Methods and systems for adjusting tumor gene mutation load by tumor percentage and coverage

Abstract

To provide a method and system for adjusting tumor gene mutation levels based on tumor percentage and coverage. [Solution] A method for detecting tumor mutational load (TMB) in a subject is presented herein. In one embodiment, the method includes the steps of determining an observed mutation count from sequence information obtained from nucleic acids in a sample derived from the subject, and determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters. The method also includes the steps of determining a predicted mutation proportion and / or a predicted distribution of the predicted mutation proportion, taking the sequencing parameters into consideration to generate a prediction result, and adjusting the observed mutation count, taking the prediction result into consideration to generate an adjusted result, thereby detecting TMB in the subject. Other embodiments relate to a method for selecting a customized therapy for treating cancer in a subject, and a method for treating cancer in a subject. Yet another embodiment includes related systems and computer-readable media used to detect TMB in a subject.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims the interests of and relies on the filing dates of U.S. Provisional Patent Applications No. 62 / 702,280 filed July 23, 2018, No. 62 / 741,770 filed October 5, 2018, No. 62 / 782,894 filed December 20, 2018, and No. 62 / 824,246 filed March 26, 2019, whose entire disclosures are incorporated herein by reference. [Background technology]

[0002] background A tumor is an abnormal growth of cells. DNA is often released into bodily fluids as extracellular free DNA and / or circulating tumor DNA when, for example, normal cells and / or cancer cells die. Tumors can be benign or malignant. In most cases, malignant tumors are referred to as cancer.

[0003] Cancer is the leading cause of disease worldwide. Tens of millions of people are diagnosed with cancer each year, and more than half ultimately die from it. In many countries, cancer is the second leading cause of death after cardiovascular disease. For many cancers, early detection is associated with improved outcomes.

[0004] Cancer is typically caused by the accumulation of mutations in an individual's normal cells, at least some of which lead to improper control of cell division. Such mutations generally include single nucleotide variants (SNVs), gene fusions, insertions and deletions (indels), base transpositions, translocations, and inversions. The number of mutations within a cancer can be an indicator of its susceptibility to cancer immunotherapy.

[0005] Cancer is often detected by tumor biopsy, subsequent cytopathology, biomarker analysis, or DNA extraction from cells. However, it has recently been proposed that cancer can also be detected from extracellular free nucleic acids in bodily fluids such as blood or urine (e.g., circulating nucleic acids, circulating tumor nucleic acids, exosomes, nucleic acids from apoptotic cells and / or necrotic cells) (see, e.g., Siravegna et al., Nature Reviews, 14: 531-548 (2017)). Such tests have the advantage of being non-invasive and being able to be performed without identifying suspicious cancer cells for biopsy and nucleic acid sampling from the entire tumor. However, such tests are complicated by the fact that the amount of nucleic acid released into bodily fluids is small and volatile, and the recovery rate of nucleic acids in an analyzable form from such fluids is also low and volatile. These sources of variability can make the accuracy of comparisons of tumor mutation burden (TMB) between samples ambiguous. TMB is a measure of mutations present in tumor cells within the tumor genome. TMB is a type of biomarker that can be used to assess whether certain types of cancer therapies, such as immunotherapy (IO), are beneficial for individuals diagnosed with or suspected of having cancer. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Siravegna et al., Nature Reviews, 14: 531-548 (2017) [Overview of the project] [Means for solving the problem]

[0007] Summary of the Invention This application discloses a method, a computer-readable medium, and a system that are useful in determining and analyzing the amount of tumor gene mutations (TMB) in patient samples and that serve as a guide for cancer treatment decisions. Traditionally, TMB obtained by counting the mutation rate is often inaccurate because the sensitivity of the assay for calling mutations decreases when the tumor fraction (e.g., mutant allele fraction (MAF)) and / or coverage is low. Thus, in certain embodiments, the observed TMB is adjusted considering various metrics of assay sensitivity such as tumor fraction (setting the MAF of the mutations in a given sample), coverage, and / or the like. Without such adjustment, samples that are TMB-High but have a low tumor fraction and / or low coverage would generally be misreported as TMB-Low. Such an outcome can have downstream significance for the patient when treatment decisions are made based on such results. Thus, prior to implementation of the adjustment methods and related embodiments disclosed herein, the average mutation count in control samples generally depends on the maximum (max) MAF and coverage. After implementation of these methods and related embodiments, the average mutation count between control or comparison samples is essentially independent of either the maximum (max) MAF or coverage.

[0008] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in the art from the following detailed description, which shows merely illustrative embodiments of the present disclosure. As will be understood, the present disclosure is capable of other and different embodiments and some of the details thereof can be modified in various obvious respects without departing from the present disclosure. Accordingly, the figures and description are to be regarded as illustrative in nature and not as restrictive.

[0009] In one aspect, the present disclosure provides a method for determining the tumor mutational burden (TMB) in a subject, comprising: (a) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject; (b) determining the tumor fraction and / or coverage of the nucleic acids to generate sequencing parameters; (c) determining a predicted mutation fraction and / or a predicted distribution of the predicted mutation fraction in consideration of the sequencing parameters to generate a prediction result; and (d) adjusting the observed mutation count in consideration of the prediction result to generate an adjusted result, thereby determining the TMB in the subject. In some embodiments of the methods and related implementations disclosed herein, the observed mutation counts are adjusted based on the likelihood of neoantigens associated with these counts. Depending on the haplotype of a given subject, for example, certain mutations or clusters of mutations may be more neoantigenic in that particular subject than in other subjects. In some embodiments, the methods disclosed herein include, for example, determining the TMB in a given subject at multiple time points to evaluate or monitor, over time, the course of treatment in the subject.

[0010] In another aspect, the present disclosure provides a method for determining the tumor mutational burden (TMB) in a subject, comprising: (a) providing a sample derived from the subject; (b) amplifying the nucleic acids in the sample to generate amplified nucleic acids; (c) sequencing the amplified nucleic acids to generate sequence information; (d) determining an observed mutation count from the sequence information; (e) determining the tumor fraction and / or coverage of the nucleic acids to generate sequencing parameters; (f) determining a predicted mutation fraction and / or a predicted distribution of the predicted mutation fraction in consideration of the sequencing parameters to generate a prediction result; and (g) adjusting the observed mutation count in consideration of the prediction result to generate an adjusted result, thereby determining the TMB in the subject.

[0011] In another aspect, the Disclosure provides a method for selecting one or more customized therapies to treat cancer in a subject, comprising: (a) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject; (b) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; (c) determining the predicted mutation proportion and / or predicted distribution of the predicted mutation proportion taking the sequencing parameters into consideration to generate a prediction result; (d) adjusting the observed mutation count taking the prediction result into consideration to generate an adjusted result; and (e) comparing the adjusted result with one or more comparison results pointing to one or more therapies to identify one or more customized therapies for the subject.

[0012] In another aspect, the Disclosure provides a method for treating cancer in a subject, comprising: (a) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject; (b) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; (c) determining a predicted mutation rate and / or a predicted distribution of the predicted mutation rate, taking the sequencing parameters into consideration, to generate a prediction result; (d) adjusting the observed mutation count, taking the prediction result into consideration, to generate an adjusted result; (e) comparing the adjusted result with one or more comparison results indicating one or more therapies to identify one or more customized therapies for the subject; and (f) administering at least one of the identified customized therapies to the subject, if the adjusted result and the comparison results substantially match, thereby treating the cancer in the subject.

[0013] In another aspect, the Disclosure provides a method for treating cancer in a subject, comprising administering one or more customized therapies to the subject to thereby treat the cancer in the subject, wherein the customized therapies are identified by (a) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject; (b) determining the tumor proportion and / or coverage of the nucleic acids to generate sequencing parameters; (c) determining a predicted mutation rate and / or a predicted distribution of the predicted mutation rate taking the sequencing parameters into consideration to generate a predicted result; (d) adjusting the observed mutation count taking the predicted result into consideration to generate an adjusted result; (e) comparing the adjusted result with one or more comparison results indicating one or more therapies; and (f) identifying one or more customized therapies for the subject if the adjusted result and the comparison results substantially match.

[0014] In some embodiments, the observed variant count and / or tumor percentage includes multiple synonymous variants, multiple non-synonymous variants, and / or multiple non-coding variants identified in the nucleic acid. In some embodiments, the observed variant count and / or tumor percentage includes multiple variants selected from the group consisting of single nucleotide variants (SNVs), insertions or deletions (indels), copy number variants (CNVs), fusions, transpositions, translocations, frameshifts, duplications, repeat extensions, and epigenetic variants. In some embodiments, driver variants and / or non-tumor-related variants (e.g., clonal hematopoietic variants - CH variants) are excluded from the observed variant count and / or tumor percentage.

[0015] In some embodiments, the method includes using pooled evidence of one or more possible mutations below the detection limit with respect to a given single nucleotide variant (SNV) or a given insertion or deletion (indel) to determine the observed mutation count.

[0016] In some embodiments, the method includes generating a predicted mutation rate, which is the observed proportion of the actual mutation count. In some embodiments, the observed mutation count and / or tumor rate includes a plurality of somatic mutations identified in nucleic acids. In some embodiments, one or more known cancer driver and / or passenger mutations are excluded from the observed mutation count.

[0017] In some embodiments, the method includes comparing sequence information with one or more reference sequences to identify observed mutation counts.

[0018] In some embodiments, the reference sequence includes at least a subsequence of hg19 and / or hg38.

[0019] In some embodiments, the tumor percentage includes the maximum mutant allele percentage (MAF) of all somatic mutations identified in the nucleic acid. In some embodiments, the tumor percentage is less than about 0.05%, less than about 0.1%, less than about 0.2%, less than about 0.5%, less than about 1%, less than about 2%, less than about 3%, less than 4%, or less than 5% of all nucleic acids in the sample.

[0020] In some embodiments, the method includes identifying a number of unique cfDNA fragments containing a given nucleotide position in a nucleic acid to determine coverage. In some embodiments, the method includes identifying the median number of unique extracellular free DNA (cfDNA) molecules constituting a given nucleotide position in a nucleic acid to determine coverage.

[0021] In some embodiments, coverage is a cfDNA fragment between 10 and 50,000 at a given nucleotide position in the nucleic acid present in the sample.

[0022] In some embodiments, the predicted mutation rate and / or the predicted distribution of the predicted mutation rate include a confidence interval of about 95% or greater with respect to the mutation rate. In some embodiments, the method includes using the upper limit of the 95% confidence interval for the predicted mutation rate to generate a lower limit for the observed mutation count.

[0023] In some embodiments, the method includes determining the predicted mutation ratio by calculating the probability that a mutation in a given mutational allele ratio (MAF) is identified across the predicted MAF distribution. In some embodiments, the method includes generating the MAF by multiplying the predicted relative MAF distribution by the tumor ratio. As used herein, the term "MAF" is not limited to the ratio and may, in certain embodiments, also include the mutant molecule count. In some embodiments, the predicted MAF distribution is calculated using the following binomial ratio confidence interval:

[0024]

number

[0025]

number

[0026] In some embodiments, the method is based on the following equation:

[0027] Percentage of observed mutations = Σ MAF (P(Variation Call | MAF) × P(Variation in MAF))

[0028] (In the formula, P is the probability, and MAF is the proportion of mutated alleles.) This includes determining the prediction results using [this method].

[0029] In some embodiments, the predictive distribution of relative MAF is obtained from one or more control sample datasets. In some embodiments, the control sample dataset includes at least about 25, at least about 50, at least about 100, at least about 200, at least about 300, at least about 400, at least about 500, at least about 600, at least about 700, at least about 800, at least about 900, at least about 1,000, at least about 5,000, at least about 10,000, at least about 15,000, at least about 20,000, at least about 25,000, at least about 30,000, or more control samples. In some embodiments, the control samples used in the control sample dataset may be cancer type specific and / or treatment specific.

[0030] In some embodiments, the maximum MAF observed in the control sample dataset constitutes approximately 0.5%, 1%, 2%, 5%, or 10%.

[0031] In some embodiments, the method calculates the relative MAF using the following equation:

[0032] F = 1 / (1 + (P_50 / relative - MAF) n )

[0033] (In the formula, F is the cumulative distribution function, P_50 is the median of the relative MAF, relative -MAF is the relative MAF, and n is the exponent for fitting the shape of the relative distribution.) This includes fitting the curve using [a specific method / tool].

[0034] In some embodiments, the method includes generating an adjusted result by dividing the observed mutation count in the sample by the predicted mutation fraction or the upper / lower limit of the confidence interval for the predicted mutation fraction. In some embodiments, the adjusted result includes multiple mutations detected in the nucleic acid across a range of mutant allele fractions. In some embodiments, the method includes generating an adjusted result by dividing the observed mutation count by the predicted result. In some embodiments, the adjusted result includes an estimate of the most likely actual mutation count. In some embodiments, the adjusted result includes an estimate of the least likely actual mutation count. In some embodiments, the adjusted result includes an adjusted mutation count. In some embodiments, the adjusted mutation count is greater than or equal to the observed mutation count. In some embodiments, the TMB score is determined by dividing the adjusted mutation count / adjusted result by the product of the size of the target genomic region being analyzed and an exome calibration factor. In certain embodiments, the exome calibrator is a value of at least 1.01, 1.02, 1.03, 1.04, 1.05, 1.06, 1.07, 1.08, 1.09, 1.10, or greater, while in other embodiments, the exome calibrator includes a value less than 1.0 (e.g., about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, or 0.9). In some embodiments, the exome calibrator is determined from the exome mutation rate of samples in a cancer database (e.g., TCGA). In certain embodiments, an exome calibrator and exome mutation rate specific to a given cancer type under consideration are determined from samples having that cancer type.

[0035] In some embodiments, the method further includes classifying the sample as a TMB-High sample. In some embodiments, the TMB score of the sample is a threshold TMBスコア If the TMB score of the sample is greater than the threshold, the sample is classified as a TMB-High sample. TMBスコア If it is smaller than the threshold, the sample is classified as a TMB-Low sample. In certain embodiments, for example, the threshold TMBスコアThis could be approximately 5, 10, 15, 20, 25, 30, 35, 40, or another selected threshold.

[0036] In some embodiments, the method includes obtaining sequence information from nucleic acids in a sample derived from a target. In some embodiments, the sequence information is obtained from a targeted segment of the nucleic acid. In some embodiments, the targeted segment includes different and / or overlapping genomic regions between about 1 and about 100,000.

[0037] In some embodiments, the method includes obtaining a sample derived from a subject. In some embodiments, the sample is selected from the group consisting of tissue, blood, plasma, serum, sputum, urine, semen, vaginal fluid, feces, synovial fluid, cerebrospinal fluid, and saliva. In some embodiments, the sample includes tissue. In some embodiments, the sample includes blood, plasma, and / or serum. In some embodiments, the subject is a mammalian subject. In some embodiments, the mammalian subject is a human subject. In some embodiments, the nucleic acid includes extracellular free nucleic acid. In some embodiments, the nucleic acid includes intracellular nucleic acid. In some embodiments, the nucleic acid includes circulating tumor nucleic acid. In some embodiments, the nucleic acid is obtained from circulating tumor cells. In some embodiments, the nucleic acid includes deoxyribonucleic acid (DNA) and / or ribonucleic acid (RNA).

[0038] In some embodiments, the method includes amplifying at least one segment of nucleic acid in a sample to produce at least one amplified nucleic acid. In some embodiments, the method includes sequencing the amplified nucleic acid to produce sequence information. In some embodiments, the method includes sequencing at least about 50,000, about 100,000, about 150,000, about 200,000, about 250,000, about 500,000, about 750,000, about 1,000,000, about 1,500,000, about 2,000,000 nucleotides or more of the nucleic acid to produce sequence information. In some embodiments, the sequencing is selected from the group consisting of targeted sequencing, intron sequencing, exome sequencing, and whole-genome sequencing.

[0039] In some embodiments, the customized therapy includes at least one immunotherapy. In some embodiments, the immunotherapy includes at least one checkpoint inhibitor antibody. In some embodiments, the immunotherapy includes antibodies against PD-1, PD-2, PD-L1, PD-L2, CTLA-40, OX40, B7.1, B7He, LAG3, CD137, KIR, CCR5, CD27, or CD40. In some embodiments, the immunotherapy includes the administration of pro-inflammatory cytokines to at least one tumor type. In some embodiments, the immunotherapy includes the administration of T cells to at least one tumor type.

[0040] In some embodiments, cancers include biliary tract cancer, bladder cancer, transitional cell carcinoma, urothelial carcinoma, brain cancer, glioma, astrocytoma, breast cancer, metaplastic carcinoma, cervical cancer, cervical squamous cell carcinoma, rectal cancer, colorectal cancer, colon cancer, hereditary nonpolyposis colorectal cancer, colorectal adenocarcinoma, gastrointestinal stromal tumor (GIST), endometrial cancer, endometrial stromal sarcoma, esophageal cancer, esophageal squamous cell carcinoma, esophageal adenocarcinoma, intraocular melanoma, uveal melanoma, gallbladder cancer, gallbladder adenocarcinoma, renal cell carcinoma, renal clear cell carcinoma, transitional cell carcinoma, urothelial carcinoma, Wilms' tumor, leukemia, acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), chronic myelomonocytic leukemia (CMML), liver cancer, hepatoma, It includes at least one tumor type selected from the group consisting of patoma, hepatocellular carcinoma, cholangiocarcinoma, hepatoblastoma, lung cancer, non-small cell lung cancer (NSCLC), mesothelioma, B-cell lymphoma, non-Hodgkin lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, T-cell lymphoma, non-Hodgkin lymphoma, progenitor T-lymphoblastic lymphoma / leukemia, peripheral T-cell lymphoma, multiple myeloma, nasopharyngeal carcinoma (NPC), neuroblastoma, oropharyngeal cancer, oral squamous cell carcinoma, osteosarcoma, ovarian cancer, pancreatic cancer, pancreatic ductal adenocarcinoma, pseudopapillary neoplasm, acinar cell carcinoma, prostate cancer, prostate adenocarcinoma, skin cancer, melanoma, malignant melanoma, cutaneous melanoma, small intestine cancer, gastric cancer, gastric cancer, gastrointestinal stromal tumor (GIST), uterine cancer, and uterine sarcoma.

[0041] In some embodiments, the sequence information includes a sequence reading of a nucleic acid generated by a nucleic acid sequencer. In some embodiments, the nucleic acid sequencer generates a sequence reading by performing pyrosequencing, single-molecule sequencing, nanopore sequencing, semiconductor sequencing, synthesis sequencing, ligation sequencing, or hybridization sequencing on the nucleic acid.

[0042] In some embodiments, the method further includes the step of selectively enriching one or more regions from the genome or transcriptome before sequencing. In some embodiments, the method further includes the step of amplifying the selectively enriched regions before sequencing. In some embodiments, sequence information is obtained from a targeted segment of nucleic acid, which is obtained by selectively enriching one or more regions from the genome or transcriptome before sequencing. In some embodiments, the method further includes the step of amplifying the obtained targeted segment before sequencing.

[0043] In some embodiments, the method further includes the step of attaching one or more adapters containing molecular barcodes to the nucleic acid before sequencing. In some embodiments, the nucleic acid is uniquely barcoded. In some embodiments, the nucleic acid is non-uniquely barcoded. In some embodiments, the adapter contains molecular barcodes between 2 and 1,000,000. In some embodiments, the adapter contains molecular barcodes between 2 and 100. In some embodiments, the adapter contains molecular barcodes between 2 and 200. In some embodiments, the adapter contains molecular barcodes between 2 and 100. In some embodiments, the method includes randomly attaching adapters containing molecular barcodes to each end of the nucleic acid. In some embodiments, the adapters are attached to the nucleic acid by blunt-end ligation or sticky-end ligation. In some embodiments, the adapters are adapters having a T-tail and / or a C-tail.

[0044] In some embodiments, the method further includes the step of grouping sequence reads into families of sequence reads, each family comprising sequence reads generated from nucleic acids in a sample.

[0045] In some embodiments, at least part of the method is performed by a computer. In some embodiments, the method further includes the step of creating an electronic report presenting one or more TMB scores.

[0046] In another embodiment, the Disclosure provides a system including a computer-readable medium containing non-temporary computer-executable instructions, or a controller capable of accessing such medium, which, when executed by at least one electronic processor, performs at least the steps of: (i) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from a subject; (ii) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; (iii) determining a predicted mutation proportion and / or predicted distribution of the predicted mutation proportion taking the sequencing parameters into consideration to generate a prediction result; and (iv) adjusting the observed mutation count taking the prediction result into consideration to generate an adjusted result, thereby determining the tumor mutational load (TMB) in the subject.

[0047] In some embodiments, the system includes a nucleic acid sequencer operably connected to a controller, which is configured to derive sequence information from nucleic acids in a sample of interest. In some embodiments, the nucleic acid sequencer is configured to generate sequenced reads by performing pyrosequencing, single-molecule sequencing, nanopore sequencing, semiconductor sequencing, synthesis sequencing, ligation sequencing, or hybridization sequencing on the nucleic acids. In some embodiments, the nucleic acid sequencer or another system component is configured to group the sequence reads generated by the nucleic acid sequencer into families of sequence reads, each family containing sequence reads generated from nucleic acids in the sample.

[0048] In some embodiments, the system includes a database operably connected to a controller, the database includes one or more comparison results pointing to one or more therapies, and the electronic processor further performs at least (v) the step of comparing a modified result with one or more comparison results, wherein the modified result and the comparison results substantially match, indicating an expected response to the therapy for the subject.

