Microsatellite instability signatures

Abstract

To provide microsatellite instability signatures.SOLUTION: Provided herein are methods for determining microsatellite instability (MSI) based on tracts of nucleotide repeats and somatic mutations in nucleic acid from patient samples. The methods provided herein include determining genomic signature scores and MSI scores for classification of samples as MSI or microsatellite-stable (MSS). The methods provided herein are useful for selecting patients for immunotherapy.SELECTED DRAWING: None

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit under 35 U.S.C. § 119(e) of currently pending U.S. application Ser. No. 62 / 988,571, filed March 12, 2020. The disclosure of the prior application is considered part of the disclosure of this application and is incorporated by reference into the disclosure of this application. FIELD OF THEINVENTION

[0002] The present invention relates generally to the detection of microsatellite instability (MSI), and more specifically to the analysis of nucleotide tract repeat lengths and genomic signatures of somatic mutations. Incorporating sequence tables

[0003] The data in the attached sequence listing is incorporated into this application by reference. The attached sequence listing text file, named PGDX3100-1 WO_SL.txt, was created on Mar. 8, 2021 and is 750 bytes. The file can be accessed using Microsoft Word on a computer using the Windows OS. [Background technology]

[0004] Background information Microsatellite instability (MSI) in tumors is a biomarker that guides the selection of immunotherapies, such as checkpoint inhibitor therapy, for the treatment of patients. Microsatellite instability can be detected by analysis of next generation sequencing (NGS) data and / or analysis of amplified fragment lengths. These assays examine the length distribution of sequenced or amplified microsatellites. However, microsatellites are notoriously difficult to amplify or sequence with high accuracy.

[0005] Individual substitutions can be accurately detected with detection limits lower than the length of individual microsatellites, for example, somatic mutations are often located in regions of the genome that are easier to sequence precisely than microsatellites. Summary of the Invention [Means for solving the problem]

[0006] The present invention relates to the detection of microsatellite instability using analysis of microsatellite tracts and somatic mutations.

[0007] In some embodiments, provided herein is a method of determining microsatellite instability (MSI), comprising: (i) determining the presence of somatic allele lengths in a plurality of nucleotide repeat tracts of sequenced DNA in a sample obtained from a subject; (ii) determining the presence of somatic mutations in DNA in regions outside the plurality of nucleotide repeat tracts; (iii) determining a match between a genomic signature of somatic mutations found outside the plurality of nucleotide repeat tracts and a genomic signature of mismatch repair deficiency; (iv) applying rules to the result of the determining step to obtain an MSI score; and (v) classifying the sample as microsatellite instability-high (MSI-H) or microsatellite stable (MSS) based on the MSI score.

[0008] In some embodiments, the methods provided herein further comprise generating a report comprising the MSI score and a classification of the sample as microsatellite instability-high (MSI-H) or microsatellite stable (MSS). In some embodiments, the sample is a tumor sample.

[0009] In some embodiments, the plurality of nucleotide repeat tracts of the methods provided herein comprise mononucleotide repeats, dinucleotide repeats, trinucleotide repeats, tetranucleotide repeats, pentanucleotide repeats, hexanucleotide repeats, heptanucleotide repeats, octanucleotide repeats, or combinations thereof. In some embodiments, the plurality of nucleotide repeat tracts comprise mononucleotide repeats. In some embodiments, the nucleotide repeat tract in the MSI-H sample is shorter than the nucleotide repeat tract in the reference sample. In some embodiments, the nucleotide repeat tract in the MSI-H sample is at least 2 base pairs shorter than the nucleotide repeat tract in the reference sample. In some embodiments, the reference sample is a matched normal sample or an MSS sample.

[0010] In some embodiments, the methods provided herein further comprise determining the frequency of shorter tracts of nucleotide repeats. In some embodiments, the plurality of nucleotide repeat tracts of the methods provided herein comprises one or more microsatellite markers. In some embodiments, the one or more microsatellite markers are selected from human genome microsatellite markers. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; D5S346; D2S123; D17S250; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; D5S346; D2S123; D17S250; BAT40; and any combination thereof.

[0011] In some embodiments, the genomic signature of the somatic mutation of the method provided herein comprises 11 base pairs surrounding the somatic mutation. In some embodiments, the method provided herein further comprises assigning a genomic signature score to the somatic mutation. In some embodiments, the somatic mutation is classified as associated with an MSI-H sample or an MSS sample based on the genomic signature score.

[0012] In some embodiments, the methods provided herein further comprise selecting or administering a treatment to the patient based on the MSI score. In some embodiments, the MSI score indicates that the patient sample is microsatellite instability-high (MSI-H) and the treatment comprises an immune checkpoint inhibitor. In some embodiments, the immune checkpoint inhibitor comprises an antibody. In some embodiments, the antibody is selected from the group consisting of an anti-PD-1 antibody; an anti-IDO antibody; an anti-CTLA-4 antibody; an anti-PD-L1 antibody; and an anti-LAG-3 antibody. In some embodiments, the checkpoint inhibitor is pembrolizumab (KEYTRUDA®), nivolumab (OPDIVO®), atezolizumab (TECENTRIQ®), or ipilimumab (YERVOY®).

[0013] In some embodiments, the sequenced DNA comprises one or more sequenced genomes, one or more sequenced exomes, or regions of one or more sequenced genomes or one or more sequenced exomes. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 shows the data analysis in more than two dimensions.

[0015] [Diagram 2] Figure 2 shows that high dimensional data requires a large amount of training data.

[0016] [Diagram 3] Figure 3 shows PGDx Cerebro, a machine learning approach for somatic mutation discovery.

[0017] [Figure 4] FIG. 4 shows detection of short alleles at mononucleotide tracts.

[0018] [Diagram 5] Figure 5 shows detection of mismatch repair deficiency mutation signatures (https: / / cancer.sanger.ac.uk / cosmic / signatures_v2; Catalogue of Somatic Mutations in Cancer (COSMIC) available at the Wellcome Sanger Institute).

[0019] [Figure 6] Figure 6 shows the match between exome signature scores and PCR. Representative exome data is shown.

[0020] [Figure 7] FIG. 7 shows the results obtained by combining the detection of short alleles in mononucleotide tracts with the detection of mismatch repair deficiency mutation signatures.

[0021] [Figure 8] FIG. 8 shows the linear separability of the data obtained using the ensemble method of MSI detection. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0022] Detailed Description of the Invention Many cancers are accompanied by the accumulation of mutations resulting from mismatch repair (MMR) deficiency. An important marker of MMR deficiency is microsatellite instability (MSI). The presence of MMR deficiency or MSI can serve as a marker of responsiveness to immunotherapy, such as checkpoint inhibitor therapy. The methods provided herein are useful for detecting MSI and selecting treatment regimens.

[0023] In some embodiments, provided herein is a method of determining microsatellite instability (MSI), comprising: (i) determining the presence of somatic allele lengths in a plurality of nucleotide repeat tracts of sequenced DNA in a sample obtained from a subject; (ii) determining the presence of somatic mutations in DNA in regions outside the plurality of nucleotide repeat tracts; (iii) determining a match between a genomic signature of somatic mutations found outside the plurality of nucleotide repeat tracts and a genomic signature of mismatch repair deficiency; (iv) applying rules to the result of the determining step to obtain an MSI score; and (v) classifying the sample as microsatellite instability-high (MSI-H) or microsatellite stable (MSS) based on the MSI score.

[0024] The methods provided herein may include ensemble methods. As used herein, "ensemble methods" refers to using results or data from two or more classification systems to make a final classification. The methods provided herein can use multiple classification systems to make a final classification. Any number of classification systems can be used to make a final classification, such as 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, and any number or range therebetween or more classification systems. In some embodiments, the methods provided herein use data from two classification systems. In some embodiments, the two classification systems are tract-based and mutation signature-based classification systems.

