DNA methylation and gene expression as determinants for genomic range cell free DNA fragmentation

Abstract

Analysis of cell free DNA (cfDNA) fragment ends in cancer patients provides a direct relationship between epigenetic alterations and cfDNA fragmentation for non-invasive disease detection and patient treatment. In certain embodiments, methods for cancer diagnosis and treatment include detecting methylation and expression differences in a patient that affect cell free (cfDNA) size and coverage. In certain aspects, include methods for determining circulating cfDNA fragmentation in a provided sample, including determining a genomic sequence to identify cfDNA endpoints; analyzing frequencies of cfDNA breaking points at a plurality of positions in the genome sequence; as a result, a circulating cfDNA fragmentation is determined.

Description

[0001] This application claims priority to U.S. Provisional Application No. 63 / 521,666, filed June 17, 2023, the contents of which are incorporated herein by reference in their entirety. Statement on Federally Funded Research

[0002] This invention was made with government support under grants granted by the National Institutes of Health (NIH) CA006973, CA062924, CA121113, and CA233259. The government holds certain rights to this invention. Technical Field

[0003] This disclosure relates to methods and compositions for evaluating and / or treating mammals (e.g., humans) suffering from cancer. In particular, the methods involve analyzing epigenetic changes and cfDNA fragmentation for non-invasive disease detection. Background Technology

[0004] Cellular free DNA (cfDNA) has become a focus of research in blood-based biomarkers for the early detection and monitoring of cancer. Typically, cellular DNA is packaged and condensed within chromosomes by being encased around a histone core. 1,2 During and after cell death, DNA is degraded by DNases, partly to prevent the release of unbound DNA, which can function as autoantigens. 3-6 The DNA fragments tightly wrapped around the histone core, commonly known as nucleosomes, appear to be protected from further degradation. 7 These fragments are typically characterized in cfDNA, which can be obtained through a simple blood draw. The characterization and origin of cfDNA fragmentation in blood remain poorly understood. Summary of the Invention

[0005] Methods for cancer diagnosis and treatment include detecting methylation and expression differences in patients that affect the size and coverage of cell-free DNA (cfDNA).

[0006] In some aspects, the present invention provides a method for determining the fragmentation of circulating cell-free DNA (cfDNA) in a sample, comprising detecting a genomic sequence to identify cfDNA end positions; analyzing the frequency of cfDNA breakpoints at multiple locations in the genomic sequence; thereby determining the fragmentation of circulating cell-free DNA (cfDNA).

[0007] In some aspects, a method is provided for determining the fragmentation of circulating cell-free DNA (cfDNA) in a sample, comprising: a) detecting the cfDNA sequence to identify the genomic location and end location of the cfDNA fragment; b) analyzing the frequency of cfDNA breakpoints at multiple locations in the genomic sequence; thereby determining the fragmentation of circulating cell-free DNA (cfDNA).

[0008] In some implementations of the above methods, whole-genome sequencing or obtaining whole-genome sequences from a database and summarizing the cfDNA sequences are used to detect the genome sequence.

[0009] In some implementations, the analysis of cfDNA breakpoint frequency includes calculating the ratio of the number of cfDNA fragments that start or end at a particular location to the number of fragments that start or end within 50 bp around that particular location.

[0010] In some embodiments of the above method, the analysis of cfDNA breakpoint frequency includes calculating the ratio of the number of cfDNA fragments starting at a specific location to the number of fragments that are start or end points within 50 bp around that specific location, or the ratio of the number of cfDNA fragments ending at that specific location to the number of fragments that are start or end points within 50 bp around that specific location.

[0011] In some embodiments, cfDNA fragments containing similar terminal position sequences contain similar motifs. In some embodiments, the motif includes a thymine (T) or adenine (A) preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by a guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence.

[0012] In some embodiments, the motif includes a thymine (T) or adenine (A) preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by a guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence, or wherein these sequences are reverse complementary sequences at the end of the cfDNA fragment.

[0013] In some embodiments, the frequency of the motif is increased in healthy subjects compared to subjects with cancer. For example, in these embodiments, the frequency of the motif in healthy subjects is increased by at least 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 12%, 14%, 16%, 18%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, or 200% compared to the frequency of the motif in subjects with cancer. In some embodiments, the frequency of the motif is decreased in healthy subjects compared to subjects with cancer. For example, in these embodiments, the frequency of the motif in healthy subjects is reduced by at least 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, 100, 150, or 200 percent relative to the frequency of the motif in subjects with cancer.

[0014] In some embodiments, the frequency of the A / T|CC motif is greater than that of the A / T|CG motif. In some embodiments, the A / T|CG motif is located near the histone H1 linker, or concentrated between 100 and 200 base pairs from the histone H1 linker. In some embodiments, the internal regions of the cfDNA fragment are rich in adenine (A) and thymine (T).

[0015] In some embodiments, the method further includes mapping cfDNA fragments to the genome and comparing the terminal sequences of the cfDNA fragments with methylated and unmethylated CpG sites of cfDNA from healthy subjects.

[0016] In some implementations, methylated CpG sites are enriched at the ends of the A / T|CG cfDNA fragment sequence.

[0017] In some implementations, quantitatively assessing the enrichment of cfDNA fragment ends at CpGs involves calculating, for each CpG, the fraction of cfDNA fragments that begin or end at a dinucleotide position within the CpG divided by the number of cfDNA fragments that begin or end within 50 bp around each CpG.

[0018] In some embodiments, the cfDNA fragments from the X chromosome of healthy subjects contain an increased number of cfDNA fragments ending in CG, and the location of CCG-terminated cfDNA fragments on the CpG islands of the X chromosome is reduced in female subjects compared to male subjects. In some embodiments, the cfDNA sequence coverage enriched with cfDNA fragment terminal sequences includes increased methylation on methylated CpG island regions compared to cfDNA fragment terminal sequences with reduced or lower occurrence rates.

[0019] In some implementations, gene expression at the transcription start site (TSS) is negatively correlated with cfDNA coverage at the TSS.

[0020] The subject, such as a mammal, particularly a human, may have previously received cancer treatment to treat cancer. The method may further include administering cancer treatment after determining circulating cell-free DNA (cfDNA) fragmentation; for example, the administered cancer treatment may be surgery, adjuvant chemotherapy, neoadjuvant chemotherapy, radiotherapy, hormone therapy, cytotoxic therapy, immunotherapy, adoptive T-cell therapy, targeted therapy, or a combination thereof. The method may also include monitoring for the presence of cancer in mammals after administering cancer treatment and / or after determining circulating cell-free DNA (cfDNA) fragmentation.

[0021] In another aspect, a method is provided for diagnosing cancer and treating a subject, the method suitably including obtaining a sample from the subject, detecting changes in the size of circulating cell-free DNA (cfDNA) fragments compared to normal and tumor-derived cfDNA controls, detecting CpG methylation at the ends of cfDNA fragments and assessing fragment end characterization at CG and CCG sites via low-coverage whole-genome cfDNA analysis; diagnosing the subject with cancer; and treating the subject with one or more of chemotherapy agents, radiation therapy, surgery, and combinations thereof. In some embodiments, an increase in cfDNA fragments ending in N|CCG compared to normal controls is a diagnosis of cancer. In some embodiments, a decrease in cfDNA fragments ending in N|CCG indicates a subject without cancer. In some embodiments, an increase in cfDNA fragments ending in CG compared to normal controls is a diagnosis of cancer. In some embodiments, the cfDNA fragments originate from genomic regions with increased methylation compared to normal genomic methylation controls. In some embodiments, the method further includes incorporating the fragment end position distribution at CG and CCG sites into a gradient boosting tree machine learning model.

[0022] In another aspect, a method for diagnosing and treating a subject diagnosed with cancer includes: detecting a genomic sequence to identify genomically widespread CpG methylation; analyzing the frequency of cfDNA breakpoints at multiple locations in the genomic sequence and determining circulating cell-free DNA (cfDNA) fragmentation, wherein enrichment of cfDNA fragment ends recurring at CpG sites is associated with higher genomically widespread methylation levels and smaller cfDNA fragments, which is a diagnosis of cancer; diagnosing the subject with cancer if hypomethylation and / or increased gene expression or a reduction in cfDNA fragment size is detected; and treating the subject with one or more of chemotherapy, radiation, surgery, or a combination thereof. In some embodiments, the genomic sequence is detected by whole-genome sequencing, or by obtaining a whole-genome sequence from a database and summarizing the cfDNA sequences.

[0023] In some implementations, the analysis of cfDNA breakpoint frequency includes calculating the ratio of the number of cfDNA fragments that start or end at a particular location to the number of fragments that start or end within 5 bp around that location.

[0024] In some embodiments, cfDNA fragments containing similar terminal position sequences have similar motifs. In some embodiments, the motif includes a thymine (T) or adenine (A) preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by a guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence.

[0025] In some embodiments, the frequency of the motif is increased in healthy subjects compared to subjects with cancer. In some embodiments, the frequency of the A / T|CC motif is greater than the frequency of the A / T|CG motif. In some embodiments, the A / T|CG motif is located near the histone H1 linker or concentrated between 100 and 200 base pairs from the histone H1 linker. In some embodiments, the internal regions of the cfDNA fragment are enriched with adenine and thymine. In some embodiments, the method further includes mapping the cfDNA fragment to the genome and comparing the terminal sequences of the cfDNA fragment with methylated and unmethylated CpG sites of cfDNA from healthy subjects. In some embodiments, methylated CpG sites are enriched at the ends of the A / T|CG cfDNA fragment sequence. In some embodiments, quantitative assessment of the enrichment of cfDNA fragment ends at CpGs includes calculating a fraction for each CpG of the number of cfDNA fragments starting or ending at a CpG dinucleotide position divided by the number of cfDNA fragments starting or ending within 50 bp around each CpG.

[0026] In some implementations, the cfDNA fragments from the X chromosome of healthy subjects contain an increased number of cfDNA fragments ending in CG, and the positions of CCG-ending cfDNA fragments on the CpG islands of the X chromosome are reduced in female subjects compared to male subjects.

[0027] In some embodiments, cfDNA sequence coverage enriched with cfDNA fragment terminal sequences includes increased methylation on methylated CpG island regions compared to cfDNA fragment terminal sequences with reduced or lower occurrence rates. In some embodiments, gene expression at the transcription start site (TSS) is negatively correlated with cfDNA coverage at the TSS. The subject, such as a mammal, particularly a human, may have previously received cancer treatment to treat cancer. The cancer treatment may be surgery, adjuvant chemotherapy, neoadjuvant chemotherapy, radiotherapy, hormone therapy, cytotoxic therapy, immunotherapy, adoptive T-cell therapy, targeted therapy, or a combination thereof. The method may also include monitoring the presence of cancer in the mammal after administration of cancer treatment.

[0028] In another aspect, a method for determining a mammalian cfDNA fragmentation profile includes processing cfDNA fragments obtained from a sample obtained from the mammal into a sequencing library, performing whole-genome sequencing (e.g., low-coverage whole-genome sequencing) on ​​the sequencing library to obtain sequencing fragments, mapping the sequencing fragments onto the genome to obtain windows of mapped sequences, and analyzing the windows of the mapped sequences to determine the length of the cfDNA fragments. The mapped sequences may include tens to thousands of windows. The windows of the mapped sequences may be non-overlapping windows. Each window of the mapped sequences may include up to 0.25 million, 0.5 million, 1 million, 2 million, 3 million, 4 million, 5 million, or 6 million or more base pairs. The cfDNA fragmentation profile can be determined within each window. The cfDNA fragmentation profile may include the median fragment size. The cfDNA fragmentation profile may include the fragment size distribution. The cfDNA fragmentation profile may include the ratio of small cfDNA fragments to large cfDNA fragments within the mapped sequence windows. The cfDNA fragmentation profile may cover the entire genome. The cfDNA fragmentation profile may cover subgenomic regions (e.g., regions within chromosomal portions). In some embodiments, the reference cfDNA fragmentation profile may be a cfDNA fragmentation profile of a healthy mammal. In some embodiments, the reference cfDNA fragmentation profile may be generated by determining the cfDNA fragmentation profile in a sample obtained from a healthy mammal. In some embodiments, the reference DNA fragmentation profile may be a reference nucleosome cfDNA fragmentation profile. In some embodiments, the cfDNA fragmentation profile may include a median fragment size, wherein the median fragment size of the cfDNA fragmentation profile is shorter than the median fragment size of the reference cfDNA fragmentation profile. In some embodiments, the cfDNA fragmentation profile may include a fragment size distribution, wherein the fragment size distribution of the cfDNA fragmentation profile differs from the fragment size distribution of the reference cfDNA fragmentation profile by at least 4, 6, 8, 10, 12, 14, 16, 18, or 20 nucleotides.

[0029] In some embodiments, the cfDNA fragmentation profile may include the ratio of small cfDNA fragments to large cfDNA fragments within a mapped sequence window, wherein the length of the small cfDNA fragments is up to 40, 60, 80, or 100 bp to 110, 120, 130, 140, or 150 bp, while the length of the large cfDNA fragments is 151 to 220 bp, and the correlation of the fragment ratios in the cfDNA fragmentation profile is lower than the correlation of the fragment ratios in a reference cfDNA fragmentation profile. In some embodiments, the cfDNA fragmentation profile may include the ratio of small cfDNA fragments to large cfDNA fragments within a mapped sequence window, wherein the length of the small cfDNA fragments is up to 100 to 150 bp, while the length of the large cfDNA fragments is 151 to 220 bp, and the correlation of the fragment ratios in the cfDNA fragmentation profile is lower than the correlation of the fragment ratios in a reference cfDNA fragmentation profile.

[0030] In some embodiments, the cfDNA fragmentation profile may include sequence coverage of small cfDNA fragments within a transgenomic window. In some embodiments, the cfDNA fragmentation profile may include sequence coverage of large cfDNA fragments within a transgenomic window. In some embodiments, the cfDNA fragmentation profile may include sequence coverage of both small and large cfDNA fragments within a transgenomic window. The comparison step may include comparing the cfDNA fragmentation profile with a reference cfDNA fragmentation profile across the entire genome. The comparison step may also include comparing the cfDNA fragmentation profile with a reference cfDNA fragmentation profile at subgenomic regions. The mammal may have previously been treated with cancer therapy to treat cancer.

[0031] In another aspect, a method is provided for diagnosing and treating a subject diagnosed with cancer, comprising: a) detecting a genomic sequence to identify genomically widespread CpG methylation; b) analyzing the frequency of cfDNA breakpoints at multiple locations in the genomic sequence and determining circulating cell-free DNA (cfDNA) fragmentation, wherein enrichment of cfDNA fragment ends recurring at CpG sites is associated with higher genomically widespread methylation levels and smaller cfDNA fragments, which is a diagnosis of cancer; c) diagnosing the subject with cancer if hypomethylation and / or increased gene expression or a reduction in cfDNA fragment size is detected; and d) treating the subject with one or more of chemotherapy, radiation, surgery, or a combination thereof.

[0032] In some embodiments, the genome sequence is detected by whole-genome sequencing or by obtaining whole-genome sequences from a database and summarizing the cfDNA sequences. In some embodiments, analysis of cfDNA breakpoint frequency includes calculating the ratio of the number of cfDNA fragments starting or ending at a specific location to the number of fragments starting or ending within 50 bp around that specific location. In some embodiments, cfDNA fragments containing similar end-position sequences contain similar motifs. In some embodiments, the motif includes thymine or adenine preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by a guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence. In some embodiments, the frequency of the motif is increased in healthy subjects compared to subjects with cancer. In some embodiments, the A / T|CG motif is located near the histone H1 linker or concentrated between 100 and 200 base pairs from the histone H1 linker. In some embodiments, the internal region of the cfDNA fragment is rich in adenine and thymine. In some embodiments, the method further includes mapping cfDNA fragments to the genome and comparing the terminal sequences of the cfDNA fragments with methylated and unmethylated CpG sites of cfDNA from healthy subjects. In some embodiments, methylated CpG sites are enriched at the ends of A / T|CG cfDNA fragment sequences. In some embodiments, quantitatively assessing the enrichment of cfDNA fragment ends at CpGs includes calculating, for each CpG, a fraction of the number of cfDNA fragments starting or ending at a CpG dinucleotide position divided by the number of cfDNA fragments starting or ending within 50 bp around each CpG. In some embodiments, cfDNA fragments from the X chromosome in healthy subjects contain an increased number of CG-terminated cfDNA fragments, and the location of CCG-terminated cfDNA fragments at X chromosome CpG islands is reduced in female subjects compared to male subjects. In some embodiments, the cfDNA sequence coverage enriched with cfDNA fragment terminal sequences includes increased methylation on methylated CpG island regions compared to cfDNA fragment terminal sequences with reduced or lower occurrence rates. In some implementations, gene expression at the transcription start site (TSS) is negatively correlated with the coverage of cfDNA at the TSS.

[0033] definition Unless otherwise defined, all terms used in this application (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms (such as those defined in common dictionaries) should be interpreted as consistent with their meaning in the relevant field context and will not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0034] As used herein, the singular forms “a / an,” “a type,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, the terms “comprising,” “including,” “having,” “containing,” “containing,” or variations thereof, as used in the specific embodiments and / or claims, are intended to be inclusive in a manner similar to the word “comprising.”

