligos that are used to supplement a RNA oligo array bait set (e.g., a commercially available array bait set), and in other implementations individual oligo baits are RNA oligos that are used to supplement a RNA oligo bait set that is a collection of individually designed and synthesized oligos.

[0314] In yet another implementation, the individual oligo baits are RNA oligos that are used to supplement a DNA oligo array bait set (e.g., a commercially available array bait set), and in other implementations individual oligo baits are RNA oligos that are used to supplement a DNA oligo bait set that is a collection of individually designed and synthesized oligos.

[0315] In one implementation, oligo baits are designed to target sequences in genes of particular interest, such as to achieve increased sequencing coverage of expanded gene sets.

[0316] In another implementation, oligo baits are designed to target sequences representing a subset of the genome, and are mixed and used as a pool instead of, or in addition to, array baits.

[0317] In one implementation, a first set of oligo baits is designed to target areas of poor sequencing coverage, and a second set of oligo baits is designed to target genes of particular interest. Then both sets of oligo baits are combined and, optionally, mixed with a standard array bait set to be used for sequencing.

[0318] In one implementation, an oligo bait mix is used, e.g., to simultaneously sequence targeted gene panels and to screen a panel of single nucleotide polymorphisms (SNPs) created, such as for the purpose of looking for genomic rearrangements and copy number alterations (equivalent of arrayed CGH (Comprehensive Genomic Hybridization)). For example, a panel of SNPs can first be created by the array method as array baits, and then additional DNA oligonucleotide baits can be designed to target areas of poor sequencing coverage to a targeted set of genes. Sequencing of the collection of SNPs can then be repeated with the original array bait set plus the additional oligo baits to achieve total intended sequencing coverage.

[0319] In some implementations, oligo baits are added to a standard array bait set to achieve more thorough sequencing coverage. In one implementation, oligo baits are designed to target areas of poor sequencing coverage following an initial sequencing round with a standard array bait set.

[0320] In another implementation, oligo baits are designed to target sequences in genes of particular interest. These oligo baits can be added to a standard array bait set or to existing oligo / array hybrid bait sets to achieve, e.g., increased sequencing coverage of expanded gene sets without going through an entire array bait pool re-design cycle.

[0321] Oligo baits can be obtained from a commercial source, such as NimbleGen (Roche) or Integrated DNA Technologies (IDT) for DNA oligos. Oligos can also be obtained from Agilent Technologies. Protocols for enrichment are publicly available, e.g., SureSelect Target Enrichment System.

[0322] Baits can be produced by methods described in US 2010 / 0029498 and Gnirke, A. et al. (2009) Nal. Biotechnol. 27(2):182-189. For example, biotinylated RNA baits can be produced by obtaining a pool of synthetic long oligonucleotides, originally synthesized on a microarray, and amplifying the oligonucleotides to produce the bait sequences. In some implementations, the baits are produced by adding an RNA polymerase promoter sequence at one end of the bait sequences, and synthesizing RNA sequences using RNA polymerase. In one implementation, libraries of synthetic oligodeoxynucleotides can be obtained from commercial suppliers, such as Agilent Technologies, Inc., and amplified using known nucleic acid amplification methods.

[0323] For example, a large collection of baits can be generated from a custom pool of synthetic oligonucleotides originally synthesized on an oligonucleotide array, e.g., an Agilent programmable DNA microarray. Accordingly, at least about 2,500, 5,000, 10,000, 20,000, 3,000, 40,000, 50,000, or 60,000 unique oligonucleotides can be synthesized simultaneously.

[0324] In one implementation, a minimal set of unique olignonucleotides are chosen and additional copies (e.g., alternating between reverse complements and the original forward strands) are added until the maximum capacity of the synthetic oligonucleotide array has been reached, e.g., for baits designed to capture a pre-selected set of targets (e.g., pre-selected set of exons). In another implementation, the target is represented at least twice, e.g., by synthesizing both forward and reverse-complemented oligonucleotides. Synthesizing forward and reverse-complemented oligonucleotides for a given target can provide better redundancy at the synthesis step than synthesizing the very same sequence twice. In yet another implementation, the PCR product or bait is the same for forward and reverse-complemented oligonucleotides.

[0325] The oligonucleotides from the chips are synthesized once, and then can be amplified to create a set of oligonucleotides that can be used many times. This approach generates a universal reagent that can be used as bait for a large number of selection experiments, thereby amortizing the chip cost to be a small fraction of the sequencing cost. Alternatively, bait sequences can be produced using known nucleic acid amplification methods, such as PCR, using human DNA or pooled human DNA samples as the template.

[0326] Following synthesis, the oligonucleotides can be liberated (e.g., stripped) from the array by chemical cleavage followed by removal of the protection groups and PCR amplified into double-stranded DNA using universal primers. A second round of PCR can be used to incorporate a promoter (e.g., T7, SP6, or T3 promoter) site into the amplicon, which is used to transcribe the DNA into single-stranded RNA.

[0327] In one implementation, the baits are tiled along the sequences (e.g., exons) without gaps or overlaps. For example, the baits can start at the "left"-most coding base in the strand of the reference genome sequence shown in the UCSC genome browser (e.g., 5' to 3' or 3' to 5' along the coding sequence, depending on the orientation of the gene) and additional baits are added until all coding bases are covered. In another implementation, at least two, three, four, or five baits for each target are designed, overlapping by at least about 15, 30, 45, or 60 bases. After oligonucleotide synthesis and PCR amplification using universal primers, one of the tails of the double-stranded DNA can be enzymatically followed by the degradation of one of the strands. The single-stranded products can be hybridized, made fully double stranded by filling in, and amplified by PCR. In this manner, it is possible to produce baits that contain at least about 300, 400, 500, or 600 contiguous target-specific bases which is more than can be chemically synthesized. Such long baits can be useful for applications that require high specificity and sensitivity, or for applications that do not necessarily benefit from limiting the length of the baits (e.g., capture of long contiguous genomic regions).

[0328] In one implementation, the coverage of each target can be assessed and targets that yield similar coverage can be grouped. Distinct sets of bait sequences can be created for each group of targets, further improving the representation. In another implementation, oligonucleotides from microarray chips are tested for efficacy of hybridization, and a production round of microarray chips ordered on which oligonucleotides are grouped by their capture efficacy, thus compensating for variation in bait efficacy. In yet another implementation, oligonucleotide pools can be aggregated to form a relatively small number of composite pools, such that there is little variation in capture efficacy among them.

[0329] The baits described herein can be labeled with a tag, e.g., an affinity tag. Exemplary affinity tags include, but not limited to, biotin molecules, magnetic particles, haptens, or other tag molecules that permit isolation of baits tagged with the tag molecule. Such molecules and methods of attaching them to nucleic acids (e.g., the baits used in the methods disclosed herein) are well known in the art. Exemplary methods for making biotinylated baits are described, e.g., in Gnirke A. et al., Nat. Biotechnol. 2009; 27(2):182-9.

[0330] Also known in the art are molecules, particles or devices that bind to or are capable of separating the set of tagged baits from the hybridization mixture. In one implementation, the molecule, particle, or device binds to the tag (e.g., the affinity tag). In one implementation, the molecule, particle, or device is an avidin molecule, a magnet, or an antibody or antigen-binding fragment thereof. In one implementation, the tagged baits are separated using a magnetic bead coated with streptavidin molecules.

[0331] Exemplary methods to prepare oligonucleotide libraries are described, e.g., in Gnirke A. et al., Nat. Biotechnol. 2009; 27(2):182-9, and Blumenstiel B. et al., Curr. Protoc. Hum. Genet. 2010; Chapter 18: Unit 18.4.Hvbridization Conditions

[0332] The methods featured herein include the step of contacting the library (e.g., the nucleic acid library) with a plurality of baits to provide a selected library catch. The contacting step can be effected in solution hybridization. In certain implementations, the method includes repeating the hybridization step by one or more additional rounds of solution hybridization. In some implementations, the methods further include subjecting the library catch to one or more additional rounds of solution hybridization with the same or different collection of baits.

[0333] In other implementations, the methods featured in the invention further include amplifying the library catch (e.g., by PCR). In other implementations, the library catch is not amplified.

[0334] In yet other implementations, the methods further include the step of subjecting the library catch to genotyping, thereby identifying the genotype of the selected nucleic acids.

