Methods and related embodiments for analyzing molecular responses

Abstract

To provide methods and related embodiments for analyzing molecular responses. [Solution] A method for determining a molecular response score is provided herein. The molecular response score can be used to monitor and guide the administration of a treatment to a subject. The method may include determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined before the administration of the therapy and the second plurality of sequence reads are determined after the administration of the therapy, and classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic cells or germline cells.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims the benefit of priority under U.S. Provisional Patent Application No. 63 / 157,592, filed on 5 March 2021, and U.S. Provisional Patent Application No. 63 / 173,193, filed on 9 April 2021, each of which is incorporated herein by reference in whole for all purposes. [Background technology]

[0002] background Molecular response is the calculation of changes in circulating tumor DNA (ctDNA) levels observed in samples collected from subjects at different time points. In certain cases, the calculation is based on the proportion of somatic variants in 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 according to the 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. Therefore, there is still a need for a method to accurately determine the molecular response score of subjects with cancer. [Overview of the Initiative]

[0003] Brief summary In one embodiment, the present disclosure provides a method for determining molecular response scores using at least partially a computer. The method includes determining a first set of sequence reads and a second set of sequence reads associated with a subject, wherein the first set of sequence reads is determined before therapy and the second set of sequence reads is determined after therapy; classifying multiple variants in the first set of sequence reads and the second set of sequence reads as somatic or germline; determining a weighted average of the first and second MAFs for at least one variant of the multiple variants classified as somatic, based on the first and second MAFs; determining the ratio of the weighted average of the first and second MAFs for the subject; determining a confidence interval based on the ratio of the weighted average of the first and second MAFs; and outputting the ratio of the weighted average of the first and second MAFs and the confidence interval as a molecular response score.

[0004] In one embodiment, the Disclosure provides a method for determining a molecular response score using at least part computer. The method includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined before therapy is administered and the second plurality of sequence reads are determined after therapy is administered; classifying a plurality of variants in the first plurality of sequence reads and the second plurality of sequence reads as somatic or germline; determining an MAF ratio for at least one variant of the plurality of variants classified as somatic, based on a first mutant allele proportion (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 embodiment, the present disclosure provides a method for determining molecular response scores using at least partially a computer. The method comprises determining a first set of sequence reads and a second set of sequence reads associated with a subject, wherein the first set of sequence reads is determined before therapy is administered and the second set of sequence reads is determined after therapy is administered; classifying multiple variants within the first set of sequence reads as somatic or germline; classifying multiple variants within the second set of sequence reads as somatic or germline; reclassifying at least one of the variants to resolve any classification discrepancies between the first set of sequence reads and the second set of sequence reads; determining the proportion of a first mutant allele based on at least a portion of the first set of sequence reads for at least one of the variants classified or reclassified as somatic; determining the proportion of a second mutant allele based on at least a portion of the second set of sequence reads for at least one of the variants classified or reclassified as somatic; and determining a molecular response score based on the proportion of the first mutant allele and the proportion of the second mutant allele.

[0006] In one embodiment, the present disclosure provides a method for determining a molecular response score using at least partially a computer. The method includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined before therapy is administered and the second plurality of sequence reads are determined after therapy is administered; 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 an undetermined potential clonal hematopoietic (CHIP) variant; removing at least one CHIP variant from the plurality of variants; determining a first mutant allele ratio based on at least a portion of the first plurality of sequence reads for at least one variant of the plurality of variants classified as somatic; determining a second mutant allele ratio based on at least a portion of the second plurality of sequence reads for at least one variant of the plurality of variants classified as somatic; 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 a molecular response score using at least part computer. The method involves determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined before therapy is administered and the second plurality of sequence reads are determined after therapy is administered; classifying a plurality of variants within the first plurality of sequence reads as somatic or germline; and classifying a plurality of variants within the second plurality of sequence reads as somatic or germline. To resolve the classification mismatch between the first set of sequence reads and the second set of sequence reads, reclassify at least one of the multiple variants, determine at least one of the multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant, remove at least one CHIP variant from the multiple variants, and for at least one of the multiple variants classified or reclassified as somatic, determine the proportion of the first mutant allele based on at least a portion of the first set of sequence reads, and the somatically classified or reclassified The method includes: determining the proportion of a second mutant allele based on at least a portion of a second set of sequence reads for at least one of a set of mutants; determining the proportion of a first mutant allele and a second mutant allele based on the MAF ratio for at least one of a set of mutants classified or reclassified as somatic cells; 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 using at least part computer. The method involves determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined before therapy is administered and the second plurality of sequence reads are determined after therapy is administered; classifying a plurality of variants within the first plurality of sequence reads as somatic or germline; classifying a plurality of variants within 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 an undetermined potential clonal hematopoietic (CHIP) variant; removing at least one CHIP variant from the plurality of variants; and at least one variant of the plurality of variants classified as somatic. The method includes: determining a first mutant allele proportion (MAF) for a differentiator based on at least a portion of a first set of sequence reads; determining a second MAF for at least one variant of a set of somatic variants based on at least a portion of a second set 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 a set of somatic variants based on the first and second MAFs; determining the ratio of the weighted average of the first MAF to the weighted average of the second MAF for a given 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 Disclosure provides a method for determining a molecular response score using at least part computer. The method includes determining a first plurality of sequence reads and a second plurality of sequence reads associated with a subject, wherein the first plurality of sequence reads are determined at a first time point before therapy is administered and the second plurality of sequence reads are determined at a second time point after therapy is administered; 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 first representative value of the first MAF and a second representative value of the second MAF for at least one variant of the plurality of variants classified as somatic, based on the first MAF at a first time point and the second MAF at a second time point; determining the ratio of the first representative value at the first time point to 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.

[0010] In one embodiment, the Disclosure provides a method for determining a molecular response score for a subject having cancer, at least partially using a computer. The method includes (a) using a computer to determine mutant allele frequencies (MAFs) for a plurality of variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the subject at first and second time points to generate first and second sets of MAFs for each of the plurality of variants. The method also includes (b) using a computer to calculate the ratios of the first and second MAFs for each of 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) using a computer to determine a molecular response score for a subject having cancer by calculating a weighted mean and confidence interval of the MAF ratios.

[0011] In another embodiment, the Disclosure provides a method for treating a target cancer. The method includes (a) determining mutant allele frequencies (MAFs) for a plurality of variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from a subject at first and second time points to generate first and second sets of MAFs for each of the plurality of variants. The method includes (b) calculating the ratio of the first and second MAFs for each of the plurality of variants to generate a set of MAF ratios and the corresponding standard deviation for each MAF ratio in the set of MAF ratios. The method also includes (c) calculating a weighted mean and confidence interval of the MAF ratios to determine the molecular response score of the subject. In addition, the method includes (d) treating the target cancer by administering one or more therapies to the subject based on at least the molecular response score.

[0012] In another embodiment, the Disclosure provides a method for treating a target cancer. The method comprises administering one or more therapies to the target based on at least a molecular response score for the target. The molecular response score is generated by (a) by computer determining mutant allele frequencies (MAFs) for multiple variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the target at first and second time points to generate first and second sets of MAFs for each of the multiple variants; (b) by computer calculating the ratio of the first and second MAFs for each of the multiple variants to generate a set of MAF ratios and corresponding standard deviations for each MAF ratio in the set of MAF ratios; and (c) by computer calculating a weighted mean and confidence interval of the MAF ratios to determine the molecular response score for the target.

[0013] In another embodiment, the Disclosure provides a method for identifying clonal hematopoietic variants in a subject having cancer, using at least partially a computer. The method includes (a) generating a set of tumor burden changes by determining tumor burden changes (R) for each of several variants relative to tumor percentage changes P(R) from sequence information created from one or more targeted nucleic acids associated with oncogenic types in samples obtained from the subject at first and second time points. The method also includes (b) identifying one or more resistance signatures corresponding to one or more clonal hematopoietic variants from the set of tumor burden changes by computer, thereby identifying clonal hematopoietic variants in a subject having cancer.

[0014] In another embodiment, the Disclosure provides a method for identifying clonal hematopoietic variants in a subject having cancer, using at least partially a computer. The method includes (a) a computer calculating a probability density function for tumor percentage change P(R) for each of several variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the subject at first and second time points. The method also includes (b) a computer grouping one or more variants by P(R) into one or more clones, and (c) a computer creating updated P(R) for each of the clones. In addition, the method also includes (d) a computer identifying one or more clones having a percentage change between first and second time points at or above a predetermined threshold, thereby identifying clonal hematopoietic variants in a subject having cancer. In some of these embodiments, the method includes determining the likelihood that a given pair of variants exhibit the same percentage change, merging the most likely pair of variants into a single clone, and updating P(R) for the single clone.

[0015] In another embodiment, the Disclosure provides a method for identifying germline variants in a subject having cancer, using at least partially a computer. The method comprises (a) determining the mutant allele frequency (MAF) for a given variant from sequence information created from one or more targeted nucleic acids associated with oncologies in a sample obtained from the subject, using a computer. The method also comprises (b) determining, by computer, that the MAF of a given variant increases the maximum MAF of a sample, and that the sample has the maximum proportion (max) of diploid genes. Identifying a given mutant as a germline mutant if it contains frac_diploid and / or if the MAF of a given mutant is at least approximately 2 times, 3 times, 4 times, 5 times, 6 times, 7 times, 8 times, 9 times, or 1 or more other MAF determined from a sample obtained from the subject, thereby identifying germline mutants in a subject with cancer.

[0016] In some embodiments, the methods disclosed herein include comparing the molecular response score of a subject with cancer to a predetermined cutoff point to identify whether the subject is likely to be a responder to one or more therapies for cancer if the molecular response score is below the predetermined cutoff point, or whether the subject is likely to be a non-responder to one or more therapies for cancer if the molecular response score is above the predetermined cutoff point. In some embodiments, 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 a subject, taking the molecular response score into consideration. In some embodiments, the methods disclosed herein include discontinuing the administration of one or more therapies for cancer to a subject, taking the molecular response score into consideration. In some embodiments, the methods disclosed herein include recommending one or more therapies. In some embodiments, the methods disclosed herein include recommending discontinuation of one or more therapies. In some embodiments, the methods disclosed herein include using the molecular response score as a prognostic biomarker and / or predictive biomarker for a subject.

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

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

[0019] In some embodiments, the ratio includes a second MAF relative to a first MAF for each variant in a plurality of variants. In some embodiments, the method disclosed herein includes calculating a weighted average of the MAF ratios using the following formula: sum[weight*ratio] / sum[weight] (In the formula, the weight is 1 / range for a given mutant among multiple mutants) 2The range is the difference between the first and second MAF values ​​for a given mutant among multiple mutants, and the ratio is the 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] (In the formula, the ratio 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 determining a molecular response score for a subject having cancer using one or more additional genomic data sources. In some embodiments, the additional genomic data sources include one or more coverages, off-target coverages, epigenetic signatures, and / or microsatellite instability scores. In some embodiments, the epigenetic signature includes cfNA fragment length, location, and / or endpoint density distribution. In some embodiments, the epigenetic signature includes an epigenetic status or state indicated 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, smoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone mutations.

[0021] The present application discloses methods, computer-readable media, and systems useful for determining a molecular response score in a subject having 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 description that follows, and in part will be understood 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 means of 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] Figure 1 shows an exemplary method.

[0024] [Figure 2] Figure 2 shows an exemplary method.

[0025] [Figure 3] Figure 3 shows an exemplary method.

[0026] [Figure 4] Figure 4 shows an exemplary method.

[0027] [Figure 5] Figure 5 shows an exemplary method.

[0028] [Figure 6] Figure 6A shows an exemplary method.

[0029] Figure 6B shows an exemplary method.

[0030] [Figure 7] Figure 7 shows an exemplary method.

[0031] [Figure 8] Figure 8 shows an exemplary method.

[0032] [Figure 9] Figure 9 shows an exemplary method.

[0033] [Figure 10] Figure 10 shows an exemplary method.

[0034] [Figure 11] Figure 11 shows an exemplary method.

[0035] [Figure 12] Figure 12A shows an exemplary method.

[0036] Figure 12B shows an exemplary method.

[0037] [Figure 13] Figure 13 shows an exemplary method.

[0038] [Figure 14] Figure 14 shows an exemplary method.

[0039] [Figure 15] Figure 15 shows an exemplary method.

[0040] [Figure 16] Figure 16 shows an exemplary method.

[0041] [Figure 17] Figure 17 shows an exemplary method.

[0042] [Figure 18] Figure 18 shows an exemplary method.

[0043] [Figure 19] Figure 19 shows an exemplary method.

[0044] [Figure 20] Figure 20 shows an exemplary method.

[0045] [Figure 21] Figure 21 shows an exemplary method.

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

[0047] [Figure 23] Figure 23 shows an example of somatic classification discrepancies that can distort MR results.

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

[0049] [Figure 25]Figures 25A–25C show that tumor signals may outperform a small number of variants when using the mean ratio, m(rVAF), or the maximum ratio, R(maxVAF). (A) MR scores are categorized as increasing, decreasing, or within the precision limits ("nearly 0% change"). (B) Patient molecular response scores for each method are shown. (C) Graph of baseline evaluable variants with R(mVAF) only (Y axis) versus all evaluable variants with R(mVAF). Black circles indicate evaluable, and lighter circles (seen on the line crossing the x axis) indicate non-evaluable.

[0050] [Figure 26] Figures 26A to 26C show examples where the certainty of the molecular response score increases with increasing the number of mutants (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 replicas (null distribution) (B), along with hypothetical examples of mutant trajectories.

[0052] [Figure 28] Figure 28 shows an example of determining the molecular response score. [Modes for carrying out the invention]

[0053] Detailed explanation The disclosed methods and compositions can be more readily understood by referring to the detailed descriptions of the following specific embodiments and the examples contained therein, as well as the drawings and their preceding and succeeding descriptions.

[0054] The disclosed methods and compositions are not limited to specific synthesis methods, analytical techniques, or reagents unless otherwise specified, and should therefore be understood to be subject to change. I. Definition

[0055] It should also be understood that the terms used herein are intended solely to describe specific embodiments and are not intended to limit them. 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 the descriptions of methods, computer-readable media, and systems and in the claims, the following terms and their grammatical variations are used according to the definitions set forth below.

[0056] Where used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly indicates otherwise. Thus, for example, a reference to “method” includes one or more methods and / or processes of the kind described herein and / or which would be apparent to those skilled in the art by reading this disclosure, etc. It will also be understood that there is an implicit “about” before temperatures, concentrations, times, the number of bases or base pairs, coverage, etc., discussed in this disclosure, so as to whether there are few substantial equivalents within the scope of this disclosure. In this application, unless otherwise specified, the use of the singular forms includes the plural forms. Furthermore, 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” when applied to a value or element of interest of one or more means a value or element similar to the reference value or element described. In certain embodiments, the term “about” or “approximately” means a range of values ​​or elements that fall in either direction (greater than or less than) 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less than or equal to 1 of the reference value or element described, unless otherwise specified or evident from the context (except when such a number exceeds 100% of the 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. An adapter may include a sequencing primer binding site at both ends, which includes a nucleic acid primer binding site that enables amplification of the nucleic acid molecule to which the adapter is adjacent, and / or a primer binding site for sequencing applications, such as various next-generation sequencing (NGS) applications. An adapter may also include a binding site for a capture probe, such as an oligonucleotide bound to a flow cell support. An adapter may also include a nucleic acid tag as described herein. A nucleic acid tag is typically positioned relative to the amplification primer and sequencing primer binding sites so that the nucleic acid tag is included in the amplicon and sequencing read of a given nucleic acid molecule. The same or different adapters can be ligated to each end of a nucleic acid molecule. In certain embodiments, the same adapter is ligated to each end of a nucleic acid molecule, except that the nucleic acid tag is different. In some embodiments, the adapter is a Y-shaped adapter having a blunt or tailed end at one end, as described herein, for binding to a nucleic acid molecule, and this blunt or tailed end also has one or more complementary nucleotides. In yet other exemplary embodiments, the adapter is a bell-shaped adapter with a blunt or tailed end for binding to the nucleic acid molecule to be analyzed. Other exemplary adapters include T-tail adapters and C-tail adapters.

[0059] Administering (giving): As used herein, “administering” or “giving” a therapeutic agent (e.g., an immunological therapeutic agent) to a subject means giving, imparting, or bringing the composition into contact with the subject. Administration can be achieved by any of several routes, including, for example, topical, oral, subcutaneous, intramuscular, intraperitoneal, intravenous, intrathecal, and intradermal.

[0060] Alleles: As used herein, “allele” or “allelemic variant” refers to a specific genetic variant at a defined genomic location or locus. Allelemic variants are typically 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 typically have a frequency of 0.5 or 1. Somatic variants; however, acquired variants typically have a frequency of <0.5. Major and secondary alleles of a locus refer to nucleic acids having a locus that is occupied by a nucleotide of the reference sequence and a variant nucleotide different from the reference sequence, respectively. Measurements at a locus can take the form of allele proportions (AF), which measure the frequency with which an allele is observed in a sample.

