Methods for Analyzing Molecular Responses and Related Embodiments

Abstract

Provided herein is a method for determining molecular response score.Molecular response score can be used to monitor and guide the administration of treatment to subject.This method can include: determining a first plurality of sequence reads and a second plurality of sequence reads associated with subject, said first plurality of sequence reads being determined before administering therapy, and said second plurality of sequence reads being determined after administering therapy; classifying a plurality of variants in said first plurality of sequence reads and said second plurality of sequence reads as somatic or germline.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Provisional Patent Application No. 63 / 157,592, filed March 5, 2021, and U.S. Provisional Patent Application No. 63 / 173,193, filed April 9, 2021, each of which is incorporated by reference in its entirety for all purposes. [Background technology]

[0002] background Molecular response is the calculation of the change in circulating tumor DNA (ctDNA) levels observed in samples collected from a subject at different time points. In certain cases, the calculation is based on the percentage of somatic variants in the total cell-free DNA (cfDNA) in the sample. In other cases, the calculation is based on the concentration of ctDNA in the sample (i.e., normalized per cfDNA concentration in the feed). A common problem associated with these approaches is that these relatively simple calculations of molecular response often result in erroneous or inaccurate molecular response scores. Thus, there remains a need for methods to accurately determine the molecular response score of subjects with cancer. Summary of the Invention

[0003] A brief summary In one aspect, the disclosure provides a method for determining a molecular response score using, at least in part, a computer. The method includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, the first plurality of sequence reads being determined prior to administering the therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; for at least one variant of the plurality of variants classified as somatic, determining a weighted average of the first mutant allele fraction (MAF) and a weighted average of the second MAF based on the first MAF and the second MAF; determining a ratio of the weighted average of the first MAF to the weighted average of the second MAF for the subject; determining a confidence interval based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF; and outputting the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval as a molecular response score.

[0004] In one embodiment, the present disclosure provides a method for determining molecular response score at least partially using a computer.The method includes: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined before administering a therapy, and the second plurality of sequence reads being determined after administering a therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; for at least one variant of the plurality of variants classified as somatic, determine a mutant allele fraction (MAF) ratio based on a first MAF and a second MAF; determining a weighted average of the MAF ratios for the subject; determining a confidence interval associated with the weighted average of the MAF ratios based on the weighted average of the MAF ratios; and outputting the weighted average of the MAF ratios and the confidence interval as a molecular response score.

[0005] In one aspect, the disclosure provides a method for determining a molecular response score using, at least in part, a computer. The method includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, the first plurality of sequence reads being determined prior to administering the therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying a plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification discrepancy between the first plurality of sequence reads and the second plurality of sequence reads; determining a first mutant allele fraction for at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the first plurality of sequence reads; determining a second mutant allele fraction for at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the second plurality of sequence reads; and determining a molecular response score based on the first mutant allele fraction and the second mutant allele fraction.

[0006] In one embodiment, the present disclosure provides a method for determining a molecular response score at least partially using a computer. The method includes: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined before administering a therapy, and the second plurality of sequence reads being determined after administering a therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining at least one variant of the plurality of variants as a clonal hematopoietic potential of undetermined (CHIP) variant; removing at least one CHIP variant from the plurality of variants; determining a first mutant allele ratio for at least one variant of the plurality of variants classified as somatic based on at least a portion of the first plurality of sequence reads; determining a second mutant allele ratio for at least one variant of the plurality of variants classified as somatic based on at least a portion of the second plurality of sequence reads; and determining a molecular response score based on the first mutant allele ratio and the second mutant allele ratio.

[0007] In one embodiment, the present disclosure provides a method for determining molecular response score by at least partially using a computer.The method includes: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined before administering a therapy, and the second plurality of sequence reads being determined after administering a therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying a plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification discrepancy between the first plurality of sequence reads and the second plurality of sequence reads; determining at least one variant of the plurality of variants as a clonal hematopoietic of undetermined potential (CHIP) variant; removing at least one CHIP variant from the plurality of variants; determining a first mutant allele fraction for at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the first plurality of sequence reads; and determining a first mutant allele fraction for at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the first plurality of sequence reads. determining a second mutant allele fraction for at least one variant of the plurality of variants based on at least a portion of the second plurality of sequence reads; determining a first mutant allele fraction and a second mutant allele fraction for at least one variant of the plurality of variants classified or reclassified as somatic based on a MAF ratio; determining a weighted average of the MAF ratios for the subject; determining a confidence interval associated with the weighted average of the MAF ratios based on the weighted average of the MAF ratios; and outputting the weighted average of the MAF ratios and the confidence interval as a molecular response score.

[0008] In one embodiment, the present disclosure provides a method for determining a molecular response score at least partially using a computer. The method includes: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined before administering a therapy, and the second plurality of sequence reads being determined after administering a therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying a plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification discrepancy between the first plurality of sequence reads and the second plurality of sequence reads; determining at least one variant of the plurality of variants as a clonal hematopoietic potential undetermined (CHIP) variant; removing at least one CHIP variant from the plurality of variants; and classifying at least one variant of the plurality of variants as somatic. determining a first mutant allele fraction (MAF) for the variant based on at least a portion of the first plurality of sequence reads; determining a second MAF for at least one variant of the plurality of variants classified as somatic based on at least a portion of the second plurality of sequence reads; determining a weighted average of the first MAF and a weighted average of the second MAF for at least one variant of the plurality of variants classified as somatic based on the first MAF and the second MAF; determining a ratio of the weighted average of the first MAF to the weighted average of the second MAF for the subject; determining a confidence interval based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF; and outputting the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval as a molecular response score.

[0009] In one embodiment, the present disclosure provides a method for determining molecular response score at least partially using a computer.The method includes: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined at a first time point before administering a therapy, and the second plurality of sequence reads being determined at a second time point after administering a therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; for at least one variant of the plurality of variants classified as somatic, determine a first representative value of the first MAF and a second representative value of the second MAF based on a first mutant allele fraction (MAF) at the first time point and a second MAF at the second time point; determine the ratio of the first representative value at the first time point and the second representative value at the second time point; and output the ratio of the first representative value at the first time point and the second representative value at the second time point as a molecular response score.

[0010] In one embodiment, the present disclosure provides a method for determining a molecular response score for a subject having cancer, at least in part, by using a computer.The method includes: (a) determining, by a computer, mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from the subject at a first time point and a second time point, to generate a first and second set of MAFs for each variant in the plurality of variants.The method includes: (b) calculating, by a computer, the ratios of the first and second MAFs for each variant in the plurality of variants, to generate a set of MAF ratios and corresponding standard deviations for each MAF ratio in the set of MAF ratios.Furthermore, the method also includes: (c) calculating, by a computer, the weighted average and confidence interval of the MAF ratios, to determine the molecular response score for the subject having cancer.

[0011] In another embodiment, the present disclosure provides a method for treating a cancer of a subject. The method includes: (a) determining mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from a subject at a first time point and a second time point to generate a set of first and second MAFs for each variant of the plurality of variants. The method includes: (b) calculating the ratios of the first and second MAFs for each variant in the plurality of variants to generate a set of MAF ratios and corresponding standard deviations for each MAF ratio in the set of MAF ratios. The method also includes: (c) calculating the weighted average and confidence interval of the MAF ratios to determine a molecular response score of the subject. In addition, the method includes: (d) treating the cancer of the subject by administering one or more therapies to the subject based on at least the molecular response score.

[0012] In another embodiment, the present disclosure provides a method for treating cancer of a subject.The method comprises administering one or more therapies to the subject based on at least the molecular response score of the subject.The molecular response score is generated by: (a) determining the mutant allele frequency (MAF) for a plurality of variants from sequence information generated from the targeting nucleic acid associated with one or more cancer types in the sample obtained from the subject at a first time point and a second time point by a computer, to generate a first and second set of MAFs for each variant of the plurality of variants; (b) calculating the ratio of the first and second MAFs for each variant of the plurality of variants by a computer, to generate a set of MAF ratios and corresponding standard deviations for each MAF ratio of the set of MAF ratios; and (c) calculating the weighted average and confidence interval of the MAF ratios by a computer, to determine the molecular response score of the subject.

[0013] In another embodiment, the present disclosure provides a method for identifying clonal hematopoietic variants in a subject with cancer, at least in part using a computer. The method includes (a) determining, by a computer, a tumor burden change (R) relative to a tumor proportion change P(R) for each of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from the subject at a first time point and a second time point to generate a set of tumor burden changes. The method also includes (b) identifying, by a computer, one or more resistance signatures corresponding to one or more clonal hematopoietic variants from the set of tumor burden changes, thereby identifying clonal hematopoietic variants in a subject with cancer.

[0014] In another aspect, the disclosure provides a method for identifying clonal hematopoietic variants in a subject with cancer, at least in part using a computer. The method includes (a) a computer calculating a probability density function for tumor percentage change P(R) for each of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from the subject at a first time point and a second time point. The method also includes (b) a computer grouping one or more variants into one or more clones by P(R), and (c) a computer generating an updated P(R) for each of the clones. In addition, the method also includes (d) a computer identifying one or more clones that have a percentage change between the first time point and the second time point that is at or above a predetermined threshold, thereby identifying a clonal hematopoietic variant in the subject with cancer. In some of these embodiments, the method includes determining the likelihood that a given pair of variants exhibits the same percentage change, merging the most likely pair of variants into one clone, and updating P(R) for the one clone.

[0015] In another embodiment, the present disclosure provides a method for identifying germline variants in a subject with cancer, at least in part using a computer. The method includes: (a) determining, by a computer, the mutant allele frequency (MAF) for a given variant from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from the subject. The method also includes: (b) identifying, by a computer, a given variant as a germline variant if the MAF of the given variant increases the maximum MAF of the sample, the sample containing the maximum proportion of diploid genes (max frac_diploid), and / or if the MAF of the given variant is at least about 2-fold greater, 3-fold greater, 4-fold greater, 5-fold greater, 6-fold greater, 7-fold greater, 8-fold greater, 9-fold greater, or one or more other MAFs determined from a sample obtained from the subject, thereby identifying a germline variant in a subject with cancer.

[0016] In some embodiments, the methods disclosed herein include comparing the molecular response score of a subject having cancer to a predefined cut-off point to identify that the subject is likely to be a responder to one or more therapies for cancer if the molecular response score is below the predefined cut-off point, or that the subject is likely to be a non-responder to one or more therapies for cancer if the molecular response score is above the predefined cut-off point. In some embodiments, the one or more therapies include one or more immunotherapies. In some embodiments, the methods disclosed herein include administering one or more therapies for cancer to the subject in consideration of the molecular response score. In some embodiments, the methods disclosed herein include discontinuing administering one or more therapies for cancer to the subject in consideration of the molecular response score. In some embodiments, the methods disclosed herein include recommending one or more therapies. In some embodiments, the methods disclosed herein include recommending discontinuing one or more therapies. In some embodiments, the methods disclosed herein include using the molecular response score as a prognostic and / or predictive biomarker for the subject.

[0017] In some embodiments, the methods disclosed herein include using molecular counts to calculate a standard deviation for each MAF ratio in the set of MAF ratios. In some embodiments, the methods disclosed herein include propagating the variance through each MAF ratio in the set of MAF ratios. In some embodiments, the methods disclosed herein include excluding one or more germline and / or clonal hematopoietic variants when determining mutant allele frequencies (MAFs) for the plurality of variants. In some embodiments, the plurality of variants includes somatic nucleic acid variants. In some embodiments, the methods disclosed herein include excluding one or more somatic variants having a MAF that is less than about 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, or 0.9% at both the first time point and the second time point. In some embodiments, the first time point includes a pre-treatment time point and the second time point includes a treatment start time point or a treatment post-treatment time point.

[0018] In some embodiments, the method disclosed herein comprises generating sequence information from nucleic acid molecules obtained from one or more tissues or cells in a sample.In some embodiments, the method disclosed herein comprises generating sequence information from cell-free nucleic acid (cfNA) in a sample obtained from a subject.In some embodiments, cfNA comprises circulating tumor DNA (ctDNA).

[0019] In some embodiments, the ratio comprises the second MAF to the first MAF for each variant in the plurality of variants. In some embodiments, the methods disclosed herein comprise calculating a weighted average of the MAF ratios using the following formula: Sum[weight*ratio] / sum[weight], (where the weight is 1 / range for a given variant among multiple variants) 2(wherein the range is the difference between the values ​​of the first and second MAF for a given variant in the plurality of variants, and the ratio is a given MAF ratio within the set of MAF ratios). In some embodiments, the methods disclosed herein include calculating the confidence interval using the following formula: Weighted average of MAF ratios + / - sqrt [ratio variance], (where the specific variance is 1 / sum[weight]).

[0020] In some embodiments, the variants include one or more single nucleotide variants (SNVs), insertion / deletion mutations (indels), gene amplifications, and / or gene fusions. In some embodiments, the methods disclosed herein include using one or more additional genomic data sources to determine a molecular response score for a subject with cancer. In some embodiments, the additional genomic data sources include one or more coverage, off-target coverage, epigenetic signatures, and / or microsatellite instability scores. In some embodiments, the epigenetic signatures include cfNA fragment length, location, and / or endpoint density distribution. In some embodiments, the epigenetic signatures include an epigenetic landscape or state exhibited by one or more epigenetic loci within a given targeted genomic region. In some embodiments, the epigenetic landscape or state includes the presence or absence of methylation, hydroxymethylation, acetylation, ubiquitination, phosphorylation, sumoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone mutations.

[0021] The present application discloses a method, computer readable medium, and system useful for determining a molecular response score of a subject with cancer. Related methods for identifying clonal hematopoietic and / or germline variants are also disclosed. Additional advantages of the disclosed methods, systems, and / or compositions will be set forth in part in the following description, and will be understood in part from the description, or may be learned by practice of the disclosed methods and compositions. The advantages of the disclosed methods and compositions will be realized and attained by the elements and combinations particularly pointed out in the appended claims. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention as claimed.

[0022] BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate several embodiments of the disclosed methods and compositions and, together with the description, serve to explain the principles of the disclosed methods and compositions. [Brief description of the drawings]

[0023] [Figure 1] FIG. 1 illustrates an exemplary method.

[0024] [Diagram 2] FIG. 2 illustrates an exemplary method.

[0025] [Diagram 3] FIG. 3 illustrates an exemplary method.

[0026] [Figure 4] FIG. 4 illustrates an exemplary method.

[0027] [Diagram 5] FIG. 5 illustrates an exemplary method.

[0028] [Figure 6] FIG. 6A illustrates an exemplary method.

[0029] FIG. 6B illustrates an exemplary method.

[0030] [Figure 7] FIG. 7 illustrates an exemplary method.

[0031] [Figure 8] FIG. 8 illustrates an exemplary method.

[0032] [Figure 9] FIG. 9 illustrates an exemplary method.

[0033] [Figure 10] FIG. 10 illustrates an exemplary method.

[0034] [Figure 11] FIG. 11 illustrates an exemplary method.

[0035] [Figure 12] FIG. 12A illustrates an exemplary method.

[0036] FIG. 12B illustrates an exemplary method.

[0037] [Figure 13] FIG. 13 illustrates an exemplary method.

[0038] [Figure 14] FIG. 14 illustrates an exemplary method.

[0039] [Figure 15] FIG. 15 illustrates an exemplary method.

[0040] [Figure 16] FIG. 16 illustrates an exemplary method.

[0041] [Figure 17] FIG. 17 illustrates an exemplary method.

[0042] [Figure 18] FIG. 18 illustrates an exemplary method.

[0043] [Figure 19] FIG. 19 illustrates an exemplary method.

[0044] [Figure 20] FIG. 20 illustrates an exemplary method.

[0045] [Figure 21] FIG. 21 illustrates an exemplary method.

[0046] [Figure 22] FIG. 22 shows the number of somatic variants detected per sample in panel space.

[0047] [Figure 23] FIG. 23 shows an example of somatic typing discrepancy that can skew MR results.

[0048] [Figure 24] Figures 24A-F show an example of mutant precision as determined by mutant molecular counting (MMC=VAF*molecular coverage). (A) Mutants have a range of molecular coverage depending on sample input and panel design. Probability of mutant detection (B) and VAF precision (C) depend on both VAF and molecular coverage (color, mapping to (A)). MMC (D) is a better metric of mutant precision because it determines the probability of mutant detection (E). VAF precision (F).

[0049] [Diagram 25]Figures 25A-C show that when using the average of the ratios, m(rVAF), or the maximum ratio, R(maxVAF), the tumor signal can outweigh the minority of variants. (A) MR scores are classified as increased, decreased, or within precision limits ("near 0% change"). (B) Patient molecular response scores per method are shown. (C) Graph of baseline evaluable variants (y-axis) for R(mVAF) only vs. R(mVAF) all evaluable variants. Solid circles are evaluable, lighter circles (seen in the line across the x-axis) are not evaluable.

[0050] [Figure 26] Figures 26A-C show examples of increasing certainty of molecular response scores with increasing number of variants (A), molecular coverage (B), and maximum VAF (C).

[0051] [Figure 27] Figures 27A and 27B show histograms of molecular response scores for clinical samples (A) and technical replicates (null distribution) (B), along with a hypothetical example of a mutant trajectory.

[0052] [Figure 28] FIG. 28 shows an example of the determination of a molecular response score. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0053] Detailed Description The disclosed methods and compositions may be understood more readily by reference to the following detailed description of specific embodiments and the examples contained therein, as well as the figures and their preceding and following descriptions.

[0054] It is to be understood that the disclosed methods and compositions are not limited to particular synthetic methods, particular analytical techniques, or particular reagents, unless otherwise specified, and as such may vary. I. Definition

[0055] It should also be understood that the terms used herein are for the purpose of describing specific embodiments only and are not intended to be limiting.Furthermore, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this disclosure belongs.In describing and claiming the method, computer readable medium, and system, the following terms and their grammatical variations are used according to the definitions set forth below.

[0056] As used herein and in the appended claims, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Thus, for example, reference to a "method" includes one or more methods and / or steps of the type described herein and / or that would be apparent to one of skill in the art upon reading this disclosure, and so forth. It is also understood that there is an implicit "about" before temperatures, concentrations, times, numbers of bases or base pairs, coverage, and the like discussed in this disclosure, such that minor and insubstantial equivalents are within the scope of this disclosure. In this application, the use of the singular includes the plural, unless otherwise indicated. Also, the use of "comprise," "comprises," "comprising," "contain," "contains," "containing," "include," "includes," and "including" is not intended to be limiting.

[0057] About: As used herein, "about" or "approximately" as applied to one or more of a given value or element refers to a value or element similar to the stated reference value or element. In certain embodiments, the term "about" or "approximately" refers to a range of values ​​or elements that falls within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater or less) of the stated reference value or element, unless otherwise stated or clear from the context (except when such number exceeds 100% of possible values ​​or elements).

[0058] Adapter: As used herein, "adapter" 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 typically at least partially double-stranded and used to ligate to either or both ends of a given sample nucleic acid molecule. The adapter may include a nucleic acid primer binding site that allows amplification of the nucleic acid molecule flanked by the adapter at both ends, and / or a sequencing primer binding site that includes a primer binding site for sequencing applications, e.g., various next generation sequencing (NGS) applications. The adapter may also include a binding site for a capture probe, such as an oligonucleotide bound to a flow cell support. The adapter may also include a nucleic acid tag as described herein. The nucleic acid tag is typically positioned relative to the amplification primer and sequencing primer binding sites such that the nucleic acid tag is included in the amplicon and sequencing reads of a given nucleic acid molecule. The same or different adapters can be ligated to each end of the nucleic acid molecule. In certain embodiments, the same adapter is ligated to each end of the nucleic acid molecule, except that the nucleic acid tags are different. In some embodiments, the adaptor is a Y-shaped adaptor having one end blunt or tailed as described herein for binding to a nucleic acid molecule, also blunt or tailed with one or more complementary nucleotides. In yet other exemplary embodiments, the adaptor is a bell-shaped adaptor that includes a blunt or tailed end for binding to a nucleic acid molecule to be analyzed. Other exemplary adaptors include T-tail adaptors and C-tail adaptors.

[0059] Administer: As used herein, "administering" or "administering" a therapeutic agent (e.g., an immunological therapeutic agent) to a subject means giving, providing, or contacting the composition with the subject. Administration can be accomplished by any of a number of routes, including, for example, topical, oral, subcutaneous, intramuscular, intraperitoneal, intravenous, intrathecal, and intradermal.

[0060] Allele: As used herein, "allele" or "allelic variant" refers to a specific genetic variant at a defined genomic location or locus. Allelic variants are usually presented at a frequency of 50% (0.5) or 100%, depending on whether the allele is heterozygous or homozygous. For example, germline variants are inherited and usually have a frequency of 0.5 or 1. Somatic variants; however, are acquired variants and usually have a frequency of <0.5. Major and minor alleles of a locus refer to nucleic acids having a locus where the locus is occupied by a nucleotide of a reference sequence and a variant nucleotide different from the reference sequence, respectively. Measurements at a locus can take the form of allelic fraction (AF), which measures the frequency with which an allele is observed in a sample.

[0061] Amplification: As used herein, "amplify" or "amplification" in reference to nucleic acids refers to the production of multiple copies of a polynucleotide or a portion of a polynucleotide, typically starting from a small amount of polynucleotide (e.g., a single polynucleotide molecule), where the amplification product or amplicon is generally detectable. Amplification of polynucleotides encompasses a variety of chemical and enzymatic processes.

[0062] Barcode: As used herein, "barcode" in the context of nucleic acids refers to a nucleic acid molecule that contains a sequence that can function as a molecular identifier. For example, individual "barcode" sequences are typically added to each DNA fragment during next generation sequencing (NGS) library preparation so that each read can be identified and sorted prior to final data analysis.

[0063] Cancer type: As used herein, "cancer," "cancer type," or "tumor type" refers to a type or subtype of cancer, as defined, for example, by histopathology. Cancer types are classified based on their occurrence in a given tissue (e.g., hematological cancer, central nervous system (CNS), brain cancer, lung cancer (small cell and non-small cell), skin cancer, nasal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, intestinal cancer, rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer, stomach cancer, breast cancer, prostate cancer, ovarian cancer, lung cancer, intestinal cancer, soft tissue cancer, neuroendocrine cancer, gastroesophageal cancer, head and neck cancer, gynecological cancer, colorectal cancer, urothelial cancer, solid cancer, heterogeneous cancer, homogenous cancer, and the like). Cancers may be defined by any conventional criteria, such as cancers of unknown primary origin, and / or of the same cellular lineage (e.g., carcinoma, sarcoma, lymphoma, cholangiocarcinoma, leukemia, mesothelioma, melanoma, or glioblastoma) and / or cancer markers such as Her2, CA15-3, CA19-9, CA-125, CEA, AFP, PSA, HCG, hormone receptors, and NMP-22. Cancers may also be classified by stage (e.g., stage 1, 2, 3, or 4) and whether primary or secondary.