[0049] In some embodiments, the system includes a sample preparation component operably connected to a controller, which is configured to prepare nucleic acids in a sample to be sequenced by a nucleic acid sequencer. In some embodiments, the sample preparation component is configured to selectively enrich regions from the genome or transcriptome in the sample. In some embodiments, the sample preparation component is configured to attach one or more adapters containing molecular barcodes to the nucleic acids.

[0050] In some embodiments, the system includes a nucleic acid amplification component operably connected to a controller, which is configured to amplify nucleic acids in a sample derived from the target. In some embodiments, the nucleic acid amplification component is configured to amplify a region selectively enriched from the genome or transcriptome in the sample.

[0051] In some embodiments, the system includes a material transfer component operably connected to a controller, which is configured to move one or more materials between at least a nucleic acid sequencer and a sample preparation component.

[0052] In another embodiment, the disclosure provides a computer-readable medium containing non-temporary, computer-executable instructions that, when executed by at least one electronic processor, perform at least the steps of: (i) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from a subject; (ii) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; (iii) determining the predicted mutation proportion and / or predicted distribution of the predicted mutation proportion considering the sequencing parameters to generate a prediction result; and (iv) adjusting the observed mutation count considering the prediction result to generate an adjusted result, thereby determining the tumor mutational load (TMB) in the subject. In some embodiments, the adjusted mutation count is determined using the predicted mutation proportion.

[0053] The systems and methods performed using computer-readable media disclosed herein include several different embodiments. In some embodiments, for example, the observed variant count and / or tumor percentage includes multiple synonymous variants, multiple non-synonymous variants, and / or multiple non-coding variants identified in the nucleic acid. In certain embodiments, the observed variant count and / or tumor percentage includes multiple variants selected from the group consisting of single nucleotide variants (SNVs), insertions or deletions (indels), copy number variants (CNVs), fusions, transpositions, translocations, frameshifts, duplications, repeat extensions, and epigenetic variants. In other exemplary embodiments, driver variants and / or non-tumor-related variants (e.g., clonal hematopoietic variants) are excluded from the observed variant count and / or tumor percentage. If necessary, the observed variant count is determined using pooled evidence of one or more possible variants below the detection limit with respect to a given single nucleotide variant (SNV) or a given insertion or deletion (indel).

[0054] In some embodiments, a predicted mutation rate is used to determine an adjusted mutation count. Generally, the observed mutation count and / or tumor rate include multiple somatic mutations identified in the nucleic acid. In some of these embodiments, one or more known cancer driver and / or passenger mutations are excluded from the observed mutation count. In certain embodiments, the observed mutation count is determined by comparing sequence information with one or more reference sequences (e.g., hg19, hg38, and / or at least a subsequence of similar sequences).

[0055] In certain embodiments, the tumor percentage includes the maximum mutant allele percentage (MAF) of all somatic mutations identified in the nucleic acid. Generally, the tumor percentage is less than about 0.05%, less than 0.1%, less than 0.2%, less than 0.5%, less than 1%, less than 2%, less than 3%, less than 4%, or less than 5% of all nucleic acids in the sample. In some embodiments, coverage is determined by identifying the median number of unique extracellular free DNA (cfDNA) molecules constituting a given nucleotide position in the nucleic acid. In some embodiments, for example, coverage is between 10 and 50,000 cfDNA fragments at a given nucleotide position in the nucleic acid present in the sample.

[0056] In some embodiments, the predicted mutation rate and / or the predicted distribution of the predicted mutation rate include a confidence interval of approximately 95% or greater with respect to the mutation rate. In certain embodiments, the upper limit of the 95% confidence interval for the predicted mutation rate is used to generate a lower limit for the observed mutation count. In some embodiments, the predicted mutation rate is determined by calculating the probability that a mutation in a given mutation allele rate (MAF) is identified across the distribution of predicted MAFs. If necessary, the MAF is generated by multiplying the distribution of relative MAFs by the tumor rate. In some embodiments, the distribution of predicted MAFs is calculated using the confidence intervals of the following binomial proportions:

number

[0057] In certain embodiments, the predictive distribution of MAF is obtained from the relative MAF observed in at least one control sample dataset. In some embodiments, the comparison results include at least about 25, at least about 50, at least about 100, at least about 200, at least about 300, at least about 400, at least about 500, at least about 600, at least about 700, at least about 800, at least about 900, at least about 1,000, at least about 5,000, at least about 10,000, at least about 15,000, at least about 20,000, at least about 25,000, at least about 30,000, or more control samples. In certain embodiments, the control samples in the control sample dataset constitute about 0.5%, about 1%, about 2%, about 5% or about 10% of the maximum MAF. Some implementations of the systems or computer-readable media disclosed herein transform the relative MAF using the following equation: F = 1 / (1 + (P_50 / relative - MAF) n ) This involves fitting a curve using the formula (wherein F is the cumulative distribution function, P_50 is the median of the relative MAF, relative-MAF is the relative MAF, and n is an index for fitting the shape of the relative distribution). In some embodiments, adjusted results are generated by dividing the observed variant count by the predicted variant proportion in the sample or the upper / lower limit of the confidence interval for the predicted variant proportion. In certain embodiments, the adjusted results include multiple variants detected in the nucleic acid across a range of variant allele proportions. In some embodiments, adjusted results are generated by dividing the observed variant count by the predicted result. In certain embodiments, the adjusted result includes an estimate of the most likely actual variant count. In some embodiments, the adjusted result includes an estimate of the least likely actual variant count. In certain embodiments, the adjusted result includes an adjusted variant count. In some of these embodiments, the adjusted variant count is greater than or equal to the observed variant count. In some embodiments, the TMB score is determined by dividing the adjusted variant count / adjusted result by the product of the size of the target genomic region being analyzed and the exome calibrator. In certain embodiments, the exome calibrator is a value of at least 1.01, 1.02, 1.03, 1.04, 1.05, 1.06, 1.07, 1.08, 1.09, 1.10, or greater, while in other embodiments, the exome calibrator includes a value less than 1.0 (e.g., about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, or 0.9). In some embodiments, the exome calibrator is determined from the exome mutation rate of samples in a cancer database (e.g., TCGA). In certain embodiments, an exome calibrator and exome mutation rate specific to a given cancer type under consideration are determined from samples having that cancer type. In some embodiments, the TMB score of a given sample is a threshold. TMBスコア If the TMB score is greater than the threshold, the sample is generally classified as a TMB-High sample. In contrast, if the TMB score of the sample is greater than the threshold, TMBスコア If the threshold is smaller than the threshold, the sample is generally classified as a TMB-Low sample. In some embodiments, TMBスコアThis could be, for example, approximately 5, 10, 15, 20, 25, 30, 35, 40, or another selected threshold.

[0058] In another aspect, the Disclosure provides a system including a communication interface for obtaining sequencing information from one or more nucleic acids in a sample derived from a subject via a communication network, and a computer communicating with the communication interface, comprising at least one computer processor, and a computer-readable medium containing machine-executable code, which, when executed by the at least one computer processor, performs a method comprising: (i) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from a subject; (ii) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; (iii) determining a predicted mutation proportion and / or predicted distribution of the predicted mutation proportion considering the sequencing parameters to generate a prediction result; and (iv) adjusting the observed mutation count considering the prediction result to generate an adjusted result, thereby detecting the tumor mutational load (TMB) in the subject.

[0059] In some embodiments, sequencing information is provided by a nucleic acid sequencer. In some embodiments, the nucleic acid sequencer generates sequencing reads by performing pyrosequencing, single-molecule sequencing, nanopore sequencing, semiconductor sequencing, synthesis sequencing, ligation sequencing, or hybridization sequencing on nucleic acids. In some embodiments, the nucleic acid sequencer generates sequencing reads using a clone single-molecule array derived from a sequencing library. In some embodiments, the nucleic acid sequencer includes a chip having an array of microwells for sequencing a sequencing library to generate sequencing reads.

[0060] In some embodiments, the computer-readable medium includes a storage device, a hard drive, or a computer server. In some embodiments, the communication network includes one or more computer servers capable of performing distributed computing. In some embodiments, the distributed computing is cloud computing. In some embodiments, the computer is located on a computer server located remotely from the nucleic acid sequencer.

[0061] In some embodiments, the computer-readable medium further includes an electronic display that communicates with the computer over a network, and the electronic display includes a user interface for displaying the results when (i) to (iv) are performed. In some embodiments, the user interface is a graphical user interface (GUI) or a web-based user interface. In some embodiments, the electronic display is in a personal computer. In some embodiments, the electronic display is in an Internet-connected computer. In some embodiments, the Internet-connected computer is located remotely from the computer. In some embodiments, the computer-readable medium includes a storage device, a hard drive, or a computer server. In some embodiments, the communication network includes a telecommunications network, the Internet, an extranet, or an intranet.

[0062] In some embodiments, somatic mutations with a max MAF of less than approximately 1%, less than 2%, less than 3%, less than 4%, less than 5%, less than 6%, less than 7%, less than 8%, less than 9%, less than 10%, less than 15%, less than 20%, less than 25%, or less than 30% are excluded from the observed mutation count and / or tumor percentage. In some embodiments, the predicted / adjusted mutation count is adjusted using an exome calibrator. In some embodiments, the adjusted mutation count is divided by the exome calibrator.

[0063] In some embodiments, the mutation count is determined within a set of genes or genomic regions containing the genes listed in Table 2.

[0064] In another embodiment, the Disclosure provides a method for characterizing a sample of extracellular free nucleic acid molecules from a subject having or suspected of having cancer, comprising the step of performing an assay on the sample to determine whether genetic variations exist in at least 100, 200, 300, 400, or 500 genes or genomic regions selected from those listed in Table 2. In some embodiments, the method is performed on 1,000 or fewer genes. In some embodiments, the method further comprises the step of performing a cancer treatment on a subject in which the presence of genetic variations has been determined in at least one of the assayed genes from Table 2. In some embodiments, the method further comprises the step of isolating nucleic acid molecules from the sample and enriching nucleic acid molecules corresponding to at least 100, 200, 300, 400, or 500 genes with probes containing segments derived from at least 100, 200, 300, 400, or 500 genes.

[0065] In yet another aspect, the Disclosure provides a method for analyzing a sample of extracellular free DNA from a subject having or suspected of having cancer, comprising the steps of: selectively enriching at least 100, 200, 300, 400, or 500 genomic regions from the group of genes listed in Table 2 to generate an enriched library; amplifying the enriched library and performing a sequencing reaction; and analyzing the presence of gene variants in the genomic regions.

[0066] The methods and related systems and computer-readable media implementations disclosed herein include a variety of embodiments. These methods and related embodiments generally involve generating a TMB score for a sample. Certain applications utilize subclonal filters. Some of these embodiments exclude somatic mutations from the observed mutation count and / or tumor percentage by filtering out low-MAF somatic mutations. Certain embodiments use, for example, subclonal filters to exclude somatic mutations whose max MAF is less than about 1%, less than 2%, less than 3%, less than 4%, less than 5%, less than 6%, less than 7%, less than 8%, less than 9%, less than 10%, less than 15%, less than 20%, less than 25%, or less than 30%. Some embodiments weight somatic mutations based on clonality (i.e., give some weight to each mutation, not just low-MAF mutations) instead of excluding / filtering out somatic mutations. In other exemplary embodiments, the predicted and / or adjusted mutation counts are adjusted using an exome calibrator. In some of these embodiments, the adjusted mutation count is divided by the exome calibrator. In certain embodiments, the exome calibrator is a value of at least 1.01, 1.02, 1.03, 1.04, 1.05, 1.06, 1.07, 1.08, 1.09, 1.10, or greater, while in other embodiments, the exome calibrator includes a value less than 1.0 (e.g., about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, or 0.9). In some embodiments, the exome calibrator is determined from the exome mutation rate of samples in a cancer database (e.g., TCGA). In certain embodiments, an exome calibrator and exome mutation rate specific to a given cancer type under consideration are determined from samples having that cancer type.

[0067] In some embodiments, tumor proportions are estimated using alternative methods other than using max MAF. In some embodiments, tumor proportions are estimated using, for example, coverage of a predetermined set of genomic regions and standard maximum likelihood estimation. In certain embodiments, tumor proportions are determined using a predetermined set of genomic regions in the sample and differences in the cfDNA fragment size distribution of regions with copy number variations. If necessary, tumor proportions are estimated by adjusting the MAFs of germline and / or somatic variants for copy number variations observed in the sample. Generally, germline variants are excluded from the determined TMB score. In some embodiments, germline / somatic status is determined using a beta-binomial distribution model that estimates the mean and variance of MAFs of common germline SNPs located near the candidate variant under consideration. Additional details relating to the beta-binomial distribution model, which may be adapted as necessary for use in performing the methods and related embodiments disclosed herein, are also described, for example, in PCT / US2018 / 052087, filed September 20, 2018, incorporated by reference. Furthermore, coverage is also determined using various methods. In some embodiments, coverage is incorporated as the median number of unique cfDNA fragments per base in a given sample. In certain embodiments of these, for example, the model can be made base-specific, and this method involves calculating the sensitivity at each base using the number of unique cfDNA fragments at that base. Generally, coverage may be at least 10, 50, 100, 500, 1000, 5000, 10,000, 20,000, 50,000, or more cfDNA fragments at a given base position.

[0068] In some embodiments, the somatic mutation count (observed mutation count) excludes (i) driver mutations that do not represent the background exome mutation rate or TMB, (ii) mutations that are more likely to originate from clonal hematopoiesis rather than the tumor under consideration, and / or (iii) resistance mutations. In some embodiments, certain types of somatic mutations are weighted instead of being excluded or filtered out from the observed mutation count. In certain embodiments, clonal hematopoietic mutations are identified by (i) using a curated list of mutations frequently observed in hematological cancers based on literature and cancer databases (e.g., COSMIC), (ii) using their context in the sample (e.g., MAF) (e.g., the presence of similar MAFs or other clonal hematopoietic variants within a range of similar MAFs), and / or by analyzing clonal hematopoietic mutations in a database of previously tested samples, and / or by sequencing DNA in patient samples derived from blood (e.g., white blood cells). Clonal hematopoiesis (e.g., "clonal hematopoiesis of indeterminate potential" or "CHIP") mutations can also be identified based on other factors such as the patient's age, methylation status, mutation enrichment in a particular population, methods for estimating tumor proportions as disclosed herein, and / or similar. In certain embodiments, resistance mutations can be identified, for example, by using a curated list of mutations frequently observed in patient samples, based on literature and cancer databases, and / or by analyzing a database of previously tested samples. Resistance mechanisms or any other processes may introduce mutations in multiple genes or clusters of gene mutations (e.g., BRCA1 / 2 reverse mutations in prostate cancer treated with KRAS or PARP inhibitors in colorectal cancer).In some embodiments, for example, if the number of mutations in a particular gene is significantly higher than the predicted number of mutations in that gene based on the overall mutation count and gene size in a given panel, at least a small number of mutations in that particular gene are excluded from the mutation count. In certain embodiments, the somatic mutation count also includes SNVs and indels at sites not reported as part of the particular panel being used.

[0069] The TMB score is generally used to predict whether a patient will respond to immunotherapy. For example, the TMB score can be used in conjunction with the presence of specific types of mutations in certain genes across a set of analyzed genes (e.g., loss of function in STK11, KEAP1, PTEN, etc.) and / or the patient's mutational signature (e.g., the patient has a C>T translocation, etc.) to predict the likelihood of a response to therapy. Generally, loss of function / driver mutations in genes take precedence, and the patient will not respond to therapy. That is, if loss of function in certain genes such as STK11, KEAP1, and PTEN is observed, the patient is likely not to respond to immunotherapy, regardless of the TMB score, but is not limited to these.

[0070] In some embodiments, mutations in specific genes (e.g., DNA repair system genes - MLH1, MLH2 & MLH3, MSH2 & MSH3, MSH6, PMS1 & PMS2; polymerase E) and certain factors such as the methylation status of certain regions, particularly the promoter regions of DNA repair system genes, can be combined with the TMB score to classify samples as TMB-High or TMB-Low.

[0071] Essentially, any therapy or combination of therapies is administered to a given patient as needed, depending on specific circumstances such as TMB score, cancer type and / or staging classification. Examples of such therapies include immunotherapies, e.g., CAR-T cell therapy, vaccines (e.g., general or patient-specific novel antigens), genetic therapies based on oligonucleotides or vectors, abscopal effect interventions (e.g., irradiation therapy), immunotherapeutic treatments / drugs (e.g., anti-TIGIT), immune checkpoint inhibitors and / or antibodies (e.g., anti-TIGIT; antibodies against PD-1, PD-2, PD-L1, PD-L2, CTLA-40, OX40, B7.1, B7He, LAG3, CD137, KIR, CCR5, CD27, or CD40), and / or combination therapies (e.g., immunotherapy + PARPi + chemotherapy, etc.), among many other therapies further illustrated herein or otherwise known to those skilled in the art.

[0072] The TMB determination described herein may be combined with other assessments or techniques as needed to make further decisions regarding treatment. Some examples of these include assessing mechanistic defects in cells (e.g., lack of response to a given therapeutic agent), assessing increased aneuploidy (e.g., to measure a potential decrease in response to IO therapy), determining other non-TMB characteristics of the patient's condition (e.g., age, haplotype, ethnicity, sex, etc.), combining TMB score determination with human leukocyte antigen (HLA) loss, HLA sequencing (e.g., as a mechanism for predicting new antigens), transcriptome, immunorepertory sequencing, and / or other analytical techniques to predict a lack of therapeutic response or the possibility of such a response.

[0073] In some embodiments, the results of the systems and methods disclosed herein are used as input for preparing a report. The report may be in paper or electronic format. For example, the adjusted results obtained by the methods and systems disclosed herein may be directly displayed in such a report. Alternatively, or in addition to, diagnostic information or therapeutic recommendations based on the adjusted results may be included in the report.

[0074] Various steps of the methods disclosed herein, or steps performed by the systems disclosed herein, may be performed at the same or different times, in the same or different geographical locations, for example, in different countries, and / or by the same or different people.

[0075] In another aspect, the Disclosure provides a method for classifying a subject as a candidate for immunotherapy, comprising: a) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject; b) determining the tumor proportion and / or coverage of the nucleic acid to generate sequencing parameters; c) determining the predicted mutation proportion and / or predicted distribution of the predicted mutation proportion considering the sequencing parameters to generate prediction results; d) adjusting the observed mutation count considering the prediction results to generate adjusted results; e) determining a TMB score using the adjusted results; and f) setting the TMB score to a threshold TMBスコア In comparison, this provides a method that includes the step of classifying the subject into a candidate for immunotherapy.

[0076] In another aspect, the disclosure relates to a method for generating novel antigen-orphan immune receptor information using at least part computer (i.e., some or all of the steps are performed using a computer). The method includes (a) receiving by computer sequence information obtained from nucleic acids in a blood sample obtained from a subject diagnosed with cancer, wherein at least a first portion of the sequence information includes sequencing reads obtained from extracellular free nucleic acids (cfNA) in the blood sample, and at least a second portion of the sequence information includes sequencing reads obtained from nucleic acids derived from one or more immune cells in the blood sample. The method also includes (b) determining a tumor mutational burden (TMB) score for the subject from at least the first portion of the sequence information, and (c) identifying one or more clonal types of the immune repertoire in at least the second portion of the sequence information. Furthermore, the method also includes (d) correlating the TMB score with one or more clonal types to identify one or more novel antigen-orphan immune receptors in the subject, thereby generating novel antigen-orphan immune receptor information.

[0077] In some embodiments, the method further includes the step of identifying one or more customized therapies for a subject using novel antigen orphan immune receptor information. In certain embodiments of these embodiments, the method further includes the step of administering one or more customized therapies to a subject. In some embodiments, the method includes identifying one or more variants in a first portion of sequence information to determine a TMB score, wherein the variants include one or more mutations selected from the group consisting of single nucleotide variants (SNVs), insertions or deletions (indels), copy number variants (CNVs), fusions, transpositions, translocations, frameshifts, duplications, repeat variants, and epigenetic variants. In some embodiments of these embodiments, the repeat variants include one or more microsatellite variants. In certain embodiments of these embodiments, the method further includes the step of identifying one or more other variants in a second portion of sequence information to determine a TMB score, wherein the other variants include one or more somatic mutations in nucleic acids derived from one or more immune cells in a blood sample.

[0078] In another aspect, the disclosure relates to a method for analyzing a number of analytes in a blood sample from a subject diagnosed with cancer. The method includes (a) isolating a first set of extracellular free nucleic acids (cfNA) and a second set of nucleic acids derived from one or more immune cells in the blood sample; and (b) amplifying one or more regions of the second set of nucleic acids encoding at least a portion of the alpha and / or beta subunits of a T cell receptor to produce a second set of enriched nucleic acids. The method also includes (c) sequencing one or more regions of the first set of cfNA and one or more regions of the second set of enriched nucleic acids to produce sequence information; and (d) determining the tumor mutational burden (TMB) score of the subject from the sequence information. Furthermore, the method also includes (e) identifying one or more clonal types of the immune repertoire from the sequence information; and (f) correlating the TMB score with one or more clonal types to identify one or more novel antigen orphan immune receptors in the subject, thereby analyzing a number of analytes in a blood sample from a subject diagnosed with cancer.