[0061] Amplification: As used herein, “amplify” or “amplify” in relation to nucleic acids typically refers to the generation of multiple copies of a polynucleotide or a portion of a polynucleotide, starting with a small amount of polynucleotide (e.g., a single polynucleotide molecule), and the amplified 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 containing 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 before final data analysis.

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

[0064] Cell-free nucleic acids: As used herein, “cell-free nucleic acids” refers to nucleic acids that are not contained within cells or otherwise not bound to cells, or, in some embodiments, nucleic acids that remain in the sample after the removal of intact cells. Cell-free nucleic acids may include, for example, all unencapsulated nucleic acids derived from bodily fluids from a subject (e.g., blood, plasma, serum, urine, cerebrospinal fluid (CSF), etc.). Cell-free nucleic acids include DNA (cfDNA), RNA (cfRNA), and hybrids thereof, including genomic DNA, mitochondrial DNA, circulating DNA, siRNA, miRNA, circulating RNA (cRNA), tRNA, rRNA, nucleolar small RNA (snoRNA), Piwi-interacting RNA (piRNA), long non-coding RNA (long ncRNA), and / or fragments of any of these. Cell-free nucleic acids may be double-stranded, single-stranded, or hybrids thereof. Cell-free nucleic acids may be released into bodily fluids via secretory or cell death processes, such as cell necrosis, apoptosis, etc. Cell-free nucleic acids may be found in efferosomes or exosomes. Some cell-free nucleic acids are released into bodily fluids from cancer cells, such as circulating tumor DNA (ctDNA). Others are released from healthy cells. CtDNA can be unencapsulated tumor-derived fragmented DNA. Another example of cell-free nucleic acids is fetal DNA that circulates freely in the maternal bloodstream, also known as cell-free fetal DNA (cffDNA). Cell-free nucleic acids can have one or more epigenetic modifications; for example, cell-free nucleic acids can be acetylated, 5-methylated, ubiquitinated, phosphorylated, SUMOylated, ribosylated, and / or citrullinated.

[0065] Classifier: As used herein, “classifier” generally refers to an algorithmic computer code that takes 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, “clone” in relation to nucleic acids means a group of nucleic acids containing nucleotide sequences that are substantially or completely identical to one another, at least at a given locus of interest (e.g., a target variant).

[0067] Undetermined Potential Clonal Hematopoiesis: As used herein, “undetermined potential clonal hematopoiesis,” “clonal hematopoietic variant,” or “CHIP” refers to hematopoiesis in an individual involving the proliferation of hematopoietic stem cells that include one or more somatic mutations (e.g., hematological cancer-associated mutations and / or non-cancerous mutations) but otherwise lack the diagnostic criteria for hematological malignancies, such as definitive morphological evidence of dysplasia. CHIP is a common age-related 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 exists a certain probability that the value of a given parameter is within that range.

[0069] Copy number variant: As used herein, “copy number variant,” “CNV,” or “copy number polymorphism” refers to a phenomenon in which sections of the genome are repeated, and the number of repeats within the genome differs 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 having a hydrogen group at the 2' position of the sugar moiety. DNA typically consists of a chain of nucleotides containing 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 having a hydroxyl group at the 2' position of the sugar moiety. RNA typically consists of a chain of nucleotides containing 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 bind specifically 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 is joined to a second nucleic acid strand consisting of nucleotides complementary to the nucleotides of the first strand, the two strands join to form a double helix. As used herein, “nucleic acid sequencing data,” “nucleic acid sequencing information,” “sequence information,” “nucleic acid sequence,” “nucleotide sequence,” “genome sequence,” “gene sequence,” or “fragment sequence,” or “nucleic acid sequencing read” means any information or data that indicates the order and identity of nucleotide bases (e.g., adenine, guanine, cytosine, and thymine or uracil) in a nucleic acid molecule such as DNA or RNA (e.g., whole genome, whole transcriptome, exome, oligonucleotide, polynucleotide, or fragment).Please understand that this instruction intends to refer to sequence information obtained using all available technologies, platforms, or techniques, 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 electron signaling-based systems.

[0072] Detection: As used herein, “detect,” “to detect,” or “detect” refers to the act of determining the presence or presence of one or more targeted nucleic acids (e.g., nucleic acids having targeted mutations or other markers) in a sample.

[0073] Enriched Sample: As used herein, “enriched sample” refers to a sample in which a specific region of interest has been enriched. A sample can be enriched by amplifying the region of interest or by using a single-stranded DNA / RNA probe or a double-stranded DNA probe that can hybridize to the nucleic acid molecule of interest (e.g., SureSelect® probe, Agilent Technologies). In some embodiments, an enriched sample refers to a subset or portion of an enriched processed sample, which contains nucleic acid molecules derived from a cell-free polynucleotide or polynucleotide sample.

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

[0075] Epigenetic locus: As used herein, “epigenetic locus” or “epigenetic site” means a fixed location on a chromosome that exhibits a different situation or state, whether or not, with alteration of the nucleotide sequence. To avoid misunderstanding, a given epigenetic locus may coincide with a given nucleotide location or genomic region that also exhibits genetic or sequence variation (e.g., mutation). For example, a given epigenetic locus may be acetylated, methylated (e.g., modified with 5-methylcytosine (5mC), modified with 5-hydroxymethylcytosine (5hmC), etc.), ubiquitinated, phosphorylated, smoylated, ribosylated, citrullinated, or not, and may have post-translational modifications of histone or other histone mutations, etc.

[0076] Epigenetic Signature: As used herein, “epigenetic signature” means an epigenetic situation or state indicated by one or more epigenetic loci in a given DNA molecule. For example, a DNA molecule or cfDNA fragment containing a given genomic region or locus (e.g., a CTCF-binding region) may also exhibit an epigenetic pattern containing a certain number of epigenetic loci in which a portion of those DNA molecules is methylated, whereas in other examples, the corresponding epigenetic loci in other DNA molecules or cfDNA fragments containing the same genomic region are not methylated.

[0077] Germline mutation: As used herein, “germline mutation” means 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 cancer cells or at least inhibit their growth, preferably to reduce further cancer growth, shrink the size of the cancer, and / or to stimulate the immune system to eliminate the 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 modulate the duration and amplitude of the physiological immune response in peripheral tissues to minimize associated tissue damage (see, for example, Pardoll, Nature Reviews Cancer 12, 252-264 (2012)). Exemplary agents include antibodies against PD-1, PD-2, PD-L1, PD-L2, CTLA-4, OX40, B7.1, B7He, LAG3, CD137, KIR, CCR5, CD27, CD40, or CD47. Other exemplary agents include pro-inflammatory cytokines, such as IL-1β, IL-6, and TNF-α. Other exemplary agents are T cells activated against tumors, for example, 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 that involves the insertion or deletion of a nucleotide position within the genome in question.

[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 highest MAF of all somatic mutants present or observed in a given sample.

[0081] Mutant Allele Frequency: As used herein, "mutant allele frequency," "mutant allele frequency," "mutant allele ratio," "mutant allele ratio," "MAF," or "VAF" refers to the frequency at which mutant alleles occur in a given population of nucleic acids, such as samples obtained from a subject. MAF is generally expressed as a ratio or percentage.

[0082] Molecular response: As used herein, “molecular response” refers to changes in the frequency, level, or amount of one or more circulating tumor DNA (ctDNA) variant alleles observed between samples taken from a given subject at different time points.

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

[0084] Molecular non-responders: As used herein, “molecular non-responder” or “non-responder” refers to subjects with a molecular response score that does not show an increase or change of 1 or more in the frequency, level, or amount of circulating tumor DNA (ctDNA) variant alleles observed between samples taken from the subject at different time points. A threshold specifying the level of decrease (or increase) can be used to determine whether a subject is a molecular responder or a molecular non-responder. For example, a molecular responder may be a subject associated with a decrease exceeding a certain percentage change in VAF, while a non-responder may be a subject associated with an increase, no change, or decrease 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, single nucleotide variants (SNVs), copy number variants or mutations (CNVs) / abnormalities, insertions or deletions (indels), shortenings, gene fusions, transversions, translocations, frameshifts, duplications, repeat extensions, and epigenetic variants. Mutations can be germline or somatic. In some embodiments, the reference sequence for comparison is the wild-type genome sequence of the target species, typically providing the test sample, a 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 that offer improved throughput compared to conventional Sanger electrophoresis and capillary electrophoresis-based approaches, such as the ability to produce hundreds of thousands of relatively small sequence reads at once. Some examples of next-generation sequencing technologies include, but are not limited to, synthesis sequencing, ligation sequencing, and hybridization sequencing.

[0087] Nucleic acid tags: As used herein, “nucleic acid tags” refers to short nucleic acids (e.g., less than about 500, about 100, about 50, or about 10 nucleotides in length) used to label nucleic acids from different samples (e.g., representing sample indices), different nucleic acid molecules in the same sample of different types (e.g., representing molecular tags), or nucleic acids that have undergone different treatments. Nucleic acid tags may be single-stranded, double-stranded, or at least partially double-stranded. Nucleic acid tags may be the same length or of varying lengths, as required. Nucleic acid tags may also contain double-stranded molecules with one or more blunt ends, may contain 5' or 3' single-stranded regions (e.g., overhangs), and / or may contain one or more other single-stranded regions at other locations within a given molecule. Nucleic acid tags may 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 origin, morphology, or treatment of a given nucleic acid sample. Nucleic acid tags can also be used to enable pooling and / or parallel processing of multiple samples containing nucleic acids with different nucleic acid tags and / or sample indices, where the nucleic acids are subsequently deconvoluted by reading the nucleic acid tags. Nucleic acid tags may also be called molecular identifiers or tags, sample identifiers, index tags, and / or barcodes. Additionally or alternatively, nucleic acid tags can be used to distinguish different molecules within 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, each nucleic acid molecule can be tagged using a limited number of tags so that different molecules can be distinguished based on start / stop locations 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 so that it is unlikely (e.g., less than about 10%, less than about 5%, less than about 1%, or less than about 0.1%) that any two molecules have the same start / stop locations and also have the same nucleic acid tag.Some nucleic acid tags include a sample, the morphology of nucleic acid molecules within the sample, and multiple molecular identifiers for labeling nucleic acid molecules within the morphology that have the same start and stop positions. Such nucleic acid tags can be referred to using exemplary morphology "A1i", where uppercase letters indicate the sample type, Arabic numerals indicate the morphology of molecules within the sample, and lowercase Roman numerals indicate molecules within the morphology.

[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 nucleoside bonds. Typically, a polynucleotide contains at least three nucleosides. Oligonucleotides are often in size ranging from a few monomer units, e.g., 3-4 units, to several hundred monomer units. Whenever a polynucleotide is represented by a sequence of letters such as “ATGCCTG,” unless otherwise specified, the nucleotides are in 5'→3' order from left to right, and in the case of DNA, it will be understood that “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 base itself, a nucleoside, or a nucleotide containing a base, as is standard in the art.

[0089] Reference Sample: As used herein, “reference sample” or “reference cfNA sample” refers to a sample of known composition and / or a sample known to have or lack certain characteristics (e.g., known nucleic acid variants, known cellular origin, known tumor percentage, known coverage, etc.) that is analyzed together with or compared to a test sample for purposes such as evaluating the accuracy of an analytical procedure and classifying the test sample. A reference sample dataset typically contains at least about 25 to at least about 30,000 or more reference samples. In some embodiments, the reference sample dataset includes approximately 50, 75, 100, 150, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 2,500, 5,000, 7,500, 10,000, 15,000, 20,000, 25,000, 50,000, 100,000, 1,000,000, or more reference samples.

[0090] Reference Sequence: As used herein, “reference sequence” or “reference genome” refers to a known sequence used for comparison with an experimentally determined sequence. For example, a known sequence may be an entire genome, a chromosome, or any segment thereof. A reference sequence typically contains at least about 20, at least about 50, at least about 100, at least about 200, at least about 250, at least about 300, at least about 350, at least about 400, at least about 450, at least about 500, at least about 1000, or more nucleotides. A reference sequence may align with a single contiguous sequence of genome or chromosome, or it may include discontinuous segments that align with different regions of genome or chromosome. An exemplary reference sequence includes, for example, the human genome, such as hG19 and hG38.

[0091] Sample: As used herein, “Sample” means any biological sample that can be analyzed by the methods and / or systems disclosed herein. In certain embodiments of this disclosure, Sample is a fluid sample, such as whole blood or a fraction thereof, lymph, urine, and / or cerebrospinal fluid, among other types of fluids supplied with cell-free (circulating) nucleic acids that are not contained within cells or otherwise bound to cells. In certain embodiments, a fluid sample is a plasma sample, which is the fluid portion of whole blood excluding cells such as red blood cells and white blood cells. In some embodiments, a fluid sample is a serum sample, i.e., fibrinogen-free plasma. In some embodiments of this disclosure, Sample is a “non-fluid sample” or “non-plasma sample,” i.e., a biological sample other than a “fluid sample,” such as a cell and / or tissue sample supplied with nucleic acids other than cell-free nucleic acids.

[0092] Sensitivity: As used herein, “sensitivity” in the context of a given assay or method means the ability of the assay or method to detect and distinguish between targeted (e.g., cfDNA fragments derived from tumor cells) and untargeted (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 biomolecule, such as nucleic acids, such as DNA or RNA. Exemplary sequencing methods include, but are not limited to, targeted sequencing, single-molecule real-time sequencing, exon or exome sequencing, intron sequencing, electron microscope-based sequencing, panel sequencing, transistor-mediated sequencing, direct sequencing, random shotgun sequencing, Sungardideoxy termination sequencing, whole-genome sequencing, hybridization sequencing, pyrosequencing, capillary electrophoresis, gel electrophoresis, double-strand sequencing, cycle sequencing, and single-nucleotide extension sequencing. Examples of sequencing methods include solid-phase sequencing, high-throughput sequencing, large-scale parallel 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 using genetic analyzers, such as those commercially available from Illumina, Inc., Pacific Biosciences, Inc., or Applied Biosystems / Thermo Fisher Scientific.

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

[0095] Somatic Mutation: As used herein, “somatic mutation” means a genomic mutation that occurs after conception. Somatic mutations can occur in any cell of the body except germ cells and are therefore not transmitted to offspring.

[0096] Specificity: As used herein, “specificity” in the context of a diagnostic analysis or assay means the degree to which the analysis or assay detects the target analyte intended by the analysis or assay, excluding other components of a given sample.

[0097] Subclone: ​​As used herein, “subclone” in relation to nucleic acids means a subpopulation of nucleic acids (i.e., a subset of a population of nucleic acids) that contains nucleotide sequences that are substantially or completely identical to one another, 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 animals such as mammalian species (e.g., humans) or avian species (e.g., birds), or other living organisms such as plants. More specifically, Subject may be vertebrates, mammals such as mice, primates, monkeys, or humans. Animals include livestock (e.g., production cattle, dairy cows, poultry, horses, pigs, etc.), sporting animals, and companion animals (e.g., pets or support animals). Subject may be a healthy individual, an individual with or suspected of having a disease or being predisposed to a disease, or an individual in need of or suspected of needing therapy. The terms “individual” or “patient” are intended to be interchangeable with “Subject.”

[0099] For example, a subject could be an individual diagnosed with cancer, scheduled to undergo cancer therapy, and / or who has undergone at least one cancer therapy. A subject could be in remission from cancer. Another example is a subject diagnosed with an autoimmune disease. Another example is a female individual who is pregnant or planning to become pregnant and who has been diagnosed with or is suspected of having a disease, such as cancer or an autoimmune disease.

[0100] Threshold: As used herein, “threshold” refers to an independently determined value used to characterize or classify experimentally determined values.

[0101] Tumor Percentage: As used herein, “tumor percentage” refers to an estimate of the proportion of nucleic acid molecules in a given sample that originate from tumors. For example, the tumor percentage of a sample may be a measure derived from the maximum somatic mutant allele frequency (max MAF) of the sample, the coverage of the sample, the length of cfNA fragments in the sample, the epigenetic status, or other characteristics, or any other selected characteristic of the sample. In some embodiments, the tumor percentage of a sample is equal to the maximum MAF of the sample.

[0102] Value: As used herein, “Value” generally refers to an entry in a dataset and can be anything that characterizes the feature that the value refers to. This includes, but is not limited to, numbers, words or phrases, symbols (e.g., + or -) or degrees.

[0103] Any methods and materials similar or equivalent to those described herein may be used in carrying out or testing the methods and compositions, but particularly useful methods, devices, and materials are those described herein. Publications cited herein and the materials from which they are cited are incorporated herein by reference. Nothing herein should be construed as an acknowledgment that the present invention has no prior rights to such disclosure by prior art. No reference is acknowledged to constitute prior art. Discussions of references state the claims of their authors, and the applicant reserves the right to challenge the accuracy and appropriateness of the cited documents. While several publications are referenced herein, it will be clearly understood that such references do not acknowledge that any of these documents form part of the common general knowledge in the art.

[0104] Throughout this description and claims, the word “comprise” and variations such as “comprising” and “comprises” mean “including but not limited to,” and are not intended to exclude, for example, other additives, components, integers or processes. In particular, in methods described as including one or more steps or operations, each step is specifically intended to include those listed (unless the step includes a limiting term such as “consisting of”), and this means that each step is not intended to exclude, for example, other additives, components, integers or processes not listed in that step.