[0335] More specifically, a mixture of several thousand bait sequences can effectively hybridize to complementary nucleic acids in a group of nucleic acids and that such hybridized nucleic acids (the subgroup of nucleic acids) can be effectively separated and recovered. In one implementation, the methods described herein use a set of bait sequences containing more than about 1,000 bait sequences, more than about 2,000 bait sequences, more than about 3,000 bait sequences, more than about 4,000 bait sequences, more than about 5,000 bait sequences, more than about 6,000 bait sequences, more than about 7,000 bait sequences, more than about 8,000 bait sequences, more than about 9,000 bait sequences, more than about 10,000 bait sequences, more than about 15,000 bait sequences, more than about 20,000 bait sequences, more than about 30,000 bait sequences, more than about 40,000 bait sequences, or more than about 50,000 bait sequences.

[0336] In some implementation, the selection process is repeated on the selected subgroup of nucleic acids, e.g., in order to increase the enrichment of selected nucleic acids. For example, after one round of hybridization, a several thousand fold enrichment of nucleic acids can be observed. After a second round, the enrichment can rise, e.g., to about 15,000-fold average enrichment, which can provide hundreds-fold coverage of the target in a single sequencer run. Thus, for experiments that require enrichment factors not achievable in a single round of hybrid selection, the methods typically include subjecting the isolated subgroup of nucleic acids (i.e., a portion or all of the target sequences) to one or more additional rounds of solution hybridization with the set of bait sequences.

[0337] Sequential hybrid selection with two different bait sequences (bait 1, bait 2) can be used to isolate and sequence the "intersection", i.e., the subgroup of DNA sequences that binds to bait 1 and to bait 2, e.g., used for applications that include but are not limited to enriching for interchromosomal. For example, selection of DNA from a tumor sample with a bait specific for sequences on chromosome 1 followed by selection from the product of the first selection of sequences that hybridize to a bait specific for chromosome 2 may enrich for sequences at chromosomal translocation junctions that contain sequences from both chromosomes.

[0338] The molarity of the selected subgroup of nucleic acids can be controlled such that the molarity of any particular nucleic acid is within a small variation of the average molarity of all selected nucleic acids in the subgroup of nucleic acids. Methods for controlling and optimizing the evenness of target representation include, but are not limited to, rational design of bait sequences based on physicochemical as well as empirical rules of probe design well known in the art, and pools of baits where sequences known or suspected to underperform are overrepresented to compensate for their intrinsic weaknesses. In some implementations, at least about 50%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, or 95% of the isolated subgroup of nucleic acids is within about 20-fold, 15-fold, 10-fold, 5-fold, 3-fold, or 2-fold of the mean molarity. In one implementation, at least about 50% of the isolated subgroup of nucleic acids is within about 3-fold of the mean molarity. In another implementation, at least about 90% of the isolated subgroup of nucleic acids is within about 10-fold of the mean molarity.

[0339] Variations in efficiency of selection can be further adjusted by altering the concentration of the baits. In one implementation, the efficiency of selection is adjusted by leveling the efficiency of individual baits within a group (e.g., a first, second or third plurality of baits) by adjusting the relative abundance of the baits, or the density of the binding entity (e.g., the hapten or affinity tag density) in reference to differential sequence capture efficiency observed when using an equimolar mix of baits, and then introducing a differential excess as much of internally-leveled group 1 to the overall bait mix relative to internally-leveled group 2.

[0340] In certain implementations, the methods described herein can achieve an even coverage of the target sequences. In one implementation, the percent of target bases having at least about 50% of the expected coverage is at least about 60%, 70%, 80%, or 90%, e.g., for short targets such as protein-coding exons. In another implementation, the percent of target bases having at least about 50% of the expected coverage is at least about 80%, 90%, or 95%, e.g., for targets that are long compared to the length of the capture baits, such as genomic regions.

[0341] Prior to hybridization, baits can be denatured according to methods well known in the art. In general, hybridization steps comprise adding an excess of blocking DNA to the labeled bait composition, contacting the blocked bait composition under hybridizing conditions with the target sequences to be detected, washing away unhybridized baits, and detecting the binding of the bait composition to the target.

[0342] Baits are hybridized or annealed to the target sequences under hybridizing conditions. "Hybridizing conditions" are conditions that facilitate annealing between a bait and target nucleic acid. Since annealing of different baits will vary depending on probe length, base concentration and the like, annealing is facilitated by varying bait concentration, hybridization temperature, salt concentration and other factors well known in the art.

[0343] Hybridization conditions are facilitated by varying the concentrations, base compositions, complexities, and lengths of the baits, as well as salt concentrations, temperatures, and length of incubation. For example, hybridizations can be performed in hybridization buffer containing 5x SSPE, 5x Denhardt's, 5 mM EDTA and 0.1% SDS and blocking DNA to suppress non-specific hybridization. RNase inhibitors can be used if the bait is RNA. In general, hybridization conditions, as described above, include temperatures of about 25 °C to about 65 °C, typically about 65 °C, and incubation lengths of about 0.5 hours to about 96 hours, typically about 66 hours. Additional exemplary hybridization conditions are in Example 12A-12C and Table 14 h\erein.

[0344] The methods described herein are adaptable to standard liquid handling methods and devices. In some implementations, the method is carried out using automated liquid handling technology as is known in the art, such as devices that handle multiwell plates (see e.g., Gnirke, A. et al. (2009) Nat Biotechnol. 27(2):182-189), This can include, but not limited to, automated library construction, and steps of solution hybridization including set-up and post-solution hybridization washes. For example, an apparatus can be used for carrying out such automated methods for the bead-capture and washing steps after the solution hybridization reaction. Exemplary apparatus can include, but not limited to, the following positions: a position for a multi-well plate containing streptavidin-coated magnetic beads, a position for the multiwall plate containing the solution hybrid-selection reactions, I / O controlled heat blocks to preheat reagents and to carry out washing steps at a user-defined temperature, a position for a rack of pipet tips, a position with magnets laid out in certain configurations that facilitate separation of supernatants from magnet-immobilized beads, a washing station that washes pipet tips and disposed of waste, and positions for other solutions and reagents such as low and high-stringency washing buffers or the solution for alkaline elution of the final catch. In one implementation, the apparatus is designed to process up to 96 hybrid selections from the bead-capture step through the catch neutralization step in parallel. In another implementation, one or more positions have a dual function. In yet another implementation, the user is prompted by the protocol to exchange one plate for another.

[0345] The directly selected nucleic acids can be concatenated and sheared, which is done to overcome the limitations of short sequencing reads. In one implementation, each exon-sized sequencing target is captured with a single bait molecule that is about the same size as the target and has endpoints near the endpoints of the target. Only hybrids that form double strand molecules having approximately 100 or more contiguous base pairs survive stringent post-hybridization washes. As a result, the selected subgroup of nucleic acids (i.e., the "catch") is enriched for randomly sheared genomic DNA fragments whose ends are near the ends of the bait molecules. Mere end-sequencing of the "catch" with very short sequencing reads can give higher coverage near the end (or even outside) of the target and lower coverage near the middle.

[0346] Concatenating "catch" molecules by ligation and followed by random shearing and shotgun sequencing is one method to get sequence coverage along the entire length of the target sequence. This method produces higher percentage of sequenced bases that are on target (as opposed to near target) than end sequencing with very short reads. Methods for concatenating molecules by co-ligation are well known in the art. Concatenation can be performed by simple blunt end ligation. "Sticky" ends for efficient ligation can be produced by a variety of methods including PCR amplification of the "catch" with PCR primers that have restriction sites near their 5' ends followed by digestion with the corresponding restriction enzyme (e.g., NotI) or by strategies similar to those commonly used for ligation-independent cloning of PCR products such as partial "chew-back" by T4 DNA polymerase (Aslanidis and de Jong, Nucleic Acids Res. 18:6069-6074, 1990) or treatment of uracil-containing PCR products with UDG glycosylase and lyase endo VIII (e.g., New England Biolabs cat. E5500S).

[0347] In another implementation, a staggered set of bait molecules is used to target a region, obtaining frequent bait ends throughout the target region. In some implementations, merely end-sequenced "catch" (i.e., without concatenation and shearing) provides fairly uniform sequence coverage along the entire region that is covered by bait including the actual sequencing target (e.g., an exon). As staggering the bait molecules widens the segment covered by bait, the sequenced bases are distributed over a wider area. As a result, the ratio of sequence on target to near target is lower than for selections with non-overlapping baits that often require only a single bait per target.

[0348] In another implementation, end sequencing with slightly longer reads (e.g., 76 bases) is the typical method for sequencing short selected targets (e.g., exons). Unlike end sequencing with very short reads, this method leads to a unimodal coverage profile without a dip in coverage in the middle. This method is easier to perform than the concatenate and shear method described above, results in relatively even coverage along the targets, and generates a high percentage of sequenced bases fall on bait and on target proper.