[0035] The terms “about” or “approximately” indicate within an acceptable margin of error for a particular value, as determined by one of ordinary skill in the art, which will depend in part on how the value is measured or determined. For example, according to convention in the art, “about” may mean within or above one standard deviation. Alternatively, “about” may mean a range of up to 20%, up to 10%, up to 5%, or up to 1% of a given value or range. Or, particularly for biological systems or methods, this term may mean within an order of magnitude of 5 times, or even within an order of magnitude of 2 times. When specific numerical values ​​are described in this application and claims, unless otherwise stated, the term “about” should be considered to indicate within an acceptable margin of error for that particular value.

[0036] The terms “aligned,” “aligned,” “mapped,” or “aligned” or “mapped” refer to one or more sequences that are identified as matching a known sequence from a reference genome in terms of the sequence of their nucleic acid molecules. This alignment can be performed manually or by computer algorithms, examples of which include the High Efficiency Local Alignment of Nucleotide Data (ELAND) computer program distributed as part of the Illumina Genomics Analysts workflow. The matching of sequence reads in an alignment can be 100% sequence matching or less than 100% (imperfect matching).

[0037] As used herein, the term “cancer” refers to (as is known in the art) a disease, condition, trait, genotype, or phenotype characterized by uncontrolled cell growth or replication; including liver cancer (including hepatocellular carcinoma (HCC)), lung cancer (including non-small cell lung cancer), gastric cancer, colorectal cancer, and leukemias such as acute myeloid leukemia (AML), chronic myeloid leukemia (CML), acute lymphoblastic leukemia (ALL), and chronic lymphocytic leukemia; AIDS-related cancers such as Kaposi's sarcoma; breast cancer; bone cancers such as osteosarcoma, chondrosarcoma, Ewing sarcoma, fibrosarcoma, giant cell tumor, ameloblastoma, and chordoma; and brain cancers such as meningioma, glioblastoma, and low-grade astrocytoma. Tumors, oligodendrogliomas, pituitary tumors, schwannomas, and metastatic brain cancers; head and neck cancers, including various lymphomas such as mantle cell lymphoma, non-Hodgkin's lymphoma, adenomas, squamous cell carcinoma, laryngeal cancer, gallbladder and bile duct cancer, retinal cancers such as retinoblastoma, esophageal cancer, gastric cancer, multiple myeloma, ovarian cancer, uterine cancer, thyroid cancer, testicular cancer, endometrial cancer, melanoma, bladder cancer, prostate cancer, pancreatic cancer, sarcoma, Wilms' tumor, cervical cancer, head and neck cancer, skin cancer, nasopharyngeal carcinoma, liposarcoma, epithelial carcinoma, renal cell carcinoma, gallbladder adenocarcinoma, parotid gland adenocarcinoma, endometrial sarcoma, multidrug-resistant cancers; and proliferative diseases and conditions, such as neovascularization associated with tumor angiogenesis.

[0038] The terms "cell-free nucleic acid," "cell-free DNA," or "cfDNA" refer to nucleic acid fragments that circulate within an individual's body (e.g., in the blood) and originate from one or more healthy cells and / or one or more cancer cells. In addition, cfDNA can also originate from other sources, such as viruses, fetuses, etc.

[0039] The term "cfDNA sequence coverage" refers to the average number of cfDNA molecules covering a specific location.

[0040] The term “circulating tumor DNA” or “ctDNA” refers to nucleic acid fragments from tumor cells or other types of cancer cells that may be released into a subject’s bloodstream due to biological processes such as apoptosis or necrosis of cells during death, or may be actively released by surviving tumor cells.

[0041] The terms “included,” “comprising,” or “included,” and variations thereof, used herein with respect to elements defined or described in articles, compositions, apparatuses, methods, processes, systems, etc., are inclusive or open-ended, allowing for other elements, and thus indicate that the defined or described articles, compositions, apparatuses, methods, processes, systems, etc., include those specified elements—or, where applicable, their equivalents—and may include other elements while still falling within the scope / definition of the defined articles, compositions, apparatuses, methods, processes, systems, etc.

[0042] "Diagnosis" or "being diagnosed" refers to recognizing the presence or nature of a pathological condition. Diagnostic methods differ in sensitivity and specificity. The "sensitivity" of a diagnostic test is the percentage of diseased individuals who test positive (the "true positive" percentage). Diseased individuals who are not detected are called "false negatives." Diseased individuals who test negative are called "true negatives." The "specificity" of a diagnostic test is 1 minus the false positive rate, where the "false positive" rate is defined as the proportion of diseased individuals who test positive. While a particular diagnostic method may not provide a definitive diagnosis of a disease, it is sufficient if the method provides a positive indication that aids in diagnosis.

[0043] The term "effective amount" as used in this article refers to the amount that provides therapeutic or preventative benefits.

[0044] The terms “fragmentation profile,” “position-dependent differences in fragmentation patterns,” and “position-dependent differences in fragment size and coverage on the genome” used herein are equivalent and can be used interchangeably. In some embodiments, determining the cfDNA fragmentation profile of a mammal can be used to identify mammals with cancer. For example, low-coverage whole-genome sequencing can be performed on cfDNA fragments obtained from a mammal (e.g., from samples obtained from a mammal), and the sequenced fragments can be mapped to the genome (e.g., in a non-overlapping window) and evaluated to determine the cfDNA fragmentation profile. As described herein, the cfDNA fragmentation profile of mammals with cancer is more heterogeneous (e.g., in terms of fragment length) than that of healthy mammals (e.g., mammals without cancer). Therefore, this disclosure also provides methods and materials for evaluating, monitoring, and / or treating mammals (e.g., humans) with or suspected of having cancer. In some embodiments, this document provides methods and materials for identifying mammals with cancer. For example, samples obtained from a mammal (e.g., blood samples) can be evaluated to determine the presence of cancer in the mammal and optionally the tissue of origin of the cancer, at least in part, based on the mammal’s cfDNA fragmentation profile. In some embodiments, methods and materials are provided for monitoring a mammal for cancer. For example, a sample obtained from a mammal (e.g., a blood sample) can be evaluated to determine the presence of cancer in the mammal, at least in part, based on the mammal's cfDNA fragmentation profile. In some embodiments, methods and materials are provided for identifying a mammal for cancer and administering one or more cancer treatments to the mammal. For example, a sample obtained from a mammal (e.g., a blood sample) can be evaluated to determine whether the mammal has cancer, at least in part, based on the mammal's cfDNA fragmentation profile, and one or more cancer treatments can be administered to the mammal.

[0045] The term "genomic nucleic acid" or "genomic DNA" refers to nucleic acid containing chromosomal DNA derived from one or more healthy cells (e.g., non-tumor cells) or tumor cells. In various implementations, genomic DNA can be extracted from cells derived from blood cell lineages, such as white blood cells (WBCs).

[0046] "Optional" or "optionally" means that the event or situation described below may or may not occur, and the description includes both the cases in which the event or situation occurs and the cases in which it does not occur.

[0047] In this specification and the appended claims, the term "or" is generally used in its meaning as including "and / or" unless otherwise expressly stated in the content.

[0048] "Parenteral" administration of immunogenic compositions includes techniques such as subcutaneous (sc), intravenous (iv), intramuscular (im), or intrasternal injection or infusion.

[0049] In this document, the terms "patient," "individual," or "subject" are used interchangeably to refer to a mammalian subject to treatment, preferably a human patient. In some embodiments, the methods of the present invention are used for laboratory animals, veterinary applications, and the development of animal models of disease, including but not limited to rodents (including mice, rats, and hamsters) and primates.

[0050] As used in this article, the term "reference genome" can refer to a digitized or previously determined database of nucleic acid sequences, a collection representing a species or individual. A reference genome can be a collection of nucleic acid sequences from multiple individuals, samples, or organisms and does not necessarily represent the nucleic acid composition of a single individual. Reference genomes can be used to map sequencing reads from a sample to chromosomal locations. For example, the National Center for Biotechnology Information (NCBI, ncbi.nlm.nih.gov) provides reference genomes for humans and many other organisms.

[0051] The term “read” or “read sequence” refers to any nucleotide sequence that includes a sequence read obtained from an individual, and / or a nucleotide sequence derived from an initial sequence read obtained from a sample obtained from an individual.

[0052] The terms “sample,” “patient sample,” “biological sample,” etc., encompass a wide range of sample types obtained from patients, individuals, or subjects and used for diagnostic, prognostic, and / or monitoring testing. Patient samples may be obtained from healthy subjects, patients with illness, or patients with lung cancer. In some embodiments, the “provided” sample may be obtained by the person (or machine) performing the test, or may be obtained by others and transferred to the person (or machine) performing the test. Furthermore, samples obtained from patients may be fractionated, with only a portion used for diagnosis. Additionally, the sample, or a portion thereof, may be stored under conditions for subsequent analysis. This definition specifically covers liquid samples of blood and other biological sources (including, but not limited to, peripheral blood, serum, plasma, cord blood, amniotic fluid, cerebrospinal fluid, urine, saliva, feces, and synovial fluid), and solid tissue samples, such as biopsy specimens or tissue cultures or cells and their progeny derived therefrom. In some embodiments, the sample includes cerebrospinal fluid. In a particular embodiment, the sample includes a blood sample. In another embodiment, the sample includes a plasma sample. In yet another embodiment, a serum sample is used. The definition of "sample" also includes samples that have been processed in any way after acquisition, such as by centrifugation, filtration, precipitation, dialysis, chromatography, reagent treatment, washing, or enrichment of specific cell populations. The term also covers clinical samples and includes cultured cells, cell supernatants, tissue samples, organs, etc. Samples may also include freshly frozen and / or formalin-fixed, paraffin-embedded tissue blocks, such as tissue blocks prepared from clinical or pathological biopsies for pathological analysis or immunohistochemical studies.

[0053] The term "sequence read" refers to a nucleotide sequence read derived from a sample obtained from an individual. Sequence reads can be obtained using various methods known in the art.

[0054] As defined herein, a "therapeuticly effective" amount of a compound or agent (i.e., an effective dose) means an amount sufficient to produce a therapeutically (e.g., clinically) desired effect. These compositions may be administered once or more daily to once or more weekly, including every other day. Those skilled in the art will understand that certain factors can influence the dosage and timing required for effective treatment of a subject, including but not limited to the severity of the disease or condition, prior treatment, the subject's overall health and / or age, and any other pre-existing conditions. Furthermore, treatment of a subject using a therapeutically effective amount of the compounds of the present invention may include a single treatment or a series of treatments.

[0055] The terms “treat,” “treating,” “treatment,” and similar terms used in this article refer to the reduction or relief of a disease and / or related symptoms. It should be understood that treating a disease or condition does not eliminate, but does not require, the complete eradication of, the disease, condition, or related symptoms.

[0056] Genes: All genes, gene names, and gene products disclosed herein are intended to correspond to homologous genes in any species to which the compositions and methods disclosed herein are applicable. It should be understood that when a gene or gene product of a particular species is disclosed, the disclosure is merely illustrative and should not be construed as limiting unless the context in which it appears explicitly indicates otherwise. Thus, for example, the genes or gene products disclosed herein are intended to cover homologous genes and / or orthologous genes and gene products from other species.

[0057] Scope: In this disclosure, various aspects of the invention can be described in the form of scope. It should be understood that the scope description is merely for convenience and brevity and should not be construed as a strict limitation on the scope of the invention. Therefore, the description of a scope should be considered as a specific disclosure of all possible sub-scopes within that scope and the respective numerical values. For example, a description of a scope such as "1 to 6" should be considered as a specific disclosure of sub-scopes such as "1 to 3", "1 to 4", "1 to 5", "2 to 4", "2 to 6", "3 to 6", etc., and the respective numerical values ​​within that scope, such as 1, 2, 2.7, 3, 4, 5, 5.3, and 6. Its applicability is regardless of the size of the scope.

[0058] Any composition or method provided herein may be used in combination with one or more of the other compositions and methods provided herein. Attached Figure Description

[0059] This patent or application document contains at least one color illustration. Upon request and payment of the necessary fees, the Office may provide a copy of this patent or application disclosure containing the color illustration.

