Methods for multimodal epigenetic sequencing assays

Abstract

In some aspects, methods are provided herein that involve epigenetic signatures, including any of the following features: methylation profile, nucleosome dynamics profile, or fragmentation profile, or any combination thereof.In other aspects, the present disclosure is directed to methods that involve epigenetic signatures (epigenetic signature determination methods, epigenetic signature discovery methods, diagnostic methods, and therapeutic methods), as well as systems, kits, and components that are useful therefor.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Provisional Patent Application No. 63 / 316,277, filed March 3, 2022, the contents of which are incorporated by reference herein in their entirety.

[0002] Technical Field The present disclosure is directed in some aspects to a multimodal epigenetic signature that includes features of any of a methylation profile, a nucleosome dynamics profile, or a fragmentation profile, or any combination thereof. In other aspects, the present disclosure is directed to methods involving said epigenetic signature, as well as systems, kits, and components useful therefor. [Background technology]

[0003] background Techniques for non-invasively detecting an individual's biological state, such as a disease state and / or response to treatment, are highly desirable. Under normal conditions, nucleic acids are shed into the systemic circulation, for example, via apoptosis, and circulate as cell-free nucleic acids, such as cell-free DNA (cfDNA). Nucleic acids may also be shed into the systemic circulation due to or derived from abnormal cells, such as cancer cells. cfDNA has been a source of biological material that can be obtained non-invasively to study an individual's biological state. However, it remains a great challenge to identify relevant and robust cfDNA markers for detecting an individual's biological state. Summary of the Invention

[0004] Quick Overview In some aspects, provided herein is a method of determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual to determine an epigenetic signature, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows.

[0005] In some embodiments, provided herein is a method of determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a methylation sequencing technique performed on the sample obtained from the individual to determine an epigenetic signature, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows.

[0006] In some aspects, provided herein is a method of generating an epigenetic signature from a sample obtained from an individual, the method comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; extracting features from the sequencing data, the features comprising information from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate an epigenetic signature based on the plurality of features; and outputting the generated epigenetic signature.

[0007] In some embodiments, provided herein is a method of generating an epigenetic signature from a sample obtained from an individual, the method comprising: receiving sequencing data obtained from a methylation sequencing technique performed on the sample obtained from the individual; extracting features from the sequencing data, the features comprising information from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate an epigenetic signature based on the plurality of features; and outputting the generated epigenetic signature.

[0008] In some aspects, provided herein is a method of diagnosing a disease in an individual, the method comprising: determining an epigenetic signature from data obtained from a non-destructive methylation sequencing technique performed on a sample obtained from the individual, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; and comparing the epigenetic signature to a disease epigenetic signature and diagnosing the individual's disease based on the epigenetic signature.

[0009] In some embodiments, provided herein is a method of diagnosing a disease in an individual, the method comprising: determining an epigenetic signature from data obtained from a methylation sequencing technique performed on a sample obtained from the individual, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; and comparing the epigenetic signature to a disease epigenetic signature and diagnosing the disease in the individual based on the epigenetic signature.

[0010] In some embodiments, the method further comprises diagnosing a disease in an individual based on the epigenetic signature compared to a disease epigenetic signature. In some embodiments, methods of treating a disease in an individual are provided herein, the methods comprising diagnosing the individual as having a disease according to the methods of diagnosing a disease in an individual provided herein, and administering an agent to treat the disease in the individual.

[0011] In some aspects, provided herein is a method of identifying a disease epigenetic signature indicative of an individual having a disease, the method comprising receiving sequencing data from a plurality of individuals having the disease and a plurality of individuals not having the disease, the sequencing data being obtained from a non-destructive methylation sequencing technique performed on samples obtained from the individuals; and extracting features from the sequencing data, the features being one of the following profiles: a methylation profile comprising information derived from one or more methylation sites; (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome occupancy. The method includes extracting features from two or more of: a nucleosome dynamics profile comprising information derived from one or more of nucleosome fuzziness, or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model, wherein the extracted features from each of a plurality of individuals are populated with an associated classification of the individual having or not having the disease; training the machine learning model to identify a disease epigenetic signature using the extracted features; and outputting the disease epigenetic signature.