[0079] In another aspect, the disclosure relates to a method for analyzing nucleic acids in a blood sample derived from a subject using at least part computer. The method includes (a) receiving sequence information obtained from nucleic acids in a blood sample derived from a subject by computer, wherein at least a first portion of the sequence information includes sequencing reads obtained from extracellular free nucleic acids (cfNA) in the blood sample, and at least a second portion of the sequence information includes sequencing reads obtained from nucleic acids derived from one or more immune cells in the blood sample. The method also includes (b) identifying one or more variants in the first portion of the sequence information and one or more clone types of an immunorepertory in the second portion of the sequence information, thereby analyzing the nucleic acids in the blood sample derived from the subject.

[0080] In some embodiments, the method includes (i) determining the observed mutation count in a first portion of sequence information; (ii) determining the tumor percentage and / or coverage of at least cfNA in the first portion of sequence information to generate sequencing parameters; (iii) determining the predicted mutation percentage and / or predicted distribution of the predicted mutation percentage taking the sequencing parameters into account to generate prediction results; and (iv) adjusting the observed mutation count taking the prediction results into account to generate adjusted results, thereby determining the tumor mutational load (TMB) in the subject. In a particular embodiment, the method includes (i) quantifying a plurality of different repeat lengths present in each of a plurality of repeating nucleic acid loci derived from a first and / or second portion of sequence information to generate a site score for each of the plurality of repeating nucleic acid loci, wherein the sequence information includes a population of repeating nucleic acid loci; (ii) calling a given repeating nucleic acid locus unstable if the site score for the repeating nucleic acid locus exceeds a site-specific training threshold for the given repeating nucleic acid locus to generate a repeating nucleic acid instability score including a plurality of unstable repeating nucleic acid loci derived from the plurality of repeating nucleic acid loci; and (iii) classifying the repeating nucleic acid instability state of a blood sample as unstable if the repeating nucleic acid instability score exceeds a population training threshold for the population of repeating nucleic acid loci in a blood sample to determine the repeating nucleic acid instability state of the blood sample.

[0081] In a particular embodiment, the method includes the steps of (i) quantifying multiple different repeat lengths present in each of a plurality of repeating deoxyribonucleic acid (DNA) loci derived from a first and / or second portion of sequence information to generate a site score for each of the plurality of repeating DNA loci, wherein the sequence information includes a population of repeating DNA loci; (ii) for each of the plurality of repeating DNA loci, comparing the site score of the given repeating DNA locus with a site-specific training threshold for the given repeating DNA locus; and (iii) a given repeat (iv) If the site score of a sex DNA locus exceeds a site-specific training threshold for a given repetitive DNA locus, the given repetitive DNA locus is called unstable, and a repetitive DNA instability score is generated that includes multiple unstable repetitive DNA loci derived from multiple repetitive DNA loci; (iv) If the repetitive DNA instability score exceeds a population training threshold for a population of repetitive DNA loci in a blood sample, the repetitive DNA instability state of the blood sample is classified as unstable, thereby determining the repetitive DNA instability state of the blood sample. In some embodiments, the method includes: (i) quantifying multiple different repeat lengths present in each of a plurality of microsatellite loci derived from a first and / or second portion of sequence information to generate a site score for each of the plurality of microsatellite loci, wherein the sequence information includes a population of microsatellite loci; (ii) calling a given repetitive nucleic acid locus unstable if the site score of the repetitive nucleic acid locus exceeds a site-specific training threshold for the given repetitive nucleic acid locus to generate a repetitive nucleic acid instability score including a plurality of unstable repetitive nucleic acid loci derived from the plurality of repetitive nucleic acid loci; and (iii) classifying the repetitive nucleic acid instability state of a blood sample as unstable if the repetitive nucleic acid instability score exceeds a population training threshold for the population of repetitive nucleic acid loci in a blood sample to determine the repetitive nucleic acid instability state of the blood sample.

[0082] In some embodiments, the method includes: (i) quantifying multiple different repeat lengths present in each of a plurality of repeating deoxyribonucleic acid (DNA) loci derived from a first and / or second portion of sequence information to generate a site score for each of the plurality of repeating DNA loci, wherein the sequence information includes a population of repeating DNA loci; (ii) comparing the site score of a given repeating DNA locus with a site-specific training threshold for a given repeating DNA locus for each of the plurality of repeating DNA loci; (iii) if the site score of a given repeating DNA locus exceeds a site-specific training threshold for a given repeating DNA locus, calling the given repeating DNA locus unstable to generate a repeating DNA instability score including a plurality of unstable repeating DNA loci derived from the plurality of repeating DNA loci; and (iv) if the repeating DNA instability score exceeds a population training threshold for a population of repeating DNA loci in a blood sample, classifying the repeating DNA instability state of the blood sample as unstable to determine the repeating DNA instability state of the blood sample.In another embodiment, the method is: (i) to quantify multiple different repeat lengths present in each of a plurality of microsatellite loci derived from a first and / or second portion of sequence information to generate a site score for each of the plurality of microsatellite loci, wherein the sequence information comprises a population of microsatellite loci; (ii) for each of the plurality of microsatellite loci, to compare the site score of the given microsatellite locus with a site-specific training threshold for the given microsatellite locus; and (iii) a given microsatellite locus (iv) If the site score of a locus exceeds a site-specific training threshold for a given microsatellite locus, the given microsatellite locus is called unstable, and a microsatellite instability score is generated that includes multiple unstable microsatellite loci derived from multiple microsatellite loci; (iv) If the microsatellite instability score exceeds a population training threshold for a population of microsatellite loci in a blood sample, the microsatellite instability (MSI) state of the blood sample is classified as unstable, thereby determining the MSI state of the blood sample.

[0083] In some embodiments, the method involves capturing multiple sets of target regions of extracellular free DNA (cfDNA), including a set of sequence-variable target regions and an epigenetic target region set, thereby resulting in a set of captured cfDNA molecules, where cfDNA molecules corresponding to the sequence-variable target region set are captured within the set of captured cfDNA molecules with a higher capture yield than cfDNA molecules corresponding to the epigenetic target region set. In certain embodiments, cfNA comprises extracellular free DNA (cfDNA). In some embodiments, nucleic acids derived from one or more immune cells in a blood sample comprise mRNA and / or gDNA. In some embodiments, nucleic acids derived from one or more immune cells in a blood sample encode at least a portion of an immunological polypeptide selected from the group consisting of antibodies, B cell receptors, and T cell receptors. In some embodiments, the method involves obtaining cfNA from a plasma fraction or serum fraction of a blood sample. In certain embodiments, the method involves obtaining nucleic acids derived from one or more immune cells from a buffy coat fraction of a blood sample.

[0084] In another aspect, the disclosure relates to a system including a communication interface that obtains sequencing information from one or more nucleic acids in a sample derived from a subject via a communication network. The system includes a computer that communicates with the communication interface, and includes at least one computer processor, and a computer-readable medium including machine-executable code that, when executed by the at least one computer processor, performs a method comprising: (i) receiving sequence information obtained from nucleic acids in a blood sample obtained from a subject diagnosed with cancer, wherein at least a first portion of the sequence information includes a sequencing read obtained from extracellular free nucleic acid (cfNA) in the blood sample, and at least a second portion of the sequence information includes a sequencing read obtained from nucleic acids derived from one or more immune cells in the blood sample; (ii) determining a tumor mutational burden (TMB) score of the subject from at least the first portion of the sequence information; (iii) identifying one or more clonal types of the immune repertoire in at least the second portion of the sequence information; and (iv) correlating the TMB score with one or more clonal types to identify one or more novel antigen orphan immune receptors in the subject.

[0085] The accompanying drawings incorporated herein, and constituting part thereof, illustrate a particular embodiment and, together with the specification, serve to illustrate certain principles of the methods, computer-readable media, and systems disclosed herein. The descriptions presented herein will be better understood when read in conjunction with the accompanying drawings, which are included as examples and are not limiting. Throughout the drawings, it will be understood that similar components are identified by similar reference numbers unless otherwise indicated in the context. It will also be understood that some or all of the drawings may be illustrative diagrams and do not necessarily indicate the actual relative size or location of the elements shown. [Brief explanation of the drawing]

[0086] [Figure 1]Figure 1 is a flowchart schematically illustrating the steps of an exemplary method for adjusting the TMB according to some embodiments of the present invention.

[0087] [Figure 2] Figure 2 is a schematic diagram of an exemplary system suitable for use in a particular embodiment of the present invention.

[0088] [Figure 3] Figure 3 is a flowchart illustrating a combined workflow for extracellular free DNA analysis and immunorepertory sequencing from the same test sample, according to one embodiment of the present disclosure. The combined results of immunorepertory profiling and TMB analysis lead to an enhancement of the response score to immunotherapy.

[0089] [Figure 4] Figures 4A and 4B are schematic diagrams illustrating a sample preparation method for a TCR gDNA assay (Figure 4A) and an immune receptor RNA assay (Figure 4B) according to one embodiment of the present disclosure.

[0090] [Figure 5] Figure 5 is a flowchart showing a combined workflow for extracellular free DNA analysis and immunorepertory sequencing derived from the same test sample, according to one embodiment of the present disclosure.

[0091] [Figure 6] Figure 6 shows TMB score plots for samples with a small tumor percentage and low coverage, with (o) and without (Δ) the TMB correction or adjustment methods disclosed herein.

[0092] [Figure 7] Figure 7 is a flowchart illustrating an exemplary TMB workflow according to some embodiments of the present invention.

[0093] [Figure 8]Figures 8A and 8B are plots showing variants correlated with non-synonymous coding single nucleotide variants (SNVs). Specifically, Figure 8A is a plot showing synonymous SNVs correlated with non-synonymous SNVs (Pearson's r = 0.90; number of non-synonymous SNVs (x axis); number of synonymous SNVs (y axis)). Figure 8B is a plot showing indels correlated with non-synonymous SNVs (Pearson's r = 0.71; number of non-synonymous SNVs (x axis); number of indels (y axis)). Figure 8C is a plot showing the correlation between the rate of intronic SNVs and the rate of exonic SNVs (Pearson's r = 0.89; rate of intronic SNVs (x axis); rate of exonic SNVs (y axis)).

[0094] [Figure 9] Figures 9A-9D are violin plots showing that tumor shedding correction in large-panel assays eliminates the dependence of mutation count on tumor shedding (Figure 9A (max MAF bins (%) (x axis); mutation count (y axis))) and input cfDNA (Figure 9B (molecular coverage (×1000) (x axis); mutation count (y axis))) and consequently yields pTMB that is largely independent of these input metrics (Figures 9C (max MAF bins (%) (x axis); TMB (mut / Mb) (y axis)) and 9D (molecular coverage (×1000) (x axis); TMB (mut / Mb) (y axis))).

[0095] [Figure 10] Figures 10A (TMB(mut / Mb) (x-axis); sample percentage (%) (y-axis)) and 10B (TMB(mut / Mb) (x-axis); tumor type (y-axis)) are plots showing the TMB distribution across cohorts and between tumor types, respectively.

[0096] [Figure 11]Figures 11A (Principal Component (PC) 1 (x-axis); PC2 (y-axis)) and 11B (TMB (mut / Mb) (x-axis); Ethnicity (y-axis)) are plots showing principal component analysis (PCA) clustering and TMB scores, respectively.

[0097] [Figure 12] Figure 12 is a plot showing the correlation between TMB and oncogenic mutations (TMB(mut / Mb)(x-axis); driver mutations(y-axis)).

[0098] [Figure 13] Figure 13A is a plot showing that the clonality and chromosomal instability of somatic mutations vary significantly across the TMB landscape. Figure 13B shows that high-scoring microsatellite instability (MSI-High) was detected in a subset of TMB-High samples. [Modes for carrying out the invention]

[0099] definition To facilitate understanding of this disclosure, certain terms are defined below. Additional definitions of the following terms and other terms may be provided throughout this specification. If the definitions below conflict with definitions in any application or patent incorporated by reference, the definitions provided in this application should be used to understand the meaning of those terms.

[0100] As used herein and in the appended claims, the singular forms “one (a),” “one (an),” and “the” include multiple references unless the context explicitly states otherwise. Thus, for example, a reference to “one (a) method” includes one or more methods, and / or the types described herein and / or steps that would be apparent to a person skilled in the art by reading this disclosure.

[0101] It should be understood that the terminology used herein is solely for the purpose of describing specific embodiments and is not limiting. Furthermore, unless otherwise defined, all scientific and technical terms used herein have the same meaning as those generally understood by those skilled in the art in which this disclosure relates. When describing and claiming methods, computer-readable media, and systems, the following terms and their grammatical variations are used according to the definitions below.

[0102] About: As used herein, “about” or “approximately” means a value or element similar to the given reference value or element when applied to one or more values ​​or elements of interest. In certain embodiments, unless otherwise specified or the context makes it clear that “about” or “approximately” means a range of values ​​or elements that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less than 1% of the given reference value or element in either direction (greater or less) (unless such number exceeds 100% of the possible values ​​or elements).

[0103] Adjusted tumor mutation burden: As used herein, “adjusted tumor mutation burden,” “adjusted tumor mutation burden,” or “adjusted result” refers to the observed mutation count adjusted to account for tumor percentage, coverage, and / or other sequencing parameters.

[0104] To administer: As used herein, “to administer” or “to administer” a therapeutic agent (e.g., an immunological therapeutic agent) to a subject means to give, apply, or bring the composition into contact with the subject. Administration can be achieved by any of several routes, including, for example, local, oral, subcutaneous, intramuscular, intraperitoneal, intravenous, intrathecal, and intradermal.

[0105] Adapter: As used herein, “adapter” generally refers to a short nucleic acid (e.g., less than about 500 nucleotides, less than about 100 nucleotides, or less than about 50 nucleotides in length) that is at least partially double-stranded and used for ligating to one or both ends of a given sample nucleic acid molecule. An adapter may include sequencing primer-binding sites at both ends, which include nucleic acid primer-binding sites to enable amplification of the nucleic acid molecule adjacent to the adapter, and / or primer-binding sites for application to sequencing, such as various next-generation sequencing (NGS) applications. An adapter may also include binding sites for capture probes, such as oligonucleotides attached to a flow cell support. An adapter may also include nucleic acid tags as described herein. A nucleic acid tag is generally positioned relative to amplification primers and sequencing primer-binding sites such that the nucleic acid tag is included in the amplicon and sequence reading of a given nucleic acid molecule. The same or different adapters can be ligated to each end of a nucleic acid molecule. In some embodiments, the same adapter is ligated to each end of a nucleic acid molecule, except that the nucleic acid tag is different. In some embodiments, the adapter is a Y-shaped adapter with one end blunt-ended or tail-added as described herein for conjugation with a nucleic acid molecule, the nucleic acid molecule also blunt-ended or tail-added with one or more complementary nucleotides. In yet other embodiments, the adapter is a bell-shaped adapter with a blunt or tail-added end for conjugation with the nucleic acid molecule to be analyzed. Other examples of adapters include adapters having a T-tail and adapters having a C-tail.

[0106] Amplification: As used herein, "amplification" or "amplification" means, with respect to nucleic acids, producing a large number of copies of a polynucleotide or a portion of a polynucleotide, generally starting from a small amount of polynucleotide (e.g., a single polynucleotide molecule), such that the amplified product or amplicon is generally detectable. Amplification of polynucleotides encompasses a variety of chemical and enzymatic processes.

[0107] Barcode: As used herein, “barcode” or “molecular barcode” refers, in the case of nucleic acids, to a nucleic acid molecule containing a sequence that can function as a molecular identifier. For example, typically, individual “barcode” sequences can be attached to each DNA fragment during next-generation sequencing (NGS) library preparation, so that each read can be identified and sorted before final data analysis.

[0108] Cancer type: As used herein, “cancer type” means a type or subtype of cancer as defined, for example, by histopathological examination. Cancer types are defined according to any conventional criteria, for example, based on their occurrence in a given tissue (e.g., hematological cancer, central nervous system (CNS) cancer, brain cancer, lung cancer (small cell and non-small cell), skin cancer, nasal cancer, pharyngeal cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, bowel cancer, rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer, stomach cancer, breast cancer, prostate cancer, ovarian cancer, lung cancer, intestinal cancer). Cancer can be defined based on factors such as its origin (e.g., carcinoma, sarcoma, lymphoma, cholangiocarcinoma, leukemia, mesothelioma, melanoma, or glioblastoma), soft tissue cancer, neuroendocrine cancer, gastroesophageal cancer, head and neck cancer, gynecological cancer, colorectal cancer, urothelial carcinoma, solid tumor, xenocarcinoma, allocarcinoma), unknown primary origin, and / or based on those belonging to the same cell lineage (e.g., carcinoma, sarcoma, lymphoma, cholangiocarcinoma, leukemia, mesothelioma, melanoma, or glioblastoma), and / or based on cancer markers such as Her2, CA15-3, CA19-9, CA-125, CEA, AFP, PSA, HCG, hormone receptors, and NMP-22. Cancer can also be classified by its stage (e.g., stage 1, 2, 3, or 4) and whether it is primary or secondary.

[0109] Extracellular free nucleic acids: As used herein, “extracellular free nucleic acids” refers to nucleic acids that are not contained within cells or are not otherwise bound to cells, or, in some embodiments, nucleic acids remaining in the sample after the removal of intact cells. Extracellular free nucleic acids may include all unencapsulated nucleic acids sourced from, for example, body fluids of the subject (e.g., blood, plasma, serum, urine, cerebrospinal fluid (CSF)). Extracellular free nucleic acids include DNA (cfDNA), RNA (cfRNA), and hybrids thereof, including genomic DNA, mitochondrial DNA, circulating DNA, siRNA, miRNA, circulating RNA (cRNA), tRNA, rRNA, small nucleolar RNA (snoRNA), Piwi-interacting RNA (piRNA), high molecular weight non-coding RNA (high molecular weight ncRNA), and / or fragments of any of these. Extracellular free nucleic acids may be double-stranded, single-stranded, or hybrids thereof. Extracellular free nucleic acids may be released into body fluids through secretion or cell death processes, such as cell necrosis, apoptosis, etc. Some extracellular free nucleic acids are released into the body fluid from cancer cells, such as circulating tumor DNA (ctDNA). Other extracellular free nucleic acids are released from healthy cells. CtDNA may be unencapsulated tumor-derived fragmented DNA. Extracellular free nucleic acids may have one or more epigenetic modifications, for example, they may be acetylated, 5-methylated, ubiquitinated, phosphorylated, SUMOylated, ribosylated, and / or citrullinated.

[0110] Intracellular nucleic acids: As used herein, “intracellular nucleic acids” means nucleic acids that are present in at least one or more cells at the time the sample is obtained or taken from the subject, even if they are later removed as part of a given analytical process (e.g., by cell lysis).

[0111] Clonal hematopoietic mutation: As used herein, “clonal hematopoietic mutation” refers to the acquisition of somatic properties of a genomic mutation in hematopoietic stem cells and / or progenitor cells that leads to increased clonality.

[0112] Clone: ​​As used herein, “clone” refers to a unique nucleotide sequence resulting from a mutation or rearrangement process of the nucleotide sequence encoding the immune cell receptor (for example, a unique nucleotide sequence encoding the CDR3 sequence of the T cell receptor (TCR) polypeptide chain).

[0113] Comparative Results: As used herein, “comparative results” means results or a set of results that can be used to compare a given test sample or result to one or more possible characteristics of the test sample or result, and / or one or more possible prognostic outcomes and / or one or more customized therapies for a subject from whom the test sample was obtained or otherwise derived. Comparative results are generally obtained from a set of reference samples (e.g., from subjects having the same type of cancer as the test subject and / or from subjects receiving or having received the same therapy as the test subject). In certain embodiments, for example, adjusted TMB scores are compared to the comparative results to identify substantial matches between the adjusted TMB scores determined for the test sample and the TMB scores determined for the set of reference samples. Generally, the TMB scores determined for the set of reference samples point to one or more customized therapies. Thus, once substantial matches are identified, the corresponding customized therapies are also identified as potential therapy pathways for the subject from whom the test sample was obtained.

[0114] Confidence interval: As used herein, “confidence interval” means a range of values ​​within which a given parameter’s value falls with a predetermined probability.

[0115] Control Sample: As used herein, “control sample” means a sample of known composition and / or known characteristics and / or parameters (e.g., known tumor percentage, known coverage, known TMB, and / or similar) that is analyzed together with or compared to the test sample to evaluate the accuracy of the analytical procedure. In some embodiments, the control samples used in the control sample dataset may be specific to the type of cancer and / or the treatment.

[0116] Control sample dataset: As used herein, “control sample dataset” means a set threshold 腫瘍割合 This refers to a dataset of control samples with a larger tumor percentage.

[0117] Copy number variant: As used herein, “copy number variant,” “CNV,” or “copy number variation” refers to the phenomenon in which a section of the genome is repeated, the number of repeats within the genome varies among individuals in the population under consideration, and also varies between two conditions or states in an individual (for example, a CNV may vary in an individual before and after receiving therapy).

[0118] Coverage: As used herein, "coverage" refers to the number of nucleic acid molecules that represent a particular base position.

[0119] Customized therapy: As used herein, “customized therapy” means therapy that results in a desired therapy outcome for a subject or population of subjects having or falling within a defined range of TMB scores.