[0025] As used herein, the term "MSI-H" means "high microsatellite instability". As used herein, the term "MSI" means "microsatellite unstable" or "microsatellite unstable" as indicated by the context. As used herein, the term "MSI-L" means "low microsatellite instability". As used herein, the term "MSS" means "microsatellite stable". The MSI status can be used to classify tumors and / or tumor samples into tumor groups including microsatellite unstable and microsatellite stable tumors. As used herein, MSI tumors are tumors or samples with a higher degree of microsatellite instability compared to normal or non-tumor tissues, or compared to MSI-L or MSS tumors or samples. Thus, MSI tumor samples are samples with a higher degree of microsatellite instability compared to normal or non-tumor samples or samples from MSI-L or MSS tumors. As used herein, the terms "MSI" and "MSI-H" may be used interchangeably when referring to the MSI status of a tumor or tumor sample, unless the context clearly indicates otherwise.

[0026] The methods provided herein can be used to analyze any number of nucleotide repeat tracts.Exemplary numbers of nucleotide repeat tracts that may be analyzed include at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 31, at least 32, at least 33, at least 34, at least 35, at least 36, at least 37, at least 38, at least 39, at least 40, at least 41, at least 42, at least 43, at least 44, at least 45, at least 46, at least 47, at least 48, at least 49, at least 50, at least 51, at least 52, at least 53, at least 54, at least 55, at least 56, at least 57, at least 58, at least 59, at least 6 0, at least 61, at least 62, at least 63, at least 64, at least 65, at least 66, at least 67, at least 68, at least 69, at least 70, at least 71, at least 72, at least 73, at least 74, at least 75, at least 76, at least 77, at least 78, at least 79, at least 80, at least 81, at least 82, at least 83, at least 84, at least 85, at least 86, at least 87, at least 88, at least 89, at least 90, at least At least 91, at least 92, at least 93, at least 94, at least 95, at least 96, at least 97, at least 98, at least 99, at least 100, at least 150, at least 200, at least 250, at least 300, at least 350, at least 400, at least 450, at least 500, at least 600, at least 700, at least 800, at least 900, at least 1000, and any number or range therebetween or more nucleotide repeat tracts. Sequencing

[0027] In some embodiments, the method provided herein comprises sequencing DNA in the sample obtained from the patient.Any sequencing method can be used, including, for example, Sanger sequencing using labeled terminator or primer and gel separation in slab or capillary system, and next generation sequencing (NGS).Exemplary NGS methodologies include Roche 454 sequencer, Life Technologies SOLiD® system, Life Technologies Ion Torrent, and Illumina system, such as Illumina Genome Analyzer II, Illumina MiSeq, Illumina HiSeq, Illumina NextSeq, and Illumina NovaSeq instrument.

[0028] In some embodiments, the sequenced DNA in the sample comprises one or more sequenced genomes or tracts thereof. In some embodiments, the sequenced DNA in the sample comprises one or more sequenced exomes or tracts thereof. As used herein, the term "exome sequencing" refers to sequencing all protein-coding exons of genes in a genome. Exome sequencing can include target enrichment methods, such as, for example, array-based capture and in-solution capture of nucleic acids. Nucleotide repeats

[0029] In some embodiments, the methods provided herein include determining the presence of somatic allele lengths in multiple nucleotide repeat tracts of DNA from a patient sample. The nucleotide repeats may constitute microsatellites or short tandem repeats (STRs). Thus, as used herein, the terms "microsatellite" and "microsatellite marker" refer to polymorphic DNA loci that contain repeated nucleotide sequences. The nucleotide repeats may be of any length. Any microsatellite marker or combination of microsatellite markers may be analyzed by the methods provided herein. In some embodiments, the microsatellite markers are present in the human genome, although one skilled in the art will understand that microsatellite markers of any species may be analyzed by the methods provided herein. Microsatellite markers for analysis may be identified, for example, by scanning a reference genome for nucleotide repeats.

[0030] Typically, the DNA motif of the nucleotide repeat is repeated, for example, 5 to 50 times, although fewer or more repeats are possible. Furthermore, the nucleotide repeat tract can have any number of repeated nucleotides. For example, the nucleotide repeat can include 1 to 6 base pairs or up to 10 base pairs as the repeating DNA motif. In some embodiments, the multiple nucleotide repeat tract includes mononucleotide repeats, dinucleotide repeats, trinucleotide repeats, tetranucleotide repeats, pentanucleotide repeats, hexanucleotide repeats, heptanucleotide repeats, octanucleotide repeats, or combinations thereof. In some embodiments, the multiple nucleotide repeat tract includes mononucleotide repeats. Microsatellite markers for analysis can be identified, for example, by scanning a reference genome for mononucleotide repeats and / or repeats of any other length.

[0031] The nucleotide repeat of DNA from a patient sample may differ in length from the nucleotide repeat in a reference sample by any number of nucleotides. The difference in nucleotide repeat length may be due to a different number of repeat DNA motifs. As an example, the nucleotide repeat length of an MSI or MSI-H sample or tumor may be shorter than the nucleotide repeat length of a reference sample or tumor. As another example, the nucleotide repeat length in an MSI or MSI-H sample or tumor may be longer than the nucleotide repeat length in a reference sample or tumor. As yet another example, partial repeats of a repeat DNA motif, deletions within a repeat DNA motif, insertions within a repeat DNA motif, substitutions within a repeat DNA motif, or any combination thereof may change the length of the nucleotide repeat. The nucleotide repeat length may be longer or shorter in an MSI or MSI-H sample or tumor compared to a reference sample or tumor by any number of nucleotides or nucleotide repeats. In some embodiments, the tract of nucleotide repeats in the MSI or MSI-H sample is shorter than the tract of nucleotide repeats in one or more reference samples. In some embodiments, the tract of nucleotide repeats in the MSI or MSI-H sample is at least 2 base pairs shorter than the tract of nucleotide repeats in one or more reference samples. In some embodiments, the one or more reference samples are matched normal samples or MSS samples.

[0032] In some embodiments, the methods provided herein further comprise determining the frequency of shorter tracts of nucleotide repeats. For example, the frequency of shorter tracts of nucleotide repeats may be greater in MSI or MSI-H samples or tumors compared to the frequency of shorter tracts of nucleotide repeats in one or more reference samples. As an example, any altered allele, such as a shorter or longer tract of nucleotide repeats, may be greater than 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%. , 100%, 125%, 150%, 175%, 200%, 225%, 250%, 275%, 300%, 325%, 350%, 375%, 400%, 425%, 450%, 475%, 500%, 550%, 600%, 650%, 700%, 750%, 800%, 850%, 900%, 950%, 1000%, and any number or range therebetween, may occur more frequently than the reference allele.

[0033] The proportion of analyzed unstable nucleotide repeat tracts can be used to classify a sample as an MSI or MSI-H sample. As used herein, the term "unstable" when referring to a nucleotide repeat tract means that the length of the nucleotide repeat tract in the sample can vary. For example, the nucleotide repeat tract can be shorter or longer. In some embodiments, the length of the nucleotide repeat tract in the sample varies compared to the length of the nucleotide repeat tract in the reference sample. In some embodiments, the length of the nucleotide repeat tract varies compared to the length of the germline allele of the nucleotide repeat tract. In some embodiments, the length of the nucleotide repeat tract in the sample is shorter compared to the length of the nucleotide repeat tract in the reference sample. In some embodiments, the length of the nucleotide repeat tract in the sample is shorter compared to the length of the germline allele of the nucleotide repeat tract. In some embodiments, the length of the nucleotide repeat tract in the sample is longer compared to the length of the nucleotide repeat tract in the reference sample. In some embodiments, the length of the nucleotide repeat tract in the sample is longer as compared to the length of a germline allele of the nucleotide repeat tract.