[0012] In some aspects, provided herein is a method of identifying a disease epigenetic signature indicative of an individual having a disease, the method comprising receiving sequencing data from a plurality of individuals with the disease and a plurality of individuals without the disease, the sequencing data being obtained from a methylation sequencing technique performed on samples obtained from the individuals; and extracting features from the sequencing data, the features being one of the following profiles: a methylation profile comprising information derived from one or more methylation sites; (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome location information. The method includes extracting features from two or more of: a nucleosome dynamics profile comprising information derived from one or more of a nucleosome fuzziness, or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model, wherein the extracted features from each of a plurality of individuals are populated with an associated classification of the individual as having or not having the disease; training the machine learning model to identify a disease epigenetic signature using the extracted features; and outputting the disease epigenetic signature.

[0013] In some embodiments, each of the one or more methylation sites in the methylation profile is selected from the group consisting of: cg18081940, cg23089825, cg16395183, cg19811148, cg07790615, cg20996351, cg04977528, cg24465685, cg20428713, cg13678973, cg25339566, cg16596317, cg23786625, cg11328303, cg19578660, cg02272851, cg10298052, cg13585930, cg23575688 , cg12394201, cg08149193, cg18854419, cg07603330, cg10658542, cg1309 9890, cg22302985, cg13596497, cg14507533, cg25366582, cg22396555, cg1 0566012, cg05168229, cg10795666, cg25078444, cg16038120, cg23883632 , cg18380808, cg13615592, cg00250422, cg19691260, cg16558770, cg15681 853, cg03397724, cg10514097, cg06674117, cg16047279, cg12127472, cg0 8843809, cg08697732, cg06384763, cg04203646, cg17112426, cg08278741 , cg14587524, cg26087117, cg18320766, cg08063125, cg10004780, cg1892 1980, cg02514318, cg20002504, cg18897632, cg15313459, cg19370054, cg1 6564824, cg02631468, cg01471196, cg23770904, cg18412834, cg24080247 , cg11549874, cg13155421, cg19442495, cg22536150, cg05413061, cg23346 462, cg09477895, cg13605674, cg13314965, cg09417547, cg00181669, cg2 3967169, cg10237419, cg21077559, cg27600205, cg19755714, cg18797590,cg00699993, cg06485940, cg27661394, cg00939495, cg11036833, cg23915769, cg0 7224726, cg02022733, cg03640756, cg15361590, cg04598517, cg06782035, cg1395 4457, cg25482900, cg20952257, cg14062050, cg01881524, cg11538641, cg1138734 0, cg05389236, cg19419054, cg10575547, cg17240815, cg24772267, cg00920327, c g00772257, cg26253500, cg23244488, cg22778435, cg26065247, cg02088996, cg19 868631, cg22280038, cg07803375, cg20230721, cg03333330, cg21517947, cg10406 295, cg05166490, cg07739205, cg20980783, cg06617456, cg01568998, cg13407456, cg23758305, cg20675505, cg07585876, cg03734437, and cg13410764.

[0014] In some embodiments, the one or more methylation sites of the methylation profile comprise methylation sites in one or more gene promoter regions.

[0015] In some embodiments, the methylation profile comprises quantitative information from at least one of the one or more methylation sites. In some embodiments, the quantitative information is based on the β value from at least one methylation site. In some embodiments, the quantitative information is based on the CHALM ratio from at least one methylation site.

[0016] In some embodiments, the nucleosome dynamics information is based on nucleosomes at genomic loci. In some embodiments, the nucleosome position information is based on a window protection score (WPS). In some embodiments, the WPS is an average WPS. In some embodiments, the nucleosome occupancy is based on the frequency with which nucleosomes occupy a genomic region. In some embodiments, the nucleosome occupancy is obtained by normalized read coverage measured in counts per million. In some embodiments, the nucleosome fuzziness is based on deviations of nucleosome positions from preferred nucleosome positions. In some embodiments, the fragmentation profile is based on one or more base length windows occupying a range of 30-250 bases in length. In some embodiments, the base length window is at least 10 bases in length. In some embodiments, the nucleosome dynamics information is obtained via DANPOS.

[0017] In some embodiments, the epigenetic signature indicates whether an individual has a disease. In some embodiments, the epigenetic signature comprises features from a methylation profile and a nucleosome dynamics profile. In some embodiments, the epigenetic signature comprises features from a methylation profile and a fragmentation profile. In some embodiments, the epigenetic signature comprises features from a nucleosome dynamics profile and a fragmentation profile. In some embodiments, the epigenetic signature comprises features from a methylation profile, a nucleosome dynamics profile, and a fragmentation profile. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome position information. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome occupancy. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome fuzziness.