[0064] Cell-free nucleic acid: As used herein, "cell-free nucleic acid" refers to nucleic acid that is not contained within or otherwise bound to a cell, or in some embodiments, nucleic acid remaining in a sample after removal of intact cells. Cell-free nucleic acid may include, for example, all unencapsulated nucleic acid derived from bodily fluids (e.g., blood, plasma, serum, urine, cerebrospinal fluid (CSF), etc.) from a subject. Cell-free nucleic acid includes 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), long non-coding RNA (long ncRNA), and / or fragments of any of these. Cell-free nucleic acid may be double-stranded, single-stranded, or hybrids thereof. Cell-free nucleic acid may be released into bodily fluids via secretion or cell death processes, such as cell necrosis, apoptosis, etc. Cell-free nucleic acid may be found in efferosomes or exosomes. Some cell-free nucleic acids are released into body fluids from cancer cells, such as circulating tumor DNA (ctDNA). Others are released from healthy cells. CtDNA can be non-encapsulated tumor-derived fragmented DNA. Another example of cell-free nucleic acid is fetal DNA, which circulates freely in maternal bloodstream, also called cell-free fetal DNA (cffDNA). Cell-free nucleic acid can have one or more epigenetic modifications, for example, cell-free nucleic acid can be acetylated, 5-methylated, ubiquitinated, phosphorylated, sumoylated, ribosylated, and / or citrullinated.

[0065] Classifier: As used herein, a "classifier" generally refers to algorithmic computer code that receives test data as input and produces as output a classification of the input data as belonging to one or another class (e.g., tumor DNA or non-tumor DNA).

[0066] Clone: ​​As used herein, with reference to nucleic acids, "clone" refers to a population of nucleic acids that contain nucleotide sequences that are substantially or completely identical to each other at least at a given locus of interest (e.g., a targeted mutant).

[0067] Undetermined potential clonal hematopoiesis: As used herein, "undetermined potential clonal hematopoiesis", "clonal hematopoietic variant" or "CHIP" refers to hematopoiesis in an individual that comprises a proliferation of hematopoietic stem cells that contain one or more somatic mutations (e.g., hematologic cancer-associated mutations and / or non-cancer-associated mutations) but otherwise lack diagnostic criteria for hematologic malignancy such as definitive morphological evidence of dysplasia. CHIP is a common age-associated phenomenon in which hematopoietic stem cells contribute to the formation of genetically distinct subpopulations of blood cells.

[0068] Confidence interval: As used herein, "confidence interval" or "confidence level" means a range of values ​​defined such that there is a certain probability that the value of a given parameter falls within that range of values.

[0069] Copy number variant: As used herein, "copy number variant", "CNV" or "copy number variation" refers to the phenomenon in which sections of the genome are repeated and the number of repeats in the genome varies among individuals in the population under consideration.

[0070] Coverage: As used herein, "coverage" refers to the number of nucleic acid molecules that represent a particular base position.

[0071] Deoxyribonucleic acid or ribonucleic acid: As used herein, "deoxyribonucleic acid" or "DNA" refers to a natural or modified nucleotide with a hydrogen group at the 2' position of the sugar moiety. DNA typically comprises a chain of nucleotides that contains four types of nucleotide bases: adenine (A), thymine (T), cytosine (C) and guanine (G). As used herein, "ribonucleic acid" or "RNA" refers to a natural or modified nucleotide with a hydroxyl group at the 2' position of the sugar moiety. RNA typically comprises a chain of nucleotides that contains four types of nucleotide bases: A, uracil (U), G and C. As used herein, the term "nucleotide" refers to a natural or modified nucleotide. Certain nucleotide pairs specifically bind to each other in a complementary manner (called 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 binds to a second nucleic acid strand that is composed of nucleotides complementary to those of the first strand, the two strands combine to form a duplex. As used herein, "nucleic acid sequencing data," "nucleic acid sequencing information," "sequence information," "nucleic acid sequence," "nucleotide sequence," "genomic sequence," "gene sequence," or "fragment sequence," or "nucleic acid sequencing read" refers to any information or data that indicates the order and identity of nucleotide bases (e.g., adenine, guanine, cytosine and thymine or uracil) in a molecule (e.g., whole genome, whole transcriptome, exome, oligonucleotide, polynucleotide, or fragment) of a nucleic acid such as DNA or RNA.It should be understood that the present teachings contemplate sequence information obtained using all of the various techniques, platforms or technologies available, 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 signaling-based systems.

[0072] Detection: As used herein, "detect," "detecting," or "detection" refers to the act of determining the presence or absence of one or more targeted nucleic acids (e.g., nucleic acids having a targeted mutation or other marker) in a sample.

[0073] Enriched sample: As used herein, "enriched sample" refers to a sample that is enriched in a specific region of interest. The sample can be enriched by amplifying the region of interest or by using single-stranded DNA / RNA probes or double-stranded DNA probes that can hybridize to the nucleic acid molecules of interest (e.g., SureSelect® probes, Agilent Technologies). In some embodiments, the enriched sample refers to a subset or portion of the enriched processed sample, which contains cell-free polynucleotides or nucleic acid molecules from the sample of polynucleotides.

[0074] Epigenetic information: As used herein, "epigenetic information" in the context of a DNA polymer means one or more epigenetic patterns or signatures exhibited in that polymer.

[0075] Epigenetic locus: As used herein, "epigenetic locus" or "epigenetic site" refers to a fixed location on a chromosome that exhibits a different state or condition without a change or alteration in nucleotide sequence. For the avoidance of doubt, a given epigenetic locus may correspond to a given nucleotide position or genomic region that also exhibits genetic or sequence variation (e.g., mutation). For example, a given epigenetic locus may or may not be acetylated, methylated (e.g., modified with 5-methylcytosine (5mC), modified with 5-hydroxymethylcytosine (5hmC), etc.), ubiquitinated, phosphorylated, sumoylated, ribosylated, citrullinated, have histone post-translational modifications or other histone mutations, etc.

[0076] Epigenetic signature: As used herein, "epigenetic signature" refers to the epigenetic status or state that is indicated by one or more epigenetic loci in a given DNA molecule.For example, a DNA molecule or cfDNA fragment that includes a given genomic region or locus (such as a CTCF binding region) may also show an epigenetic pattern that includes a certain number of epigenetic loci in which some of those DNA molecules are methylated, while in other examples, the corresponding epigenetic loci in other DNA molecules or cfDNA fragments that include the same genomic region are not methylated.

[0077] Germline mutation: As used herein, "germline mutation" refers to a mutation in nucleic acid in germ cells that is present before conception.

[0078] Immunotherapy: As used herein, "immunotherapy" refers to treatment with one or more agents that act to kill or at least inhibit the growth of cancer cells, preferably to reduce further growth of cancer, reduce the size of cancer, and / or stimulate the immune system to eliminate cancer. Some such agents bind to targets present on cancer cells, some bind to targets present on immune cells but not on cancer cells, and some bind 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 pathways in the immune system that maintain self-tolerance and regulate the duration and amplitude of physiological immune responses in peripheral tissues to minimize associated tissue damage (see, e.g., Pardoll, Nature Reviews Cancer 12, 252-264 (2012)). Exemplary agents include antibodies against PD-1, PD-2, PD-L1, PD-L2, CTLA-4, OX40, B7.1, B7He, LAG3, CD137, KIR, CCR5, CD27, CD40 or CD47. Other exemplary agents include proinflammatory cytokines, such as IL-1β, IL-6 and TNF-α. Other exemplary agents are T cells activated against tumors, such as T cells activated by expressing chimeric antigens that target tumor antigens recognized by T cells.

[0079] Indel: As used herein, "indel" refers to a mutation involving the insertion or deletion of a nucleotide position in the genome of a subject.

[0080] Maximum mutant allele frequency: As used herein, "maximum mutant allele frequency", "maximum mutant allele frequency", "maximum MAF", "MAX MAF", "maximum VAF", "max-MAF" or "MAX VAF" refers to the maximum or greatest MAF of all somatic variants present or observed in a given sample.

[0081] Mutant allele frequency: As used herein, "mutant allele frequency," "mutant allele frequency," "mutant allele proportion," "mutant allele proportion," "MAF" or "VAF" refers to the frequency with which a mutant allele occurs in a given population of nucleic acids, such as a sample obtained from a subject. MAF is generally expressed as a proportion or percentage.

[0082] Molecular response: As used herein, "molecular response" refers to changes in one or more circulating tumor DNA (ctDNA) variant allele frequencies, levels, or amounts observed between samples taken from a given subject at different time points.

[0083] Molecular responder: As used herein, a "molecular responder" or "responder" refers to a subject having a molecular response score indicating a decrease in one or more circulating tumor DNA (ctDNA) variant allele frequencies, levels, or amounts observed between samples taken from a subject at different time points.

[0084] Molecular non-responder: As used herein, "molecular non-responder" or "non-responder" refers to a subject with a molecular response score that does not show an increase or change in one or more circulating tumor DNA (ctDNA) variant allele frequencies, levels, or amounts observed between samples taken from the subject at different time points. A threshold that specifies the level of decrease (or increase) can be utilized to determine whether a subject is a molecular responder or a molecular non-responder. For example, a molecular responder can be a subject associated with a decrease of more than a certain percentage change in VAF, and a non-responder can be a subject associated with an increase, no change, or decrease of less than a certain percentage change in VAF.

[0085] Mutation: As used herein, "mutation", "nucleic acid variant", "variant" or "genetic abnormality" refers to a variation from a known reference sequence, including, for example, mutations such as single nucleotide variants (SNVs), copy number variants or variations (CNVs) / aberrations, insertions or deletions (indels), truncations, gene fusions, transversions, translocations, frameshifts, duplications, repeat expansions, and epigenetic variants. Mutations can be germline or somatic mutations. In some embodiments, the reference sequence for comparison is the wild-type genomic sequence of the species of subject that provides the test sample, typically the human genome. In certain cases, the mutation or variant is a "tumor-associated genetic variant" that causes or at least contributes to tumorigenesis.

[0086] Next generation sequencing: As used herein, "next generation sequencing" or "NGS" refers to sequencing technologies with increased throughput compared to traditional Sanger electrophoresis and capillary electrophoresis-based approaches, e.g., with the ability to generate hundreds of thousands of relatively small sequence reads at a time. Some examples of next generation sequencing technologies include, but are not limited to, sequencing by synthesis, sequencing by ligation, and sequencing by hybridization.

[0087] Nucleic acid tag: As used herein, "nucleic acid tag" refers to a short nucleic acid (e.g., less than about 500, about 100, about 50 or about 10 nucleotides in length) used to label nucleic acid molecules to distinguish between nucleic acids from different samples (e.g., representing a sample index) or different nucleic acid molecules in the same sample of different types (e.g., representing a molecular tag) or nucleic acids that have been subjected to different processing. Nucleic acid tags can be single-stranded, double-stranded or at least partially double-stranded. Nucleic acid tags have the same length or various lengths as desired. Nucleic acid tags can also include double-stranded molecules with one or more blunt ends, can include 5' or 3' single-stranded regions (e.g., overhangs), and / or can include one or more other single-stranded regions at other locations within a given molecule. Nucleic acid tags can be attached to one or both ends of other nucleic acids (e.g., sample nucleic acids to be amplified and / or sequenced). Nucleic acid tags can be decoded to reveal information such as the sample of origin, form or processing of a given nucleic acid. Nucleic acid tags can also be used to allow pooling and / or parallel processing of multiple samples containing nucleic acids with different nucleic acid tags and / or sample indexes, with the nucleic acids subsequently deconvoluted by reading the nucleic acid tags. Nucleic acid tags can also be referred to as molecular identifiers or tags, sample identifiers, index tags, and / or barcodes. Additionally or alternatively, nucleic acid tags can be used to distinguish different molecules in the same sample. This includes, for example, uniquely tagging each different nucleic acid molecule in a given sample, or not uniquely tagging such molecules. For non-unique tagging applications, a limited number of tags can be used to tag each nucleic acid molecule, for example, so that different molecules can be distinguished based on start / stop positions that map to a selected reference genome in combination with at least one nucleic acid tag. Typically, a sufficient number of different nucleic acid tags are used such that the probability that any two molecules have the same start / stop positions and also have the same nucleic acid tag is low (e.g., less than about 10%, less than about 5%, less than about 1%, or less than about 0.1% probability).Some nucleic acid tags include multiple molecular identifiers for labeling samples, forms of nucleic acid molecules within the samples, and nucleic acid molecules within forms that have the same start and stop positions. Such nucleic acid tags can be referenced using the exemplary format "A1i," where a capital letter indicates the sample type, an Arabic numeral indicates the form of the molecule within the sample, and a lowercase Roman numeral indicates the molecule within the form.

[0088] Polynucleotide: As used herein, "polynucleotide", "nucleic acid", "nucleic acid molecule", or "oligonucleotide" refers to a linear polymer of nucleosides (including deoxyribonucleosides, ribonucleosides, or analogs thereof) linked by internucleoside linkages. Typically, a polynucleotide contains at least three nucleosides. Oligonucleotides often range in size from a few monomeric units, e.g., 3-4, to hundreds of monomeric units. Whenever a polynucleotide is represented by a sequence of letters, such as "ATGCCTG", it will be understood that the nucleotides are in 5'→3' order from left to right, unless otherwise noted, and that, for 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 refer to the bases themselves, nucleosides, or nucleotides that contain the bases, as is standard in the art.

[0089] Reference Sample: As used herein, a "reference sample" or "reference cfNA sample" refers to a sample of known composition and / or a sample known to have or lack a particular characteristic (e.g., known nucleic acid variant, known cellular origin, known tumor fraction, known coverage, etc.) that is analyzed with or compared to a test sample to assess the accuracy of an analytical procedure, classify the test sample, etc. A reference sample dataset typically includes at least about 25 to at least about 30,000 or more reference samples. In some embodiments, the reference sample dataset comprises about 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 reference samples.

[0090] Reference sequence: As used herein, "reference sequence" or "reference genome" refers to a known sequence used for the purpose of comparison with an experimentally determined sequence. For example, the known sequence can be a whole genome, a chromosome, or any segment thereof. A reference sequence typically comprises 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 can be aligned with a single continuous sequence of a genome or chromosome, or can comprise discontinuous segments that are aligned with different regions of a genome or chromosome. Exemplary reference sequences include human genomes, such as, for example, hG19 and hG38.

[0091] Sample: As used herein, "sample" refers to any biological sample that can be analyzed by the methods and / or systems disclosed herein. In certain embodiments of the present disclosure, the sample is a bodily fluid sample, such as whole blood or a fraction thereof, lymph, urine, and / or cerebrospinal fluid, among other bodily fluid types that provide acellular (circulating not contained within cells or otherwise bound to cells) nucleic acids. In certain embodiments, the bodily fluid sample is a plasma sample, which is the fluid portion of whole blood that excludes cells such as red blood cells and white blood cells. In some embodiments, the bodily fluid sample is a serum sample, i.e., plasma that lacks fibrinogen. In some embodiments of the present disclosure, the sample is a "non-bodily fluid sample" or "non-plasma sample", i.e., a biological sample other than a "bodily fluid sample", such as a cell and / or tissue sample, from which nucleic acids other than acellular nucleic acids are provided.

[0092] Sensitivity: As used herein, "sensitivity" in the context of a given assay or method refers to the ability of the assay or method to detect and distinguish between targeted (e.g., cfDNA fragments derived from tumor cells) and non-targeted (e.g., cfDNA fragments derived from non-tumor cells) analytes.

[0093] 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 biological molecule, e.g., a nucleic acid 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 microscope-based sequencing, panel sequencing, transistor-mediated sequencing, direct sequencing, random shotgun sequencing, Sanger dideoxy termination sequencing, whole genome sequencing, sequencing by hybridization, pyrosequencing, capillary electrophoresis, gel electrophoresis, double-stranded sequencing, cycle sequencing, single-base extension sequencing, and the like. The sequencing methods include sequencing, solid-phase sequencing, high-throughput sequencing, massively parallel signature sequencing, emulsion PCR, co-amplification-PCR at low denaturation temperature (COLD-PCR), multiplex PCR, reversible dye terminator sequencing, paired-end sequencing, near-term sequencing, exonuclease sequencing, ligation sequencing, short-read sequencing, single molecule sequencing, synthesis sequencing, 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 by a genetic analyzer, such as a genetic analyzer commercially available from Illumina, Inc., Pacific Biosciences, Inc., or Applied Biosystems / Thermo Fisher Scientific, among others.

[0094] Single nucleotide variant: As used herein, "single nucleotide variant" or "SNV" refers to a single nucleotide mutation or variation that occurs at a specific location in the genome.

[0095] Somatic mutation: As used herein, "somatic mutation" refers to a mutation in the genome that occurs after conception. Somatic mutations can occur in any cell of the body except germ cells and therefore cannot be passed on to offspring.

[0096] Specificity: As used herein, "specificity" in the context of a diagnostic analysis or assay refers to the degree to which the analysis or assay detects its intended target analyte to the exclusion of other constituents of a given sample.

[0097] Subclone: ​​As used herein, in reference to a nucleic acid, a "subclone" refers to a subpopulation of nucleic acids (i.e., a subset of a population of nucleic acids) that contain nucleotide sequences that are substantially or completely identical to each other at least at a given locus of interest (e.g., a target variant). For example, a subclonal can refer to a subset of cancer cells.

[0098] Subject: As used herein, "subject" refers to an animal, such as a mammalian species (e.g., human) or avian (e.g., bird) species, or other organism, such as a plant. More specifically, the subject may be a vertebrate, e.g., a mammal, such as a mouse, a primate, a monkey, or a human. Animals include livestock (e.g., production cattle, dairy cattle, poultry, horses, pigs, etc.), sports animals, and companion animals (e.g., pets or support animals). A subject may be a healthy individual, an individual having or suspected of having a disease or predisposition to a disease, or an individual in need of therapy or suspected of needing therapy. The terms "individual" or "patient" are intended to be interchangeable with "subject."

[0099] For example, the subject may be an individual who has been diagnosed with cancer, who is going to undergo cancer therapy, and / or who has undergone at least one cancer therapy.The subject may be in remission of cancer.As another example, the subject may be an individual who has been diagnosed with autoimmune disease.As another example, the subject may be a female individual who is pregnant or planning pregnancy, and may be diagnosed or suspected with disease, such as cancer, autoimmune disease.

[0100] Threshold: As used herein, a "threshold" refers to a discretely determined value used to characterize or classify experimentally determined values.

[0101] Tumor fraction: As used herein, "tumor fraction" refers to the estimated fraction of nucleic acid molecules derived from tumor in a given sample.For example, the tumor fraction of a sample can be the maximum somatic mutant allele frequency (max MAF) of the sample or the coverage of the sample, or the length of the cfNA fragment in the sample, epigenetic status, or other characteristics, or any other selected characteristic of the sample.In some embodiments, the tumor fraction of a sample is equal to the maximum MAF of the sample.

[0102] Value: As used herein, a "value" generally refers to an entry in a dataset and can be anything that characterizes the feature to which it refers, including but not limited to a number, a word or phrase, a symbol (e.g., + or -), or a degree.

[0103] Any method and material similar or equivalent to the methods and materials described herein can be used in carrying out or testing the methods and compositions, but particularly useful methods, devices, and materials are as described. Publications cited herein and the materials for which they are cited are specifically incorporated herein by reference. Nothing herein should be construed as an admission that the invention is not entitled to antedate such disclosure by prior invention. No admission is made that any reference constitutes prior art. The discussion of a reference states what its author asserts, and applicants reserve the right to challenge the accuracy and pertinence of the cited documents. It will be clearly understood that although several publications are referred to herein, such reference is not an admission that any of these documents form part of the common general knowledge in the art.

[0104] Throughout the description and claims of this specification, the word "comprise" and variations of words such as "comprising" and "comprises" mean "including but not limited to" and are not intended to exclude, for example, other additives, components, integers or steps. In particular, in methods described as including one or more steps or operations, it is specifically contemplated that each step includes what is recited (unless that step includes a limiting term such as "consisting of"), meaning that each step is not intended to exclude, for example, other additives, components, integers or steps not recited in that step.

[0105] Where the invention provides a process comprising multiple successive steps, the invention can also provide a process in which these different steps may be carried out at very different times by different people in different places (e.g. in different countries). II. Molecular Response Scoring

[0106] In the embodiment shown in FIG. 1, a method 100 for determining a molecular response (MR) score is disclosed. The disclosed method may have a wide variety of applications in the manipulation, preparation, identification, quantification and / or analysis of cell-free nucleic acids. Molecular response is an assessment of the change in circulating tumor DNA (ctDNA) burden during treatment (usually 3-10 weeks) compared to a pre-treatment baseline. Molecular response is related to patient response to therapy and long-term outcomes across solid tumors and therapy types. Molecular response can also be used to predict clinical response earlier than radiographic and / or RECIST response. Multiple methods have been used to calculate molecular response, and there is no consensus on which method is best.

[0107] A method and system for assessing response to treatment using a molecular response (MR) score is described. In one embodiment, baseline (pre-treatment) gene expression data can be obtained for a number of patients prior to treatment, and on-treatment gene expression data can be obtained for a number of patients during treatment. In one embodiment, the baseline gene expression data (e.g., variant data) and / or on-treatment gene expression data can be analyzed to determine a molecular response (MR) score. The MR score can indicate that the patient is a responder or non-responder to the treatment. In an embodiment, the mutant allele fraction (MAF) can be determined as part of the MR score. In one embodiment, the variance of each MAF can be incorporated into the molecular response score determination. This ensures that the molecular response score includes the correct variance, providing a significant improvement in making the correct conclusions from the molecular response score. The improvement is even more significant when the molecular response score is a ratio, since ratios are sensitive to the variance of the denominator. Variance can be incorporated into the molecular response score by mathematically deriving the molecular response variance, or by simulating or sampling from the variance distribution of each variant to determine the molecular response variance. a. cfDNA isolation and extraction

[0108] As shown in FIG. 1, at a first time T0, baseline cfDNA may be obtained from one or more baseline samples obtained from one or more subjects prior to treatment at step 101, and at a second time T1, on-treatment cfDNA may be obtained from one or more on-treatment samples obtained from one or more subjects after treatment at step 102. Treatment may / can be performed at any time after time T0. For example, treatment may be performed minutes, hours, days, etc. after time T0. As a further example, treatment may be performed 30 minutes after time T0, 1 to 2 hours after time T0, 1 to 2 days after time T0, 1 to 2 weeks after time T0, 1 to 2 months after time T0, 6 months to 1 year after time T0, 1 to 2 years after time T0, etc. Time T1 can be any amount of time after time T0, including, for example, 1 to 24 hours, 1 to 180 days, 1 to 12 weeks, 6 to 12 months, and the like.