[0034] about 50, about 51, about 52, about 53, about 54, about 55, about 56, about 57, about 58, about 59, about 60, about 61, about 62, about 63, about 64, about 65, about 66, about 67, about 68, about 69, about 70, about 71, about 72, about 73, about 74, about 75, about 76, about 77, about 78, about 79, about 80, about 81, about 82, about 83, about 84, about 85, about 86, about 87, about 88, about 89, about 90, about 91, about 92, about 93, The presence of about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, about 11, about 12, about 13, about 14, about 15 or more shorter nucleotide repeat tracts out of about 94, about 95, about 96, about 97, about 98, about 99, about 100 or more analyzed nucleotide repeat tracts can be used to classify a sample as an MSI or MSI-H sample. In some embodiments, the presence of about 8 nucleotide repeat tracts with shorter alleles out of about 68 analyzed nucleotide repeat tracts is used to classify a sample as an MSI or MSI-H sample. In some embodiments, the presence of about 68 nucleotide repeat tracts with shorter alleles out of about 11 analyzed nucleotide repeat tracts is used to classify a sample as an MSI or MSI-H sample. In some embodiments, if all nucleotide repeat tracts analyzed have the shorter allele, the sample is classified as an MSI or MSI-H sample. In some embodiments, the sample is a tumor sample.

[0035] In some embodiments, the plurality of nucleotide repeat tracts comprises one or more microsatellite markers. Exemplary microsatellite markers include BAT-25, BAT-26, MONO-27, NR-21, NR-24, BAT-40, TGFβRII, IGFIIR, hMSH3, BAX, and dinucleotide loci, such as D2S123, D9S283, D9S1851, D2S123, D17S250, and D18S58. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; D5S346; D2S123; D17S250; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; D5S346; D2S123; D17S250; BAT40; and any combination thereof. Somatic mutation detection

[0036] In some embodiments, the methods provided herein include determining the presence of somatic mutations in DNA from a patient sample. The presence of somatic mutations can be determined in DNA in regions outside of a multiple nucleotide repeat tract.

[0037] Looking at one dimension of the data may not be sufficient to separate the data into distinct classes. Rather, it is often necessary to look at multiple dimensions of the data to be able to separate the data into the correct classes (Figure 1).

[0038] Somatic mutation detection may require many dimensions due to laboratory artifacts, sequencing artifacts, and alignment artifacts that may confound analysis. Laboratory artifacts can be detected, for example, by interstrand bias, genomic context, and combination of reference alleles and alternative alleles. Sequence artifacts can be detected, for example, by low base quality or by genomic context. Exemplary parameters that may contribute to the detection of alignment artifacts include, for example, mapping quality, aligner match, and position distribution. A further confounding aspect in the detection of somatic mutations is the possibility of germline mutations. Germline mutations may exist, for example, in individuals who are not identified as having mutations, i.e., normal individuals. In addition, germline mutations can be found, for example, in databases.

[0039] High-dimensional data may require a large number of training data. For example, without sufficient and representative training data, classifiers may be inadequately adapted, leading to overfitting (Figure 2). Machine learning methods can fit large data sets much more easily and accurately than humans. A machine model called Cerebro has been developed that examines more than 70 dimensions. Furthermore, for exome data, more than 2 million labeled candidate mutations are used in this model.

[0040] The features of Cerebro for use in somatic mutation detection are shown in FIG. 3 (Wood et al., Sci. Transl. Med. 2018, Vol. 10, eaar7939). Cerebro uses machine learning models to score candidate mutations and separate true somatic mutations from germline mutations and artifacts. For example, dual alignment provides data from two alignments, thereby increasing the data for statistical analysis per mutation. Further features considered by Cerebro include analysis of mutation context, such as %GC and DUST (a measure of sequence complexity), sample coverage, such as distinct coverage and mutant allele frequency (MAF), sequence and / or alignment quality, such as mapping quality and base quality, and coverage comparison, including analysis of tumor and normal samples and application of statistics such as Fisher's exact test (FET). Thus, in some embodiments, the detection of somatic mutations in the methods provided herein includes the use of Cerebro. Genomic signatures of somatic mutations

[0041] In some embodiments, the methods provided herein include determining a genomic signature of the somatic mutation. In some embodiments, the methods provided herein include determining the match between the genomic signature of the somatic mutation found outside the multiple nucleotide repeat tract and the genomic signature of the mismatch repair deficiency. The genomic signature may include any sequence context of the somatic mutation, including sequences adjacent to, surrounding, or including the somatic mutation. The sequences adjacent to the somatic mutation may be located 5' and / or 3' of the somatic mutation and may extend any number of base pairs from the somatic mutation. As used herein, the term "adjacent" refers to the sequence immediately adjacent to the somatic mutation. In some embodiments, the sequences that constitute the sequence context or genomic signature of the somatic mutation may be located away from the somatic mutation. For example, sequences that constitute the sequence context or genomic signature of a somatic mutation may not be located directly adjacent to the somatic mutation, but may be located 1 base pair, 2 base pairs, 3 base pairs, 4 base pairs, 5 base pairs, 6 base pairs, 7 base pairs, 8 base pairs, 9 base pairs, 10 base pairs, 11 base pairs, 12 base pairs, 13 base pairs, 14 base pairs, 15 base pairs, 16 base pairs, 17 base pairs, 18 base pairs, 19 base pairs, 20 base pairs, 25 base pairs, 30 base pairs, 35 base pairs, 40 base pairs, 45 base pairs, 50 base pairs, 55 base pairs, 60 base pairs, 65 base pairs, 70 base pairs, 75 base pairs, 80 base pairs, 85 base pairs, 90 base pairs, 95 base pairs, 10 ... base pairs, 125 base pairs, 150 base pairs, 200 base pairs, 250 base pairs, 300 base pairs, 350 base pairs, 400 base pairs, 450 base pairs, 500 base pairs, 600 base pairs, 700 base pairs, 800 base pairs, 900 base pairs, 1000 base pairs, 5000 base pairs, 10 kilobase pairs (kbp), 20 kbp, 30 kbp, 40 kbp, 50 kbp, 60 kbp, 70 kbp, 80 kbp, 90 kbp, 100 kbp, 200 kbp, 300 kbp, 400 kbp, 500 kbp, 600 kbp, 700 kbp, 800 kbp, 900 kbp, 1 megabase pair (Mbp), and any number or range of distances therebetween.Such sequences may be located on either side of the somatic mutation, i.e., 5' or 3', or on both sides, i.e., 5' and 3'. In the case of sequences located 5' or 3' or both 5' and 3' of the somatic mutation, the sequence context may include the somatic mutation. In some embodiments, in the case of sequences located both 5' and 3' of the somatic mutation, the sequence context surrounds the somatic mutation.

[0042] A sequence context or genomic signature can be 1 base pair, 2 base pairs, 3 base pairs, 4 base pairs, 5 base pairs, 6 base pairs, 7 base pairs, 8 base pairs, 9 base pairs, 10 base pairs, 11 base pairs, 12 base pairs, 13 base pairs, 14 base pairs, 15 base pairs, 16 base pairs, 17 base pairs, 18 base pairs, 19 base pairs, 20 base pairs, 25 base pairs, 30 base pairs, 35 base pairs, 40 base pairs, 45 base pairs, 50 base pairs, 55 base pairs, 60 base pairs, 6 It may include any number of base pairs, such as 5 base pairs, 70 base pairs, 75 base pairs, 80 base pairs, 85 base pairs, 90 base pairs, 95 base pairs, 100 base pairs, 125 base pairs, 150 base pairs, 200 base pairs, 250 base pairs, 300 base pairs, 350 base pairs, 400 base pairs, 450 base pairs, 500 base pairs, 600 base pairs, 700 base pairs, 800 base pairs, 900 base pairs, 1000 base pairs, and any number or range therebetween. In some embodiments, the genomic signature of the somatic mutation includes 11 base pairs surrounding the somatic mutation.