[0120] Deoxyribonucleic acid or ribonucleic acid: As used herein, “deoxyribonucleic acid” or “DNA” means a natural or modified nucleotide having a hydrogen group at the 2' position of the sugar moiety. DNA generally consists of a chain of nucleotides composed of four types of nucleotide bases: adenine (A), thymine (T), cytosine (C), and guanine (G). As used herein, “ribonucleic acid” or “RNA” means a natural or modified nucleotide having a hydroxyl group at the 2' position of the sugar moiety. RNA generally consists of a chain of nucleotides composed of four types of nucleotides: A, uracil (U), G, and C. As used herein, the term “nucleotide” means a natural or modified nucleotide. Certain pairs of nucleotides bind specifically to each other in a complementary manner (referred to as complementary base pairing). In DNA, adenine (A) pairs with thymine (T), and cytosine (C) pairs with guanine (G). In RNA, adenine (A) pairs with uracil (U), and cytosine (C) pairs with guanine (G). When a first nucleic acid strand is joined to a second nucleic acid strand composed of nucleotides complementary to the nucleotides of the first strand, these two strands join to form a double helix. As used herein, “nucleic acid sequencing data,” “nucleic acid sequencing information,” “sequence information,” “nucleic acid sequence,” “nucleotide sequence,” “genome sequence,” “gene sequence,” or “fragment sequence,” or “nucleic acid sequencing read” refers to any information or data indicating the order and identity of nucleotide bases (e.g., adenine, guanine, cytosine, and thymine or uracil) within a nucleic acid molecule such as DNA or RNA (e.g., whole genome, whole transcriptome, exome, oligonucleotide, polynucleotide, or fragment).It should be understood that this instruction intends to refer to sequence information obtained using all available techniques, platforms, or technologies, including, but not limited to, capillary electrophoresis, microarrays, ligation-based systems, polymerase-based systems, hybridization-based systems, direct or indirect nucleotide identification systems, pyrosequencing, ion or pH-based detection systems, and electronic signature-based systems.

[0121] Driver mutation: As used herein, “driver mutation” means a mutation that drives the progression of cancer.

[0122] Predicted distribution of predicted mutation rates: As used herein, “predicted distribution of predicted mutation rates” refers to the range of predicted mutation rates determined by a statistical distribution model, such as the binomial distribution.

[0123] Expected mutation count: As used herein, "expected mutational count," "adjusted mutation count," or "adjusted mutational count" refers to the adjusted mutation count.

[0124] Predicted mutation rate (f): As used herein, “predicted mutation rate” refers to the proportion of actual somatic mutations called in a sample, which is derived from the sensitivity of the bioinformatics analysis and the relative MAF distribution derived from a database of relative MAFs of all mutations determined by bioinformatics analysis in a control sample dataset.

[0125] Prediction Results: As used herein, “prediction results” means estimated, likely, or expected outcomes.

[0126] Immune repertoire: As used herein, “immune repertoire” refers to the total number of T cell receptors and B cell receptors (i.e., immunoglobulins) that constitute the adaptive immune system of interest.

[0127] Immunotherapy: As used herein, “immunotherapy” means treatment with one or more agents that act to kill cancer cells or at least inhibit their growth, preferably reducing further cancer growth, shrinking the size of the tumor, and / or stimulating the immune system to eliminate the cancer. Some such agents bind to targets present on cancer cells, some to targets present on immune cells but not on cancer cells, and some to targets present on both cancer cells and immune cells. Such agents include, but are not limited to, checkpoint inhibitors and / or antibodies. Checkpoint inhibitors are inhibitors of immune system pathways that maintain self-tolerance and modulate the duration and amplitude of the physiological immune response in peripheral tissues in such a way that secondary tissue damage is minimized (see, for example, Pardoll, Nature Reviews Cancer 12, 252-264 (2012)). Exemplary agents include antibodies against PD-1, PD-2, PD-L1, PD-L2, CTLA-40, OX40, B7.1, B7He, LAG3, CD137, KIR, CCR5, CD27, or CD40. Other exemplary agents include pro-inflammatory cytokines such as IL-1β, IL-6, and TNF-α. Other exemplary agents are tumor-activated T cells, such as T cells activated by the expression of chimeric antigens that target tumor antigens recognized by T cells.

[0128] Indel: As used herein, “indel” refers to a mutation involving the insertion or deletion of a nucleotide within the genome in question.

[0129] To indicate: As used herein, “to indicate” means that a first element (e.g., the TMB score) is associated with a second element (e.g., a given therapy).

[0130] Maximum MAF: As used herein, "maximum MAF" or "max MAF" refers to the maximum MAF of all somatic cell variants in the sample.

[0131] Minor Allele Frequency: As used herein, “minor allele frequency” refers to the frequency of a non-major allele (e.g., an allele that is not the most common) in a given population of nucleic acids, such as a sample obtained from a subject. Generally, gene variants with low minor allele frequencies are relatively infrequent in samples.

[0132] Mutant Allele Ratio: As used herein, “mutation ratio,” “mutation load,” or “MAF” refers to the proportion of nucleic acid molecules in a given sample that have an allele alteration or mutation at a given location within the genome. MAF is generally expressed as a ratio or percentage. For example, MAF is generally less than about 0.5, less than 0.1, less than 0.05, or less than 0.01 (i.e., less than 50%, less than 10%, less than 5%, or less than 1%) of all somatic variants or alleles present at a given locus.

[0133] Mutation: As used herein, “mutation” refers to a variation from a known reference sequence, including, for example, single nucleotide variants (SNVs), copy number variants or variations (CNVs) / abnormalities, insertions or deletions (indels), gene fusions, transpositions, translocations, frameshifts, duplications, repeat extensions, and epigenetic variants. Mutations may be germline or somatic mutations. In some embodiments, the reference sequence for comparison is the wild-type genome sequence of the species under consideration, generally the human genome, from which the test sample is provided.

[0134] Mutation Caller: As used herein, “mutation caller” means an algorithm (typically incorporated into software or otherwise executed by a computer) used to identify mutations in test sample data (e.g., sequence information obtained from a subject).

[0135] Mutation count: As used herein, "mutation count" or "mutational count" refers to the number of somatic mutations in the whole genome, exome, or targeted region of a nucleic acid sample.

[0136] Novel Antigen Orphan Immunoreceptor Information: As used herein, “novel antigen orphan immunoreceptor information” refers to information relating to an antigen that was not previously recognized by the immune system of a given object. Generally, novel antigen orphan immunoreceptor information is obtained from modified polypeptides or encoding polynucleotides formed as a result of one or more tumor-associated mutations. In certain embodiments, novel antigen orphan immunoreceptor information is derived from sequence information relating to these modified polypeptides or encoding polynucleotides.

[0137] Neoplasm: As used herein, the terms “neoplasm” and “tumor” are interchangeable. These terms refer to the abnormal growth of cells in an object. Neoplasms or tumors may be benign, potentially malignant, or malignant. Malignant tumors are referred to as cancer or carcinomatous tumors.

[0138] Next-generation sequencing: As used herein, “next-generation sequencing” or “NGS” refers to sequencing techniques that have increased throughput compared to conventional Sanger-based and capillary electrophoresis-based methods, and that have the ability to generate, for example, hundreds of thousands of relatively small sequence reads at once. Some examples of next-generation sequencing techniques, but not limited to these, include synthesis sequencing, ligation sequencing, and hybridization sequencing.

[0139] Non-synonymous mutation: As used herein, “non-synonymous mutation” means a mutation that alters the amino acid sequence of the encoded polypeptide.

[0140] Nucleic acid tags: As used herein, “nucleic acid tags” means short nucleic acids (e.g., less than about 500 nucleotides, less than about 100 nucleotides, less than about 500 nucleotides, or less than about 10 nucleotides in length) used to distinguish nucleic acids from different samples of different types or subjected to different processing (e.g., to represent a sample index), or to distinguish them from different nucleic acid molecules in the same sample (e.g., to represent a molecular barcode). Nucleic acid tags comprise a predetermined, fixed, non-random, random, or semi-random oligonucleotide sequence. Such nucleic acid tags can be used to label different nucleic acid molecules or different nucleic acid samples or subsamples. Nucleic acid tags may be single-stranded, double-stranded, or at least partially double-stranded. Nucleic acid tags may have the same length or a varying length as needed. Nucleic acid tags may comprise a double-stranded molecule having one or more blunt ends, a double-stranded molecule containing a 5' or 3' single-stranded region (e.g., a protrusion), and / or one or more other single-stranded regions at other positions within a given molecule. Nucleic acid tags can be attached to one end of another nucleic acid (e.g., a sample nucleic acid to be amplified and / or sequenced), or to both ends. Nucleic acid tags can be decoded to reveal information about a given nucleic acid, such as its origin, morphology, or processing. For example, nucleic acid tags can be used to enable pooling and / or parallel processing of a large number of samples containing nucleic acids with different molecular barcodes and / or sample indices, in which case the nucleic acids are subsequently deconvoluted by detecting (e.g., reading) the nucleic acid tags. Nucleic acid tags can also be referred to as identifiers (e.g., molecular identifiers, sample identifiers). In addition, or instead, nucleic acid tags can be used as molecular identifiers (e.g., to distinguish different molecules or parent molecules in the same sample or sub-sample from different amplicons). This includes, for example, uniquely tagging different nucleic acid molecules in a given sample, or non-uniquely tagging such molecules.In non-unique tagging applications, each nucleic acid molecule can be tagged using a limited number of tags (i.e., molecular barcodes), and thus different molecules can be distinguished based on a combination of their endogenous sequence information (e.g., start and / or stop locations where they map to a selected reference genome, subsequences at one or both ends of the sequence, and / or sequence length) and at least one molecular barcode. Generally, a sufficient number of different molecular barcodes are used so that the probability of any two molecules having the same endogenous sequence information (e.g., start and / or stop locations, subsequences at one or both ends of the sequence, and / or length) and also having the same molecular barcode is low (e.g., an expected probability of less than about 10%, less than about 5%, less than about 1%, or less than about 0.1%).

[0141] Mutation count: As used herein, "mutation count" or "mutational count" refers to the number of somatic mutations determined by the bioinformatics analysis described herein.

[0142] Passenger mutation: As used herein, “passenger mutation” means a mutation that occurs in a cell that does not result in a change in fitness but simultaneously or later acquires a driver mutation.

[0143] Polynucleotide: As used herein, “polynucleotide,” “nucleic acid,” “nucleic acid molecule,” or “oligonucleotide” refers to a linear polymer of nucleosides joined by nucleoside linkages (including deoxyribonucleosides, ribonucleosides, or analogs thereof). Generally, a polynucleotide contains at least three nucleosides. The size range of oligonucleotides is often from a small number of monomeric units, e.g., 3-4, to several hundred monomeric units. Whenever a polynucleotide is represented by a sequence of letters such as “ATGCCTG,” it will be understood that, unless otherwise specified, the nucleotides are in the 5'→3' order from left to right, and in the case of DNA, “A” represents deoxyadenosine, “C” represents deoxycytidine, “G” represents deoxyguanosine, and “T” represents deoxythymidine. The letters A, C, G, and T may be used to indicate the base itself, a nucleoside, or a nucleotide containing a base, as is standard practice in the art.

[0144] Processing: As used herein, the terms “processing,” “calculating,” and “comparing” may be used interchangeably. In certain applications, this term refers to determining differences, such as differences in numbers or sequences. For example, gene expression, copy number variations (CNVs), indels, and / or single nucleotide variant (SNV) values ​​or sequences may be processed.

[0145] Relative MAF: As used herein, “relative MAF” refers to an estimate of the MAF of a particular variant compared to the maximum (max) MAF in the sample.

[0146] Reference Sequence: As used herein, “reference sequence” refers to a known sequence used for comparison with a sequence determined experimentally. For example, a known sequence may be an entire genome, a chromosome, or any segment thereof. A reference generally contains at least about 20, at least about 50, at least about 100, at least about 200, at least about 250, at least about 300, at least about 350, at least about 400, at least about 450, at least about 500, at least about 1000, or more nucleotides. A reference sequence may be one that can be aligned with a single contiguous sequence of genome or chromosome, or it may include non-contiguous segments that can be aligned with various regions of genome or chromosome. An example of a reference sequence is the human genome, such as hG19 and hG38.

[0147] Sample: As used herein, “sample” means anything that can be analyzed by the methods and / or systems disclosed herein.

[0148] Limit of Detection (LoD): As used herein, “Limit of Detection” means the minimum amount of substance (e.g., nucleic acid) in a sample that can be measured by a given assay or analytical technique.

[0149] Sensitivity: As used herein, "sensitivity" means the probability of detecting the presence of a mutation in a given MAF and coverage.

[0150] Sequencing: As used herein, “sequencing” refers to any of several techniques used to determine the sequence (e.g., identity and order of monomeric units) of a biomolecule, such as nucleic acids, such as DNA or RNA. Exemplary sequencing methods include, but are not limited to, targeted sequencing, single-molecule real-time sequencing, exon or exome sequencing, intron sequencing, electron microscopy-based sequencing, panel sequencing, transistor-mediated sequencing, direct sequencing, random shotgun sequencing, Sangerdideoxy termination sequencing, whole-genome sequencing, hybridization sequencing, pyrosequencing, capillary electrophoresis, double-strand sequencing, cycle sequencing, and single-nucleotide extension sequencing. Examples of sequencing methods include solid-phase sequencing, high-throughput sequencing, large-scale parallel-process signature sequencing, emulsion PCR, co-amplification-PCR at low denaturation temperature (COLD-PCR), multiplex PCR, sequencing by reversible dye terminator, paired-end sequencing, short-read sequencing, exonuclease sequencing, sequencing by ligation, short-read sequencing, single-molecule sequencing, sequencing by synthesis, real-time sequencing, reverse-terminator sequencing, nanopore sequencing, 454 sequencing, Solexa Genome Analyzer sequencing, SOLiD® sequencing, MS-PET sequencing, and combinations thereof. In some embodiments, sequencing can be performed using genetic analysis instruments, such as those commercially available from Illumina, Inc., Pacific Biosciences, Inc., or Applied Biosystems / Thermo Fisher Scientific, among many others.

[0151] Sequence information: As used herein, “sequence information” means, with respect to nucleic acid polymers, the order and identity of monomeric units (e.g., nucleotides) within the polymer.

[0152] Single nucleotide variant: As used herein, “single nucleotide variant” or “SNV” means a single nucleotide mutation or variation that occurs at a specific location within the genome.

[0153] Somatic mutation: As used herein, “somatic mutation” means a mutation in the genome that occurs after conception. Somatic mutations can occur in any cell of the body other than germ cells and are therefore not passed on to offspring.

[0154] Substantial Match: As used herein, “substantial match” means that at least one value or element is at least substantially equal to at least one second value or element. In certain embodiments, a customized therapy is identified, for example, when the adjusted result (e.g., adjusted TMB score) and the comparison result (e.g., TMB score determined from one or more control or reference samples) are at least substantially or substantially matched.

[0155] Subject: As used herein, “Subject” refers to animals such as mammalian species (e.g., humans) or bird species (e.g., birds), or other living organisms such as plants. More specifically, Subject may be vertebrates, mammals such as mice, primates, monkeys, or humans. Animals include farm animals (e.g., beef cattle, dairy cows, poultry, horses, pigs, etc.), sports animals, and companion animals (e.g., pets or support animals). Subject may be a healthy individual, an individual with a disease or predisposition to a disease or suspected to have a disease, or an individual in need of treatment or suspected to need treatment. The terms “individual” or “patient” are intended to be interchangeable with “Subject.”

[0156] For example, subjects may be individuals diagnosed with cancer, individuals scheduled to undergo cancer therapy, and / or individuals who have undergone at least one cancer therapy. Subjects may be in a state of cancer remission. As another example, subjects may be individuals diagnosed with an autoimmune disease. As yet another example, subjects may be pregnant or planning to become pregnant women who may have been diagnosed with or suspect they have a disease, such as cancer or an autoimmune disease.

[0157] Synonymous variant: As used herein, “synonymous variant” means a variant that does not alter the amino acid sequence of the encoded polypeptide.

[0158] Threshold: As used herein, “threshold” refers to a predetermined value used to characterize the same parameter value, experimentally determined for different samples, in relation to a threshold.

[0159] threshold 最大(max)MAF :When used herein, "threshold" 最大(max)MAF " refers to a predetermined value of max MAF used to characterize the max MAF value determined experimentally for different samples. In some embodiments, for example, a threshold 最大(max)MAF This could be approximately 0.5%, 1%, 2%, 5%, 10%, or another selected threshold.

[0160] threshold TMBスコア :When used herein, "threshold" TMBスコア " refers to a predetermined value of the TMB score used to characterize the TMB score values ​​determined experimentally for different samples. In a particular embodiment, for example, a threshold TMBスコア This could be approximately 5, 10, 15, 20, 25, 30, 35, 40, or another selected threshold.

[0161] threshold 腫瘍割合 :When used herein, "threshold" 腫瘍割合" refers to a predetermined value of the tumor percentage used to characterize the tumor percentage values ​​determined experimentally for different samples. In a particular embodiment, for example, a threshold 腫瘍割合 This could be approximately 1%, 2%, 3%, 4%, 5%, or another selected threshold.

[0162] threshold f :When used herein, "threshold" f " refers to a predetermined value of the predicted mutation fraction (f) used to characterize the predicted mutation fraction values ​​determined experimentally for different samples. In some embodiments, for example, a threshold f This could be greater than approximately 0.05, greater than approximately 0.1, greater than approximately 0.2, greater than approximately 0.3, greater than approximately 0.4, greater than approximately 0.5, greater than approximately 0.6, greater than approximately 0.7, greater than approximately 0.8, greater than approximately 0.9, or another selected threshold.

[0163] Tumor Percentage: As used herein, “tumor percentage” refers to an estimate of the proportion of nucleic acid molecules in a given sample that originate from tumors. For example, the tumor percentage of a sample may be a measure derived from the sample’s max MAF, the pattern of the sample’s sequencing coverage, the length of cfDNA fragments in the sample, or any other selected sample feature. In some examples, the tumor percentage of a sample is equal to the sample’s max MAF.

[0164] Tumor Mutation Load: As used herein, the terms “tumor mutation burden (TMB),” “tumor mutational burden (TMB),” “cancer mutation burden,” “mutational load,” or “mutation load” are interchangeable. These terms refer to the total number of mutations present in a sequenced portion of the tumor genome, e.g., somatic mutations. TMB may refer to the number of coding, base substitutions, indels, or other mutations per megabase of the tumor genome or exome or targeted region of the genome being investigated. These may indicate the detection, evaluation, calculation, or prediction of sensitivity and / or resistance to cancer therapeutic agents or drugs, e.g., immune checkpoint inhibitors, antibodies. Tumors with high levels of TMB may express more novel antigens, a type of cancer-specific antigen, potentially enabling a more robust immune response and, therefore, a longer-lasting response to immunotherapy. Since the immune system relies on a sufficient number of novel antigens to respond appropriately, the number of somatic mutations can act as a surrogate for determining the number of novel antigens in a tumor. TMB can be used to estimate the robustness of the immune response to drug treatment and the effectiveness of drug treatment in a subject. Germline variants and somatic variants can be distinguished using bioinformatics to identify antigenic somatic variants.

[0165] Variant: As used herein, “variant” may be referred to as an allele. Variants typically exist at a frequency of 50% (0.5) or 100% (1), depending on whether the allele is heterozygous or homozygous. For example, germline variants are hereditary and typically have a frequency of 0.5 or 1. Somatic variants, however, are acquired variants and typically have a frequency of <0.5. Major and non-major alleles at a locus refer to nucleic acids having a locus occupied by nucleotides of a reference sequence and a locus occupied by variant nucleotides different from the reference sequence, respectively. Measurements at a locus may take the form of allele proportions (AF), which measure the frequency with which an allele is observed in a sample. Detailed explanation

[0166] Introduction

[0167] Cancer encompasses a large group of genetic disorders that share common characteristics: abnormal cell growth and the potential for metastasis beyond the origin of cells within the body. The molecular basis for these diseases lies in mutations and / or epigenetic changes that lead to alterations in cellular phenotype, regardless of whether these detrimental changes are acquired through inheritance or are somatic. Troublingly, these molecular changes generally vary not only among patients with the same type of cancer, but even within a given patient's own tumor.

[0168] Given the variability of mutations observed in the majority of cancers, one of the challenges in cancer care is identifying the therapy most likely to elicit a response from a patient, taking into account their individualized cancer type. Various biomarkers are used to match cancer patients with appropriate treatment, including cancer immunotherapy. One such response biomarker is tumor mutational load (TMB), which is a quantitative measure of the total number of mutations per coding region in a given cancer genome. To date, the application of this response biomarker has been limited in part to the means by which TMB can be measured and analyzed.

[0169] This disclosure provides a method, computer-readable medium, and system that are useful in determining and analyzing TMB in patient samples and that can serve as a guide for cancer treatment decisions. Traditionally, when the tumor proportion (e.g., the proportion of mutant alleles (MAF)) and / or coverage is low, the sensitivity of the assay for calling mutations decreases, so TMB obtained by counting mutation rates is often inaccurate. Therefore, in certain embodiments, observed TMB is adjusted to take into account various measures of assay sensitivity, e.g., tumor proportion (setting the MAF of mutations in a given sample), coverage, and / or similar. Without such adjustment, for example, a sample that is actually TMB-High but has a low tumor proportion and / or low coverage is more likely to be misreported as TMB-Low. Such outcomes can have significant downstream implications for the patient if treatment decisions are made based on such results.