[0109] As described herein, polynucleotides can include any type of nucleic acid, such as DNA and / or RNA. For example, when polynucleotides are DNA, they can be genomic DNA, complementary DNA (cDNA), or any other deoxyribonucleic acid. Polynucleotides can be cell-free nucleic acids, such as cell-free DNA (cfDNA). For example, polynucleotides can be circulating cfDNA. Circulating cfDNA can include DNA that is discharged from body cells via apoptosis or necrosis. cfDNA discharged via apoptosis or necrosis can originate from normal (e.g., healthy) body cells. When there is abnormal tissue growth, such as cancer, tumor DNA can be discharged. Circulating cfDNA can include circulating tumor DNA (ctDNA). i. Sample

[0110] Isolation and extraction of cell-free polynucleotides can be performed by collecting samples using various techniques. The sample can be any biological sample isolated from a subject. Samples can include body tissue, whole blood, platelets, serum, plasma, feces, red blood cells, white blood cells or leucocytes, endothelial cells, tissue biopsies (e.g., biopsies from known or suspected solid tumors), cerebrospinal fluid, synovial fluid, lymphatic fluid, ascites fluid, interstitial or extracellular fluid (e.g., fluid from intercellular spaces), gingival exudate, crevicular fluid, bone marrow, pleural fluid, cerebrospinal fluid, saliva, mucous membrane, sputum, semen, sweat, and urine. Samples are preferably body fluids, particularly blood and its fractions, as well as urine. Such samples include nucleic acids discharged from tumors. Nucleic acids can include DNA and RNA and can be in double-stranded and single-stranded forms. Sample may be in the form that is first isolated from subject, or may be further processed to remove or add components such as cells, to enrich one component with respect to another, or to convert one form of nucleic acid into another, such as RNA into DNA, or single-stranded nucleic acid into double-stranded.Thus, for example, the body fluid sample for analysis is plasma or serum that contains cell-free nucleic acid, such as cell-free DNA (cfDNA).

[0111] In some embodiments, the sample volume of bodily fluid taken from the subject depends on the desired read depth of the sequenced region. Exemplary volumes are about 0.4-40 ml, about 5-20 ml, about 10-20 ml. For example, the volume can be about 0.5 ml, about 1 ml, about 5 ml, about 10 ml, about 20 ml, about 30 ml, about 40 ml, or more milliliters. The amount of blood sampled is typically about 5 ml to about 20 ml.

[0112] Samples can contain varying amounts of nucleic acid. Typically, the amount of nucleic acid in a given sample is equivalent to multiple genome equivalents. For example, a sample of about 30 ng of DNA will contain approximately 10,000 (10 4 ) haploid human genome equivalents, and for cfDNA, approximately 200 billion (2x1011 ) 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.

[0113] In some embodiments, the sample comprises nucleic acid from different sources, such as cells and acellular sources (e.g., blood samples, etc.). Typically, the sample comprises nucleic acid having a mutation. For example, the sample optionally comprises DNA carrying germline mutations and / or somatic mutations. Typically, the sample comprises DNA carrying cancer-associated mutations (e.g., cancer-associated somatic mutations). In some embodiments of the present disclosure, the cell-free nucleic acid in the subject can be derived from a tumor. For example, the cell-free DNA isolated from the subject can comprise ctDNA.

[0114] Exemplary amounts of cell-free nucleic acid in a sample prior to amplification typically range from about 1 femtogram (fg) to about 1 microgram (μg), e.g., from about 1 picogram (pg) to about 200 nanograms (ng), from about 1 ng to about 100 ng, from 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 cell-free nucleic acid molecules. Optionally, 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 cell-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 cell-free nucleic acid molecule. In some embodiments, the method comprises obtaining from about 1 fg to about 200 ng of cell-free nucleic acid molecule from the sample.

[0115] Cell-free nucleic acids typically have a size distribution of about 100 to about 500 nucleotides in length, with molecules of about 110 to about 230 nucleotides in length accounting for about 90% of the molecules in a sample, with a mode of about 168 nucleotides in length and a second minor peak in the range of about 240 to about 440 nucleotides in length. In certain embodiments, the cell-free nucleic acids are about 160 to about 180 nucleotides in length, or about 320 to about 360 nucleotides in length, or about 440 to about 480 nucleotides in length.

[0116] In some embodiments, the cell-free nucleic acids are isolated from the body fluid through a partitioning step in which the cell-free nucleic acids as found in solution are separated from intact cells and other insoluble components of the body fluid. In some of these embodiments, the partitioning includes techniques such as centrifugation or filtration. Alternatively, the cells in the body fluid are lysed and the cell-free and cellular nucleic acids are processed together. Generally, after the addition of buffer and washing steps, the cell-free nucleic acids are precipitated, for example with alcohol. In certain embodiments, additional clean-up steps, such as silica-based columns, are used to remove contaminants or salts. Non-specific bulk carrier nucleic acids are added throughout the reaction as needed to optimize certain aspects of the exemplary procedure, such as, for example, yield. After such processing, the sample typically contains various forms of nucleic acids, including double-stranded DNA, single-stranded DNA and / or single-stranded RNA. If necessary, the single-stranded DNA and / or single-stranded RNA are converted to double-stranded form for inclusion in subsequent processing and analysis steps. Further details regarding cfDNA partitions and associated analysis of epigenetic modifications, optionally adapted for use in practicing the methods disclosed herein, are described, for example, in International Publication No. WO 2018 / 119452, filed December 22, 2017, which is incorporated by reference. ii. Nucleic acid tags

[0117] In certain embodiments, tags providing molecular identifiers or barcodes are incorporated or otherwise attached to the adaptors by chemical synthesis, ligation, or overlap extension PCR, among other methods. In some embodiments, the assignment of unique or non-unique identifiers or molecular barcodes during a reaction is described, for example, in U.S. Patent Application No. 20010053519, U.S. Patent Application No. 20030152490, U.S. Patent Application No. 20110160078, and U.S. Patent Nos. 6,582,908, 7,537,898, and 9,598,731, each of which is incorporated by reference.

[0118] The tags are randomly or non-randomly linked (e.g., ligated) to the sample nucleic acid. In some embodiments, the tags are introduced with an expected ratio of identifiers (e.g., unique and / or non-unique barcode combinations) to microwells. For example, identifiers can be loaded such that more than about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 500, 1000, 5000, 10000, 50,000, 100,000, 500,000, 1,000,000, 10,000,000, 50,000,000 or 1,000,000,000 identifiers are loaded per genome sample. In some embodiments, the identifiers are loaded such that less than about 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 500, 1000, 5000, 10000, 50,000, 100,000, 500,000, 1,000,000, 10,000,000, 50,000,000 or 1,000,000,000 identifiers are loaded per genomic sample. In certain embodiments, the average number of identifiers loaded per genomic sample is less than or greater than about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 500, 1000, 5000, 10000, 50,000, 100,000, 500,000, 1,000,000, 10,000,000, 50,000,000 or 1,000,000,000 identifiers per genomic sample. Identifiers are generally unique or non-unique.

[0119] One exemplary format uses about 2 to about 1,000,000 different tags, or about 5 to about 150 different tags, or about 20 to about 50 different tags linked to both ends of a target nucleic acid molecule. For 20-50 x 20-50 tags, a total of 400-2500 tags are created. Such a number of tags is typically sufficient so that different molecules with the same start and stop points have a high probability (e.g., at least 94%, 99.5%, 99.99%, 99.999%) that they will receive different combinations of tags.

[0120] In some embodiments, the identifier is a predetermined, random, or semi-random sequence oligonucleotide. In other embodiments, multiple barcodes may be used such that the barcodes are not necessarily unique to each other among the multiple barcodes. In these embodiments, the barcode is generally attached (e.g., by ligation or PCR amplification) to individual molecules, thereby creating a unique sequence that can be tracked individually for the combination of the barcode and the sequence to which it can be attached. As described herein, detection of a non-uniquely tagged barcode in combination with sequence data of the beginning (start) and ending (stop) portions of the sequence read typically allows assignment of a unique identity to a particular molecule. The length or number of base pairs of the individual sequence reads is also used to assign a unique identity to a given molecule, as necessary. As described herein, fragments from a single strand of nucleic acid that have been assigned a unique identity may thereby allow subsequent identification of fragments from the parental and / or complementary strands. iii. Nucleic acid amplification

[0121] The sample nucleic acid adjacent to the adaptor is typically amplified by PCR and other amplification methods using a nucleic acid primer that binds to the primer binding site in the adaptor adjacent to the DNA molecule to be amplified. In some embodiments, the amplification method includes cycles of extension, denaturation and annealing resulting from thermal cycling, or can be isothermal, for example, as in the case of transcription-mediated amplification. Other exemplary amplification methods that are optionally utilized include ligase chain reaction, strand displacement amplification, nucleic acid sequence-based amplification, and self-sustaining sequence-based replication, among other approaches.

[0122] One or more amplification cycles are generally applied to introduce the sample index / tag into the nucleic acid molecule using conventional nucleic acid amplification methods. The amplification is typically performed in one or more reaction mixtures. In some embodiments, the molecular tag and the sample index / tag are introduced before and / or after the sequence capture step is performed. In some embodiments, only the molecular tag is introduced before the probe capture step, and the sample index / tag is introduced after the sequence capture step is performed. In certain embodiments, both the molecular tag and the sample index / tag are introduced before performing the probe-based capture step. In some embodiments, the sample index / tag is introduced after the sequence capture step (i.e., enrichment of the nucleic acid) is performed. Typically, the sequence capture protocol includes introducing a single-stranded nucleic acid molecule that is complementary to a targeted nucleic acid sequence, such as a coding sequence of a genomic region and a mutation in such region associated with a cancer type. Typically, the amplification reaction creates multiple non-uniquely or uniquely tagged nucleic acid amplicons having molecular tags and sample indexes / tags ranging in size from about 200 nucleotides (nt) to about 700 nt, 250 nt to about 350 nt, or about 320 nt to about 550 nt. In some embodiments, the amplicons have a size of about 300 nt. In some embodiments, the amplicons have a size of about 500 nt. iv. Nucleic acid enrichment

[0123] In some embodiments, sequences are enriched prior to sequencing the nucleic acids. Enrichment is performed for specific target regions or non-specifically ("target sequences"), as appropriate. As an example, enrichment can be performed non-specifically based on size selection methods that are sequence fragment size specific rather than sequence specific. In some embodiments, targeted regions of interest can be enriched with nucleic acid capture probes ("baits") selected for one or more bait set panels using differential tiling and capture schemes. Differential tiling and capture schemes generally use different relative concentrations of bait sets to differentially tile (e.g., with different "resolutions") across genome sections related to the baits according to a set of constraints (e.g., sequencer constraints such as sequencing load, usefulness of each bait, etc.) to capture targeted nucleic acids at a desired level for downstream sequencing. These targeted genome sections of interest optionally include natural or synthetic nucleotide sequences of nucleic acid constructs. In some embodiments, biotin-labeled beads with probes for one or more sections of interest can be used to capture target sequences and optionally subsequently amplify those sections to enrich the regions of interest.

[0124] Sequence capture typically involves the use of oligonucleotide probes that hybridize to target nucleic acid sequences. In certain embodiments, the probe set strategy involves tiling probes across a section of interest. Such probes can be, for example, about 60 to about 120 nucleotides long. Sets can have a depth of about 2x, 3x, 4x, 5x, 6x, 8x, 9x, 10x, 15x, 20x, 50x or more. The effectiveness of sequence capture generally depends in part on the length of the sequence in the target molecule that is complementary (or nearly complementary) to the sequence of the probe. B. Nucleic acid sequencing

[0125] As shown in Figure 1, after cfDNA is extracted and isolated from a sample in steps 101 and 102, the cfDNA can be sequenced in steps 103 and 104. With or without pre-amplification, sample nucleic acids, optionally flanked by adapters, are generally subject to sequencing. Optionally, the sequencing method or commercially available format may include, for example, Sanger sequencing, high throughput sequencing, bisulfite sequencing, pyrosequencing, sequencing by synthesis, single molecule sequencing, nanopore-based sequencing, semiconductor sequencing, sequencing by ligation, sequencing by hybridization, RNA-Seq (Illumina), digital gene expression (Helicos), next generation sequencing (NGS), single molecule sequencing by synthesis (SMSS) (Helicos), massively parallel sequencing, clonal single molecule array (Solexa), shotgun sequencing, Ion Torrent, Oxford Nanopore, Roche Genia, primer walking, PacBio, SOLiD, Ion Torrent, or sequencing using nanopore platform. Sequencing reactions can be carried out in a variety of sample processing units, which may include multiple lanes, multiple channels, multiple wells, or other means of processing multiple sample sets substantially simultaneously. The sample processing unit may also contain multiple sample chambers to allow for the processing of multiple runs simultaneously.

[0126] Sequencing reactions can be performed on one or more nucleic acid fragment types or fragments that are known to contain markers for cancer or other diseases. Sequencing reactions can also be performed on any nucleic acid fragments present in a sample. Sequencing reactions can provide at least about 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9% or 100% sequence coverage of the genome. In other cases, the sequence coverage of the genome can be less than about 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9% or 100% of the genome.

[0127] Simultaneous sequencing reaction can be carried out using multiplex sequencing technology.In some embodiments, cell-free polynucleotide is sequenced at least about 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000 or 100,000 sequencing reactions.In other embodiments, cell-free polynucleotide is sequenced at less than about 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000 or 100,000 sequencing reactions.Sequencing reactions are typically carried out sequentially or simultaneously.Subsequent data analysis is generally carried out on all or part of sequencing reactions. In some embodiments, data analysis is performed on at least about 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, or 100,000 sequencing reactions. In other embodiments, data analysis may be performed on less than about 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, or 100,000 sequencing reactions. Exemplary read depths are about 1000 to about 50,000 reads per locus (base position).

[0128] In some embodiments, a nucleic acid population is prepared for sequencing by enzymatically forming blunt ends on double-stranded nucleic acids with single-stranded overhangs at one or both ends. In these embodiments, the population is typically treated with an enzyme having 5'-3' DNA polymerase activity and 3'-5' exonuclease activity in the presence of nucleotides (e.g., A, C, G, and T or U). Exemplary enzymes or catalytic fragments thereof that are optionally used include Klenow large fragment and T4 polymerase. For 5' overhangs, the enzyme typically extends the recessed 3' end on the opposing strand until it is flush with the 5' end to generate a blunt end. For 3' overhangs, the enzyme generally digests from the 3' end to the 5' end of the opposite strand, and sometimes beyond. If this digestion proceeds beyond the 5' end of the opposing strand, the gap can be filled by an enzyme with the same polymerase activity as used for the 5' overhang. The formation of blunt ends on double-stranded nucleic acids facilitates, for example, the binding of adaptors and subsequent amplification.

[0129] In some embodiments, the population of nucleic acids undergoes additional processing, such as converting single-stranded nucleic acids to double strands and / or converting RNA to DNA. These forms of nucleic acid are also optionally ligated to adaptors and amplified.

[0130] With or without prior amplification, the nucleic acid subjected to the process of forming blunt ends described above, and optionally other nucleic acids in the sample, can be sequenced to generate sequenced nucleic acids. A sequenced nucleic acid can refer to either the sequence (i.e., sequence information) of a nucleic acid or the nucleic acid whose sequence has been determined. Sequencing can be performed to provide sequence data for individual nucleic acid molecules in a sample, directly or indirectly from the consensus sequence of the amplification products of the individual nucleic acid molecules in the sample.

[0131] In some embodiments, the double-stranded nucleic acid with single-stranded overhangs in the sample after blunt end formation is ligated to an adaptor containing a barcode at both ends, and sequencing determines the nucleic acid sequence as well as the in-line barcode introduced by the adaptor. The blunt-ended DNA molecule is optionally ligated at least partially to the blunt end of a double-stranded adaptor (e.g., a Y-shaped or bell-shaped adaptor). Alternatively, the blunt ends of the sample nucleic acid and adaptor can be tailed with complementary nucleotides to facilitate ligation (e.g., sticky end ligation).

[0132] A nucleic acid sample is typically contacted with a sufficient number of adapters so that the probability that any two copies of the same nucleic acid will receive the same combination of adapter barcodes from adapters ligated at both ends is low (e.g., less than 1% or 0.1%). The use of adapters in this manner allows for the identification of a family of nucleic acid sequences that have the same start and stop points on the reference nucleic acid and are ligated to the same combination of barcodes. Such a family represents the sequence of the amplification product of the template / parent nucleic acid in the sample before amplification. The sequences of the family members can be edited to obtain a consensus nucleotide or a complete consensus sequence of the nucleic acid molecules in the original sample, modified by blunt end formation and adapter ligation. In other words, a nucleotide occupying a particular position of the nucleic acid in the sample is determined to be the consensus of the nucleotides occupying its corresponding position in the family member sequences. A family may include sequences of one or both strands of a double-stranded nucleic acid. When a family member includes sequences of both strands from a double-stranded nucleic acid, the sequences of one strand are converted to their complements in order to edit all sequences to obtain a consensus nucleotide or sequence. Some families include only a single member sequence. In this case, this sequence can be the sequence of the nucleic acid in the sample before amplification. Alternatively, families with only a single member sequence can be excluded from further analysis.

[0133] Nucleotide variations of sequenced nucleic acids can be determined by comparing the sequenced nucleic acids to a reference sequence. The reference sequence is often a known sequence, such as a known whole genome sequence or partial genome sequence from a subject (e.g., a whole genome sequence of a human subject). The reference sequence can be, for example, hG19 or hG38. The sequenced nucleic acid can represent a sequence determined directly for a nucleic acid in a sample, or a consensus of sequences of an amplification product of such a nucleic acid, as described above. The comparison can be performed at one or more designated positions on the reference sequence. A subset of sequenced nucleic acids can be identified that includes positions that correspond to the designated positions of the reference sequence when the respective sequences are maximally aligned. Within such a subset, it can be determined which sequenced nucleic acids, if any, include nucleotide variations at the designated positions, the length of a given cfDNA fragment based on where its endpoints (i.e., which are the 5' and 3' terminal nucleotides) map to the reference sequence, the offset of the midpoint of the given cfDNA fragment from the midpoint of the genomic region in the cfDNA fragment, and, if necessary, which include the reference nucleotide (i.e., the same as the reference sequence). A variant nucleotide can be called at a specified position if there is a number of sequenced nucleic acids in the subset that contain a nucleotide variant that exceeds a selected threshold. The threshold can be a simple number, for example, at least 1, 2, 3, 4, 5, 6, 7, 9 or 10 sequenced nucleic acids in the subset that contain the nucleotide variant, or a ratio of at least 0.5, 1, 2, 3, 4, 5, 10, 15 or 20 sequenced nucleic acids in the subset that contain the nucleotide variant, among other possibilities. The comparison can be repeated for any specified position of interest in the reference sequence. Sometimes, the comparison can be performed for specified positions that occupy at least about 20, 100, 200 or 300 consecutive positions on the reference sequence, for example, about 20-500 or about 50-300 consecutive positions.

[0134] Further details regarding nucleic acid sequencing, including the formats and uses described herein, can also be found in, 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); Astier et al., J Am Chem. Soc.,128(5):1705-10(2006), U.S. Patent No. 6,210,891, U.S. Patent No. 6,258,568, U.S. Patent No. 6,833,246, U.S. Patent No. 7,115,400, U.S. Patent No. 6,969,488, U.S. Patent No. 5,912,148, U.S. Patent No. 6,130,073, U.S. Patent No. 7,169,560, U.S. Patent No. 7,282,337, Nos. 7,482,120, 7,501,245, 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. i. Sequencing Panel

[0135] To improve the likelihood of detecting genomic regions of interest, and optionally tumors that exhibit mutations, the section of DNA sequenced may include a panel of genes or genomic sections that include known genomic regions. Selection of limited sections (e.g., limited panels) for sequencing can reduce the total sequencing required (e.g., the total amount of sequenced nucleotides). Sequencing panels can target multiple different genes or regions, for example, to detect a single cancer, a set of cancers, or all cancers. Alternatively, DNA can be sequenced by whole genome sequencing (WGS) or other unbiased sequencing methods without the use of a sequencing panel. Examples of panels and targets suitable for use in panels can be found in the epigenetic targets described in U.S. Provisional Patent Application No. 62 / 799,637, filed January 31, 2019, which is incorporated by reference in its entirety.

[0136] In some embodiments, a panel targeting multiple different genes or genomic regions (e.g., transcription factor binding regions, distal regulatory elements (DREs), repetitive elements, intron-exon junctions, transcription start sites (TSSs), etc.) is selected such that a determined percentage of subjects with cancer exhibits genetic variants or tumor markers in one or more different genes in the panel. The panel may be selected to limit the region for sequencing to a fixed number of base pairs. The panel may be selected to sequence a desired amount of DNA. The panel may further be selected to achieve a desired sequence read depth. The panel may be selected to achieve a desired sequence read depth or sequence read coverage for a certain amount of sequenced base pairs. The panel may be selected to achieve a theoretical sensitivity, theoretical specificity, and / or theoretical accuracy for detecting one or more genetic variants in a sample.

[0137] Probes for detecting panels of regions can include probes for detecting genomic regions of interest (hotspot regions) as well as nucleosome recognition probes (e.g., KRAS codons 12 and 13) and can be designed to optimize capture based on analysis of cfDNA coverage and fragment size variation influenced by nucleosome binding patterns and GC sequence composition. Regions as used herein can also include non-hotspot regions optimized based on nucleosome positions and GC models. Panels can include multiple subpanels including subpanels for identifying tissue of origin (e.g., use of published literature to define 50-100 baits representing genes with the most diverse transcriptional profiles across tissues (not necessarily promoters)), whole genome scaffolds (e.g., to identify ultraconserved genomic content and sparsely tiling across chromosomes with a handful of probes for copy number baselining purposes), transcription start sites (TSSs) / CpG islands (e.g., to capture differentially methylated regions (e.g., differentially methylated regions (DMRs)) in promoters of tumor suppressor genes (e.g., SEPT9 / VIM in colorectal cancer). In some embodiments, the marker for tissue of origin is a tissue-specific epigenetic marker.