[0105] If the present invention provides a process comprising a plurality of sequential steps, the present invention can also provide a process in which these different steps can be carried out by different people in different locations (e.g., different countries) at very different times. II. Molecular Response Scoring

[0106] In an embodiment shown in Figure 1, a method 100 for determining a molecular response (MR) score is disclosed. The method of this disclosure 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) loading during treatment (typically 3 to 10 weeks) compared to baseline before treatment. Molecular response is associated with the patient's 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 responses. Multiple methods have been used to calculate molecular response, and there is no consensus on which method is best.

[0107] Methods and systems for evaluating response to treatment using a molecular response (MR) score are described. In one embodiment, baseline (pre-treatment) gene expression data can be obtained for multiple patients before treatment, and intra-treatment gene expression data can be obtained for multiple patients during treatment. In one embodiment, the baseline gene expression data (e.g., variant data) and / or intra-treatment gene expression data can be analyzed to determine the molecular response (MR) score. The MR score may indicate whether a patient is a responder or non-responder to the treatment. In an embodiment, the proportion of mutant alleles (MAF) may be determined as part of the MR score. In one embodiment, the variance of each MAF may be incorporated into the determination of the molecular response score. This ensures that the molecular response score contains accurate variance and provides a significant improvement in drawing correct conclusions from the molecular response score. The improvement is even more pronounced when the molecular response score is a ratio, as the ratio is sensitive to the variance in the denominator. Variance can be incorporated into the molecular response score by mathematically deriving the molecular response variance, or by simulation or sampling from the variance distribution of each variant to determine the molecular response variance. a. Isolation and extraction of cfDNA

[0108] As shown in Figure 1, at a first time T0, baseline cfDNA may be obtained from one or more baseline samples obtained from one or more subjects before treatment in step 101, and at a second time T1, treatment cfDNA may be obtained from one or more treatment samples obtained from one or more subjects after treatment in step 102. Treatment may be performed at any point after time T0. For example, treatment may be performed a few minutes, a few hours, a few days, etc., after time T0. Further examples include treatment 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, such as 1 to 24 hours, 1 to 180 days, 1 to 12 weeks, 6 to 12 months, etc.

[0109] As described herein, polynucleotides can include any type of nucleic acid, such as DNA and / or RNA. For example, if the polynucleotide is DNA, it can be genomic DNA, complementary DNA (cDNA), or any other deoxyribonucleic acid. A polynucleotide may also be a cell-free nucleic acid, such as cell-free DNA (cfDNA). For example, a polynucleotide may be circulating cfDNA. Circulating cfDNA may contain DNA that has been expelled from somatic cells via apoptosis or necrosis. cfDNA expulsion via apoptosis or necrosis may originate from normal (e.g., healthy) somatic cells. Tumor DNA may be expelled in cases of abnormal tissue growth, such as cancer. Circulating cfDNA may contain circulating tumor DNA (ctDNA). i. Sample

[0110] The isolation and extraction of cell-free polynucleotides can be performed by collecting samples using various techniques. Samples may include body tissues, 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, lymph, ascites fluid, interstitial fluid or extracellular fluid (e.g., fluids from intercellular spaces), gingival exudate, crevicular fluid, bone marrow, pleural fluid, cerebrospinal fluid, saliva, mucous membranes, sputum, semen, sweat, and urine. Preferably, samples are body fluids, particularly blood and its fractions, as well as urine. Such samples may contain nucleic acids excreted from tumors. Nucleic acids can include DNA and RNA and may be in double-stranded and single-stranded forms. The sample may be in the form initially isolated from the subject, or it may have been subjected to further processing to remove or add components such as cells, concentrate one component against another, or convert one form of nucleic acid to another, for example, RNA to DNA, or single-stranded nucleic acid to double-stranded nucleic acid. Therefore, for example, a bodily fluid sample for analysis may be plasma or serum containing cell-free nucleic acid, such as cell-free DNA (cfDNA).

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

[0112] A sample can contain varying amounts of nucleic acids. Typically, the amount of nucleic acid in a given sample is equivalent to several genome equivalents. For example, a sample with approximately 30 ng of DNA is equivalent to approximately 10,000 (10) genome equivalents. 4 ) haploid human genome equivalents, and in the case of cfDNA, approximately 200 billion (2 x 10⁶)11 It may contain ) individual polynucleotide molecules. Similarly, a sample of about 100 ng of DNA may contain about 30,000 haploid human genome equivalents, and in the case of cfDNA, about 600 billion individual molecules.

[0113] In some embodiments, the sample comprises nucleic acids from different sources, e.g., cellular and cell-free sources (e.g., blood samples). Typically, the sample comprises nucleic acids with mutations. For example, the sample may optionally include DNA harboring germline mutations and / or somatic mutations. Typically, the sample comprises DNA harboring cancer-associated mutations (e.g., cancer-associated somatic mutations). In some embodiments of this disclosure, cell-free nucleic acids in the subject may originate from a tumor. For example, cell-free DNA isolated from the subject may include ctDNA.

[0114] Exemplary amounts of cell-free nucleic acids in the sample before amplification are typically in the range of about 1 femtogram (fg) to about 1 microgram (μg), e.g., about 1 picogram (pg) to about 200 nanograms (ng), about 1 ng to about 100 ng, or about 10 ng to about 1000 ng. In some embodiments, the sample contains cell-free nucleic acid molecules up to about 600 ng, about 500 ng, about 400 ng, about 300 ng, about 200 ng, about 100 ng, about 50 ng, or about 20 ng. If necessary, the amount is at least about 1 fg, at least about 10 fg, at least about 100 fg, at least about 1 pg, at least about 10 pg, at least about 100 pg, at least about 1 ng, at least about 10 ng, at least about 100 ng, at least about 150 ng, or at least about 200 ng of 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 molecules. In some embodiments, the method involves obtaining about 1 fg to about 200 ng of cell-free nucleic acid molecules from a sample.

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

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

[0117] In certain embodiments, tags providing molecular identifiers or barcodes are incorporated into or otherwise attached to the adapter by other means, including chemical synthesis, ligation, or overlap extension PCR. In some embodiments, the assignment of unique or non-unique identifiers or molecular barcodes during the reaction is described, for example, in U.S. Patent Applications 20010053519, 20030152490, 20110160078, and U.S. Patents 6,582,908, 7,537,898, and 9,598,731, which are incorporated by reference, respectively.

[0118] The tags are attached to the sample nucleic acids randomly or non-randomly (e.g., ligated). In some embodiments, the tags are introduced in an expected ratio of identifiers (e.g., unique and / or non-unique barcode combinations) per microwell. For example, identifiers may be loaded so that approximately 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 500, 1000, 5000, 100,000, 500,000, 1,000,000, 10,000,000, 50,000,000 or more identifiers are loaded per genomic sample. In some embodiments, identifiers are loaded so that approximately 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 fewer than 1,000,000,000 identifiers are loaded per genome sample. In a particular embodiment, the average number of identifiers loaded per genome sample is less than or greater than approximately 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 20, 50, 100, 500, 1000, 5000, 100,000, 500,000, 1,000,000, 10,000,000, 50,000,000, or 1,000,000,000 identifiers per genome sample. Identifiers are generally unique or non-unique.

[0119] One exemplary format uses approximately 2 to 1,000,000 different tags, or approximately 5 to 150 different tags, or approximately 20 to 50 different tags, attached to both ends of a target nucleic acid molecule. In the case of 20 to 50 × 20 to 50 tags, a total of 400 to 2,500 tags are produced. Such a number of tags is typically sufficient to ensure that different molecules with the same start and stop points receive different combinations of tags with a high probability (e.g., at least 94%, 99.5%, 99.99%, 99.999%).

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

[0121] Sample nucleic acids adjacent to the adapter are typically amplified by PCR and other amplification methods that use nucleic acid primers that bind to primer binding sites in the adapter adjacent to the DNA molecule to be amplified. In some embodiments, the amplification method may involve a cycle of extension, denaturation, and annealing resulting from thermal cycling, or it may be isothermal, as in the case of transcription-mediated amplification. Other exemplary amplification methods that may be used as needed include, among other approaches, ligase chain reaction, strand displacement amplification, nucleic acid sequence-based amplification, and autologous sequence-based replication.

[0122] Conventional nucleic acid amplification methods typically involve one or more amplification cycles to introduce a sample index / tag into a nucleic acid molecule. Amplification is typically performed in one or more reaction mixtures. In some embodiments, the molecular tag and 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 probe capture, and the sample index / tag is introduced after the sequence capture step is performed. In certain embodiments, both the molecular tag and sample index / tag are introduced before the probe-based capture step is performed. In some embodiments, the sample index / tag is introduced after the sequence capture step (i.e., nucleic acid enrichment) is performed. Typically, the sequence capture protocol involves introducing a single-stranded nucleic acid molecule complementary to a targeted nucleic acid sequence, such as a coding sequence for a genomic region and mutations in such regions associated with oncogenicity. Typically, the amplification reaction produces multiple non-uniquely or uniquely tagged nucleic acid amplicons with molecular tags and sample index / tags ranging in size from approximately 200 nucleotides (nt) to approximately 700 nt, 250 nt to approximately 350 nt, or approximately 320 nt to approximately 550 nt. In some embodiments, the amplicons have a size of approximately 300 nt. In some embodiments, the amplicons have a size of approximately 500 nt. iv. Nucleic acid enrichment

[0123] In some embodiments, sequences are enriched before sequencing the nucleic acids. The enrichment may be performed on a specific target region or non-specifically ("target sequence"), as needed. For example, enrichment may be performed non-specifically based on a size selection method that is sequence fragment size specific rather than sequence specific. In some embodiments, the targeted region of interest may be enriched with nucleic acid capture probes ("baits") selected for one or more bait set panels using a differential tiling and capture scheme. A differential tiling and capture scheme typically uses bait sets of different relative concentrations to differentially tile (e.g., in different "dissolutions") across genomic sections associated with the baits, according to a set of constraints (e.g., sequencing load, usefulness of each bait, etc.) and capture the targeted nucleic acids at desired levels for downstream sequencing. These targeted genomic sections of interest may optionally contain native or synthetic nucleotide sequences of the nucleic acid construct. In some embodiments, biotin-labeled beads having probes for one or more sections of interest may be used to capture the target sequences and, if necessary, subsequently amplify those sections to enrich the region of interest.

[0124] Sequence capture typically involves the use of oligonucleotide probes that hybridize to a target nucleic acid sequence. In certain embodiments, probe set strategies involve tiling probes across a desired section. Such probes may be, for example, about 60 to about 120 nucleotides long. Sets can have depths of about 2×, 3×, 4×, 5×, 6×, 8×, 9×, 10×, 15×, 20×, 50× 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 probe sequence. b. Nucleic acid sequencing

[0125] As shown in Figure 1, after extracting and isolating cfDNA from the sample in steps 101 and 102, the cfDNA can be sequenced in steps 103 and 104. Sample nucleic acids adjacent to the adapter are generally sequenced as needed, with or without pre-amplification. Sequence determination methods or commercially available formats that may be used as needed include, for example, Sanger sequencing, high-throughput sequencing, bisulfite sequencing, pyrosequencing, synthesis sequencing, single-molecule sequencing, nanopore-based sequencing, semiconductor sequencing, ligation sequencing, hybridization sequencing, RNA-Seq (Illumina), digital gene expression (Helicos), next-generation sequencing (NGS), synthesis single-molecule sequencing (SMSS) (Helicos), large-scale parallel sequencing, cloned single-molecule array (Solexa), shotgun sequencing, Ion Torrent, Oxford Nanopore, Roche Genia, primer walking, PacBio, SOLiD, Ion Torrent, or sequencing using a nanopore platform. Sequence determination reactions can be carried out in various sample processing units that may include multiple lanes, multiple channels, multiple wells, or other means of processing multiple sample sets substantially simultaneously. The sample processing unit may also include multiple sample chambers to enable simultaneous processing of multiple runs.

[0126] Sequencing reactions can be performed on one or more nucleic acid fragment types or fragments known to contain markers for cancer or other diseases. Sequencing reactions can also be performed on any nucleic acid fragment present in the sample. Sequencing reactions may provide sequence coverage of at least about 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9%, or 100% of the genome. In other cases, genome sequence coverage may 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 reactions can be carried out using multiple sequencing techniques. In some embodiments, cell-free polynucleotides are sequenced in at least about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. In other embodiments, cell-free polynucleotides are sequenced in about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or fewer than 100,000 sequencing reactions. Sequencing reactions are typically performed sequentially or simultaneously. Subsequent data analysis is generally performed on all or part of the sequencing reactions. In some embodiments, data analysis is performed for at least about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. In other embodiments, data analysis may be performed for less than about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, or 100,000 sequencing reactions. An exemplary read depth is about 1,000 to about 50,000 reads per locus (base location).

[0128] In some embodiments, nucleic acid populations are prepared for sequencing by enzymatically forming blunt ends on double-stranded nucleic acids having single-stranded overhangs at one or both ends. In these embodiments, the 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). Optionally, exemplary enzymes or catalytic fragments thereof used include Krenow large fragments and T4 polymerase. For the 5' overhang, the enzyme typically produces a blunt end by extending the recessed 3' end on the opposing strand until it is coplanar with the 5' end. For the 3' overhang, the enzyme generally digests from the 3' end to the 5' end of the opposite strand, and possibly beyond. If this digestion proceeds beyond the 5' end of the opposing strand, the gap can be filled with an enzyme having the same polymerase activity used for the 5' overhang. The formation of blunt ends on double-stranded nucleic acids facilitates, for example, adapter binding and subsequent amplification.

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

[0130] Regardless of whether prior amplification has been performed, sequenced nucleic acids can be generated by sequencing the nucleic acids targeted in the blunt-end formation process described above, and, if necessary, other nucleic acids in the sample. Sequenced nucleic acids can refer to either the sequence of the nucleic acid (i.e., sequence information) or the nucleic acid whose sequence has been determined. Sequencing can be performed to provide sequence data for individual nucleic acid molecules in the sample, either directly or indirectly, from the consensus sequences of the amplified products of individual nucleic acid molecules in the sample.

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

[0132] Nucleic acid samples are typically brought into contact with a sufficient number of adapters such that the probability of any two copies of the same nucleic acid receiving the same combination of adapter barcodes 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 families 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 families represent the sequences of the amplified product of the template / parent nucleic acid in the sample before amplification. The sequences of family members can be edited to obtain consensus nucleotides or complete consensus sequences 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 in the nucleic acid in the sample is determined to be the consensus of the nucleotides occupying its corresponding position in the family member sequence. Families may contain sequences from one or both strands of a double-stranded nucleic acid. If a family member contains sequences from both strands derived from a double-stranded nucleic acid, the sequences from one strand are converted to their complements for the purpose of editing all sequences to obtain consensus nucleotides or sequences. Some families contain only a single member sequence. In this case, this sequence can be used as the nucleic acid sequence in the sample before amplification. Alternatively, families containing only a single member sequence can be excluded from subsequent analysis.

[0133] Nucleotide mutations in sequenced nucleic acids can be determined by comparing the sequenced nucleic acid with a reference sequence. The reference sequence is often a known sequence, such as a known whole-genome or partial-genome sequence from the subject (e.g., the whole-genome sequence of a human subject). The reference sequence may be, for example, hG19 or hG38. The sequenced nucleic acid can represent a sequence directly determined for the nucleic acid in the sample, as described above, or a consensus of sequences of amplified products of such nucleic acids. The comparison can be performed at one or more specified positions on the reference sequence. A subset of sequenced nucleic acids can be identified that contains positions corresponding to specified positions on the reference sequence when each sequence is maximally aligned. Within such a subset, it can be determined which sequenced nucleic acids, if present, contain nucleotide mutations at specified positions, the length of a given cfDNA fragment based on where its endpoints (i.e., 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 contain the reference nucleotide (i.e., the same as the reference sequence). If there are enough sequenced nucleic acids in a subset containing nucleotide variants to exceed a selected threshold, the variant nucleotide can be called at a specified position. 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 containing nucleotide variants, or among other possibilities, a ratio of at least 0.5, 1, 2, 3, 4, 5, 10, 15, or 20 sequenced nucleic acids in the subset containing nucleotide variants. The comparison can be repeated for any specified position of interest in the reference sequence. Sometimes, the comparison can be performed for specified positions occupying at least approximately 20, 100, 200, or 300 consecutive positions on the reference sequence, for example, approximately 20–500 or approximately 50–300 consecutive positions.