[0018] In some embodiments, the non-destructive methylation sequencing technique is an EM-seq technique. In some embodiments, the non-destructive methylation sequencing technique is performed based on the target gene position. In some embodiments, the method further comprises performing the non-destructive methylation sequencing technique.

[0019] In some embodiments, the data obtained from the non-destructive methylation sequencing technique comprises a plurality of sequence reads. In some embodiments, the method further comprises processing the plurality of sequence reads to remove low quality reads and / or remove adapter contamination and / or filter based on the size of the sequence reads. In some embodiments, the method further comprises aligning the plurality of sequence reads to a reference genome.

[0020] In some embodiments, the machine learning model comprises a support vector machine model, a random forest machine model, or a logistic regression machine model. In some embodiments, the method further comprises a cross-validation procedure.

[0021] In some embodiments, the sample is a cell-free DNA sample. In some embodiments, the method further comprises obtaining the sample.

[0022] In some embodiments, the disease is cancer. In some embodiments, the cancer is colorectal cancer. In some embodiments, the individual is a human. In some embodiments, the individual is suspected of having the disease. [Brief description of the drawings]

[0023] [Figure 1] 1 shows a schematic diagram of an exemplary workflow 100 of certain methods provided herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] Detailed Description In certain aspects, provided herein are multimodal epigenetic signatures that include features derived from any combination of two or more of methylation profiles, nucleosome dynamics profiles (including any features thereof, such as nucleosome position information, nucleosome occupancy, and nucleosome fuzziness), and fragmentation profiles, as well as methods of multimodal use thereof. The present disclosure of the present application is based on the inventors' unique perspective and unexpected findings regarding multimodal analysis that provides significant improvements in determining an individual's condition, such as a disease state, using the epigenetic signatures and methods taught herein. Specifically, the inventors have developed flexible methods for using non-destructive methylation sequencing techniques to obtain information to generate any combination of methylation profiles, nucleosome dynamics profiles, and fragmentation profiles from a single assay. In combination with machine learning techniques, the description herein provides an unexpectedly flexible, accurate, sensitive, and robust measure of an individual's biological state. For example, the inventors have demonstrated that an epigenetic signature comprising a methylation profile and a nucleosome dynamics profile significantly improves the sensitivity of colon cancer detection (see Example 1). The flexibility offered by the multimodal epigenetic signatures provided herein allows such discoveries to be extended to a wide variety of human diseases with distinct epigenetic footprints.

[0025] Thus, in some aspects, provided herein is a method for determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual to determine an epigenetic signature, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows.

[0026] In some aspects, provided herein are methods for determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a methylation sequencing technique performed on the sample obtained from the individual to determine an epigenetic signature, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows.

[0027] In some aspects, provided herein are methods for determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a non-destructive methylation sequencing technique and one or more additional sequencing techniques (e.g., deep sequencing) performed on the sample obtained from the individual to determine an epigenetic signature, wherein the epigenetic signature comprises features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows.

[0028] In another aspect, provided herein is a method for generating an epigenetic signature from a sample obtained from an individual, the method comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; extracting features from the sequencing data, the features comprising information from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate an epigenetic signature based on the plurality of features; and outputting the generated epigenetic signature.

[0029] In another aspect, provided herein is a method for diagnosing a disease in an individual, the method comprising: determining an epigenetic signature from data obtained from a non-destructive methylation sequencing technique performed on a sample obtained from the individual, the epigenetic signature comprising features obtained from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution in one or more base length windows; and diagnosing the disease in the individual based on the epigenetic signature compared to the disease epigenetic signature. In some embodiments, the method further comprises diagnosing the disease in the individual based on the epigenetic signature compared to the disease epigenetic signature.

[0030] In another aspect, provided herein is a method of treating a disease in an individual, the method comprising diagnosing the individual as having a disease as described in any claim herein and administering an agent to treat the disease in the individual.