[0043] The method provided herein may include assigning a genomic signature score to the somatic mutation. In some embodiments, the somatic mutation is classified as associated with MSI-H or MSI samples or MSS samples based on the genomic signature score. For example, a set of position-specific weight matrices (PWMs) can be used to evaluate the log-likelihood that a particular single base substitution (SBS) originates from an MSI or MSI-H tumor, given a reference allele, alternative alleles, and an 11nt context or genomic signature (Example 2 below) centered on the reference base.

[0044] The genomic signature of a somatic mutation may include any structural context. Thus, the method provided herein may include determining the nucleic acid structural context of a somatic mutation. As used herein, "structural context" refers to primary structure, secondary structure, tertiary structure, quaternary structure, and any combination thereof. For example, the primary structure may include sequence context. The secondary structure may include interactions between bases, such as the formation of double-stranded regions, stem-loops, hairpin loops, tetraloops, and pseudoknots. The tertiary structure may include large-scale folding and interactions of secondary structures, such as the interactions of stem loops, hairpin loops, and other secondary structural features. The quaternary structure may include interactions between nucleic acid molecules and proteins, such as the organization of DNA into nucleosomes, and interactions between separate nucleic acids, for example.

[0045] As an example, the primary structural context or sequence may affect the propensity of a base to be mutated by a particular mutational process, such as mismatch repair deficiency. As another example, the secondary structural context may affect the propensity of a base to be mutated by a particular mutational process, such as mismatch repair deficiency. As yet another example, the distance between an exon base and a splice site, such as the nearest splice site, may affect the propensity of a base to be mutated by a particular mutational process, such as mismatch repair deficiency. The structural context may also affect the relative location of somatic mutations. For example, somatic mutations may accumulate in the center or inside of an exon. In some embodiments, the tertiary and / or quaternary structure affects the propensity of a base to be mutated by a particular mutational process, such as mismatch repair deficiency. In some embodiments, the somatic mutations result from mismatch repair deficiency. In some embodiments, the somatic mutations indicate MSI or MSI-H status. In some embodiments, the somatic mutations are MSI or MSI-H mutations.

[0046] In some embodiments, a primary structure context or sequence context is determined. In some embodiments, a secondary structure context is determined. In some embodiments, a tertiary structure context is determined. In some embodiments, a quaternary structure context is determined. Any structural context and any combination of structural contexts can be determined, such as primary structure, secondary structure, tertiary structure, and quaternary structure. In some embodiments, a primary structure context, a secondary structure context, a tertiary structure context, a quaternary structure context, a distance between an exon base and a splice site, a distance between an exon base and a nearest splice site, a position of a somatic mutation within an exon, or any combination thereof is determined. Any structural context can provide a genomic signature of a somatic mutation resulting from a specific mutation process, such as mismatch repair deficiency. In some embodiments, the structural context provides a genomic signature for a somatic mutation resulting from mismatch repair deficiency. MSI score and classification

[0047] The MSI score can be obtained by applying rules to the results of the steps of determining the presence of somatic allele lengths in the multiple nucleotide repeat tracts of DNA from a sample, such as a patient sample, the presence of somatic mutations in DNA in tracts outside the multiple nucleotide repeat tracts, and the genomic signature of somatic mutations found outside the multiple nucleotide repeat tracts. In some embodiments, the sample or the tumor from which the sample was taken is classified as microsatellite instability-high (MSI-H), microsatellite instability (MSI), or microsatellite stable (MSS) based on the MSI score. In some embodiments, the methods provided herein further include generating a report that includes the MSI score and the classification of the sample as microsatellite instability-high (MSI-H), microsatellite instability (MSI), or microsatellite stable (MSS).

[0048] For example, any microsatellite marker or combination of microsatellite markers including BAT-25, BAT-26, MONO-27, NR-21, NR-24, BAT-40, TGFβRII, IGFIIR, hMSH3, BAX and dinucleotide loci such as D2S123, D9S283, D9S1851, D2S123, D17S250 and D18S58 can be used to obtain MSI scores. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; D5S346; D2S123; D17S250; and any combination thereof. In some embodiments, the one or more microsatellite markers are selected from the group consisting of BAT-25; BAT-26; MONO-27; NR-21; NR-24; D5S346; D2S123; D17S250; BAT40; and any combination thereof. Exemplary microsatellite markers are shown in FIG.

[0049] In some embodiments, one or more microsatellite markers are selected from human genome microsatellite markers. Any human microsatellite marker or combination of human microsatellite markers can be used. For example, any mammal, such as rodents (including mice, rats, hamsters and guinea pigs), cats, dogs, rabbits, cattle, horses, goats, sheep, pigs, and other livestock, and primates (including monkeys, chimpanzees, orangutans and gorillas) can also be used. The unstable tract of the microsatellite marker can be shorter or longer than the reference allele, for example, as described above. An exemplary method of obtaining MSI score is described below (Example 3 and Example 4). Computing Devices

[0050] As one of ordinary skill in the art will recognize as necessary or best suited, the performance of the methods provided herein may include one or more computing devices, computing systems, or computers including one or more of a processor (e.g., a central processing unit (CPU), a graphics processing unit (GPU), etc.), computer-readable storage devices (e.g., main memory, static memory, etc.), or combinations thereof in communication with each other via a bus.

[0051] The processor may include any suitable processor known in the art, such as the processor XEON® E7 by Intel (Santa Clara, Calif.), or the processor sold under the trademark OPTERON™ 6200 by AMD (Sunnyvale, Calif.).

[0052] The memory preferably includes at least one tangible, non-transitory medium capable of storing one or more sets of instructions executable to cause the system to perform the functions described herein (e.g., software implementing any methodology or function found herein, or computer program as mentioned above); data (e.g., pharmaceutical data, personal data, or pharmaceutical database source images); or both. Although the computer-readable storage device may be a single medium in an exemplary embodiment, the term "computer-readable storage device" should be interpreted to include a single medium or multiple media (e.g., centralized or distributed databases, and / or associated caches and servers) that store instructions or data. Thus, the term "computer-readable storage device" should be interpreted to include, but is not limited to, solid-state memory (e.g., subscriber identity module (SIM) card, secure digital card (SD card), micro SD card, or solid-state drive (SSD)), optical and magnetic media, and any other tangible storage medium.

[0053] Any suitable service may be used for storage, such as, for example, Amazon Web Services, memory of a computing system, cloud storage, a server, or other computer readable storage.

[0054] The input / output devices in accordance with the methods provided herein may include one or more of a display unit (e.g., a liquid crystal display (LCD) or cathode ray tube (CRT) monitor), an alphanumeric input device (e.g., a keyboard), a cursor control device (e.g., a mouse or trackpad), a disk drive unit, a printer, a signal generating device (e.g., a speaker), a touch screen, a button, an accelerometer, a microphone, a cellular radio frequency antenna, a network interface device, which may be, for example, a network interface card (NIC), a Wi-Fi card, or a cellular modem, or any combination thereof.