[0349] In one implementation, the selected subgroup of nucleic acids are amplified (e.g., by PCR) prior to being analyzed by sequencing or genotyping. In another implementation, the subgroup is analyzed without an amplification step, e.g., when the selected subgroup is analyzed by sensitive analytical methods that can read single molecules.Bait Module

[0350] Methods described herein provide for optimzed sequencing of a large number of genes and gene products from samples, e.g., tumor samples, from one or more subjects by the appropriate selection of baits, e.g., baits for use in solution hybridization, for the selection of target nucleic acids to be sequenced. The efficiency of selection for various subgenomic intervals, or classes thereof, are matched according to bait sets having preselected efficiency of selection.

[0351] Thus a method (e.g., element (b) of the method recited above) comprises contacting the library with a plurality of baits to provide selected members (sometimes referred to herein as library catch).

[0352] Accordingly, a method of analyzing a sample, e.g., a tumor sample is provided. The method comprises: (a) acquiring a library comprising a plurality members from a sample, e.g., a plurality of tumor members from a tumor sample; (b) contacting the library with a bait set to provide selected members (e.g., a library catch); (c) acquiring a read for a subgenomic interval from a member, e.g., a tumor member from said library or library catch, e.g., by a method comprising sequencing, e.g., with a next generation sequencing method; (d) aligning said read by an alignment method, e.g., an alignment method described herein; and (e) assigning a nucleotide value (e.g., calling a mutation, e.g., with a Bayesian method or a method described herein) from said read for the preselected nucleotide position, thereby analyzing said tumor sample, wherein the method comprises contacting the library with at plurality, e.g., at least two, three, four, or five, of bait sets, wherein each bait set of said plurality has a unique (as opposed to the other bait sets in the plurality), preselected efficiency for selection. E.g., each unique bait set provides for a unique depth of sequencing.

[0353] In an implementation, the efficiency of selection of a first bait set in the plurality differs from the efficiency of a second bait set in the plurality by at least 2 fold. In an implementation, the first and second bait sets provide for a depth of sequencing that differs by at least 2 fold.

[0354] In an implementation, the method comprises contacting one, or a plurality of the following bait sets with the library: a) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 500X or higher sequencing depth, e.g., to sequence a mutation present in no more than 5 % of the cells from the sample; b) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 200X or higher, e.g., about 200X-about 500X, sequencing depth, e.g., to sequence a mutation present in no more than 10 % of the cells from the sample; c) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 10-100X sequencing depth, e.g., to sequence one or more subgenomic intervals (e.g., exons) that are chosen from: a) a pharmacogenomic (PGx) single nucleotide polymorphism (SNP) that may explain the ability of patient to metabolize different drugs, b) a genomic SNPs that may be used to uniquely identify (e.g., fingerprint) a patient, c) a genomic SNPs / loci that may be used to assess copy number gains / losses of genomic DNA and loss-of-heterozygosity (LOH); d) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 5-50 X sequencing depth, e.g., to detect a structural breakpoint, such as a genomic translocation or an indel. For example, detection of an intronic breakpoint requires 5-50X sequence-pair spanning depth to ensure high detection reliability. Such bait sets can be used to detect, for example, translocation / indel-prone cancer genes; or e) a bait set that selects sufficient members comprising a subgenomic interval to provide for about 0.1-300X sequencing depth, e.g., to detect copy number changes. In one implementation, the sequencing depth ranges from about 0.1-10X sequencing depth to detect copy number changes. In other implementations, the sequencing depth ranges from about 100-300X to detect a genomic SNPs / loci that is used to assess copy number gains / losses of genomic DNA or loss-of-heterozygosity (LOH). Such bait sets can be used to detect, for example, amplification / deletion-prone cancer genes.

[0355] In implementations, the method comprises the use of baits designed to capture two or more different target categories, each category having a different bait design strategies. In implementations, the hybrid capture methods and compositions disclosed herein capture a defined subset of target sequences (e.g., target members) and provide homogenous coverage of the target sequence, while minimizing coverage outside of that subset. In one implementation, the target sequences include the entire exome out of genomic DNA, or a selected subset thereof. The methods and compositions disclosed herein provide different bait sets for achieving different depths and patterns of coverage for complex target nucleic acid sequences (e.g., nucleic acid libraries).

[0356] In an implementation the method comprises providing selected members of a nucleic acid library (e.g., a library catch). The method includes: providing a library (e.g., a nucleic acid library) comprising a plurality of members, e.g., target nucleic acid members (e.g., including a plurality of tumor members, reference members, and / or PGx members); contacting the library, e.g., in a solution- or array-based reaction, with a plurality of baits (e.g., oligonucleotide baits) to form a hybridization mixture comprising a plurality of bait / member hybrids; separating the plurality of bait / member hybrids from said hybridization mixture, e.g., by contacting said hybridization mixture with a binding entity that allows for separation of said plurality of bait / member hybrid, thereby providing a library-catch (e.g., a selected or enriched subgroup of nucleic acid molecules from the library), wherein the plurality of baits includes two or more of the following: a) a first bait set that selects a high-level target (e.g., one or more tumor members that include a subgenomic interval, such a gene, an exon, or a base) for which the deepest coverage is required to enable a high level of sensitivity for an alteration (e.g., one or more mutations) that appears at a low frequency, e.g., about 5% or less (i.e., 5% of the cells from the sample harbor the alteration in their genome). In one implementation; the first bait set selects (e.g., is complementary to) a tumor member that includes an alteration (e.g., a point mutation) that requires about 500X or higher sequencing depth; b) a second bait set that selects a mid-level target (e.g., one or more tumor members that include a subgenomic interval, such as a gene, an exon, or a base) for which high coverage is required to enable high level of sensitivity for an alteration (e.g., one or more mutations) that appears at a higher frequency than the high-level target in a), e.g., a frequency of about 10% (i.e., 10% of the cells from the sample harbor the alteration in their genome). In one implementation; the second bait set selects (e.g., is complementary to) a tumor member that includes an alteration (e.g., a point mutation) that requires about 200X or higher sequencing depth; c) a third bait set that selects a low-level target (e.g., one or more PGx members that includes a subgenomic interval, such as a gene, an exon, or a base) for which low-medium coverage is required to enable high level of sensitivity, e.g., to detect heterozygous alleles. For example, detection of heterozygous alleles requires 10-100X sequencing depth to ensure high detection reliability. In one implementation, third bait set selects one or more subgenomic intervals (e.g., exons) that are chosen from: a) a pharmacogenomic (PGx) single nucleotide polymorphism (SNP) that may explain the ability of patient to metabolize different drugs, b) a genomic SNPs that may be used to uniquely identify (e.g., fingerprint) a patient, c) a genomic SNPs / loci that may be used to assess copy number gains / losses of genomic DNA and loss-of-heterozygosity (LOH); d) a fourth bait set that selects a first intron target (e.g., a member that includes an intron sequence) for which low-medium coverage is required, e.g., to detect a structural breakpoint, such as a genomic translocation or an indel. For example, detection of an intronic breakpoint requires 5-50X sequence-pair spanning depth to ensure high detection reliability. Said fourth bait sets can be used to detect, for example, translocation / indel-prone cancer genes; or e) a fifth bait set that selects a second intron target (e.g., an intron member) for which sparse coverage is required to improve the ability to detect copy number changes. For example, detection of a one-copy deletion of several terminal exons requires 0.1-10X coverage to ensure high detection reliability. Said fifth bait sets can be used to detect, for example, amplification / deletion-prone cancer genes.

[0357] Any combination of two, three, four or more of the aforesaid bait sets can be used in methods and compositions featured herein, such as, for example, a combination of the first and the second bait sets; first and third bait sets; first and fourth bait sets; first and fifth bait sets; second and third bait sets; second and fourth bait sets; second and fifth bait sets; third and fourth bait sets; third and fifth bait sets; fourth and fifth bait sets; first, second and third bait sets; first, second and fourth bait sets; first, second and fifth bait sets; first, second, third, fourth bait sets; first, second, third, fourth and fifth bait sets, and so on.

[0358] In one implementation, each of the first, second, third, fourth, or fifth bait set has a preselected efficiency for selection (e.g., capture). In one implementation, the value for efficiency of selection is the same for at least two, three, four of all five baits according to a)-e). In other implementations, the value for efficiency of selection is different for at least two, three, four of all five baits according to a)-e).