[0138] Some examples of lists of genomic locations of interest can be found in Table 1 and Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include at least a portion of at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, at least 70, at least 75, at least 80, at least 85, at least 90, at least 95, or 97 of the genes in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, or 70 of the SNVs in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all the SNVs in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, or 18 of the CNVs of Table 1. In embodiments, the genomic locations used in the methods of the present disclosure include all of the CNVs of Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 1, at least 2, at least 3, at least 4, at least 5, or 6 of the fusions of Table 1. In embodiments, the genomic locations used in the methods of the present disclosure include all of the fusions of Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include at least a portion of at least 1, at least 2, or 3 of the indels of Table 1. In embodiments, the genomic locations used in the methods of the present disclosure include all of the indels of Table 1.In one embodiment, the genomic location used in the method of the present disclosure includes all genes, SNVs, CNVs, fusions and indels in Table 1. In some embodiments, the genomic location used in the method of the present disclosure includes at least a portion of at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, at least 70, at least 75, at least 80, at least 85, at least 90, at least 95, at least 100, at least 105, at least 110, or 115 of the genes in Table 2. In an embodiment, the genomic location used in the method of the present disclosure includes all genes in Table 2. In one embodiment, the genomic location used in the method of the present disclosure includes all genes in Table 1 and Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 5, at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, at least 60, at least 65, at least 70, or 73 of the SNVs in Table 2. In an embodiment, the genomic locations used in the methods of the present disclosure include all of the SNVs in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all of the SNVs in Table 1 and Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, or 18 of the CNVs in Table 2. In an embodiment, the genomic locations used in the methods of the present disclosure include all of the CNVs in Table 2. In one embodiment, the genomic locations used in the methods of the disclosure include all of the CNVs in Table 1 and Table 2. In some embodiments, the genomic locations used in the methods of the disclosure include at least 1, at least 2, at least 3, at least 4, at least 5, or 6 of the fusions in Table 2.In an embodiment, the genomic location used in the method of the present disclosure comprises all of the fusions in Table 2. In one embodiment, the genomic location used in the method of the present disclosure comprises all of the fusions in Table 1 and Table 2. In some embodiments, the genomic location used in the method of the present disclosure comprises at least a portion of at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, or 18 indels in Table 2. In an embodiment, the genomic location used in the method of the present disclosure comprises all of the indels in Table 2. In one embodiment, the genomic location used in the method of the present disclosure comprises all of the indels in Table 1 and Table 2. In one embodiment, the genomic location used in the method of the present disclosure comprises all of the genes, SNVs, CNVs, fusions, and indels in Table 2. In one embodiment, the genomic location used in the method of the present disclosure comprises all of the genes, SNVs, CNVs, fusions, and indels in Table 1 and Table 2. Each of these genomic locations of interest can be identified as a backbone or hotspot region for a given bait set panel. [Table 1] Table 2 [Table 2-1] [Table 2-2]

[0139] In some embodiments, one or more regions in the panel include one or more loci from one or more genes for detecting residual cancer after surgery. This detection can be performed earlier than existing cancer detection methods. In some embodiments, one or more genomic locations in the panel include one or more loci from one or more genes for detecting cancer in high-risk patient populations. For example, smokers have a much higher rate of lung cancer than the general population. In addition, smokers may develop other lung conditions that make cancer detection more difficult, such as the development of irregular nodules in the lungs. In some embodiments, the methods described herein detect the response of patients (especially in high-risk patients) to cancer therapy earlier than is possible with existing methods of cancer detection.

[0140] Genomic locations can be selected for inclusion in a sequencing panel based on the number of subjects with cancer that have a tumor marker in that gene or region. Genomic locations can be selected for inclusion in a sequencing panel based on the prevalence of subjects with cancer and tumor markers present in that gene. The presence of a tumor marker in a region can indicate that a subject has cancer.

[0141] In some cases, the panel may be selected using information from one or more databases. The information about the cancer may come from a cancer tumor biopsy or a cfDNA assay. The database may include information describing the population of sequenced tumor samples. The database may include information about mRNA expression in the tumor samples. The database may include information about regulatory elements or genomic regions in the tumor samples. The information about the sequenced tumor samples may include the frequency of various genetic variants and may describe the genes or regions in which the genetic variants are present. The genetic variants may be tumor markers. A non-limiting example of such a database is COSMIC. COSMIC is a catalog of somatic mutations found in various cancers. For a particular cancer, COSMIC ranks genes based on the frequency of mutations. Genes may be selected for inclusion in a panel by having a high frequency of mutations in a given gene. For example, COSMIC shows that 33% of the population of sequenced breast cancer samples have mutations in TP53 and 22% of the population of sampled breast cancers have mutations in KRAS. Other ranked genes, including APC, have mutations found in only about 4% of the population of sequenced breast cancer samples. TP53 and KRAS may be included in a sequencing panel based on having a relatively high frequency (e.g., compared to APC, which occurs at a frequency of about 4%) among sampled breast cancers. While COSMIC is provided as a non-limiting example, any database or set of information that associates cancer with tumor markers located in genes or gene regions may be used. In another example, of 1156 biliary tract cancer samples, as provided by COSMIC, 380 samples (33%) carried a mutation in TP53. Several other genes, such as APC, have mutations in 4-8% of all samples. Thus, TP53 may be selected for inclusion in a panel based on its relatively high frequency in a population of cholangiocarcinoma samples.

[0142] Genes or genomic sections may be selected for panels in which the frequency of tumor markers is significantly greater in the sampled tumor tissue or circulating tumor DNA than is found in a given background population. Combinations of genomic locations may be selected for inclusion in a panel such that at least a majority of subjects with cancer may have a tumor marker or genomic region present in at least one of the genomic locations or genes in the panel. Combinations of genomic locations may be selected based on data indicating that for a particular cancer or set of cancers, a majority of subjects have one or more tumor markers in one or more selected regions. For example, to detect cancer 1, a panel including regions A, B, C and / or D may be selected based on data indicating that 90% of subjects with cancer 1 have tumor markers in regions A, B, C and / or D of the panel. Alternatively, tumor markers may be shown to occur independently in two or more regions in subjects with cancer such that, in combination, tumor markers in two or more regions are present in a majority of the population of subjects with cancer. For example, to detect cancer 2, a panel including regions X, Y and Z may be selected based on data showing that 90% of subjects have tumor markers in one or more regions, and in 30% of such subjects the tumor marker is only detected in region X, and for the remainder of subjects in which the tumor marker is detected the tumor marker is only detected in regions Y and / or Z. Tumor markers present at one or more genomic locations previously shown to be associated with one or more cancers may be indicative or predictive of subjects having cancer if the tumor marker is detected in one or more of those regions 50% or more of the time. Computational approaches such as models that use the conditional probability of detecting cancer given the cancer frequency for a set of tumor markers in one or more regions may be used to predict which regions, alone or in combination, may be predictive of cancer.Other approaches for panel selection include the use of databases that describe information from studies using large panels and / or comprehensive genomic profiling of tumors by whole genome sequencing (WGS, RNA-seq, Chip-seq, bisulfate sequencing, ATAC-seq, etc.). Information gleaned from the literature can also describe pathways commonly affected and mutated in specific cancers. Panel selection can be further informed by the use of ontologies that describe genetic information.

[0143] The genes included in the panel for sequencing may include the complete transcribed region, promoter region, enhancer region, regulatory element and / or downstream sequence.To further increase the likelihood of detecting tumors that show mutations, only exons may be included in the panel.The panel may include all exons of the selected gene, or only one or more exons of the selected gene.The panel may include exons from each of a plurality of different genes.The panel may include at least one exon from each of a plurality of different genes.

[0144] In some embodiments, the panel of exons from each of a plurality of different genes is selected such that a determined proportion of subjects with cancer exhibits a genetic variant in at least one exon within the panel of exons.

[0145] At least one complete exon from each different gene in the panel of genes may be sequenced. The sequenced panel may include exons from multiple genes. The panel may include exons from 2-100 different genes, 2-70 genes, 2-50 genes, 2-30 genes, 2-15 genes, or 2-10 genes.

[0146] The selected panel may include various numbers of exons. The panel may include 2-3000 exons. The panel may include 2-1000 exons. The panel may include 2-500 exons. The panel may include 2-100 exons. The panel may include 2-50 exons. The panel may include 2-50 exons. The panel may include 300 or fewer exons. The panel may include 200 or fewer exons. The panel may include 100 or fewer exons. The panel may include 50 or fewer exons. The panel may include 40 or fewer exons. The panel may include 30 or fewer exons. The panel may include 25 or fewer exons. The panel may include 20 or fewer exons. The panel may include 15 or fewer exons. The panel may include 10 or fewer exons. The panel may include 9 or fewer exons. The panel may include up to 8 exons. The panel may include up to 7 exons.

[0147] A panel may include one or more exons from a plurality of different genes. A panel may include one or more exons from each of a proportion of a plurality of different genes. A panel may include at least two exons from each of at least 25%, 50%, 75% or 90% of the different genes. A panel may include at least three exons from each of at least 25%, 50%, 75% or 90% of the different genes. A panel may include at least four exons from each of at least 25%, 50%, 75% or 90% of the different genes.

[0148] The size of a sequencing panel can vary. A sequencing panel can be larger or smaller (in terms of nucleotide size) depending on several factors including, for example, the total amount of nucleotides sequenced for a particular region within the panel or the number of unique molecules sequenced. A sequencing panel can be 5 kb to 50 kb in size. A sequencing panel can be 10 kb to 30 kb in size. A sequencing panel can be 12 kb to 20 kb in size. A sequencing panel can be 12 kb to 60 kb in size. A sequencing panel can be at least 10 kb, 12 kb, 15 kb, 20 kb, 25 kb, 30 kb, 35 kb, 40 kb, 45 kb, 50 kb, 60 kb, 70 kb, 80 kb, 90 kb, 100 kb, 110 kb, 120 kb, 130 kb, 140 kb, or 150 kb in size. Sequencing panels can be less than 100 kb, 90 kb, 80 kb, 70 kb, 60 kb or 50 kb in size.

[0149] The panel selected for sequencing may include at least 1, 5, 10, 15, 20, 25, 30, 40, 50, 60, 80 or 100 genomic positions (e.g., each includes a genomic region of interest). In some cases, the genomic positions in the panel are selected such that the size of the positions is relatively small. In some cases, the regions in the panel have a size of about 10kb or less, about 8kb or less, about 6kb or less, about 5kb or less, about 4kb or less, about 3kb or less, about 2.5kb or less, about 2kb or less, about 1.5kb or less, or about 1kb or less. In some cases, genomic locations within a panel have a size of about 0.5 kb to about 10 kb, about 0.5 kb to about 6 kb, about 1 kb to about 11 kb, about 1 kb to about 15 kb, about 1 kb to about 20 kb, about 0.1 kb to about 10 kb, or about 0.2 kb to about 1 kb. For example, a region within a panel can have a size of about 0.1 kb to about 5 kb.

[0150] The panel selected herein can allow for deep sequencing sufficient to detect low frequency genetic variants (e.g., in cell-free nucleic acid molecules obtained from a sample). The amount of genetic variants in a sample can be referred to in terms of mutant allele frequency of a given genetic variant. Mutant allele frequency can refer to the frequency with which a mutant allele (e.g., not the most common allele) occurs in a given population of nucleic acids, such as a sample. A genetic variant with a low mutant allele frequency may be present relatively infrequently in a sample. In some cases, the panel allows for detection of a genetic variant with a mutant allele frequency of at least 0.0001%, 0.001%, 0.005%, 0.01%, 0.05%, 0.1% or 0.5%. The panel can allow for detection of a genetic variant with a mutant allele frequency of 0.001% or more. The panel can allow for detection of a genetic variant with a mutant allele frequency of 0.01% or more. The panel may allow detection of genetic variants present in a sample at frequencies as low as 0.0001%, 0.001%, 0.005%, 0.01%, 0.025%, 0.05%, 0.075%, 0.1%, 0.25%, 0.5%, 0.75%, or 1.0%. The panel may allow detection of tumor markers present in a sample at frequencies of at least 0.0001%, 0.001%, 0.005%, 0.01%, 0.025%, 0.05%, 0.075%, 0.1%, 0.25%, 0.5%, 0.75%, or 1.0%. The panel may allow detection of tumor markers in a sample at frequencies as low as 1.0%. The panel may allow detection of tumor markers in a sample at frequencies as low as 0.75%. The panel may allow detection of tumor markers in a sample at frequencies as low as 0.5%. The panel may allow for detection of tumor markers in a sample at a frequency as low as 0.25%. The panel may allow for detection of tumor markers in a sample at a frequency as low as 0.1%. The panel may allow for detection of tumor markers in a sample at a frequency as low as 0.075%.The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.05%. The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.025%. The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.01%. The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.005%. The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.001%. The panel may allow for detection of a tumor marker in a sample at a frequency as low as 0.0001%. The panel may allow for detection of a tumor marker in sequenced cfDNA at a frequency as low as 1.0% to 0.0001% in the sample. The panel may allow for detection of a tumor marker in sequenced cfDNA at a frequency as low as 0.01% to 0.0001% in the sample.

[0151] Genetic variants can be represented by the percentage of the population of subjects with disease (e.g., cancer).In some cases, at least 1%, 2%, 3%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95% or 99% of the population with cancer show one or more genetic variants in at least one of the regions of the panel.For example, at least 80% of the population with cancer can show one or more genetic variants in at least one of the genomic positions in the panel.

[0152] A panel can include one or more locations that include a genomic region of interest from each of one or more genes. In some cases, a panel can include one or more locations that include a genomic region of interest from each of at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 40, 50, or 80 genes. In some cases, a panel can include one or more locations that include a genomic region of interest from each of up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 40, 50, or 80 genes. In some cases, a panel can include one or more locations that include a genomic region of interest from each of about 1 to about 80, 1 to about 50, about 3 to about 40, 5 to about 30, 10 to about 20 different genes.

[0153] The position comprising the genomic region in the panel can be selected so that one or more epigenetic modification regions are detected.The one or more epigenetic modification regions can be acetylated, methylated, ubiquitinated, phosphorylated, sumoylation, ribosylation and / or citrullination.For example, the region in the panel can be selected so that one or more methylation regions are detected.

[0154] Regions in the panel can be selected so that they contain sequences that are differentially transcribed across one or more tissues.In some cases, the location that contains a genomic region can contain sequences that are transcribed at a higher level in a particular tissue compared to other tissues.For example, the location that contains a genomic region can contain sequences that are transcribed in a particular tissue but not in other tissues.

[0155] Genomic locations in the panel may include coding and / or non-coding sequences. For example, genomic locations in the panel may include one or more sequences in exons, introns, promoters, 3' untranslated regions, 5' untranslated regions, regulatory elements, transcription start sites, and / or splice sites. In some cases, regions in the panel may include other non-coding sequences, including pseudogenes, repeat sequences, transposons, viral elements, and telomeres. In some cases, genomic locations in the panel may include sequences in non-coding RNAs, such as ribosomal RNAs, transfer RNAs, Piwi-interacting RNAs, and microRNAs.

[0156] Genomic locations within the panel can be selected to detect (diagnose) cancer with a desired level of sensitivity (e.g., through detection of one or more genetic variants). For example, regions within the panel can be selected to detect cancer (e.g., through detection of one or more genetic variants) with at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% sensitivity. Genomic locations within the panel can be selected to detect cancer with 100% sensitivity.

[0157] Genomic locations within the panel can be selected to detect (diagnose) cancer with a desired level of specificity (e.g., through detection of one or more genetic variants). For example, genomic locations within the panel can be selected to detect cancer with at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9% specificity (e.g., through detection of one or more genetic variants). Genomic locations within the panel can be selected to detect one or more genetic variants with 100% specificity.

[0158] Genomic locations within the panel can be selected to detect (diagnose) cancer with a desired positive predictive value. Positive predictive value can be increased by increasing sensitivity (e.g., the likelihood of detecting an actual positive) and / or specificity (e.g., the likelihood of not mistaking an actual negative for a positive). As a non-limiting example, genomic locations within the panel can be selected to detect one or more genetic variants with a positive predictive value of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9%. Regions within the panel can be selected to detect one or more genetic variants with a positive predictive value of 100%.

[0159] Genomic positions within the panel can be selected to detect (diagnose) cancer with a desired degree of accuracy. As used herein, the term "accuracy" can refer to the ability of a test to distinguish between disease symptoms (e.g., cancer) and health symptoms. Accuracy can be quantified using measures such as sensitivity and specificity, predictive value, likelihood ratio, area under the ROC curve, Youden's index and / or diagnostic odds ratio.

[0160] Accuracy can be presented as a percentage, referring to the ratio between the number of tests that give a correct result and the total number of tests performed. Regions within the panel can be selected to detect cancer with at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% accuracy. Genomic locations within the panel can be selected to detect cancer with 100% accuracy.

[0161] The panel can be selected to be highly sensitive and detect low frequency genetic variants.For example, the panel can be selected to detect genetic variants or tumor markers present in samples with a frequency as low as 0.01%, 0.05% or 0.001% with at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9% sensitivity.The genome position in the panel can be selected to detect tumor markers present in samples with a frequency of 1% or less with a sensitivity of 70% or more. A panel may be selected to detect tumor markers with a frequency as low as 0.1% in a sample with a sensitivity of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9%. A panel may be selected to detect tumor markers with a frequency as low as 0.01% in a sample with a sensitivity of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9%. The panel may be selected to detect tumor markers with a frequency as low as 0.001% in a sample with a sensitivity of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9%.

[0162] The panel can be selected to be highly specific and detect low frequency genetic variants. For example, the panel can be selected to detect genetic variants or tumor markers present in samples with a frequency as low as 0.01%, 0.05% or 0.001% with at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5% or 99.9% specificity. The genomic positions within the panel can be selected to detect tumor markers present in samples with a frequency of 1% or less with a specificity of 70% or more. The panel can be selected to detect tumor markers in samples with a frequency as low as 99% with a specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.1%, 99.5% or 99.9%. A panel may be selected to detect tumor markers in a sample at a frequency as low as 99% with a specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.01%, 99.5%, or 99.9%. A panel may be selected to detect tumor markers in a sample at a frequency as low as 99% with a specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.001%, 99.5%, or 99.9%.

[0163] The panel can be selected to be highly accurate and detect low frequency genetic variants. The panel can be selected to detect genetic variants or tumor markers present in samples with frequencies as low as 0.01%, 0.05%, or 0.001% with at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% accuracy. Genomic positions within the panel can be selected to detect tumor markers present in samples with frequencies as low as 1% or less with 70% or more accuracy. The panel can be selected to detect tumor markers present in samples with frequencies as low as 99% with at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.1%, 99.5%, or 99.9% accuracy. A panel may be selected to detect tumor markers at frequencies as low as 99% in samples with at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.01%, 99.5%, or 99.9% accuracy. A panel may be selected to detect tumor markers at frequencies as low as 99% in samples with at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.001%, 99.5%, or 99.9% accuracy.

[0164] The panel can be selected to be highly predictive and detect low frequency genetic variants. The panel can be selected such that a genetic variant or tumor marker present in a sample at a frequency as low as 0.01%, 0.05%, or 0.001% can have a positive predictive value of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9%.

[0165] To capture more nucleic acid molecules within a sample, the concentration of the probe or bait used in the panel may be increased (2-6 ng / μL). The concentration of the probe or bait used in the panel may be at least 2 ng / μL, 3 ng / μL, 4 ng / μL, 5 ng / μL, 6 ng / μL or more. The concentration of the probe may be about 2 ng / μL to about 3 ng / μL, about 2 ng / μL to about 4 ng / μL, about 2 ng / μL to about 5 ng / μL, about 2 ng / μL to about 6 ng / μL. The concentration of the probe or bait used in the panel may be 2 ng / μL or more, and 6 ng / μL or less. In some cases, this may allow more molecules within a biological sample to be analyzed, thereby allowing lower frequency alleles to be detected.

[0166] In one embodiment, after sequencing, the sequence reads may be assigned a quality score. The quality score may be a representation of the sequence reads that indicates whether they may be useful in subsequent analysis based on a threshold. In some cases, some sequence reads are not of sufficient quality or length to perform a subsequent mapping step. Sequence reads with a quality score of at least 90%, 95%, 99%, 99.9%, 99.99% or 99.999% may be excluded from the data set of sequence reads. In other cases, sequence reads that are assigned a quality score of at least 90%, 95%, 99%, 99.9%, 99.99% or 99.999% may be excluded from the data set. Sequence reads that meet a certain quality score threshold may be mapped to a reference genome. After mapping alignment, the sequence reads may be assigned a mapping score. The mapping score may be a representation of the sequence reads that are mapped back to the reference sequence, indicating whether each position is uniquely mappable or not. Sequence reads having a mapping score of at least 90%, 95%, 99%, 99.9%, 99.99% or 99.999% may be excluded from the data set. In other cases, sequencing reads assigned a mapping score of less than 90%, 95%, 99%, 99.9%, 99.99% or 99.999% may be excluded from the data set. c.MAF decision

[0167] As shown in FIG. 1, after cfDNA sequencing of the sample at steps 103 and / or 104, one or more mutant allele fractions (MAFs) may be determined at steps 105 and / or 106. Some or all of the MAF determinations may be performed before variant classification 107 / 108, after variant classification 107 / 108, during variant classification 107 / 108, before variant filtering 109, after variant filtering 109, during variant filtering 109, or combinations thereof. Prior to step 103, the cfDNA may be end-repaired, ligated with adapters containing molecular barcodes, amplified, and concentrated. The amplification may incorporate a sample index. In embodiments, MAF values ​​may be determined for all variants or all somatic variants. In embodiments, MAF values ​​may be determined for less than all variants or less than all somatic variants. Variant allele fraction (VAF) is used interchangeably with MAF herein. The mutant allele fraction (MAF) represents the number of mutant molecules divided by the total number of molecules at a particular genomic location (i.e., molecular coverage):

number

[0168] The maximum fraction of diploid genes ("max frac_diploid") (minimum allelic imbalance) can be determined. The fraction of diploid genes ("frac_diploid") is a measure of the level of allelic imbalance across samples as determined by copy number. Samples with high levels of allelic imbalance are prone to germline / somatic misclassification. Thus, low levels of allelic imbalance (or high frac_diploid) are an indication of the confidence of the somatic classification call.