[0134] Further details regarding nucleic acid sequencing, including the formats and applications described herein, are also incorporated in their entirety by reference, for example, Levy et al., Annual Review of Genomics and Human Genetics, 17:95-115 (2016), Liu et al., J. of Biomedicine and Biotechnology, Volume 2012, Article ID 251364:1-11 (2012), Voelkerding et al., Clinical Chem., 55:641-658 (2009), MacLean et al., Nature Rev. Microbiol., 7:287-296 (2009), and Astier et al., J Am Chem. Soc., 128(5):1705-10(2006), U.S. Patent Nos. 6,210,891, 6,258,568, 6,833,246, 7,115,400, 6,969,488, 5,912,148, 6,130,073, 7,169,560, 7,282,337, U.S. This is provided in Japanese Patent No. 7,482,120, U.S. Patent No. 7,501,245, U.S. Patent No. 6,818,395, U.S. Patent No. 6,911,345, U.S. Patent No. 7,501,245, U.S. Patent No. 7,329,492, U.S. Patent No. 7,170,050, U.S. Patent No. 7,302,146, U.S. Patent No. 7,313,308, and U.S. Patent No. 7,476,503. i. Sequencing Panel

[0135] To improve the likelihood of detecting a target genomic region and, if necessary, a tumor exhibiting mutations, the sequenced DNA section may include a panel of genes or genomic sections containing known genomic regions. Selecting a limited section for sequencing (e.g., a limited panel) can reduce the total sequencing required (e.g., the total amount of nucleotides sequenced). The sequencing panel may target multiple different genes or regions to detect, for example, a single cancer, a set of cancers, or all cancers. Alternatively, the DNA may be sequenced by whole-genome sequencing (WGS) or other unbiased sequencing methods without the use of a sequencing panel. Examples of suitable panels and targets for use in the panel 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), repeat elements, intron-exon junctions, transcription start sites (TSSs), etc.) is selected such that a determined proportion of subjects with cancer exhibits genetic variants or tumor markers in one or more different genes within 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 be further 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 given amount of sequenced base pairs. The panel may be selected to achieve theoretical sensitivity, theoretical specificity, and / or theoretical precision for detecting one or more genetic variants in a sample.

[0137] Probes for detecting a panel of regions may include probes for detecting genomic regions of interest (hotspot regions) as well as nucleosome recognition probes (e.g., KRAS codons 12 and 13), and may be designed to optimize capture based on analysis of cfDNA coverage and fragment size variations influenced by nucleosome binding patterns and GC sequence composition. Regions used herein may also include non-hotspot regions optimized based on nucleosome location and GC model. A panel may include multiple subpanels, including subpanels for identifying origin tissue (e.g., using published literature to define 50-100 baits representing genes with the most diverse transcription profiles across the tissue (not necessarily promoters)), whole-genome backbone (e.g., for identifying hyperconservative genomic contents and sparsely tiling across chromosomes with a handful of probes for copy number-based lining purposes), and transcription start site (TSS) / CpG islands (e.g., for capturing differentially methylated regions (DMRs) in promoters of tumor suppressor genes (e.g., SEPT9 / VIM in colorectal cancer)). In some embodiments, the marker for the originating tissue is a tissue-specific epigenetic marker.

[0138] Some examples of lists of target genomic locations can be found in Tables 1 and 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, at least 75, at least 80, at least 85, at least 90, at least 95, or 97 portions of the genes in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all 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 SNVs in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all 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 CNVs in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all CNVs 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, or 6 fusions in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all fusions in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include at least 1, at least 2, or at least 3 portions of indels in Table 1. In some embodiments, the genomic locations used in the methods of the present disclosure include all indels in Table 1.In one embodiment, the genomic locations used in the methods of the present disclosure include all genes, SNVs, CNVs, fusions, and indels 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, 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 portions of the genes in Table 2. In embodiments, the genomic locations used in the methods of the present disclosure include all genes in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all genes in Tables 1 and 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 SNVs in Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include all SNVs in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all SNVs in Tables 1 and 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 CNVs in Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include all CNVs in Table 2. In one embodiment, the genomic locations used in the method of the present disclosure include all CNVs in Tables 1 and 2. In some embodiments, the genomic locations used in the method of the present disclosure include at least one, at least two, at least three, at least four, at least five, or six fusions in Table 2.In some embodiments, the genomic locations used in the methods of the present disclosure include all fusions in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all fusions in Tables 1 and 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 at least 18 portions of the indels in Table 2. In some embodiments, the genomic locations used in the methods of the present disclosure include all indels in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all indels in Tables 1 and 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all genes, SNVs, CNVs, fusions, and indels in Table 2. In one embodiment, the genomic locations used in the methods of the present disclosure include all genes, SNVs, CNVs, fusions, and indels in Tables 1 and 2. Each of these target genomic locations can be identified as a mainchain region or hotspot region in a given bait set panel. [Table 1] Table 2 [Table 2-1] [Table 2-2]

[0139] In some embodiments, one or more regions within the panel include one or more loci from one or more genes for detecting residual cancer after surgery. This detection can be performed faster than existing cancer detection methods. In some embodiments, one or more genomic locations within 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. Furthermore, smokers may develop other pulmonary conditions that make cancer detection more difficult, such as the development of irregular nodules in the lungs. In some embodiments, the methods described herein can detect a patient's response to cancer therapy (particularly in high-risk patients) earlier than is possible with existing methods of cancer detection.

[0140] Genomic locations may be selected for inclusion in a sequencing panel based on several subjects with cancer that have tumor markers in that gene or region. Genomic locations may 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 particular region may indicate that a subject has cancer.

[0141] In some cases, the panel may be selected using information from one or more databases. Information about cancer may come from cancer tumor biopsies or cfDNA assays. Databases may contain information describing populations of sequenced tumor samples. Databases may contain information about mRNA expression in tumor samples. Databases may contain information about regulatory elements or genomic regions in tumor samples. Information about sequenced tumor samples may include the frequencies of various genetic variants and may describe the genes or regions in which genetic variants exist. Genetic variants may be tumor markers. A non-limiting example of such a database is COSMIC, 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 the panel by having high-frequency mutations within a given gene. For example, COSMIC shows that 33% of a population of sequenced breast cancer samples have mutations in TP53, and 22% of a sampled population of 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 the sequencing panel based on their relatively high frequency in the sampled breast cancers (compared to APC, which occurs at approximately 4% frequency, for example). Any database or set of information that associates cancer with tumor markers located in genes or gene regions may be used, although COSMIC is provided as a non-limiting example. In another example, as provided by COSMIC, 380 out of 1156 biliary tract cancer samples (33%) had mutations in TP53. Several other genes, such as APC, have mutations in 4–8% of all samples. Therefore, TP53 may be selected for inclusion in the panel based on its relatively high frequency in the population of biliary tract cancer samples.

[0142] For panels where the frequency of tumor markers is significantly higher in sampled tumor tissue or circulating tumor DNA than in a given background population, genes or genomic sections may be selected. Combinations of genomic locations may be selected to include panels such that at least the majority of subjects with cancer may have a tumor marker or genomic region present in at least one genomic location or gene within the panel. Combinations of genomic locations may be selected based on data showing that for a particular cancer or set of cancers, the 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 showing 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 of subjects with cancer so that tumor markers in two or more regions are present in the 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 a tumor marker in one or more regions, that in 30% of such subjects the tumor marker is detected only in region X, and that for the remaining subjects where the tumor marker is detected, the tumor marker is detected only in regions Y and / or Z. Tumor markers located in one or more genomic locations previously shown to be associated with one or more cancers can indicate or predict subjects with cancer if the tumor marker is detected in one or more regions of those regions for 50% or more of the time. Computational approaches, such as models that use conditional probabilities of detecting cancer considering the cancer frequency for a set of tumor markers within one or more regions, can be used to predict which regions, individually or in combination, may predict cancer.Other approaches to panel selection include the use of databases that document information from studies employing comprehensive genomic profiling of tumors by large panels and / or whole-genome sequencing (WGS, RNA-seq, Chip-seq, bisulfate sequencing, ATAC-seq, etc.). Information gathered from the literature may also describe commonly affected and mutated pathways in specific cancers. Panel selection may be further informed by the use of ontologs that describe genetic information.

[0143] The genes included in the sequencing panel may include fully transcribed regions, promoter regions, enhancer regions, regulatory elements, and / or downstream sequences. To further increase the likelihood of detecting tumors exhibiting mutations, only exons may be included in the panel. The panel may include all exons of a selected gene, or only one or more exons of a selected gene. The panel may include exons from each of several different genes. The panel may include at least one exon from each of several different genes.

[0144] In some embodiments, a panel of exons from each of several different genes is selected such that a determined proportion of subjects with cancer exhibit 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 gene panel can be sequenced. The sequenced panel may contain exons from multiple genes. The panel may contain exons from 2 to 100 different genes, 2 to 70 genes, 2 to 50 genes, 2 to 30 genes, 2 to 15 genes, or 2 to 10 genes.

[0146] The selected panel may contain a variety of exons. The panel may contain 2 to 3000 exons. The panel may contain 2 to 1000 exons. The panel may contain 2 to 500 exons. The panel may contain 2 to 100 exons. The panel may contain 2 to 50 exons. The panel may contain 300 or fewer exons. The panel may contain 200 or fewer exons. The panel may contain 100 or fewer exons. The panel may contain 50 or fewer exons. The panel may contain 40 or fewer exons. The panel may contain 30 or fewer exons. The panel may contain 25 or fewer exons. The panel may contain 20 or fewer exons. The panel may contain 15 or fewer exons. The panel may contain 10 or fewer exons. The panel may contain 9 or fewer exons. A panel may contain 8 or fewer exons. A panel may contain 7 or fewer exons.

[0147] The panel may contain one or more exons from multiple different genes. The panel may contain one or more exons from each of multiple different genes in proportions. The panel may contain at least two exons from each of at least 25%, 50%, 75%, or 90% of different genes. The panel may contain at least three exons from each of at least 25%, 50%, 75%, or 90% of different genes. The panel may contain at least four exons from each of at least 25%, 50%, 75%, or 90% of 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 may be 5kb to 50kb in size. A sequencing panel may be 10kb to 30kb in size. A sequencing panel may be 12kb to 20kb in size. A sequencing panel may be 12kb to 60kb in size. A sequencing panel may be at least 10kb, 12kb, 15kb, 20kb, 25kb, 30kb, 35kb, 40kb, 45kb, 50kb, 60kb, 70kb, 80kb, 90kb, 100kb, 110kb, 120kb, 130kb, 140kb, or 150kb in size. Sequencing panels can be 100kb, 90kb, 80kb, 70kb, 60kb, or less than 50kb in size.

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

[0150] The panels selected herein can enable 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 may be referred to in terms of the mutant allele frequency of a given genetic variant. Mutant allele frequency may refer to the frequency at which a mutant allele (e.g., not the most common allele) occurs in a given population of nucleic acids, such as a sample. Genetic variants with low mutant allele frequencies may be present in relatively low frequencies in a sample. In some cases, the panel enables the detection of genetic variants at mutant allele frequencies of at least 0.0001%, 0.001%, 0.005%, 0.01%, 0.05%, 0.1%, or 0.5%. The panel can enable the detection of genetic variants at mutant allele frequencies of 0.001% or more. The panel can enable the detection of genetic variants at mutant allele frequencies of 0.01% or more. The panel can enable the detection of genetic variants present in a sample at low frequencies of 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 can enable the 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 can enable the detection of tumor markers in a sample at a low frequency of 1.0%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.75%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.5%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.25%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.1%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.075%.The panel can enable the detection of tumor markers in a sample at a low frequency of 0.05%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.025%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.01%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.005%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.001%. The panel can enable the detection of tumor markers in a sample at a low frequency of 0.0001%. The panel can enable the detection of tumor markers in sequenced cfDNA in a sample at a low frequency of 1.0% to 0.0001%. The panel can enable the detection of tumor markers in sequenced cfDNA in a sample at a low frequency of 0.01% to 0.0001%.

[0151] Genetic variants can be expressed as a percentage of the target population having a 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 exhibit one or more genetic variants in at least one region of the panel. For example, at least 80% of the population with cancer may exhibit one or more genetic variants in at least one genomic location within the panel.

[0152] The panel may include one or more locations containing the target genomic region from each of one or more genes. In some cases, the panel may include one or more locations containing the target genomic region 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, the panel may include one or more locations containing the target genomic region 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, the panel may include one or more locations containing the target genomic region from each of about 1 to about 80, 1 to about 50, about 3 to about 40, 5 to about 30, or 10 to about 20 different genes.

[0153] Locations within the panel containing genomic regions can be selected to detect one or more epigenetic modification regions. These one or more epigenetic modification regions may be acetylated, methylated, ubiquitinated, phosphorylated, smoylated, ribosylated, and / or citrullinated. For example, regions within the panel can be selected to detect one or more methylated regions.

[0154] Regions within a panel can be selected to contain sequences that are differentially transcribed across one or more tissues. In some cases, locations containing genomic regions may contain sequences that are transcribed at a higher level in a particular tissue compared to other tissues. For example, locations containing genomic regions may contain sequences that are transcribed in a particular tissue but not in other tissues.

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

[0156] Genomic locations within the panel can be selected to detect (diagnose) cancer with a desired level of sensitivity (e.g., through the detection of one or more genetic variants). For example, regions within the panel can be selected to detect cancer (e.g., through the detection of one or more genetic variants) 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%. Genomic locations within the panel can also 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 the detection of one or more genetic variants). For example, genomic locations within the panel can be selected to detect cancer with specificity of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% (e.g., through the 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) cancers with a desired positive predictive value. Positive predictive values ​​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 positive predictive values ​​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 locations within the panel can be selected to detect (diagnose) cancer with the desired accuracy. As used herein, the term “accuracy” may refer to the ability of a test to distinguish between disease symptoms (e.g., cancer) and health symptoms. Accuracy may be quantified using measures such as sensitivity and specificity, predictive value, likelihood ratio, area under the ROC curve, Iorden’s index, and / or diagnostic odds ratio.

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

[0161] The panel may be selected to be highly sensitive and to detect low-frequency genetic variants. For example, the panel may be selected so that genetic variants or tumor markers present in the sample at low frequencies of 0.01%, 0.05%, or 0.001% can be detected 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%. Genomic locations within the panel may be selected so that tumor markers present in the sample at a frequency of 1% or less can be detected with a sensitivity of 70% or more. The panel may be selected to detect tumor markers present at a low frequency of 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%. The panel may be selected to detect tumor markers present at a low frequency of 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 present at a low frequency of 0.001% in the 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] Panels can be selected to be highly specific and to detect low-frequency genetic variants. For example, a panel can be selected to detect genetic variants or tumor markers present in a sample at low frequencies of 0.01%, 0.05%, or 0.001% with specificity of at least 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9%. Genomic locations within a panel can be selected to detect tumor markers present in a sample at frequencies of 1% or less with specificity of 70% or more. Panels can be selected to detect tumor markers in a sample at low frequencies of 99% with specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.1%, 99.5%, or 99.9%. The panel may be selected to detect tumor markers in a sample at a low frequency of 99% with specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.01%, 99.5%, or 99.9%. The panel may be selected to detect tumor markers in a sample at a low frequency of 99% with specificity of at least 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 0.001%, 99.5%, or 99.9%.

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

[0164] The panel may be selected to be highly predictive and to detect low-frequency genetic variants. The panel may be selected so that genetic variants or tumor markers present in the sample at low frequencies of 0.01%, 0.05%, or 0.001% may have positive predictive values ​​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 the 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 greater. The probe concentration may be approximately 2 ng / μL–3 ng / μL, approximately 2 ng / μL–4 ng / μL, approximately 2 ng / μL–5 ng / μL, or approximately 2 ng / μL–6 ng / μL. The concentration of the probe or bait used in the panel may be 2 ng / μL or greater, and 6 ng / μL or less. In some cases, this may allow for the analysis of more molecules within the biological sample, thereby potentially enabling the detection of lower-frequency alleles.

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

[0167] As shown in Figure 1, after cfDNA sequencing of the sample in steps 103 and / or 104, the mutant allele ratio (MAF) of 1 or more may be determined in steps 105 and / or 106. Partial or complete MAF determination may be performed before, after, or during variant classification 107 / 108, before, after, or during variant filtering 109, or during variant filtering 109, or in combination thereof. Before step 103, the cfDNA may be end-repaired, ligated with an adapter containing a molecular barcode, amplified, and enriched. Amplification may incorporate a sample index. In embodiments, the MAF value may be determined for all variants or all somatic variants. In embodiments, the MAF value may be determined for less than all variants or less than all somatic variants. The mutant allele ratio (VAF) is used herein interchangeably with MAF. The mutant allele ratio (MAF) represents the number of mutant molecules divided by the total number of molecules at a particular genomic location (e.g., molecular coverage):

number

[0168] The maximum percentage of diploid genes ("max frac_diploid") (minimum allele imbalance) can be determined. The percentage of diploid genes ("frac_diploid") is a measure of the level of allele imbalance across samples, determined by copy number. Samples with high levels of allele imbalance are prone to germline / somatic misclassification. Therefore, low levels of allele imbalance (or high frac_diploid) are an indicator of the reliability of somatic classification calls.