[0359] In some implementations, at least two, three, four, or all five bait sets have a preselected efficiency value that differ. For example, a value for efficiency of selection chosen from one of more of: (i) the first preselected efficiency has a value for first efficiency of selection that is at least about 500X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the second, third, fourth or fifth preselected efficiency of selection (e.g., about 2-3 fold greater than the value for the second efficiency of selection; about 5-6 fold greater than the value for the third efficiency of selection; about 10 fold greater than the value for the fourth efficiency of selection; about 50 to 5000-fold greater than the value for the fifth efficiency of selection); (ii) the second preselected efficiency has a value for second efficiency of selection that is at least about 200X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the third, fourth or fifth preselected efficiency of selection (e.g., about 2 fold greater than the value for the third efficiency of selection; about 4 fold greater than the value for the fourth efficiency of selection; about 20 to 2000-fold greater than the value for the fifth efficiency of selection); (iii) the third preselected efficiency has a value for third efficiency of selection that is at least about 100X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the fourth or fifth preselected efficiency of selection (e.g., about 2 fold greater than the value for the fourth efficiency of selection; about 10 to 1000-fold greater than the value for the fifth efficiency of selection); (iv) the fourth preselected efficiency has a value for fourth efficiency of selection that is at least about 50X or higher sequencing depth (e.g., has a value for efficiency of selection that is greater than the fifth preselected efficiency of selection (e.g., about 50 to 500-fold greater than the value for the fifth efficiency of selection); or (v) the fifth preselected efficiency has a value for fifth efficiency of selection that is at least about 10X to 0.1X sequencing depth.

[0360] In certain implementations, the value for efficiency of selection is modified by one or more of: differential representation of different bait sets, differential overlap of bait subsets, differential bait parameters, or mixing of different bait sets. For example, a variation in efficiency of selection (e.g., relative sequence coverage of each bait set / target category) can be adjusted by altering one or more of: (i) Differential representation of different bait sets - The bait set design to capture a given target (e.g., a target member) can be included in more / fewer number of copies to enhance / reduce relative target coverage depths; (ii) Differential overlap of bait subsets - The bait set design to capture a given target (e.g., a target member) can include a longer or shorter overlap between neighboring baits to enhance / reduce relative target coverage depths; (iii) Differential bait parameters - The bait set design to capture a given target (e.g., a target member) can include sequence modifications / shorter length to reduce capture efficiency and lower the relative target coverage depths; (iv) Mixing of different bait sets - Bait sets that are designed to capture different target sets can be mixed at different molar ratios to enhance / reduce relative target coverage depths; (v) Using different types of oligonucleotide bait sets -In certain embodiments, the bait set can include: (a) one or more chemically (e.g., non-enzymatically) synthesized (e.g., individually synthesized) baits, (b) one or more baits synthesized in an array, (c) one or more enzymatically prepared, e.g., in vitro transcribed, baits; (d) any combination of (a), (b) and / or (c), (e) one or more DNA oligonucleotides (e.g., a naturally or non-naturally occurring DNA oligonucleotide), (f) one or more RNA oligonucleotides (e.g., a naturally or non-naturally occurring RNA oligonucleotide), (g) a combination of (e) and (f), or (h) a combination of any of the above.

[0361] The different oligonucleotide combinations can be mixed at different ratios, e.g., a ratio chosen from 1:1, 1:2, 1:3, 1:4, 1:5, 1:10, 1:20, 1:50; 1:100, 1:1000, or the like. In one embodiment, the ratio of chemically-synthesized bait to array-generated bait is chosen from 1:5, 1:10, or 1:20. The DNA or RNA oligonucleotides can be naturally- or non-naturally-occurring. In certain implementations, the baits include one or more non-naturally-occurring nucleotide to, e.g., increase melting temperature. Exemplary non-naturally occurring oligonucleotides include modified DNA or RNA nucleotides. Exemplary modified nucleotides (e.g., modified RNA or DNA nucleotides) include, but are not limited to, a locked nucleic acid (LNA), wherein the ribose moiety of an LNA nucleotide is modified with an extra bridge connecting the 2' oxygen and 4' carbon; peptide nucleic acid (PNA), e.g., a PNA composed of repeating N-(2-aminoethyl)-glycine units linked by peptide bonds; a DNA or RNA oligonucleotide modified to capture low GC regions; a bicyclic nucleic acid (BNA) or a crosslinked oligonucleotide; a modified 5-methyl deoxycytidine; and 2,6-diaminopurine. Other modified DNA and RNA nucleotides are known in the art.

[0362] In certain implementations, a substantially uniform or homogeneous coverage of a target sequence (e.g., a target member) is obtained. For example, within each bait set / target category, uniformity of coverage can be optimized by modifying bait parameters, for example, by one or more of: (i) Increasing / decreasing bait representation or overlap can be used to enhance / reduce coverage of targets (e.g., target members), which are under / over-covered relative to other targets in the same category; (ii) For low coverage, hard to capture target sequences (e.g., high GC content sequences), expand the region being targeted with the bait sets to cover, e.g., adjacent sequences (e.g., less GC-rich adjancent sequences); (iii) Modifying a bait sequence can be made to reduce secondary structure of the bait and enhance its efficiency of selection; (iv) Modifying a bait length can be used to equalize melting hybridization kinetics of different baits within the same category. Bait length can be modified directly (by producing baits with varying lengths) or indirectly (by producing baits of consistent length, and replacing the bait ends with arbitrary sequence); (v) Modifying baits of different orientation for the same target region (i.e. forward and reverse strand) may have different binding efficiencies. The bait set with either orientation providing optimal coverage for each target may be selected; (vi) Modifying the amount of a binding entity, e.g., a capture tag (e.g. biotin), present on each bait may affect its binding efficiency. Increasing / decreasing the tag level of baits targeting a specific target may be used to enhance / reduce the relative target coverage; (vii) Modifying the type of nucleotide used for different baits can be altered to affect binding affinity to the target, and enhance / reduce the relative target coverage; or (viii) Using modified oligonucleotide baits, e.g., having more stable base pairing, can be used to equalize melting hybridization kinetics between areas of low or normal GC content relative to high GC content.

[0363] In other implementations, the efficiency of selection is adjusted by leveling the efficiency of individual baits within a group (e.g., a first, second or third plurality of baits) by adjusting the relative abundance of the baits, or the density of the binding entity (e.g., the hapten or affinity tag density) in reference to differential sequence capture efficiency observed when using an equimolar mix of baits, and then introducing a differential excess of internally-leveled group 1 to the overall bait mix relative to internally-leveled group 2.

[0364] In an implementation, a library catch is provided by use of a plurality of bait sets including a bait set that selects a tumor member, e.g., a nucleic acid molecule comprising a subgenomic interval from a tumor cell (also referred to herein as "a tumor bait set"). The tumor member can be any nucleotide sequence present in a tumor cell, e.g., a mutated, a wild-type, a PGx, a reference or an intron nucleotide sequence (e.g., a member), as described herein, that is present in a tumor or cancer cell. In one implementation, the tumor member includes an alteration (e.g., one or more mutations) that appears at a low frequency, e.g., about 5% or less of the cells from the tumor sample harbor the alteration in their genome. In other implementations, the tumor member includes an alteration (e.g., one or more mutations) that appears at a frequency of about 10% of the cells from the tumor sample. In other implementations, the tumor member includes a subgenomic interval from a PGx gene or gene product, an intron sequence, e.g., an intron sequence as described herein, a reference sequence, that is present in a tumor cell.

[0365] In other implementations, the method further includes detecting a non-tumor member, e.g., a nucleic acid molecule (such as a subgenomic interval) that is present in a non-tumor cell. In one implementation, the plurality of bait sets includes a bait set that selects the non-tumor member (also referred to herein as "a non-tumor bait set"). For example, the non-tumor member can be from a normal (e.g., non-cancerous) reference sample (e.g., form the same subject from whom the tumor sample was obtained); a normal adjacent tissue (NAT) or a blood sample from the same subject having or at risk of having the tumor. In other implementations, the non-tumor member is from a different subject as the tumor member (e.g., is from a normal (e.g., non-cancerous) reference sample; a normal adjacent tissue (NAT); or a blood sample), from one or more different subjects (e.g., healthy subjects or other subjects having or at risk of having the tumor). In one implementation, the non-tumor member includes a subgenomic interval from a PGx gene or gene product, an intron sequence, a reference sequence, that is present in a non-tumor cell.