[0169] In one embodiment, a total coverage profile can be used to capture fold changes and therefore tumor fraction rather than individual genes. d. Mutant classification

[0170] Sequencing in steps 103 and 104 creates a plurality of sequence reads. In steps 107 and / or 108, the plurality of sequence reads can be analyzed to determine one or more variants and classify the one or more variants. In embodiments, some or all variant classifications can be determined before MAF determination 105 / 106, after MAF determination 105 / 106, during MAF determination 105 / 106, or a combination thereof. Variants can include, for example, single nucleotide variants (SNVs), indels, fusions, and copy number variants. Any known technique for variant calling can be used. In embodiments, the plurality of sequence reads from the sample can be assembled and / or mapped and aligned to a genomic location relative to a reference genome. In some embodiments, the plurality of sequence reads (assembled or not) can then be compared to a reference genome to determine how the plurality of sequence reads of the subject differ from the sequence reads of the reference genome. Such a process can determine the presence of one or more variants in the plurality of sequence reads. In some embodiments, the molecular barcodes and / or start and stop genomic locations of a nucleic acid molecule obtained from a plurality of sequence reads can be used to identify mutant molecules for which the sequence reads belonging to that molecule differ from a reference genome. Such a process can determine the presence of one or more variants in the plurality of sequence reads.

[0171] In one embodiment, common heterozygous SNPs are used to model the behavior of local germline allele counts, and may be called somatic variants if they deviate significantly from the observed germline mutant allele rate. To model both the mean and variance of mutant allele counts at common SNPs, a beta binomial model can be used. For example, the beta binomial model described in PCT / US2018 / 052087, the entirety of which is incorporated herein by reference, can be used. This is an improvement over simpler methods such as fixed MAF cutoffs or Poisson models, as they cannot adequately represent the variance of molecular counts. e. Mutant filtering

[0172] In one embodiment, one or more filtering processes may be applied to the sequence reads in step 109 to exclude sequence reads from further analysis, as shown in Figure 1. In one embodiment, some or all of the filtering may be applied before MAF determination 105 / 106, after MAF determination 105 / 106, during MAF determination 105 / 106, before variant classification 107 / 108, after variant classification 107 / 108, during variant classification 107 / 108, or a combination thereof.

[0173] In some embodiments, one or more somatic variants having a MAF of less than about 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, or 0.9% at the first and / or second time point may be excluded from further analysis. In some embodiments, one or more somatic variants having a mutant molecule count of less than 5, 10, 15, 20, 25, or 30 at the first and / or second time point may be excluded from further analysis. In some embodiments, one or more somatic variants having a coverage of less than 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, or 1000 at the first and / or second time point may be excluded from further analysis.

[0174] In an embodiment, copy number variants can be used to exclude sequence reads from further analysis.Copy number amplification can be determined as known in the art.In step 109, method 100 can exclude copy number amplification in genes that have either insufficient probe coverage or insufficient copy number (e.g., below 95% detection limit).

[0175] As an example, CNVs can be determined by analyzing sequence reads to generate chromosomal regions of coverage. Chromosomal regions can be divided into variable length windows or bins. Read coverage can be determined for each window / bin region. In one embodiment, a quantitative measure for sequencing read coverage is a measure that indicates the number of reads derived from a DNA molecule that corresponds to a locus (e.g., a specific position, base, region, gene, or chromosome from a reference genome). To associate a read with a locus, the read can be mapped or aligned to a reference. Software for mapping or alignment (e.g., Bowtie, BWA, mrsFAST, BLAST, BLAT) can associate sequencing reads with loci. After sequence read coverage is determined, a probabilistic modeling algorithm can be applied to convert the normalized nucleic acid sequence read coverage for each window / bin region into a discrete copy number situation. In some cases, the algorithm may include one or more hidden Markov models, dynamic programming, support vector machines, Bayesian networks, trellis decoding, Viterbi decoding, expectation maximization, Kalman filtering methodology, and neural networks. The individual copy number status of each window region can be utilized to identify copy number variation in the chromosomal region. In some cases, all adjacent window / bin regions with the same copy number can be merged into a segment to report the presence or absence of copy number variation status. In some cases, the various windows / bins can be filtered before being merged with other segments. Copy number variation can be used to report a percentage score indicating how much disease material (or nucleic acid with copy number variation) is present in the cell-free polynucleotide sample.

[0176] In one embodiment, the presence of CNVs in one or more genes may be used to exclude variants from further analysis. For example, variants with a threshold number of LDT reportable genes with copy numbers equal to or greater than the gene-specific 95% limit of detection (LoD) in either the T0 or T1 samples. The threshold may be about 10 to about 30. The threshold may be, for example, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, etc. In one embodiment, the threshold may be 19.

[0177] Copy number variation may indicate the fold change of a given variant. A Gaussian model can be used to determine the ratio of the fold change between time T0 and time T1, which may be used as an estimate of the molecular response score.

[0178] In an embodiment, if a subject does not have a somatic variant or does not have a variant that meets the criteria of the variant filtering process, the subject may be classified as unevaluable. In an embodiment, a subject classified as unevaluable may be further classified as a molecular responder. In an embodiment, a subject with low ctDNA at both time T0 and time T1 may be classified as unevaluable and further classified as a molecular responder. In an embodiment, a subject with low MAF at both time T0 and time T1 may be classified as unevaluable and further classified as a molecular responder. In an embodiment, a subject with low tumor fraction at both time T0 and time T1 may be classified as unevaluable and further classified as a molecular responder. Low MAF or low tumor fraction may refer to MAF or tumor fraction below the detection limit (e.g., below the 95% detection limit) or below the quantification limit. What constitutes low may depend on the panel design, but for example, MAF f 0.1, 0.2, or 0.3% may be considered low. i. Germline Filter

[0179] In the embodiment shown in FIG. 2, a germline filter 200 may be applied to the sequence reads. Some (e.g., less than all) or all of the steps shown in FIG. 2 may be performed in any combination and in any order. Samples collected over the course of a subject's treatment (e.g., samples collected at time T0 and time T1) may have different levels of tumor shedding and allelic imbalance, meaning that variant classification at steps 107 / 108 may tend to assign different somatic classifications to the same variant in the same subject. Since the goal of the molecular response is to track somatic variants over the course of treatment, classification discrepancies may be automatically resolved to appropriately remove germline variants from consideration by reclassifying variants. For example, a variant may be classified as somatic at time T0 and germline at time T1. For example, a variant may be classified as germline at time T0 and somatic at time T1. For example, a variant may be classified as germline at time T0 and not classified at time T1. For example, a variant may be classified as somatic at time T0 and not classified at time T1. Germline filter 200 is configured to resolve such discrepancies and reassign variant classifications.

[0180] As shown in Figure 2, in step 201, for at least one variant in a sequence read, a determination may be made as to whether the variant is a deleterious variant (e.g., a frameshift mutation or a nonsense mutation) in a tumor suppressor gene (TSG). For example, the variant may be compared to a database of known TSGs. If the variant is a deleterious variant in a TSG, the variant may be classified as somatic, regardless of the classification outcome in steps 107 / 108 (e.g., the classification would be changed from germline to somatic).

[0181] If the variant is not a deleterious variant of a TSG, the germline filter 200 may determine in step 202 the maximum MAF of the variant present in the sample and the maximum proportion of diploid genes for at least one variant in the sample. In step 203, if the maximum proportion of diploid genes for the variant (at one of the at least two time points) indicates that the variant is somatic and that the MAF for the variant (at one of the at least two time points) does not increase the maximum MAF, the variant may be classified as somatic, regardless of the classification result in step 107 / 108 (e.g., the classification would be changed from germline to somatic). In step 203, if the maximum proportion of diploid genes for the variant (at one of the at least two time points) indicates that the variant is germline and that the MAF for the variant (at one of the at least two time points) would increase the maximum MAF, the variant may be classified as germline, regardless of the classification result in step 107 / 108 (e.g., the classification would be changed from somatic to germline).

[0182] If, in step 203, the maximum percentage of diploid genes for the variant indicates that the variant is somatic and that the MAF for the variant increases the maximum MAF,-or-if the maximum percentage of diploid genes for the variant indicates that the variant is germline and that the MAF for the variant does not increase the maximum MAF, the germline filter 200 may determine in step 204 whether the variant is classified as somatic in another patient sample below a threshold percentage (at one of at least two time points). The threshold percentage may be at least about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, or 9%. If the variant is classified as somatic in another patient sample below a threshold percentage, the variant may be classified as somatic, regardless of the classification result in steps 107 / 108 (e.g., the classification would be changed from germline to somatic).

[0183] If the variant is not classified as somatic in less than 5% of other patient samples in step 204, the germline filter 200 may determine in step 205 whether the MAF of the variant (at one of at least two time points) is greater than another MAF in the sample. For example, the germline filter 200 may determine whether the MAF for the variant is at least about 2-fold greater, 3-fold greater, 4-fold greater, 5-fold greater, 6-fold greater, 7-fold greater, 8-fold greater, 9-fold greater, or at least 10-fold greater than one or more other MAFs in the same sample. The one or more other MAFs in the sample may be, for example, the next highest somatic MAF of the maximum MAF in the sample. If the MAF for the variant is greater than another MAF in the sample, the variant may be classified as germline, regardless of the classification outcome in steps 107 / 108 (e.g., the classification would be changed from somatic to germline).

[0184] Germline filter 200 may determine in step 205 whether the MAF for the variant (at one of at least two time points) is greater than another MAF in another sample. For example, germline filter 200 may determine whether the MAF for the variant is at least about 2-fold greater, 3-fold greater, 4-fold greater, 5-fold greater, 6-fold greater, 7-fold greater, 8-fold greater, 9-fold greater, or at least 10-fold greater than one or more other MAFs in another sample. The one or more other MAFs in another sample may be, for example, the maximum MAF of the other sample. If the MAF for the variant is greater than another MAF in another sample, the variant may be classified as germline, regardless of the classification result in step 107 / 108 (e.g., the classification will be changed from somatic to germline).

[0185] In step 205, if the MAF for the variant is not greater than another MAF in the sample and is not greater than another MAF in another sample, the germline filter 200 may classify the variant as germline regardless of the classification result in steps 107 / 108 (e.g., the classification would be changed from somatic to germline).

[0186] Those variants classified as germline may be excluded from further analysis, including, for example, MAF determination and / or MR scoring. In some embodiments, variants are classified as CHIP variants if they are classified as CHIP in at least one patient sample. ii.CHIP filter

[0187] cfDNA may contain aggregates of cfDNA from any cell type including tumors, blood cells, etc. Undetermined potential clonal hematopoietic mutations (CHIPs) may even be present in cfDNA. Common approaches for CHIP filtering leverage recurrent CHIP genes or hotspots curated by large public or internal cohort studies. However, these approaches do not address the challenge in identifying random CHIP mutations in plasma-only approaches. The remaining unfiltered CHIP variants bias the percentage change to 1 (unchanged) and thus result in inaccurate subsequent molecular response predictions. To filter private CHIP variants (e.g., variants that are CHIP but not previously documented or poorly documented in previous databases of known CHIP variants), mutation measurements between two time points can be used to cluster variants of similar percentage change. When patients undergo treatment, progression or response results in few somatic mutations, but the CHIP variants remain stable. By clustering mutations into clones, random CHIP variants can be found in clones that are enriched in known CHIP lists or that have stable percentage differences.

[0188] Therefore, provided herein is an improvement of CHIP filtering that leverages observations between two time points (T0 and T1) to cluster genomic mutations of clones with different percentage changes. CHIP filtering can group / cluster events into clones to estimate clonal load change%. The clustering procedure can start with each single event and then merge utilizing a novel clustering heuristic. Once the clonal load change% is determined using all events, each clone can be examined based on the mutant composition and clonal load change% to determine whether the mutant is a CHIP clone.

[0189] In one embodiment, genomic mutations / variants are clustered utilizing a novel agglomerative hierarchical clustering heuristic. The heuristic quantifies statistical dissimilarity between mutations / variants and clusters via a custom dissimilarity metric. An adjustable stopping rule is utilized that continues agglomeration until a minimum (or maximum depending on the metric) acceptable dissimilarity threshold is met. In one embodiment, the custom dissimilarity metric is a modification of the Bhattacharyya distance, such that a numerical integration is performed on the product (not square-rooted) of the scaled likelihoods of mutations / variants and / or clusters considered for merging at a given step of the clustering heuristic. The likelihood is scaled to numerically integrate to 1 over the support of the integral. For SNVs and indels, the likelihood is calculated on a beta-binomial model fit of the observed count data that informs the MAF decision of the variants being clustered. The variance of the beta-binomial model is set via an adjustable parameter. For CNVs, the likelihood is calculated on a Gaussian model fit of the observed fold change estimate of the mutation of interest, with the variability of the Gaussian model also set via an adjustable parameter. The aggregation of mutations is optionally performed in a novel manner such that clustering is performed via a stepwise approach in which a first set of mutations are clustered until a stopping rule is met, then a second set of mutations is introduced, and further aggregation steps are performed, possibly according to the same dissimilarity metric and stopping rule. In some circumstances, a third set of mutations is introduced in a similar manner after application of a clustering heuristic to the second set of mutations.

[0190] In the embodiment shown in FIG. 3, the CHIP filter 300 calculates in step 301 a scaled likelihood function for each mutation / variant in the sample.

number

number

number

number

number

number

number

number

[0191]

number

number

number

number

number

number

number

[0192] The approximate confidence interval is

number

number

number

number

[0193] In step 302, the set of mutations / variants is

number

number

number

number

number

number

number

number

number

[0194] In step 303, the number of clones and the associated percentage change between time points can be reported with a confidence interval. Clones with a percentage change between the first and second time points at or above a predefined threshold can be identified. If multiple clones are identified, clones with a percentage change close to 1 and / or clones with a specific known CHIP variant can be classified as potential CHIP variants. CHIP variants can be excluded from further analysis. In some embodiments, variants can be classified as CHIP variants if they are classified as CHIP in at least one patient sample.

[0195] FIG. 4 shows an example of the application of the CHIP filter 300. FIG. 4 corresponds to an example of an aggregation procedure. In this example, three eligible mutations are identified. The leftmost panel of FIG. 4 displays the scaled likelihood function of each mutation (y-axis) on the support of R (x-axis). Assume that the mutation corresponding to the first likelihood (line 403) of the scaled likelihood function for each mutation is a known CHIP mutation. The most similar mutation in the left panel is marked with an asterisk. The middle panel displays the aggregate likelihood resulting from the merging of the first likelihood (line 403) and the second likelihood (line 401) of the scaled likelihood function of the clones in the left panel. The third likelihood (line 402) of the scaled likelihood function of each clone from the left panel has a likelihood function that is unchanged by aggregation. The right panel displays the final clonality. The composition of the second likelihood (line 401) clones is 50% CHIP, so the second likelihood (line 401) clones can be identified as putative CHIPs. This results in the final value of R being defined solely by the third likelihood (line 402) clone.

[0196] In an embodiment, FIG. 5 shows a flow chart outlining exemplary method steps for identifying clonal hematopoietic variants in a subject with cancer, according to some embodiments. As shown, the method 500 includes determining a tumor burden change (R) relative to a tumor proportion change P(R) for each of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from the subject at a first and second time point to generate a set of tumor burden changes (step 501). Furthermore, the method 500 also includes identifying one or more resistance signatures corresponding to one or more clonal hematopoietic variants from the set of tumor burden changes (step 502). f.MR score

[0197] Returning to FIG. 1, method 100 may proceed to determine an MR score in step 110. In one embodiment, the MR score may be determined using MAF values ​​associated with somatic variants remaining after variant filtering in step 109. In an embodiment, MAF values ​​of all somatic variants may be used. In an embodiment, MAF values ​​of less than all somatic variants may be used. As described in steps 105 / 106, MAF may be determined for a plurality of somatic variants from sequence reads made from targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at T0 (e.g., before treatment) and T1 (e.g., during treatment) to generate a set of first and second MAFs for somatic variants in the plurality of somatic variants. The MR score may be expressed as a proportion or percentage. As shown in FIG. 6A, the MR score may be determined according to method 600. Method 600 may include, in step 601, determining a ratio of a first MAF and a second MAF for somatic variants in a plurality of somatic variants to generate a set of MAF ratios and a corresponding standard deviation for the MAF ratios in the set of MAF ratios. In some embodiments, the standard deviation may be utilized as a basis for reporting the MR score. For example, the standard deviation of the MR score based on the individual standard deviation of at least one variant may be used to determine a confidence interval and subsequent cutoff for sample evaluability. In some embodiments, the cutoff may be at least 0.1, 0.15, 0.2, 0.3, 0.4 or 0.5. In step 602, a weighted average of MAF ratios may be determined for a subject using the following formula:

number

number

number

[0198] In one embodiment, in addition to or instead of using the weighted average of MAF ratios as the MR score, a method is disclosed that clusters variants based on MAF ratios, calculates the total MAF ratio of the clusters, and then uses either a single selected cluster ratio or the weighted average of the cluster ratios as the MR score. Clustering can be performed by combining variant pairs with overlapping MAF ratio distributions, or by other clustering methods. The single selected cluster can be one that contains known cancer driver variants or one that does not have known clonal hematopoietic variants. The weight of the cluster can also depend on the presence of known cancer driver variants or the maximum VAF or number of variants in the cluster.

[0199] As shown in FIG. 6B, the MR score may be determined according to method 610. Method 610 may include, in step 601, determining a weighted average of the first MAF and the weighted average of the second MAF for the somatic variants in the plurality of somatic variants, and the corresponding standard deviation for the weighted MAF ratio. In some embodiments, the standard deviation may be utilized as a criterion for reporting the MR score. For example, the standard deviation of the MR score based on the individual standard deviation of at least one variant may be used to determine a confidence interval and subsequent cutoff of sample evaluability. In some embodiments, the cutoff may be at least 0.1, 0.15, 0.2, 0.3, 0.4 or 0.5. In step 612, the ratio of the weighted averages of the MAFs may be determined for the subject. The confidence interval as the variance of the ratio. For example, the confidence interval may be determined using the following formula: R=A / B:var(R)~=var(B) / A^2+var(A)*B^2 / A^4 (where A and B are the weighted average MAF at time points 1 and 2, respectively).

[0200] Clusters can be weighted based on the strength of evidence. For example, max-VAF can indicate which are primary clones, the number of non-CHIP variants can weight clusters with stronger signals, and driver weight can increase weight or select clusters containing drivers for that particular cancer type or molecular subtype. The weighting applied can be, for example, applying a greater weight to variants known to be drivers for a particular cancer type or molecular subtype. In one embodiment, the weighting can be based on the maximum VAF (any sample), the number of non-CHIP variants, and / or the driver weight (tumor type specific; defined in the configuration file). In another embodiment, the weighting applied can be, for example, weighting somatic variants equally.

[0201] In one embodiment, classification as molecular responder or molecular non-responder may depend on mutant VAF and mutant weight.For example, if MR score is the ratio of average VAF, the higher VAF (i.e., more clonal mutants) are more likely to be dominant.If MR score uses mutant weight, the mutant with higher weight (e.g., driver mutant) may be dominant.

[0202] The weighted average of the obtained MAF ratios as described in FIG. 6A or the ratio of the weighted average of MAFs as described in FIG. 6B is the MR score of the subject. Such MR scores incorporate the variance of MAFs into the molecular response calculation. This ensures that the molecular response score contains the correct variance and contributes to drawing correct conclusions from the molecular response. The MR score can be considered as a "numerically stable" ratio of the average MAFs, which appropriately weights the changes in MAFs based on the accuracy of the MAFs and is less susceptible to overconfident and erroneous results when the MAFs are fluctuating around the limit of detection (LOD). The MR score can be compared to a threshold to determine whether the subject is responding or not responding to the treatment. The threshold can be, for example, about 25% to about 75% and / or can include it. In some embodiments, the weighting can be based on either the VAF accuracy (e.g., location, hotspot area, coverage depth, etc.) or prior knowledge of the importance of that variant to the tumor (e.g., known driver or resistance mutations, or variants of uncertain (or unknown) significance).

[0203] To provide a simple example to illustrate an aspect of the problem addressed by the MR scoring methods presented herein, consider a subject with one variant detected, with a MAF of 0.3% at baseline (T0) and 0.1% during treatment (T1), and coverage at that variant position of 3000 molecules. Using existing methods, the molecular response score would be:

number

[0204] To provide a simple example to illustrate an aspect of the problem addressed by the MR scoring method presented herein, consider a subject in which two variants (a and b) are detected and have a MAF of a=0.1% and b=8.0% at baseline (T0) and MAF a=0.3% and b=2.0% during treatment (T1). Using existing methods of taking the average of the ratios, the molecular response score is

number

number

[0205] To provide a simple example to illustrate an aspect of the problem addressed by the MR scoring methods presented herein, consider a subject in which two variants (a and b) are detected and have a MAF of a=0.3% and b=0.0% at baseline (T0), and MAF a=0.0% and b=0.3% on treatment (T1). If existing methods are used to evaluate only variants above 0.3% at baseline, the molecular response score would be:

number

number

[0206] The method 100 may include administering one or more therapies to the subject based at least on the molecular response score. Exemplary therapies are further disclosed herein. In some embodiments, the method 100 includes comparing the molecular response score of the subject having cancer to a predefined cutoff point to identify that the subject is likely to be a responder to one or more therapies (such as, for example, immunotherapy) for the cancer if the molecular response score is below the predefined cutoff point, or that the subject is likely to be a non-responder to one or more therapies for the cancer if the molecular response score is above the predefined cutoff point. In some embodiments, the method 100 includes administering one or more therapies for the cancer to the subject in consideration of the molecular response score. In some embodiments, the method 100 includes discontinuing administering one or more therapies for the cancer to the subject in consideration of the molecular response score. In some embodiments, the method 100 includes using the molecular response score as a prognostic and / or predictive biomarker for the subject.

[0207] In other exemplary embodiments, variance is incorporated into the molecular response calculation by simulation or sampling from the variance distribution of at least one variant to calculate molecular response variance. As further disclosed herein, some applications include weighting variants based on their importance in the tumor or the tumor's likelihood for clonal hematopoiesis. Some embodiments include integrating multiple genomic data sources to estimate tumor fraction (rather than simply relying on variant (e.g., SNV, indel, and fusion) VAF), coverage (e.g., copy number), off-target coverage, and / or methylation, among other genomic data sources.