[0170] Methods for adjusting tumor gene mutation levels

[0171] This application discloses various methods for adjusting the TMB to account for variability in tumor percentage and / or coverage in a given assay, variability which may otherwise lead to inaccurate TMB reporting. In certain embodiments, the method includes adjusting a raw somatic mutation count (e.g., SNVs and indels called by a given bioinformatics pipeline or workflow) by a model prediction of the actual percentage of mutations in a sample in a panel under consideration, called by a particular bioinformatics pipeline. In certain embodiments, the model uses the logic of a bioinformatics mutation (e.g., SNVs, indels, and / or similar) caller and a binary sampling solution to calculate the sensitivity of the mutation caller over the predicted distribution of MAFs in a sample coverage setting and / or in that sample. In some embodiments, the predicted distribution of MAFs in a sample is derived by the relative MAFs of all mutations called in a control sample in a control sample dataset. In some of these embodiments, the model calculates the predicted proportion of observed mutations and the probability distribution of said proportion, which can be summarized, for example, as a 95% confidence interval for said proportion. This can be used to output a highly sensitive (e.g., the most likely actual mutation count) and / or highly specific (e.g., the least likely actual mutation count) calculation of the predicted mutation count. In certain embodiments of these embodiments, the predicted mutation count is then divided by the size of the genomic region being analyzed to obtain the mutation rate (i.e., TMB or TMB score). Compared to the “gold standard” of TMB calculated from whole exome sequencing of tumor-derived tissue samples or TMB calculated when the tumor proportion of the sample is high, for example, the TMB calculated by performing the adjustments described herein will be more accurate than the TMB measured without these adjustments.

[0172] For further illustration, Figure 1 provides a flowchart schematically illustrating the steps of an exemplary method for adjusting TMB according to some embodiments of the present invention. As shown, Method 100 includes, in step 110, determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample derived from the subject. Method 100 also includes, in step 112, determining the tumor percentage and / or coverage of the nucleic acid to generate sequencing parameters, and in step 114, determining the predicted mutation percentage and / or predicted distribution of the predicted mutation percentage taking the sequencing parameters into account to generate a prediction result. Furthermore, Method 100 also includes, in step 116, adjusting the observed mutation count taking the prediction result into account to generate an adjusted result, thereby determining the TMB in the subject.

[0173] In certain embodiments, Method 100 includes additional upstream and / or downstream steps. In some embodiments, for example, Method 100 begins in step 102 and provides a sample derived from the subject in step 104 (e.g., providing a blood sample obtained from the subject). In these embodiments, the workflow of Method 100 generally also includes, in step 106, amplifying the nucleic acid in the sample to produce amplified nucleic acid, and in step 108, sequencing the amplified nucleic acid to produce sequence information, and then in step 110, determining the observed mutation count from the sequence information. Nucleic acid amplification (including associated sample preparation), nucleic acid sequencing, and associated data analysis are further described herein.

[0174] In some embodiments, Method 100 includes various steps downstream from the adjusted result generated in step 116. Some of these examples include, in step 118, identifying one or more customized therapies for the subject by comparing the adjusted result with one or more comparison results pointing to one or more therapies. In some embodiments, Method 100 includes reporting the results to the subject or physician. In other exemplary embodiments, Method 100 also includes, in step 120, administering at least one of the identified customized therapies to the subject (for example, to treat the subject's cancer or another disease or condition) if the adjusted result and the comparison results are substantially matched, and then terminating in step 122.

[0175] For further illustration, for any given sample, the model is input with respect to the tumor percentage (e.g., the maximum MAF of somatic mutations ("non-major AF") or another estimate of the tumor percentage) and / or several metrics of coverage (e.g., the median of unique molecules per base). The model uses a computation based on a mutation (e.g., SNVs, indels, and / or similar) calling algorithm to output the percentage of mutations in the panel space that are predicted to be called under these conditions, and the distribution of the percentage of mutations that are predicted to be observed (e.g., can be summarized using a 95% confidence interval for the range of the percentage of mutations that are predicted to be observed). In some embodiments, the predicted percentage of actual mutations in a given panel that will be called by bioinformatics analysis is calculated by assessing the probability (i.e., sensitivity) that a mutation in a given MAF is called across a range of predicted MAFs (i.e., converting the distribution of relative MAFs to MAFs by multiplying by the sample tumor percentage).

[0176] Generally, the sensitivity of a mutation caller is estimated from the probability and coverage of a mutation being called in a given MAF using the bioinformatics analysis algorithms described herein. In some embodiments, the probability can be calculated using an empirical distribution based on prior data (i.e., performing experiments to test how frequently a sample has known mutations in various MAFs with varying coverage, resulting in an "empirical" mutation call sensitivity). In some embodiments, the probability can be calculated using a binomial distribution. In some embodiments, the probability can be calculated using a multicomponent distribution based on numerous requirements for mutation calling (e.g., the number of molecules supporting the mutation, along with some other considerations of prior predictions about the mutation; the mutation call sensitivity, along with whether the mutation is present in a cancer hotspot region—a combination based on all of these components can be used). In some embodiments, the sensitivity of the mutation caller can be estimated based on specific basic requirements such as variant type (e.g., SNV, short indel, or long indel), genomic context (e.g., hotspot regions, skeletal regions, local GC content, etc.), or sample context (e.g., sequencing metrics such as GC, MAPD, coverage profile; or tumor-type sample metrics).

[0177] In some embodiments, the method includes defining a predictive distribution of MAF in a particular sample. In certain embodiments of these embodiments, the relative MAF distribution is empirically fitted to a curve across all control samples in a control sample dataset. The curve to be empirically fitted is given by the following equation: F = 1 / (1 + (P_50 / relative - MAF) n ) This is described by (wherein F is the cumulative distribution function, P_50 is the median of relative MAFs, relative-MAF is the relative MAF, and n is an index for fitting the shape of the relative distribution). In a particular embodiment, the predictive distribution of the predicted mutation rate is obtained from one or more datasets of relative mutation allele rates (MAFs) observed in at least one control sample dataset. The control sample dataset generally includes at least about 25 to at least about 30,000 or more control samples. In some embodiments, the control sample dataset includes approximately 50, 75, 100, 150, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 2,500, 5,000, 7,500, 10,000, 15,000, 20,000, 25,000, 50,000, 100,000, 1,000,000, or more control samples. In some embodiments, the maximum MAF observed in the control samples constitutes approximately 0.5%, approximately 1%, approximately 2%, approximately 5%, approximately 10%, or more. In some embodiments, the threshold of the maximum MAF (i.e., threshold) 最大(max)MAF ) is used. In these embodiments, threshold 腫瘍割合 This is generally around 1%, 2%, 3%, 4%, 5%, or another selected threshold.

[0178] In some embodiments, the prediction results are expressed by the following equation: The proportion of observed mutations (f) = Σ MAF (P(Variation Call | MAF) × P(Variation in MAF)) (In the formula, P is probability, MAF is the proportion of mutated alleles, P(mutation call|MAF) is the probability that a mutation is called in a particular MAF, which represents the sensitivity of the mutation caller in a particular MAF and coverage of the sample, and P(mutation in MAF) is the probability of a mutation in a MAF, which represents the predictive distribution of the MAF of the sample.) Use this to make a decision.

[0179] Since this proportion is essentially a binomial sampling probability, the upper and lower bounds (and confidence intervals) of the distribution for this proportion can be estimated using the confidence interval of the binomial proportion:

number

[0180] In some of these embodiments, if the tumor proportion of the sample is large and / or coverage is large, the raw somatic mutation count (observed mutation count) is then divided by this output from the model to obtain the number of mutations called and predicted in the panel. The proportion estimate can be used as the best estimate of the actual / predicted mutation count. If necessary, an upper limit of the 95% confidence interval (f) for the proportion can be set. 上限 Using this, a lower bound of the actual mutation count can be obtained (i.e., there is a 95% probability that the actual mutation count is at least equal to that mutation count), which is a highly specific output (high specificity with respect to reporting TMB-High). In certain embodiments of these, the predicted / adjusted mutational count is then divided by the size of the genomic region being analyzed to obtain the TMB score of the sample. In some embodiments, the TMB score of the sample is a threshold TMBスコア If the TMB score of the sample is greater than the threshold, the sample is classified as a TMB-High sample. TMBスコア If it is smaller than the threshold, the sample is classified as a TMB-Low sample. In certain embodiments, for example, the threshold TMBスコア This could be approximately 5, 10, 15, 20, 25, 30, 35, 40, or another selected threshold.

[0181] Generally, the observed variant count and / or tumor percentage includes multiple somatic mutations identified in the nucleic acid. In some of these embodiments, one or more known cancer driver and / or passenger mutations are excluded from the observed variant count and / or tumor percentage. In certain embodiments, the observed variant count and / or tumor percentage includes multiple synonymous mutations, multiple non-synonymous mutations, and / or multiple non-coding mutations identified in the nucleic acid. If necessary, the observed variant count and / or tumor percentage includes multiple mutations, including single nucleotide variants (SNVs), insertions or deletions (indels), copy number variants (CNVs), fusions, transpositions, translocations, frameshifts, duplications, repeat extensions, epigenetic variants, and / or similar. In certain embodiments, clonal hematopoietic mutations are excluded from the observed variant count and / or tumor percentage. If necessary, the predicted variant percentage is used as the observed percentage of the actual variant count.

[0182] In certain embodiments, alternative methods are used to incorporate tumor proportions other than, for example, using the maximum variant allele proportion (max)MAF of all somatic mutations (e.g., derived from SNVs and indel data). Some exemplary alternatives include using coverage-based tumor proportions, using cfDNA fragment length-based tumor proportions, using germline and / or somatic variant-based tumor proportions, or any combination of these methods. In some embodiments, tumor proportions can be estimated using sequencing coverage patterns of a given set of genomic regions and standard maximum likelihood methods. In some embodiments, tumor proportions can be estimated from differences in cfDNA fragment size distribution between a given set of genomic regions in a sample and regions with copy number variations. In some embodiments, tumor proportions can be estimated by adjusting the MAF of germline and / or somatic variants for copy number variations observed in the sample. In some embodiments, tumor proportions (i.e., somatic proportions) can be estimated using sequencing coverage of probes used to capture a set of cfDNA molecules in a target genomic region. In some embodiments, tumor proportions can be estimated based on the methylation status of cfDNA molecules in a sample. In some embodiments, the tumor percentage can be estimated using max MAF, but adjustments can be made for the copy number at a particular location within the genome. In some embodiments, somatic cell MAF in the sample can be combined with the tumor percentage estimate. In some embodiments, methylation can be used to estimate the tumor percentage based on identifying molecules originating from the tumor using methylation patterns, and thus the percentage of tumor molecules can be identified and used to estimate the tumor percentage. In some examples, all combinations of the embodiments described above, or combinations of at least a subset of the embodiments described above, can be used in the model to estimate the tumor percentage.

[0183] In some embodiments, the tumor percentage of a given sample may be less than approximately 0.05%, less than approximately 0.1%, less than approximately 0.2%, less than approximately 0.5%, less than 1%, less than approximately 2%, less than approximately 3%, less than approximately 4%, or less than 5% of the total nucleic acids.

[0184] In some embodiments, germline variants are excluded from the determination of the TMB score. In some embodiments, the germline / somatic status of a variant can be determined using a beta-binomial distribution model. The beta-binomial distribution is used to model the mean and variance of the mutant allele counts of common germline single nucleotide polymorphisms (SNPs) located near the candidate variant. If the candidate variant deviates from the distribution of these local germline SNPs, the variant is called a “somatic variant”; otherwise, the variant is called a “germline variant.” Methods and systems described in PCT / US2018 / 052087 and U.S. Patent Provisional Applications 62 / 726,182, 62 / 823,578, and 62 / 857,048 are incorporated by reference.

[0185] Coverage is also incorporated into the model in various ways as needed. In some embodiments, coverage can be incorporated as the median number of unique cfDNA fragments per base. In some embodiments, the model can be made base-specific, and the method includes calculating the sensitivity at each base using the number of unique cfDNA fragments at that base. In certain embodiments, the method includes determining coverage by identifying a number of unique sequencing readouts containing a given nucleotide position in a nucleic acid. In yet other exemplary embodiments, the method includes determining coverage by identifying the median number of unique extracellular free DNA (cfDNA) molecules constituting a given nucleotide position in a nucleic acid. Coverage is such that there are at least 10, 50, 100, 500, 1000, 5000, 10,000, 20,000, or 50,000 cfDNA fragments at a given base position.

[0186] The methods disclosed in this application generally involve obtaining sequence information from nucleic acids in a sample obtained from a subject. In certain embodiments, sequence information is obtained from targeted segments of nucleic acids. Basically any number of genomic regions are targeted as needed. The targeted segments are at least 10, at least 50, at least 100, at least 500, at least 1000, at least 2000, at least 5000, at least 10,000, at least 20,000, at least 50,000 or at least 100,000 (e.g., 25, 50, 75, 100, 200, 300, 400, 500, It may include different and / or overlapping genomic regions (600, 700, 800, 900, 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 15,000, 25,000, 30,000, 35,000, 40,000, 45,000, 50,000, or 100,000).

[0187] In certain embodiments, the method includes dividing the observed mutation count by the predicted result to produce an adjusted result. Generally, the adjusted result includes multiple mutations detected in the nucleic acid across a range of mutation allele percentages. Generally, the adjusted result includes the predicted highest probability actual mutation count and / or the predicted lowest probability actual mutation count. In certain embodiments, the predicted result includes the predicted mutation fraction. In certain embodiments, the adjusted result includes the adjusted mutation count. Generally, the adjusted mutation count is greater than or equal to the observed mutation count.

[0188] In other exemplary embodiments, driver mutations that do not represent the background exome mutation rate or TMB are excluded from the somatic mutation count (e.g., SNVs, indels, etc.). In some embodiments, mutations that are more likely to originate from clonal hematopoiesis rather than the tumor under consideration are excluded from the somatic mutation count. Mutations originating from clonal hematopoiesis can be identified using a curated list of mutations frequently observed in hematological cancers, based on literature and cancer databases (e.g., COSMIC). In some embodiments, clonal hematopoietic mutations can be identified using their context in the sample (e.g., MAFs) (e.g., the presence of other clonal hematopoietic variants in a range of similar MAFs) or by analyzing clonal hematopoietic mutations in a database of previously tested samples. In some embodiments, clonal hematopoietic mutations can be identified by sequencing DNA in patient samples derived from blood (e.g., white blood cells). In certain embodiments, the somatic mutation count also includes SNVs and indels at sites not reported as part of the specific panel used. For example, such mutations may occur in regions (e.g., introns) where mutations are "unreported" in a given application. This means, for example, a larger panel for counting somatic mutations, and therefore smaller sampling errors and a higher signal for mutation rates. In some embodiments, clonal hematopoietic mutations can be identified by using a probability / likelihood model that takes as input at least some of these parameters—patient age, tumor type, methylation status at its location, the size of the molecular fragments supporting the mutation in its sample and any other mutations. In some embodiments, these input parameters can be used to build a model or preference for whether each mutation is a clonal hematopoietic mutation. In some embodiments, machine learning algorithms such as logistic regression and random forests can be applied to identify clonal hematopoietic mutations.Methods and systems described in PCT / US2019 / 035214 may also be used, and these are incorporated by reference.

[0189] In other exemplary embodiments, the method involves using pooled evidence of mutations that are below the limit of detection (LOD) for any particular SNV, indel, and / or other type of mutation, but which have been shown to be more likely to be mutations at that base than the vast majority of other bases. In these embodiments, TMB-High samples generally have more of this evidence than TMB-Low samples. Some embodiments involve increasing the number of sites in a given panel that have TMB-correlated mutations, and / or further improving the estimation of tumor proportions. In certain embodiments, the estimation or determination of tumor proportions is also further improved by including sites that are potentially highly informative with respect to fragment mixes.

[0190] In some embodiments, subclonal mutations are excluded from the observed mutation count by filtering out somatic mutations with low MAF. In some embodiments, somatic mutations that are less than approximately 1%, less than 2%, less than 3%, less than 4%, less than 5%, less than 6%, less than 7%, less than 8%, less than 9%, less than 10%, less than 15%, less than 20%, less than 25%, or less than 30% of the maximum (max) MAF are excluded from the observed mutation count.

[0191] In some embodiments, the predicted / adjusted mutation count is adjusted using an exome calibrator, which is the ratio of the mutation rate of the panel being analyzed to the exome mutation rate. In some embodiments, the predicted / adjusted mutation count is divided by the exome calibrator. In some embodiments, the exome calibrator is 1.01, 1.02, 1.03, 1.04, 1.05, 1.06, 1.07, 1.08, 1.09, or at least 1.10. In some embodiments, the exome calibrator is determined from the exome mutation rate of samples in a cancer database, e.g., The Cancer Genome Atlas (TCGA). In some embodiments, the exome calibrator may be cancer type specific, and the exome mutation rate can be determined from samples (in the database) having a specific cancer type.

[0192] In some embodiments, resistance mutations are excluded from the observed mutation count. In some embodiments, resistance mutations can be identified using a curated list of frequently observed mutations in patient samples, based on literature and cancer databases (e.g., COSMIC, TCGA). In some embodiments, resistance mutations can be identified by analyzing a database of previously tested samples.

[0193] In some embodiments, resistance mechanisms or any other processes may introduce numerous mutations into a gene (e.g., BRCA1 / 2 reverse mutations in prostate cancer treated with KRAS or PARP inhibitors in CRC), and counting all of these mutations in the observed mutation count may not reflect the overall exome mutation rate. In such embodiments, a small number of mutations in a particular gene are excluded from the observed mutation count if the number of mutations in that particular gene is significantly higher than the number of mutations in that gene predicted based on the overall sample mutation count and gene size in the panel. The term "significantly higher" can be assessed based on a statistical sampling model. For example, if the sample mutation rate is 10 mutations per Mb (observed by analyzing a panel of cancer-related genes), and 5 of these mutations are present in the KRAS gene (in the panel), the mutation rate in KRAS will be much higher than that predicted based on the sample mutation rate. The number of KRAS-derived mutations counted in the observed mutation count is limited to approach the predicted rate based on the sample mutation rate.

[0194] In some embodiments, the TMB score can be used to predict whether a patient will respond to immunotherapy. In certain embodiments, the presence of a specific type of mutation in a particular gene across a set of analyzed genes (e.g., loss of function in STK11, KEAP1, PTEN, etc.) and / or a patient's mutational signature (e.g., the patient has a C>T translocation) can be used in combination with the TMB score to predict whether a patient will respond to immunotherapy. In certain embodiments, if a patient has loss of function in a particular gene, such as STK11, KEAP1, and PTEN, but not limited to these, the patient will not respond to immunotherapy regardless of the TMB score. For example, if a patient has (i) a high TMB score and (ii) loss of function in STK11, the loss of function / driver mutation takes precedence and the patient will not respond to therapy.

[0195] In some embodiments, the preparation methods disclosed herein include obtaining a sample from a subject (e.g., a human or other mammalian subject). Exemplary sample types used as needed are further described herein. Essentially any type of nucleic acid (e.g., DNA and / or RNA) can be evaluated according to the methods disclosed in this application. Some examples include extracellular free nucleic acids (e.g., tumor-derived cfDNA and / or similar), intracellular nucleic acids including circulating tumor cells (e.g., obtained by lysing intact cells in a sample), and circulating tumor nucleic acids.

[0196] In some embodiments, the TMB-corrected model cannot be applied to samples that exhibit tumor shedding below a certain cutoff and / or have a tumor percentage or either parameter with coverage below a certain cutoff, resulting in a very low predicted rate of mutations. In some embodiments, the TMB-corrected model cannot be applied to samples that predominantly contain CHIP mutations that interfere with accurate TMB estimation. Criteria for determining whether the TMB-corrected model can be applied to a sample include methods used to calculate whether the sample predominantly contains CHIP mutations. Such methods include, for example, (a) a large proportion of mutations known to be CHIP (using a curated database or other methods to identify CHIP, e.g., buffy sequencing, fragment mix); (b) evidence of solid tumors, e.g., the absence of known driver mutations for the tumor type of the sample; and (c) any combination of the above. In some embodiments, the TMB-corrected model cannot be applied to samples that contain low tumor shedding or CHIP dominance.

[0197] In these embodiments, the method generally also includes various sample or library preparation steps for preparing nucleic acids for sequencing. Many different sample preparation techniques are well known to those skilled in the art. Basically any of those techniques can be used or adapted for use in the implementation of the method herein. For example, typical steps for preparing nucleic acids for sequencing include, in addition to various purification steps for isolating the nucleic acid from other components in a given sample, tagging the nucleic acid with a molecular barcode, attaching an adapter (which may include a molecular barcode), amplifying the nucleic acid once or multiple times, enriching the targeted segment of the nucleic acid (e.g., using various target capture strategies, etc.), and / or similar. Exemplary library preparation processes are further described herein. Further details regarding nucleic acid sample / library preparation are also found, for example, in van Dijk et al., Library preparation methods for next-generation sequencing: Tone down the bias, Experimental Cell Research, 322(1): 12-20 (2014), Micic (Ed.), Sample Preparation Techniques for Soil, Plant, and Animal Samples (Springer Protocols Handbooks), 1st Ed., Humana Press (2016), and Chiu, Next-Generation Sequencing and Sequence Data Analysis, Bentham Science Publishers (2018), each of which is incorporated by reference.

[0198] Adjusted tumor mutation burden (TMB), determined by the methods disclosed herein, may be used, as necessary, to diagnose the presence of disease or condition, particularly cancer, in a subject; to characterize such disease or condition (e.g., to stage a given cancer, to determine cancer heterogeneity, etc.); to monitor response to treatment; to assess the potential risk of a given disease or condition developing; and / or to assess the prognosis of the disease or condition. Adjusted tumor mutation burden may also be used, as necessary, to characterize specific forms of cancer. Since cancers are often heterogeneous in both composition and staging, TMB data may be used to characterize specific subtypes of cancer, thereby assisting in diagnosis and treatment selection. This information may also provide subjects or healthcare practitioners with clues about the prognosis of specific types of cancer, enabling subjects and / or healthcare practitioners to adapt treatment options as the disease progresses. Some cancers become more invasive and genetically unstable as they progress. Other tumors remain benign, inactive, or quiescent.