[0055] Those skilled in the art will recognize that any suitable development environment or programming language may be used to implement the methods described herein. For example, the methods herein may be implemented using Perl, Python, C++, C#, Java, JavaScript, Visual Basic, Ruby on Rails, Groovy and Grails, or any other suitable tools. For mobile devices, it may be preferable to use native xCode or Android Java. sample

[0056] In some embodiments, the methods provided herein include sequenced DNA in a sample obtained from a patient. In some embodiments, the patient is afflicted with a tumor or cancer. In some embodiments, the sample is a tumor sample. Samples from both solid and liquid tumors can be used in the methods described herein. As used herein, the term "tumor" refers to a mass or lump of tissue formed by the accumulation of abnormal cells. Tumors can be benign (i.e., not cancerous), malignant (i.e., cancer), or pre-malignant (i.e., pre-cancerous). The terms "tumor" and "neoplasm" can be used interchangeably. Generally, a cancerous tumor is malignant.

[0057] As used herein, the term "solid tumor" refers to a mass of abnormal tissue that does not usually contain cysts or fluid. Exemplary solid tumors include, for example, sarcomas and carcinomas. As used herein, the term "liquid tumor" refers to a tumor or cancer that resides in bodily fluids, such as blood and bone marrow. Exemplary liquid tumors include hematopoietic tumors, such as leukemia and lymphoma, despite the ability of lymphoma to grow as a solid tumor, for example by growing in lymph nodes. The term "liquid tumor" can be used interchangeably with the term "blood cancer", unless the context clearly indicates otherwise.

[0058] Samples from any cancer can be analyzed by the methods provided herein. Any type of cancer can be analyzed by the methods provided herein. In some embodiments, the cancer is selected from breast cancer, pancreatic cancer, lung cancer, melanoma, skin cancer, hematopoietic cancer, leukemia, lymphoma, colon cancer, rectal cancer, kidney cancer, renal cell carcinoma, bladder cancer, oral cancer, pharyngeal cancer, thyroid cancer, head and neck cancer, brain cancer, bone cancer, muscle cancer, sarcoma, rhabdomyosarcoma, ovarian cancer, cervical cancer, uterine cancer, prostate cancer, etc. Thus, the patient of the methods provided herein can suffer from any cancer. In some embodiments, the patient is afflicted with breast cancer, pancreatic cancer, lung cancer, melanoma, skin cancer, hematopoietic cancer, leukemia, lymphoma, colon cancer, rectal cancer, kidney cancer, renal cell carcinoma, bladder cancer, oral cancer, pharyngeal cancer, thyroid cancer, head and neck cancer, brain cancer, bone cancer, muscle cancer, sarcoma, rhabdomyosarcoma, ovarian cancer, cervical cancer, uterine cancer, prostate cancer, or any other cancer.

[0059] Any sample or type of sample can be used in the methods provided herein. In some embodiments, the sample is blood, saliva, plasma, serum, urine, or other biological fluid. Further exemplary biological fluids include serous fluid, lymphatic fluid, cerebrospinal fluid, mucosal secretions, vaginal fluid, peritoneal fluid, pleural fluid, pericardial fluid, peritoneal fluid, and ascites. In some embodiments, the sample is a tissue sample. In some embodiments, the sample is a tissue sample from a cancer. In some embodiments, the sample is a cell sample. In some embodiments, the sample is a cell sample from a cancer. In some embodiments, the sample is a cancer sample. The cancer sample can be a sample from a solid tumor or a liquid tumor. Patient Selection and Treatment Methods

[0060] In some embodiments, the methods provided herein include selecting or administering a treatment to a patient based on the MSI score. In some embodiments, the patient is afflicted with cancer. In some embodiments, the MSI score indicates that the patient's sample is microsatellite instability-high (MSI-H) or microsatellite instability (MSI) and the treatment includes immunotherapy. The patient can be selected for treatment with any immunotherapy, as described below. In some embodiments, the MSI score indicates that the patient's sample is microsatellite instability-high (MSI-H) or microsatellite instability (MSI) and the treatment includes an immune checkpoint inhibitor. In some embodiments, the immune checkpoint inhibitor includes an antibody, as described below. In some embodiments, the antibody is selected from the group consisting of an anti-PD-1 antibody; an anti-IDO antibody; an anti-CTLA-4 antibody; an anti-PD-L1 antibody; and an anti-LAG-3 antibody. In some embodiments, the checkpoint inhibitor is pembrolizumab (KEYTRUDA®), nivolumab (OPDIVO®), atezolizumab (TECENTRIQ®), or ipilimumab (YERVOY®). immunotherapy

[0061] Immunotherapy includes treatment with activating immunotherapy and treatment with suppressing immunotherapy. Activating immunotherapy induces or activates an immune response, while suppressing immunotherapy reduces or suppresses an immune response. Immunotherapy may include treatment with immunomodulators, such as interleukins, cytokines, chemokines, immunomodulatory imid drugs (IMiDs), and the like. Any interleukin, cytokine, chemokine, or immunomodulatory imid drug (IMiD) can be used for immunotherapy. Exemplary interleukins for immunotherapy include IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12, IL-15, IL-18, IL-21, and IL-23. Exemplary cytokines for immunotherapy include interferons, TNF-α, TGF-β, G-CSF, and GM-CSF. Exemplary chemokines for immunotherapy include CCL 3, CCL 26, and CXCL7. Exemplary IMiDs include thalidomide and its analogs lenalidomide, pomalidomide, and apremilast. Other immunomodulatory agents include, for example, cytosine phosphate-guanosine, oligodeoxynucleotides, and glucans.

[0062] Cancer immunotherapy generally involves the stimulation of the immune system to destroy cancer cells and tumors. Exemplary cancer immunotherapy includes CAR-T cell therapy, in which a chimeric antigen receptor (CAR) is introduced into a patient's T cells to generate CAR-T cells. The CAR-T cells are then introduced into the patient's bloodstream to treat the cancer by adoptive cell transfer (ACT). CARs generally include an antigen recognition domain that can target antigens expressed on the cell surface of cancer cells and one or more signaling domains. Thus, CAR-T cells can target and destroy cancer cells that express the target antigen. Exemplary CAR-T cell therapies include tisagenlecleucel (KYMRIAH®) and axicabtageneciloreucel (YESCARTA).

[0063] Further cancer immunotherapy includes TCR therapy, another type of ACT. Similar to CAR-T cell therapy, T cells are harvested from the patient, re-engineered, and introduced into the patient. Further types of ACT include tumor infiltrating lymphocyte (TIL) therapy. Patient-derived TILs are isolated from the patient's tumor tissue, expanded in vitro, and then introduced into the patient.

[0064] Yet another type of cancer immunotherapy is treatment with monoclonal antibodies. Monoclonal antibodies for use in immunotherapy can be naked, i.e. unconjugated, or conjugated, i.e. chemotherapeutic drugs or radioactive particles can be attached. In addition to monoclonal antibodies, other molecules, such as interleukins and cytokines, can be conjugated to target cancer cells. As an example, denileukin diftitox (ONTAK) contains IL-2 bound to diphtheria toxin. Furthermore, monoclonal antibodies for cancer immunotherapy can be bispecific, i.e. designed to recognize and bind to two different proteins. Thus, bispecific monoclonal antibodies can recognize, for example, two or more antigens on the surface of cancer cells. As another example, bispecific antibodies can recognize proteins or antigens on cancer cells and proteins or antigens on immune cells, thereby promoting immune cells to attack cancer cells.

[0065] Exemplary monoclonal antibodies for treating cancer include alemtuzumab (CAMPATH), trastuzumab (HERCEPTIN®), ibritumomab tiuxetan (ZEVALIN), brentuximab vedotin (ADCETRIS®), trastuzumab emtansine (KADCYLA®), blinatumomab (BLINCYTO®), bevacizumab (AVASTIN®), and cetuximab (ERBITUX).