[0060] Figure 1A-1C This indicates motif enrichment at the ends of cfDNA fragments and at the ends of recurring cfDNA fragments. A 3bp motif is located around the fragment, one base outside the fragment, and the first two bases of the fragment. Vertical lines indicate the fragment start position. Figure 1A: Frequency of 3bp DNA motifs at the ends of DNA fragments after ultrasonic shearing, frequency of 3bp DNA motifs at the ends of cfDNA fragments, and frequency of 3bp DNA motifs at the ends of cfDNA fragments with “preferred” ends (observed in at least 5% of recurring sequences). Relative frequencies were normalized by the occurrence rate of 3bp motifs observed in the human genome. Figure 1B: Increased motif preference at recurring cfDNA fragment end positions, as measured by the ratio of the increased number of cfDNA fragments starting or ending at a specific position divided by the total number of cfDNA fragments covering a + / - 50 base window around that position. These include enrichment of T|CC and A|CC, and T|CG and A|CG at these recurring cfDNA end positions. Figure 1C: PDB entry 7COW 44X-ray crystal structure. The TCG motif is red, the nucleosome protein appears as a gray surface, and the histone H1 linker is located at the top. Bases within 5 Å of the H1 linker appear as spherical. Figure 2A -2E indicates that DNA methylation is a determining factor in cfDNA fragmentation. a. From cfDNA break counts, the frequency of CpG observed at different locations within cfDNA fragments differs between methylated and unmethylated CpG sites. Unmethylated CpGs are more evenly distributed across cfDNA fragments, while methylated CpGs are enriched at the start sites of cfDNA fragments. Figure 2B: The preference for fragments starting with CpG increases with higher levels of methylation of that CpG. The preference for cfDNA fragments starting with CpG is measured by the ratio of cfDNA fragments starting with CpG to the total number of cfDNA fragments with a + / - 50 base window surrounding the first base of that CpG. Figure 2C: An inverse relationship is observed when the CpG is preceded by cytosine: when the CpG in the motif is unmethylated, there is a preference for cfDNA starting with CCG. Figure 2D: CpGs on the X chromosome are known to be differentially methylated in male and female individuals. Since CpG island methylation leads to X chromosome inactivation, these CpGs show differences in their preference for the terminal positions of cfDNA fragments. With increasing methylation, more fragments begin with CpG, while fewer fragments begin with CCG. Figure 2E: On autosomes, we did not observe differences in methylation between male and female individuals; Figures 3A-3G illustrate the effects of CpG methylation and gene expression on the coverage and size of cfDNA fragments. Figure 3A: Sequence coverage of CpG island regions, sorted by average methylation level of CpG islands. Figure 3B: Sequence coverage of transcription start site regions, sorted by average gene expression in myeloid cell lines. Figure 3C: Average cfDNA fragment size in CpG island regions, sorted by average methylation level of CpG islands. Figure 3D: Average cfDNA fragment size in transcription start site regions, sorted by average gene expression in myeloid cell lines. Figure 3E Correlation matrix (right) of significant gene set enrichment analysis of gene expression, CpG island DNA methylation, cfDNA fragment coverage at transcription start sites, and cfDNA coverage on CpG islands (all associated with the same genes). Representative gene set analysis of KEGG neuron receptor-ligand interactions (left panel), which showed decreased expression and increased methylation levels in WBCs, indicating increased cfDNA coverage at TSS and CpG sites. Figure 3FThe following cfDNA fragment size distributions (top figure) and cumulative characterizations (bottom figure) are shown: fragments with mutations (tumor origin) compared to wild-type fragments (mainly leukocytes), fragments from highly expressed transcription start site regions compared to fragments from low-expressed transcription start site regions, and fragments from methylated CpG island regions compared to fragments from unmethylated CpG island regions. Figure 3G Data from human allogeneic xenografts (IDH1 R132H mutant compared to IDH1 wild-type human glioblastoma cell line) showed that, in the top n differentially methylated or gene-expressed regions, human cfDNA coverage was increased in regions of increased methylation, while coverage was decreased in regions of increased gene expression. Figures 4A and 4B show a comparison of the terminal motifs of cfDNA fragments in differentially methylated regions in individuals with and without pancreatic cancer. Figure 4A: Aggregation ratio of cfDNA fragments starting or ending with specific motifs containing CpG, showing differential methylation between cfDNA in healthy individuals and pancreatic cancer tissue. The signal amplification was greatest in pancreatic cancer patients (n=34), while cfDNA from other cancer patients (colorectal cancer n=27, ovarian cancer n=28, lung cancer n=39, and breast cancer n=54) showed moderate signal intensity, between that of pancreatic cancer patients and non-cancer patients (n=244). Figure 4B: The predictive value of this signal for detecting pancreatic cancer is shown as a receiver operating characteristic curve compared with DELFI; however, the best prediction is realized in the integrated model combining the DELFI method with methylation-based signal. Figure 5 The relative frequencies of nucleotides in the 144bp cfDNA fragment were confirmed. A bar chart shows the relative frequencies of nucleotides in all 144bp cfDNA fragments from 543 healthy individuals. The observed frequencies indicate a 10.4-base periodicity in nucleotide occurrence, with A and T (more flexible) alternating with C and G (less flexible). Figure 6 It was confirmed that the fraction of cfDNA fragments initiating or terminating at a specific CpG motif depends on the CpG methylation state. Increased CpG methylation leads to more fragments initiating at that position; the opposite relationship was observed in fragments initiating at CCG. This effect is independent of the genomic environment in which the CpG resides, including CpG islands, CpG shore regions, CpG shelf regions, or offshore regions. Figure 7A , 7BThe table shows the number of cfDNA fragments originating from motifs containing CpG. The top two histograms (light blue and dark blue) show the normalized amount of fragments originating at each position around the motif, obtained from a low-coverage WGS dataset of cfDNA from 543 cancer-free individuals. In the light blue histogram, the motif contains an unmethylated CpG (β value < 0.3). In the dark blue histogram, the motif contains a methylated CpG (β value > 0.7). The third histogram (brown) shows the difference between the top two histograms, indicating that the number of fragments originating from a CpG increases when the CpG is methylated. We also observed a significant decrease in the amount of cfDNA fragments originating from an additional cytosine preceding the CpG when the subsequent CpG was methylated. Figure 7A The effects of methylation on cfDNA fragments originating around a 3bp motif were shown, while Figure 7B The effects of methylation on fragments originating around a 4bp motif were shown; Figure 8 This study confirmed that naturally occurring methylation differences on the X chromosome between males and females, due to X chromosome inactivation, affect the ends of cfDNA fragments. Similar to Figures 2D and 2E, which focus on CpG islands on the X chromosome and autosomes, fragments originating from CpG shore regions on the X chromosome differ between males and females due to increased methylation at these sites on the second X chromosome. An opposite pattern was observed in the CpG rack regions and offshore CpGs on the X chromosome. In these regions, females showed fewer CpG-leading fragments and more CCG-leading fragments, indicating lower methylation of these CpGs on the female X chromosome. The opposite relationship was observed when the CpG was preceded by cytosine. As a control, no difference was found between males and females in the amount of cfDNA fragments starting with either CpG or cytosine-leading CpGs on autosomes. Figures 9A and 9B confirm the relationship between CpG island methylation, gene expression, and cfDNA coverage. Figure 9A: The scatter plot shows that decreased CpG island methylation (lower β values) is associated with lower cfDNA coverage at CpG islands, while increased CpG island methylation (higher β values) is associated with higher cfDNA coverage at CpG islands. Figure 9B: Gene expression shows an inverse relationship with cfDNA coverage around transcription start sites, where increased gene expression is associated with decreased cfDNA coverage, and decreased gene expression is associated with increased cfDNA coverage. Figures 10A-10E confirm the effects of methylation and gene expression on cfDNA coverage and fragment size in large regions around CpG islands and transcription start sites. Figure 10A: Cumulative coverage and (Figure 10C) average cfDNA fragment size around the 1000 most methylated and 1000 least methylated CpG islands (concentrated around the central part of the CpG islands), showing reduced coverage and smaller fragments around less methylated CpG islands. Figure 10B: Cumulative coverage and (Figure 10D) average cfDNA fragment size around the 1000 most expressed and 1000 least expressed genes (centered on transcription start sites), showing reduced coverage and smaller fragments around the most expressed genes. This effect is most significant directly around transcription start sites. Figure 10E Average methylation around the 1000 most methylated CpG islands and the 1000 least methylated CpG islands (centered on the center of the CpG islands); Figure 11A , 11B The effects of gene expression and CpG island methylation on cfDNA coverage and the fraction of cfDNA fragments that start or end at CpG were confirmed. Figure 11A The cumulative cfDNA coverage around the transcription start sites of the 1000 most expressed and 1000 least expressed genes across all genes, and in subgroups of genes that have only unmethylated CpG islands (β value < 0.3) or methylated CpG islands (β value > 0.7). Figure 11B When genes are sorted by the amount of methylation of their associated CpG islands (left), the fraction of cfDNA fragments originating from CpG in these genes also shows a similar increase. When genes are sorted by expression (right), the effect on the fraction of cfDNA fragments originating from CpG in these genes is shown to be much more limited. The conclusion is that methylation has an independent effect on the amount of fragments originating from CpG compared to expression. Figure 12A , 12B The number of genes expressed and methylation sites in leukocytes and cancer tissues are shown respectively. Figure 12A Healthy white blood cells showed statistically significantly fewer gene expressions than cancerous tissue. Figure 12B Leukocytes showed a statistically significant number of methylated CpG islands compared to cancerous tissue. BRCA: Breast cancer, COAD: Colonic adenocarcinoma, LIHC: Hepatocellular carcinoma, LUAD: Lung adenocarcinoma, LUSC: Lung squamous cell carcinoma, OV: Ovarian cancer, PAAD: Pancreatic adenocarcinoma; Figure 13 (including) Figure 13A and 13B Multivariate analysis revealed the effects of CpG island methylation, RNA expression, and nucleosome localization on cfDNA coverage and size. Figure 13AForest plots illustrate the use of nucleosome localization scores (WPS), RNA expression levels (RNA), and methylation status indicators (if β-methylation is positive). > 0.5, Meth = 1; otherwise 0) are the regression coefficients of the coverage model for independent predictors. RNA and WPS are both standardized to have unit standard deviation. To model the non-linear relationship between coverage and WPS, a quadratic term (WPS²) is included. ) And cubic terms (WPS³) ) Furthermore, we added interaction terms to make the relationship between coverage and WPS dependent on RNA expression levels and methylation status. This resulted in a model containing only WPS and RNA conditions (WPS + WPS). 2 + WPS 3 +RNA + WPS xRNA + WPS 2 xRNA + WPS 3 The analysis of variance (ANOVA) comparing xRNA with the complete model shown in Figure (a) was statistically significant (F). 5,18377 = 22.0, p<0.0001), indicating that methylation helps explain the changes in coverage beyond its role in nucleosome localization and RNA expression itself. Similarly, the model containing only WPS and Meth conditions (WPS + WPS) 2 + WPS 3 + Meth + WPS x Meth + WPS 2 xMeth + WPS 3 Analysis of variance (ANOVA) comparing the x-Meth model with the complete model was statistically significant and helped explain the variation in coverage beyond the effects of WPS and methylation itself (F). 5,18377 = 1271.2, p<0.0001). The corresponding ANOVAs performed using WPS as a linear or quadratic model were also significant in all analyses (p<0.0001). Figure 13B The forest plot illustrates the model results with cfDNA fragment size (bp) as the dependent variable and WPS, RNA, and Meth as independent variables. Similarly, the quadratic and cubic terms of WPS allow for a non-linear relationship between fragment size and WPS. Analysis of variance comparing the complete model (Figure b) with models without methylation and models without RNA expression is statistically significant (Meth ANOVA: F0). 5,18539 p = 1.3e-10; RNA variance analysis: F 5,18539 (p<0.0001). Detailed Implementation

[0061] This disclosure provides the association between methylation, gene expression, and cfDNA in healthy subjects and cancer patients. It further provides analysis of motif-related features in healthy subjects and cancer patients, said motifs being associated with nucleosome localization and cfDNA fragment breakpoints. The disclosure also provides how epigenetic markers lead to specific fragmentation patterns of cfDNA and how these relate to both methylation and gene expression. In the following examples section, it is demonstrated that differentially methylated CpGs in a specific sequence background can be used to identify differences in cfDNA fragmentation between healthy subjects and cancer patients.

[0062] CpG methylation

[0063] Epigenetics refers to heritable alterations in gene expression not attributable to changes in DNA. The most well-defined epigenetic alterations are DNA methylation of cytosine by DNA methyltransferases. Cytosine associated with guanine is called CpG dinucleotides, and these are typically found in CpG-rich regions called CpG islands. A CpG island is defined as a sequence range with a length of at least 200 bp, a GC content greater than 50%, and an observed / expected (Obs / Exp) CpG ratio greater than 60%. The expected number of CpG dimers in a window is calculated by multiplying the number of 'C's in the window by the number of 'G's, and then dividing by the window length. A CpG island is defined as a region greater than 500 bp with a guanine-cytosine content greater than 55%. Up to 60% of CpG islands are located in the 5′ regulatory (promoter) region of a gene. However, CpG islands located outside the promoter region can also be found in both coding and non-coding regions of a gene, which may be targets for de novo methylation in cancer and aging. DNA methylation affects many different cellular processes, including apoptosis, cell cycle, DNA damage repair, proliferative factor responses, signal transduction, and tumor structure, all of which can promote cancer development and progression.

[0064] Methylated cytosine can be present in CpG islands, shore regions, shelf regions, offshore regions, and sites surrounding the transcription sites of coding genes (-200 to -1500 bp, 5′ untranslated region (UTR) and exon 1), as well as in the gene body and 3′ UTR and other / offshore regions, as derived from genome-wide association studies. Shore regions are considered to be 0–2 kb from CpG islands, shelf regions are 2–4 kb from CpG islands, and other / offshore regions are isolated CpG sites in the genome that do not have a specific region.

[0065] Therefore, in some implementations, the genome sequence is examined to identify genome-wide CpG methylation and the frequency of cfDNA breakpoints at multiple locations in the genome sequence to determine circulating cell-free DNA (cfDNA) fragmentation, wherein enrichment of cfDNA fragment ends recurring at CpG sites is associated with higher genome-wide methylation levels and smaller cfDNA fragments, which is a diagnostic criterion for cancer.

[0066] In some embodiments, the method further includes mapping cfDNA fragments to the genome and comparing the cfDNA fragment end sequences with methylated and unmethylated CpG sites in cfDNA from healthy subjects. Methylated CpGs are enriched at the ends of A / T|CG type cfDNA fragment sequences, and quantitatively assessing the enrichment of cfDNA fragment ends at CpGs involves, for each CpG, calculating the fraction of cfDNA fragments starting or ending at a CpG dinucleotide position divided by the number of cfDNA fragments with start or end positions within 50 bp around each CpG. In some embodiments, cfDNA fragments on the X chromosome from healthy subjects include more cfDNA fragments ending in CG, while in female subjects, the number of cfDNA fragment ends with CCGs located at CpG islands on the X chromosome is reduced compared to male subjects. In some embodiments, cfDNA sequence coverage of enriched cfDNA fragment end sequences includes increased methylation on methylated CpG island regions compared to cfDNA fragment end sequences with decreased or lower occurrence rates. In some embodiments, gene expression at transcription start sites (TSS) is negatively correlated with cfDNA coverage at TSS. cfDNA fragmentation profile

[0067] A cfDNA fragmentation profile may include one or more cfDNA fragmentation patterns. A cfDNA fragmentation pattern may include any suitable cfDNA fragmentation pattern. Examples of cfDNA fragmentation patterns include, but are not limited to: median fragment size, fragment size distribution, ratio of small to large cfDNA fragments, and cfDNA fragment coverage. In some embodiments, the cfDNA fragmentation pattern includes two or more (e.g., two, three, or four) of median fragment size, fragment size distribution, ratio of small to large cfDNA fragments, and cfDNA fragment coverage. In some embodiments, the cfDNA fragmentation profile may be a whole-genome cfDNA map (e.g., a whole-genome cfDNA map across a genomic window). In some embodiments, the cfDNA fragmentation profile may be a targeted region profile. A target region may be any suitable portion of the genome (e.g., a chromosomal region). Examples of chromosomal regions for which a cfDNA fragmentation profile can be determined as described herein include, but are not limited to, portions of chromosomes (e.g., portions of 2q, 4p, 5p, 6q, 7p, 8q, 9q, 10q, 11q, 12q, and / or 14q) and chromosome arms (e.g., chromosome arms of 8q, 13q, 11q, and / or 3p). In some embodiments, the cfDNA fragmentation profile may include two or more target region maps.

[0068] In some embodiments, cfDNA fragmentation profiling can be used to identify variations (e.g., alterations) in the length of cfDNA fragments. These alterations can be genome-wide changes or changes in one or more target regions / sites. Target regions can be regions containing one or more cancer-specific alterations. In some embodiments, cfDNA fragmentation profiling can be used to identify (e.g., simultaneously identify) approximately 10 to approximately 500 alterations (e.g., approximately 25 to approximately 500, approximately 50 to approximately 500, approximately 100 to approximately 500, approximately 200 to approximately 500, approximately 300 to approximately 500, approximately 10 to approximately 400, approximately 10 to approximately 300, approximately 10 to approximately 200, approximately 10 to approximately 100, approximately 10 to approximately 50, approximately 20 to approximately 400, approximately 30 to approximately 300, approximately 40 to approximately 200, approximately 50 to approximately 100, approximately 20 to approximately 100, approximately 25 to approximately 75, approximately 50 to approximately 250, approximately 100 to approximately 200 alterations).

[0069] cfDNA fragmentation profiles can be obtained using any suitable method. In some implementations, cfDNA from mammals (e.g., mammals with or suspected of having cancer) can be processed into a sequencing library and subjected to whole-genome sequencing (e.g., low-coverage whole-genome sequencing), mapped to the genome, and analyzed to determine the length of the cfDNA fragments. The mapped sequences can be analyzed within non-overlapping windows covering the genome. The window can be of any suitable size. For example, the length of the window can range from thousands to millions of bases. As a non-limiting example, the length of the window can be approximately 5 megabases (Mb). Any suitable number of windows can be mapped. For example, tens to thousands of windows can be mapped in the genome. For example, hundreds to thousands of windows can be mapped in the genome. The cfDNA fragmentation profile can be determined within each window.

[0070] In some implementations, the methods and materials disclosed herein may also include machine learning. For example, machine learning can be used to identify mutation frequencies, altered fragmentation profiles (e.g., using cfDNA fragment coverage, cfDNA fragment size, chromosome coverage, and mtDNA).

[0071] In some embodiments, determining the cfDNA fragmentation profile in a mammal can be used to identify a mammal with cancer. For example, cfDNA fragments obtained from a mammal (e.g., from a sample obtained from a mammal) can be subjected to low-coverage whole-genome sequencing, and the sequenced fragments can be mapped to the genome and evaluated to determine the cfDNA fragmentation profile. As described herein, the cfDNA fragmentation profile of a mammal with cancer is more heterogeneous (e.g., in terms of fragment length) than that of a healthy mammal (e.g., a mammal without cancer). Therefore, methods and materials are also provided for evaluating, monitoring, and / or treating mammals (e.g., humans) with or suspected of having cancer. In some embodiments, methods and materials are provided for identifying a mammal with cancer. For example, a sample obtained from a mammal (e.g., a blood sample) can be evaluated, in part based on the cfDNA fragmentation profile of the mammal, to determine the presence of cancer, and optionally the tissue origin of the cancer. In some embodiments, methods and materials are provided for monitoring a mammal with cancer. For example, a sample obtained from a mammal (e.g., a blood sample) can be evaluated, in part based on the cfDNA fragmentation profile of the mammal, to determine the presence of cancer in the mammal. In some embodiments, methods and materials are provided for identifying a mammal with cancer and administering one or more cancer treatments to the mammal to treat the mammal. For example, a sample obtained from a mammal (e.g., a blood sample) can be evaluated to determine whether the mammal has cancer, in part based on the mammal's cfDNA fragmentation profile, and one or more cancer treatments can be administered to the mammal.

[0072] In some embodiments, cfDNA fragmentation profiling can be used to detect tumor-derived DNA. For example, tumor-derived DNA can be detected by comparing the cfDNA fragmentation profiling of a mammal with cancer or suspected cancer with a reference cfDNA fragmentation profiling (e.g., the cfDNA fragmentation profiling of a healthy mammal and / or the nucleosome DNA fragmentation profiling from healthy cells of said mammal with cancer or suspected cancer). In some embodiments, the reference cfDNA fragmentation profiling is a previously generated profile from a healthy mammal. For example, the methods of this disclosure can be used to determine a reference cfDNA fragmentation profiling in a healthy mammal, and this reference cfDNA fragmentation profiling can be stored (e.g., in a computer or other electronic storage medium) for future comparison with the cfDNA fragmentation profiling of a mammal with cancer or suspected cancer. In some embodiments, the reference cfDNA fragmentation profiling (e.g., a stored cfDNA fragmentation profiling) is determined on a genome-wide scale. In some embodiments, the reference cfDNA fragmentation profiling of a healthy mammal (e.g., a stored cfDNA fragmentation profiling) is determined on a subgenomic region scale.

[0073] In some implementations, cfDNA fragmentation profiling can be used to identify mammals (e.g., humans) with cancer (e.g., liver cancer, colorectal cancer, lung cancer, breast cancer, stomach cancer, pancreatic cancer, cholangiocarcinoma, and / or ovarian cancer).

[0074] cfDNA fragmentation profiles can include cfDNA fragment size patterns. cfDNA fragments can be any suitable size. For example, the length of a cfDNA fragment can range from approximately 50 base pairs (bp) to approximately 400 base pairs.