[0031] In another aspect, provided herein is a method for identifying a disease epigenetic signature indicative of an individual having a disease, the method comprising receiving sequencing data from a plurality of individuals having the disease and a plurality of individuals not having the disease, the sequencing data being obtained from a non-destructive methylation sequencing technique performed on samples obtained from the individuals; and extracting features from the sequencing data, the features being one of the following profiles: a methylation profile comprising information derived from one or more methylation sites; (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome occupancy. The method includes extracting features from two or more of: a nucleosome dynamics profile comprising information derived from one or more of nucleosome fuzziness, or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model, wherein the extracted features from each of a plurality of individuals are populated with an associated classification of the individual having or not having the disease; training the machine learning model to identify a disease epigenetic signature using the extracted features; and outputting the disease epigenetic signature.

[0032] A.Definition Unless otherwise defined, all technical terms, notations, and other technical and scientific terms or terminology used herein are intended to have the same meaning as commonly understood by one of ordinary skill in the art to which the subject matter of the claims pertains. In some cases, terms having a commonly understood meaning are defined herein for clarity and / or ready reference, and the inclusion of such definitions herein should not necessarily be construed as representing a substantial difference to what is commonly understood in the art.

[0033] Throughout this disclosure, various aspects of the claimed subject matter are presented in a range format. It should be understood that the description in range format is for convenience and brevity only and should not be construed as an inflexible limitation on the scope of the claimed subject matter. Thus, the description of a range should be considered to have specifically disclosed all possible subranges as well as individual numerical values ​​within that range. For example, when a range of values ​​is provided, unless the context clearly dictates otherwise, it is understood that each intervening value, to the tenth of the unit of the lower limit between the upper and lower limits of that range, and any other stated or intervening value within that stated range, is encompassed within the disclosure, subject to any specifically excluded limit within the stated range. When a stated range includes one or both limits, ranges excluding either or both of those included limits are also included in the disclosure. In some embodiments, two opposing open-ended ranges are provided for a feature, and such descriptions contemplate that a combination of these two ranges is provided herein. For example, in some embodiments, a feature is described as being greater than about 10 units and then (in a separate sentence, etc.) the feature is described as being less than about 20 units, and thus a range of about 10 units to about 20 units is described herein.

[0034] As used herein, the term "about" refers to the normal error range of each value, which is readily known in the art. Reference herein to "about" a value or parameter includes (and describes) the variation of the value or parameter itself. For example, a description that refers to "about X" includes the description of "X".

[0035] As used herein, including the appended claims, the singular forms "a," "or," and "the" include plural referents unless the context clearly dictates otherwise. For example, "a" or "an" means "at least one" or "one or more." It is understood that aspects and variations described herein include embodiments that "consist of" and / or "consist essentially of" such aspects and variations.

[0036] As used herein, a "subject" or an "individual" are used interchangeably and are a mammal. In some embodiments, a "mammal" includes humans, non-human primates, farm and zoo animals, sport animals, or pets such as dogs, rabbits, cows, pigs, hamsters, gerbils, ferrets, mice, cats, monkeys, etc. In some embodiments, the subject or individual is a human.

[0037] As used herein, "treatment" or "treating" refers to an approach to obtain beneficial or desired results, including clinical results. For purposes of the present invention, beneficial or desired clinical results include, but are not limited to, alleviating one or more symptoms caused by a disease, reducing the extent of the disease, stabilizing the disease (e.g., preventing or slowing the worsening of the disease), preventing or slowing the spread of the disease, preventing or slowing the recurrence of the disease, slowing or slowing the progression of the disease, improving the condition, providing remission (partial or total) of the disease, reducing the dosage of one or more other therapeutic agents required to treat the disease, slowing the progression of the disease, improving quality of life, and / or prolonging survival. Reduction of the pathological consequences of the disease is also encompassed by "treatment".

[0038] The method of the present invention contemplates any one or more of these therapeutic aspects. Those skilled in the art will recognize that several embodiments are possible within the scope and spirit of the present disclosure. The following description is illustrative of the present disclosure and, of course, should not be construed as limiting the scope of the invention described herein in any way.

[0039] B. Multimodal Epigenetic Signatures and Related Methods Provided herein In certain aspects, methods related to multimodal epigenetic signatures are provided herein that include features derived from any combination of two or more of the following profiles taught herein: a methylation profile that includes information derived from one or more methylation sites; a nucleosome dynamics profile that includes information derived from one or more of (a) nucleosome occupancy information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile that includes information derived from read distribution in one or more base length windows. In some embodiments, the term multimodal as used herein refers to a combination of two or more different profiles including a methylation profile, a nucleosome dynamics profile, and a fragmentation profile in the described methods and epigenetic signatures. Two or more different profiles may be combined to improve the technique, such as by machine learning techniques and cross-validation.