[0199] Adjusted TMB may also be useful in determining disease progression and / or monitoring recurrence. In certain cases, for example, successful treatment may initially increase adjusted TMB because the number of cancer cell deaths increases and nucleic acids are shed (shedding). In these cases, generally, adjusted TMB then decreases as the tumor size continues to shrink as therapy progresses. In other cases, successful treatment may also decrease TMB and / or the proportion of non-major alleles without an initial increase in tumor mutational load. Furthermore, if the cancer is found to be in remission after treatment, adjusted TMB can be used to monitor residual disease or disease recurrence in the patient.

[0200] sample

[0201] The sample may be any biological sample isolated from the subject. The sample may include body tissue, whole blood, platelets, serum, plasma, stool, red blood cells, white blood cells or leucocytes, endothelial cells, tissue biopsy material (e.g., biopsy material from a known or suspected solid tumor), cerebrospinal fluid, synovial fluid, lymph, ascites, interstitial fluid or extracellular fluid (e.g., exudate from the intercellular space), gingival exudate, sulcus exudate, bone marrow, pleural fluid, cerebrospinal fluid, saliva, mucus, sputum, semen, sweat, and urine. The sample is preferably a body fluid, particularly blood and its proportions, as well as urine. Such a sample contains nucleic acids excreted from the tumor. Nucleic acids may include DNA and RNA and may be in double-stranded or single-stranded form. The sample may be in the form originally isolated from the subject, or it may have been subjected to further processing to remove or add components such as cells, to enrich one component with another, or to convert one form of nucleic acid to another, for example, to convert RNA to DNA, or to convert single-stranded nucleic acid to double-stranded nucleic acid. Therefore, for example, the body fluid for analysis may be plasma or serum containing extracellular free nucleic acids, such as extracellular free DNA (cfDNA).

[0202] In some embodiments, the sample volume of bodily fluids obtained from the subject depends on the desired reading depth relative to the region being sequenced. Exemplary volumes are about 0.4–40 ml, about 5–20 ml, and about 10–20 ml. For example, the volume may be about 0.5 ml, about 1 ml, about 5 ml, about 10 ml, about 20 ml, about 30 ml, about 40 ml, or larger in milliliters. The volume of sampled plasma is generally between about 5 ml and about 20 ml.

[0203] A sample may contain varying amounts of nucleic acids. Generally, the amount of nucleic acid in a given sample is equivalent to the amount of many genome equivalents. For example, a sample of about 30 ng of DNA is equivalent to about 10,000 (10 4 ) haploid human genome equivalents, and in the case of cfDNA, approximately 200 billion (2 × 10⁻¹⁶). 11It may contain ) individual polynucleotide molecules. Similarly, a sample of about 100 ng of DNA may contain about 30,000 haploid human genome equivalents, and in the case of cfDNA, about 600 billion individual molecules.

[0204] In some embodiments, the sample includes nucleic acids from different sources, e.g., nucleic acids from cells and nucleic acids from an extracellular free source (e.g., a blood sample). Generally, the sample includes nucleic acids with mutations. For example, the sample may include DNA with germline mutations and / or somatic mutations, as needed. Generally, the sample includes DNA with cancer-related mutations (e.g., cancer-related somatic mutations).

[0205] Exemplary amounts of extracellular free nucleic acids in a sample before amplification generally range from about 1 femtogram (fg) to about 1 microgram (μg), for example, about 1 picogram (pg) to about 200 nanograms (ng), about 1 ng to about 100 ng, and about 10 ng to about 1000 ng. In some embodiments, the sample contains up to about 600 ng, up to about 500 ng, up to about 400 ng, up to about 300 ng, up to about 200 ng, up to about 100 ng, up to about 50 ng, or up to about 20 ng of extracellular free nucleic acid molecules. If necessary, the amount is at least about 1 fg, at least about 10 fg, at least about 100 fg, at least about 1 pg, at least about 10 pg, at least about 100 pg, at least about 1 ng, at least about 10 ng, at least about 100 ng, at least about 150 ng, or at least about 200 ng of extracellular free nucleic acid molecules. In certain embodiments, the amount is up to about 1 fg, about 10 fg, about 100 fg, about 1 pg, about 10 pg, about 100 pg, about 1 ng, about 10 ng, about 100 ng, about 150 ng, or about 200 ng of extracellular free nucleic acid molecules. In some embodiments, the method involves obtaining between about 1 fg and about 200 ng of extracellular free nucleic acid molecules from a sample.

[0206] Extracellular free nucleic acids generally have a size distribution between approximately 100 and 500 nucleotide lengths, with molecules between approximately 110 and 230 nucleotide lengths accounting for approximately 90% of the molecules in the sample, the mode being approximately 168 nucleotides, and a second secondary peak between approximately 240 and 440 nucleotide lengths. In certain embodiments, extracellular free nucleic acids range from approximately 160 to 180 nucleotide lengths, or from approximately 320 to 360 nucleotide lengths, or from approximately 440 to 480 nucleotide lengths.

[0207] In some embodiments, extracellular free nucleic acids found in solution are isolated from body fluids through a splitting step that separates them from intact cells and other insoluble components of the body fluid. In some of these embodiments, the splitting includes techniques such as centrifugation or filtration. Alternatively, cells in the body fluid are lysed, and the extracellular free nucleic acids and intracellular nucleic acids are processed together. Generally, after the addition of buffers and washing steps, the extracellular free nucleic acids are precipitated, for example, using alcohol. In certain embodiments, an additional clarification step is used, such as a silica-based column, to remove impurities or salts. Nonspecific bulk carrier nucleic acids are added as needed throughout the reaction to optimize certain aspects of the exemplary procedure, such as yield. After such processing, the sample generally contains various forms of nucleic acids, including double-stranded DNA, single-stranded DNA, and / or single-stranded RNA. If necessary, single-stranded DNA and / or single-stranded RNA are converted to double-stranded form, and thus the double-stranded form is included in subsequent processing and analysis steps.

[0208] Nucleic acid tags

[0209] In some embodiments, nucleic acid molecules (derived from a polynucleotide sample) can be tagged with a sample index and / or molecular barcode (commonly referred to as a “tag”). The tag can be incorporated into an adapter or otherwise conjugated by, among other methods, chemical synthesis, ligation (e.g., blunt-end ligation or sticky-end ligation), or overlap-extension polymerase chain reaction (PCR). Such an adapter can then be ultimately conjugated to a target nucleic acid molecule. In other embodiments, one or more rounds of amplification cycles (e.g., PCR amplification) are generally applied to introduce the sample index into the nucleic acid molecule using a conventional nucleic acid amplification method. Amplification can be performed in one or more reaction mixtures (e.g., multiple microwells in an array). The molecular barcode and / or sample index can be introduced simultaneously or in any sequential order. In some embodiments, the molecular barcode and / or sample index are introduced before and / or after the sequence capture step. In some embodiments, only the molecular barcode is introduced before probe capture, and the sample index is introduced after the sequence capture step. In some embodiments, both the molecular barcode and the sample index are introduced before the probe-based capture step. In some embodiments, the sample index is introduced after the sequence capture step. In some embodiments, the molecular barcode is incorporated into nucleic acid molecules (e.g., cfDNA molecules) in the sample via an adapter by ligation (e.g., blunt-end ligation or sticky-end ligation). In some embodiments, the sample index is incorporated into nucleic acid molecules (e.g., cfDNA molecules) in the sample by overlap extension polymerase chain reaction (PCR). Generally, sequence capture protocols involve introducing a single-stranded nucleic acid molecule complementary to a target nucleic acid sequence, such as the coding sequence of a genomic region, where mutations in such a region are associated with a type of cancer.

[0210] In some embodiments, the tag can be positioned at one or both ends of the sample nucleic acid molecule. In some embodiments, the tag is an oligonucleotide of a predetermined, random, or semi-random sequence. In some embodiments, the tag may be less than about 500 nucleotides, less than about 200 nucleotides, less than about 100 nucleotides, less than about 50 nucleotides, less than about 20 nucleotides, less than about 10 nucleotides, less than about 9 nucleotides, less than about 8 nucleotides, less than about 7 nucleotides, less than about 6 nucleotides, less than about 5 nucleotides, less than about 4 nucleotides, less than about 3 nucleotides, less than 2 nucleotides, or less than about 1 nucleotide in length. The tag may be randomly or non-randomly attached to the sample nucleic acid.

[0211] In some embodiments, each sample is uniquely tagged using a sample index or a combination of sample indices. In some embodiments, each nucleic acid molecule in a sample or subsample is uniquely tagged using a molecular barcode or a combination of molecular barcodes. In other embodiments, multiple molecular barcodes can be used such that the molecular barcodes are not necessarily unique to one another (e.g., non-unique molecular barcodes). In these embodiments, molecular barcodes are generally attached to individual molecules such that a unique sequence is created by the combination of the molecular barcode and the sequence to which it can be attached (e.g., by ligation). Detection of combinations of non-uniquely tagged molecular barcodes and endogenous sequence information (e.g., the first (start) and / or last (end) portions corresponding to the sequence of the original nucleic acid molecule in the sample, subsequences of sequence readings at one or both ends, the length of the sequence readings, and / or the length of the original nucleic acid molecule in the sample) generally makes it possible to assign a unique identity to a particular molecule. The length or number of base pairs of individual sequence readings can also be used as needed to assign a unique identity to a given molecule. As described herein, assigning a unique identity to a single-stranded nucleic acid fragment may subsequently allow for identification of the fragment from the parent and / or complementary strands.

[0212] In some embodiments, molecular barcodes are introduced with a set of identifiers (e.g., combinations of unique or non-unique molecular barcodes) and a predicted ratio of molecules in the sample. One example of the format uses approximately 2 to 1,000,000 different molecular barcodes, or approximately 5 to 150 different molecular barcodes, or approximately 20 to 50 different molecular barcodes ligated to both ends of the target molecule. Alternatively, approximately 25 to 1,000,000 different molecular barcodes can be used. For example, 20-50 × 20-50 molecular barcodes can be used. The number of such identifiers is generally sufficient to ensure a high probability (e.g., at least 94%, 99.5%, 99.99%, or 99.999%) that different molecules with the same start and end points will obtain different identifier combinations. In some embodiments, approximately 80%, 90%, 95%, or 99% of molecules have the same combination of molecular barcodes.

[0213] In some embodiments, the assignment of unique or non-unique molecular barcodes in a reaction is carried out using, for example, the methods and systems described in U.S. Patent Applications Nos. 20010053519, 20030152490, and 20110160078, and U.S. Patents Nos. 6,582,908, 7,537,898, 9,598,731, and 9,902,992, each of which is incorporated herein by reference in whole. Alternatively, in some embodiments, only endogenous sequence information (e.g., start and / or stop positions, one or both terminal subsequences of the sequence, and / or length) can be used to identify different nucleic acid molecules in a sample.

[0214] nucleic acid amplification

[0215] The sample nucleic acid, sandwiched between adapters, is typically amplified by PCR and other amplification methods using nucleic acid primers that bind to primer-binding sites within the adapters that sandwich the DNA molecule to be amplified. In some embodiments, the amplification method involves a cycle of extension, denaturation, and annealing, which can be thermocycling or isothermal, as in transcription amplification, for example. Other typical amplification methods that may be used as needed include, among other techniques, ligase chain reaction, strand displacement amplification, nucleic acid sequence-based amplification, and auto-persistent sequence-based replication.

[0216] Generally, one or more rounds of amplification cycles are applied to introduce molecular barcodes and / or sample indices into nucleic acid molecules using conventional nucleic acid amplification methods. Amplification is generally carried out in one or more reaction mixtures. Molecular barcodes and sample indices are introduced simultaneously or in any sequential order as needed. In some embodiments, molecular barcodes and sample indices are introduced before and / or after the sequence capture step. In some embodiments, only molecular barcodes are introduced before probe-based capture, and sample indices are introduced after the sequence capture step. In certain embodiments, both molecular barcodes and sample indices are introduced before the probe-based capture step. In some embodiments, sample indices are introduced after the sequence capture step. Generally, sequence capture protocols involve introducing single-stranded nucleic acid molecules complementary to a target nucleic acid sequence, such as the coding sequence of a genomic region, where mutations in such regions are associated with a type of cancer. Generally, the amplification reaction generates multiple nucleic acid amplicons tagged non-uniquely or uniquely with molecular barcodes and sample indices ranging in size from approximately 200 nucleotides (nt) to approximately 700 nt, 250 nt to approximately 350 nt, or approximately 320 nt to approximately 550 nt. In some embodiments, the amplicon size is approximately 300 nt. In some embodiments, the amplicon size is approximately 500 nt.

[0217] Nucleic acid enrichment

[0218] In some embodiments, sequences are enriched before sequencing of nucleic acids. Enrichment is performed, as necessary, on specific target regions or non-specifically ("target sequences"). In some embodiments, the target regions are enriched using a differential tiling and capture scheme with nucleic acid capture probes ("baits") selected for one or more bait set panels. The differential tiling and capture scheme generally uses bait sets of varying relative concentrations (e.g., at different "resolutions") to differentially tile across genomic regions to which the baits are bound, subjecting them to a set of constraints (e.g., sequencer constraints such as sequencing load and the usefulness of each bait) to capture the target nucleic acids at the desired level for downstream sequencing. These target genomic regions may optionally include native or synthetic nucleotide sequences of nucleic acid constructs. In some embodiments, target sequences can be captured using biotin-labeled beads with probes for one or more target regions, and these regions may then be amplified to enrich the target regions, as necessary.

[0219] Sequence capture generally involves the use of oligonucleotide probes that hybridize with a target nucleic acid sequence. In certain embodiments, probe set strategies involve tiling probes across a region of interest. Such probes may range in length from, for example, about 60 nucleotides to about 120 nucleotides. Set depths may be about 2×, 3×, 4×, 5×, 6×, 8×, 9×, 10×, 15×, 20×, 50×, or deeper. The effectiveness of sequence capture generally depends, in part, on the length of the target molecule sequence that is complementary (or nearly complementary) to the probe sequence.

[0220] Nucleic acid sequencing

[0221] Sample nucleic acids, which are sandwiched between adapters as needed and pre-amplified or unamplified, are generally subjected to sequencing. Sequencing methods or commercially available formats used as needed include, for example, Sanger sequencing, high-throughput sequencing, pyrosequencing, synthesis sequencing, single-molecule sequencing, nanopore-based sequencing, semiconductor sequencing, ligation sequencing, hybridization sequencing, RNA-Seq (Illumina), Digital Gene Expression (Helicos), next-generation sequencing (NGS), single-molecule synthesis sequencing (SMSS) (Helicos), large-scale parallel sequencing, and Clonal Single Molecule sequencing. Examples include sequencing using Array (Solexa), shotgun sequencing, Ion Torrent, Oxford Nanopore, Roche Genia, Maxim-Gilbert sequencing, primer walking, PacBio, SOLiD, Ion Torrent, or Nanopore platforms. Sequencing reactions can be carried out in various sample processing devices that may include multiple lanes, multiple channels, multiple wells, or other means for processing multiple sample sets substantially simultaneously. Sample processing devices may also include multiple sample chambers to enable simultaneous processing of multiple operations.

[0222] Sequencing can be performed on one or more nucleic acid fragment types or regions known to contain markers for cancer or other diseases. Sequencing can also be performed on any nucleic acid fragment present in the sample. Sequencing can be performed on at least about 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9%, or 100% of the genome. In other cases, sequencing can be performed on less than about 5%, less than about 10%, less than about 15%, less than about 20%, less than about 25%, less than about 30%, less than about 40%, less than about 50%, less than about 60%, less than about 70%, less than about 80%, less than about 90%, less than about 95%, less than about 99%, less than about 99.9%, or less than 100% of the genome.

[0223] Simultaneous sequencing reactions can be carried out using multiplex sequencing techniques. In some embodiments, sequencing of extracellular free polynucleotides is performed in at least about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. In other embodiments, sequencing of extracellular free polynucleotides is performed in less than about 1,000, less than about 2,000, less than about 3,000, less than about 4,000, less than about 5,000, less than about 6,000, less than about 7,000, less than about 8,000, less than about 9,000, less than about 10,000, less than about 50,000, or less than 100,000 sequencing reactions. Sequencing reactions are generally carried out sequentially or simultaneously. Subsequent data analysis is generally performed on all or part of the sequencing reactions. In some embodiments, data analysis is performed on at least about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. In other embodiments, data analysis can be performed on fewer than about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. Exemplary read depths range from about 1,000 reads to about 50,000 reads per locus (base location).

[0224] In some embodiments, nucleic acid populations are prepared for sequencing by enzymatically forming blunt ends on double-stranded nucleic acids having single-stranded overhangs at one or both ends. In these embodiments, the populations are treated in the presence of nucleotides (e.g., A, C, G, and T or U) using an enzyme generally having 5'-3' DNA polymerase activity and 3'-5' exonuclease activity. Examples of enzymes or catalytic fragments used as needed include Krenow large fragments and T4 polymerase. For 5' overhangs, the enzyme generally extends the 3' end until it aligns with the 5' end of the recessed 3' end on the opposite strand, resulting in a blunt end. For 3' overhangs, the enzyme generally digests the 3' end to and sometimes beyond the 5' end of the opposite strand. If this digestion proceeds beyond the 5' end of the opposite strand, the gap can be filled by an enzyme with the same polymerase activity used for the 5' overhang. The formation of blunt ends in double-stranded nucleic acids facilitates, for example, adapter attachment and subsequent amplification.

[0225] In some embodiments, the nucleic acid population is subjected to additional processing, such as conversion from single-stranded nucleic acids to double-stranded nucleic acids and / or conversion from RNA to DNA. These forms of nucleic acids are also ligated with adapters and amplified as needed.

[0226] Nucleic acids can be subjected to the above-described process of forming blunt ends, either by pre-amplifying or without amplification, and, if necessary, by sequencing other nucleic acids in the sample to produce sequenced nucleic acids. Sequential nucleic acids can refer to either the sequence of nucleic acids (i.e., sequence information) or nucleic acids whose sequences have been determined. Sequencing can be performed so that the sequence data of individual nucleic acid molecules in the sample is obtained directly or indirectly from the consensus sequences of the amplified products of individual nucleic acid molecules in the sample.

[0227] In some embodiments, both ends of a double-stranded nucleic acid with a single-stranded protrusion in the sample after blunt-end formation are ligated to an adapter containing a molecular barcode, and the nucleic acid sequence and the molecular barcode introduced by the adapter are determined by sequencing. The blunt-end DNA molecule is ligated to the blunt ends of a double-stranded adapter (e.g., a Y-shaped or bell-shaped adapter) at least partially, as needed. Alternatively, to facilitate ligation (e.g., with respect to sticky-end ligation), the blunt ends of the sample nucleic acid and the adapter can be tail-added using complementary nucleotides.

[0228] Generally, a nucleic acid sample is brought into contact with a sufficient number of adapters such that the probability of any two copies of the same nucleic acid receiving the same combination of adapter barcodes (i.e., molecular barcodes) derived from adapters attached to both ends is low (e.g., <1 or 0.1%). Using adapters in this way makes it possible to identify families of nucleic acid sequences that share the same start and end points in the reference nucleic acid and are linked to the same combination of molecular barcodes. Such families represent the sequences of the amplified products of the nucleic acids in the sample before amplification. As they are modified by blunt end formation and adapter attachment, the sequences of the family members can be compiled to derive the consensus nucleotide(s) or complete consensus sequence for the nucleic acid molecule in the original sample. In other words, the nucleotides occupying a predetermined position in the nucleic acid in the sample are determined to be the consensus of the nucleotides occupying that corresponding position in the family member sequence. The family may include sequences from one or both strands of a double-stranded nucleic acid. If a family of members includes sequences from both strands of a double-stranded nucleic acid, all sequences are compiled and one strand's sequence is converted to its complement for the purpose of deriving a consensus nucleotide(s) or sequence. Some families contain only a single member sequence. In this case, this sequence can be considered the sequence of the nucleic acid in the sample before amplification. Alternatively, families containing only a single member sequence can be excluded from subsequent analysis.

[0229] The nucleotide variations of a sequenced nucleic acid can be determined by comparing the sequenced nucleic acid with a reference sequence. The reference sequence is often a known sequence, such as a known whole-genome sequence or partial-genome sequence from the subject (e.g., the whole-genome sequence of a human subject). The reference sequence may be, for example, hG19 or hG38. The sequenced nucleic acid may represent a sequence directly determined for the nucleic acid in the sample, as described above, or a consensus of sequences of amplified products of such nucleic acids. The comparison can be performed at one or more specified positions on the reference sequence. A subset of sequenced nucleic acids containing positions corresponding to the specified positions on the reference sequence can be identified when each sequence is maximally aligned. Within such a subset, it can be determined which sequenced nucleic acids, if any, contain nucleotide variations at the specified positions, and, if necessary, which contain the reference nucleotide (i.e., are identical to the reference sequence), if any. If the number of sequenced nucleic acids in the subset containing nucleotide variants exceeds a selected threshold, the variant nucleotide can be called at that specified position. The threshold may be a simple number such as at least 1, 2, 3, 4, 5, 6, 7, 9, or 10 for the sequenced nucleic acids in the subset containing nucleotide variants, or it may be a ratio such as at least 0.5, 1, 2, 3, 4, 5, 10, 15, or 20 for the sequenced nucleic acids in the subset containing nucleotide variants. The comparison can be repeated for any designated position of interest in the reference sequence. Sometimes, the comparison can be performed for at least approximately 20, 100, 200, or 300 consecutive positions in the reference sequence, for example, a designated position occupying approximately 20–500 or approximately 50–300 consecutive positions.