[0075] cfDNA fragmentation profiles can include cfDNA fragment size distributions. As described herein, the cfDNA size distribution in cancerous mammals is more variable than that in healthy mammals. In some embodiments, the size distribution can be located within a target region. Healthy mammals (e.g., mammals without cancer) may have a target region cfDNA fragment size distribution of approximately 1 or less than approximately 1. In some embodiments, the target region cfDNA fragment size distribution in cancerous mammals may be longer than that in healthy mammals (e.g., 10, 15, 20, 25, 30, 35, 40, 45, 50 or more base pairs long, or any number of base pairs between these values). In some embodiments, the target region cfDNA fragment size distribution in cancerous mammals may be shorter than that in healthy mammals (e.g., 10, 15, 20, 25, 30, 35, 40, 45, 50 or more base pairs short, or any number of base pairs between these values). In some embodiments, the size distribution can be a genome-wide size distribution. Healthy mammals (e.g., mammals without cancer) may have a very similar distribution of short and long cfDNA fragments across their entire genome. In some embodiments, mammals with cancer may have one or more alterations (e.g., increases or decreases) in the size of cfDNA fragments across their entire genome. These alterations may occur in any appropriate chromosomal region of the genome. For example, the alteration may occur on a portion of a chromosome. Examples of chromosomal portions that may contain one or more alterations in the size of cfDNA fragments include, but are not limited to, portions of 2q, 4p, 5p, 6q, 7p, 8q, 9q, 10q, 11q, 12q, and 14q. For example, the alteration may span a chromosome arm (e.g., an entire chromosome arm).

[0076] cfDNA fragmentation profiles can include the ratio of small cfDNA fragments to large cfDNA fragments, and the correlation between the fragment ratio and a reference fragment ratio. As used herein, regarding the ratio of small cfDNA fragments to large cfDNA fragments, small cfDNA fragments can be approximately 100 bp to approximately 150 bp in length. As used herein, regarding the ratio of small cfDNA fragments to large cfDNA fragments, large cfDNA fragments can be approximately 151 bp to 220 bp in length. The correlation of cfDNA fragment ratios in cancerous mammals (e.g., the correlation of cfDNA fragment ratios to reference DNA fragment ratios (e.g., the correlation of cfDNA fragment ratios to DNA fragment ratios from one or more healthy mammals) may be lower than in healthy mammals (e.g., 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, or more lower). The correlation of cfDNA fragment ratios in healthy mammals (e.g., non-cancer mammals) (e.g., the correlation of cfDNA fragment ratios to reference DNA fragment ratios (e.g., the correlation of cfDNA fragment ratios to DNA fragment ratios from one or more healthy mammals) is approximately 1 (e.g., approximately 0.96). In some implementations, the correlation of fragment ratios in cancerous mammals (e.g., the correlation of cfDNA fragment ratios with reference DNA fragment ratios, such as those from one or more healthy mammals) is on average lower than the correlation of fragment ratios in healthy mammals (e.g., the correlation of cfDNA fragment ratios with reference DNA fragment ratios, such as those from one or more healthy mammals).

[0077] The cfDNA fragmentation profile can include coverage of all fragments. Coverage of all fragments can include coverage windows (e.g., non-overlapping windows). In some embodiments, coverage of all fragments can include windows of small fragments (e.g., fragments from about 100 bp to about 150 bp in length). In some embodiments, coverage of all fragments can include windows of large fragments (e.g., fragments from about 151 bp to about 220 bp in length).

[0078] In some implementations, cfDNA fragmentation profiling can be used to identify the molecular origin of a patient's cfDNA and to identify genomic and chromatin features associated with fragmentation changes.

[0079] In some embodiments, cfDNA fragmentation profiling can be used to identify the tissue origin of cancer (e.g., liver cancer, colorectal cancer, lung cancer, breast cancer, gastric cancer, pancreatic cancer, bile duct cancer, or ovarian cancer). For example, cfDNA fragmentation profiling can be used to identify localized cancers. When the cfDNA fragmentation profiling includes a target region map, one or more alterations described herein can be used to identify the tissue origin of cancer. In some embodiments, one or more alterations to a chromosomal region can be used to identify the tissue origin of cancer.

[0080] cfDNA fragmentation profiles can be obtained by any suitable method. In some embodiments, cfDNA from mammals (e.g., mammals with or suspected of having cancer) can be processed into sequencing libraries and subjected to whole-genome sequencing (e.g., low-coverage whole-genome sequencing), mapped to the genome, and analyzed to determine the length of the cfDNA fragments. The mapped sequences can be analyzed within non-overlapping windows covering the entire genome. The window can be of any suitable size. For example, the length of the window can range from thousands to millions of bases. As a non-limiting example, a window can be approximately 5 megabases (Mb) long. Any suitable number of windows can be mapped. For example, tens to thousands of windows can be mapped in the genome. For example, hundreds to thousands of windows can be mapped in the genome. The cfDNA fragmentation profile can be determined within each window.

[0081] In some implementations, the methods and materials described herein may also include machine learning. For example, machine learning can be used to identify altered fragmentation profiles (e.g., using cfDNA fragment coverage, cfDNA fragment size, chromosome coverage, and mtDNA). Various machine learning algorithms can be used to analyze fragmentation profiles. For example, stochastic gradient boosting models (GBM; see, for example, Friedman et al., 2001 Ann Stat 29:1189-1232; and Friedman et al., 2002 Comput Stat Data An 38:367-378) can be used to differentiate between healthy and cancer patients using fragmentation profiles. The GC-corrected total fragment and short fragment coverage for all 504 bins can be centered and normalized for each sample to have a mean of 0 and a unit standard deviation. Other features include Z-scores for each of the 39 autosomal arms and mitochondrial characterization (the proportion of log10 transformations mapped to mitochondrial reads). To estimate the prediction error of this method, 10-fold cross-validation can be used, as described in other literature (see, for example, Efron et al., 1997 J Am StatAssoc 92, 548-560). Feature selection is performed only on the training data in each cross-validation run, removing highly correlated (correlation > 0.9) or bins with near-zero variance. Stochastic gradient boosting machine learning can be implemented using the R package gbm. To average the prediction error originating from randomized patients to the fold, the 10-fold cross-validation procedure can be repeated.

[0082] In some implementations, the machine learning model can be a neural network (NN). In some implementations, the neural network can be a convolutional neural network (CNN), a recurrent neural network (RNN), or a deep learning neural network (DNN). In some implementations, the machine learning model is a random forest, logistic regression, or unsupervised clustering model. In other implementations of the neural network model, the model can also be a deep neural network (DNN) or a high-order neural network (HONN) with multiple local and fully connected hidden layers. For DNNs, a restricted Boltzmann machine (RBM) can be used to pre-train the neural nodes of the input and connection layers. For HONNs, a mean-covariance RBM can be used to pre-train the neural nodes of the input and connection layers.

[0083] In some embodiments, a computer system for obtaining database file access and executing one or more software programs may include a server, data storage devices, a network, and a user interface device. The server may also be a hypervisor-based system running one or more guest partitions hosting operating systems with modules containing server configuration information. In a further embodiment, the system may include a storage controller or storage server configured to manage data communication between the data storage devices and the server or other network-connected components. Alternatively, the storage controller may be connected to a network.

[0084] In some embodiments, "user interface device" is broadly referred to and intended to encompass appropriate processor-based devices, such as desktop computers, laptop computers, personal digital assistants (PDAs) or tablets, smartphones, or other mobile communication devices with network access. In further embodiments, the user interface device may access the Internet or other wide area networks (WANs) or local area networks (LANs) to access web applications or web services hosted by a server and to provide an interface for the user to input or receive information. The network may facilitate data communication between the server and the user interface device. The network may include any type of communication network, including but not limited to, direct PC-to-PC connections, local area networks (LANs), wide area networks (WANs), modem-to-modem connections, the Internet, combinations of the above, or any other communication network now known or later developed in the field of networking that allows communication between two or more computers.

[0085] In some embodiments, the computer system includes a central processing unit (“CPU”) connected to a system bus. The CPU may be a general-purpose CPU or a microprocessor, graphics processing unit (“GPU”), and / or microcontroller. The CPU can execute various logical instructions according to this embodiment. The computer system may also include random access memory (RAM), which may be synchronous RAM (SRAM), dynamic RAM (DRAM), synchronous dynamic RAM (SDRAM), or the like. The computer system may utilize RAM to store various data structures used by software applications. The computer system may also include read-only memory (ROM), which may be PROM, EPROM, EEPROM, optical storage, or the like. The ROM may store configuration information used to boot the computer system. RAM and ROM store user and system data, and both RAM and ROM are randomly accessible.

[0086] The computer system may also include I / O adapters, communication adapters, user interface adapters, and display adapters. In some embodiments, the I / O adapter and / or user interface adapter enable a user to interact with the computer system. In further embodiments, the display adapter can display a graphical user interface (GUI) associated with software or web-based applications on a display device (e.g., a monitor or touchscreen).

[0087] I / O adapters can connect one or more storage devices (such as hard drives, solid-state storage devices, flash drives, optical disc drives, floppy disk drives, and tape drives) to a computer system. Data storage may be coupled to a separate server of the computer system via a network connected to the I / O adapter. Communication adapters may be suitable for connecting the computer system to a network, which may be one or more of a LAN, WAN, and / or the Internet. User interface adapters couple user input devices (such as keyboards, pointing devices, and / or touchscreens) to the computer system. Display adapters may be driven by the CPU to control the display on a display device.

[0088] Examples of computer systems as computing devices are provided, which can be adapted to perform the functions of server and / or user interface devices. For example, various implementations and / or steps of the cancer detection model disclosed herein can be implemented using any suitable processor-based device, including but not limited to personal data assistants (PDAs), tablets, smartphones, computer game consoles, and multiprocessor servers. Furthermore, various implementations of the cancer detection methods disclosed herein can be implemented on application-specific integrated circuits (ASICs), very large-scale integrated circuit (VLSI) circuits, or other circuits. In fact, those skilled in the art can utilize any suitable structure capable of performing the logical operations according to the described implementations. For example, the computer system can be virtualized for access by multiple users and / or applications.

[0089] If the various methods, steps, and parameter calculations described herein are implemented in firmware and / or software, the aforementioned functions can be stored as one or more instructions or codes on a computer-readable medium. Examples include non-transitory computer-readable media encoding data structures and computer-readable media encoding computer programs. Computer-readable media include physical computer storage media. Storage media can be any available medium that is accessible to a computer. By way of example, and not limitation, such computer-readable media can include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other medium, such as cloud-based media, which can be used to store the required program code in the form of instructions or data structures and is accessible to a computer. Disks and optical discs include compact optical discs (CDs), laser discs, optical discs, digital versatile optical discs (DVDs), floppy disks, and Blu-ray discs. Typically, disks copy data magnetically, while optical discs copy data optically. Combinations of the above media should also be included within the scope of computer-readable media.

[0090] In addition to being stored on a computer-readable medium, instructions and / or data may also be provided as signals on a transmission medium included in a communication device. For example, a communication device may include a transceiver having signals indicating instructions and data. The instructions and data are configured to cause one or more processors to perform the functions embodied herein. Treatment

[0091] The method described herein includes identifying cancer in mammals. The method includes extracting cell-free cellular DNA (cfDNA) from a biological sample of a subject; generating a genomic library from the extracted cfDNA; sequencing individual cfDNA molecules to obtain fragmentation profiles; analyzing motif-related features in both healthy individuals and cancer patients, the motifs being associated with nucleosome localization and cfDNA fragmentation points; and analyzing epigenetic markers that cause specific fragmentation patterns in cfDNA and how these are related to both methylation and gene expression. Using this information, differentially methylated CpGs within a specific sequence background can be used to identify differences in cfDNA fragmentation between healthy individuals and cancer patients.

[0092] In another aspect, a method for diagnosing cancer and treating a subject includes: obtaining a sample from the subject, detecting changes in the size of cell-free DNA (cfDNA) fragments compared to normal and tumor-derived cfDNA controls, detecting CpG methylation at the ends of cfDNA fragments, and assessing fragment end characterization at CG and CCG sites through low-coverage whole-genome cfDNA analysis; diagnosing the subject with cancer; and treating the subject. In some embodiments, an increase in cfDNA fragments ending in N|CCG compared to normal controls is a diagnosis of cancer. In some embodiments, a decrease in cfDNA fragments ending in N|CCG compared to normal controls indicates that the subject does not have cancer. In some embodiments, an increase in cfDNA fragments ending in CG compared to normal controls is a diagnosis of cancer. In some embodiments, the cfDNA fragments originate from genomic regions with increased methylation compared to normal genomic methylation controls. In some embodiments, the method further includes integrating the distribution of fragment end positions at CG and CCG sites into a gradient boosting tree machine learning model.

[0093] In another aspect, methods for diagnosing and treating a subject diagnosed with cancer include: detecting genomic sequences to identify genome-wide CpG methylation; analyzing the frequency of cfDNA breakpoints at multiple locations in the genomic sequence and determining fragmentation of cell-free DNA (cfDNA), wherein recurring cfDNA fragments enriched at CpG sites are associated with higher genome-wide methylation levels and smaller cfDNA fragments, which is a diagnosis of cancer; diagnosing the subject with cancer if hypomethylation and / or increased gene expression or a reduction in cfDNA fragment size is detected; and treating the subject. In some embodiments, the genomic sequence is detected by whole-genome sequencing or by obtaining a whole-genome sequence from a database and compiling cfDNA sequences. In some embodiments, the analysis of cfDNA breakpoint frequencies includes calculating the ratio of the number of cfDNA fragments starting or ending at a specific location to the number of cfDNA fragments starting or ending within 50 bp around that specific location. In some embodiments, cfDNA fragments containing similar end-position sequences include similar motifs. In some embodiments, these motifs include a thymine or adenine preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by a guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence. In some embodiments, these motifs are more frequent in healthy subjects than in cancer patients. In some embodiments, the frequency of the A / T|CC motif is greater than that of the A / T|CG motif. In some embodiments, the A / T|CG motif is located near the histone H1 linker, or concentrated between 100 and 200 base pairs from the histone H1 linker. In some embodiments, the internal regions of the cfDNA fragment are enriched with adenine and thymine. In some embodiments, the method further includes mapping the cfDNA fragment to the genome and comparing the terminal sequence of the cfDNA fragment with methylated and unmethylated CpG sites in healthy subjects. In some embodiments, methylated CpG sites are enriched at the ends of A / T|CG type cfDNA fragment sequences.

[0094] In some embodiments, quantitative assessment of cfDNA fragment enrichment at CpG ends includes calculating a fraction for each CpG of cfDNA fragments starting or ending at a CpG dinucleotide position, divided by the number of cfDNA fragments starting or ending at that position within 50 bp around each CpG. In some embodiments, cfDNA fragments from the X chromosome in healthy subjects include increased cfDNA fragments ending at CG, and in female subjects, cfDNA fragments ending at CCG at X chromosome CpG island positions are reduced compared to male subjects. In some embodiments, cfDNA sequence coverage of enriched cfDNA fragment end sequences includes increased methylation at methylated CpG island regions, compared to cfDNA fragment end sequences with reduced or lower occurrence rates. The mammal may have previously been treated for cancer. The method may also include monitoring the mammal for the presence of cancer after administration of cancer treatment.

[0095] In some implementations, the subject is diagnosed with cancer, such as early-stage cancer. In other implementations, the type and stage of cancer are identified, and the subject is treated with one or more cancer therapies.

[0096] In some embodiments, the methods and materials described herein are used to assess, monitor, and / or treat mammals (e.g., humans) that have or are suspected of having cancer. In some embodiments, methods and materials are provided for identifying mammals with cancer.

[0097] In some embodiments, methods are provided for identifying mammals with cancer and administering one or more treatments to said mammals to treat them. In some embodiments, the mammal may be monitored (or optionally subjected to increased monitoring) and / or further diagnostic testing during or after cancer treatment (such as any cancer treatment described herein). In some embodiments, monitoring may include assessing mammals with or suspected of having cancer by, for example, evaluating samples obtained from the mammal (e.g., blood samples) to determine motif-related characteristics associated with nucleosome localization and cfDNA fragment breakpoints in both healthy subjects and cancer patients; analyzing epigenetic markers that elicit specific fragmentation patterns of cfDNA and how these relate to both methylation and gene expression. Using this information, differentially methylated CpGs in a specific sequence background are utilized to identify cfDNA fragmentation differences between healthy subjects and cancer patients, to identify treatment responses, and / or to identify mammals with cancer (e.g., residual cancer).

[0098] Any suitable mammal can be evaluated, monitored, and / or treated as described herein. For example, a mammal can be a mammal with liver cancer. A mammal can be a mammal suspected of having liver cancer. Mammals that can be evaluated, monitored, and / or treated as described herein include, but are not limited to, humans, primates such as monkeys, dogs, cats, horses, cattle, pigs, sheep, mice, and rats.