[0040] To illustrate the description provided herein, a schematic diagram of an exemplary workflow 100 is provided in FIG. 1. As shown, in some embodiments, the workflow 100 starts with a cell-free DNA (cfDNA) sample 102. Such a sample may be obtained from a blood sample obtained from an individual, such as an individual being evaluated for the absence of disease, and further sample processing may be performed to obtain or study a cfDNA sample. The cfDNA sample is then subjected to a non-destructive methylation sequencing technique 104, such as EM-seq. The sequencing information obtained from the non-destructive methylation sequencing technique 104 can then be analyzed based on any configuration of desired multimodal features 106, including any of a methylation profile, a nucleosome dynamics profile, and a fragmentation profile. As shown, the nucleosome dynamics profile may include information derived from any one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness. Feature identification and evaluation may be performed using a combined predictive model 108 using information obtained from a single assay (i.e., a single non-destructive methylation sequencing technique performed on a cfDNA sample) to determine an epigenetic signature 110. In some embodiments, the workflow 100 is configured for the discovery of multimodal epigenetic signatures. In some embodiments, the workflow 100 is configured for the assessment of multimodal epigenetic signatures in a sample from an individual, such as for the diagnosis of a disease such as cancer.

[0041] The multimodal epigenetic signatures taught herein provide insightful information regarding an individual's biological state, such as disease state and / or response to treatment, and can be used in a wide variety of ways. In certain aspects, methods of determining epigenetic signatures are provided herein. In other aspects, methods of generating epigenetic signatures using machine learning models are provided herein. In other aspects, methods of diagnosing disease in an individual using epigenetic signatures are provided herein. In other aspects, methods of treating disease in an individual, including diagnosing disease in an individual using epigenetic signatures are provided herein. In certain aspects, methods of identifying disease epigenetic signatures in an individual, including training a machine learning model to identify disease epigenetic signatures, are provided herein.

[0042] The multimodal epigenetic signatures provided herein may include information derived from any combination of two or more of a methylation profile, a nucleosome dynamics profile, and a fragmentation profile. As described herein, the methylation profile includes information derived from one or more methylation sites. The nucleosome dynamics profile includes information derived from any one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness. The fragmentation profile includes information derived from read distribution of one or more base length windows. In some embodiments, the epigenetic signature includes features from the methylation profile and the nucleosome dynamics profile (including features from any one or combination of nucleosome position information, nucleosome occupancy, or nucleosome fuzziness). In some embodiments, the epigenetic signature includes features from the methylation profile and the fragmentation profile. In some embodiments, the epigenetic signature comprises features from a nucleosome dynamics profile (including features from any or a combination of nucleosome position information, nucleosome occupancy, or nucleosome fuzziness) and a fragmentation profile. In some embodiments, the epigenetic signature comprises features from a methylation profile, a nucleosome dynamics profile (including features from any or a combination of nucleosome position information, nucleosome occupancy, or nucleosome fuzziness), and a fragmentation profile. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome position information. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome occupancy. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome fuzziness. In some embodiments, the epigenetic signature indicates whether an individual has a disease.

[0043] The following sections provide additional descriptions of various aspects of the epigenetic signatures and related methods taught herein. Such descriptions in a modular manner are not intended to limit the scope of the present disclosure, and based on the teachings provided herein, those skilled in the art will readily understand that certain modules can be integrated at least in part. The section headings used herein are for organizational purposes only and should not be construed as limiting the subject matter described.

[0044] I. Non-destructive methylation sequencing techniques In certain aspects, the methods provided herein include the use of non-destructive methylation sequencing techniques and / or data obtained therefrom. In some embodiments, the non-destructive methylation sequencing techniques are configured to generate sequencing information, such as sequencing reads, suitable for use in determining one or more of a methylation profile, a nucleosome dynamics profile, or a fragmentation profile from a single assay. In some embodiments, the non-destructive methylation sequencing techniques include the use of enzymes to convert nucleobases, such as through deamination of unmethylated cytosine to uracil, such that the nucleobases can be distinguished from the sequencing information.