[0169] In one embodiment, the total coverage profile may be used to capture ploidy changes, and therefore tumor percentages, rather than individual genes. d. Mutant Classification

[0170] The sequencing in steps 103 and 104 generates multiple sequence reads. In steps 107 and / or 108, the multiple sequence reads can be analyzed to determine and classify one or more variants. In embodiments, some or all variant classification may be determined before MAF determination 105 / 106, after MAF determination 105 / 106, during MAF determination 105 / 106, or a combination thereof. Variants may include, for example, single nucleotide variants (SNVs), indels, fusions, and copy number variants. Any known technique for variant calling may be used. In embodiments, multiple sequence reads from a sample may be assembled and / or mapped and aligned to genomic locations relative to a reference genome. In some embodiments, the multiple sequence reads (assembled or unassembled) may then be compared to the reference genome to determine how the multiple sequence reads of interest differ from the sequence reads of the reference genome. Such a process may determine the presence of one or more variants in the multiple sequence reads. In some embodiments, molecular barcodes and / or start and stop genome locations of nucleic acid molecules obtained from multiple sequence reads can be used to identify mutant molecules in which the sequence reads belonging to that molecule differ from those in the reference genome. Such a process can determine the presence of one or more variants in multiple sequence reads.

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

[0172] In one embodiment, as shown in Figure 1, one or more filtering processes may be applied to sequence reads in step 109 to exclude them from further analysis. In one embodiment, some or all 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 mutants having MAFs less than approximately 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 mutants 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 mutants 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 embodiments, copy number variants may be used to exclude sequence reads from further analysis. Copy number amplification may be determined as known in the art. In step 109, method 100 may exclude copy number amplification in genes having either insufficient probe coverage or insufficient copy number (e.g., below the 95% detection limit).

[0175] For example, CNV can be determined by analyzing sequence reads to create 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 of sequencing read coverage is a measure indicating the number of reads derived from DNA molecules corresponding to a locus (e.g., a specific location, base, region, gene, or chromosome from a reference genome). To associate reads with loci, reads can be mapped or aligned to a reference. Software for performing mapping or alignment (e.g., Bowtie, BWA, mrsFAST, BLAST, BLAT) can associate sequencing reads with loci. After sequence read coverage has been determined, probabilistic modeling algorithms can be applied to convert the normalized nucleic acid sequence read coverage for each window / bin region into discrete copy number situations. In some cases, this 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 situation of each window region can be used to identify copy number variations in chromosomal regions. 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. In some cases, different windows / bins can be filtered before being merged with other segments. Copy number variations can be used to report a percentage score indicating the degree of disease material (or nucleic acids with copy number variations) present in cell-free polynucleotide samples.

[0176] In one embodiment, the presence of CNVs in one or more genes may be used to exclude mutants from further analysis. For example, a mutant having a threshold number of LDT-reportable genes with a copy number greater than or equal to the gene-specific 95% limit of detection (LoD) in either a T0 or T1 sample. 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 can represent the multiplicative change of a given mutant. Using a Gaussian model, the ratio of the multiplicative change between time T0 and time T1 can be determined and may be used as an estimate of the molecular response score.

[0178] In embodiments, if a subject does not have somatic variants or does not have variants that meet the criteria for the variant filtering process, the subject may be classified as unevaluable. In embodiments, subjects classified as unevaluable may be further classified as molecular responders. In embodiments, subjects with low ctDNA at both time T0 and time T1 may be classified as unevaluable and further classified as molecular responders. In embodiments, subjects with low MAF at both time T0 and time T1 may be classified as unevaluable and further classified as molecular responders. In embodiments, subjects with low tumor percentage at both time T0 and time T1 may be classified as unevaluable and further classified as molecular responders. Low MAF or low tumor percentage may refer to MAF or tumor percentage 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 Figure 2, the germline filter 200 may be applied to sequence reads. Some (e.g., fewer than all) or all of the steps shown in Figure 2 may be performed in any combination and any order. Samples collected over the course of the treatment of the subject (e.g., samples collected at time T0 and time T1) may have different levels of tumor shedding and allelic imbalance, meaning that the variant classification in steps 107 / 108 may tend to assign different somatic classifications to the same variant of the same subject. Since the objective of the molecular response is to track somatic variants over the course of the treatment, classification discrepancies can be automatically resolved in order to appropriately remove germline variants from consideration by reclassifying the 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 not at time T1. For example, a mutant may be classified as somatic at time T0 but not at time T1. The germline filter 200 is configured to resolve such discrepancies and reassign mutant classifications.

[0180] As shown in Figure 2, in step 201, for at least one variant in the sequence read, a determination may be made as to whether the variant is a harmful 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 harmful variant of a TSG, the variant may be classified as somatic regardless of the classification result in step 107 / 108 (e.g., the classification will be changed from germline to somatic).

[0181] If the mutant is not a harmful TSG mutant, the germline filter 200 may determine in step 202 the maximum MAF of the mutant present in the sample and the maximum proportion of diploid genes for at least one mutant in the sample. In step 203, if the maximum proportion of diploid genes for the mutant (at least one of two time points) indicates that the mutant is somatic and that the MAF for the mutant (at least one of two time points) does not increase the maximum MAF, then the mutant 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 mutant (at least one of two time points) indicates that the mutant is germline and that the MAF for the mutant (at least one of two time points) would increase the maximum MAF, then the mutant 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] In step 203, if the maximum proportion of diploid genes to the mutant indicates that the mutant is somatic and that MAF to the mutant increases the maximum MAF, or if the maximum proportion of diploid genes to the mutant indicates that the mutant is germline and that MAF to the mutant does not increase the maximum MAF, the germline filter 200 may in step 204 determine whether the mutant is classified as somatic in another patient sample below a threshold percentage (at least at one of two time points). The threshold percentage may be at least about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, or 9%. If the mutant is classified as somatic in another patient sample below a threshold percentage, the mutant may be classified as somatic regardless of the classification result in step 107 / 108 (e.g., the classification will be changed from germline to somatic).

[0183] In step 204, if the mutant is less than 5% and not classified as somatic in another patient sample, the germline filter 200 may, in step 205, determine whether the MAF of the mutant is greater than another MAF in the sample (at least at one of two time points). For example, the germline filter 200 may determine whether the MAF for the mutant is at least about 2 times, 3 times, 4 times, 5 times, 6 times, 7 times, 8 times, 9 times, or at least 10 times greater than one or more other MAFs in the same sample. One or more other MAFs in the sample could be, for example, the next highest somatic MAF after the highest MAF in the sample. If the MAF for the mutant is greater than another MAF in the sample, the mutant may be classified as germline regardless of the classification result in step 107 / 108 (for example, the classification would be changed from somatic to germline).

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

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

[0186] These variants, classified as germline variants, may be excluded from further analysis, such as 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 can contain aggregates of cfDNA from any cell type, including tumors and blood cells. 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-scale public or internal cohort studies. However, these approaches do not address the challenges of identifying random CHIP mutations in plasma-only approaches. The remaining unfiltered CHIP variants bias the percentage change to 1 (invariant), resulting in inaccurate predictions of subsequent molecular responses. To filter out private CHIP variants (e.g., CHIPs that are not previously documented or are poorly documented in previous databases of known CHIP variants), mutational measurements between two time points can be used to cluster variants with similar percentage changes. As patients undergo treatment, progression or response may result in minor somatic mutations, while the CHIP variants remain stable. By clustering mutations into clones, random CHIP variants can be found in clones enriched with known CHIP lists or clones with stable percentage differences.

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

[0189] In one embodiment, genomic mutations / variants are clustered using a heuristic novel agglomerative hierarchical clustering. The heuristic quantifies the statistical dissimilarity between the mutations / variants and clusters via a custom dissimilarity metric. A tunable stopping rule is used to continue agglomerating until a minimum (or maximum depending on the metric) allowable difference threshold is met. In one embodiment, the custom dissimilarity metric is a modification of the Bhattacharya distance, and a numerical integration is performed with respect to the product (not square-rooted) of scaled likelihoods of the mutations / variants and / or clusters being considered to be merged in a given step of the clustering heuristic. The likelihood is scaled to integrate numerically to 1 rather than supporting the integration. For SNVs and indels, the likelihood is calculated with respect to a beta-binomial model approximation of the observed count data informing the MAF determination of the clustered variants. The variance of the beta-binomial model is set via a tunable parameter. For CNVs, the likelihood is calculated with respect to a Gaussian model approximation of the observed multiplicative change estimate of the mutation of interest, and the variability of the Gaussian model is also set via a tunable parameter. Mutation aggregation may occur in a novel manner, in some cases, by clustering a first set of mutations until a stopping rule is met, then introducing a second set of mutations, and further aggregation steps may be carried out according to the same dissimilarity metric and stopping rule, thereby performing clustering via a stepwise approach. In some situations, a third set of mutations may be introduced in a similar manner after the application of the clustering heuristic to the second set of mutations.

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

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 / mutants is

number

number

number

number

number

number

number

number

number

[0194] In step 303, the number of clones and the relevant percentage changes between time points can be reported along with confidence intervals. Clones with a percentage change between a first and second time point that is at or above a predetermined threshold can be identified. If multiple clones are identified, clones with a percentage change close to 1 and / or clones with specific known CHIP variants 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 those variants are classified as CHIP in at least one patient sample.

[0195] Figure 4 shows an example of the application of the CHIP filter 300. Figure 4 corresponds to an example of the agglutination procedure. In this example, three eligible mutations are identified. The leftmost panel of Figure 4 displays the scaled likelihood function of each mutation (y axis) on the support R (x axis). We 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 mutations in the left panel with the highest similarity are marked with an asterisk. The center panel displays the agglutination likelihood obtained from the merge of the first likelihood (line 403) and 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 does not change due to agglutination. The right panel displays the final clonality. Since the composition of the second likelihood (line 401) clone is 50% CHIP, the second likelihood (line 401) clone can be identified as a presumptive CHIP. This results in the final value of R being defined solely by the third likelihood (line 402) clone.

[0196] In an embodiment, Figure 5 shows a schematic flowchart illustrating exemplary method steps for identifying clonal hematopoietic variants in a subject with cancer, according to several embodiments. As shown, Method 500 includes determining a tumor burden change (R) relative to a tumor percentage change P(R) for each of several variants from sequence information created from one or more targeted nucleic acids associated with oncological types in samples obtained from the subject at first and second time points to generate a set of tumor burden changes (Step 501). Furthermore, 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 Figure 1, Method 100 may proceed to determining the 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 ​​for all somatic variants may be used. In an embodiment, fewer MAF values ​​than all somatic variants may be used. As described in steps 105 / 106, MAF may be determined for multiple somatic variants from sequence reads created from one or more targeted nucleic acids associated with oncologies in samples obtained from subjects at T0 (e.g., pre-treatment) and T1 (e.g., during treatment) to generate first and second sets of MAF for somatic variants in multiple somatic variants. The MR score may be expressed as a percentage or proportion. The MR score may be determined according to Method 600, as shown in Figure 6A. Method 600 may include, in step 601, determining the ratio of a first MAF to a second MAF for a somatic mutant in a plurality of somatic mutants in order to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios in the set of MAF ratios. In some embodiments, the standard deviations can be used as a criterion for reporting the MR score. For example, the standard deviation of the MR score based on the individual standard deviations of at least one mutant can be used to determine the 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 the MAF ratios may be determined for the subject using the following formula:

number

number

number

[0198] In one embodiment, a method is disclosed in which, in addition to or instead of a weighted average of MAF ratios as the MR score, mutants are clustered based on their MAF ratios, the total MAF ratio of the clusters is calculated, and then either a single selected cluster ratio or a weighted average of the cluster ratios is used as the MR score. Clustering may be performed by combining pairs of mutants having overlapping MAF ratio distributions or by other clustering methods. A single selected cluster may contain known cancer driver mutants or may not contain known clonal hematopoietic mutants. The cluster weights may also depend on the presence of known cancer driver mutants or the maximum VAF or number of mutants in the cluster.

[0199] As shown in Figure 6B, the MR score may be determined according to Method 610. Method 610 may include, in step 601, determining the weighted average of a first MAF and a weighted average of a second MAF for somatic mutants among a plurality of somatic mutants, as well as the corresponding standard deviations for the weighted MAF ratios. In some embodiments, the standard deviations can be used as a criterion for reporting the MR score. For example, the standard deviation of the MR score based on the individual standard deviations of at least one mutant can be used to determine the 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 612, the ratio of the weighted average MAFs may be determined for the subject. The confidence interval is 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 (In the formula, A and B are the weighted average MAF at time point 1 and time point 2, respectively.)

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

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

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

[0203] To provide a simple example illustrating the nature of the problem addressed by the MR scoring method presented herein, we consider a subject with one detected variant, having 0.3% MAF at baseline (T0) and 0.1% MAF during treatment (T1), as well as coverage of 3000 molecules at its variant location. Using existing methods, the molecular response score would be as follows:

number

[0204] To provide a simple example for illustrating the aspect of the problem addressed by the MR scoring method presented herein, consider a subject in which two variants (a and b) are detected, having MAFs of a=0.1% and b=8.0% at baseline (T0), and MAFs of a=0.3% and b=2.0% during treatment (T1). Using the existing method of averaging the ratios, the molecular response score is [Formula] . In the case of a cutoff for defining 50% "molecular responders" versus "molecular non-responders", this subject will be a "molecular non-responder". However, using the averaged ratio according to the method described herein, the molecular response score would be as follows [Formula] . Therefore, for this subject, the molecular response should be regarded as a "molecular responder".

[0205] To provide a simple example for illustrating the aspect of the problem addressed by the MR scoring method presented herein, consider a subject in which two variants (a and b) are detected, having MAFs of a=0.3% and b=0.0% at baseline (T0), and MAFs of a=0.0% and b=0.3% during treatment (T1). When only variants above 0.3% at baseline are evaluated using the existing method, the molecular response score is as follows: [Formula] In the case of a cutoff for defining 50% "molecular responders" versus "molecular non-responders", this subject will be a "molecular responder". However, including variants that arise during treatment, the molecular response score is as follows: [Formula] . Therefore, for this subject, the molecular response should be regarded as a "molecular non-responder".

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

[0207] In other exemplary embodiments, variance is incorporated into molecular response calculations by simulation or sampling from the variance distribution of at least one variant in order to compute molecular response variance. As further disclosed herein, some applications include weighting variants based on their importance in tumors or the likelihood of tumors for clonal hematopoiesis. Some embodiments include integrating multiple genomic data sources to estimate tumor proportion (rather than relying solely on variants (e.g., SNVs, indels, and fusions) VAFs), coverage (e.g., copy number), off-target coverage, and / or methylation, among other genomic data sources.

[0208] In some embodiments, the method 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 coverages, off-target coverages, epigenetic signatures, tumor mutation loads, and / or microsatellite instability scores. For data sources, a calculation of tumor proportions based on those data sources may be performed, and the calculated tumor proportions may be combined across the data sources (e.g., using a weighted average to incorporate the confidence of the data sources in the tumor proportions for that particular sample), and then the overall tumor proportion estimates 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 an epigenetic status or state indicated by one or more epigenetic loci within a given targeted genomic region. In some embodiments, the epigenetic situation or state includes the presence or absence of methylation, hydroxymethylation, acetylation, ubiquitination, phosphorylation, smoylation, ribosylation, citrullination, and / or histone post-translational modifications or other histone mutations.

[0209] Although this method is described in the context of Figure 1 and the first time T0 and second time T1, it should be understood that more than three time points are possible, for example, for longitudinal monitoring. As shown in Figure 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 from any subsequent time T nIn this process, the treated cfDNA may be obtained from one or more treated samples obtained from one or more subjects after treatment. Time T1 can be any amount of time after time T0, e.g., 1 to 24 hours, 1 to 180 days, 1 to 12 weeks, 1 to 25 weeks, 1 to 30 weeks, etc. Furthermore, method 100 may use times T0, T1, ..., T n This can be applied to any combination of the above. For example, samples may be taken at time T1 and time T2, and samples taken at both times are samples being treated. In another example, samples may be taken at time T1 and time T2, with the sample taken at time T1 representing the sample being treated and the sample taken at time T2 representing the untreated sample.

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

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

[0212] Figure 8 illustrates an exemplary practical application of molecular response scores for patient stratification. Patients with advanced cancer may have a baseline MAF determined at time T0 before treatment. After 4–10 weeks of treatment, patients with advanced cancer may have an intra-treatment MAF determined at time T1. The obtained molecular response score may indicate a decrease in the patient's ctDNA, in which case the patient should continue treatment with the primary investigational drug. The obtained molecular response score may indicate an increase in the patient's ctDNA, 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 increased, the patient should receive one or more additional therapies, changes in therapy, or changes in the dose of the primary investigational drug.

[0213] Figure 9 illustrates an exemplary practical application of molecular response scores for clinical trial enrichment. Patients with advanced cancer eligible for standard of care (SOC) treatment may have a baseline MAF determined at time T0 before SOC treatment. After 4–10 weeks of SOC treatment, patients with advanced cancer may have an intra-treatment MAF determined at time T1. The resulting molecular response score may indicate a decrease in the patient's ctDNA, in which case the patient should continue treatment with SOC. The resulting molecular response score may indicate an increase in the patient's ctDNA, in which case the patient may be determined to be eligible for treatment with the investigational drug.

[0214] Figure 10 shows a practical application of molecular response scores for prospective patient stratification and escalation for the MSKCC trial of osimertinib + / - chemotherapy in patients with EGFR-positive non-small cell lung cancer (NSCLC). Patients newly diagnosed with EGFR-positive NSCLC may have a baseline MAF determined at time T0 before using osimertinib. After one cycle of osimertinib, patients may have an intra-treatment MAF determined on day 1 of cycle 2 of osimertinib. The resulting molecular response score may indicate no EGFR driver detected, based solely on the EGFR driver; in this case, the patient should continue treatment with osimertinib alone. The resulting molecular response score may indicate an EGFR driver detected, based solely on the EGFR driver; in this case, the patient should continue treatment with osimertinib, carboplatin, and pemetrexed.