[0208] In some embodiments, the method includes using one or more additional genomic data sources to determine a molecular response score for a subject with cancer. In some embodiments, the additional genomic data sources include one or more coverage, off-target coverage, epigenetic signature, tumor mutation burden, and / or microsatellite instability score. For a data source, there may be a calculation of tumor fraction based on the data source, and the calculated tumor fraction may be combined across data sources (e.g., using a weighted average to incorporate the confidence of the data source in the tumor fraction of that particular sample), and then the overall tumor fraction estimate in the sample may be combined to calculate the overall molecular response. In some embodiments, the epigenetic signature includes cfNA fragment length, location, and / or endpoint density distribution. In some embodiments, the epigenetic signature includes the epigenetic status or state represented by one or more epigenetic loci in a given targeted genomic region. In some embodiments, the epigenetic status or state includes the presence or absence of methylation, hydroxymethylation, acetylation, ubiquitination, phosphorylation, sumoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone alterations.

[0209] Although the method is described in the context of FIG. 1 and a first time T0 and a second time T1, it should be understood that more than two time points are contemplated, e.g., for longitudinal monitoring. As shown in FIG. 7, at the first time T0, baseline cfDNA may be obtained from one or more baseline samples obtained from one or more subjects before treatment and at the second time T1, or at any subsequent time T nIn the method, the on-treatment cfDNA may be obtained from one or more on-treatment samples obtained from one or more subjects after treatment. Time T1 can be any amount of time after time T0, including, for example, 1-24 hours, 1-180 days, 1-12 weeks, 1-25 weeks, 1-30 weeks, etc. Additionally, method 100 can include any number of times after times T0, T1, ..., T n For example, samples may be taken at time T1 and time T2, with the samples taken at both times being samples during treatment. In another example, samples may be taken at time T1 and time T2, with the sample taken at time T1 representing a sample during treatment and the sample taken at time T2 representing a sample without treatment.

[0210] In one embodiment, the dosage of therapy administered to the subject can be adjusted based on the molecular response score. For example, the molecular response score can indicate that the subject is not responding to the first treatment, and the dosage of the first treatment can be increased in response. In an embodiment, an alternative therapy can be identified based on the molecular response score. For example, the molecular response score can indicate that the subject is not responding to the first treatment, and then the subject can receive a second treatment instead of or in addition to the first treatment. In one embodiment, the molecular response score can be determined for subjects in a clinical trial, and the molecular response score can be determined for subjects receiving a placebo and subjects receiving a treatment. The molecular response scores of the two categories of subjects can be compared to evaluate the treatment.

[0211] In another example, placebo and treatment may be generalized to two arms of a clinical trial comparing different combinations of drugs. The threshold or cutoff may be specific to the use case: the use case may require clearance (MR=0), or the use case may require a certain level of decrease or increase in ctDNA levels.

[0212] FIG. 8 illustrates an exemplary practical application of molecular response scores for patient stratification. An advanced cancer patient may have a baseline MAF determined at time T0, prior to treatment. After 4-10 weeks of treatment, an advanced cancer patient may have an on-treatment MAF determined at time T1. The resulting molecular response score may indicate that the patient's ctDNA is decreasing, in which case the patient should continue treatment with the primary investigational drug. The resulting molecular response score may indicate that the patient's ctDNA is increasing, in which case the patient should continue treatment with the primary investigational drug (or placebo) if the patient is in the control group. If instead the patient's ctDNA is increasing, the patient should receive one or more therapies added to their treatment plan, a change in therapy, or a change in dose of the primary investigational drug.

[0213] FIG. 9 illustrates an exemplary practical application of the molecular response score for clinical trial enrichment. An advanced cancer patient eligible for standard of care (SOC) treatment may have a baseline MAF determined at time T0 prior to SOC treatment. After 4-10 weeks of SOC treatment, an advanced cancer patient may have an on-treatment MAF determined at time T1. The resulting molecular response score may indicate that the patient's ctDNA is decreasing, in which case the patient should continue treatment with the SOC treatment. The resulting molecular response score may indicate that the patient's ctDNA is increasing, in which case the patient may be determined eligible for treatment with a clinical trial drug.

[0214] FIG. 10 shows a practical application of the molecular response score for prospective patient stratification and escalation for the MSKCC trial of osimertinib + / - chemotherapy in EGFR-positive non-small cell lung cancer (NSCLC) patients. A patient newly diagnosed with EGFR-positive NSCLC may have a baseline MAF determined at time T0 before using osimertinib. After one cycle of osimertinib, the patient may have an on-treatment MAF determined on day 1 of cycle 2 of osimertinib. The resulting molecular response score, based on the EGFR driver alone, may indicate that an EGFR driver is not detected, in which case the patient should continue treatment with osimertinib alone. The resulting molecular response score, based on the EGFR driver alone, may indicate that an EGFR driver is detected, in which case the patient should continue treatment with osimertinib, carboplatin and pemetrexed.

[0215] Aspects of these methods are further illustrated in FIG. 11. As shown, method 1100 includes determining mutant allele frequencies (MAFs) of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in samples obtained from a subject at a first (e.g., before treatment) and a second (e.g., during treatment) time point to generate a set of first and second MAFs for variants of the plurality of variants (step 1101). Method 1100 also includes calculating ratios of the first and second MAFs for variants in the plurality of variants to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios in the set of MAF ratios (step 1102). Additionally, method 1100 also includes calculating a weighted average (step 1103) and confidence interval of the MAF ratios to determine a molecular response score for a subject with cancer.

[0216] In some embodiments, the method 1100 includes comparing a molecular response score of a subject having cancer to a predefined cutoff point and identifying that the subject is likely to be a responder to one or more therapies for the cancer (such as, for example, immunotherapy) if the molecular response score is below the predefined cutoff point, or that the subject is likely to be a non-responder to one or more therapies for the cancer if the molecular response score is at or above the predefined cutoff point. In some embodiments, the method 1100 includes administering to the subject one or more therapies for the cancer in consideration of the molecular response score. In some embodiments, the method 1100 includes discontinuing administration of one or more therapies for the cancer to the subject in consideration of the molecular response score. In some embodiments, the method 1100 includes using the molecular response score as a prognostic and / or predictive biomarker for the subject.

[0217] In some embodiments, method 1100 includes using molecular counts to calculate a standard deviation for the MAF ratios in the set of MAF ratios. In some embodiments, method 1100 includes propagating the variance through the MAF ratios in the set of MAF ratios. In some embodiments, method 1100 includes excluding one or more germline and / or clonal hematopoietic variants when determining mutant allele frequencies (MAFs) for a plurality of variants. Examples of methods for excluding germline and CHIP variants are described further herein. In some embodiments, method 1100 includes excluding one or more somatic variants having a MAF that is less than about 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, or 0.9% at the first and / or second time points. In some embodiments, the method comprises excluding one or more somatic variants having a mutant molecular count of less than 5, 10, 15, 20, 25 or 30 at the first and / or second time point. In some embodiments, the method comprises excluding one or more somatic variants having a coverage of less than 300, 400, 500, 600, 700, 800, 900 or 1000 at the first and / or second time point. In some of these embodiments, the first time point comprises a pre-treatment time point and the second time point comprises a treatment initiation time point or a post-treatment time point.

[0218] In some embodiments, the method disclosed herein comprises generating sequence information from nucleic acid molecules obtained from one or more tissues or cells in a sample.In some embodiments, the method disclosed herein comprises generating sequence information from cell-free nucleic acid (cfNA) in a sample obtained from a subject.In some embodiments, cfNA comprises circulating tumor DNA (ctDNA).

[0219] In some embodiments, the ratio comprises a second MAF to a first MAF for a variant in the plurality of variants. In some embodiments, the method 1100 comprises calculating a weighted average of the MAF ratios using the following formula: Sum[weight*ratio] / sum[weight], (where the weight is 1 / range for a given variant among multiple variants) 2 (wherein the range is the difference between the values ​​of the first and second MAF for a given variant in the plurality of variants, and the ratio is a given MAF ratio within the set of MAF ratios). In some embodiments, the method 1100 includes calculating the confidence interval using the following formula: Weighted average of MAF ratios + / - sqrt [ratio variance], (where the specific variance is 1 / sum[weight]).

[0220] In some embodiments, the variants include one or more single nucleotide variants (SNVs), insertion / deletion mutations (indels), gene amplifications, and / or gene fusions. In some embodiments, the method 1100 includes determining a molecular response score for a subject with cancer using one or more additional genomic data sources. In some embodiments, the additional genomic data sources include one or more coverage, off-target coverage, epigenetic signatures, and / or microsatellite instability scores. In some embodiments, the epigenetic signatures include cfNA fragment lengths, positions, and / or endpoint density distributions. In some embodiments, the epigenetic signatures include epigenetic status or conditions represented by one or more epigenetic loci within a given targeted genomic region. In some embodiments, the epigenetic status or state includes the presence or absence of methylation, hydroxymethylation, acetylation, ubiquitination, phosphorylation, sumoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone alterations.

[0221] To further illustrate, FIG. 12A is a flow chart that generally illustrates an exemplary method 1200. As shown, the method 1200 includes determining mutant allele frequencies (MAFs) of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in samples obtained from a subject at a first and second time point to generate a set of first and second MAFs for variants of the plurality of variants (step 1201). The method 1200 also includes calculating ratios of the first and second MAFs for variants in the plurality of variants to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios in the set of MAF ratios (step 1202), and calculating weighted averages and confidence intervals of the MAF ratios to determine a molecular response score for the subject (step 1203). In some embodiments, the standard deviation can be used as an estimate of the confidence interval. In some embodiments, the standard deviation can be used as a basis for reporting the molecular response score. Additionally, method 1200 also includes administering one or more therapies to the subject based at least on the molecular response score (step 1204). Exemplary therapies are further disclosed herein.

[0222] FIG. 12B is a flow chart that generally illustrates an exemplary method 1210. As shown, the method 1210 includes determining mutant allele frequencies (MAFs) of a plurality of variants from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from a subject at a first and a second time point to generate a set of first and second MAFs for variants of the plurality of variants (step 1211). The method 1210 includes, in step 1212, determining a representative value obtained from the MAFs of the considered somatic variants for a time point (i.e., the first time point and the second time point). It is understood that the representative value can be one of, but not limited to, the mean, the median, or the mode. The method 1210 includes, in step 1213, determining a ratio between the representative value at the first time point and the representative value at the second time point. The method 1210 may include calculating the standard deviation of the central tendency ratio using the standard deviation of the considered MAFs. In some embodiments, the representative value can be the mean or the median. In some embodiments, the representative value may be the mean. In some embodiments, the representative value may be the median. In some embodiments, the method 1210 comprises, in step 1212, determining the mean of the MAF of the somatic variant considered for each time point (i.e., the first time point and the second time point), and, in step 1213, calculating the ratio of the mean obtained at the first time point to the mean obtained at the second time point, and calculating the standard deviation of the mean ratio using the standard deviation of each MAF considered. In some embodiments, the molecular response score can be calculated from the ratio of the mean obtained at the first time point to the mean obtained at the second time point. In some embodiments, the method 1210 comprises, in step 1212, determining the median of the MAF of the somatic variant considered for each time point (i.e., the first time point and the second time point), and, in step 1213, calculating the ratio of the median obtained at the first time point to the median obtained at the second time point, and calculating the standard deviation of the median ratio using the standard deviation of each MAF considered. In some embodiments, the molecular response score can be calculated from the ratio of the median value obtained at the first time point to the median value obtained at the second time point.In some embodiments, the standard deviation can be used as an estimate of the confidence interval. In some embodiments, the standard deviation can be used as a basis for reporting the molecular response score. Further, the method 1210 also includes administering one or more therapies to the subject based at least on the molecular response score (step 1214). Exemplary therapies are further disclosed herein.

[0223] Typically, methods for determining molecular response scores include filtering out CHIP variants. For example, molecular response is typically measured by the allele frequency of genomic alterations (e.g., small variants between two time points) that represent tumor percentage changes. Given that cfDNA signals are aggregations of signals from essentially any cell type, including tumors, blood cells, etc., many studies have shown the presence of undetermined potential clonal hematopoietic (CHIP) variants in cfDNA samples. Common approaches for CHIP filtering frequently leverage recurrent CHIP genes or hotspots curated by various data sources. However, it remains a challenge to identify random CHIP mutations using a plasma-only approach. The remaining unfiltered CHIP variants typically bias the percentage change to 1 (unchanged), thus resulting in inaccurate molecular response predictions or scores. Thus, in some embodiments, the methods disclosed herein use a model that leverages observations between two time points to cluster genomic mutations of clones with distinct percentage changes. To group mutations, these approaches typically leverage the mutant allele counts and total counts for the variants from two time points and construct a probability density function of the tumor fraction change R as P(R).

[0224] As a further illustration, Figure 13 is a flow chart outlining exemplary method steps for identifying clonal hematopoietic variants in a subject with cancer, according to some embodiments. As shown, method 1300 includes calculating a probability density function of tumor percentage change P(R) of variants of a plurality of variants from sequence information generated from targeting nucleic acids associated with one or more cancer types in samples obtained from a subject at a first and second time point (step 1301). As further shown, method 1300 also includes grouping one or more variants into one or more clones by P(R) (step 1302), generating an updated P(R) for clones of clones (step 1303), and identifying one or more clones having a percentage change between the first and second time points at or above a predetermined threshold (step 1304).

[0225] In other embodiments, the present disclosure provides a method for identifying and excluding germline variants, or otherwise resolving somatic classification discrepancies, when determining molecular response scores.For example, one problem is that samples collected over the course of a patient's treatment typically have different levels of tumor shedding and allelic imbalance, which means that the somatic variant caller of a given bioinformatics pipeline may arrive at different somatic classifications for the same variant in the same patient.Since the purpose of molecular response determination is to track somatic variants over the course of treatment, any classification discrepancies should be resolved in order to properly remove germline variants from consideration.

[0226] To illustrate, Figure 14 is a flow chart outlining exemplary method steps for identifying variants in a subject with cancer, according to some embodiments. As shown, method 1400 includes determining a mutant allele frequency (MAF) for a given variant from sequence information generated from a targeting nucleic acid associated with one or more cancer types in a sample obtained from a subject (step 1401). Method 1400 can utilize the determined MAF for a given variant to identify the given variant as a germline or somatic variant. In some embodiments, method 1400 can utilize the baseline MAF and subsequent on-treatment MAF of a given variant to classify the given variant as a germline or somatic variant, or to change a previous classification of the given variant. Method 1400 may also include identifying a given variant as a germline variant if the MAF of the given variant increases (at one of at least two time points) the maximum MAF of a sample containing the maximum proportion of diploid genes (max frac_diploid) (i.e., has the least allelic imbalance) and / or if the MAF of the given variant is at least about 2-fold greater, 3-fold greater, 4-fold greater, 5-fold greater, 6-fold greater, 7-fold greater, 8-fold greater, 9-fold greater, or at least 10-fold greater than one or more other MAFs (e.g., maximum MAF in the sample) determined from a sample obtained from the subject or another patient sample. In some embodiments, a given variant is classified as somatic if it does not increase the maximum MAF of the sample (e.g., compared to another MAF) at one of at least two time points where max frac_diploid is somatic. In some embodiments, a given variant is classified as germline when it increases the maximum MAF and the sample with max frac_diploid is germline. In some embodiments, the method 1400 includes classifying a given variant as somatic if the variant is determined to be a deleterious variant (e.g., a frameshift mutation or a nonsense mutation) in a tumor suppressor gene (TSG).In some embodiments, a given variant is classified as somatic if it is found in less than about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, or 9% of any given sample. In some embodiments, a given variant is classified as germline if the associated discrepancy is not resolved by method 1400. In these embodiments, variants are typically excluded from further consideration in determining a given molecular response score. In some embodiments, variants are classified as CHIP variants if they are classified as CHIP in at least one patient sample.

[0227] 15 is a flow chart that generally illustrates a method 1500 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 1501). As further illustrated, the method 1500 also includes classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline (step 1502), determining a MAF ratio (step 1503), determining a weighted average of the MAF ratios (step 1504), determining a confidence interval associated with the weighted average of the MAF ratios (step 1505), and outputting the weighted average of the MAF ratios and the confidence interval (step 1506). It is understood that the first plurality of sequence reads can be determined before administering a therapy, and the second plurality of sequence reads can be determined after administering a therapy. The classification of the plurality of variants in the first and second plurality of sequence reads as somatic or germline in step 1502 may be performed as described herein, for example, as described with respect to FIG. 2. In one embodiment, at least two variants of the plurality of variants are classified as somatic. It is also understood that the determination of the MAF ratio (step 1503) may be determined for at least one variant of the plurality of variants classified as somatic and based on the first MAF and the second MAF. The first MAF may be determined using the variant in the first plurality of sequence reads at a pre-treatment time point, and the second MAF may be determined using the same variant in the second plurality of sequence reads at a post-treatment time point. The first MAF and the second MAF may be determined for the same variant in both the first and second plurality of sequence reads. It is further understood that the determination of the weighted average of the MAF ratio (step 1504) may be for a subject. It is further understood that the determination of the confidence interval associated with the weighted average of the MAF ratios (step 1505) may be based on the weighted average of the MAF ratios. Finally, it is understood that the weighted average of the MAF ratios and the confidence interval can be output as a molecular response score.

[0228] 16 is a flow chart that generally illustrates a method 1600 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 1601). As further illustrated, the method 1600 also includes classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline (step 1602), determining a weighted average of the first MAF and a weighted average of the second MAF (step 1603), determining a ratio of the weighted average of the first MAF and the weighted average of the second MAF (step 1604), determining a confidence interval (step 1605), and outputting the ratio of the weighted average of the first MAF and the weighted average of the second MAF and the confidence interval (step 1606). It is understood that the first plurality of sequence reads can be determined before administering a therapy, and the second plurality of sequence reads can be determined after administering a therapy. The classification of the plurality of variants in the first and second plurality of sequence reads as somatic or germline in step 1602 may be performed as described herein, for example, as described with respect to FIG. 2. In one embodiment, at least two variants of the plurality of variants are classified as somatic. A first MAF may be determined using a variant in the first plurality of sequence reads at a pre-treatment time point, and a second MAF may be determined using the same variant in the second plurality of sequence reads at a post-treatment time point. A first MAF and a second MAF may be determined for the same variant in both the first and second plurality of sequence reads. It is also understood that the determination of the weighted average of the first MAF and the weighted average of the second MAF (step 1603) may be determined for at least one variant of the plurality of variants classified as somatic and based on the first MAF and the second MAF. It is further understood that the determination of the ratio of the weighted average of the first MAF to the weighted average of the second MAF (step 1604) can be for the subject. It is further understood that the determination of the confidence interval (step 1605) can be based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF. Finally, it is understood that the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval can be output as a molecular response score.

[0229] 17 is a flow chart that generally illustrates a method 1700 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 1701). As further shown, the method 1700 also includes classifying a plurality of variants in the first plurality of sequence reads as somatic or germline (step 1702), classifying a plurality of variants in the second plurality of sequence reads as somatic or germline (step 1703), reclassifying at least one variant of the plurality of variants to resolve a classification discrepancy between the first plurality of sequence reads and the second plurality of sequence reads (step 1704), determining a first mutant allele fraction (MAF) (step 1705), determining a second MAF (step 1706), and determining a molecular response score (1707). It is understood that the first plurality of sequence reads can be determined before administering a therapy and the second plurality of sequence reads can be determined after administering a therapy. The classification of the plurality of variants in the first and second plurality of sequence reads as somatic or germline in step 1703 may be performed as described herein, for example, as described with respect to FIG. 2. The first MAF may be determined using the variants in the first plurality of sequence reads at a time point before treatment, and the second MAF may be determined using the same variants in the second plurality of sequence reads at a time point after treatment. The first MAF and the second MAF may be determined for the same variant in both the first and second plurality of sequence reads. In one embodiment, at least two variants of the plurality of variants are classified as somatic. It is also understood that the determination of the first MAF (step 1705) may be for at least one variant of the plurality of variants that is classified as somatic and is based on at least a portion of the first plurality of sequence reads. It is further understood that the determination of a second MAF (step 1706) may be for at least one variant of the plurality of variants classified or reclassified as somatic and may be based on at least a portion of the second plurality of sequence reads. Finally, it is understood that a molecular response may be determined based on the first MAF and the second MAF.

[0230] 18 is a flow chart that generally illustrates a method 1800 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 1801). As further illustrated, the method 1800 also includes classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline (step 1802), determining at least one variant of the plurality of variants as a clonal hematopoietic potential undetermined (CHIP) variant (step 1803), removing at least one CHIP variant (step 1804), determining a first mutant allele fraction (MAF) (step 1805), determining a second MAF (step 1806), and determining a molecular response score (step 1807). It is understood that the first plurality of sequence reads can be determined before administering a therapy and the second plurality of sequence reads can be determined after administering a therapy. The classification of the plurality of variants in the first and second plurality of sequence reads as somatic or germline in step 1802 may be performed as described herein, for example, as described with respect to FIG. 2. In one embodiment, at least two variants of the plurality of variants are classified as somatic. A first MAF may be determined using a variant in the first plurality of sequence reads at a time point before treatment, and a second MAF may be determined using the same variant in the second plurality of sequence reads at a time point after treatment. A first MAF and a second MAF may be determined for the same variant in both the first and second plurality of sequence reads. It is also understood that the removal of at least one CHIP variant (step 1804) may be derived from a plurality of variants. It is further understood that the determination of the first MAF (step 1805) may be for at least one variant of the plurality of variants classified as somatic and based on at least a portion of the first plurality of sequence reads. It is further understood that the second MAF determination (step 1806) can be for at least one variant of the plurality of variants that is classified as somatic and is based on at least a portion of the second plurality of sequence reads.Finally, it is understood that the determination of the molecular response score (step 1807) can be based on the first MAF and the second MAF.