[0230] For further details regarding nucleic acid sequencing, including the formats and applications described herein, see, for example, Levy et al., Annual Review of Genomics and Human Genetics, 17: 95-115 (2016), Liu et al., J. of Biomedicine and Biotechnology, Volume 2012, Article ID 251364: 1-11 (2012), Voelkerding et al., Clinical Chem., 55: 641-658 (2009), MacLean et al., Nature Rev. Microbiol., 7: 287-296 (2009), and Astier et al., J Am Chem Soc., 128 (5): 1705-10 (2006), U.S. Patent Nos. 6,210,891, 6,258,568, 6,833,246, 7,115,400, 6,969,488, 5,912,148, 6,130,073, 7,169,560, 7,282,337, 7,482,120, 7,50 U.S. Patents 1,245, 6,818,395, 6,911,345, 7,501,245, 7,329,492, 7,170,050, 7,302,146, 7,313,308, and 7,476,503 are also provided, each of which is incorporated herein by reference in its entirety.

[0231] Comparison results

[0232] The adjusted tumor mutational load (TMB) of a given subject, determined according to the method disclosed in this application, is generally compared to a database of comparative results (e.g., TMB) from a reference population to identify customized or targeted therapies for that subject. In some embodiments, the TMB of the test subject and the comparative TMB are measured, for example, across the entire genome or the entire exome, while in other embodiments, their TMBs are measured, for example, based on a subset of the genome or exome or a targeted region, and they are estimated as needed to determine the TMB for the whole genome or the whole exome. Generally, the reference population includes patients with the same type of cancer as the test subject and / or patients who are or have received the same therapy as the test subject. In some embodiments, the TMB of the test subject and the comparative TMB are measured by determining mutation counts or loadings in a predetermined or selected set of genes or genomic regions. Essentially any gene (e.g., oncogenes) is selected as needed for such analysis. In certain embodiments of these designs, the selected genes or genomic regions include at least about 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 1,500, 2,000 or more selected genes or genomic regions. In some embodiments of these designs, the selected genes or genomic regions include, as may be, one or more genes listed in Table 1. [Table 1]

[0233] In certain embodiments, the selected gene or genomic region may include, as necessary, one or more genes listed in Table 2. [Table 2-1] [Table 2-2]

[0234] cancer

[0235] In certain embodiments, the methods and systems disclosed herein are used to identify customized therapies for treating a given disease or condition in a patient. Generally, the disease under consideration is a type of cancer. Non-specific examples of such cancers include biliary tract cancer, bladder cancer, transitional cell carcinoma, urothelial carcinoma, brain cancer, glioma, astrocytoma, breast cancer, metaplastic carcinoma, cervical cancer, cervical squamous cell carcinoma, rectal cancer, colorectal cancer, colon cancer, hereditary nonpolyposis colorectal cancer, colorectal adenocarcinoma, gastrointestinal stromal tumor (GIST), endometrial cancer, endometrial stromal sarcoma, esophageal cancer, esophageal squamous cell carcinoma, esophageal adenocarcinoma, intraocular melanoma, uveal melanoma, gallbladder cancer, gallbladder adenocarcinoma, renal cell carcinoma, clear cell carcinoma, transitional cell carcinoma, urothelial carcinoma, Wilms' tumor, leukemia, acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), chronic myelomonocytic leukemia (CLL). Examples of cancers include MML, liver cancer, hepatoma, hepatocellular carcinoma, cholangiocarcinoma, hepatoblastoma, lung cancer, non-small cell lung cancer (NSCLC), mesothelioma, B-cell lymphoma, non-Hodgkin lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, T-cell lymphoma, non-Hodgkin lymphoma, progenitor T-lymphoblastic lymphoma / leukemia, peripheral T-cell lymphoma, multiple myeloma, nasopharyngeal carcinoma (NPC), neuroblastoma, oropharyngeal cancer, oral squamous cell carcinoma, osteosarcoma, ovarian cancer, pancreatic cancer, pancreatic ductal adenocarcinoma, pseudopapillary neoplasm, acinar cell carcinoma, prostate cancer, prostate adenocarcinoma, skin cancer, melanoma, malignant melanoma, cutaneous melanoma, small intestine cancer, gastric cancer, gastrointestinal stromal tumor (GIST), uterine cancer, or uterine sarcoma.

[0236] Customized therapy and associated procedures

[0237] In some embodiments, the methods disclosed herein relate to identifying and administering customized therapies for a patient having a given modified TMB. Essentially any cancer therapy (e.g., surgery, radiotherapy, chemotherapy, and / or similar) is included as part of these methods. Generally, a customized therapy includes at least one immunotherapy (or immunotherapy agent). Immunotherapy generally refers to a method of enhancing the immune response to a given type of cancer. In certain embodiments, immunotherapy refers to a method of enhancing the T-cell response to a tumor or cancer.

[0238] In some embodiments, immunotherapy or immunotherapeutic agents target immune checkpoint molecules. Certain tumors can evade the immune system by selecting immune checkpoint pathways. Therefore, targeting immune checkpoints has emerged as an effective method to block the ability of tumors to evade the immune system and activate anti-tumor immunity against certain cancers. Pardoll, Nature Reviews Cancer, 2012, 12: 252-264.

[0239] In certain embodiments, an immune checkpoint molecule is an inhibitory molecule that reduces the signaling involved in the T cell response to an antigen. For example, CTLA4 is expressed on T cells and plays a role in downregulating T cell activation by binding to CD80 (also known as B7.1) or CD86 (also known as B7.2) on antigen-presenting cells. PD-1 is another inhibitory checkpoint molecule expressed on T cells. PD-1 limits T cell activity in peripheral tissues during inflammatory responses. Furthermore, ligands of PD-1 (PD-L1 or PD-L2) are generally upregulated on the surface of many different tumors, resulting in downregulation of the anti-tumor immune response in the tumor microenvironment. In certain embodiments, the inhibitory immune checkpoint molecule is CTLA4 or PD-1. In other embodiments, the inhibitory immune checkpoint molecule is a ligand of PD-1, such as PD-L1 or PD-L2. In other embodiments, the inhibitory immune checkpoint molecule is a ligand of CTLA4, such as CD80 or CD86. In other embodiments, the inhibitory immune checkpoint molecule is lymphocyte activation gene 3 (LAG3), killer cell immunoglobulin-like receptor (KIR), T cell membrane protein 3 (TIM3), galectin 9 (GAL9), or adenosine A2a receptor (A2aR).

[0240] Antagonists targeting these immune checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Therefore, in certain embodiments, the immunotherapy or immunotherapy agent is an antagonist of an inhibitory immune checkpoint molecule. In certain embodiments, the inhibitory immune checkpoint molecule is PD-1. In certain embodiments, the inhibitory immune checkpoint molecule is PD-L1. In certain embodiments, the antagonist of the inhibitory immune checkpoint molecule is an antibody (e.g., a monoclonal antibody). In certain embodiments, the antibody or monoclonal antibody is an anti-CTLA4 antibody, an anti-PD-1 antibody, an anti-PD-L1 antibody, or an anti-PD-L2 antibody. In certain embodiments, the antibody is a monoclonal anti-PD-1 antibody. In some embodiments, the antibody is a monoclonal anti-PD-L1 antibody. In certain embodiments, the monoclonal antibody is a combination of an anti-CTLA4 antibody and an anti-PD-1 antibody, a combination of an anti-CTLA4 antibody and an anti-PD-L1 antibody, or a combination of an anti-PD-L1 antibody and an anti-PD-1 antibody. In certain embodiments, the anti-PD-1 antibody is one or more of pembrolizumab (Keytruda®) or nivolumab (Opdivo®). In certain embodiments, the anti-CTLA4 antibody is ipilimumab (Yervoy®). In certain embodiments, the anti-PD-L1 antibody is one or more of atezolizumab (Tecentriq®), avelumab (Bavencio®), or durvalumab (Imfinzi®).

[0241] In certain embodiments, the immunotherapy or immunotherapeutic agent is an antagonist (e.g., an antibody) to CD80, CD86, LAG3, KIR, TIM3, GAL9, TIGIT, or A2aR. In other embodiments, the antagonist is a soluble version of the inhibitory immune checkpoint molecule, such as a soluble fusion protein containing the extracellular domain of the inhibitory immune checkpoint molecule and the Fc domain of the antibody. In certain embodiments, the soluble fusion protein contains the extracellular domain of CTLA4, PD-1, PD-L1, or PD-L2. In some embodiments, the soluble fusion protein contains the extracellular domain of CD80, CD86, LAG3, KIR, TIM3, GAL9, or A2aR. In one embodiment, the soluble fusion protein contains the extracellular domain of PD-L2 or LAG3.

[0242] In certain embodiments, the immune checkpoint molecule is a costimulatory molecule that amplifies signals involved in the T cell response to an antigen. For example, CD28 is a costimulatory receptor expressed on T cells. When a T cell binds to an antigen through its T cell receptor, CD28 binds to CD80 (also known as B7.1) or CD86 (also known as B7.2) on the antigen-presenting cell, amplifying T cell receptor signaling and promoting T cell activation. Since CD28 binds to the same ligands (CD80 and CD86) as CTLA4, CTLA4 can counteract or regulate the costimulatory signaling mediated by CD28. In certain embodiments, the immune checkpoint molecule is a costimulatory molecule selected from CD28, an inducible T cell costimulator (ICOS), CD137, OX40, or CD27. In other embodiments, the immune checkpoint molecule is a ligand for a co-stimulatory molecule, including, for example, CD80, CD86, B7RP1, B7-H3, B7-H4, CD137L, OX40L, or CD70.

[0243] Agonists targeting these co-stimulatory checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Therefore, in certain embodiments, the immunotherapy or immunotherapeutic agent is a co-stimulatory checkpoint molecule agonist. In certain embodiments, the co-stimulatory checkpoint molecule agonist is an agonist antibody, preferably a monoclonal antibody. In certain embodiments, the agonist antibody or monoclonal antibody is an anti-CD28 antibody. In other embodiments, the agonist antibody or monoclonal antibody is an anti-ICOS antibody, anti-CD137 antibody, anti-OX40 antibody, or anti-CD27 antibody. In other embodiments, the agonist antibody or monoclonal antibody is an anti-CD80 antibody, anti-CD86 antibody, anti-B7RP1 antibody, anti-B7-H3 antibody, anti-B7-H4 antibody, anti-CD137L antibody, anti-OX40L antibody, or anti-CD70 antibody.

[0244] In certain embodiments, the customized therapies described herein are generally administered parenterally (e.g., intravenously or subcutaneously). Pharmaceutical compositions containing immunotherapeutic agents are generally administered intravenously. Certain therapeutic agents are administered orally. However, customized therapies (e.g., immunotherapeutic agents) may also be administered by any method known in the art, including, for example, buccal, sublingual, rectal, vaginal, urethral, ​​ophthalmic, intraocular, intranasal, and / or intraauricular, and these administrations may include tablets, capsules, granules, aqueous suspensions, gels, sprays, suppositories, ointments, and the like.

[0245] System and computer-readable media

[0246] This disclosure also provides various systems and computer program products or machine-readable media. In some embodiments, for example, the methods described herein are implemented or facilitated, at least in part, using systems, distributed computing hardware and applications (e.g., cloud computing services), electronic communication networks, communication interfaces, computer program products, machine-readable media, electronic storage media, software (e.g., machine-executable code or logical instructions) and / or similar, as necessary. For illustrative purposes, Figure 2 shows a schematic diagram of an exemplary system suitable for use in the execution of at least the embodiments of the methods disclosed herein. As shown, the system 200 includes at least one controller or computer, e.g., a server 202 (e.g., a search engine server) including a processor 204 and a storage device, storage device, or storage device component 206, as well as one or more other communication devices 214 and 216 (e.g., client-side computer terminals, telephones, tablets, laptops, and other mobile devices) located remotely from the remote server 202 and communicating with the remote server 202 through an electronic communication network 212 such as the Internet or other internetworks. Communication devices 214 and 216 generally include, for example, an electronic display (e.g., an Internet-connected computer) that communicates with a server 202 computer through a network 212, and the electronic display includes a user interface (e.g., a graphical user interface (GUI), a web-based user interface, and / or similar) for displaying the results when the method described herein is performed. In certain embodiments, the communication network also includes the physical transfer of data from one location to another using, for example, a hard drive, a thumb drive, or other data storage mechanism.System 200 also includes a program product 208 stored on a machine-readable medium, such as a computer or one or more of various types of storage devices, such as storage device 206 of Server 202, which is readable by Server 202, for example, to facilitate a guided search application or other executable by one or more other communication devices, such as 214 (figuratively shown as a desktop or personal computer) and 216 (figuratively shown as a tablet computer). In some embodiments, System 200 also includes at least one database server, such as Server 210 (e.g., control sample or comparative result data pointing to a customized therapy), which is associated with an online website storing data that is searchable either directly or through the search engine server 202, for example. System 200 also includes one or more other servers, which are optionally located away from Server 202, each of which is optionally associated with one or more database servers 210 that are optionally located away from each of the other servers or are local to them. The other servers can benefit geographically dispersed users and enhance geographically distributed operations.

[0247] As will be understood by those skilled in the art, the storage device 206 of server 202 may include, as necessary, volatile and / or non-volatile storage devices, including, for example, RAM, ROM, and magnetic or optical disks. Although a single server is illustrated, it will be understood by those skilled in the art that the configuration of server 202 illustrated is merely illustrative, and other types of servers or computers configured according to various other methods or architectures may also be used. Server 202 schematically shown in Figure 2 represents a server or server cluster or server farm, and is not limited to any individual physical server. A server site may be located as a server farm or server cluster managed by a server hosting provider. The number of servers and their architectures and configurations may be increased based on usage, demand, and capacity requirements for system 200. As will also be understood by those skilled in the art, in these embodiments, other user communication devices 214 and 216 may be, for example, laptops, desktops, tablets, personal digital assistants (PDAs), mobile phones, servers, or other types of computers. As is known and understood to those skilled in the art, network 212 may include the Internet, intranet, telecommunications network, extranet, or World Wide Web, which consists of multiple computers / servers communicating with one or more other computers through a communications network and / or a portion of a local or other area network.

[0248] As will be further understood by those skilled in the art, the exemplary program product or machine-readable medium 208 may take the form of microcode, programs, cloud computing formats, routines, and / or symbolic languages ​​that, if necessary, provide one or more sets of ordered operations that control and direct the functions of the hardware. The program product 208 in the exemplary embodiment does not necessarily have to reside entirely in a volatile storage device, and may, if necessary, be selectively loaded according to various systems of methods known and understood by those skilled in the art.

[0249] As will be further understood by those skilled in the art, the terms “computer-readable medium” or “machine-readable medium” refer to any medium involved in presenting instructions for execution to a processor. For illustrative purposes, the terms “computer-readable medium” or “machine-readable medium” include, for example, distributed media, cloud computing formats, intermediate storage media, computer execution memory devices, and any other media or devices that can store program products 608 that perform the functionality or processes of various embodiments of this disclosure for computer reading. “Computer-readable medium” or “machine-readable medium” can take many forms, including, but are not limited to, non-volatile media, volatile media, and transmission media. Examples of non-volatile media include optical disks or magnetic disks. Examples of volatile media include dynamic memory, such as the main memory of a given system. Examples of transmission media include coaxial cables, copper wires, and optical fibers, including wires including buses. Transmission media can also take the form of sound waves or light waves, such as those generated between radio and infrared data communications. Examples of computer-readable media include floppy disks, flexible disks, hard disks, magnetic tapes, flash drives, or any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tapes, holes, RAM, PROMs, and any other physical media having EPROM patterns, FLASH®-EPROMs, any other memory chips or cartridges, carriers, or any other media that a computer can read.

[0250] The program product 208 is copied, if necessary, from a computer-readable medium to a hard disk or similar intermediate storage medium. When the program product 208 or a portion thereof is implemented, it is loaded, if necessary, into the executable storage device of one or more computers from their distributed media, their intermediate storage medium, etc., which are configured to operate in accordance with the functionality or methods of various embodiments. All such operations are, for example, well known to those skilled in the art of computer systems.

[0251] For further illustrative purposes, in a particular embodiment, the application provides a system comprising one or more processors and one or more storage device components communicating with the processors. The storage device components generally include one or more instructions that, when executed, cause the processor to provide information that triggers the display of at least one mutation count, adjusted results / TMB, comparison results, customized therapies, and / or similar (e.g., via communication devices 214, 216, etc.), and / or receive information from other system components and / or system users (e.g., via communication devices 214, 216, etc.).

[0252] In some embodiments, when the program product 208 is executed by the electronic processor 204, at least: (i) determining an observed mutation count from sequence information obtained from one or more nucleic acids in a sample from a subject; (ii) determining the tumor fraction and / or coverage of the nucleic acids to generate sequencing parameters; (iii) determining a predicted mutation fraction and / or a predicted distribution of the predicted mutation fraction considering the sequencing parameters to generate a prediction result; (iv) adjusting the observed mutation count considering the prediction result to generate an adjusted result, thereby detecting the tumor mutational burden (TMB) in the subject; and optionally, (v) comparing the adjusted result with one or more comparison results, wherein a substantial match between the adjusted result and the comparison results indicates an expected response to a therapy for the subject. The non-transitory computer-executable instructions include performing the steps.

[0253] System 200 generally also includes additional system components configured to implement various aspects of the methods described herein. In some of these embodiments, one or more of these additional system components are located remotely from the remote server 202 and communicate with the remote server 202 through the electronic communication network 212, while in other embodiments, one or more of these additional system components are located locally and communicate with the server 202 (i.e., in the absence of the electronic communication network 212) or directly communicate with, for example, the desktop computer 214.

[0254] In some embodiments, for example, additional system components include a sample preparation component 218 operably connected to the controller 202 (either directly or indirectly (e.g., via an electronic communication network 212)). The sample preparation component 218 is configured to prepare nucleic acids in a sample to be amplified and / or sequenced by a nucleic acid amplification component (e.g., a thermal cycler, etc.) and / or a nucleic acid sequencer (e.g., prepare a library of nucleic acids). In certain ones of these embodiments, the sample preparation component 218 is configured to isolate nucleic acids in the sample from other components, attach one or more adapters containing molecular barcodes to the nucleic acids as described herein, selectively enrich one or more regions from a genome or transcriptome prior to sequencing, and / or perform similar things.

[0255] In certain embodiments, the system 200 also includes a nucleic acid amplification component 220 (e.g., a thermal cycler, etc.) operably connected to the controller 202 (either directly or indirectly (e.g., via an electronic communication network 212)). The nucleic acid amplification component 220 is configured to amplify nucleic acids in a sample derived from a subject. For example, the nucleic acid amplification component 220 is configured to amplify, as needed, regions selectively enriched from a genome or transcriptome in a sample as described herein.

[0256] System 200 also generally includes at least one nucleic acid sequencer 222 operably connected to the controller 202 (directly or indirectly, e.g., via the electronic communication network 212). The nucleic acid sequencer 222 is configured to yield sequence information from nucleic acids (e.g., amplified nucleic acids) in a sample of interest. Essentially any type of nucleic acid sequencer can be adapted for use in these systems. For example, the nucleic acid sequencer 222 is configured to generate sequencing reads by performing pyrosequencing, single-molecule sequencing, nanopore sequencing, semiconductor sequencing, synthesis sequencing, ligation sequencing, hybridization sequencing, or other techniques on the nucleic acids, as needed. If necessary, the nucleic acid sequencer 222 is configured to group sequence reads into families, each family containing sequence reads generated from nucleic acids in a given sample. In some embodiments, the nucleic acid sequencer 222 uses a clone single-molecule array derived from a sequencing library to generate sequencing reads. In a particular embodiment, the nucleic acid sequencer 222 includes at least one chip having an array of microwells for sequencing a sequencing library and generating sequencing reads.

[0257] To facilitate full or partial system automation, the system 200 also generally includes a material transfer component 224 operably connected to the controller 202 (directly or indirectly, e.g., via the electronic communication network 212). The material transfer component 224 is configured to transfer one or more materials (e.g., nucleic acid samples, amplicons, reagents, and / or similar) to and from the nucleic acid sequencer 222, sample preparation component 218, and nucleic acid amplification component 220. Additional details regarding computer systems and networks, databases, and computer program products can be found, for example, in Peterson, Computer Networks: A Systems Approach, Morgan Kaufmann, 5th Ed. (2011), and Kurose, Computer Networking: A Top-Down Approach, Pearson, 7 th Ed. (2016), Elmasri, Fundamentals of Database Systems, Addison Wesley, 6th Ed. (2010), Coronel, Database Systems: Design, Implementation, & Management, Cengage Learning, 11 th Ed. (2014), Tucker, Programming Languages, McGraw-Hill Science / Engineering / Math, 2nd Ed. (2006), and Rhoton, Cloud Computing Architected: Solution Design. This information is also available in the Handbook, Recursive Press (2011), and both are incorporated herein by reference in their entirety.