[0045] In some embodiments, the methods provided herein further comprise performing a non-destructive methylation sequencing technique. In some embodiments, the non-destructive methylation sequencing technique is an enzymatic methyl sequencing technique (EM-seq) technique. In some embodiments, the non-destructive methylation sequencing technique comprises (a) enzymatically modifying methylated cytosines (such as 5-methylcytosine (5mc) and 5-hydroxymethylcytosine (5hmC)) to prevent deamination in a further enzymatic step, (b) enzymatically converting unmethylated cytosines to uracil, (c) performing PCR amplification (thereby converting uracil to thymine), and (d) determining the sequence using a next-generation sequencing technique. Various techniques for performing non-destructive methylation sequencing techniques have been described in the art. See, for example, Vaisvila et al., Genome Res, 31, 2021, which is incorporated herein in its entirety. In some embodiments, enzymatically modifying methylated cytosines is performed using TET2 and / or T4-BGT. In some embodiments, the non-destructive methylation sequencing technique includes enzymatically converting unmethylated cytosines to uracils using APOBEC3A. In some embodiments, the non-destructive methylation sequencing technique includes subjecting a sample including genomic DNA, such as a cifDNA sample, to a next-generation sequencing library preparation technique. In some embodiments, the next-generation sequencing library preparation technique includes shearing the genomic DNA to obtain a DNA size of less than about 500 base pairs, such as less than about any of 450 base pairs, 400 base pairs, 350 base pairs, or 300 base pairs. In some embodiments, the next-generation sequencing library preparation technique includes a step of final preparation of the sheared DNA. In some embodiments, the next-generation sequencing library preparation technique includes a step of adaptor ligation. In some embodiments, the next-generation sequencing library preparation technique includes a step of cleaning up the adaptor-ligated DNA.In some embodiments, the cleaned and ligated DNA is subjected to an oxidizing enzyme, such as TET2 and / or T4-BGT, to modify methylated cytosines (5-methylcytosine and 5-hydroxymethylcytosine). In some embodiments, the next generation sequencing library preparation technique comprises a step of enzymatically cleaning the oxidized DNA. In some embodiments, the oxidized DNA is further subjected to enzymatic cytosine deamination (such as using APOBEC3A). In some embodiments, the next generation sequencing library preparation technique comprises a step of PCR amplification of the deaminated DNA. In some embodiments, the next generation sequencing library preparation technique comprises a step of sequencing and quantification. In some embodiments, the method comprises, for example, adding a control to the sample containing genomic DNA before performing any enzymatic conversion step.

[0046] In some embodiments, the non-destructive methylation sequencing technique is performed based on the target gene location. In some embodiments, the non-destructive methylation sequencing technique is performed across the entire genome.

[0047] In some embodiments, the data obtained from the non-destructive methylation sequencing technique comprises a plurality of sequence reads. In some embodiments, the non-destructive methylation sequencing technique is performed to a sequencing depth of about 50x to about 500x. In some embodiments, the non-destructive methylation sequencing technique is performed to a sequencing depth of at least about 50x, such as 75x, 100x, 125x, 150x, 175x, 200x, 225x, 250x, 275x, 300x, 325x, 350x, 375x, 400x, 425x, 450x, 475x, or 500x. In some embodiments, non-destructive methylation sequencing techniques are performed to a sequencing depth of about 50x, 75x, 100x, 125x, 150x, 175x, 200x, 225x, 250x, 275x, 300x, 325x, 350x, 375x, 400x, 425x, 450x, 475x, or 500x.

[0048] In some embodiments, the method further comprises processing the plurality of sequence reads to remove low quality reads and / or remove adapter contamination and / or filter based on size of the sequence reads. In some embodiments, the method further comprises aligning the plurality of sequence reads to a reference genome.

[0049] In some embodiments, the methods provided herein include a non-destructive methylation sequencing technique in combination with one or more additional sequencing techniques. In some embodiments, the one or more additional sequencing techniques include next-generation sequencing, such as deep sequencing, droplet digital PCR, and / or pyrosequencing. In some embodiments, the sequencing investigates DNA mutations (e.g., cfDNA mutations), RNA, micoRNA, or any combination thereof. For example, the method may include performing non-destructive methylation sequencing and deep sequencing (e.g., to evaluate mutations). In some embodiments, the method includes performing non-destructive methylation sequencing to obtain a methylation profile that includes information derived from one or more methylation sites, and performing another sequencing technique (e.g., deep sequencing) to obtain a nucleosome dynamics profile that includes information derived from any one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness. In some embodiments, the method includes performing non-destructive methylation sequencing to obtain a methylation profile including information derived from one or more methylation sites, and performing one or more additional sequencing techniques (e.g., deep sequencing) to obtain a fragmentation profile including information derived from read distribution in one or more base length windows. In some embodiments, the method includes performing non-destructive methylation sequencing to obtain a methylation profile including information derived from one or more methylation sites, performing one or more additional sequencing techniques (e.g., deep sequencing) to obtain a nucleosome dynamics profile including information derived from one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness, and performing another sequencing technique (e.g., deep sequencing) to obtain a fragmentation profile including information derived from read distribution in one or more base length windows.