[0231] 19 is a flow chart that generally illustrates a method 1900 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 1901). As further shown, the method 1900 also includes steps of classifying a plurality of variants in the first plurality of sequence reads as somatic or germline (step 1902), classifying a plurality of variants in the second plurality of sequence reads as somatic or germline (step 1903), reclassifying at least one variant of the plurality of variants to resolve classification discrepancies between the first plurality of sequence reads and the second plurality of sequence reads (step 1904), and classifying at least one variant of the plurality of variants as undetermined potential clonal hematopoiesis (CHIP) (step 1905). ) variants (step 1905), removing at least one CHIP variant (step 1906), determining a first mutant allele fraction (MAF) (step 1907), determining a second MAF (step 1908), determining a MAF ratio (step 1909), determining a weighted average of the MAF ratios (step 1910), determining a confidence interval associated with the weighted average of the MAF ratios (step 1911), and outputting the weighted average of the MAF ratios and the confidence interval (step 1912). It is understood that the first plurality of sequence reads may be determined before administering the therapy and the second plurality of sequence reads may be determined after administering the therapy. Classifying the plurality of variants in the first plurality of sequence reads in step 1902 and classifying the second plurality of sequence reads as somatic or germline in step 1903 may be performed as described herein, for example, as described with respect to FIG. 2. In one embodiment, at least two variants of the plurality of variants are classified as somatic.It is also understood that the removal of at least one CHIP variant (step 1906) can be derived from a plurality of variants.A first MAF can be determined using a variant in a first plurality of sequence reads at a time point before treatment, and a second MAF can be determined using the same variant in a second plurality of sequence reads at a time point after treatment.For the same variant in both the first plurality of sequence reads and the second plurality of sequence reads, the first MAF and the second MAF can be determined.The classification discrepancy may be a variant classified as somatic in the first plurality of sequence reads and classified as germline in the second plurality of sequence reads. The classification discrepancy may be a variant classified as germline in the first plurality of sequence reads and classified as somatic in the second plurality of sequence reads. It is further understood that the first MAF determination (step 1907) may be for at least one variant of the plurality of variants classified or reclassified as somatic and based on at least a portion of the first plurality of sequence reads. It is further understood that the second MAF determination (1908) may be for at least one variant of the plurality of variants classified or reclassified as somatic and based on at least a portion of the second plurality of sequence reads. It is also understood that the MAF ratio (1909) determination may be for at least one variant of the plurality of variants classified or reclassified as somatic and based on the first mutant allele proportion and the second mutant allele proportion. It is further understood that the determination of the MAF ratio (step 1910) can be subjective. It is further understood that the determination of the confidence interval associated with the weighted average of the MAF ratios (step 1911) can be based on the weighted average of the MAF ratios. Finally, it is understood that the weighted average of the MAF ratios and the confidence interval can be output as a molecular response score.

[0232] 20 is a flow chart that generally illustrates a method 2000 that includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject (step 2001). As further shown, the method 2000 also includes classifying a plurality of variants in the first plurality of sequence reads as somatic or germline (step 2002), classifying a plurality of variants in the second plurality of sequence reads as somatic or germline (step 2003), reclassifying at least one variant of the plurality of variants to resolve classification discrepancies between the first and second plurality of sequence reads (step 2004), and determining at least one variant of the plurality of variants as a clonal hematopoietic potential undetermined (CHIP) variant (step 2005). 2005), removing at least one CHIP variant (step 2006), determining a first mutant allele fraction (MAF) (step 2007), determining a second MAF (step 2008), determining a weighted average of the first MAF and a weighted average of the second MAF (step 2009), determining a ratio of the weighted average of the first MAF to the weighted average of the second MAF (step 2010), determining a confidence interval (2011), and outputting the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval (step 2012). It is understood that the first plurality of sequence reads is determined prior to administering the therapy and the second plurality of sequence reads is determined after administering the therapy. Classifying the plurality of variants in the first plurality of sequence reads in step 2002 and classifying the second plurality of sequence reads as somatic or germline in step 2003 can be performed as described herein, for example, as described with respect to FIG. 2. In one embodiment, at least two variants of the plurality of variants are classified as somatic. It is also understood that the removal of at least one CHIP variant (step 2006) can be derived from multiple variants. The first MAF can be determined using the variants in the first plurality of sequence reads at a pre-treatment time point, and the second MAF can be determined using the same variants in the second plurality of sequence reads at a post-treatment time point.A first MAF and a second MAF can be determined for the same variant in both the first and second plurality of sequence reads. It is further understood that the determination of the first MAF (step 2007) can be for at least one variant of the plurality of variants that is classified or reclassified as somatic and based on at least a portion of the first plurality of sequence reads. It is further understood that the determination of the second MAF (step 2008) can be for at least one variant of the plurality of variants that is classified or reclassified as somatic and based on at least a portion of the second plurality of sequence reads. It is also understood that the determination of the weighted average of the first MAF and the weighted average of the second MAF (step 2009) can be for at least one variant of the plurality of variants that is classified as somatic and based on the first and second MAFs. It is further understood that the determination of the ratio of the weighted average of the first MAF and the weighted average of the second MAF (step 2010) can be for a subject. It is further understood that the determination of the confidence interval (step 2011) may be based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF. Finally, it is understood that the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval may be output as the molecular response score. III. Cancer and Other Diseases

[0233] In certain embodiments, the methods and aspects disclosed herein are used for long-term monitoring of patients with a given disease, disorder or condition.The disclosed methods can be used to follow the response of patients to one or more treatments over time.Typically, the disease under consideration is a type of cancer. Non-limiting 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, ocular melanoma, uveal melanoma, gallbladder cancer, gallbladder adenocarcinoma, renal cell carcinoma, clear cell renal cell carcinoma, transitional cell carcinoma, urothelial carcinoma, Wilms' tumor, leukemia, acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic (CLL), chronic myelogenous (CML), chronic myelomonocytic (CMML), liver cancer, ... carcinoma), hepatoma, hepatocellular carcinoma, cholangiocarcinoma, hepatoblastoma, lung cancer, non-small cell lung cancer (NSCLC), mesothelioma, B-cell lymphoma, non-Hodgkin's lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, T-cell lymphoma, non-Hodgkin's lymphoma, precursor T-lymphoblastic lymphoma / leukemia, peripheral T-cell lymphoma, multiple myeloma, nasopharyngeal carcinoma (NPC), neuroblastoma, oropharyngeal carcinoma, 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, stomach cancer, gastric carcinoma, gastrointestinal stromal tumor (GIST), uterine cancer, or uterine sarcoma.

[0234] Non-limiting examples of other gene-based diseases, disorders, or conditions that may be optionally evaluated using the methods and systems disclosed herein include achondroplasia, alpha 1-antitrypsin deficiency, antiphospholipid syndrome, autism, autosomal dominant polycystic kidney disease, Charcot-Marie-Tooth (CMT), cri-cat syndrome, Crohn's disease, cystic fibrosis, Delcam's disease, Down's syndrome, Duane's syndrome, Duchenne muscular dystrophy, factor V Leiden thrombocytosis, familial hypercholesterolemia, familial meningeal fever, fragile X syndrome, Gaucher's disease, hemochromatosis, hemophilia, holoprosencephaly, Huntington's disease, Klinefelter's syndrome, Marfan's syndrome, myotonic dystrophy, neurofibromatosis, Noonan's syndrome, osteogenesis imperfecta, Parkinson's disease, phenylketonuria, Poland syndrome, These include porphyria, progeria, retinitis pigmentosa, severe combined immunodeficiency (SCID), sickle cell disease, spinal muscular atrophy, Tay-Sachs disease, thalassemia, trimethylaminuria, Turner syndrome, velar-cardiofacial syndrome, WAGR syndrome, and Wilson's disease. IV. Customized Therapy and Related Administration

[0235] In some embodiments, the methods disclosed herein relate to identifying the patient with a given disease, disorder or condition and administering therapy. Essentially, any cancer therapy (e.g., surgical therapy, radiation therapy, chemotherapy, etc.) is included as part of these methods. In certain embodiments, the therapy administered to the subject can include at least one chemotherapeutic agent. In some embodiments, the chemotherapy agent may include alkylating agents (e.g., but not limited to, chlorambucil, cyclophosphamide, cisplatin and carboplatin), nitrosoureas (e.g., but not limited to, carmustine and lomustine), antimetabolites (e.g., but not limited to, fluorouracil, methotrexate and fludarabine), plant alkaloids and natural products (e.g., but not limited to, vincristine, paclitaxel and topotecan), antitumor antibiotics (e.g., but not limited to, bleomycin, doxorubicin and mitoxantrone), hormonal agents (e.g., but not limited to, prednisone, dexamethasone, tamoxifen and leuprolide) and biological response modifiers (e.g., but not limited to, herceptin and avastin, erbitux and rituxan). In some embodiments, the chemotherapy administered to the subject may include FOLFOX or FOLFIRI. In certain embodiments, the therapy administered to the subject may include at least one PARP inhibitor. In certain embodiments, the PARP inhibitor may include, among others, OLAPARIB, TALAZOPARIB, RUCAPARIB, NIRAPARIB (brand name ZEJULA). Typically, the therapy includes at least one immunotherapy (or immunotherapeutic agent). Immunotherapy generally refers to a method of enhancing immune response to a given cancer type. In certain embodiments, immunotherapy refers to a method of enhancing T cell response to tumor or cancer.

[0236] In some embodiments, the immunotherapeutic agent or immunotherapeutic agent targets immune checkpoint molecules. Certain tumors can evade the immune system by exploiting immune checkpoint pathways. Thus, targeting immune checkpoints has emerged as an effective approach to combat 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.

[0237] In certain embodiments, the immune checkpoint molecule is an inhibitory molecule that reduces signals involved in T cell responses to antigens. For example, CTLA4 is expressed on T cells and plays a role in downregulating T cell activation by binding to CD80 (aka B7.1) or CD86 (aka B7.2) on antigen presenting cells. PD-1 is another inhibitory checkpoint molecule expressed on T cells. PD-1 limits the activity of T cells in peripheral tissues during inflammatory responses. In addition, the ligands of PD-1 (PD-L1 or PD-L2) are commonly upregulated on the surface of many different tumors, resulting in downregulation of anti-tumor immune responses 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).

[0238] Antagonists targeting these immune checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Thus, in certain embodiments, the immunotherapy or immunotherapeutic 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 an inhibitory immune checkpoint molecule is an antibody (e.g., a monoclonal antibody). In certain embodiments, the antibody or monoclonal antibody is an anti-CTLA4, anti-PD-1, anti-PD-L1, or 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, an anti-CTLA4 antibody and an anti-PD-L1 antibody, or 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 tezolizumab (Tecentriq®), avelumab (Bavencio®), or durvalumab (Imfinzi®).

[0239] In certain embodiments, the immunotherapy or immunotherapeutic agent is an antagonist (e.g., an antibody) against CD80, CD86, LAG3, KIR, TIM3, GAL9, or A2aR. In other embodiments, the antagonist is a soluble version of an inhibitory immune checkpoint molecule, such as a soluble fusion protein comprising the extracellular domain of an inhibitory immune checkpoint molecule and the Fc domain of an antibody. In certain embodiments, the soluble fusion protein comprises the extracellular domain of CTLA 4, PD-1, PD-L1, or PD-L2. In some embodiments, the soluble fusion protein comprises the extracellular domain of CD80, CD86, LAG3, KIR, TIM3, GAL9, or A2aR. In one embodiment, the soluble fusion protein comprises the extracellular domain of PD-L2 or LAG3.

[0240] In certain embodiments, the immune checkpoint molecule is a costimulatory molecule that amplifies signals involved in T cell response to antigen. For example, CD28 is a costimulatory receptor expressed on T cells. When T cells bind to antigen through their T cell receptor, CD28 binds to CD80 (aka B7.1) or CD86 (aka B7.2) on antigen-presenting cells, amplifying T cell receptor signaling and promoting T cell activation. Because CD28 binds to the same ligands (CD80 and CD86) as CTLA4, CTLA4 can counteract or modulate the costimulatory signaling mediated by CD28. In certain embodiments, the immune checkpoint molecule is a costimulatory molecule selected from CD28, inducible T cell costimulatory factor (ICOS), CD137, OX40, or CD27. In other embodiments, the immune checkpoint molecule is a ligand of a costimulatory molecule, including, for example, CD80, CD86, B7RP1, B7-H3, B7-H4, CD137L, OX40L, or CD70.

[0241] Agonists that target these costimulatory checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Thus, in certain embodiments, the immunotherapy or immunotherapeutic agent is an agonist of a costimulatory checkpoint molecule. In certain embodiments, the agonist of a costimulatory checkpoint molecule 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, anti-CD137, anti-OX40, or anti-CD27 antibody. In other embodiments, the agonist antibody or monoclonal antibody is an anti-CD80, anti-CD86, anti-B7RP1, anti-B7-H3, anti-B7-H4, anti-CD137L, anti-OX40L, or anti-CD70 antibody.

[0242] Therapeutic options for treating particular genetically-based diseases, disorders, or conditions other than cancer are generally well known to those of skill in the art and will be apparent given the particular disease, disorder, or condition under consideration.

[0243] In certain embodiments, the customized therapy described herein is typically administered parenterally (e.g., intravenously or subcutaneously). Pharmaceutical compositions containing immunotherapeutic agents are typically administered intravenously. Certain therapeutic agents are administered orally. However, customized therapy (e.g., immunotherapeutic agents, etc.) can also be administered by any method known in the art, including, for example, buccal, sublingual, rectal, vaginal, urethral, ​​topical, intraocular, intranasal, and / or intraauricular, and administration can include tablets, capsules, granules, aqueous suspensions, gels, sprays, suppositories, salves, ointments, etc. V. SYSTEM AND COMPUTER READABLE MEDIUM

[0244] The present disclosure also provides various systems and computer program products or machine-readable media. In some embodiments, for example, the methods described herein are at least partially performed or facilitated as needed 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 the like. For illustration purposes, FIG. 21 provides a schematic diagram of an exemplary system suitable for implementing at least aspects of the methods disclosed in the present application. As shown, the system 2100 includes at least one controller or computer, such as a server 2102 (e.g., a search engine server), including a processor 2104 and memory, storage, or memory components 1506, and one or more other communication devices 2114 and 2116 (e.g., client-side computer terminals, phones, tablets, laptops, other mobile devices, etc.) that are located remotely from and communicate with the remote server 2102 via an electronic communication network 2112, such as the Internet or other internetwork. The communications devices 2114 and 2116 typically include, for example, an electronic display (e.g., an Internet-enabled computer, etc.) in communication with the server 2102 computer over the network 2112, the electronic display including a user interface (e.g., a graphical user interface (GUI), a web-based user interface, and / or the like) for displaying results from performing the methods described herein. In certain embodiments, the communications network also encompasses the physical transfer of data from one location to another, for example, using a hard drive, thumb drive, or other data storage mechanism.The system 2100 also includes a program product 1508 stored in a computer or machine readable medium, such as one or more various types of memory, such as, for example, memory 2106 of the server 2102, readable by the server 2102 to facilitate a guided search application or other executable file executable by one or more other communication devices, such as, for example, 2114 (shown diagrammatically as a desktop or personal computer) and 2116 (shown diagrammatically as a tablet computer). In some embodiments, the system 2100 also optionally includes at least one database server, such as, for example, server 2110 associated with an online website that stores data (e.g., classifier scores, control sample or comparator result data, indexed customized therapies, etc.) searchable directly or via the search engine server 2102. The system 2100 also optionally includes one or more other servers located remotely from the server 2102, each of which is associated with one or more database servers 2110 located remotely from each of the other servers or locally relative to each of the other servers, as appropriate. Other servers may beneficially serve geographically dispersed users and facilitate geographically distributed operations.

[0245] As will be appreciated by those skilled in the art, the memory 2106 of the server 2102 may include volatile and / or non-volatile memory, including, for example, RAM, ROM, and magnetic or optical disks, among others, as appropriate. Although illustrated as a single server, it will also be appreciated by those skilled in the art that the illustrated configuration of the server 2102 is given by way of example only, and that other types of servers or computers configured according to various other methodologies or architectures may also be used. The server 2102 shown generally in FIG. 21 represents a server or a server cluster or a server farm, and is not limited to an individual physical server. The server site may be deployed as a server farm or server cluster managed by a server hosting provider. The number of servers and their architecture and configuration may be increased based on the use, demand, and capacity requirements of the system 2100. Also, as will be appreciated by those skilled in the art, the other user communication devices 2114 and 2116 in these embodiments may be, for example, laptops, desktops, tablets, personal digital assistants (PDAs), mobile phones, servers, or other types of computers. As will be known and understood by those skilled in the art, the network 2112 may include portions of the Internet, an intranet, a telecommunications network, an extranet, or the World Wide Web, and / or a local or other area network of multiple computers / servers in communication with one or more other computers via a communications network.

[0246] As will be further appreciated by those skilled in the art, the exemplary program product or machine-readable medium 2108 is in the form of microcode, programs, cloud computing formats, routines, and / or symbolic languages ​​that provide one or more sets of sequenced operations that control and direct the operation of the hardware, as appropriate. According to an exemplary embodiment, the program product 2108 also need not reside entirely in volatile memory, but can be selectively loaded as needed according to various methods as will be known and understood by those skilled in the art.

[0247] As will be further understood by those skilled in the art, the term "computer-readable medium" or "machine-readable medium" refers to any medium that participates in providing instructions to a processor for execution. By way of example, the term "computer-readable medium" or "machine-readable medium" encompasses distribution media, cloud computing formats, intermediate storage media, computer execution memory, and any other medium or device that can store a program product 2108 that implements the functions or processes of various embodiments of the present disclosure, for example, for reading by a computer. A "computer-readable medium" or "machine-readable medium" can take many forms, including, but not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media include, for example, optical or magnetic disks. Volatile media include dynamic memory, such as the main memory of a given system. Transmission media include coaxial cables, copper wire, and fiber optics, including the wires that comprise a bus. Transmission media can also take the form of acoustic or light waves, such as those created during radio wave and infrared data communications, among others. Exemplary forms of computer-readable media include a floppy disk, a flexible disk, a hard disk, a magnetic tape, a flash drive, or any other magnetic medium, a CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with a pattern of holes, RAM, PROM, and EPROM, FLASH-EPROM, any other memory chip or cartridge, a carrier wave, or any other medium from which a computer can read.

[0248] The program product 2108 is copied from the computer readable medium to a hard disk or similar intermediate storage medium, as appropriate. When the program product 2108, or portions thereof, are executed, they are loaded, as appropriate, from their distribution medium, their intermediate storage medium, etc., into one or more computer executable memories, and configure the computer to operate according to the functions or methods of the various embodiments. All such operations are well known to those skilled in the art of, for example, computer systems.

[0249] To further illustrate, in certain embodiments, the present application provides a system that includes one or more processors and one or more memory components in communication with the processor. The memory components typically include one or more instructions that, when executed, cause the processor to provide information (e.g., via communication devices 2114, 2116, etc.) such as sequence information, epigenetic information, classifier scores, cfDNA characteristic data, cfDNA fragment distribution set data, test results, control or comparator results, customized therapies, etc., and / or receive information from other system components and / or system users (e.g., via communication devices 2114, 2116, etc.).

[0250] In some embodiments, the program product 2108 includes non-transitory computer-executable instructions that, when executed by the electronic processor 2104, perform at least: determining mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from the targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at a first time point and a second time point to generate a set of first and second MAFs for at least one variant of the plurality of variants; calculating ratios of the first and second MAFs for at least one variant of the plurality of variants to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios of the set of MAF ratios; and calculating weighted averages and confidence intervals of the MAF ratios to determine a molecular response score for the subject with cancer. Additional computer-readable medium embodiments are described herein.

[0251] The system 2100 also typically includes additional system components configured to implement various aspects of the methods described herein. In some of these embodiments, these one or more additional system components are located remotely from and communicate with the remote server 2102 via an electronic communications network 2112, while in other embodiments, these one or more additional system components are located locally and communicate with the server 2102 (i.e., when the electronic communications network 2112 is not present) or directly with, for example, a desktop computer 2114.

[0252] In some embodiments, for example, additional system components include a sample preparation component 2118 operably connected to the controller 2102 (either directly or indirectly (e.g., via electronic communications network 2112)). The sample preparation component 2118 is configured to prepare nucleic acids in the 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 of these embodiments, the sample preparation component 2118 is configured to isolate nucleic acids from other components in the sample, attach one or more adaptors comprising barcodes to the nucleic acids as described herein, selectively enrich one or more regions from a genome or transcriptome prior to sequencing, and / or the like.

[0253] In certain embodiments, the system 2100 also includes a nucleic acid amplification component 2120 (e.g., a thermal cycler, etc.) operably connected (directly or indirectly, e.g., via electronic communication network 2112) to the controller 2102. The nucleic acid amplification component 2120 is configured to amplify nucleic acids in a sample from a subject. For example, the nucleic acid amplification component 2120 is configured to amplify selectively enriched regions from a genome or transcriptome in the sample as desired, as described herein.

[0254] The system 2100 also typically includes at least one nucleic acid sequencer 2122 operably connected (directly or indirectly, e.g., via electronic communication network 2112) to the controller 2102. The nucleic acid sequencer 2122 is configured to provide sequence information from nucleic acids (e.g., amplified nucleic acids) in a sample from a subject. Essentially any type of nucleic acid sequencer can be adapted for use in these systems. For example, the nucleic acid sequencer 2122 is optionally configured to perform bisulfite sequencing, pyrosequencing, single molecule sequencing, nanopore sequencing, semiconductor sequencing, sequencing by ligation, sequencing by hybridization, or other techniques on nucleic acids to create sequencing reads. Optionally, the nucleic acid sequencer 2122 is configured to group sequence reads into families of sequence reads, each family including sequence reads created from nucleic acids in a given sample. In some embodiments, the nucleic acid sequencer 2122 creates sequencing reads using clonal single molecule arrays derived from a sequencing library. In certain embodiments, the nucleic acid sequencer 2122 includes at least one chip having an array of microwells for sequencing a sequencing library to generate sequencing reads.

[0255] To facilitate full or partial system automation, the system 2100 also typically includes a material transfer component 2124 operably connected (directly or indirectly (e.g., via electronic communications network 2112)) to the controller 2102. The material transfer component 2124 is configured to transfer one or more materials (e.g., nucleic acid samples, amplicons, reagents, etc.) to and / or from the nucleic acid sequencer 2122, the sample preparation component 2118, and the nucleic acid amplification component 2120.