[0099] Any suitable mammalian sample can be evaluated as described herein (e.g., to assess DNA fragmentation patterns). In some embodiments, the sample may include DNA (e.g., genomic DNA). In some embodiments, the sample may include cfDNA (e.g., circulating tumor DNA (ctDNA)). In some embodiments, the sample may be a liquid sample (e.g., liquid biopsy). Examples of samples that may contain DNA and / or peptides include, but are not limited to, blood (e.g., whole blood, serum, or plasma), amniotic fluid, tissue, urine, cerebrospinal fluid, saliva, sputum, bronchoalveolar lavage fluid, bile, lymph, cyst fluid, feces, ascites, cervical smears, breast milk, and exhaled breath condensate.

[0100] Samples from mammals to be evaluated as described herein may contain any appropriate amount of cfDNA. In some embodiments, the sample may include a small amount of DNA. For example, cfDNA fragmentation profiling can be performed from samples containing more fragments than other cfDNA analysis methods (e.g., in Phallen). et al , 2017 Sci Transl Med 9; Cohen et al 2018 Science 359:926; Newman et al , 2014 Nat Med 20:548; and Newman et al , 2016 Nat Biotechnol Those described in 34:547 typically require less DNA.

[0101] In some embodiments, the sample may be processed (e.g., to isolate and / or purify DNA and / or peptides). For example, DNA isolation and / or purification may include cell lysis (e.g., using detergents and / or surfactants), protein removal (e.g., using proteases), and / or RNA removal (e.g., using RNases). As another example, peptide isolation and / or purification may include cell lysis (e.g., using detergents and / or surfactants), DNA removal (e.g., using DNases), and / or RNA removal (e.g., using RNases).

[0102] Cancer can be at any stage. In some implementations, cancer can be early-stage cancer. In some implementations, cancer can be asymptomatic cancer. In some implementations, cancer can be residual disease and / or recurrence (e.g., after surgical resection and / or cancer treatment). Cancer can be of any type. Types of cancer that can be evaluated, monitored, and / or treated include, but are not limited to, colorectal cancer, lung cancer, breast cancer, stomach cancer, pancreatic cancer, biliary tract cancer, and ovarian cancer.

[0103] In treating mammals with or suspected of having liver cancer as described herein, the mammal may be administered one or more cancer treatments. The cancer treatment can be any suitable cancer therapy. One or more cancer treatments may be administered to the mammal at any appropriate frequency (e.g., once or multiple times over a period ranging from days to weeks). Examples of cancer treatments include, but are not limited to, adjuvant chemotherapy, neoadjuvant chemotherapy, radiation therapy, hormone therapy, cytotoxic therapy, immunotherapy, adoptive T-cell therapy (e.g., chimeric antigen receptors and / or T cells with wild-type or modified T-cell receptors), targeted therapy such as administration of kinase inhibitors (e.g., kinase inhibitors targeting specific gene damage, such as translocations or mutations) (e.g., kinase inhibitors, antibodies, bispecific antibodies), signal transduction inhibitors, bispecific antibodies or antibody fragments (e.g., BiTEs), monoclonal antibodies, immune checkpoint inhibitors, surgery (e.g., surgical resection), or any combination of the foregoing. In some embodiments, the cancer treatment may reduce the severity of the cancer, alleviate the symptoms of the cancer, and / or reduce the number of cancer cells in the mammal.

[0104] In some implementations, cancer treatment may include immune checkpoint inhibitors. Non-limiting examples of immune checkpoint inhibitors include nivolumab (Opdivo), pembrolizumab (Keytruda), atezolizumab (Tecentriq), avelumab (Bavencio), durvalumab (Imfinzi), ipilimumab (Yervoy), etc.

[0105] Cancer treatment generally includes a variety of combination therapies, combining both chemotherapy and radiation-based treatments. Combination chemotherapy includes, for example, cisplatin (CDDP), carboplatin, procarbazine, nitrogen mustard, cyclophosphamide, camptothecin, ifosfamide, melphalan, chlorambucil, busulfan, nitrosourea, actinomycin D, daunorubicin, doxorubicin, bleomycin, procainoxine, mitomycin, etoposide (VP16), tamoxifen, raloxifene, estrogen receptor binders, paclitaxel, gemcitabine, vinorelbine, farnesyltransferase inhibitors, antiplatinum, 5-fluorouracil, vincristine, vinblastine, methotrexate, temozolomide (in aqueous solution of DTIC), or any analogues or derivatives thereof. The combination of chemotherapy and biological therapy is called biochemotherapy. Chemotherapy can also be administered in low, continuous doses, a process known as metronome chemotherapy.

[0106] Further combination chemotherapy includes, for example, alkylating agents such as thiotepa and cyclophosphamide; alkyl sulfonates such as busulfan, indomethacin, and piperazofan; aziridines such as phenyldopa, carboquinone, metodopa, and eulidopa; ethyleneimine and methylmelamine derivatives, including hexamethylmelamine, triethylenemelamine, triethylenephosphamide, triethylenethiophosphamide, and trimethylolamine; acetyltransferases (especially bratasine and bratasineone); camptothecin (including the synthetic analog topotecan); lichenin; carlistatin; CC-1065 (including its synthetic analogs adolaxine, carzecillin, and bizecillin); nodules (especially nodulesine 1 and nodulesine 8); salicylate; and pyruvicin (including the synthetic analog KW-2189). And CB1-TM1); Acanthopanax senticosine; sphagnum moss; sphagnum moss; spongiform septicemia; nitrogen mustard such as chlorambucil, naphthylambucil, chlorphosphamide, estradiol, ifosfamide, bis(2-chloroethyl)methylamine, methoxynitroglycerin hydrochloride, melphalan, neo-embezzin, benzyl mustard cholesterol, prednisolone, trolophosphamide, uramustine; nitrosourea such as carmustine, chloramphenicol, formustine, lomustine, nimustine, ramustine; antibiotics such as enediyne antibiotics (e.g., kachimycin, kachimycin γII, kachimycin ωII) Danendomycin, including danendomycin A; bisphosphonates, such as clophosphonate; lamivudine; and new carcinogen chromophores and related chromogens, aclarubicin, actinomycin, authrarnycin, diazoserine, bleomycin, actinomycin C, carabicin, erythromycin, carcinogen, chromopycin, actinomycin D, danomycin, detoxin, 6-diazo-5-oxo-L-leucine, doxorubicin (including morpholino-doxorubicin, cyanomorpholino-doxorubicin) (e.g., 2-pyrrolodoxomil, deoxydoxomil), epirubicin, esopycin, idarubicin, mesorcinol, mitomycin C, mycophenolic acid, nogamycin, oligomycin, pepromycin, pofiromycin, puromycin, quelamycin, rodorubicin, streptomycin, streptozotocin, tuberculin, ubenimex, fenestrated statin, zolarubicin; antimetabolites such as methotrexate and 5-fluorouracil (5-FU), folic acid analogues, etc.) Examples of purine analogs include: folate, pteroxate, trimesat; purine analogs such as fludarabine, 6-mercaptopurine, thioimidine, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmoflurane, araciclovir, dideoxyuridine, deoxyfluorouridine, enoxabin, fluorouridine; androgens such as capprostol, drotaldophenone propionate, cyclothionol, meandrolone, testrolide; antiadrenergics such as mitotane, triplostertan; folic acid supplements such as folic acid; aceglucuronolactone; aldehydephosphoramide glycoside; aminopyruvate; enuracil; acridine; amustine (bestrabucil); and bismuth subsalicylate.Edatraxate; defofamine; colchicine; seriquinone; elfornithine; eletine; epothilone; ethoxyphenidine; gallium nitrate; hydroxyurea; lentinan; lonidainine; maytansinoids such as maytansine and anserine; mitoxantrone; mitoxantrone; mopidanmol; nitraerine; pentostatin; phenamet; pirarubicin; loxoantrone; podophyllin; 2-ethylhydrazine; methylbenzylhydrazine; PSK polysaccharide complex; razosen; rhizomycin; cizonan; germanium spiroamine; tenuazonic acid; triamine; 2,2',2”-trichlorotriethylamine; crescent toxin (especially T-2 toxin, verracurin A) A) Lactin A and Serpentin); Ureaplasma; Vinpocetine; Dacarbazine; Mannitol Nitrogen Mustard; Dibromomannitol; Dibromoeutherol; Guanospermum; Gacytosine; Arabin (“Ara-C”); Cyclophosphamide; Taxanes such as paclitaxel and docetaxel; 6-Thioguanine; Mercaptopurine; Platinum complexes such as cisplatin, oxaliplatin and carboplatin; Vincristine; Platinum; Etoposide (VP-16); Ifosfamide; Mitoxantrone; Vincristine; Vinorelbine; Mitoxantrone; Teniposide; Edatraxa (edatrexate); Daunomycin; Amistetine; Xeloda; Ibandronate; Irinotecan (e.g., CPT-11); Topoisomerase inhibitors RFS 2000; difluoromethylornithine (DMFO); retinoids such as retinoic acid; capecitabine; carboplatin; procarbazine; prilocytic styramine; gemcitabine; vinorelbine; farnesyltransferase inhibitors; antiplatin; and pharmaceutically acceptable salts, acids, or derivatives of any of the foregoing.

[0107] Immunotherapy typically relies on immune effector cells and molecules to target and destroy cancer cells. Immune effectors can be, for example, specific antibodies against certain markers on the surface of tumor cells. Antibodies can act as therapeutic effectors on their own, or they can recruit other cells to actually kill the cells. Antibodies can also be conjugated to drugs or toxins (chemotherapeutic agents, radionuclides, ricin A chain, cholera toxin, pertussis toxin, etc.) and act solely as a target. Alternatively, effectors can be lymphocytes carrying surface molecules that interact directly or indirectly with tumor cell targets. Various effector cells include cytotoxic T cells and NK cells, as well as genetically engineered variants of these cell types modified to express chimeric antigen receptors.

[0108] Immunotherapy may include suppressor T regulatory cells (Tregs), myeloid-derived suppressor cells (MDSCs), and cancer-associated fibroblasts (CAFs). In some embodiments, immunotherapy is a tumor vaccine (e.g., whole-tumor cell vaccines, peptides, and recombinant tumor-associated antigen vaccines) or adoptive cell therapy (ACT) (e.g., T cells, natural killer cells, TILs, and LAK cells). T cells may be engineered with a tumor antigen-specific chimeric antigen receptor (CAR) or T cell receptor (TCR). As used herein, a chimeric antigen receptor (or CAR) may refer to any engineered receptor that specifically targets a target antigen, and when expressed in a T cell, the CAR can confer specificity to that T cell. Once T cells expressing a chimeric antigen receptor have been constructed using standard molecular techniques, they can be introduced into a patient, as in, for example, adoptive cell transfer techniques. In some aspects, these T cells are activated CD4 and / or CD8 T cells in an individual, characterized by CD4 and / or CD8 T cells that produce γ-IFN, and / or with enhanced cytotoxic activity relative to prior combination administration. The CD4 and / or CD8 T cells may exhibit enhanced release of cytokines selected from the group consisting of IFN-γ, TNF-α, and interleukins. The CD4 and / or CD8 T cells may be effector memory T cells. In some embodiments, the CD4 and / or CD8 effector memory T cells are characterized by expression of CD44. high CD62L low .

[0109] Immunotherapy can be a cancer vaccine containing one or more cancer antigens, particularly proteins or immunogenic fragments thereof, DNA or RNA encoding said cancer antigen (particularly proteins or immunogenic fragments thereof), cancer cell lysates, and / or protein preparations derived from tumor cells. As used herein, a cancer antigen refers to an antigenic substance present in cancer cells. In principle, any protein produced in cancer cells that has an abnormal structure due to mutation can serve as a cancer antigen. In principle, a cancer antigen can be the product of mutated oncogenes and tumor suppressor genes, the product of other mutated genes, overexpressed or aberrantly expressed cellular proteins, cancer antigens produced by oncogenic viruses, carcinoembryonic antigens, altered cell surface glycolipids and glycoproteins, or cell type-specific differentiation antigens. Examples of cancer antigens include aberrant products of the ras and p53 genes. Other examples include tissue differentiation antigens, mutated protein antigens, oncogenic viral antigens, testicular cancer antigens, and vascular or matrix-specific antigens. Tissue differentiation antigens are antigens specific to a certain type of tissue.

[0110] Immunotherapy can be an antibody, such as part of a polyclonal antibody preparation, or it can be a monoclonal antibody. The antibody can be a humanized antibody, a chimeric antibody, an antibody fragment, a bispecific antibody, or a single-chain antibody. The antibodies disclosed herein include antibody fragments, such as, but not limited to, Fab, Fab', and F(ab')2, Fd, single-chain Fv (scFv), single-chain antibodies, disulfide-linked Fv (sdfv), and fragments containing VL or VH domains. In some aspects, the antibody or a fragment thereof specifically binds to epidermal growth factor receptor (EGFR1, Erb-B1), HER2 / neu (Erb-B2), CD20, vascular endothelial growth factor (VEGF), insulin-like growth factor receptor (IGF-1R), TRAIL receptor, epithelial cell adhesion molecule, carcinoembryonic antigen, prostate-specific membrane antigen, mucin-1, CD30, CD33, or CD40.

[0111] Examples of monoclonal antibodies include, but are not limited to, trastuzumab (anti-HER2 / neu antibody); pertuzumab (anti-HER2 mAb); cetuximab (anti-epidermal growth factor receptor EGFR chimeric monoclonal antibody); panitumumab (anti-EGFR antibody); nimotuzumab (anti-EGFR antibody); zarumumab (anti-EGFR mAb); nimotuzumab (anti-EGFR mAb); MDX-210 (humanized anti-HER-2 bispecific antibody); MDX-447 (humanized anti-EGF receptor bispecific antibody); rituximab (chimeric mouse / human anti-CD20 mAb); octotuzumab (anti-CD20 mAb); octotuzumab (anti-CD20 mAb); tosimomumab-I131 (anti-CD20 mAb); tiimumab (anti-CD20 mAb); bevacizumab (anti-VEGF). mAb); Ramucirumab (anti-VEGFR2 mAb); Ranibizumab (anti-VEGF mAb); Aflibercept (VEGFR1 and VEGFR2 extracellular domains fused with IgG1 Fc); AMG386 (angiopoietin-1 and -2 binding peptides fused with IgG1 Fc); Darrottozumab (anti-IGF-1R mAb); Gelatumab (anti-CD33 mAb); Alemumab (anti-Campath-1 / CD52 mAb); Bentuximab (anti-CD30 mAb); Caputuximab (bispecific mAb targeting epithelial cell adhesion molecules and CD3); Napotumab (anti-5T4 mAb); Gelatumab (anti-carbonic anhydrase ix); or Fatuximab (anti-folate receptor).Other examples include antibodies such as Panorex™ (17-1A) (mouse monoclonal antibody); Panorex (MAb17-1A) (chimeric mouse monoclonal antibody); BEC2 (anti-idiotype mAb, mimicking GD epitopes) (used in combination with BCG); Oncolym (Lym-1 monoclonal antibody); SMART M195 Ab, humanized 13' 1 LYM-1 (Oncolym); Ovarex (B43.13, anti-idiotype mouse mAb); 3622W94 mAb, which binds to the pan-cancer antigen EGP40 (17-1A) on adenocarcinoma; Zenapax (dacizumab) (SMART anti-Tac (IL-2 receptor); SMART M195 Ab (humanized antibody, humanized); NovoMAb-G2 (pan-cancer specific antibody); TNT (chimeric mAb against histone antigen); Gliomab-H (monoclonal antibody, humanized Ab); GNI-250 mAb; EMD-72000 (chimeric EGF antagonist); LymphoCide (epazolizumab) (humanized IL-12 antibody); and MDX-260 bispecific antibody targeting GD-2, ANA antibody, SMARTIDIO antibody, SMART ABL 364 antibody, or ImmuRAIT-CEA.Other examples of antibodies include Zanulimumab (anti-CD4 mAb), Keliximab (anti-CD4 mAb); Ipilimumab (MDX-101; anti-CTLA-4 mAb); Texilimumab (anti-CTLA-4 mAb); Daxetizumab (anti-CD25 / IL-2R mAb); Baliximab (anti-CD25 / IL-2R mAb); MDX-1106 (anti-PD-1 mAb); Anti-GITR antibody; GC1008 (anti-TGF-β antibody); Metilimumab / CAT-192 (anti-TGF-β antibody); Ledelimumab / CAT-152 (anti-TGF-β antibody); ID11 (anti-TGF-β antibody); Denoxaparin (anti-RANKL mAb); BMS-663513 (humanized anti-4-1BB mAb); SGN-40 (humanized anti-CD40). mAb); CP870, 893 (humanized anti-CD40 mAb); infliximab (chimeric anti-TNF mAb); adalimumab (human anti-TNF mAb); sertozumab (humanized Fab anti-TNF); golimumab (anti-TNF); etanercept (TNFR extracellular domain fused with IgG1 Fc); beracip (CTLA-4 extracellular domain fused with Fc); abatacept (CTLA-4 extracellular domain fused with Fc); belimumab (anti-B lymphocyte stimulating factor); morotumab-CD3 (anti-CD3 mAb); oxizumab (anti-CD3 mAb); telizumab (anti-CD3 mAb); tocilizumab (anti-IL-6R mAb); REGN88 (anti-IL-6R mAb); ustekinumab (anti-IL-12 / 23 monoclonal antibody); brenumab (anti-IL-12 / 23) mAb); nastatinumab (anti-α4 integrin); vedolizumab (anti-α4β7 integrin mAb); T1h (anti-CD6 mAb); epazolizumab (anti-CD22 mAb); efalizumab (anti-CD11a mAb); and asceticip (extracellular domain of transmembrane activator and calcium-regulated ligand interactor fused with Fc).