[0066] Further cancer immunotherapies include cancer vaccines, which induce an immune response against cancer cells. Yet another cancer immunotherapy is "checkpoint inhibitor therapy," which is described further below. Checkpoint inhibitor therapy

[0067] Checkpoint inhibitor therapy is a form of cancer treatment that uses or targets immune checkpoints that affect the function of the immune system. Immune checkpoints can be stimulatory or inhibitory. Tumors can use these checkpoints to protect themselves from immune system attacks. Checkpoint therapy can block inhibitory checkpoints and restore immune system function. Checkpoint proteins include programmed cell death 1 protein (PDCD1, PD-1; also known as CD279) and its ligand, PD-1 ligand 1 (PD-L1, CD274), cytotoxic T lymphocyte-associated protein 4 (CTLA-4), A2AR (adenosine A2A receptor), B7-H3 (or CD276), B7-H4 (or VTCN1), BTLA (B and T lymphocyte attenuator, or CD272), IDO (indoleamine 2,3-dioxygenase), KIR (killer cell immunoglobulin-like receptor), LAG3 (lymphocyte activation gene 3), TIM-3 (T cell immunoglobulin domain and mucin domain 3), and VISTA (V domain Ig inhibitor of T cell activation).

[0068] Programmed cell death protein 1, also known as PD-1 and CD279 (cluster of differentiation 279), are cell surface receptors that play a key role in downregulating the immune system by suppressing T cell inflammatory activity and promoting self-tolerance. Without being limited by theory, PD-1 is an immune checkpoint that defends against autoimmunity through a dual mechanism of promoting apoptosis (programmed cell death) in antigen-specific T cells in lymph nodes while simultaneously reducing apoptosis in regulatory T cells (anti-inflammatory, inhibitory T cells). PD-1 has two ligands, PD-L1 and PD-L2, which are members of the B7 family. PD-L1 protein is upregulated on macrophages and dendritic cells (DCs) in response to LPS and GM-CSF treatment, and on T and B cells upon TCR and B cell receptor signaling, whereas in resting mice, for example, PD-L1 mRNA can be detected in the heart, lungs, thymus, spleen and kidney. PD-L1 is expressed in almost all mouse tumor cell lines, including PA1 myeloma, P815 mastocytoma, and B16 melanoma, upon treatment with IFN-γ. PD-L2 expression is more restricted, being expressed mainly by DCs and a few tumor lines.

[0069] PD-L1 is expressed in some cancers. Monoclonal antibodies targeting PD-1 can enhance the immune system for the treatment of cancer. Many tumor cells express PD-L1, an immunosuppressive PD-1 ligand; inhibition of the interaction between PD-1 and PD-L1 can enhance T cell responses in vitro and mediate preclinical antitumor activity.

[0070] CTLA4 or CTLA-4 (cytotoxic T-lymphocyte-associated protein 4), also known as CD152 (cluster of differentiation 152), is a protein receptor that functions as an immune checkpoint and downregulates immune responses. CTLA4 is constitutively expressed in regulatory T cells, but is generally upregulated in conventional T cells after activation, especially in cancer. CTLA4 is a member of the immunoglobulin superfamily that is expressed by activated T cells and transmits inhibitory signals to T cells. CTLA4 is homologous to the T cell costimulatory protein CD28, and both molecules bind to CD80 and CD86, also called B7-1 and B7-2, respectively, on antigen-presenting cells. Without being limited by theory, CTLA-4 binds to CD80 and CD86 with higher affinity and avidity than CD28, and thus can outcompete CD28 for its ligand. CTLA4 transmits inhibitory signals to T cells, while CD28 transmits stimulatory signals. CTLA4 is also found in regulatory T cells, where it contributes to their inhibitory function. T cell activation via the T cell receptor and CD28 leads to increased expression of CTLA-4.

[0071] Several checkpoint inhibitors can be used to treat cancer. PD-1 inhibitors include pembrolizumab (KEYTRUDA®) and nivolumab (OPDIVO®). PD-L1 inhibitors include, for example, atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®) and durvalumab (IMFINZI®). CTLA-4 inhibitors include, for example, iprimumab (YERVOY®). Other checkpoint inhibitors include, for example, anti-B7-H3 antibodies (MGA271), anti-KIR antibodies (lirilumab), and anti-LAG3 antibodies (BMS-986016). Any checkpoint inhibitor can be used in the methods described herein. Furthermore, the response to any checkpoint inhibitor can be determined or predicted using the methods described herein. In some embodiments, the checkpoint inhibitor is pembrolizumab (KEYTRUDA®), nivolumab (OPDIVO®), atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®), durvalumab (IMFINZI®), or ipilimumab (YERVOY®).

[0072] As used herein, the singular forms "a," "an," and "the" include the plural unless the context clearly dictates otherwise. Thus, for example, reference to a "method" includes one or more methods, and / or steps of the type described herein that will be apparent to those skilled in the art upon reading this disclosure, and so forth.

[0073] As used herein, "about" when referring to a measurable value, such as an amount, duration, and the like, is meant to encompass a variation of ±20%, or ±10%, or ±5%, or even ±1% from the specified value, where such variations are appropriate for the disclosed method or for carrying out the disclosed method. The term "about" can be used interchangeably with the term "approximately," unless the context clearly contradicts.

[0074] 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 disclosure belongs.

[0075] As used herein, the term "protein" refers to any polymeric chain of amino acids. The terms "peptide" and "polypeptide" are used interchangeably with the term "protein" and also refer to a polymeric chain of amino acids. The term "protein" encompasses natural or artificial proteins, protein fragments, and polypeptide analogs of protein sequences. Proteins may be monomeric or polymeric. The term "protein" encompasses fragments and variants thereof (including fragments of variants), unless contradicted by context.

[0076] As used herein, the term "nucleic acid" refers to any deoxyribonucleic acid (DNA) molecule, ribonucleic acid (RNA) molecule, or nucleic acid analog. DNA or RNA molecules can be double-stranded or single-stranded and can be of any size. Exemplary nucleic acids include, but are not limited to, chromosomal DNA, plasmid DNA, cDNA, cell-free DNA (cfDNA), mRNA, tRNA, rRNA, siRNA, microRNA (miRNA or miR), hnRNA. Exemplary nucleic acid analogs include peptide nucleic acid, morpholino and locked nucleic acid, glycol nucleic acid, and threose nucleic acid.

[0077] As used herein, the term "patient" refers to any individual or subject on which the methods disclosed herein are carried out. The term "patient" can be used interchangeably with the terms "individual" or "subject". The patient can be a human, but it will be understood by those skilled in the art that the patient can also be an animal. Thus, other animals, including mammals, such as rodents (including mice, rats, hamsters and guinea pigs), cats, dogs, rabbits, cows, horses, goats, sheep, pigs, and other domestic animals, and primates (including monkeys, chimpanzees, orangutans and gorillas), are included within the definition of a patient.

[0078] As used herein, the terms "treat", "treatment", "therapy", "therapeutic" and the like refer to obtaining a desired pharmacological and / or physiological effect, including, but not limited to, obtaining a beneficial or desired result with respect to a disease, disorder, or condition, such as alleviating a disease or disorder, delaying or slowing the progression, reducing the effects or symptoms, inhibiting the onset of a disease or disorder, amelioration, therapeutic benefit and / or prophylactic benefit. As used herein, "treatment" encompasses any treatment of a disease in a mammal, particularly a human, and includes (a) preventing the disease from occurring in a subject susceptible to or at risk of developing the disease, but not yet diagnosed as having it, (b) inhibiting the disease, i.e., arresting its development, and (c) relieving the disease, i.e., regressing the disease. Therapeutic benefit includes eradication or amelioration of the underlying disease being treated. Therapeutic benefit is also achieved by eradication or amelioration of one or more physiological symptoms associated with the underlying disease, such that an improvement is observed in the subject, even though the subject may still be afflicted by the underlying disease. In some cases, for preventive benefit, treatment is administered to subjects who are at risk of developing a particular disease, or who report one or more physiological symptoms of the disease, even if the disease may not have been diagnosed.The method of the present disclosure can be used with any mammal or other animal.In some cases, treatment can result in the reduction or cessation of symptoms.Preventive effects include delaying or eliminating the appearance of a disease or condition, delaying or eliminating the symptoms of a disease or condition, delaying, halting, or reversing the progression of a disease or condition, or any combination thereof.