[0256] Further details regarding computer systems and networks, databases, and computer program products are also provided in, for example, Peterson, Computer Networks: A Systems Approach, Morgan Kaufmann, 5th Ed. (2011), Kurose, Computer Networking: A Top-Down Approach, Pearson, 7th Ed. (2016), Elmasri, Fundamentals of Database Systems, Addison Wesley, 6th Ed. (2010), Coronel, Database Systems: Design, Implementation, & Management, Cengage Learning, 11th Ed. (2014), Tucker, Programming Languages, McGraw-Hill Science / Engineering / Math, 2nd Ed. (2006), and Rhoton, Cloud Computing Architected: Solution Design Handbook, Recursive Press (2011), each of which is incorporated by reference in its entirety. EXAMPLES

[0257] VI. Working Examples A. Example 1: Comparison of Molecular Response Calculations for Prediction of Patient Outcome 1. Background Molecular response (MR), estimated as the change in circulating tumor (ctDNA) burden between an early treatment sample (usually 2-9 weeks after treatment initiation) and pre-treatment baseline, has been shown in many retrospective studies to predict patient response and outcome across solid tumors and therapy types. However, there is no consensus regarding the best method to assess molecular response. Therefore, we aimed to evaluate several molecular response calculations and determine the optimal method for predicting outcome in individual patients with advanced cancer. 2. Method

[0258] Aggregated results from >4,000 patient sample pairs (3–10 weeks apart), >1000 patient sample technical replicates, >100 devised sample dilutions, and in silico simulations were analyzed using the cfDNA NGS Assay Clinical Platform (Guardant Health, Inc., Redwood City, CA, USA). Baseline and on-treatment patient sample pairs were collected from advanced cancer patients with >12 tumor types including lung, colon, and breast. MR calculations included variant allele fractions (VAFs) of somatic SNVs, indels, and fusions. Methods including maximum VAF ratio (RmaxVAF), mean VAF ratio (RmVAF), and mean VAF ratio (mVAF) were compared. Analytical precision, reproducibility, and limit of detection (LoD) were assessed. 3.Results

[0259] Comparison of methods to calculate net change in ctDNA burden for over 1500 sample pairs showed high correlation (ρ ranged from 0.93 to 0.98) and categorical agreement split at the median (93%). Therefore, selecting the optimal method based on outcome prediction would require prohibitively large patient cohorts. Analytical evaluation and in silico simulations can predict the behavior of each method. Simulation of changes in tumor fraction of real pretreatment samples found that RmVAF or RmaxVAF are more accurate than mVAFR, which can be distorted by low VAF ratios. Nearly 25% of sample pairs had tumor driver or resistance mutations that were not maxVAF, suggesting that tumor dynamics are better captured by mVAF than maxVAF. Newly detected treatment variants may be important signals of elevated ctDNA levels, affecting MR in roughly 2% of sample pairs.

[0260] Importantly, the MR accuracy of all methods drops off as the maxVAF approaches or falls below the variant LoD, due to both stochastic detection at low VAF and the higher CV of variants. Thus, the assay variant LoD is a key determinant of the proportion of patients who can undergo MR evaluation. Technical replicates identified variant criteria where a 50% change in tumor proportion was significantly different than technical variability, allowing us to define the limits of analytical reporting. 4. Conclusion

[0261] Comparison of MR methods in a large set of patient samples and simulations supports RmVAF including the newly detected mutations. B. Example 2 1. Introduction

[0262] Molecular response (MR) is an assessment of the change in circulating tumor (ctDNA) burden early during treatment (usually 3-10 weeks) compared to pre-treatment baseline. In many retrospective studies, molecular response has been associated with patient response to therapy as well as long-term outcomes across solid tumors and therapy types.

[0263] Molecular response has also been shown to predict clinical response earlier than radiographic and / or RECIST response. Multiple methods have been used to calculate molecular response, and there is no consensus regarding which method is best.

[0264] In this example, several molecular response calculations were evaluated to determine the optimal method for predicting outcomes in individual patients with advanced cancer. 2. Method

[0265] Paired samples from over 1,500 patient plasma samples spaced 3–10 weeks apart were processed using a cfDNA NGS assay clinical platform (Guardant Health, Inc., Redwood City, CA, USA) with a median unique coverage of approximately 4600 molecules sequenced at 20,000× read depth. Somatic and germline SNVs, small indels, and fusions were subsetted into a 74-gene panel space to mimic clinical applications of molecular response. Technical replicates of over 140 patient samples were processed with either panel and subsetted into the 74-gene panel space. Three previously published molecular response methods were evaluated (see Table 3).

[0266] Table 3 [Table 3] 3.Results i. Molecular response calculations capture ctDNA VAF changes for SNVs, indels, and fusions

[0267] Figure 22 shows the number of somatic variants detected per sample in the 74 cancer-related gene panel space. The number of somatic SNVs, indels, and fusions were counted toward the molecular response calculation per sample for the top three cancer types in this study. The median mutation variant count is 4, 5, and 3 for Breast, CRC, and NSCLC, respectively. ii. Resolving somatic typing with paired samples improves tumor signal

[0268] Figure 23 shows an example of somatic classification discrepancy that can skew MR results. Classification discrepancy for rare somatic conditions (<0.8% of variants) can occur in high tumor proportions and allelic imbalance. Unresolved ALK skews MR scores against universally reduced VAF.

[0269] Table 4 shows an example of resolving somatic classification discrepancies between patient samples to improve variant accuracy. Somatic classification discrepancies in patient sample pairs were resolved by an algorithm based on variant characteristics. Accuracy was evaluated against manual resolution by subject matter experts.

[0270] Table 4. [Table 4] iii. The variant is included in the molecular response calculation based on detection and VAF accuracy.

[0271] Figure 24 shows an example of mutant precision as determined by mutant molecular count (MMC=VAF*molecular coverage). (Figure 24A) Mutants have a range of molecular coverage depending on sample input and panel design. Probability of mutant detection (Figure 24B) and VAF precision (Figure 24C) depend on both VAF and molecular coverage (color, mapping to (Figure 24A)). MMC (Figure 24D) is a better metric of mutant precision because it determines probability of mutant detection (Figure 24E) and VAF precision (Figure 24F). Mutants with low MMC at both time points should be excluded from the molecular response to better clarify signal from noise. iv. Molecular responses are generally consistent between methods, but R(mVAF) is more robust across patients

[0272] Figure 25 shows that when using the mean of the ratios, m(rVAF), or the maximum ratio, R(maxVAF), the tumor signal can be outweighed by a minority of variants. (Figure 25A) MR scores are classified as increasing, decreasing, or within the precision limits ("nearly 0% change"). Only 8% of patients change between increasing and decreasing in any way, indicating high categorical correlation (X2 p<0.001). MR correlations range from ρ=0.42 to 0.86 (p<0.001). (Figure 25B) m(rVAF) tends to overestimate MR when several VAFs are low (red). R(maxVAF) can be skewed by a single maximum variant (purple) that deviates from the majority. 20% of sample pairs have tumor driver or resistance mutations that are not maxVAF, suggesting that tumor dynamics are well captured by mVAF. (C) Excluding new treatment variants results in a lower MR assessable rate, excluding the signal of new variants. v. Patients with low signal ctDNA level changes are identified as not evaluable for molecular response

[0273] FIG. 26 shows examples of increasing certainty of molecular response scores as the number of variants (FIG. 26A), molecular coverage (FIG. 26B), and maximum VAF (FIG. 26C) increase.

[0274] In the absence of somatic variants (approximately 7% of patients) or in the absence of somatic variants fulfilling the inclusion criteria (16%), the sample pairs are not evaluable for molecular response using the VAF-based method. The certainty of the molecular response score is also calculated theoretically using a statistical model of VAF accuracy. Sample pairs beyond the acceptable limit of uncertainty (black line) are not evaluable for MR (3%). This results in approximately 74% of sample pairs evaluable for MR. vi. The range of molecular response scores in clinical patient samples reflects a strong biological signal

[0275] In clinical patient sample pairs, the molecular response distribution shows a range of scores from 100% decrease to over 100% increase (Figure 27A).

[0276] Technical replicates provide a null molecular response distribution that peaks at 0% change (Figure 27B). 4. Conclusion

[0277] Each component of the molecular response calculation is critical to the accurate assessment of MR, including germline and low-fidelity variant filtering, global formulation, and evaluable criteria. Comparison of molecular response methods in large sets of patient samples and simulations supports the ratio of mean VAFs that include newly detected mutations. C. Example 3

[0278] Figure 28 shows an example of a sample pair for MR calculation. Starting with all SNVs, Indels, and fusions detected in either sample, common germline variants are removed. Then, discrepancies in somatic / germline classification of variants are resolved to give a single classification. (In this example, there were no discrepancies.) Next, germline variants are filtered out, followed by CHIP variants (in this example, ATM.R3008H is the CHIP variant that was removed). Variants that do not meet MMC or coverage-based inclusion thresholds are then removed. In this example, after these filtering steps, three somatic variants (PDGFRA, RET, and TP53) remain. Finally, the MR score is calculated from these remaining variants. In this example, the baseline mean VAF is 22.2% and the treatment mean VAF is 2.7%, giving a MR score of 12%, which is a ctDNA reduction of 88%.

[0279] All patents, patent applications, websites, other publications or documents, accession numbers, etc. cited herein are incorporated by reference in their entirety for all purposes to the same extent as if each individual item was specifically and individually indicated to be incorporated by reference. Where different versions of a sequence are associated with an accession number at different times, the version associated with the accession number at the effective filing date of this application is meant. By effective filing date, if applicable, is meant the earlier of the actual filing date or the filing date of the priority application that references the registration number. Similarly, where different versions of a publication, website, etc. are published at different times, the version last published at the effective filing date of the application is meant, unless otherwise indicated.

[0280] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the methods and compositions described herein which equivalents are intended to be encompassed by the following claims.

Claims

1. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining a weighted average of the first mutant allele fraction (MAF) and the second MAF for at least one variant of the plurality of variants classified as somatic based on the first MAF and the second MAF; determining a ratio of the weighted average of the first MAF to the weighted average of the second MAF for the subject; determining a confidence interval based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF; and outputting the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval as a molecular response score.

2. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining a mutant allele fraction (MAF) ratio for at least one variant of the plurality of variants classified as somatic, based on a first MAF and a second MAF; determining a weighted average of the MAF ratios for the subjects; determining a confidence interval associated with the weighted average of the MAF ratios based on the weighted average of the MAF ratios; and and outputting the weighted average of the MAF ratios and the confidence interval as a molecular response score.

3. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying the plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads; determining a first mutant allele fraction (MAF) for at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the first plurality of sequence reads; determining a second MAF for the at least one variant of the plurality of variants classified or reclassified as somatic based on at least a portion of the second plurality of sequence reads; determining a molecular response score based on the first MAF and the second MAF.

4. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining at least one variant of the plurality of variants as a potential clonal hematopoietic undefined (CHIP) variant; removing said at least one CHIP variant from said plurality of variants; determining a first mutant allele fraction (MAF) for at least one variant of the plurality of variants classified as somatic based on at least a portion of the first plurality of sequence reads; determining a second MAF for at least one variant of the plurality of variants classified as somatic based on at least a portion of the second plurality of sequence reads; determining a molecular response score based on the first MAF and the second MAF.

5. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying the plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads; determining at least one variant of the plurality of variants as a potential clonal hematopoietic undefined (CHIP) variant; removing said at least one CHIP variant from said plurality of variants; determining a first mutant allele fraction (MAF) for at least one variant of the plurality of variants classified as somatic based on at least a portion of the first plurality of sequence reads; determining a second MAF for at least one variant of the plurality of variants classified as somatic based on at least a portion of the second plurality of sequence reads; determining a MAF ratio for at least one variant of the plurality of variants classified as a somatic cell based on the first MAF and the second MAF; determining a weighted average of the MAF ratios for the subjects; determining a confidence interval associated with the weighted average of the MAF ratios based on the weighted average of the MAF ratios; and and outputting the weighted average of the MAF ratios and the confidence interval as a molecular response score.

6. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined prior to administering a therapy and the second plurality of sequence reads being determined after administering the therapy; classifying a plurality of variants in the first plurality of sequence reads as somatic or germline; classifying the plurality of variants in the second plurality of sequence reads as somatic or germline; reclassifying at least one variant of the plurality of variants to resolve a classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads; determining at least one variant of the plurality of variants as a potential clonal hematopoietic undefined (CHIP) variant; removing said at least one CHIP variant from said plurality of variants; determining a first mutant allele fraction (MAF) for at least one variant of the plurality of variants classified as somatic based on at least a portion of the first plurality of sequence reads; determining a second MAF for at least one variant of the plurality of variants classified as somatic based on at least a portion of the second plurality of sequence reads; determining a weighted average of the first MAF and a weighted average of the second MAF for the at least one variant of the plurality of variants classified as somatic based on the first MAF and the second MAF; determining a ratio of the weighted average of the first MAF to the weighted average of the second MAF for the subject; determining a confidence interval based on the ratio of the weighted average of the first MAF to the weighted average of the second MAF; and outputting the ratio of the weighted average of the first MAF to the weighted average of the second MAF and the confidence interval as a molecular response score.

7. 1. A method comprising: determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, the first plurality of sequence reads being determined at a first time point prior to administration of a therapy, and the second plurality of sequence reads being determined at a second time point after administration of the therapy; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining, for at least one variant of the plurality of variants classified as somatic, a first representative value of the first mutant allele fraction (MAF) and a second representative value of the second MAF based on a first MAF at the first time point and a second MAF at the second time point; determining a ratio between the first representative value at the first time point and the second representative value at the second time point; and outputting the ratio of the first representative value at the first time point to the second representative value at the second time point as a molecular response score.

8. The method of claim 7 , wherein the representative value is one or more of a mean, a median, or a mode.

9. 1. A method of determining a molecular response score for a subject having cancer, using at least in part a computer, comprising: (a) determining, by the computer, mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at a first time point and a second time point to generate a first and a second set of MAFs for at least one variant in the plurality of variants; (b) calculating, by the computer, a ratio of the first and second MAFs for at least one variant in the plurality of variants to generate a set of MAF ratios and a corresponding standard deviation for each MAF ratio in the set of MAF ratios; (c) determining, by the computer, the molecular response score of the subject with cancer by calculating a weighted average and confidence interval of the MAF ratio.

10. 1. A method of obtaining a molecular response score as an indication for treating cancer in a subject, comprising: (a) determining mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at a first time point and a second time point to generate a first and a second set of MAFs for at least one variant in the plurality of variants; (b) calculating a ratio of the first and second MAFs for at least one variant in the plurality of variants to generate a set of MAF ratios and a corresponding standard deviation for each MAF ratio in the set of MAF ratios; (c) calculating a weighted average and confidence interval of the MAF ratios to determine the molecular response score for the subject; and based on at least said molecular response score, one or more therapies are administered to said subject, thereby indicating treating said cancer in said subject.

11. 1. A method of obtaining a molecular response score as an indication for treating cancer in a subject, the method comprising: indicating that one or more therapies are to be administered to the subject based at least on the molecular response score of the subject; and wherein the molecular response score is: (a) determining, by a computer, mutant allele frequencies (MAFs) for a plurality of variants from sequence information generated from targeted nucleic acids associated with one or more cancer types in samples obtained from said subject at a first time point and a second time point to generate a first and a second set of MAFs for at least one variant in said plurality of variants; (b) calculating, by the computer, a ratio of the first and second MAFs for at least one variant in the plurality of variants to generate a set of MAF ratios and a corresponding standard deviation for each MAF ratio in the set of MAF ratios; and (c) calculating by the computer a weighted average and confidence interval of the MAF ratio to determine the molecular response score for the subject; method.

12. 1. A method for identifying clonal hematopoietic variants in a subject with cancer, using at least in part a computational approach, comprising: (a) determining, by the computer, a tumor burden change (R) relative to a tumor fraction change P(R) for each of a plurality of variants from sequence information generated from targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at a first time point and a second time point to generate a set of tumor burden changes; (b) identifying, by the computer, one or more resistance signatures corresponding to one or more clonal hematopoietic variants from the set of tumor burden changes, thereby identifying said clonal hematopoietic variants in said subject with said cancer.

13. 1. A method for identifying clonal hematopoietic variants in a subject with cancer using, at least in part, a computer, comprising: (a) calculating, by the computer, a probability density function for tumor fraction change P(R) for each of a plurality of variants from sequence information generated from targeted nucleic acids associated with one or more cancer types in samples obtained from the subject at a first time point and a second time point; (b) grouping, by said computer, one or more mutations by P(R) into one or more clones; (c) generating, by the computer, an updated P(R) for each of the clones; and (d) identifying, by said computer, one or more clones having a percent change between said first and second time points that is at or above a predetermined threshold, thereby identifying said clonal hematopoietic variants in said subject with cancer.

14. 14. The method of claim 13, comprising determining the likelihood that a given pair of mutants exhibits the same fractional change, merging the most likely pair of mutants into a single clone, and updating the P(R) for the single clone.

15. 1. A method for identifying germline variants in a subject having cancer, using at least in part a computer, comprising: (a) determining a mutant allele frequency (MAF) for a given variant from sequence information generated by the computer from targeted nucleic acids associated with one or more cancer types in a sample obtained from the subject; (b) identifying, by the computer, the given variant as a germline variant if the MAF of the given variant increases the maximum MAF of the sample, which sample contains the maximum fraction of diploid genes (max frac_diploid), and / or if the MAF of the given variant is at least about 2-fold greater, 3-fold greater, 4-fold greater, 5-fold greater, 6-fold greater, 7-fold greater, 8-fold greater, 9-fold greater, or more than one or more other MAFs determined from the sample obtained from the subject, thereby identifying the germline variant in the subject with cancer.

16. 2. The method of any one of the preceding claims, further comprising comparing the molecular response score of the subject having the cancer to a pre-defined cut-off point and identifying the subject as likely to be a responder to one or more therapies for the cancer if the molecular response score is below the pre-defined cut-off point, or identifying the subject as likely to be a non-responder to the one or more therapies for the cancer if the molecular response score is at or above the pre-defined cut-off point.

17. 2. The method of any one of the preceding claims, wherein the one or more therapies comprise one or more immunotherapies.

18. 2. The method of any one of the preceding claims, wherein the molecular response score indicates that one or more therapies for the cancer should be administered to the subject.

19. 2. The method of any one of the preceding claims, wherein the molecular response score indicates that administration of one or more therapies for the cancer to the subject should be discontinued.

20. 2. The method of any one of the preceding claims, comprising using said molecular response score as a prognostic and / or predictive biomarker for said subject.

21. 2. The method of any one of the preceding claims, comprising calculating the standard deviation for each MAF ratio in the set of MAF ratios using molecular counting.

22. 2. The method of any one of the preceding claims, comprising propagating a dispersion through each MAF ratio in the set of MAF ratios.

23. 2. The method of any one of the preceding claims, further comprising excluding one or more germline and / or clonal hematopoietic variants when determining the mutant allele frequencies (MAFs) for the plurality of variants.

24. 2. The method of any one of the preceding claims, wherein the plurality of variants comprises somatic nucleic acid variants.

25. 2. The method of any one of the preceding claims, further comprising filtering out one or more somatic variants having a MAF that is less than about 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8% or 0.9% at both the first and second time points.

26. 2. The method of any one of the preceding claims, wherein the first time point comprises a time point before treatment and the second time point comprises a time point at the start of treatment or a time point after treatment.

27. 13. The method of any one of the preceding claims, comprising generating said sequence information from nucleic acid molecules obtained from one or more tissues or cells in said sample.

28. 2. The method of any one of the preceding claims, comprising generating said sequence information from cell-free nucleic acid (cfNA) in said sample obtained from said subject.

29. 2. The method of any one of the preceding claims, wherein the cfNA comprises circulating tumor DNA (ctDNA).

30. 2. The method of any one of the preceding claims, wherein the ratio comprises the second MAF to the first MAF for at least one mutant in the plurality of mutants.

31. The formula: [Equation 38] where the weight is 1 / range for a given variant in the plurality of variants. ２ and the range is the difference between the first and second MAF values ​​for a given variant in the plurality of variants, and the ratio is a given MAF ratio among a set of MAF ratios.

2. The method of claim 1, further comprising calculating the weighted average of the MAF ratios using

32. The formula: [0039] (where ratio variance is 1 / sum[weight]) 2. The method of claim 1, further comprising calculating the confidence interval using

33. calculating the ratio of the weighted averages of the MAFs according to the following formula: (weighted average of MAF ratios at time 2) / (weighted average of MAF ratios at time 1) (where the MAF of a mutant not detected at a given time point is set to 0) 2. A method according to any one of the preceding claims, comprising using

34. 2. The method of any one of the preceding claims, wherein the variants comprise one or more single nucleotide variants (SNVs), insertion / deletion mutations (indels), gene amplifications, and / or gene fusions.

35. 2. The method of any one of the preceding claims, further comprising determining the molecular response score for the subject with cancer using one or more additional genomic data sources.

36. 2. The method of any one of the preceding claims, wherein the additional genomic data sources comprise one or more of coverage, off-target coverage, epigenetic signatures, and / or microsatellite instability scores.

37. 2. The method of any one of the preceding claims, wherein the epigenetic signature comprises a length, location and / or endpoint density distribution of cfNA fragments.

38. 2. The method of any one of the preceding claims, wherein the epigenetic signature comprises an epigenetic status or state represented by one or more epigenetic loci within a given targeted genomic region.

39. 10. The method of any one of the preceding claims, wherein the epigenetic status or state comprises the presence or absence of methylation, hydroxymethylation, acetylation, ubiquitination, phosphorylation, sumoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone alterations.

40. 2. The method of any one of the preceding claims, further comprising determining one or more copy number variants prior to administering said therapy and determining one or more copy number variants after administering said therapy, and removing said one or more copy number variants from said plurality of variants based on a number of said one or more copy number variants that exceed a threshold.

41. determining a first representative value obtained from the MAF of a somatic variant at a first time point and a second representative value obtained from the MAF of said somatic variant at a second time point; calculating a ratio between the representative value at the first time point and the representative value at the second time point; calculating a standard deviation of the ratio between the representative value at the first time point and the representative value at the second time point; 2. The method of any one of the preceding claims, further comprising outputting the ratio of the representative value at the first time point to the representative value at the second time point as another molecular response score.

42. obtaining a plurality of sequence reads at each of a plurality of time points; and determining a MAF ratio between any two of the plurality of time points.

10. A method according to any one of the preceding claims.

43. A method according to any one of the preceding claims, wherein the molecular response score indicates that one or more therapies should be recommended.

44. Apparatus configured to perform a method according to any preceding claim.

45. A computer-readable medium comprising non-transitory computer-executable instructions which, when executed by at least one electronic processor, perform a method according to any of the preceding claims.

46. A system configured to carry out a method according to any preceding claim.

Images