[0112] When monitoring mammals with or suspected of having cancer, as described herein (e.g., identifying differences in cfDNA fragmentation between healthy individuals and cancer patients based in part on CpG differentially methylated in a specific sequence background), monitoring can occur before, during, and / or after cancer treatment. The monitoring methods provided herein can be used to determine the efficacy of one or more cancer treatments and / or to select mammals for enhanced monitoring. In some embodiments, monitoring may include identifying cfDNA fragmentation profiles as described herein. For example, cfDNA fragmentation profiles may be obtained before administering one or more cancer treatments to a mammal with or suspected of having cancer, one or more cancer treatments may be administered to the mammal, and one or more samples may be analyzed during cancer treatment. In some embodiments, cfDNA fragmentation profiles may change during cancer treatment (e.g., any cancer treatment described herein). For example, a cfDNA fragmentation profile indicating that a mammal has cancer may change to a cfDNA fragmentation profile indicating that the mammal does not have cancer. Such a change in cfDNA fragmentation profile may indicate that the cancer treatment is effective. Conversely, a cfDNA fragmentation profile may remain unchanged (e.g., identical or substantially identical) during cancer treatment (e.g., any cancer treatment described herein). Such a static cfDNA fragmentation profile may indicate that the cancer treatment is ineffective.

[0113] In some embodiments, monitoring may also include conventional techniques capable of monitoring one or more cancer treatments (e.g., the efficacy of one or more cancer treatments). In some embodiments, mammals selected for enhanced monitoring may be administered diagnostic tests (e.g., any of the diagnostic tests disclosed herein) at a higher frequency compared to mammals not selected for enhanced monitoring. For example, mammals selected for enhanced monitoring may administer diagnostic tests twice daily, daily, every two weeks, weekly, every two months, monthly, quarterly, semi-annually, annually, or any of these frequencies. In some embodiments, mammals selected for enhanced monitoring may be administered one or more additional diagnostic tests compared to mammals not selected for enhanced monitoring. For example, mammals selected for enhanced monitoring may be administered two diagnostic tests, while mammals not selected for enhanced monitoring may be administered only a single diagnostic test (or no diagnostic test). In some embodiments, mammals selected for enhanced monitoring may also be selected for further diagnostic testing. Once a tumor or cancer (e.g., cancer cells) is identified (e.g., by any of the various methods disclosed herein), enhanced surveillance (e.g., assessing the progression of the tumor or cancer in the mammal and / or assessing the development of one or more cancer biomarkers (e.g., mutations)) and further diagnostic testing (e.g., to determine the size and / or exact location of the tumor or cancer (e.g., tissue origin)) may be beneficial for mammals selected for enhanced surveillance. In some embodiments, after the detection of cancer biomarkers and / or differentially methylated CpGs in a specific sequence background to identify differences in cfDNA fragmentation between healthy individuals and cancer patients has not improved or worsened, one or more cancer treatments may be administered to the selected mammals for enhanced surveillance. Any cancer treatment disclosed herein or known in the art may be applied. For example, for mammals selected for enhanced surveillance, further surveillance may be conducted, and if cancer cells persist during enhanced surveillance, cancer treatment may be administered. Additionally, or alternatively, cancer treatment may be administered to mammals selected for enhanced surveillance, and further surveillance may be conducted as cancer treatment progresses. In some embodiments, after cancer treatment has been administered to mammals selected for enhanced surveillance, enhanced surveillance will reveal one or more cancer biomarkers (e.g., mutations). In some implementations, one or more such cancer biomarkers will provide a reason for administering different cancer treatments (e.g., during cancer treatment, cancer cells may develop resistance mutations that make them resistant to the original cancer treatment).

[0114] When a mammal is identified as having the cancer described herein (e.g., at least in part based on the mammal's cfDNA fragmentation profile), this identification can be made before and / or during cancer treatment. The methods provided herein for identifying a mammal with cancer can be used as initial diagnostic tools to identify a mammal (e.g., with cancer prior to any treatment procedure) and / or to select the mammal for further diagnostic testing. In some embodiments, once a mammal is identified as having cancer, further testing can be applied to the mammal and / or selection can be made for further diagnostic testing. In some embodiments, the methods provided herein can be used to select the mammal for further diagnostic testing during a period prior to when conventional techniques can diagnose early-stage cancer in the mammal. For example, the methods provided herein for selecting mammals for further diagnostic testing can be used when a mammal has not yet been diagnosed with cancer by conventional methods, and / or when it is unknown whether the mammal carries cancer. In some embodiments, mammals selected for further diagnostic testing can be subjected to diagnostic tests (e.g., any of the diagnostic tests disclosed herein) at a higher frequency compared to mammals not selected for further diagnostic testing. For example, mammals selected for further diagnostic testing may be given the diagnostic tests at frequencies such as twice daily, once daily, once every two weeks, once weekly, once every two months, once monthly, once quarterly, once every six months, once a year, or any of these frequencies. In some embodiments, mammals selected for further diagnostic testing may be given one or more additional diagnostic tests compared to mammals not selected for further diagnostic testing. For example, mammals selected for further diagnostic testing may be given two diagnostic tests, while mammals not selected for further diagnostic testing may be given only one diagnostic test (or no diagnostic test). In some embodiments, the diagnostic testing method can identify the presence of cancer of the same type as the initially detected cancer (e.g., having the same tissue or origin) (e.g., at least partially based on the mammal's cfDNA fragmentation profile). Additionally, or alternatively, the diagnostic testing method can identify the presence of cancer of a different type than the initially detected cancer. In some embodiments, the diagnostic testing method is a scan. In some implementations, the scan is a computed tomography (CT), CT angiography (CTA), esophageal contrast (barium meal), barium enema, magnetic resonance imaging (MRI), PET scan, ultrasound examination (e.g., endobronchial ultrasound, endoscopic ultrasound), X-ray, or DEXA scan.

[0115] In some embodiments, diagnostic testing methods include physical examinations such as anoscopy, bronchoscopy (e.g., autofluorescence bronchoscopy, white light bronchoscopy, navigation bronchoscopy), colonoscopy, digital mammography, endoscopic retrograde cholangiopancreatography (ERCP), esophagogastric and duodenal endoscopy, mammography, cervical smear examination, pelvic examination, positron emission tomography (PET-CT), and computed tomography (PET-CT). In some embodiments, mammals selected for further diagnostic testing may also be selected for enhanced surveillance. Once a tumor or cancer (e.g., cancer cells) is identified (e.g., by any of the various methods disclosed herein), enhanced surveillance of the mammal (e.g., to assess the progression of the tumor or cancer in the mammal and / or to assess the development of one or more cancer biomarkers, such as mutations) and further diagnostic testing (e.g., to determine the size and / or exact location of the tumor or cancer) would be beneficial. In some embodiments, cancer treatment is administered to mammals selected for further diagnostic testing after detection of cancer biomarkers and / or after the mammal's cfDNA fragmentation profile has not improved or worsened. Any cancer treatment method disclosed herein or known in the art can be applied. For example, further diagnostic tests can be administered to mammals selected for further diagnostic testing, and cancer treatment can be administered if a tumor or cancer is confirmed. Alternatively, cancer treatment can be administered to mammals selected for further diagnostic testing, and the treatment can be further monitored as the cancer treatment progresses. In some embodiments, after cancer treatment is administered to mammals selected for further diagnostic testing, the additional tests will reveal one or more cancer biomarkers. In some embodiments, these one or more cancer biomarkers (e.g., mutations) will provide a reason for administering different cancer treatments (e.g., during cancer treatment, resistance mutations may appear in cancer cells, and cancer cells carrying these resistance mutations develop resistance to the original cancer treatment). Example

[0116] Example 1: DNA methylation and gene expression as determinants of cell-free DNA fragmentation across the genome With the development of high-throughput sequencing methods, it has become possible to study the characterization of cfDNA fragments, including features associated with basal nucleosomes. In healthy individuals, the localization of nucleosomes and the larger chromatin structures show striking similarities to those found in myeloid cells and lymphocytes. 8-10 Similarly, the methylation patterns of cfDNA from cancer-free individuals are very similar to those of DNA methylation in leukocytes. 11 Although epigenetic changes are related to genome packaging 12 and chromatin structure 13 and gene expression 14Closely related, but to date, research on the relationship between methylation, expression, and cfDNA fragmentation is limited. 15-17 Still limited.

[0117] In this study, motif-related features associated with nucleosome localization and breakpoints in cfDNA fragments were analyzed in both healthy individuals and cancer patients. This demonstrated how epigenetic markers lead to specific cfDNA fragmentation patterns and how these relate to methylation and gene expression. Using this information, it was shown that differentially methylated CpGs in a specific sequence context can be used to identify differences in cfDNA fragmentation between healthy individuals and cancer patients. method

[0118] Study population To analyze motif frequencies, recurring ends, and their relationship with gene expression and methylation, low-coverage whole-genome sequencing (WGS) of cfDNA from 787 cancer-free individuals and 182 cancer individuals was used (1-2 x). 32 Previously, cfDNA methylation from cancer-free individuals was analyzed using the Illumina Infinium Methylation EPIC microarray¹¹, and the data was available from the NCBI Gene Expression Comprehensive Database (GEO) (dataset number GSE122126). Cellular expression factors contributing to cfDNA methylation in cancer-free individuals were previously validated. 9,11 This indicates that most cfDNA originates from myeloid-derived cells. As previously published, average gene expression from multiple myeloid cell lines was used. 33 .

[0119] cfDNA sample processing Whole-genome libraries from cancer patients and cancer-free individuals were sequenced using 100-bp paired-end runs (200 cycles) on an Illumina HiSeq2500 platform at 1–2x coverage per genome. Before alignment, adapter sequences were filtered from reads using FastP software. 34 Bowtie2 software was used. 35 The sequenced reads were compared with the hg19 human reference genome and analyzed using Sambamba. 36 Remove duplicate reads. Only retain reads with a MapQ score of at least 30 or higher. After alignment, use bedtools. 37 Each alignment is converted into a gene region representing the sequenced DNA fragment.

[0120] motif frequencies around the ends of cfDNA fragments The expected frequencies of 3-bp motifs in the human genome were calculated by statistically analyzing the occurrence rate of each 3-bp motif in the human genome (hg19). To calculate the empirical frequencies of 3-bp motifs at the ends of cleaved (sonic) fragments, we used published sequencing data from 10 lymphoblastoid cell lines. 38 Genomic DNA from these lymphoblastic cell lines was fragmented using a Covaris M220 focused ultrasound system. For this analysis, the data were reanalyzed in the same manner as the cfDNA sample analysis. For 543 low-coverage WGS from these 10 lymphoblastic cell lines and cancer-free individuals, the number of 3-bp motifs at the fragment initiation and the number of anticomplementary 3-bp motifs around the fragment ends were counted. Each 3-bp motif consists of one base outside the fragment, followed by the first two bases of the fragment. Using these absolute values, the relative frequencies of 64 3-bp motifs were calculated.

[0121] To quantify the preference for cfDNA fragment termination at specific genomic locations, the ratio of cfDNA fragments ending at that location (the end of a repetitive fragment) to the number of cfDNA fragments overlapping that location by 1 bp or more within + / - 50 bp (nearby fragments) was calculated. This ratio of repetitive fragment ends to nearby fragments was calculated by aggregating cfDNA fragments from all 543 cancer-free individuals. This calculation was repeated for each location in the hg19 reference genome, where the number of nearby fragments consisted of 200 or higher high-quality alignments (MAPQ > 30) to capture repetitions in non-repetitive regions of the genome. Locations with high repetition were defined as those where the ratio of repetitive fragment ends to nearby fragments was at least 5%.

[0122] X-ray crystal structure of nucleosomes To identify DNA-binding nucleosome structures, the term "nucleosome" was searched in the Protein Database (PDB; rcsb.org), and the results were filtered for structures derived from X-ray diffraction or cryo-electron microscopy (cryo-EM), yielding 427 entries. DNA sequences from these structures (648) were downloaded, and sequences with a length of at least 167 bases were selected. This identified 80 unique sequences from 51 PDB entries. These entries were visually inspected, removing those with fewer than 167 identifiable bases or whose interaction with the H1 linker was interfered with by other DNA-binding proteins. This left 17 structures (Table 3). Motifs were considered well-positioned if they were within 5 Å of the H1 linker, or if bases on the same strand within 167 bases of the H1 linker were within 5 Å of the H1 linker.

[0123] Correlating cfDNA fragment patterns with CpG methylation To determine whether cfDNA fragmentation patterns from cancer-free individuals are affected by methylation, raw data from the Illumina Infinium Methylation EPIC array were analyzed. This data came from eight different cfDNA experiments, encompassing four biologically distinct cohorts (young men, older men, young women, and older women). 11 The Infinium array was processed using standard procedures. 39 A numerical score (β value) between 0 (unmethylated) and 1 (methylated) was obtained for each CpG and averaged across the samples to summarize the overall methylation level. CpG sites were labeled as unmethylated if the average β value was less than 0.3 and as methylated if the average β value was greater than 0.7. For 543 cfDNA plasma samples from cancer-free individuals (Table 1), the positions of Infinium CpG sites within cfDNA fragments were recorded using a 1-base index. CpG sites were grouped according to the average β value of the Infinium array. The number of fragments starting or ending at a CpG site within a CpG group was calculated, and this frequency was normalized by the number of fragments overlapping within a 50 bp window of the start or end position. Fragments were further classified based on their 3 bp terminal motif and whether the cfDNA fragment was located in a CpG island, shore region, shelf region, or offshore region.

[0124] CpG islands and cfDNA sequence coverage and fragment size at transcription start sites (TSS) To summarize the cfDNA fragment length at a CpG island, the average length of the cfDNA fragment at the start or end of the CpG island was calculated. This is conventionally referred to as position 0. This summarization step was repeated in increments of 1 bp at positions x bases from the CpG island, from x = -3000 bp to x = +3000 bp, and from x = -500000 bp to x = +500000 bp. This procedure was repeated for each CpG island. The average fragment length at the TSS was also summarized in a similar manner, where position 0 represents the TSS. Due to the use of low-coverage WGS, cfDNA fragments were pooled from all 543 cancer-free samples. Methylation data described previously were used. 11 Gene expression data from myeloid cell lines 33 These regions were sorted to visualize patterns associated with CpG island methylation and gene expression (Fig. 3A, 3B, 3C, 3D).

[0125] Gene set enrichment analysis Use from molecular signature database 43 Hallmark 41 and KEGG 42 Gene set enrichment analysis was performed on the gene set. 40Genes in RNA sequencing were graded using the mean normalized counts from healthy PBMC samples. In methylation analysis, genes were graded using the mean β value of all CpGs covering the transcript. These metrics were the same as those previously used to grade genes for visualizing cfDNA coverage around CpG islands and transcription start sites. In coverage-based analyses, normalized fragment counts of overlap with transcription start sites or CpG islands falling within the transcript were used. Each graded set was permuted 10,000 times to obtain the minimum unadjusted p-value of 1e-4. All gene sets (unadjusted p-value <.1) that showed moderate significance in all comparisons in any direction were selected for inclusion in a heatmap displaying gene set enrichment scores. Figure 3G ).

[0126] Multivariate model A generalized linear model was used to assess the relationship between the average cfDNA fragment size and total coverage at the transcript level and RNA expression, WPS, and methylation. For methylation, we calculated the mean β value per CpG island in 97 blood samples processed on the Infinium chip (see [link to Infinium chip]). Study population The R package `annotatr` (version 1.28.0) was used to map CpG islands to transcripts based on their proximity to the TSS. Transcripts were considered methylated if the mean β value was 0.5 or higher, otherwise unmethylated. The mean RNA expression levels (mean TPM) of the six myeloid cell lines were converted to log10 (mean TPM+1) and then centered and normalized by the overall mean and standard deviation of all transcripts. The WPS values ​​of each transcript within the range of +1 to +10 bases from the TSS were summarized, centered, and normalized. The total cfDNA coverage at each base within the range of -10 bp to -1 bp from the TSS in 543 non-cancer cells was calculated and averaged, while the mean fragment size was calculated within the range of -1480 bp to -1471 bp from the TSS. For all 10 bp genomic intervals within 2500 bp from the TSS, the intervals used to summarize cfDNA coverage, fragment size, and WPS were evaluated. The intervals that generated the measurement with the highest absolute correlation to RNA expression were selected for regression analysis. Based on the above quantitative summary, the expected normalized coverage (DV) of transcript i is provided by the following formula: .