[0079] As used herein, the terms "sample" and "biological sample" refer to any sample suitable for the method provided herein. The sample used in the method can be obtained from a tissue sample or body fluid from a subject, or tissue obtained by a biopsy procedure (e.g., needle biopsy) or surgical procedure. In certain embodiments, the biological sample of the method is, for example, a sample of body fluid, such as cerebrospinal fluid (CSF), blood, serum, plasma, urine, saliva, tears and ascites. The sample of body fluid can be collected by any suitable method known to those skilled in the art. EXAMPLES

[0080] Example 1 This example illustrates mutational signature modeling of MSI-associated mutations.

[0081] Mutagenesis processes tend to have certain "signatures," or propensities to mutate certain reference alleles to specific alternative alleles when those reference alleles are surrounded by certain sequence contexts. Microsatellite instability (MSI) is caused by mismatch repair deficiency with at least four known signatures that are publicly detailed in a collection of known cancer signatures. Using these signatures, a model was created based on a position-specific weight matrix (PWM) that can indicate the likelihood of a mutation originating from an MSI or MSI-H tumor versus a microsatellite stable (MSS) tumor. This model can then be used to score mutations within a tumor, and the sum of the tumor mutation scores results in a "tumor signature score" that can be used to determine whether the tumor itself is MSI or MSS. This score can be used alone or in an ensemble method to perform MSI / MSS classification.

[0082] The model for this method is fitted to publicly available mutation data (TCGA MC3; Ellrott, K et al., Scalable Open Science Approach for Mutation Calling of Tumor Exomes Using Multiple Genomic Pipelines. Cell Systems 2018). The model is a set of position-specific weight matrices whose inputs are specific substitutions (e.g., C>A, C>G, T>A) and 11 bp context (e.g., CATGCCCAGTC (SEQ ID NO: 1)) centered on a reference allele. The output of the model is a score that is positive if the mutation is more likely to come from an MSI tumor than an MSS tumor, and negative if the mutation is more likely to come from an MSS tumor than an MSI tumor. Software was written to perform model fitting and scoring.

[0083] This model was fitted to published exome somatic substitution data and MSI / MSS labels were assigned to individual exomes using the published MOSAIC MSI classification markers (Hause R et al. Classification and characterization of microsatellite instability across 18 cancer types. Nat Med 2016). Mutations from the MSI exomes were used to create a set of six position probability matrices (PPMs) (one for each possible substitution of a pyrimidine reference base, i.e., C>A, C>G, C>T, T>A, T>C and T>G). The same was done separately with mutations from the MSS exomes. These two sets of PPMs were combined into one set of six PWMs by dividing, entry-wise, the MSI PPM of a given substitution by the corresponding MSS PPM of that same substitution and then storing the logarithm of each quotient in the various entries of the PWMs. The PPMs and PWMs used are based on the frequency of overlapping triplets, rather than the single letters used by traditional PPM / PWM, to better account for the influence that adjacent bases may have on the mutation signature. Overlapping triplets can be used as an optional feature to further enhance traditional PWM implementations. Example 2

[0084] This example describes MSI signature scoring of individual mutations and tumors.

[0085] Without being limited by theory, microsatellite instability is caused by mismatch repair deficiency, which can give rise to specific somatic mutations in specific contexts (Figure 5). We created a model to estimate the likelihood of substitutions originating from MSI or MSI-H tumors or MSS tumors.

[0086] We created a set of position-specific weight matrices (PWMs) that evaluate the log-likelihood that a particular single-base substitution (SBS) originates from an MSI tumor, given the reference allele, the alternative allele, and the 11-nt context centered around the reference base. This set of matrices was developed from TCGA MC3 data, using the MOSAIC classification to determine whether a given case is MSI / MSI-H or MSS. The matrix set is used to score all SBSs in the tumor, and the scores are summed from those SBSs to obtain a signature score for the tumor itself. This signature score is then combined with a tract-based score (the proportion of mononucleotide tracts that are considered to have somatic deletions) to obtain an overall MSI score, which is used to determine the classification of the tumor. Structure of the position-specific weight matrix

[0087] The PWM was constructed from background and foreground position probability matrices (PPMs). The background matrix was developed from permutations for the MSS case, and the foreground matrix was developed from permutations for the MSI case. The PWM was created by dividing the probabilities in the foreground matrix by the corresponding probabilities in the background matrix and calculating the common (base 10) logarithm of the quotient.

[0088] Each column of the PWM represents a 3-nt substring of the sequence context of a mutation, as opposed to the 1-nt corresponding columns of conventional PWM. We use overlapping 3-nt substrings (trimers) instead of individual bases to better represent the interactions of adjacent nucleotides in the sequence context. A set of position-specific weight matrices

[0089] Requiring the reference base to be a pyrimidine (as is done in the COSMIC signature), there are six possible substitutions: C>A, C>G, C>T, T>A, T>C, and T>G. One PWM is maintained for each of these six substitutions. Additionally, the percentage of substitutions in the MSI case, C>A, C>G, etc., is recorded. The same is done for MSS substitutions. Similar to the calculation of the PWM, the MSI substitution percentage is divided by the MSS substitution percentage and the base 10 logarithm is calculated to obtain a percentage score. For example, 1.26% of MSI substitutions are C>G, while 12.01% of MSS substitutions are C>G. Thus, the percentage score for C>G substitutions is log(0.0126 / 0.1201)=-0.9792, indicating that C>G substitutions are approximately 10 times more likely to come from MSS cases than MSI cases. Scoring Permutations

[0090] A substitution is defined by a reference allele, an alternative allele, and an 11 nt context centered on the reference base. By convention, the reference base is assumed to be the pyrimidine base of the base pair, and the substitution is modified accordingly (e.g., a G>T substitution on the forward strand is treated as a C>A substitution with the context extracted from the reverse strand). The reference allele and the alternative allele determine the ratio score used, and the context is scored using a PWM from the set of PWMs that correspond to the particular substitution. For example, a T>G substitution with a context of CAAAGTGAGGA (SEQ ID NO:2) is scored using a T>G PWM and receives a ratio score of -0.1704, while the context has a score of -0.7253. The signature score of a mutation is the sum of the ratio score and the context score. In the previous example, the signature score is -0.8957, indicating that the mutation is more likely to come from an MSS case (since the score is negative; MSI substitutions are more likely to be positive). Scoring patient samples

[0091] A tumor's signature score is determined by summing the mutation signature scores for all somatic mutations found in the tumor sample. Tumors with a signature score above 0 are more likely to be MSI, and tumors with a signature score below 0 are more likely to be MSS. Example 3

[0092] This example illustrates combining signature scores with tract-based scores for the determination of MSI.

[0093] Determination of microsatellite instability using detection of short alleles at mononucleotide tracts is shown in Figure 4. Exemplary mononucleotide tract loci are shown, including BAT25, BAT26, MONO27, NR21 and NR24. Detection of mismatch repair deficiency using mutational signatures is shown in Figure 5. Exemplary mutational signatures are shown.