[0258] Sequencing of the immune repertoire

[0259] In addition to the TMB analysis described herein, this application further provides a method for sequencing the immune repertoire. Immune receptors (T cell receptors (TCRs) for T cells and immunoglobulins (Ig) for B cells) are the unique "barcodes" of these lymphocytes (in healthy humans there are 10 types of receptors). 6 (Across species). When the immune system recognizes a pathogen (i.e., vaccines, pathogens, and novel antigens derived from cancer mutations), it triggers an immune response that allows for the proliferation of antigen-specific T cells into clones. Sequencing the uniqueness of immune receptors such as TCRs makes it possible to track the immune response in patient samples. In certain embodiments, profiling the immune repertoire allows for complementary analysis to TMB analysis. High TMB scores are likely to correlate with a patient's response to immunotherapy (anti-PD1, anti-PDL1, anti-CTLA4, etc.), while measuring the immune repertoire is an actual functional outcome in response to the high presence of novel antigens present in cancer. Furthermore, since some cells, such as leukemia cells and other cancer cells, may have distinct receptor clonal signatures, sequencing the immune repertoire allows for the identification of hematological malignancies. This can also be extended to minimal residual disease (MRD) studies for hematological malignancies by tracking the receptors of the original cancer cells.

[0260] For further illustration, Figure 3 presents a flowchart illustrating the steps of an exemplary method for performing sequencing of extracellular free nucleic acids (using TMB analysis) in combination with sequencing of an immunorepertory derived from the same sample, according to certain embodiments of the present disclosure. As shown, Method 300 includes the step of obtaining a test sample (e.g., a blood sample) from the subject. The test sample is subjected to an isolation or extraction step to produce a plasma fraction containing extracellular free nucleic acids and a buffy coat fraction containing lymphocytes. In the embodiments shown, ctDNA in the plasma sample portion is subjected to a ProK-based extraction step as part of a process to identify mutations or variants in the ctDNA. Optionally, the method includes a targeted panel and a methyl-binding domain (MBD) splitting step. Additional details regarding the analysis of epigenetic modifications, which may be adapted as necessary for use in the implementation of the methods disclosed herein, are described, for example, in WO2018 / 119452, application filed December 22, 2017, incorporated by reference. As shown, ctDNA analysis also includes various library preparation steps (end repair, ligation, polymerase chain reaction (PCR) amplification, etc.) and enrichment steps. Method 300 includes a TCR clonal typing assay that includes a genomic DNA (gDNA) extraction step and library preparation steps (e.g., TCR barcoding, purification, multiplexed nested PCR, and further purification and quantification steps). As also shown, exemplary Method 300 includes an immune receptor discovery assay / hematological malignancy detection and minimal residual disease (MRD) assay that includes, as optional, RNA extraction, reverse transcription and template switching, and purification steps together with various PCR and purification steps, in addition to or instead of the TCR clonal typing assay. The products of the various assays of Method 300 are pooled, sequenced, and further analyzed by the bioinformatics pipeline (BIP) described herein. Further descriptions of sample preparation methods, including the PCR step, are shown in Figures 4A and 4B for the TCR clonal typing assay and the immune receptor discovery assay, respectively.

[0261] For further illustration, Figure 5 schematically shows Method 500, which includes analysis of the buffy coat and plasma portion of a blood sample. As shown, buffy coat analysis reveals the immune repertoire and other immunogenes of clonal hematopoiesis (e.g., clonal hematopoiesis of indeterminate potential). The assay includes evaluating for potential (or CHIP) mutations, which are further evaluated by the bioinformatics pipeline described herein. The assay includes evaluating the clonal types of T cell receptors (TCRs) and B cell receptors (BCRs), as well as evaluating clonal augmentation or novel antigen receptors. With respect to the plasma portion of the sample, the ctDNA library is evaluated using the bioinformatics pipeline described herein, which includes CHIP-related information obtained from the buffy coat sample proportion of the sample. As part of the evaluation of the plasma sample proportion, single nucleotide variants (SNVs), microsatellite instability (MSI), and fusions are evaluated to generate a TMB score. As also shown, the data collected between the two shown assay pathways can be used to correlate mutations in order to identify novel antigen orphan receptors. As also shown, the results of the two shown assay pathways can also be used to generate an improved score for identifying a potential response to immunotherapy.

[0262] In some embodiments, for example, the buffy coat is subjected to genomic DNA isolation using a standard kit (e.g., commercially available from Qiagen or other suppliers), followed by two nested PCRs to enrich amplicons targeting TCR alpha and beta subunits outside highly variable regions such as CDR3. The second of the two PCRs requires primers that not only amplify the TCR gDNA but also attach partial adapters for sequencing (e.g., SP5 and SP7 adapters). These partial adapters are used in the final third "index" PCR of this exemplary embodiment to attach full-length P5 and P7 oligos to the desired TCR amplicons, thereby creating a library that can be sequenced, for example, on an Illumina sequencer. [Examples]

[0263] (Example 1) Adjustment of TMB score in samples with a small tumor percentage and low coverage.

[0264] A certain volume of patient samples with a high known tumor proportion and high coverage was diluted 3-4 times with non-tumor cfDNA (i.e., normal cfDNA) to obtain samples with a low tumor proportion and low coverage. The diluted samples were processed and analyzed using a blood-based DNA assay developed by Guardant Health, Inc. (Redwood City, CA). The mutation count was estimated by bioinformatics analysis, and the corrected TMB score was determined by applying a TMB correction model. The observed mutation count, max MAF, and coverage of the diluted samples were provided as input parameters. This model allowed for the determination of the predicted mutation fraction (f) and the upper limit of the 95% confidence level of the predicted mutation fraction (f). 上限), and the adjusted mutation count and TMB score were estimated. As shown in Table 3, the TMB score of the diluted samples reported by the model (23.5 mutations (mut) per megabase Mb) is very close to the TMB score of the original sample with a high tumor percentage (23.2 mutations per Mb). Figure 6 shows the TMB score plots for diluted samples with and without the TMB-corrected model applied (o). [Table 3]

[0265] (Example 2) Landscape and genomic correlation of tumor gene mutations based on ctDNA across 6 solid tumor types

[0266] Introduction

[0267] Tumor mutational load (TMB) is a predictive biomarker of response to immune checkpoint inhibitor (ICI) therapy. Current panel-based TMB algorithms collect signals from specific types of somatic variants (e.g., non-synonymous coding SNVs). Since many TMB-high patients do not respond to ICIs, we investigated whether additional variant types and other genomic correlations could refine TMB calculations. Furthermore, while plasma TMB yields a better yield of reportable TMB results compared to tissue, TMB in low tumor DNA shedders tends to be underestimated. In this embodiment, we improve the TMB algorithm using additional genomic features and adjustments for low DNA shedding by testing thousands of samples from numerous cancer types using a highly sensitive 500-gene cfDNA sequencing platform (large panel assay).

[0268] method

[0269] A cfDNA-based TMB algorithm was developed that is robust to varying tumor shedding levels. cfDNA-based TMB was evaluated in over 1,000 plasma samples across six solid tumor types, including lung, colorectal, and prostate. The contribution of silent SNVs and indels to the TMB score was investigated. The correlation between TMB, tumor type, patient ethnicity, and molecular subtypes of lung tumors was investigated. Finally, the landscape of TMB as well as additional genomic features: subclonality, chromosomal instability, and microsatellite instability (MSI) were also investigated. Additional details regarding the large panel assay are presented in Table 4. Furthermore, the TMB workflow used in this example is schematically shown in Figure 7.

Table 4

[0270] Results

[0271] 1. Performance of the large panel assay

[0272] Table 5 presents a summary of the analytical validation performance and details of the large panel assay based on 30 ng of cfDNA input.

Table 5

[0273] Somatic / germline lineage status was determined using a beta-binomial statistical model of deviation from the proportion of local germline variant alleles. This is further described, for example, in Nance et al. (2018) A novel approach to differentiate somatic vs. germline variants in liquid biopsies using a betabinomial model. AACR, Poster 4272, which is incorporated by reference. This method did not rely on a database of common germline variants (e.g., dbSNP).

[0274] 2. Panel-based TMB components

[0275] Figures 8A, B, and C show plots of variants correlated with non-synonymous coded SNVs across the cohort: (Figure 8A) synonymous SNVs, (Figure 8B) indels, and (Figure 8C) intronic SNVs (Pearson r = 0.90, 0.71, 0.89). The variability of mutation counts (box plots) was consistent with predictions based on sample-specific mutation rates (black lines: thick: median, thin: IQR).

[0276] 3. Adjusted plasma TMB is largely independent of input.

[0277] Figures 9A–9D are plots showing that large-panel assay tumor shedding correction removed the dependence of mutation count on tumor shedding (Figure 9A) and input cfDNA (Figure 9B), resulting in plasma tumor mutational burden (pTMB) that is largely independent of these input metrics (Figures 9C and 9D). Figures 9A and 9B show the mutation count distribution across tumor shedding bins (estimated by maximum somatic MAF (max MAF)) or input volume (resulting in molecular coverage). Figures 9C and 9D show the corrected TMB scores across the same bins. Violin plots show median and interquartile range (IQR). The black line shows the trend of the median.

[0278] 4. TMB across tumor types

[0279] Figures 10A and 10B are plots showing the TMB distribution across cohorts and between tumor types: (Figure 10A) Across cohorts, TMB scores had a long-tailed distribution, which is consistent with previous observations in tissue-based and plasma-based TMB, with a median of 10 mut / Mb across tumor types and an upper tertile of 14 mut / Mb (black line). Figure 10B shows that the median TMB varied among colorectal cancer, lung cancer, and prostate cancer samples with a trend consistent with previous tissue-based observations.

[0280] 5. TMB is not dependent on ethnicity.

[0281] Figures 11A and 11B are plots showing that median TMB does not vary between ethnicities. Principal component analysis (PCA) of common germline variants was used to cluster the samples and define race-based clusters. Clusters were labeled using sets of samples with known race (Asian / Pacific Islander: A; Black / African American: B; White: W). For each race-based cluster, 23 tumor-type matched samples were randomly selected. Figure 11A shows the PCA clustering, and Figure 11B shows the TMB scores. Contrary to some tissue TMB pipelines, median TMB was constant across clusters (using comparisons based on non-parametric sampling of medians, A vs. W: p=0.78, B vs. W: 0.77, A vs. B: 0.99).

[0282] 6. Correlation between TMB and oncogenic mutations

[0283] Figure 12 is a plot showing that TMB (i.e., plasma-based TMB (pTMB)) in lung samples differs depending on the driver status. Among lung samples, median TMB was lower in EGFR-driven and ALK-driven lung tumors (median 8 mut / Mb, p<0.01 for median equivalence) and higher in tumors with KRAS or PIK3CA hotspot mutations (14 mut / Mb, p=0.04) compared to all lung tumors (12 mut / Mb), and higher in tumors with KRAS or PIK3CA hotspot mutations. Lung tumors with BRAF drivers had a median TMB similar to all lung tumors (11 mut / Mb, p=0.25). Loss-of-function (LOF) mutations in STK11, KEAP1, or PTEN are putative negative predictors of ICI response. Median TMB was slightly higher in tumors with these mutations than in all lung tumors (15 mut / Mb, p<0.01), suggesting that these latter events may be clinical biomarkers independent of TMB.

[0284] 7. The TMB landscape exhibits fluctuating clonal structure, chromosomal instability, and MSI status.

[0285] Figure 13A is a plot showing that the clonality and chromosomal instability of somatic mutations varied greatly across the TMB landscape: various subclonal mutation fractions existed across the range of TMB scores observed in the cohort. This is illustrated by 100 randomly selected samples across tumor types, ordered by TMB score, where black represents clonal mutations (MAF is ≥10% of the sample's max MAF) and light gray represents subclonal mutations (defined as MAF <10% of the sample's max MAF). The percentage of all mutations that were subclonal is shown in the bars below (black = highly clonal; light gray = highly subclonal). The subclonal fraction did not correlate with the TMB score. Similarly, chromosomal instability, measured either by the number of genes in which amplification was detected (#CNV: shown below the subclonal fraction) or the percentage of the panel space that was diploid (diploid fraction: shown below (#CNV)), also did not correlate with the TMB score. Figure 13B shows that MSI-High was detected in a subset of TMB-High samples.

[0286] conclusion

[0287] Panel-based TMB scores can influence syntactic and non-coding mutations to enhance the signaling of mutational loadings across the exome. As more patient outcome data becomes available, TMB algorithms and orthogonal biomarkers of tumor genomic immunogenicity will further evolve to improve guidance regarding patients' responses to immunotherapy. Highly sensitive panel sequencing for TMB across a large panel space, as well as the ability to detect copy number variations and MSI status, will be crucial for the development and clinical application of biomarkers.

[0288] (Example 3) Adjustment of TMB scores in samples with low tumor percentage and low coverage using a TMB correction model.

[0289] Patient samples were processed and analyzed using a blood-based DNA assay developed by Guardant Health, Inc. (Redwood City, CA). Bioinformatics analysis identified 13 somatic mutations in these samples. In this embodiment, only SNVs and indels were considered as somatic mutations. In this embodiment, three of the 13 somatic mutations were excluded from the mutation count because two of these three mutations were non-tumor-related mutations (e.g., clonal hematopoietic mutations), and the other mutations were driver mutations. The mutation count, max MAF, and coverage of the samples were provided as input parameters to a TMB-corrected model. This model allowed for the prediction mutation fraction (f) and the upper limit of the 95% confidence level of the prediction mutation fraction (f). 上限 The adjusted mutation count and TMB score were estimated. Here, the TMB score was determined by dividing the adjusted mutation count by the product of the exome calibrator and the size of the genomic region being analyzed (approximately 1 Mb in this example). Table 6 shows the TMB scores reported from the TMB-corrected model. [Table 6]

[0290] (Example 4) Samples that have not been evaluated using the TMB-corrected model (small tumor proportion)

[0291] Patient samples were processed and analyzed using a blood-based DNA assay developed by Guardant Health, Inc. (Redwood City, CA). Bioinformatics analysis identified one somatic mutation in this sample. In this embodiment, only SNVs and indels were considered as somatic mutations. The maximum (max) MAF of this sample was determined to be 0.1%. In this embodiment, the maximum (max) MAF was considered as the tumor percentage. The tumor percentage of this sample was below the tumor percentage cutoff. Therefore, evaluation using the TMB-corrected model was not performed for this sample. The data for this example are further summarized in Table 7. [Table 7]

[0292] While the foregoing disclosure includes some details as examples and illustrations for clarity and understanding, it will be apparent to those skilled in the art that various variations in form and detail can be made without departing from the true scope of this disclosure and can be implemented within the scope of the appended claims. For example, all features, steps, elements, or other embodiments of methods, systems, computer-readable media, and / or components can be used in various combinations.

[0293] All patents, patent applications, websites, other publications or documents, accession numbers, etc., cited herein are incorporated by reference in whole for any purpose, just as each individual item is specifically and individually indicated to be incorporated by reference. If different versions of an arrangement are associated with an accession number at different times, the version associated with the accession number on the effective filing date of this application is intended. The effective filing date means prior to the actual filing date or, where applicable, the filing date of the priority application referencing the accession number. Similarly, if different versions of a publication, website, etc., are published at different times, unless otherwise specified, the version published most recently on the effective filing date of this application is intended. The present invention provides, for example, the following items: (Item 1) A method for determining tumor gene mutational load (TMB) in a subject, (a) A step of determining the observed mutation count from sequence information obtained from one or more nucleic acids in the sample derived from the subject, (b) A step of determining the tumor percentage and / or coverage of the nucleic acid and generating sequencing parameters, (c) A step of determining the predicted mutation rate and / or the predicted distribution of the predicted mutation rate, taking into account the sequencing parameters, and generating a prediction result, (d) Adjusting the observed mutation count in consideration of the prediction results to generate an adjusted result, thereby determining the TMB in the subject. A method that includes this. (Item 2) A method for determining tumor gene mutational load (TMB) in a subject, (a) A step of providing a sample derived from the subject, (b) A step of amplifying the nucleic acid in the sample to produce amplified nucleic acid, (c) A step of sequencing the amplified nucleic acid to generate sequence information, (d) A step of determining the observed mutation count from the sequence information, (e) A step of determining the tumor percentage and / or coverage of the nucleic acid and generating sequencing parameters, (f) A step of determining the predicted mutation rate and / or the predicted distribution of the predicted mutation rate, taking into account the sequencing parameters, and generating a prediction result; (g) Adjusting the observed mutation count in consideration of the prediction results to generate an adjusted result, thereby determining the TMB in the subject. A method that includes this. (Item 3) A method for selecting one or more customized therapies to treat cancer in a subject, (a) A step of determining the observed mutation count from sequence information obtained from one or more nucleic acids in the sample derived from the subject, (b) A step of determining the tumor percentage and / or coverage of the nucleic acid and generating sequencing parameters, (c) A step of determining the predicted mutation rate and / or the predicted distribution of the predicted mutation rate, taking into account the sequencing parameters, and generating a prediction result, (d) A step of adjusting the observed mutation count in consideration of the prediction results to generate the adjusted result, (e) The step of identifying one or more customized therapies for the subject by comparing the adjusted results with one or more comparative results indicating one or more therapies. A method that includes this. (Item 4) A method for treating cancer in a subject, (a) A step of determining the observed mutation count from sequence information obtained from one or more nucleic acids in the sample derived from the subject, (b) A step of determining the tumor percentage and / or coverage of the nucleic acid and generating sequencing parameters, (c) A step of determining the predicted mutation rate and / or the predicted distribution of the predicted mutation rate, taking into account the sequencing parameters, and generating a prediction result, (d) A step of adjusting the observed mutation count in consideration of the prediction results to generate the adjusted result, (e) The step of identifying one or more customized therapies for the subject by comparing the adjusted results with one or more comparative results indicating one or more therapies, (f) If the adjusted result and the comparative result substantially match, administer at least one of the identified customized therapies to the subject to treat the cancer in the subject. A method that includes this. (Item 5) A method for treating cancer in a subject, comprising administering one or more customized therapies to the subject to thereby treat the cancer in the subject, wherein the customized therapy is (a) A step of determining the observed mutation count from sequence information obtained from one or more nucleic acids in the sample derived from the subject, (b) A step of determining the tumor percentage and / or coverage of the nucleic acid and generating sequencing parameters, (c) A step of determining the predicted mutation rate and / or the predicted distribution of the predicted mutation rate, taking into account the sequencing parameters, and generating a prediction result, (d) A step of adjusting the observed mutation count in consideration of the prediction results to generate the adjusted result, (e) The step of comparing the adjusted result with one or more comparative results indicating one or more therapies, (f) If the adjusted result and the comparative result substantially match, the step of identifying one or more customized therapies for the subject. A method identified by (Item 6) The method according to any one of the items, wherein the observed mutation count and / or the tumor percentage includes a plurality of synonymous mutations, a plurality of non-synonymous mutations, and / or a plurality of non-coding mutations identified in the nucleic acid. (Item 7) The method according to any one of the above items, wherein the observed mutation count and / or the tumor percentage includes multiple mutations selected from the group consisting of single nucleotide variants (SNVs), insertions or deletions (indels), copy number variants (CNVs), fusions, translocations, frameshifts, duplications, repeat extensions, and epigenetic variants. (Item 8) The method according to any one of the items, wherein driver mutations and / or clonal hematopoietic mutations are excluded from the observed mutation count and / or the tumor percentage. (Item 9) The method according to any one of the preceding items, comprising using pooled evidence of one or more possible mutations below the detection limit with respect to a given single nucleotide variant (SNV) or a given insertion or deletion (indel) in order to determine the observed mutation count. (Item 10) The method according to any one of the above items, comprising using the predicted mutation rate as the observed mutation rate in the actual mutation count. (Item 11) The method according to any one of the items, wherein the observed mutation count and / or the tumor percentage include a plurality of somatic mutations identified in the nucleic acid. (Item 12) The method according to item 11, wherein one or more known cancer driver and / or passenger mutations are excluded from the observed mutation count. (Item 13) The method according to any one of the items, comprising comparing the sequence information with one or more reference sequences to identify the observed mutation count. (Item 14) The method according to item 13, wherein the reference sequence includes at least a subsequence of hg19 and / or hg38. (Item 15) The method according to any one of the above items, wherein the tumor percentage includes the maximum mutant allele percentage (MAF) of all somatic mutations identified in the nucleic acid. (Item 16) The method according to any one of the above items, wherein the tumor percentage is less than approximately 0.05%, less than approximately 0.1%, less than approximately 0.2%, less than approximately 0.5%, less than approximately 1%, less than approximately 2%, less than approximately 3%, less than approximately 4%, or less than approximately 5% of all nucleic acids in the sample. (Item 17) The method according to any one of the above items, comprising identifying a plurality of unique cfDNA fragments containing a given nucleotide position in the nucleic acid in order to determine the coverage. (Item 18) The method according to any one of the items, comprising identifying the median number of unique extracellular free DNA (cfDNA) molecules constituting a given nucleotide position in the nucleic acid in order to determine the coverage. (Item 19) The method according to any one of the above items, wherein the coverage is a cfDNA fragment between 10 and 50,000 nucleotides at a given nucleotide position in the nucleic acid present in the sample. (Item 20) The method according to any one of the items, wherein the predicted variance rate and / or the predicted distribution of the predicted variance rate includes a confidence interval of about 95% or greater with respect to the variance rate. (Item 21) The method of any one of the items, comprising generating a lower limit for the observed mutation count using the upper limit of the 95% confidence interval for the predicted mutation rate. (Item 22) The method according to any one of the above items, comprising determining the predicted mutation ratio by calculating the probability that a mutation in a given mutation ratio (MAF) is identified across the distribution of the predicted MAF. (Item 23) The method according to item 22, comprising generating MAF by multiplying the predicted relative MAF distribution by the tumor percentage. (Item 24) The distribution of the predicted MAF is calculated using the following confidence intervals for the binomial proportions, according to the method described in item 22 or 23:

number

Claims

[Claim 1] The invention described herein.

Images