[0366] In one implementation, the tumor bait set is chosen from one, two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, or all A-M of the following: A. A bait set that selects an exon sequence that includes a single nucleotide alteration associated with a cancerous phenotype; B. A bait set that selects an in-frame deletion of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more codons from a reference nucleotide (e.g., chromosome) sequence; C. A bait set that selects an intragenic deletion; D. A bait set that selects an intragenic insertion; E. A bait set that selects a deletion of a full gene; F. A bait set that selects an inversion, e.g., an intrachromosomal inversion; G. A bait set that selects an interchromosal translocation; H. A bait set that selects a tandem duplication, e.g., an intrachromosomal tandem duplication; I. A bait set that selects a nucleotide sequence of interest flanked by adjacent non-repetitive sequences; J. A bait set that selects one or more subgenomic intervals corresponding to a fusion sequence, e.g., a preselected pair of subgenomic intervals (e.g., a preselected pair of exons) corresponding to a fusion sequence (e.g., a fusion transcript or a cancer associated alternative spliced form of non-fusion transcript); K. A bait set that selects a subgenomic interval adjacent to a nucleotide sequence that includes an undesirable feature, e.g., a nucleotide sequence of high GC content, a nucleotide sequence including one or more repeated elements and / or inverted repeats; L. A bait set that selects a rearrangement, e.g., a genomic rearrangement (e.g., a rearrangement that includes an intron sequence, e.g., a 5' or 3'-UTR); or M. A bait set that selects a subgenomic interval that includes an exon adjacent to a cancer associated gene fusion.

[0367] Additional implementations of the bait sets and methods of using them are as follows: In one implementation, the bait set selects a member by hybridization (e.g., a bait or plurality of baits in the bait set is complementary to one or more members, e.g., target members, such as first-fifth members, tumor or non-tumor members, as described herein).

[0368] In one implementation, the library (e.g., the nucleic acid library) includes a plurality of members, e.g., target nucleic acid members from 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30 or more subjects. In one implementation, the subject is human having, or at risk of having, a cancer or tumor.

[0369] In certain implementations, the method includes sequencing tumor members from tumor samples from at least X subjects, (wherein X = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 30, 40, or more subjects). In one implementation, the subject is human having, or at risk of having, a cancer or tumor. The method includes sequencing at least 5, 10, 15, 20, 30, 40, 50, 75 or more genes or gene products described herein (e.g., genes or gene products from Table 1, 1A, 2, 3, or 4) from at least X subjects.

[0370] In other implementations or in addition to the aforesaid implementations, the method or includes sequencing a reference subgenomic interval from a gene or gene product from the same subject as the tumor sample, e.g., a wild-type or a non-mutated nucleotide sequence of a gene or gene product described herein (e.g., genes or gene products from Table 1, 1A, 2, 3, or 4). In one implementation, the reference gene or gene product is from the same subject or a different subject as the tumor sample (e.g., is from one or more of the same or a different tumor sample; a normal (e.g., non-cancerous) sample; a normal adjacent tissue (NAT); or a blood sample), from the same subject having or at risk of having the tumor, or from a different subject.

[0371] In one implementation, the member (e.g., any of the members described herein) comprises a subgenomic interval. In one implementation, the subgenomic interval includes an intragenic region or an intergenic region. In one implementation, the subgenomic interval includes a gene or fragment thereof, an exon or a fragment thereof, or a preselected nucleotide position (e.g., a base). In another implementation, the subgenomic interval includes an exon or an intron, or a fragment thereof, typically an exon or a fragment thereof. In one implementation, the subgenomic interval includes a coding region or a non-coding region, e.g., a promoter, an enhancer, a 5' untranslated region (5' UTR), or a 3' untranslated region (3' UTR), or a fragment thereof.

[0372] In another implementation, the subgenomic interval of the member (e.g., any of the members described herein) includes an alteration (e.g., one or more mutations) associated, e.g., positively or negatively, with a cancerous phenotype (e.g., one or more of cancer risk, cancer progression, cancer treatment or resistance to cancer treatment). In yet another implementation, the subgenomic interval includes an alteration, e.g., a point or a single mutation, a deletion mutation (e.g., an in-frame deletion, an intragenic deletion, a full gene deletion), an insertion mutation (e.g., intragenic insertion), an inversion mutation (e.g., an intra-chromosomal inversion), a linking mutation, a linked insertion mutation, an inverted duplication mutation, a tandem duplication (e.g., an intrachromosomal tandem duplication), a translocation (e.g., a chromosomal translocation, a non-reciprocal translocation), a rearrangement (e.g., a genomic rearrangement (e.g., a rearrangement of one or more introns, or a fragment thereof; a rearranged intron can include a a 5'- and / or 3'- UTR); a change in gene copy number; a change in gene expression; a change in RNA levels, or a combination thereof. In one implementation, the subgenomic interval of the first or the second member includes an alteration of a gene or gene product according to Table 1, 1A, 3, or 4.

[0373] In one implementation, the tumor member includes one or more alterations (e.g., one or more altered or mutated subgenomic intervals from gene or gene products from a tumor sample). In some implementations, the bait set (e.g., any of the bait sets described herein) selects (e.g., is complementary to) a tumor member, e.g., a nucleic acid molecule (e.g., a subgenomic interval, such as a gene, an exon, or a fragment thereof), that includes an alteration (e.g., one or more mutations) associated. e.g., positively or negatively, with a cancerous phenotype.

[0374] In an implementation, the member is associated with a cancerous phenotype, e.g., one or more of cancer risk, cancer progression, cancer treatment, or resistance to cancer treatment. The association with the cancerous phenotype can include one or more of: a genetic risk factor for cancer, a positive treatment response predictor, a negative treatment response predictor, a positive prognostic factor, a negative prognostic factor, or a diagnostic factor. In one implementation, the cancerous phenotype associated with the tumor member is the same tumor type as detected by histological analysis of the sample. In other implementations, the cancerous phenotype associated with the tumor member is from a different tumor type as detected by histological analysis of the sample.

[0375] In certain implementations, the subgenomic interval includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is predictive of a positive clinical outcome, and / or responsiveness to therapy. In other implementations, the subgenomic interval includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is predictive of a negative clinical outcome, and / or responsiveness to therapy. In certain implementations, the subgenomic interval of the nucleic acid sample includes a nucleotide sequence, wherein the presence or absence of a preselected allelic variant is indicative of a genetic (e.g., a germline risk) factor for developing cancer (e.g., the gene or gene product is chosen from one or more of BRCA1, BRCA2, EGFR, HRAS, KIT, MPL, ALK, PTEN, RET, APC, CDKN2A, MLH1, MSH2, MSH6, NF1, NF2, RB1, TP53, VHL or WT1).

[0376] In other implementations, the member is not associated with the cancerous phenotype. In certain implementations, the subgenomic interval of the member (e.g., any of the members described herein) includes a nucleic acid molecule (in the same or a different subgenomic interval) not associated with the cancerous phenotype for the tumor of the type from the sample.

[0377] In one implementation, the subgenomic interval of the member (e.g., any of the members described herein) includes a wild-type or a non-mutated nucleotide sequence of a gene or gene product (e.g., an exon sequence or a fragment thereof). In one implementation, the subgenomic interval of the first or the second member includes a wild-type or a non-mutated nucleotide sequence of a gene or gene product that when mutated is associated with a cancerous phenotype (e.g., a wild type or a non-mutated sequence of a gene or gene product as described herein, e.g., a gene or gene product described herein in Table 1, 1A, 3 or 4). Members containing the wild-type or non-mutated gene or gene product sequence are also referred to herein as "reference members." For example, the subgenomic interval is from one or more of: a wild type allele of a heterozygous mutation; a normal (e.g., non-cancerous) reference sample (e.g., from the same subject from whom the tumor sample was obtained); a normal adjacent tissue (NAT) or a blood sample from the same subject having or at risk of having the tumor. In other implementations, the subgenomic interval is from a different subject as the tumor member (e.g., is from one or more of the same or a different tumor sample from a different subject; a normal (e.g., non-cancerous) reference sample; a normal adjacent tissue (NAT); or a blood sample), from one or more different subjects (e.g., healthy subjects or other subjects having or at risk of having the tumor).

[0378] In one implementation, the first bait set, or the tumor bait set, selects (e.g., is complementary to) a subgenomic interval that includes a point mutation that appear at a frequency of about 5% or less (i.e. 5% of the cells from which the sample was prepared harbor this mutation in their genome), e.g., requires about 500X or higher sequencing depth to ensure high detection reliability.