[0127] The expected fragment length was modeled in a similar manner. Coefficients from these models were estimated using generalized linear models with identity connection functions in R (version 4.3.2). Using analysis of variance (ANOVA), we assessed whether RNA expression helped explain—after adjusting for methylation and WPS—changes in coverage by testing the main effects of RNA expression and its interaction with methylation. We performed a similar ANOVA to assess whether methylation explained the changes in coverage after adjusting for the effects of RNA and WPS on coverage. Forest plots were generated for each model using sjPlot (version 2.8.15) to visualize the estimated model coefficients with 95% confidence intervals.

[0128] Monte Carlo simulation of human cfDNA coverage in xenograft models For each of the six xenografts (three IDH1 R132H mutant xenografts and three IDH1 wild-type xenografts), we calculated the cfDNA coverage for the top 500, 1000, 2000, 3000, 4000, and 5000 most differentially methylated CpG islands or most differentially expressed genes. These coverages were normalized for the total size of these regions. By comparing the IDH1 mutant and wild-type, we determined whether the direction of difference in normalized coverage was consistent with prior expectations that for each of the four possible comparisons (hypermethylated regions in the IDH mutant, hypermethylated regions in the IDH wild-type, high expression in the IDH mutant, and high expression in the IDH wild-type), we would observe higher coverage in methylated regions and lower coverage in expressed regions. To assess the likelihood of observing empirical consistency under the null hypothesis—that there is no difference in cfDNA coverage between IDH1 mutant and wild-type tumors in mice—we rearranged the mutant and wild-type labels and assessed consistency as described above. This was repeated 10,000 times. We repeated this process for each of the 19 possible sample label rearrangements, obtaining the consistency distribution under the null hypothesis. The p-value was calculated as the proportion of rearrangements where the consistency was as high or higher than that obtained from class labels that were never rearranged. We repeated these analyses for each of the six region or gene list sizes mentioned above, for a total of 24 comparisons.

[0129] Tumor-specific cfDNA methylation patterns based on differentially methylated CpG Data on differentially methylated CpG from individuals with and without cancer were evaluated using publicly available data. Published data on pancreatic cancer. 30A large number of differentially methylated regions were identified. Using these differentially methylated CpGs, different subgroups were formed based on the direction of differential methylation (methylation in non-cancer vs. unmethylation in pancreatic cancer; unmethylation in non-cancer vs. methylation in pancreatic cancer) and based on 3bp and 4bp motifs. For each sample, a total of 16 different features were extracted, based on the ratio of polymerized cfDNA fragments starting or ending at these motifs to the number of polymerized fragments overlapping a 101bp window around that motif (Figure 4A). These features were used to generate a machine learning model, for which performance was compared with the original DELFI method 8. The ensemble model combining the methylation-based method with DELFI showed a synergistic improvement in performance compared to the two models alone (Figure 4B).

[0130] Statistical methods All t-tests are Welch two-sample t-tests unless otherwise stated. Statistical analysis was performed using R, version 4.2.0. result

[0131] The frequency and composition of cfDNA start and end sequences were investigated; these characteristics have been considered non-random and potentially related to endogenous DNAase I. 8,19 The cutting of cfDNA is relevant. To rigorously identify cfDNA end positions, cfDNA sequence data were pooled from low-coverage whole-genome sequencing from a healthy subject cohort (n=543), and the frequency of cfDNA breakpoints was investigated at each possible location in the genome. Only reads with high sequence and mapping quality were considered, and the ratio of the number of cfDNA fragments starting or ending at a specific location to the number of fragments starting or ending within 50 bp of that location was calculated. This method is necessary for identifying recurring fragment end sequences and for interpreting differences in cfDNA fragment size and cross-genome coverage. 8 .

[0132] Fragments of cfDNA with more frequently observed terminal positions were evaluated and found to be enriched for specific motifs. These typically include thymine or adenine preceding the initiation of the cfDNA fragment, and two cytosines (A / T|CC) or cytosine followed by guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment (Fig. 1A). It is speculated that recurring terminal sequences in cfDNA fragments likely represent nucleosome-protected sites, and the frequency of these base motifs was found to be further increased at “preferred” repeating ends in healthy individuals (Fig. 1A, 1B). The occurrence rates of cfDNA fragment ends with preferred A / T|CC and A / T|CG were observed to be much higher than theoretically expected in the genome (26.5× for A / T|CC and 5.5× for A / T|CG) (Fig. 1A) (p<2.2e-16, one-sample t-test), while the frequencies of ends from fragments generated by sonication of genomic DNA were close to theoretical abundance.

[0133] To understand the underlying basis for the enrichment of A / T|CG terminal sequences, the available X-ray crystal and cryo-electron microscopy (cryo-EM) structures of nucleosome-bound DNA were examined. It was found that in 82% of the unique structures, the A / T|CG motif was located near the H1 linker or concentrated at a distance of 167 bp, typical of cfDNA molecules (Figure 1C). In contrast to the fragment terminal sequences, adenine and thymine were observed to be enriched in the internal regions of the cfDNA fragments, with these nucleotides exhibiting a periodicity of 10–11 bp in frequency along the fragment length. Figure 5 These observations are consistent with current predictions of DNA winding around the nucleosome core, and the necessity of alternating between rigid DNA regions (rich in C and G) and more flexible regions (rich in A and T) to wind DNA around the nucleosome nearly two times. 20 .

[0134] Given the prevalence of CG-containing cfDNA fragment ends, the question arose whether epigenetic markers at these locations would influence cfDNA fragmentation. cfDNA fragments were mapped to the genome, and their ends were assessed, with the assessment relating to methylated and unmethylated CpG sites on previously identified cfDNA from healthy individuals obtained from an array evaluating methylation at 850K CpG sites (Moss J. et al., Nat Commun 9, 5068 (2018)). Methylated CpG sites were observed to be enriched at the ends of A / T|CG cfDNA fragments, while unmethylated CpG sites were relatively evenly distributed across the lengths of these molecules. Figure 2A , 7A7B). To quantitatively assess the enrichment of cfDNA fragment ends at CpGs, for each CpG, the fraction of the cfDNA fragment starting or ending at that dinucleotide was calculated by dividing it by the number of fragments with start or end positions within 50 bp around each CpG. It was observed that the fraction of preferred ends increased by up to 2.4-fold with higher methylation levels (p < 2.2e-16, t-test) (Figure 2B). Furthermore, CG cfDNA fragment ends were enriched by up to 2.2-fold at genome-wide methylated CpG locations, including CpG islands, shore regions, shelf regions, and offshore regions. Figure 6 This indicates that the enrichment of methylated CpG fragments is a common feature of cfDNA in these regions.

[0135] It was observed that methylated CG terminal sequences were preferentially enriched even when overlapping with frequently observed CC fragment terminal sequences. When N|CC sequences were followed by guanine (G) to form N|CCG sequences, the frequency of typical N|CC terminal motifs decreased because they competed with overlapping C|CG motifs, which were enriched when CpG sites containing these sequences were methylated (Fig. 2B, Fig. 2C). The overall effect of this competition resulted in a significant decrease in N|CCG fragment terminal sequences at methylated CpG sites, even exceeding the corresponding increase in N|CG fragment terminal sequences, as seen, for example, a 3.7-fold decrease in T|CCG terminal sequences (t-test, p < 2.2e-16) compared to a 2.2-fold increase in C|CG terminal motifs (t-test, p < 2.2e-16).

[0136] To provide additional biological evidence regarding the association between methylated CG and cfDNA fragmentation, cfDNA fragments derived from the X chromosome in healthy individuals were compared, since it is generally accepted that in females, one of the two X chromosomes is inactivated by methylation of CpG islands, while these regions on the single X chromosome in males are not methylated. 21,22 Consistent with the observation of methylation-induced fragment end enrichment, cfDNA fragments ending in CG were enriched at CpG islands on the female X chromosome compared to males, while fragments ending in CCG were preferentially reduced (for CG ends, on average 0.39% males vs 0.48% females, t-test, p < 2.2e-16; for CCG ends, on average 0.68% males vs 0.54% females, t-test, p < 2.2e-16); however, these differences were not observed on autosomes in males and females (t-test, p = 0.58 and p = 0.80) (Fig. 2D). Although the above trend continued in the CpG shore region, higher CG fragment end enrichment was observed in males compared to females in the CpG shelf region and the open ocean region, consistent with previously reported higher methylation in these regions on the male X chromosome ( Figure 8 ) 12,23 .

[0137] In addition to the enrichment of cfDNA fragment ends at epigenetic marker sites, a correlation was observed between cfDNA sequence coverage (the average number of overlapping cfDNA molecules at a specific location) and methylation levels (r = 0.6, p < 2.2e-16, Pearson correlation test) (Fig. 9A), and regions spanning methylated CpG islands showed up to 1.7 times higher coverage compared to unmethylated ones (p < 2.2e-16, t-test) (Fig. 3A). Given the relationship between CpG island methylation and gene expression, the relationship between gene expression at transcription start sites (TSS) and cfDNA fragmentation patterns was also assessed. A negative correlation was found between cfDNA coverage at TSS and the expression levels of nearby genes (r = -0.48, p < 2.2e-16, Pearson correlation test) (Fig. 9B). The overall level of cfDNA fragments overlapping with the TSS of unexpressed genes was up to 3.7 times higher than that in regions of expressed genes, likely due to the lack of nucleosome destabilization by transcription factors (p < 2.2e-16, t-test) (Fig. 3B). Consistent with the higher cfDNA coverage, variations in cfDNA fragment size were also observed in these regions, including 4-5 bp shorter fragments in the 800-1000 bp upstream of the TSS of highly expressed genes compared to unexpressed genes (164.5 bp vs. 168.6 bp, p < 2.2e-16, t-test), or in regions surrounding unmethylated CpG islands compared to highly methylated CpG regions (165.1 bp vs. 167.3 bp, p < 2.2e-16, t-test) (Fig. 3C, 3D). Further examination of the broader regions surrounding the TSS or CpG islands revealed differences between expressed / unexpressed or unmethylated / methylated genes in regions up to 500 kb around these areas (Figures 10A-10D).

[0138] Analysis of gene-adjacent cfDNA fragments in the KEGG and Hallmark gene sets showed that cfDNA coverage was correlated with expression and CpG methylation of all significant gene sets found in blood cells. 11 Correlation (p<0.1, gene set enrichment analysis) (Figure 4A, Table 7). These include higher cfDNA coverage in CpG islands and TSS regions when methylation levels are increased and gene expression levels are decreased, and lower cfDNA coverage when methylation levels are decreased and expression is increased. For example, gene pathways that are not normally expressed in blood cells (including neuronal receptor-ligand interactions or olfactory receptor transduction) are often methylated in regions containing CpG islands or TSS and are presented more prominently in cfDNA fragments. Figure 3FConversely, genes utilized in hematopoiesis (including E2f transcription factor targets and blood cell metabolism-related genes) were highly expressed, more frequently unmethylated, and exhibited lower cfDNA levels in the CpG or TSS regions of these genes (Figure 4B). Overall, cfDNA coverage was found to be correlated with both CpG methylation and the expression of neighboring genes. Figure 11A However, the enrichment of cfDNA fragment ends at recurring CpG sites (compared to gene expression) is more correlated with methylation levels. Figure 11B These results indicate that DNA methylation is a fundamental feature affecting cfDNA fragmentation, and it is associated with gene expression but independent of gene expression.

[0139] To directly and independently analyze the effects of methylation or gene expression on cfDNA coverage, we evaluated the fragmentation coverage of human cfDNA in the plasma of mice with implanted human tumors, which had or did not have the knock-in of the IDH1 chromatin modifier (containing the mutation at R132), a mutation known to be activating in our previous work (Parsons, DW). et al.Science 321, 1807–1812 (2008)), and caused widespread genome-wide alterations in methylation and expression. Duncan, CG et al. Genome Res. 22, 2339–2355 (2012).; Wei, S. et al.Oncogene 37, 5160–5174 (2018); Brennan, CW et al.Cell 155, 462–477 (2013). Mice were injected with the U87 glioblastoma cell line, which is wild-type against IDH1 ( n =3) or modified with the same genotype to contain the R132H mutation ( n =3), and evaluation was performed 20–30 days after tumor implantation. After screening human-derived cfDNA fragments from mouse plasma, high coverage of human cfDNA was observed in regions of increased methylation levels; while low cfDNA coverage was observed in regions of increased expression. Figure 3G () p <0.053 (Monte Carlo simulation), consistent with our previous analysis. This tightly controlled analysis provides a direct causal association between genome-wide variations in epigenetic traits and cfDNA fragmentation.

[0140] Because numerous reports have indicated that the overall size of cfDNA in cancer patients is smaller than that in healthy individuals. 24,25 This suggests the process of tumor development. 26Genomic variations in DNA methylation and gene expression may influence cfDNA fragmentation in cancer patients. To clearly compare tumor-derived cfDNA with wild-type cfDNA, samples from 98 cancer patients were used. 8,27 Ultra-high sensitivity NGS targeted next-generation sequencing was used to detect changes in cfDNA fragment size between tumor-derived and WBC-derived cfDNA. An average change of 3.9 bp was found in tumor-derived cfDNA from these patients, similar to the cfDNA size differences observed in healthy cfDNA at high and low expression TSS regions and at methylated versus unmethylated CpG sites. Figure 3E As we observed, tumor cells typically have an increased number of expressed genes and are hypomethylated compared to white blood cells. Figure 12A , 12B Similar to previous studies) 26,28,29 These results support the idea that changes in gene expression and methylation in cancer cells may be partly responsible for the overall smaller cfDNA fragments observed in cancer patients.

[0141] To identify the impact of CpG methylation differences between healthy individuals and cancer patients on cfDNA fragment ends, regions previously identified by comparing simplified representative bisulfite sequencing (RRBS) data from laser capture microdissection of pancreatic ductal adenocarcinoma and normal pancreatic tissue were evaluated, utilizing methyl DNA immunoprecipitation (MeDIP). 30 These regions were validated in cf DNA. Then, through pancreatic cancer ( n =34), colorectal cancer ( n =27), ovarian cancer ( n =28), lung cancer ( n =39) or breast cancer ( n =54) patients and those from healthy individuals ( n =244) 8 Low-coverage whole-genome cfDNA analysis was conducted to assess fragment end characterization at CG and CCG sites. Regions of increased CpG methylation in non-cancerous tissues showed preferential reduction in cfDNA fragments ending in N|CCG in non-cancerous individuals, compared to the abundance of these fragments in patients with pancreatic cancer or other cancers. Figure 5 A). Conversely, in regions of elevated CpG methylation in pancreatic cancer, an increase in cfDNA fragments ending with CG was observed in cancer patients compared to levels in cancer-free individuals. Figure 5A). The strongest signal was observed in pancreatic cancer patients across all cases, indicating improved performance when using tumor-specific methylation sites in this cancer type. Incorporating the fragment end position distribution of these CG and CCG sites into a gradient boosting tree machine learning model successfully distinguished pancreatic cancer patients from those without pancreatic cancer (AUC = 0.87 with cross-validation). This method was further enhanced by combining genome-wide fragmentation analysis (DELFI) that integrates fragment coverage and size. 8 The combination further improved the sensitivity of the combined method (AUC=0.94, 95% CI=0.90-0.96). These observations suggest that cfDNA fragment end characterization at CG and CCG sites can be used to detect cancer individuals using genome-wide DNA methylation.

[0142] Multiple regression model analysis assessing DNA methylation (Fig. 3A, 3C), gene expression (Fig. 3B, 3D), nucleosome localization (Supplementary Fig. 13), and the interactions between these factors showed that each of these elements independently affected cfDNA coverage and fragment length. Figure 3F The relationship between methylation and coverage is qualitatively similar in additional models incorporating the interaction between DNA methylation and nucleosome localization, as well as in more complex models involving the three-factor interaction of DNA methylation, gene expression, and nucleosome localization (Figure 13). These results highlight that DNA methylation is a fundamental feature influencing cf DNA fragmentation. discuss

[0143] This study particularly demonstrates the profound impact of genome-wide CpG methylation on cfDNA fragmentation. Repeated cfDNA sequences with CG fragment ends are enriched at methylation sites, increasing at N|CG sites and decreasing at competing N|CCG sites in a manner dependent on methylation level. Structural analysis of DNA bound to nucleosomes shows that CG sequences are typically located near histone H1 linkers. These observations, combined with previous molecular dynamics simulations... 31 It has been demonstrated that methylation of the CG sequence provides a more stable interaction between methylated DNA and the H1 adapter, thereby protecting the nucleosome-binding cfDNA fragment from degradation.

[0144] Methylated CpG not only affect the end positions of cfDNA fragments but also lead to a greater amount of circulating cfDNA in these regions. In TSS regions with decreased expression (both at the single-gene level and gene pathway level), cfDNA fragmentation is similarly affected. cfDNA fragment size is altered by both methylation and gene expression changes, and can be observed near CG and TSS sites as well as at locations hundreds or thousands of base pairs away. These determinants of cfDNA size, coupled with the overall increase in gene expression and decrease in methylation observed in human cancer in this study, provide mechanistic evidence for the overall shortening of cfDNA fragment length observed in cancer patients.