[0094] After examining data from 725 tumors, we established classification rules that combine signature score information with tract-based scores in an ensemble approach to detect microsatellite instability. The two component scores are highly correlated, with all tumors with tract-based scores below 10% having signature scores below +10, and all cases with tract-based scores above 25% having signature scores above 0. Neither component score was entirely determinative of what was likely to be the correct classification (in some cases, PCR orthogonal data were available, in others, analysis of mutation data was used). A boundary segment between points (0.1, 20) and (0.25, -10) was established in a plane with tract-based scores on the x-axis and signature scores on the y-axis. The slope (-200) and y-intercept (40) of the line containing this segment indicate that points with 200x+y>=40 are classified as MSI, and points with 200x+y<40 are classified as MSS. There were also two "indeterminate" regions where no data were available: (a) the region with x<=0.1 and y>=20, and (b) the region with x>=0.25 and y<=-10. Tumors with component scores that fall into the indeterminate tract were not assigned a classification. A reduction in the area of ​​the indeterminate region can change the slope, intercept, and / or endpoint of the border segment. Example 4

[0095] This example describes the detection of microsatellite instability using PGDx elio™ tissue complete.

[0096] The PGDx elio™ tissue complete pipeline identifies microsatellite instability based on genomic context signatures of selected mononucleotide tracts and sequence variations. A linear classifier combines the frequency of unstable tracts with the signature of observed variations to determine an overall case status of microsatellite instability-high (MSI-H), microsatellite stable (MSS), or indeterminate.

[0097] To determine unstable tracts, a peak-finding algorithm was used to determine the observed tract lengths for 68 mononucleotide tracts across the tract of interest (ROI). The allele lengths of individual tracts were compared to the reference length to determine unstable tracts that were >2bp shorter than the reference length. Although tract extensions have been reported, unstable mononucleotide tracts are typically short in length, and shifts to longer lengths are rare. No systematic extensions across mononucleotide tracts in the samples were observed. The large number of mononucleotide microsatellite tracts assessed with PGDx elio™ tissue complete ensured that extensions of any single tract, or low frequency of tracts, did not preclude detection of microsatellite instability. Low coverage tracts were excluded from the analysis.

[0098] A set of position-specific weight matrices was used to obtain genomic signature scores. These matrices encode the likelihood of observing a mutation in an MSI-H case versus an MSS case, given the alternative alleles of the mutation and the surrounding 11-bp genomic context. The matrices were developed using MSI-H and MSS context cases from The Cancer Genome Atlas (TCGA). Candidate single-base substitutions were filtered based on germline status, functional outcome, mapping quality, and sequence quality metrics. The remaining sequence variants were assigned a genomic signature score indicating whether the variant is associated with the MSI-H phenotype or not. The likelihood scores from the candidate single-base substitutions were added to obtain an overall genomic signature score for the sample. The genomic signature score was combined with the tract instability frequency to obtain an overall MSI score. This overall MSI score was determined by adding the proportion of unstable tracts multiplied by 200 to the genomic signature score. An overall MSI score >40 was classified as MSI-H, otherwise the sample was classified as MSS. Samples with significantly different MSI status between the tract and signature score are reported as indeterminate.

[0099] In other studies, it was found that the signature scores obtained using exome data appear to be in agreement with the results obtained using PCR for colorectal, gastric, renal and uterine cancer samples (Figure 6). The results obtained by combining the analysis of tracts of nucleotide repeats to determine MSI with the genomic signature of somatic mutations are shown in Figures 7 and 8, which depict samples classified as microsatellite unstable (MSI) or microsatellite stable (MSS). The data further show that the data are linearly separable on both measures alone, i.e. MSI tract frequency and signature score.

[0100] Taken together, these data indicate that an ensemble method combining MSI tract frequency and somatic mutation signature scores for detecting microsatellite instability (MSI) can classify samples as MSI / MSI-H or MSS with greater accuracy than methods relying on a single parameter or score.

[0101] Although the invention has been described with reference to the above examples, it will be understood that modifications and variations are encompassed within the spirit and scope of the invention. Accordingly, the invention is limited only by the scope of the following claims. The present invention provides, for example, the following items. (Item 1) 1. A method for determining microsatellite instability (MSI), comprising: (i) determining the presence of somatic allele lengths in a multi-nucleotide repeat tract of sequenced DNA in a sample obtained from a subject; (ii) determining the presence of somatic mutations in the DNA in regions outside the multiple nucleotide repeat tract; (iii) determining a match between the genomic signature of the somatic mutation found outside the plurality of nucleotide repeat tracts and a genomic signature of a mismatch repair deficiency; (iv) applying rules to the result of said determining step to obtain an MSI score; (v) classifying the sample as microsatellite instability-high (MSI-H) or microsatellite stable (MSS) based on the MSI score; A method comprising: (Item 2) 2. The method of claim 1, further comprising generating a report comprising the MSI score and a classification of the sample as microsatellite instability-high (MSI-H) or microsatellite stable (MSS). (Item 3) 2. The method of claim 1, wherein the sample is a tumor sample. (Item 4) 2. The method of claim 1, wherein the plurality of nucleotide repeat tracts comprises mononucleotide repeats, dinucleotide repeats, trinucleotide repeats, tetranucleotide repeats, pentanucleotide repeats, hexanucleotide repeats, heptanucleotide repeats, octanucleotide repeats, or a combination thereof. (Item 5) 5. The method of claim 4, wherein the multiple nucleotide repeat tracts comprise mononucleotide repeats. (Item 6) 2. The method of claim 1, wherein the nucleotide repeat tract in the MSI-H sample is shorter than the nucleotide repeat tract in the reference sample. (Item 7) 2. The method of claim 1, wherein the nucleotide repeat tract in the MSI-H sample is at least 2 base pairs shorter than the nucleotide repeat tract in the reference sample. (Item 8) 7. The method of claim 6, wherein the reference sample is a matched normal sample or an MSS sample. (Item 9) 7. The method of claim 6, further comprising determining the frequency of shorter tracts of nucleotide repeats. (Item 10) 2. The method of claim 1, wherein the multiple nucleotide repeat tracts comprise one or more microsatellite markers. (Item 11) The one or more microsatellite markers are human genomic microsatellite markers. 11. The method according to item 10, wherein the compound is selected from the group consisting of tomarkers. (Item 12) 2. The method of claim 1, wherein the genomic signature of the somatic mutation comprises 11 base pairs surrounding the somatic mutation. (Item 13) 13. The method of claim 12, further comprising the step of assigning a genomic signature score to the somatic mutation. (Item 14) 14. The method of claim 13, wherein somatic mutations are classified as associated with an MSI-H sample or an MSS sample based on the genomic signature score. (Item 15) 2. The method of claim 1, further comprising the step of selecting or administering a treatment to the patient based on the MSI score. (Item 16) 16. The method of claim 15, wherein the MSI score indicates that the sample from the patient is microsatellite instability-high (MSI-H) and the treatment comprises an immune checkpoint inhibitor. (Item 17) 17. The method of claim 16, wherein the immune checkpoint inhibitor comprises an antibody. (Item 18) 18. The method of claim 17, wherein the antibody is selected from the group consisting of an anti-PD-1 antibody; an anti-IDO antibody; an anti-CTLA-4 antibody; an anti-PD-L1 antibody; and an anti-LAG-3 antibody. (Item 19) 18. The method of item 17, wherein the checkpoint inhibitor is pembrolizumab, nivolumab, atezolizumab, or ipilimumab. (Item 20) 2. The method of claim 1, wherein the sequenced DNA in the sample comprises one or more sequenced genomes, one or more sequenced exomes, or regions of one or more sequenced genomes or one or more sequenced exomes.

Claims

[Claim 1] The invention described in the specification.

Images