[0379] In other implementations, the first bait set, or the tumor bait set, selects (e.g., is complementary to) a tumor or reference member chosen from one, two, three, four, five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more of: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH 1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53 gene or gene product. In one implementation, the first bait set, or the tumor bait set, selects (e.g., is complementary to) one, two, three, four, five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, thirty-five codons chosen from one or more of: codon 315 of the ABL1 gene; codon 1114, 1338, 1450 or 1556 of APC; codon 600 of BRAF; codon 32, 33, 34, 37, 41 or 45 of CTNNB1; codon 719, 746-750, 768, 790, 858 or 861 of EGFR; codon 835 of FLT3; codon 12, 13, or 61 of HRAS; codon 617 of JAK2; codon 816 of KIT; codon 12, 13, or 61 of KRAS; codon 88, 542, 545, 546, 1047, or 1049 of PIK3CA; codon 130, 173, 233, or 267 of PTEN; codon 918 of RET; codon 175, 245, 248, 273, or 306 of TP53.

[0380] In one implementation, the first bait set, or the tumor bait set, selects one or more subgenomic intervals that are frequently mutated in certain types of cancer, e.g., at least 5, 10, 20, 30 or more subgenomic intervals from a Priority 1 Cancer gene or gene product according to Table 1 or 1A.

[0381] In other implementations, the second bait set selects (e.g., is complementary to) a tumor member that includes an alteration (e.g., a point mutation) that appears at a frequency of 10%, e.g., requires about 200X or higher sequencing depth to ensure high detection reliability.

[0382] In other implementations, the second bait set selects (e.g., is complementary to) a tumor member chosen one, two, three, four, five, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHA5, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1, TSC1, TSC2, USP9X, VHL, or WT1 gene or gene product.

[0383] In one implementation, the second bait set, or the tumor bait set, selects one or more subgenomic intervals (e.g., exons) that are chosen from at least 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100 or more of the Cancer genes or gene products according to Table 1 or 1A.

[0384] In certain implementations, the first or the second bait set, or the tumor bait set, selects a wild-type and / or a non-mutated nucleotide sequence, e.g., a reference member that has a wild-type or a non-mutated nucleotide sequence, e.g., a wild-type and / or a non-mutated nucleotide sequence of a subgenomic interval of a gene or gene product as described herein, e.g., as described in Table 1, 1A, 3 or 4.

[0385] In one implementation, the first or the second bait set, or the tumor set, selects a member, e.g., a reference member, that has a wild-type or a non-mutated nucleotide sequence of a gene or gene product (e.g., an exon sequence or a fragment thereof) that when mutated is associated, e.g., positively or negatively, with a cancerous phenotype.

[0386] In one implementation, the reference member is from the same subject as the tumor member (e.g., is from one or more of the same or a different tumor sample; a wild-type heterozygous allele of the mutated member; a normal (e.g., non-cancerous) reference sample; a normal adjacent tissue (NAT); or a blood sample), from the same subject having or at risk of having the tumor. In other implementations, the reference member is from a different subject as the tumor member (e.g., is from one or more of the same or a different tumor sample from a different subject; a normal (e.g., non-cancerous) reference sample; a normal adjacent tissue (NAT); or a blood sample), from one or more different subjects having or at risk of having the tumor.

[0387] In one implementation, the first or second bait set, or the tumor bait set, selects an exon sequence that includes a single nucleotide alteration associated with a cancerous phenotype. For example, the first bait set, or the tumor bait set, can include a nucleotide sequence complementary to nucleotides 25,398,215-25,398,334 of chromosome 12, and contains a base complementary to a C-T substitution at position 25,398,286, which represents a G12S mutation in the KRAS gene.

[0388] In another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by an in-frame deletion of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more codons from a reference nucleotide (e.g., chromosome) sequence. In one implementation, the first bait set, or the tumor bait set, includes (or consists of) two discontinuous nucleotide sequences of a reference chromosome sequence, in their reference 5' to 3' orientation, separated on the reference chromosome sequence by a gap of any of 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51, 54, 57, 60 or more nucleotides. For example the first bait set, or the tumor bait set, can include a nucleotide sequence that is complementary to nucleotides 55,242,400 to 55,242,535 of chromosome 7, but lacks nucleotides 55,242,464 to 55,242,479, which represents an in-frame deletion of codons 746-750 of the EGFR gene.

[0389] In yet another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by an intragenic deletion. In one implementation, the first bait set, or the tumor bait set, includes (or consists of) two discontinuous segments of a reference nucleotide (e.g., chromosome) sequence, in their reference 5' to 3' orientation, separated by 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60 nucleotides from the reference chromosome sequence. For example, the first bait set, or the tumor bait set, can include a nucleotide sequence that is complementary to nucleotides 9,675,214 to 89,675,274 of chromosome 10, followed by bases 89,675,277 to 89,675,337 of chromosome 10, which represents a deletion of the dinucleotide sequence "CA" from codon 64 of the PTEN gene.

[0390] In yet another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by an intragenic insertion. In one implementation, the first bait set, or the tumor bait set, includes (or consists of) two continuous segments of a reference nucleotide (e.g., chromosome) sequence, separated by a non-reference sequence of 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60 nucleotides. For example, the first bait set, or the tumor bait set, can include a nucleotide sequence that is complementary to nucleotides 89,692,864 to 89,692,924 of chromosome 10, followed by a nucleotide sequence "GGNATG", followed by nucleotides 89,692,925 to 89,692,980 of chromosome 10, which represents the insertion of amino acid residues "Gly-Met" following codon 136 of the PTEN gene.

[0391] In another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by a deletion of a full gene. In one implementation, the first bait set, or the tumor bait set, includes (or consists of) two discontinuous segments of a reference nucleotide (e.g., chromosome) sequence, in their reference 5' to 3' orientation, separated by 500, 1000, 1500, 2000, 2500, 3000, 4000, 5000 or more nucleotides from the reference chromosome sequence. For example, the first bait set, or the tumor bait set, can include a nucleotide sequence complementary to bases 21,961,007 to 21,961,067 of chromosome 9 adjacent to bases 22,001,175 to 22,001.235 of chromosome 9, which represents the deletion of the CDKN2A gene.

[0392] In another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by an inversion, e.g., an intrachromosomal inversion. In one implementation, the first bait set, or the tumor bait set, includes a nucleotide sequence complementary to two discontinuous segments of a reference nucleotide (e.g., chromosome) sequence, one of which is inverted from its reference orientation, e.g., to capturing a member that results from an inversion. For example, the first bait set, or the tumor bait set, can include nucleotides 42,522,893 to 42,522,953 of chromosome 2, juxtaposed with nucleotides 29,449,993 to 29,449,933 of chromosome 2, which represents an inversion producing an EML4:ALK fusion.

[0393] In another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by an interchromosal translocation. In one implementation, the first bait set, or the tumor bait set, includes a nucleotide sequence complementary to two discontinuous segments of a reference nucleotide (e.g., genomic) sequence, originating from different reference chromosome sequences, e.g., to capture a member that results from an interchromosomal translocation. For example, the first bait set, or the tumor bait set, can include nucleotides 23,632,552 to 23,632,612 of chromosome 22, juxtaposed with nucleotides 133,681,793 to 133,681,853 of chromosome 9, which represents the presence of a chromosomal translocation resulting in a BCR-ABL fusion.

[0394] In yet another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized by a tandem duplication, e.g., an intrachromosomal tandem duplication. In one implementation, the first bait set, or the tumor bait set, includes a nucleotide sequence complementary to one segment of a reference nucleotide (e.g., chromosome) sequence, of at least 3, 6, 9, 12, 15, 18, 21, 24, 27, or 30 nucleotides in length, repeated at least once, e.g., twice, three times, four times, or five times, in its reference orientation, e.g., to capture a member has a tandem duplication. For example, a bait can include bases 28,608,259 to 28,608,285 of chromosome 13 repeated twice in the same orientation, which represents an internal tandem duplication (ITD) mutation in the FLT3 gene.

[0395] In yet another implementation, the first or the second bait set, or the tumor bait set, selects a tumor member characterized bv a nucleotide sequence of interest flanked by adjacent non-repetitive sequences. In one implementation, the first bait set, or the tumor bait set, includes at least two non-contiguous nucleotide sequences. A first nucleotide sequence complementary to the 5' flanking region of the sequence of interest, and a second nucleotide sequence complementary to the 3' flanking region of the sequence of interest. For example, a first and second pair of baits can include a first nucleotide sequence complementary to nucleotides 51,288,380 to 51,288,500 (bait 1) and a second nucleotide sequence complementary to nucleotides 51,288,560 to 51,288,680 (bait 2) of chromosome 2, which can capture members containing the microsatellite marker sequence D2S123.