[0215] Embodiments of these methods are further illustrated in Figure 11. As shown, Method 1100 includes determining mutant allele frequencies (MAFs) of multiple variants from sequence information derived from one or more targeted nucleic acids associated with cancer types in samples obtained from a subject at a first (e.g., pre-treatment) and second (e.g., during treatment) time point to generate first and second sets of MAFs for the variants of the multiple variants (Step 1101). Method 1100 also includes calculating the ratios of the first and second MAFs for the variants in the multiple variants to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios in the set of MAF ratios (Step 1102). Furthermore, Method 1100 also includes calculating a weighted average of the MAF ratios (Step 1103) and confidence intervals to determine a molecular response score for a subject with cancer.

[0216] In some embodiments, Method 1100 includes comparing the molecular response score of a subject with cancer to a predetermined cutoff point to identify whether the subject is likely to be a responder to one or more therapies for cancer (e.g., immunotherapy) if the molecular response score is below the predetermined cutoff point, or whether the subject is likely to be a non-responder to one or more therapies for cancer if the molecular response score is above or at the predetermined cutoff point. In some embodiments, Method 1100 includes administering one or more therapies for cancer to the subject, taking the molecular response score into consideration. In some embodiments, Method 1100 includes discontinuing the administration of one or more therapies for cancer to the subject, taking the molecular response score into consideration. In some embodiments, Method 1100 includes using the molecular response score as a prognostic biomarker and / or predictive biomarker for the subject.

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

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

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

[0220] In some embodiments, the variant comprises one or more single nucleotide variants (SNVs), insertion / deletion mutations (indels), gene amplifications, and / or gene fusions. In some embodiments, the method 1100 comprises determining a molecular response score for a subject having cancer using one or more additional genomic data sources. In some embodiments, the additional genomic data sources comprise one or more of coverage, off-target coverage, epigenetic signature, and / or microsatellite instability score. In some embodiments, the epigenetic signature comprises cfNA fragment length, position and / or endpoint density distribution. In some embodiments, the epigenetic signature comprises an epigenetic situation or state indicated by one or more epigenetic loci within a given targeted genomic region. In some embodiments, the epigenetic situation 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 variations.

[0221] To further illustrate, Figure 12A is a flowchart schematically illustrating an exemplary method 1200. As shown, method 1200 includes determining mutant allele frequencies (MAFs) of multiple variants from sequence information created from targeted nucleic acids associated with one or more oncologies in samples obtained from the subject at first and second time points to generate first and second sets of MAFs for the variants of the multiple variants (step 1201). Method 1200 also includes calculating the ratios of the first and second MAFs for the variants in the multiple 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 the weighted mean and confidence interval of the MAF ratios to determine the molecular response score of 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 criterion for reporting the molecular response score. Furthermore, Method 1200 also includes administering one or more therapies to a subject based on at least a molecular response score (Step 1204). Exemplary therapies are further disclosed herein.

[0222] Figure 12B is a flowchart schematically illustrating exemplary method 1210. As shown, method 1210 includes generating first and second sets of MAFs for the variants of multiple variants by determining the mutant allele frequencies (MAFs) of multiple variants from sequence information created from targeted nucleic acids associated with one or more oncologies in samples obtained from subjects at first and second time points (step 1211). Method 1210 includes, in step 1212, determining representative values ​​obtained from the MAFs of somatic variants considered for a given time point (i.e., first and second time points). It is understood that the representative values ​​may be, but are not limited to, the mean, median, or mode. Method 1210 includes, in step 1213, determining the ratio of the representative values ​​at the first time point to the representative values ​​at the second time point. Method 1210 may include calculating the standard deviation of the central tendency ratio using the standard deviations of the MAFs considered. In some embodiments, the representative values ​​may be the mean or median. In some embodiments, the representative value may be the mean. In some embodiments, the representative value may be the median. In some embodiments, method 1210 includes, in step 1212, determining the mean MAF of the somatic mutants considered for each time point (i.e., a first time point and a 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, method 1210 includes, in step 1212, determining the median MAF of the somatic mutants considered for each time point (i.e., a first time point and a 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 a first time point to the median value obtained at a 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 criterion for reporting the molecular response score. Furthermore, method 1210 also includes administering one or more therapies to a subject based on at least the molecular response score (step 1214). Exemplary therapies are further disclosed herein.

[0223] Typically, methods for determining molecular response scores involve excluding CHIP variants. For example, molecular response is typically measured by the allele frequency of genomic changes (e.g., small variants between two time points) that represent tumor proportion changes. Given that cfDNA signaling is an aggregation 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 repeating CHIP genes or hotspots curated by various data sources. However, identifying random CHIP mutations using plasma-only approaches remains challenging. The remaining unfiltered CHIP variants typically bias the proportion change to 1 (invariant), and thus result in inaccurate molecular response predictions or scores. Therefore, in some embodiments, the methods disclosed herein utilize models that cluster clonal genomic mutations with distinct proportion changes by leveraging observations between two time points. To group mutations, these approaches typically utilize mutant allele counts and total counts for the mutants from two time points in time, constructing a probability density function of the tumor percentage change R as P(R).

[0224] As a further example, Figure 13 is a flowchart schematically illustrating exemplary method steps for identifying clonal hematopoietic variants in a subject with cancer, according to several embodiments. As shown, Method 1300 includes calculating a probability density function of the tumor percentage change P(R) of multiple variants from sequence information created from targeted nucleic acids associated with one or more oncologies in samples obtained from the subject at first and second time points (Step 1301). As further shown, Method 1300 also includes grouping one or more variants by P(R) into one or more clones (Step 1302), creating updated P(R) for the clones of the clones (Step 1303), and identifying one or more clones having a percentage change between first and second time points of a predetermined threshold or above (Step 1304).

[0225] In other embodiments, the disclosure provides methods 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 allele imbalance, meaning that a somatic variant caller in a given bioinformatics pipeline may arrive at different somatic classifications for the same variant in the same patient. Since the objective of molecular response determination is to track somatic variants over the course of treatment, any classification discrepancies should be resolved in order to appropriately exclude germline variants from consideration.

[0226] For example, Figure 14 is a flowchart illustrating exemplary method steps for identifying variants in a subject with cancer according to several embodiments. As shown, Method 1400 includes determining the mutant allele frequency (MAF) for a given variant from sequence information created from one or more targeted nucleic acids associated with oncological types in a sample obtained from the subject (Step 1401). Using the MAF determined for a given variant, Method 1400 can identify the given variant as germline or somatic variant. In some embodiments, Method 1400 can use the baseline MAF and subsequent intra-treatment MAF of a given variant to classify the given variant as germline or somatic variant, or to change the previous classification of the given variant. Method 1400 may also include identifying a given mutant as a germline mutant if the MAF of the given mutant increases the maximum MAF of a sample containing the maximum proportion of diploid genes (max frac_diploid) (i.e., the least allele imbalance) at at least one of two time points, and / or if the MAF of the given mutant is at least about 2 times, 3 times, 4 times, 5 times, 6 times, 7 times, 8 times, 9 times, or at least 10 times greater than other MAFs (e.g., maximum MAF in the sample) of 1 or more determined from a sample obtained from the subject or another patient sample. In some embodiments, a given mutant is classified as somatic if it does not increase the maximum MAF of the sample (e.g., compared to another MAF) at at least one of two time points in which max frac_diploid is somatic. In some embodiments, a given mutant is classified as germline if it increases the maximum MAF and the sample having max frac_diploid is germline. In some embodiments, Method 1400 includes classifying a given variant as somatic when it is determined that the variant is a harmful variant (e.g., a frameshift mutation or a nonsense mutation) within a tumor suppressor gene (TSG).In some embodiments, a given variant is classified as somatic if it is found in less than approximately 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 relevant discrepancies are not resolved by method 1400. In these embodiments, the variant is typically excluded from further consideration when determining a given molecular response score. In some embodiments, a variant is classified as a CHIP variant if those variants are classified as CHIP in at least one patient sample.

[0227] Figure 15 is a schematic flowchart illustrating Method 1500, which includes determining a first set of multiple sequence reads and a second set of multiple sequence reads relevant to the subject (step 1501). As further shown, Method 1500 also includes classifying multiple variants in the first set of multiple sequence reads and the second set of multiple sequence reads as somatic or germline (step 1502), determining the MAF ratio (step 1503), determining the weighted mean of the MAF ratio (step 1504), determining the confidence interval associated with the weighted mean of the MAF ratio (step 1505), and outputting the weighted mean of the MAF ratio and the confidence interval (step 1506). It is understood that the first set of multiple sequence reads may be determined before therapy, and the second set of multiple sequence reads may be determined after therapy. The classification of the multiple variants in the first and second multiple sequence reads as somatic or germline in step 1502 may be carried out as described herein, for example with respect to Figure 2. In one embodiment, at least two variants of the multiple 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 multiple variants that is classified as somatic and based on the first and second MAFs. The first MAF may be determined using the variants in the first multiple sequence reads at the time before treatment, and the second MAF may be determined using the same variants in the second multiple sequence reads at the time after treatment. The first and second MAFs can be determined for the same variants in both the first and second multiple sequence reads. It is further understood that the determination of the weighted average of the MAF ratios (step 1504) may be for the subject. Furthermore, it is understood that the determination of confidence intervals related to the weighted average of MAF ratios (step 1505) may also be based on the weighted average of MAF ratios. Finally, it is understood that the weighted average of MAF ratios and confidence intervals can be output as numerator response scores.

[0228] Figure 16 is a flowchart illustrating Method 1600, which includes determining a first set of multiple sequence reads and a second set of multiple sequence reads relevant to the subject (step 1601). As further shown, Method 1600 also includes classifying multiple variants in the first set of multiple sequence reads and the second set of multiple sequence reads as somatic or germline (step 1602), determining the weighted average of the first and second MAFs (step 1603), determining the ratio of the weighted average of the first and second MAFs (step 1604), determining the confidence interval (step 1605), and outputting the ratio of the weighted average of the first and second MAFs and the confidence interval (step 1606). It is understood that the first set of multiple sequence reads may be determined before therapy, and the second set of multiple sequence reads may be determined after therapy. The classification of the multiple variants in the first and second multiple sequence reads as somatic or germline in step 1602 may be carried out as described herein, for example with respect to Figure 2. In one embodiment, at least two variants of the multiple variants are classified as somatic. The first MAF may be determined using the variants in the first multiple sequence reads at the time before treatment, and the second MAF may be determined using the same variants in the second multiple sequence reads at the time after treatment. The first and second MAFs can be determined for the same variants in both the first and second multiple sequence reads. It is also understood that the determination of the weighted average of the first and second MAFs (step 1603) may be determined for at least one variant of the multiple variants that are classified as somatic and are 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 to the weighted average of the second MAF (step 1604) may be for the subject. Furthermore, it is understood that the determination of the confidence interval (step 1605) 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 numerator response score.

[0229] Figure 17 is a flowchart schematically illustrating Method 1700, which includes determining a first and second set of sequence reads relevant to the subject (Step 1701). As further shown, Method 1700 also includes classifying multiple variants within the first set of sequence reads as somatic or germline (Step 1702), classifying multiple variants within the second set of sequence reads as somatic or germline (Step 1703), reclassifying at least one variant of the multiple variants to resolve classification discrepancies between the first and second set of sequence reads (Step 1704), determining a first mutant allele ratio (MAF) (Step 1705), determining a second MAF (Step 1706), and determining a molecular response score (Step 1707). It is understood that the first set of sequence reads may be determined before therapy is administered, and the second set of sequence reads may be determined after therapy is administered. The classification of the multiple variants in the first and second multiple sequence reads as somatic or germline in step 1703 may be performed as described herein, for example with respect to Figure 2. The first MAF may be determined using the variants in the first multiple sequence reads at the time before treatment, and the second MAF may be determined using the same variants in the second multiple sequence reads at the time after treatment. The first and second MAFs can be determined for the same variants in both the first and second multiple sequence reads. In one embodiment, at least two of the multiple 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 multiple variants that is classified as somatic and is based on at least a portion of the first multiple sequence reads. It is further understood that the determination of the second MAF (step 1706) may be for at least one variant of a 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 the molecular response may be determined based on the first and second MAFs.

[0230] Figure 18 is a schematic flowchart illustrating Method 1800, which includes determining a first set of multiple sequence reads and a second set of multiple sequence reads relevant to the subject (Step 1801). As further shown, Method 1800 also includes classifying multiple variants in the first set of multiple sequence reads and the second set of multiple sequence reads as somatic or germline (Step 1802), determining at least one variant of the multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant (Step 1803), removing at least one CHIP variant (Step 1804), determining a first mutant allele ratio (MAF) (Step 1805), determining a second MAF (Step 1806), and determining a molecular response score (Step 1807). It is understood that the first set of multiple sequence reads may be determined before therapy, and the second set of multiple sequence reads may be determined after therapy. The classification of the multiple variants in the first and second multiple sequence reads as somatic or germline in step 1802 may be carried out as described herein, for example with respect to Figure 2. In one embodiment, at least two variants of the multiple variants are classified as somatic. The first MAF may be determined using the variants in the first multiple sequence reads at the time before treatment, and the second MAF may be determined using the same variants in the second multiple sequence reads at the time after treatment. The first and second MAFs can be determined for the same variants in both the first and second multiple sequence reads. It is also understood that the removal of at least one CHIP variant (step 1804) may be derived from multiple variants. It is further understood that the determination of the first MAF (step 1805) may be for at least one variant of the multiple variants, classified as somatic and based on at least a portion of the first multiple sequence reads. Furthermore, it is understood that the determination of the second MAF (step 1806) is classified as somatic and may be for at least one variant of multiple variants based on at least a portion of the second set 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] Figure 19 is a flowchart schematically illustrating Method 1900, which includes determining a first set of sequence reads and a second set of sequence reads relevant to the subject (Step 1901). As further shown, Method 1900 also includes classifying multiple variants within the first set of sequence reads as somatic or germline (Step 1902), classifying multiple variants within the second set of sequence reads as somatic or germline (Step 1903), reclassifying at least one variant of the multiple variants to resolve a classification mismatch between the first set of sequence reads and the second set of sequence reads (Step 1904), and classifying at least one variant of the multiple variants as undetermined potential clonal hematopoiesis (CHIP). The process includes determining a variant (step 1905), removing at least one CHIP variant (step 1906), determining the first mutant allele proportion (MAF) (step 1907), determining the second MAF (step 1908), determining the MAF ratio (step 1909), determining the weighted average of the MAF ratios (step 1910), determining the 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 therapy, and the second plurality of sequence reads may be determined after therapy. Classifying the plurality of variants within 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 done as described herein, for example with respect to Figure 2. In one embodiment, at least two of the multiple variants are classified as somatic. It is also understood that the removal of at least one CHIP variant (step 1906) may be derived from multiple variants. A first MAF may be determined using the variants in a first set of sequence reads at the time before treatment, and a second MAF may be determined using the same variants in a second set of sequence reads at the time after treatment. The first and second MAFs can be determined for the same variants in both the first and second set of sequence reads.A classification mismatch may be a variant that is classified as somatic in a first set of sequence reads and as germline in a second set of sequence reads. A classification mismatch may be a variant that is classified as germline in a first set of sequence reads and as somatic in a second set of sequence reads. It is further understood that the determination of the first MAF (step 1907) may be for at least one variant of the multiple variants, which has been classified or reclassified as somatic and is based on at least a portion of the first set of sequence reads. Furthermore, it is understood that the determination of the second MAF (1908) may be for at least one variant of the multiple variants, which has been classified or reclassified as somatic and is based on at least a portion of the second set of sequence reads. It is also understood that the determination of the MAF ratio (1909) may be for at least one variant of the multiple variants, which has been classified or reclassified as somatic and is based on the first mutant allele ratio and the second mutant allele ratio. It is further understood that the determination of the MAF ratio (step 1910) may be for the subject. Furthermore, it is understood that the determination of the confidence interval related to the weighted average of the MAF ratios (step 1911) 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 may be output as the numerator response score.