[0145] Incorporating cfDNA fragment end features at CpG sites into a cross-validated machine learning model provides a method for detecting cancer independently of other cfDNA features. This method shows complementarity with DELFI cfDNA fragmentation analysis, resulting in improved performance. Understanding the methylation and expression differences affecting cfDNA fragment size and coverage in cancer patients will improve future methods for assessing genome-wide cfDNA fragmentation. Combining methylation and expression changes with other genome-wide epigenetic markers provides complementary insights into the origin and mechanisms of action of cfDNA fragmentation. References

[0146] 1.Richmond, TJ, Finch, JT, Rushton, B., Rhodes, D.&Klug, A.Structure of the nucleosome core particle at 7 Å resolution. Nature 311, 532–537 (1984). 2.Richmond, TJ&Davey, CA The structure of DNA in the nucleosomecore. Nature 423, 145–150 (2003). 3.Ceppellini, R., Polli, E.&Celada, F. A DNA-Reacting Factor in Serum of a Patient with Lupus Erythematosus Diffusus.*. P Soc Exp Biol Med 96, 572–574 (1957). 4.Robbins, W. C., Holman, H. R., Deicher, H.&Kunkel, H. G. ComplementFixation with Cell Nuclei and DNA in Lupus Erythematosus. P Soc Exp Biol Med 96, 575–579 (1957). 5.Miescher, P.&Strässle, R. New Serological Methods for the Detectionof the L. E. Factor*. Vox Sang 2, 283–287 (1957). 6.Seligmann, M. [Demonstration in the blood of patients withdisseminated lupus erythematosus a substance determining a precipitationreaction with desoxyribonucleic acid]. Comptes Rendus Hebdomadaires Des Seances De L'academie Des Sci 245, 243–5 (1957). 7.Barra, G. B. et al. EDTA-mediated inhibition of DNases protectscirculating cell-free DNA from ex vivo degradation in blood samples. Clin Biochem 48, 976–981 (2015). 8.Cristiano, S. et al. Genome-wide cell-free DNA fragmentation inpatients with cancer. Nature. 570, 385–389 (2019). 9.Snyder, M. W., Kircher, M., Hill, A. J., Daza, R. M.&Shendure, J.Cell-free DNA Comprises an In Vivo Nucleosome Footprint that Informs ItsTissues-Of-Origin. Cell 164, 57–68 (2016). 10.Foda, Z. H. et al. Detecting liver cancer using cell-free DNAfragmentomes. Cancer Discov 13, 616–631 (2022). 11.Moss, J. et al. Comprehensive human cell-type methylation atlasreveals origins of circulating cell-free DNA in health and disease. Nat Commun 9, 5068 (2018). 12.Fortin, J.-P.&Hansen, K. D. Reconstructing A / B compartments asrevealed by Hi-C using long-range correlations in epigenetic data. Genome Biol 16, 180 (2015). 13.Collings, C. K., Waddell, P. J.&Anderson, J. N. Effects of DNAmethylation on nucleosome stability. Nucleic Acids Res 41, 2918–2931 (2013). 14.Keshet, I., Yisraeli, J.&Cedar, H. Effect of regional DNAmethylation on gene expression. Proc National Acad Sci 82, 2560–2564 (1985). 15.Ulz, P. et al.Inferring expressed genes by whole-genome sequencingof plasma DNA. Nat Genet 48, 1273–1278 (2016). 16.Esfahani, M. S. et al. Inferring gene expression from cell-free DNAfragmentation profiles. Nat Biotechnol 40, 585–597 (2022). 17.Zhou, Q. et al. Epigenetic analysis of cell-free DNA byfragmentomic profiling. Proc National Acad Sci 119, e2209852119 (2022). 18.Chan, K. C. A. et al. Second generation noninvasive fetal genomeanalysis reveals de novo mutations, single-base parental inheritance, andpreferred DNA ends. Proc National Acad Sci 113, E8159–E8168 (2016). 19.Serpas, L. et al. Dnase1l3 deletion causes aberrations in lengthand end-motif frequencies in plasma DNA. Proc National Acad Sci 116, 641–649(2019). 20.Trifonov, E. N. Cracking the chromatin code: Precise rule ofnucleosome positioning. Phys Life Rev 8, 39–50 (2011). 21.Norris, D. P., Brockdorff, N.&Rastan, S. Methylation status ofCpG-rich islands on active and inactive mouse X chromosomes. Mamm Genome 1,78–83 (1991). 22.Tribioli, C. et al. Methylation and sequence analysis around Eagisites: identification of 28 new CpG islands in XQ24-XQ28. Nucleic Acids Res 20, 727–733 (1992). 23.Duncan, C. G. et al. Dosage compensation and DNA methylationlandscape of the X chromosome in mouse liver. Sci Rep-uk 8, 10138 (2018). 24.Jiang, P. et al. Lengthening and shortening of plasma DNA inhepatocellular carcinoma patients. Proc National Acad Sci 112, E1317–E1325(2015). 25.Giacona, M. B. et al. Cell-Free DNA in Human Blood Plasma. Pancreas 17, 89–97 (1998). 26.Jones, P. A.&Baylin, S. B. The fundamental role of epigeneticevents in cancer. Nat Rev Genet 3, 415–428 (2002). 27.Phallen, J. et al. Direct detection of early-stage cancers usingcirculating tumor DNA. Sci Transl Med9, eaan2415 (2017). 28.Baylin, S. B. et al. Abnormal patterns of DNA methylation in humanneoplasia: potential consequences for tumor progression. Cancer Cells Cold Spring Harb, NY, 1989 3, 383–90 (1991). 29.Gama-Sosa, M. A. et al. The 5-methylcytosine content of DNA fromhuman tumors. Nucleic Acids Res 11, 6883–6894 (1983). 30.Shen, S. Y. et al. Sensitive tumour detection and classificationusing plasma cell-free DNA methylomes. Nature 563, 579–583 (2018). 31.Li, S., Peng, Y., Landsman, D.&Panchenko, A. R. DNA methylationcues in nucleosome geometry, stability and unwrapping. Nucleic Acids Res 50,1864–1874 (2022). 32.Mathios, D. et al. Detection and characterization of lung cancerusing cell-free DNA fragmentomes. Nat Commun 12, 5060 (2021). 33.Uhlén, M. et al. Tissue-based map of the human proteome. Science 347, 1260419 (2015). 34.Chen, S., Zhou, Y., Chen, Y.&Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018). 35.Langmead, B.&Salzberg, S. L. Fast gapped-read alignment withBowtie 2. Nat Methods 9, 357–359 (2012). 36.Tarasov, A., Vilella, A. J., Cuppen, E., Nijman, I. J.&Prins, P.Sambamba: fast processing of NGS alignment formats. Bioinformatics 31, 2032–2034 (2015). 37.Quinlan, A. R.&Hall, I. M. BEDTools: a flexible suite of utilitiesfor comparing genomic features. Bioinformatics 26, 841–842 (2010). 38.Papp, E. et al. Integrated Genomic, Epigenomic, and ExpressionAnalyses of Ovarian Cancer Cell Lines. Cell Reports 25, 2617–2633 (2018). 39.Aryee, M. J. et al. Minfi: a flexible and comprehensiveBioconductor package for the analysis of Infinium DNA methylationmicroarrays. Bioinformatics 30, 1363–1369 (2014). 40.Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc National Acad Sci 102, 15545–15550 (2005). 41. Liberzon, A. et al. The Molecular Signatures Database Hallmark GeneSet Collection. Cell Syst 1, 417–425 (2015). 42. Kanehisa, M. & Goto, S. Kegg: Kyoto Encyclopedia of Genes andGenomes. Nucleic Acids Res 28, 27–30 (2000). 43. Liberzon, A. et al. Molecular signatures database (MSigDB)3.0. Bioinformatics 27, 1739–1740 (2011). 44. Adhireksan, Z. et al. Engineering nucleosomes for generating diverse chromatin assemblies. Nucleic Acids Res 49, gkab070 (2021). Other implementation plans

[0147] Based on the above description, it is evident that various changes and modifications can be made to the disclosure described herein to suit different uses and conditions. Such embodiments are also within the scope of the appended claims.

[0148] All references to sequences, patents, and publications in this specification are incorporated herein by reference to the extent that each individual patent and publication is specifically and individually identified and incorporated by reference. By citing various references herein, the applicant does not acknowledge any particular reference as "prior art" to its disclosure.

Claims

1. A method for determining circulating cell-free DNA (cfDNA) fragmentation in a sample, comprising: determining cfDNA sequences to identify genomic locations and end positions of cfDNA fragments; analyzing frequencies of cfDNA breakpoints at multiple positions in the genomic sequence; thereby, determining circulating cell-free DNA (cfDNA) fragmentation.

2. The method of claim 1, wherein the genomic sequence is determined by whole genome sequencing, or by obtaining a whole genome sequence from a database and pooling cfDNA sequences.

3. The method of claim 1, wherein the analysis of frequencies of cfDNA breakpoints comprises calculating a ratio of the number of cfDNA fragments that start at a particular position or the number of cfDNA fragments that end at a particular position relative to the number of fragments that have the particular position within 50 bp as a start or end position.

4. The method of claim 3, wherein cfDNA fragments that contain similar end position sequences have similar motifs.

5. The method of claim 4, wherein the motifs comprise a thymine or adenine preceding the start of the cfDNA fragment sequence, and two cytosines (A / T|CC) or one cytosine followed by one guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence, or wherein these sequences are the reverse complements at the end of the cfDNA fragment.

6. The method of claim 5, wherein the frequency of the motifs is increased or decreased in healthy subjects relative to subjects with cancer.

7. The method of claim 6, wherein the frequency of the A / T|CC motif is greater than the frequency of the A / T|CG motif.

8. The method of claim 7, wherein the A / T|CG motif is proximal to a histone H1 linker, or is separated from the histone H1 linker by between 100 base pairs and 200 base pairs.

9. The method of any one of claims 1-8, wherein internal regions of cfDNA fragments are enriched for adenine and thymine.

10. The method of claim 1, further comprising mapping cfDNA fragments to the genome and comparing the cfDNA fragment end sequences to methylated and non-methylated CpG sites from cfDNA of healthy subjects.

11. The method of claim 10, wherein methylated CpGs are enriched at the end of A / T|CG cfDNA fragment sequences.

12. The method of claim 11, wherein quantitative assessment of cfDNA fragment end positions at CpG enrichment comprises calculating for each CpG a fraction of cfDNA fragments that start at the CpG dinucleotide and a fraction of cfDNA fragments that end at the CpG dinucleotide divided by the number of cfDNA fragments that have the CpG within 50 bp as a start or end position.

13. The method of any one of claims 1-12, wherein methylated CpGs are reduced at the end of cfDNA fragments that start with a CCG trinucleotide (containing a CpG at positions 2 and 3).

14. The method according to any one of claims 1 to 13, wherein the quantitative assessment of the reduction of cfDNA fragment ends ending in CCG comprises calculating for each CCG the fraction of cfDNA fragments ending in the CCG trinucleotide over the number of cfDNA fragments in the region surrounding the first cytosine of the CCG trinucleotide (preferably 50 bp upstream or downstream of the particular position); a similar quantitative assessment of the reduction of cfDNA fragment ends ending in the reverse complement of the CCG trinucleotide (CGG) can be done by calculating the fraction of cfDNA fragments ending in the CGG trinucleotide over the number of cfDNA fragments in the region surrounding the last nucleotide of the CGG trinucleotide.

15. The method according to any one of claims 1 to 14, wherein cfDNA fragments obtained from the X chromosome of healthy subjects comprise an increased number of cfDNA fragments ending in CG and a reduced number of cfDNA fragments ending in CCG at the position of the X chromosome CpG island in female subjects relative to male subjects.

16. The method according to any one of claims 1 to 15, wherein the increased cfDNA fragment sequence coverage comprises increased methylation at the region of the methylated CpG island.

17. The method according to any one of claims 1 to 16, wherein gene expression at a transcription start site (TSS) is negatively correlated with cfDNA coverage at the TSS.

18. The method according to any one of claims 1 to 17, wherein reduced CpG island methylation is associated with an increased number of smaller cfDNA fragments in the region surrounding the CpG island.

19. The method according to any one of claims 1 to 18, wherein increased gene expression at a transcription start site is associated with an increased number of smaller cfDNA fragments in the region surrounding the transcription start site.

20. The method according to any one of claims 1 to 19, further comprising treating the subject with one or more chemotherapeutic agents, radiation therapy, surgery, and combinations thereof.

21. A method of diagnosing cancer and treating a subject, comprising: obtaining a sample from the subject; determining changes in circulating cell-free DNA (cfDNA) fragment size relative to normal and tumor-derived cfDNA controls; determining CpG methylation on cfDNA fragment ends via low-coverage whole genome cfDNA analysis by assessing cfDNA fragment ends at CG and CCG sites; diagnosing the subject with cancer; and treating the subject with one or more chemotherapeutic agents, radiation therapy, surgery, and combinations thereof.

22. The method according to claim 21, wherein an increase in cfDNA fragments ending in N|CCG in regions of reduced methylation in cancer compared to cfDNA from healthy individuals is diagnostic of cancer.

23. The method according to claim 21 or 22, wherein a decrease in cfDNA fragments ending in N|CCG in regions of increased methylation in cancer compared to cfDNA from healthy individuals is diagnostic of cancer.

24. The method of any one of claims 21 to 23, wherein an increase in cfDNA fragments ending in CG in a region of increased methylation in cancer compared to cfDNA from a healthy individual is diagnostic of cancer.

25. The method of any one of claims 21 to 24, wherein a decrease in cfDNA fragments ending in CG in a region of decreased methylation in cancer compared to cfDNA from a healthy individual is diagnostic of cancer.

26. The method of any one of claims 21 to 25, further comprising incorporating the distribution of fragment end positions at CG and CCG sites into a gradient boosting tree machine learning model.

27. The method of any one of claims 21 to 26, further comprising incorporating the distribution of fragment end positions at CG and CCG sites and fragment size distribution into a gradient boosting tree machine learning model.

28. A method of diagnosing and treating a subject diagnosed with cancer, comprising: determining a genomic sequence to identify genome-wide CpG methylation; analyzing the frequency of cfDNA breakpoint at a plurality of positions in the genomic sequence and determining circulating cell-free DNA (cfDNA) fragmentation, wherein a recurring cfDNA fragment end at a CpG site is enriched with higher genome-wide methylation level and smaller cfDNA fragments associated with cancer diagnosis; diagnosing the subject with cancer if hypomethylation and / or increased gene expression or decrease in cfDNA fragment size is detected; and treating the subject with one or more chemotherapeutic agents, radiation therapy, surgery, or a combination thereof.

29. The method of claim 28, wherein the genomic sequence is determined by whole genome sequencing or obtaining a whole genome sequence from a database and pooling cfDNA sequences.

30. The method of claim 29, wherein the analysis of the frequency of cfDNA breakpoints comprises calculating the ratio of the number of cfDNA fragments that start or end at a particular position relative to the number of fragments that are start or end positions within 50 bp around the particular position.

31. The method of any one of claims 28 to 30, wherein cfDNA fragments comprising similar end position sequences comprise similar motifs.

32. The method of claim 31, wherein the motif comprises a thymine or adenine preceding the start of the cfDNA fragment sequence and two cytosines (A / T|CC) or one cytosine followed by one guanine (A / T|CG) as the first two nucleotides of the cfDNA fragment sequence.

33. The method of any one of claims 28 to 32, wherein the frequency of the motif is increased in healthy subjects compared to subjects with cancer.

34. The method of claim 33, wherein the frequency of the A / T|CC motif is greater than the frequency of the A / T|CG motif.

35. The method of claim 33 or 34, wherein the A / T|CG motif is located proximal to a histone H1 linker or is concentrated between 100 base pairs to 200 base pairs from a histone H1 linker.

36. The method of any one of claims 28 to 35, wherein the internal regions of cfDNA fragments are enriched for adenine and thymine.

37. The method of any one of claims 28 to 36, further comprising mapping cfDNA fragments to the genome and comparing cfDNA fragment end sequences to methylated and unmethylated CpG sites from cfDNA of healthy subjects.

38. The method of claim 37, wherein methylated CpGs are enriched at the ends of A / T|CG cfDNA fragment sequences.

39. The method of any one of claims 28 to 38, wherein quantitative assessment of enrichment of cfDNA fragment ends at CpGs comprises, for each CpG, calculating the fraction of cfDNA fragments that begin or end at the CpG dinucleotide position divided by the number of cfDNA fragments within 50 bp around each CpG that have a beginning or ending position.

40. The method of any one of claims 28 to 39, wherein cfDNA fragments obtained from the X chromosome in healthy subjects comprise an increase in cfDNA fragments ending in CG, and a decrease in cfDNA fragments ending in CCG at X chromosome CpG island positions in female subjects compared to male subjects.

41. The method of any one of claims 28 to 40, wherein cfDNA sequence coverage of enriched cfDNA fragment end sequences comprises an increase in methylation at methylated CpG island regions compared to a decrease or lower occurrence of cfDNA fragment end sequences.

42. The method of any one of claims 28 to 41, wherein gene expression at a transcription start site (TSS) is negatively correlated with cfDNA coverage at the TSS.

Images