[0396] In another implementation, the first or the second bait set, or the tumor bait set, selects (e.g., is complementary to) a preselected pair of subgenomic intervals (e.g., a preselected pair of exons) corresponding to a fusion sequence (e.g., a fusion transcript or a cancer associated alternative spliced form of non-fusion transcript).

[0397] In other implementations, the first or the second bait set, or the tumor bait set, selects a subgenomic interval adjacent to a nucleotide sequence that includes an undesirable feature, e.g., a nucleotide sequence of high GC content. a nucleotide sequence including one or more repeated elements and / or inverted repeats. In one implementation, the first bait set, or the tumor bait set, selects a subgenomic interval that includes a repeated element, but does not hybridize to the repeated element (e.g., does not hybridize to the repeated elements in a BRCA2 gene).

[0398] In other implementations, the first, the second, or the tumor, bait set selects a subgenomic interval that includes an exon adjacent to a cancer associated gene fusion, to thereby facilitate the capture of nucleic acid sequences (e.g., cDNA fragments) adjacent to the gene fusion.

[0399] In other implementations, the first, the second, or the tumor, bait set selects a subgenomic interval that is from one or more genes or gene products shown in Table 1, 1A, 3 or 4, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of the cancer types described in Tables 1. 1A, 3 or 4.

[0400] In other implementations, the first bait set, or the tumor bait set, selects an ABL-1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a soft-tissue malignancy chosen from one or more of CML, ALL or T-ALL. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of havine. one or more of CML, ALL or T-ALL.

[0401] In other implementations, the first bait set, or the tumor bait set, selects an AKT1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, colorectal, ovarian, or non-small cell lung carcinoma (NSCLC). In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of breast, colorectal, ovarian, or non-small cell lung carcinoma (NSCLC).

[0402] In other implementations, the first bait set, or the tumor bait set, selects an ALK gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of ALCL, NSCLC or neuroblastoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having. one or more of ALCL, NSCLC or neuroblastoma.

[0403] In other implementations, the first bait set, or the tumor bait set, selects an APC gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, pancreatic, desmoid, hepatoblastoma, glioma, or other CNS cancers or tumors. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of colorectal, pancreatic, desmoid. hepatoblastoma, glioma, or other CNS cancers or tumors.

[0404] In other implementations, the first bait set, or the tumor bait set, selects a BRAF gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, colorectal cancer, lung cancer, other epithelial malignancies, or hematological malignancies including AML or ALL. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of melanoma, colorectal cancer, lung cancer, other epithelial malignancies. or hematological malignancies including AML or ALL.

[0405] In other implementations, the first bait set, or the tumor bait set, selects a CDKN2A gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, pancreatic, or other tumor types. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of melanoma, pancreatic, or other tumor types.

[0406] In other implementations, the first bait set, or the tumor bait set, selects a CEBPA gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of AML or MDS. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of AML or MDS.

[0407] In other implementations, the first bait set, or the tumor bait set, selects a CTNNB 1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, ovarian, hepatoblastoma, or pleomorphic alivary adenoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of colorectal, ovarian, hepatoblastoma, or pleomorphic salivary adenoma.

[0408] In other implementations, the first bait set, or the tumor bait set, selects an EGFR gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of glioma, lung cancer, or NSCLC. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of glioma, lung cancer, or NSCLC.

[0409] In other implementations, the first bait set, or the tumor bait set, selects an ERBB2 gene or gene product, or a subgenomic interval thereof, that is associated, e.g., positively or negatively, with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, ovarian, NSCLC, gastric or other solid tumors. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of breast, ovarian, NSCLC, gastric or other solid tumor types.

[0410] In other implementations, the first bait set, or the tumor bait set, selects an ESR1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, ovarian or endometrial tumors. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of breast, ovarian or endometrial tumors.

[0411] In other implementations, the first bait set, or the tumor bait set, selects an FGFR1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of MPD or NHL. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of MPD or NHL.

[0412] In other implementations, the first bait set, or the tumor bait set, selects an FGFR2 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g.. a cancer chosen from one or more of gastric, NSCLC or endometrial tumors. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of gastric, NSCLC or endometrial tumors.

[0413] In other implementations, the first bait set, or the tumor bait set, selects an FGFR3 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of bladder cancer, multiple myeloma or T-cell lymphoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of bladder cancer, multiple myeloma or T-cell lymphoma.

[0414] In other implementations, the first bait set, or the tumor bait set, selects an FLT3 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, colorectal, papillary thyroid, ovarian, non small-cell lung cancer (NSCLC), cholangiocarcinoma, or pilocytic astrocytoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of melanoma, colorectal, papillary thyroid, ovarian, non small-cell lung cancer (NSCLC), cholangiocarcinoma, or pilocytic astrocytoma.

[0415] In other implementations, the first bait set, or the tumor bait set, selects an HRAS gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of rhadomvosarcoma, ganglioneuroblastoma, bladder, sarcomas, or other cancer types. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of rhadomyosarcoma. ganglioneuroblastoma, bladder, sarcomas, or other cancer types.

[0416] In other implementations, the first bait set, or the tumor bait set, selects a JAK2 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of ALL, AML, MPD or CML. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of ALL. AML, MPD or CML.

[0417] In other implementations, the first bait set, or the tumor bait set, selects a KIT gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of gastrointestinal stromal tumors (GIST), AML, TGCT, mastocytosis, mucosal melanoma, or epithelioma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of gastrointestinal stromal tumors (GIST), AML, TGCT, mastocytosis, mucosal melanoma, or epithelioma.

[0418] In other implementations, the first bait set, or the tumor bait set, selects a KRAS gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of pancreatic, colon, colorectal, lung, thyroid, or AML. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of pancreatic, colon, colorectal, lung, thyroid, or AML.

[0419] In other implementations, the first bait set, or the tumor bait set, selects a MET gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of renal or head-neck squamous cell carcinoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of renal or head-neck squamous cell carcinoma.

[0420] In other implementations, the first bait set, or the tumor bait set, selects an MLL gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of AML or ALL. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of AML or ALL.

[0421] In other implementations, the first bait set selects (e.g., is complementary to) an NF1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of neurofibroma or glioma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of neurofibroma or glioma.

[0422] In other implementations, the first bait set, or the tumor bait set, selects a NOTCH1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a T-ALL cancer. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, a T-ALL cancer.

[0423] In other implementations, the first bait set, or the tumor bait set, selects an NPM1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of NHL, APL or AML. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of NHL, APL or AML.

[0424] In other implementations, the first bait set, or the tumor bait set, selects an NRAS gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of melanoma, colorectal cancer, multiple myeloma, AML, or thyroid cancer. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of melanoma, colorectal cancer, multiple myeloma, AML, or thyroid cancer.

[0425] In other implementations, the first bait set, or the tumor bait set, selects a PDGFRA gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of GIST or idiopathic hypereosinophilic syndrome. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of GIST or idiopathic hypereosinophilic syndrome.

[0426] In other implementations, the first bait set, or the tumor bait set, selects a PIK3CA gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, gastric, gliobastoma, or breast cancer. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of colorectal, gastric, gliobastoma, or breast cancer.

[0427] In other implementations, the first bait set, or the tumor bait set, selects a PTEN gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of colorectal, glioma, prostate, or endometrial cancer. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of colorectal, glioma, prostate, or endometrial cancer.

[0428] In other implementations, the first bait set, or the tumor bait set, selects an RB1 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of retinoblastoma, sarcoma, breast, or small cell lung carcinoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of retinoblastoma, sarcoma, breast, or small cell lung carcinoma.

[0429] In other implementations, the first bait set, or the tumor bait set, selects a RET gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of medullary thyroid, papillary thyroid, or pheochromocytoma. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of medullary thyroid, papillary thyroid, or pheochromocytoma.

[0430] In other implementations, the first bait set, or the tumor bait set, selects a TP53 gene or gene product, or a subgenomic interval thereof, that is associated with a cancerous phenotype, e.g., a cancer chosen from one or more of breast, colorectal, lung, sarcoma, adrenocortical, glioma, or other tumor types. In one implementation, the library, e.g., the nucleic acid library, is obtained from a sample from a subject having, or at risk of having, one or more of breast, colorectal, lung, sarcoma, adrenocortical. glioma, or other tumor types.

[0431] In one implementation, the first bait set, or the tumor bait set, selects a gene or gene product, or a subgenomic interval thereof, that is a positive predictor of therapeutic response. Examples of a positive predictor of a therapeutic response include, but are not limited to, an activating mutation in the EGFR gene that predicts responsiveness to small molecule EGFR TKIs (e.g., Iressalgefitinib) in NSCLC patients; presence of an EML4 / ALK fusion gene predicts responsiveness to ALK inhibitors (e.g. PF-02341066) in NSCLC patients; presence of a BRAF V600E mutation predicts responsiveness to BRAF inhibition (e.g. PLX-4032) in melanoma patients.