[0050] Suitable sequencing techniques useful for the non-destructive methylation sequencing techniques described herein are well known in the art. In some embodiments, such sequencing techniques involve (i) amplification and detection, or (ii) direct detection, by a variety of methods, such as (a) PCR (sequence-specific amplification), such as TaqMan®, (b) DNA sequencing of untreated and treated DNA, (c) sequencing by ligation of dye-modified probes (including circular ligation and cleavage), (d) pyrosequencing, (e) single molecule sequencing, (f) mass spectrometry, or (g) Southern blot analysis.

[0051] In some embodiments, restriction enzyme digestion of PCR products amplified from enzymatically converted DNA may be used, for example the method described by Sadri and Hornsby (Sadri et al., 1996, Nucl. Acids Res. 24:5058-5059), or COBRA (combined bisulfite restriction analysis) (Xiong and Laird, 1997, Nucleic Acids Res. 25:2532-2534). COBRA analysis is a quantitative methylation assay useful for determining DNA methylation levels at specific loci in small amounts of genomic DNA. Briefly, restriction enzyme digestion is used to reveal methylation-dependent sequence differences in PCR products of enzymatically converted DNA. PCR amplification of the converted DNA is then performed using primers specific for the CpG sites of interest, followed by restriction endonuclease digestion, gel electrophoresis, and specifically labeled hybridization probes. The methylation level of the original DNA sample is expressed as the relative amounts of digested and undigested PCR products in a linearly quantitative manner across DNA methylation levels.

[0052] In some embodiments, the methylation profile of selected CpG sites is determined using methylation-specific PCR (MSP), which allows for the assessment of the methylation status of virtually any group of CpG sites within a CpG island, independent of the use of methylation-sensitive restriction enzymes (described in Herman et al., 1996, Proc. Nat. Acad. Sci. USA, 93, 9821-9826; U.S. Pat. No. 5,786,146; U.S. Pat. No. 6,017,704; U.S. Pat. No. 6,200,756; U.S. Pat. No. 6,265,171 (Herman and Baylin); U.S. Patent Publication No. 2010 / 0144836 (Van England et al.) which are incorporated herein by reference in their entirety). Briefly, DNA is enzymatically deaminated to convert unmethylated cytosines to uracils, but not methylated cytosines, and then amplified with primers specific for methylated versus unmethylated DNA. In some embodiments, typical reagents for MSP analysis (e.g., as may be found in a typical MSP-based kit) include, but are not limited to, methylated and unmethylated PCR primers for specific genes (or methylation-modified DNA sequences or CpG islands), optimized PCR buffers and deoxynucleotides, and specific probes. Quantitative multiplexed methylation-specific PCR (QM-PCR) as described by Fackler et al., 2004, Cancer Res. 64(13)4442-4452, or Fackler et al., 2006, Clin. Cancer Res.12(11Pt1)3306-3310 may be used.

[0053] In some embodiments, the non-destructive methylation sequencing technique comprises the MethyLight method and / or the heavy methyl method. The MethyLight assay and the heavy methyl assay are high-throughput quantitative methylation assays that utilize fluorescence-based real-time PCR (Taq Man®) technology that does not require further manipulation after the PCR step (Eads, CA et al., 2000, Nucleic Acid Res. 28,e 32; Cottrell et al., 2007, J. Urology 177, 1753; U.S. Patent No. 6,331,393 (Laird et al.), the contents of which are incorporated herein by reference in their entirety).

[0054] In some embodiments, the non-destructive methylation sequencing technique comprises the Ms-SNuPE technique, which is a quantitative method based on the enzymatic deamination of DNA to assess methylation differences at specific CpG sites, followed by single nucleotide primer extension (Gonzalgo and Jones, 1997, Nucleic Acids Res. 25, 2529-2531).