[0232] Figure 20 is a flowchart schematically illustrating Method 2000, which includes determining a first set of multiple sequence reads and a second set of multiple sequence reads relevant to the subject (Step 2001). As further shown, Method 2000 also includes classifying multiple variants within the first set of multiple sequence reads as somatic or germline (Step 2002), classifying multiple variants within the second set of multiple sequence reads as somatic or germline (Step 2003), reclassifying at least one variant of the multiple variants to resolve classification discrepancies between the first set of multiple sequence reads and the second set of multiple sequence reads (Step 2004), and determining at least one variant of the multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant (Step 2002). 05) includes removing at least one CHIP variant (step 2006), determining the first mutant allele proportion (MAF) (step 2007), determining the second MAF (step 2008), determining the weighted average of the first MAF and the weighted average of the second MAF (step 2009), determining the ratio of the weighted average of the first MAF to the weighted average of the second MAF (step 2010), determining the confidence interval (step 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 multiple sequence reads are determined before therapy and the second multiple sequence reads are determined after therapy. Classifying multiple variants within a first set of sequence reads in step 2002 and classifying a second set of sequence reads as somatic or germline in step 2003 may be carried out as described herein, for example with respect to Figure 2. In one embodiment, at least two of the multiple variants are classified as somatic. It is also understood that the removal of at least one CHIP variant (step 2006) may be derived from multiple variants. The first MAF may be determined using the variants in the first set of sequence reads at the time before treatment, and the second MAF may be determined using the same variants in the second set of sequence reads at the time after treatment.The first MAF and the second MAF can be determined for the same variant in both the first and second set of sequence reads. It is further understood that the determination of the first MAF (step 2007) may be for at least one variant of the set of variants that are classified or reclassified as somatic and are based on at least a portion of the first set of sequence reads. Furthermore, it is understood that the determination of the second MAF (step 2008) may be for at least one variant of the set of variants that are classified or reclassified as somatic and are based on at least a portion of the second set 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) may be for at least one variant of the set of variants that are classified as somatic and are 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 to the weighted average of the second MAF (step 2010) may be for the subject. Furthermore, it is 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 numerator response score. III. Cancer and Other Diseases

[0233] In certain embodiments, the methods and embodiments disclosed herein are used for long-term monitoring of patients with a given disease, disorder, or condition. The disclosed methods may be used to track a patient's response to one or more treatments over time. Typically, the disease under consideration is a type of cancer. Such cancers are not limited to the following: 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 lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), chronic myelomonocytic leukemia (CMML), liver cancer, liver cancer Examples include carcinoma, hepatoma, hepatocellular carcinoma, cholangiocarcinoma, hepatoblastoma, lung cancer, non-small cell lung cancer (NSCLC), mesothelioma, B-cell lymphoma, non-Hodgkin lymphoma, diffuse large B-cell lymphoma, mantle cell lymphoma, T-cell lymphoma, non-Hodgkin lymphoma, precursor T-lymphoblastic lymphoma / leukemia, peripheral T-cell lymphoma, multiple myeloma, nasopharyngeal cancer (NPC), neuroblastoma, oropharyngeal cancer, oral squamous cell carcinoma, osteosarcoma, ovarian cancer, pancreatic cancer, pancreatic ductal adenocarcinoma, pseudopapillary neoplasm, and 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-exclusive examples of other gene-based diseases, disorders, or conditions that may be evaluated as needed 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), cat-crying syndrome, Crohn's disease, cystic fibrosis, Delcam disease, Down syndrome, Duane syndrome, Duchenne muscular dystrophy, factor V-Leiden thrombocytosis, familial hypercholesterolemia, familial meningeal fever, fragile X syndrome, Gaucher disease, hemochromatosis, hemophilia, holoprosencephalopathy, Huntington's disease, Klinefelter syndrome, Marfan syndrome, myotonic dystrophy, neurofibromatosis, Noonan syndrome, osteogenesis imperfecta, Parkinson's disease, phenylketonuria, and Poland syndrome. Examples of anomalies include porphyria, progeria, retinitis pigmentosa, severe combined immunodeficiency (SCID), sickle cell disease, spinal muscular atrophy, Tay-Sachs disease, thalassemia, trimethylaminuria, Turner syndrome, palatocardiofacial syndrome, WAGR syndrome, and Wilson's disease. IV. Customized Therapies and Related Doses

[0235] In some embodiments, the methods disclosed herein relate to identifying and treating a patient having a given disease, disorder, or condition. Essentially, any cancer therapy (e.g., surgical therapy, radiotherapy, chemotherapy, etc.) is included as part of these methods. In certain embodiments, the therapy administered to the subject may include at least one chemotherapeutic agent. In some embodiments, the chemotherapeutic agents may include alkylating agents (e.g., chlorambucil, cyclophosphamide, cisplatin, and carboplatin), nitrosoureas (e.g., carmustine and lomustine), antimetabolites (e.g., fluorouracil, methotrexate, and fludarabine), plant alkaloids and natural products (e.g., vincristine, paclitaxel, and topotecan), antitumor antibiotics (e.g., bleomycin, doxorubicin, and mitoxantrone), hormonal agents (e.g., prednisone, dexamethasone, tamoxifen, and leuprolide), and biological response modifiers (e.g., Herceptin and Avastin, Erbitux, and Rituxan). In some embodiments, the chemotherapy administered to the target may include FOLFOX or FOLFIRI. In certain embodiments, the therapy administered to a subject may include at least one PARP inhibitor. In certain embodiments, the PARP inhibitor may include, among others, OLAPARIB, TALAZOPARIB, RUCAPARIB, and NIRAPARIB (trade name ZEJULA). Typically, the therapy includes at least one immunotherapy (or immunotherapy agent). Immunotherapy generally refers to a method of enhancing the immune response to a given type of cancer. In certain embodiments, immunotherapy refers to a method of enhancing the T-cell response to a tumor or cancer.

[0236] In some embodiments, immunotherapeutic agents or immunotherapeutic agents target immune checkpoint molecules. Certain tumors can evade the immune system by utilizing immune checkpoint pathways. Therefore, targeting immune checkpoints has emerged as an effective approach to counteract the tumor's ability to evade the immune system and activate anti-tumor immunity against specific cancers. Pardoll, Nature Reviews Cancer, 2012, 12:252-264.

[0237] In certain embodiments, an immune checkpoint molecule is an inhibitory molecule that reduces the signaling involved in the T cell response to an antigen. For example, CTLA4 is expressed on T cells and plays a role in downregulating T cell activation by binding to CD80 (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 T cell activity in peripheral tissues during inflammatory responses. Furthermore, ligands of PD-1 (PD-L1 or PD-L2) are generally upregulated on the surface of many different tumors, resulting in downregulation of the anti-tumor immune response in the tumor microenvironment. In certain embodiments, the inhibitory immune checkpoint molecule is CTLA4 or PD-1. In other embodiments, the inhibitory immune checkpoint molecule is a ligand of PD-1, such as PD-L1 or PD-L2. In other embodiments, the inhibitory immune checkpoint molecule is a ligand of CTLA4, such as CD80 or CD86. In other embodiments, the inhibitory immune checkpoint molecule is lymphocyte activation gene 3 (LAG3), killer cell immunoglobulin-like receptor (KIR), T cell membrane protein 3 (TIM3), galectin 9 (GAL9), or adenosine A2a receptor (A2aR).

[0238] Antagonists targeting these immune checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Therefore, in certain embodiments, the immunotherapy or immunotherapy agent is an antagonist of an inhibitory immune checkpoint molecule. In certain embodiments, the inhibitory immune checkpoint molecule is PD-1. In certain embodiments, the inhibitory immune checkpoint molecule is PD-L1. In certain embodiments, the antagonist of 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 immunotherapy agent is an antagonist (e.g., an antibody) to CD80, CD86, LAG3, KIR, TIM3, GAL9, or A2aR. In other embodiments, the antagonist is a soluble version of an inhibitory immune checkpoint molecule, e.g., 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, an immune checkpoint molecule is a costimulatory molecule that amplifies signals involved in the T cell response to an antigen. For example, CD28 is a costimulatory receptor expressed on T cells. When a T cell binds to an antigen via its T cell receptor, CD28 binds to CD80 (aka B7.1) or CD86 (aka B7.2) on the antigen-presenting cell to amplify T cell receptor signaling and promote T cell activation. Since 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 for a costimulatory molecule, including, for example, CD80, CD86, B7RP1, B7-H3, B7-H4, CD137L, OX40L, or CD70.

[0241] Agonists targeting these co-stimulatory checkpoint molecules can be used to enhance antigen-specific T cell responses to certain cancers. Therefore, in certain embodiments, the immunotherapy or immunotherapy agent is a co-stimulatory checkpoint molecule agonist. In certain embodiments, the co-stimulatory checkpoint molecule agonist is an agonist antibody, preferably a monoclonal antibody. In certain embodiments, the agonist antibody or monoclonal antibody is an anti-CD28 antibody. In other embodiments, the agonist antibody or monoclonal antibody is an anti-ICOS, 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] Treatment options for certain gene-based diseases, disorders, or conditions other than cancer are generally well known to those skilled in the art and would be apparent when considering the specific disease, disorder, or condition under consideration.

[0243] In certain embodiments, the customized therapies described herein are 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 therapies (e.g., immunotherapeutic agents) may also be administered by any method known in the art, including, for example, buccal, sublingual, rectal, vaginal, urethral, ​​topical, intraocular, intranasal, and / or intraaural, and administration may include tablets, capsules, granules, aqueous suspensions, gels, sprays, suppositories, ointments, etc. V. Systems and computer-readable media

[0244] This disclosure also provides a variety of systems and computer program products or machine-readable media. In some embodiments, for example, the methods described herein are performed or facilitated, at least in part, 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), etc. For illustrative purposes, Figure 21 provides a schematic diagram of an exemplary system suitable for carrying out at least a part of the methods disclosed in this application. As shown in the figure, system 2100 includes at least one controller or computer, e.g., server 2102 (e.g., search engine server), which includes a processor 2104 and memory, storage, or memory component 1506, and one or more other communication devices 2114 and 2116 (e.g., client-side computer terminals, telephones, tablets, laptops, other mobile devices, etc.) located away from the remote server 2102 and communicating with the remote server via an electronic communication network 2112 such as the Internet or other internetwork. Communication devices 2114 and 2116 typically include, for example, an electronic display (e.g., an Internet-enabled computer) that communicates with a server 2102 computer on a network 2112, and the electronic display includes a user interface (e.g., a graphical user interface (GUI), a web-based user interface, and / or similar) for displaying the results of performing the methods described herein. In certain embodiments, the communication network also includes the physical transfer of data from one location to another using, for example, a hard drive, a thumb drive, or other data storage mechanism.System 2100 also includes a program product 1508 stored in a computer or machine-readable medium, such as one or more different types of memory, such as memory 2106 of server 2102, which is readable by server 2102, to facilitate guided search applications or other executable files that can be run by one or more other communication devices, such as 2114 (figuratively shown as a desktop or personal computer) and 2116 (figuratively shown as a tablet computer). In some embodiments, System 2100 also optionally includes at least one database server, such as server 2110, associated with an online website storing data that can be searched, for example, directly or via search engine server 2102 (e.g., classifier scores, control sample or comparator result data, indexed customized therapies, etc.). System 2100 also optionally includes one or more other servers located remotely from server 2102, each of which may optionally be associated with one or more database servers 2110 located remotely from each of the other servers or locally to each of the other servers. Other servers can provide beneficial services to geographically dispersed users and enhance geographically distributed operations.

[0245] As will be understood by those skilled in the art, the memory 2106 of server 2102 includes, for example, volatile and / or non-volatile memory, including, in particular, RAM, ROM, and magnetic or optical disks, as appropriate. Although illustrated as a single server, the illustrated configuration of server 2102 is given only as an example, and it will be understood by those skilled in the art that other types of servers or computers configured according to various other methodologies or architectures may also be used. Server 2102 schematically shown in Figure 21 represents a server or server cluster or server farm, and is not limited to individual physical servers. The server site may be deployed as a server farm or server cluster managed by a server hosting provider. The number of servers, as well as their architecture and configuration, may be increased based on the usage, demand, and capacity requirements of system 2100. Also, as will be understood by those skilled in the art, other user communication devices 2114 and 2116 in these embodiments can 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, network 2112 may include the Internet, intranet, telecommunications network, extranet, or World Wide Web and / or part of local or other area networks of multiple computers / servers communicating with one or more other computers via a communication network.

[0246] As will be further understood by those skilled in the art, the exemplary program product or machine-readable medium 2108 is, as necessary, in the form of microcode, programs, cloud computing formats, routines, and / or symbolic languages ​​that provide one or more sets of ordered operations that control the functions of the hardware and direct its operation. According to the exemplary embodiment, the program product 2108 also does not need to reside entirely in volatile memory and can be loaded selectively as necessary in a variety of ways known and understood by those skilled in the art.

[0247] As will be further understood by those skilled in the art, the terms “computer-readable medium” or “machine-readable medium” refer to any medium involved in providing instructions to a processor for execution. Exemplarily, the terms “computer-readable medium” or “machine-readable medium” encompass distribution media, cloud computing formats, intermediate storage media, computer running memory, and any other media or devices capable of storing program products 2108 implementing functions or processes of various embodiments of this disclosure for, for example, computer reading. “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 disks or magnetic disks. Volatile media include dynamic memory, such as the main memory of a given system. Transmission media include coaxial cables, copper wires, and optical fibers, including wires with buses. Transmission media can also take the form of acoustic or light waves, such as those created during radio and infrared data communications, among other things. Examples of computer-readable media include floppy disks, flexible disks, hard disks, magnetic tapes, flash drives, or any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tapes, any other physical media having a pattern of holes, RAM, PROMs, and EPROMs, FLASH®-EPROMs, any other memory chips or cartridges, carriers, or any other media that can be read by a computer.

[0248] Program product 2108 is copied, if necessary, from a computer-readable medium to a hard disk or similar intermediate storage medium. When program product 2108 or any part thereof is executed, it is loaded, if necessary, from their distribution medium, their intermediate storage medium, etc., into the running memory of one or more computers and configured to operate according to the functions or methods of various embodiments. All such operations are well known, for example, to those skilled in the art of computer systems.

[0249] To further illustrate, in certain embodiments, the present application provides a system comprising one or more processors and one or more memory components communicating with the processors. The memory components typically include one or more instructions, when executed, that cause the processors to provide information (e.g., via communication devices 2114, 2116, etc.) to display, 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, when executed by the electronic processor 2104, includes non-temporary computer-executable instructions that perform at least the following: determine mutant allele frequencies (MAFs) for multiple variants from sequence information created from targeted nucleic acids associated with one or more cancer types in samples obtained from a subject at first and second time points to construct first and second sets of MAFs for at least one of the multiple variants; calculate the ratio of the first and second MAFs for at least one of the multiple variants to generate a set of MAF ratios and corresponding standard deviations for the MAF ratios in the set of MAF ratios; and calculate a weighted mean and confidence interval of the MAF ratios to determine a molecular response score for a subject with cancer. Embodiments of additional computer-readable media are described herein.

[0251] System 2100 also typically includes additional system components configured to implement various aspects of the methods described herein. In some of these embodiments, one or more of these additional system components are located remotely from the remote server 2102 via the electronic communications network 2112 and communicate with the remote server, while in other embodiments, one or more of these additional system components are located locally and communicate with the server 2102 (i.e., if 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 (directly or indirectly (e.g., via the electronic communication network 2112)). The sample preparation component 2118 is configured to prepare nucleic acids in a sample that are amplified and / or sequenced by a nucleic acid amplification component (e.g., a thermal cycler) and / or a nucleic acid sequencer (e.g., to prepare a library of nucleic acids). In these particular embodiments, the sample preparation component 2118 is configured to isolate nucleic acids from other components in the sample, to attach one or more adapters containing barcodes to the nucleic acids described herein, and to selectively enrich one or more regions from the genome or transcriptome prior to sequencing, and / or similarly.

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

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

[0255] To facilitate full or partial system automation, system 2100 also typically includes a material transfer component 2124 operably connected to controller 2102 (directly or indirectly, e.g., via electronic communication network 2112). 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 nucleic acid sequencer 2122, sample preparation component 2118, and nucleic acid amplification component 2120.

[0256] For further details on computer systems and networks, databases, and computer program products, see, 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 It is also provided in Systems, Addison Wesley, 6th Ed. (2010), Coronel, Database Systems: Design, Implementation, & Management, Engagement 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), and these are all incorporated by reference. [Examples]

[0257] VI. Examples A. Example 1: Comparison of molecular response calculations for predicting patient outcomes 1. Background Molecular response (MR), estimated as the change in circulating tumor (ctDNA) load between the initial treatment sample (typically 2–9 weeks after the start of treatment) and the pre-treatment baseline, has been shown in numerous retrospective studies to predict patient response and outcomes across solid tumors and therapy types. However, there is no consensus on the best method for assessing molecular response. Therefore, we aimed to evaluate several molecular response calculations and determine the optimal method for predicting outcomes in individual patients with advanced cancer. 2. Method

[0258] Aggregated results from over 4,000 patient sample pairs (separated 3–10 weeks apart), over 1,000 technical replications of patient samples, over 100 refined sample dilutions, and in silico simulations were analyzed using the cfDNA NGS assay clinical platform (Guardant Health, Inc., Redwood City, California, USA). Baseline and treatment-stage patient sample pairs were collected from patients with advanced cancer having over 12 tumor types, including lung, colon, and breast cancer. MR calculations included the percentage of variant alleles (VAF) for 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 evaluated. 3.Results

[0259] A comparison of methods for calculating the net change in ctDNA loadings for over 1500 sample pairs showed high correlation (ρ in the range of 0.93–0.98) and categorically agreed at a median (93%). Therefore, selecting the optimal method based on outcome prediction would require an extraordinarily large patient cohort. Analytical evaluation and in silico simulations can predict the behavior of each method. Simulations of changes in tumor proportion in actual pre-treatment samples found that RmVAF or RmaxVAF were more accurate than mVAFR, which can be distorted by a low VAF ratio. Nearly 25% of sample pairs had tumor drivers or resistance mutations other than maxVAF, suggesting that tumor dynamics are better captured by mVAF than maxVAF. Newly detected variants during treatment can be a significant signal of elevated ctDNA levels, affecting MR in approximately 2% of sample pairs.