[0432] In other implementations, the first bait set, or the tumor bait set, selects a gene or gene product, or a subgenomic interval thereof, that is a negative predictor of therapeutic response. Examples of a negative predictor of a therapeutic response include, but are not limited to, an activating mutation in the KRAS gene that predict lack of response to anti-EGFR monoclonal antibodies (cetuximab, panitumumab) in CRC patients; and the presence of an M351T mutation in the BCR / Abl fusion gene predicts resistance to Gleevec / imatinib in CML patients.

[0433] In other implementations, the first bait set, or the tumor bait set, selects a gene or gene product, or a subgenomic interval thereof, that is a prognostic factor. Examples of prognostic factors include, but are not limited to, the presence of an insertion mutation in the FLT3 gene, which is a negative prognostic for relapse in AML patients; the presence of specific RET gene mutations, e.g. M918T, which are negative prognostic factors for survival in medullary thyroid carcinoma patients.

[0434] In other implementations, the first bait set, or the tumor bait set, selects a gene or gene product, or a subgenomic interval thereof, that is a diagnostic factor. Examples of prognostic factors include, but are not limited to, the presence of a BCR / Abl fusion gene, which is diagnostic for CML; and the presence of a SMARCB1 mutation, which is diagnostic of Rhabdoid tumor of the kidney.

[0435] In yet other implementations, the first or second bait set, or the tumor bait set, selects a nucleic acid molecule (e.g., a subgenomic interval) that includes an alteration that is associated with tumor progression and / or resistance, and has a late onset in cancer progression ...

Claims

1. A method of analyzing a tumor nucleic acid sample, comprising: (a) acquiring a library comprising a plurality of nucleic acid molecules from the tumor nucleic acid sample; (b) contacting the library with a plurality of bait sets to provide a library catch being enriched for preselected subgenomic intervals of the tumor nucleic acid sample; (c) acquiring reads for the subgenomic intervals from the tumor nucleic acid sample from said library catch by a next generation sequencing method; (d) aligning said reads to a reference sequence by an alignment method; and (e) assigning a nucleotide value from said reads for a preselected nucleotide position, thereby analysing said tumor nucleic acid sample; wherein each bait set is a plurality of nucleic acid molecules which can hybridize to and thereby capture a target nucleic acid; wherein each bait set provides for a level or depth of sequence coverage that is adjusted for selection for its target subgenomic interval; and wherein the plurality of bait sets comprises a first bait set and a second bait set, wherein the first and second bait sets provide for a depth of sequencing that differs by at least 2 fold, and wherein the plurality of bait sets comprise at least two, three, four, or five of the following: (i) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an exon sequence that includes a single nucleotide alteration associated with a cancerous phenotype; (ii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an in-frame deletion of one or more codons from a reference nucleotide sequence; (iii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an intragenic deletion; (iv) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an intragenic insertion; (v) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a deletion of a full gene; (vi) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an inversion; (vii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising an interchromosomal translocation; (viii) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a tandem duplication; (ix) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a fusion sequence; (x) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a genomic rearrangement that comprises an intron sequence; or (xi) a bait set that hybridizes to and thereby captures a target nucleic acid in a subgenomic interval comprising a genomic rearrangement that includes an intron sequence from a 5' or 3'-UTR.

2. The method of claim 1, wherein the plurality of nucleic acid molecules of the library from step (a) comprises fragmented DNA.

3. The method of claim 2, wherein the fragmented DNA is sheared or enzymatically prepared genomic DNA.

4. The method of any one of claims 1 to 3, wherein the plurality of nucleic acid molecules of the library from step (a) comprises a sequence not derived from a subject, e.g., an adapter sequence, a primer sequence or a barcode sequence.

5. The method of any one of claims 1 to 4, wherein the plurality of bait sets from step (b) each comprise a binding entity that allows for the capture and separation of a hybrid formed by a bait and a nucleic acid hybridized to the bait.

6. The method of any one of claims 1 to 5, wherein the plurality of bait sets from step (b) is suitable for solution phase hybridization.

7. The method of any one of claims 1 to 6, wherein the tumor nucleic acid sample comprises one or more mutations associated with a cancerous phenotype, e.g., cancer risk, cancer progression, cancer treatment, or resistance to cancer treatment.

8. The method of any one of claims 1 to 7, wherein the amount of nucleic acid used to generate library of step (a) comprises less than 5 micrograms, less than 1 microgram, less than 500 ng, less than 100 ng, or less than 50 ng of fragmented DNA.

9. The method of claim 8, wherein the amount of nucleic acid sample used to generate the library of step (a) comprises less than 100 ng of fragmented DNA.

10. The method of claim 8, wherein the amount of nucleic acid sample used to generate the library of step (a) comprises less than 50 ng of fragmented DNA.

11. The method of any one of claims 1 to 10, wherein the method comprises sequencing a subgenomic interval chosen from: at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty or more genes or gene products from the sample, wherein the genes or gene products are chosen from: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, BRAF, CCND1, CDK4, CDKN2A, CEBPA, CTNNB1, EGFR, ERBB2, ESR1, FGFR1, FGFR2, FGFR3, FLT3, HRAS, JAK2, KIT, KRAS, MAP2K1, MAP2K2, MET, MLL, MYC, NF1, NOTCH1, NPM1, NRAS, NTRK3, PDGFRA, PIK3CA, PIK3CG, PIK3R1, PTCH1, PTCH2, PTEN, RB1, RET, SMO, STK11, SUFU, or TP53.

12. The method of any one of claims 1 to 10, wherein the method comprises sequencing a subgenomic interval chosen from at least five, six, seven, eight, nine, ten, fifteen, twenty, twenty-five, thirty, thirty-five, forty, forty-five, fifty, fifty-five, sixty, sixty-five, seventy, seventy-five, eighty, eighty-five, ninety, ninety-five, one hundred, one hundred and five, one hundred and ten, one hundred and fifteen, one hundred and twenty or more of subgenomic intervals from a mutated or wild type gene or gene product chosen from at least five or more of: ABL2, ARAF, ARFRP1, ARID1A, ATM, ATR, AURKA, AURKB, BAP1, BCL2, BCL2A1, BCL2L1, BCL2L2, BCL6, BRCA1, BRCA2, CBL, CARD11, CBL, CCND2, CCND3, CCNE1, CD79A, CD79B, CDH1, CDH2, CDH20, CDH5, CDK6, CDK8, CDKN2B, CDKN2C, CHEK1, CHEK2, CRKL, CRLF2, DNMT3A, DOT1L, EPHA3, EPHAS, EPHA6, EPHA7, EPHB1, EPHB4, EPHB6, ERBB3, ERBB4, ERG, ETV1, ETV4, ETV5, ETV6, EWSR1, EZH2, FANCA, FBXW7, FGFR4, FLT1, FLT4, FOXP4, GATA1, GNA11, GNAQ, GNAS, GPR124, GUCY1A2, HOXA3, HSP90AA1, IDH1, IDH2, IGF1R, IGF2R, IKBKE, IKZF1, INHBA, IRS2, JAK1, JAK3, JUN, KDM6A, KDR, LRP1B, LRP6, LTK, MAP2K4, MCL1, MDM2, MDM4, MEN1, MITF, MLH1, MPL, MRE11A, MSH2, MSH6, MTOR, MUTYH, MYCL1, MYCN, NF2, NKX2-1, NTRK1, NTRK2, PAK3, PAX5, PDGFRB, PKHD1, PLCG1, PRKDC, PTPN11, PTPRD, RAF1, RARA, RICTOR, RPTOR, RUNX1, SMAD2, SMAD3, SMAD4, SMARCA4, SMARCB1, SOX10, SOX2, SRC, TBX22, TET2, TGFBR2, TMPRSS2, TNFAIP3, TNK, TNKS2, TOP1, TSC1, TSC2, USP9X, VHL, or WT1.

13. The method of any one of claims 1 to 12, wherein the method further comprises amplifying the nucleic acid molecules in the library catch.

14. The method of any one of claims 1 to 13, wherein the alignment method of step (d) includes: selecting a rearrangement reference sequence for alignment with a read, wherein said rearrangement reference sequence is preselected to align with a preselected rearrangement, e.g., wherein the reference sequence is not identical to the genomic rearrangement; and comparing, e.g., aligning, a read with said preselected rearrangement reference sequence.