[0260] Importantly, the MR accuracy of all methods decreases when maxVAF approaches or falls below the mutant LoD, due to both stochastic detection at low VAF and higher CV of the mutants. Therefore, the assay mutant LoD is a critical determinant of the proportion of patients who can undergo MR evaluation. Through technical replication, we were able to identify mutant criteria where a 50% change in tumor proportion significantly differs from the technical variability and define the limits of analytical reporting. 4. Conclusion

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

[0262] Molecular response (MR) is an assessment of the change in circulating tumor (ctDNA) load during the initial (typically 3–10 weeks) period of treatment compared to baseline before treatment. 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 responses have also been shown to predict clinical responses earlier than radiographic and / or RECIST responses. Multiple methods have been used to calculate molecular responses, and there is no consensus on which method is best.

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

[0265] Paired samples from over 1,500 patient plasma samples placed at intervals of 3 to 10 weeks were sequenced using the cfDNA NGS assay clinical platform (Guardant), which has a median coverage of approximately 4,600 unique molecules at a reading depth of 20,000 ×. The samples were processed using Health, Inc. (Redwood City, California, USA). Somatic and germline SNVs, small indels, and fusions were subsetted into a 74-cancer-related gene panel space to mimic the clinical application of molecular responses. Technical replicas of over 140 patient samples were processed with one of the panels 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 changes in ctDNA VAFs of SNVs, indels, and fusions.

[0267] Figure 22 shows the number of somatic variants detected per sample in a panel space of 74 cancer-related genes. The number of somatic SNVs, indels, and fusions was counted for molecular response calculations per sample for the top three cancer types in this study. The median mutant counts were 4, 5, and 3 for breast cancer, CRC, and NSCLC, respectively. ii. Resolving somatic classification by sample pairing improves tumor signaling.

[0268] Figure 23 shows examples of somatic classification discrepancies that can distort MR results. Rare somatic classification discrepancies (less than 0.8% of variants) can occur with high tumor proportions and allele imbalances. Unresolved ALK distorts MR scores for universally reduced VAF.

[0269] Table 4 shows an example of resolving discrepancies in somatic classification between patient samples, improving variant accuracy. Discrepancies in somatic classification between patient sample pairs were resolved using a variant-characteristic algorithm. Accuracy was evaluated against manual resolution by experts in the relevant field.

[0270] Table 4. [Table 4] iii. The variants are included in molecular response calculations based on detection and VAF accuracy.

[0271] Figure 24 shows an example of mutant precision determined by mutant molecular count (MMC = VAF * molecular coverage). (Figure 24A) Mutants have a range of molecular coverage depending on the sample input and panel design. The 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 for mutant precision as it determines the probability of mutant detection (Figure 24E) and VAF precision (Figure 24F). To make the signal from noise clearer, mutants with low MMC at both time points should be excluded from the molecular response. iv. Molecular responses are generally consistent across methods, but R(mVAF) is more robust among patients.

[0272] Figure 25 shows that tumor signals may outweigh a small number of variants when using the mean ratio, m(rVAF), or the maximum ratio, R(maxVAF). (Figure 25A) MR scores are categorized as increasing, decreasing, or within the limits of precision ("nearly 0% change"). Only 8% of patients change between increasing and decreasing in any way, showing a 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 some VAFs are low (red). R(maxVAF) can be distorted by a single maximum variant (purple) that deviates from the majority. 20% of sample pairs have tumor drivers or resistance mutations that are not maxVAF, suggesting that tumor dynamics are better captured by mVAF. (C) Excluding newly treated variants results in a lower MR evaluable rate and excludes signals from new variants. Patients with low signaling of v.ctDNA level changes are identified as being unable to assess molecular responses.

[0273] Figure 26 shows an example where the certainty of the molecular response score increases as the number of variants (Figure 26A), molecular coverage (Figure 26B), and maximum VAF (Figure 26C) increase.

[0274] If somatic variants are absent (approximately 7% of patients) or if somatic variants that meet the inclusion criteria are absent (16%), the sample pair cannot be evaluated for molecular response using VAF-based methods. Furthermore, the certainty of the molecular response score is theoretically calculated using a statistical model of VAF precision. Sample pairs exceeding the acceptable limit of uncertainty (black line) cannot be evaluated for MR (3%). This results in approximately 74% of sample pairs being evaluable for MR. vi. The range of molecular response scores in clinical patient samples reflects strong biological signals.

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

[0276] The technical replica provides a null molecule response distribution that peaks at 0% change (Figure 27B). 4. Conclusion

[0277] Each component of molecular response calculations is crucial for accurate assessment of MR, including germline and low-precision mutant filtering, overall formulation, and evaluable criteria. Comparisons of molecular response methods in large sets of patient samples and simulations support the ratio of mean VAFs, including newly detected mutations. C. Example 3

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

[0279] All patents, patent applications, websites, other publications or documents, accession numbers, etc., referenced herein are incorporated by reference in whole for all purposes to the same extent as each individual item is incorporated by reference specifically and individually. Where different versions of an array are associated with an accession number at different points in time, the version associated with the accession number on the effective filing date of this application is meant. The effective filing date means, where applicable, the earlier of the actual filing date or the filing date of the priority application referring to the registration number. Similarly, where different versions of a publication, website, etc., are published at different points in time, unless otherwise specified, the version last published on the effective filing date of this application is meant.

[0280] Those skilled in the art will recognize, or can confirm by routine experimentation, many equivalents to specific embodiments of the methods and compositions described herein. Such equivalents are intended to be covered by the following claims. The present invention provides, for example, the following items: (Item 1) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The process involves classifying the multiple variants in the first and second sequence reads as somatic cells or germline cells, For at least one of the plurality of mutants classified as somatic, the weighted average of the first and second mutant allele proportions (MAFs) is determined based on the first and second MAFs. With respect to the aforementioned subject, the ratio of the weighted average of the first MAF to the weighted average of the second MAF is determined, The confidence interval is determined based on the ratio of the weighted average of the first MAF and the weighted average of the second MAF. A method comprising 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. (Item 2) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The process involves classifying the multiple variants in the first and second sequence reads as somatic cells or germline cells, For at least one of the multiple mutants classified as somatic, the MAF ratio is determined based on the first mutant allele ratio (MAF) and the second MAF. The weighted average of the MAF ratio for the aforementioned subject is determined, Based on the weighted average of the MAF ratios, a confidence interval related to the weighted average of the MAF ratios is determined. A method comprising outputting the weighted average of the MAF ratio and the confidence interval as a molecular response score. (Item 3) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The first step involves classifying multiple variants within multiple sequence reads as either somatic or germline, Classifying the multiple variants within the second set of sequence reads as somatic or germline, In order to resolve the classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads, at least one variant among the plurality of variants is reclassified, For at least one of the plurality of variants classified or reclassified as somatic, the first mutant allele proportion (MAF) is determined based on at least a portion of the first plurality of sequence reads, For at least one of the plurality of variants classified or reclassified as somatic, a second MAF is determined based on at least a portion of the second plurality of sequence reads, A method comprising determining a molecular response score based on the first MAF and the second MAF. (Item 4) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The process involves classifying the multiple variants in the first and second sequence reads as somatic cells or germline cells, To determine at least one of the aforementioned multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant, Removing at least one CHIP variant from the aforementioned group of variants, For at least one of the plurality of mutants classified as somatic, the first mutant allele ratio (MAF) is determined based on at least a portion of the first plurality of sequence reads, For at least one of the plurality of mutants classified as somatic, a second MAF is determined based on at least a portion of the second plurality of sequence reads, A method comprising determining a molecular response score based on the first MAF and the second MAF. (Item 5) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The first step involves classifying multiple variants within multiple sequence reads as either somatic or germline, Classifying the multiple variants within the second set of sequence reads as somatic or germline, In order to resolve the classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads, at least one variant among the plurality of variants is reclassified, To determine at least one of the aforementioned multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant, Removing at least one CHIP variant from the aforementioned group of variants, For at least one of the plurality of mutants classified as somatic, the first mutant allele ratio (MAF) is determined based on at least a portion of the first plurality of sequence reads, For at least one of the plurality of mutants classified as somatic, a second MAF is determined based on at least a portion of the second plurality of sequence reads, For at least one of the multiple mutants classified as somatic cells, the first MAF and the second MAF-MAF ratio are determined, The weighted average of the MAF ratio for the aforementioned subject is determined, Based on the weighted average of the MAF ratios, a confidence interval related to the weighted average of the MAF ratios is determined. A method comprising outputting the weighted average of the MAF ratio and the confidence interval as a molecular response score. (Item 6) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined before the administration of the therapy, and the second plurality of sequence reads are determined after the administration of the therapy. The first step involves classifying multiple variants within multiple sequence reads as either somatic or germline, Classifying the multiple variants within the second set of sequence reads as somatic or germline, In order to resolve the classification mismatch between the first plurality of sequence reads and the second plurality of sequence reads, at least one variant among the plurality of variants is reclassified, To determine at least one of the aforementioned multiple variants as an undetermined potential clonal hematopoietic (CHIP) variant, Removing at least one CHIP variant from the aforementioned group of variants, For at least one of the plurality of mutants classified as somatic, the first mutant allele ratio (MAF) is determined based on at least a portion of the first plurality of sequence reads, For at least one of the plurality of mutants classified as somatic, a second MAF is determined based on at least a portion of the second plurality of sequence reads, For at least one mutant of several mutants classified as somatic, the weighted average of the first MAF and the weighted average of the second MAF are determined based on the first MAF and the second MAF. With respect to the aforementioned subject, the ratio of the weighted average of the first MAF to the weighted average of the second MAF is determined, The confidence interval is determined based on the ratio of the weighted average of the first MAF and the weighted average of the second MAF. A method comprising 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. (Item 7) It is a method, Determining a first plurality of sequence reads and a second plurality of sequence reads associated with the subject, wherein the first plurality of sequence reads are determined at a first time point before the administration of therapy, and the second plurality of sequence reads are determined at a second time point after the administration of therapy. The process involves classifying the multiple variants in the first and second sequence reads as either somatic or germline, For at least one of the plurality of mutants classified as somatic, a first representative value of the first MAF and a second representative value of the second MAF are determined based on the first MAF at the first time point and the second MAF at the second time point. Determining the ratio between the first representative value at the first time point and the second representative value at the second time point, A method comprising 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. (Item 8) The method according to item 7, wherein the aforementioned representative value is greater than or equal to one of the mean, median, or mode. (Item 9) A method for determining a molecular response score for a subject with cancer, using at least part computer assistance, (a) Using a computer, determine mutant allele frequencies (MAFs) for multiple variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the subject at first and second time points, and create first and second sets of MAFs for at least one of the multiple variants; (b) The computer calculates the ratio of the first and second MAFs to at least one of the plurality of mutants, and generates a set of MAF ratios and the corresponding standard deviation for each MAF ratio in the set of MAF ratios, (c) A method comprising determining the molecular response score of the subject having cancer by calculating a weighted mean and confidence interval of the MAF ratio using the computer. (Item 10) A method for treating the target cancer, (a) Determining mutant allele frequencies (MAFs) for multiple mutants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the subject at the first and second time points, and creating first and second sets of MAFs for at least one of the multiple mutants, (b) Calculate the ratio of the first and second MAFs for at least one mutant among the plurality of mutants to generate a set of MAF ratios and the corresponding standard deviation for each MAF ratio in the set of MAF ratios, (c) Calculate the weighted average and confidence interval of the MAF ratio to determine the molecular response score of the subject, (d) A method comprising treating the cancer in the subject by administering one or more therapies to the subject based on at least the molecular response score. (Item 11) A method for treating a target cancer, comprising administering to the target one or more therapies based on at least the molecular response score of the target, wherein the molecular response score is (a) Using a computer, determine mutant allele frequencies (MAFs) for multiple variants from sequence information created from one or more targeted nucleic acids associated with cancer types in samples obtained from the subject at first and second time points, and generate first and second sets of MAFs for at least one of the multiple variants. (b) The computer calculates the ratio of the first and second MAFs to at least one of the plurality of mutants, and generates a set of MAF ratios and the corresponding standard deviation for each MAF ratio in the set of MAF ratios, (c) A method for treating a target cancer, which is generated by the computer calculating the weighted mean and confidence interval of the MAF ratio to determine the molecular response score of the target. (Item 12) A method for identifying clonal hematopoietic variants in subjects with cancer, using at least partially computer-aided technology, (a) The computer determines the tumor burden change (R) for each of the multiple variants relative to the tumor percentage change P(R) from sequence information created from one or more cancer-related nucleic acids in the sample obtained from the subject at a first time point and a second time point, and generates a set of tumor burden changes, (b) A method comprising identifying the clonal hematopoietic variant in the subject having the cancer by using the computer to identify one or more resistance signatures corresponding to one or more clonal hematopoietic variants from the set of tumor burden changes. (Item 13) A method for identifying clonal hematopoietic variants in subjects with cancer, using a computer at least partially; (a) The computer calculates a probability density function for the change in tumor percentage P(R) for each of the multiple variants from sequence information created from one or more targeted nucleic acids associated with cancer types in the sample obtained from the subject at a first time point and a second time point, (b) The computer groups one or more variants of P(R) into one or more clones, (c) The computer creates an updated P(R) for each of the clones, (d) A method comprising using the computer to identify one or more clones having a predetermined threshold or a percentage change between the first time point and the second time point, thereby identifying the clonal hematopoietic variant in the subject having cancer. (Item 14) The method according to item 13, comprising determining the likelihood that a given pair of mutants exhibit the same proportion change, merging the most likely pair of mutants into a single clone, and updating P(R) for the single clone. (Item 15) A method for identifying germline variants in a subject with cancer, using a computer at least partially; (a) The computer determines the mutant allele frequency (MAF) for a given mutant from sequence information generated from one or more targeted nucleic acids associated with cancer types in a sample obtained from the subject, (b) A method comprising: (b) Identifying a given variant as a germline variant if the computer determines that the MAF of the given variant increases the maximum MAF of the sample and the sample contains the maximum proportion of diploid genes (max frac_diploid), and / or that the MAF of the given variant is at least about 2 times, 3 times, 4 times, 5 times, 6 times, 7 times, 8 times, 9 times, or more than 1 or more other MAF determined from the sample obtained from the subject, thereby identifying the germline variant in the subject having cancer. (Item 16) The method of any one of the preceding items, further comprising comparing the molecular response score of the subject having the cancer with a predetermined cutoff point, and identifying that if the molecular response score is below the predetermined cutoff point, the subject is likely to be a responder to one or more therapies for the cancer, or if the molecular response score is above or at the predetermined cutoff point, the subject is likely to be a non-responder to one or more therapies for the cancer. (Item 17) The method according to any one of the preceding items, wherein the aforementioned 1 or more therapies include 1 or more immunotherapies. (Item 18) The method according to any one of the preceding items, further comprising administering one or more therapies for the cancer to the subject, taking into account the molecular response score. (Item 19) The method of any one of the preceding items, further comprising discontinuing the administration of one or more therapies for the cancer to the subject, taking into consideration the molecular response score. (Item 20) The method according to any one of the preceding items, comprising using the molecular response score as a prognostic biomarker and / or predictive biomarker for the subject. (Item 21) The method according to any one of the preceding items, comprising using a molecular count to calculate the standard deviation for each MAF ratio in the set of MAF ratios. (Item 22) The method of any one of the preceding items, comprising propagating the variance through each MAF ratio in the set of MAF ratios. (Item 23) The method according to any one of the preceding items, further comprising excluding one or more germline and / or clonal hematopoietic variants when determining the mutant allele frequency (MAF) for the plurality of variants. (Item 24) The method of any one of the preceding items, wherein the plurality of variants comprises somatic nucleic acid variants. (Item 25) The method of any one of the preceding items, further comprising 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. (Item 26) The method of any one of the preceding items, 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. (Item 27) The method of any one of the preceding items, comprising generating the sequence information from nucleic acid molecules obtained from one or more tissues or cells in the sample. (Item 28) The method of any one of the preceding items, comprising generating the sequence information from cell-free nucleic acid (cfNA) in the sample obtained from the subject. (Item 29) The method of any one of the preceding items, wherein the cfNA comprises circulating tumor DNA (ctDNA). (Item 30) The method of any one of the preceding items, wherein the ratio comprises the second MAF over the first MAF for at least one variant of the plurality of variants. (Item 31) The following formula: [Equation] Sum[weight * ratio] / Sum[weight], (wherein the weight is 1 / range for a given variant of the plurality of variants 2 , the range is the difference between the first and second MAF values for the given variant of the plurality of variants, and the ratio is a given MAF ratio in the set of MAF ratios) The method of any one of the preceding items, comprising calculating the weighted average of the MAF ratios using (Item 32) The following formula:

number

Claims

[Claim 1] The invention described in the specification.

Images