[0055] In some embodiments, a method is provided for quantifying the average methylation density of a target sequence in a population of genomic DNA. In some examples, a quantitative amplification method (e.g., quantitative PCR or quantitative linear amplification) is used. For quantitative amplification methods, see, for example, U.S. Patent Nos. 6,180,349, 6,033,854, and 5,972,602, as well as, for example, DeGraves, et al., 34(1)Biotechniques 106-15(2003), Deiman B, et al., 20(2)Mol. Biotechnol. 163-79(2002), and Gibson et al., 6 Genome Res. 995-1001(1996).

[0056] In some embodiments, the methods provided herein include sequence-based analysis. For example, when one particular gene sequence from a sample is determined to be hypermethylated or hypomethylated compared to its counterpart, the amount of genomic sequence can be determined. This amount can then be compared to a standard control value and used to determine the presence of liver cancer in the sample. In many cases, it is desirable to amplify the nucleic acid sequence using any of several nucleic acid amplification procedures well known in the art. Specifically, nucleic acid amplification is the chemical or enzymatic synthesis of a nucleic acid copy that contains a sequence that is complementary to the nucleic acid sequence (template) being amplified. The methods and kits may use any nucleic acid amplification or detection method known to those of skill in the art, such as those described in U.S. Pat. Nos. 5,525,462 (Takarada et al.), 6,114,117 (Hepp et al.), 6,127,120 (Graham et al.), 6,344,317 (Urnovitz), 6,448,001 (Oku), 6,528,632 (Catanzariti et al.), and PCT Publication No. WO2005 / 111209 (Nakajima et al.), all of which are incorporated herein by reference in their entireties.

[0057] In some embodiments, nucleic acid is amplified by PCR amplification using methodology known to those skilled in the art. However, those skilled in the art will recognize that amplification can be achieved by any known method, such as ligase chain reaction (LCR), Q-replication amplification, rolling circle amplification, transcription amplification, self-sustained sequence replication, nucleic acid sequence-based amplification (NASBA), each of which provides sufficient amplification. Branched DNA technology is also optionally used to qualitatively demonstrate the presence of the sequence of this technology that represents a specific methylation pattern, or to quantitatively determine the amount of this specific genomic sequence in a sample. Nolte has reviewed branched DNA signal amplification for direct quantification of nucleic acid sequences in clinical samples (Nolte, 1998, Adv. Clin. Chem. 33: 201-235).

[0058] PCR processes are well known in the art and include, for example, reverse transcription PCR, ligation-mediated PCR, digital PCR (dPCR), or droplet digital PCR (ddPCR). For a review of PCR methods and protocols, see, for example, Innis et al., eds., PCR Protocols, A Guide to Methods and Application, Academic Press, Inc., San Diego, Calif. 1990, U.S. Pat. No. 4,683,202 (Mullis). PCR reagents and protocols are also available from commercial vendors such as Roche Molecular Systems. In some instances, PCR is performed as an automated process using a thermostable enzyme. In this process, the temperature of the reaction mixture is automatically cycled through the denaturation, primer annealing, and extension reaction regions. Machines specially adapted for this purpose are commercially available.

[0059] Suitable next generation sequencing techniques are widely available. Examples are the 454 Life Sciences platform (Roche, Branford, CT) (Marguies et al., 2005 Nature, 437, 376-380), Illumina's Genome Analyzer, the GoldenGate Methylation Assay, or the Infinium Methylation Assays, i.e., the Infinium HumanMethylation 27K BeadArray or the VeraCode GoldenGate Methylation Array (Illumina, San Diego, CA; Bibkova et al., 2006, Genome Res. 16, 383-393; U.S. Patents 6,306,597 and 7,598,035 (Maceviz), 7,232,656 (Balasubramanian et al.)), Bio-Rad's QX200™ Droplet Digital™ PCR System, or DNA Sequencing by Ligation, the SOLiD System (Applied Biosystems / Life Sciences, Inc.). Technologies, U.S. Patents 6,797,470, 7,083,917, 7,166,434, 7,320,865, 7,332,285, 7,364,858, and 7,429,453 (Barany et al.), Helicos True Single Molecule DNA sequencing technology (Harris et al., 2008 Science, 320, 106-109, U.S. Patents 7,037,687 and 7,645,596 (Williams et al.), 7,169,560 (Lapidus et al.), 7,769,400 (Harris)), Pacific Biosciences' Single Molecule, Real-Time (SMRT™) technology, and sequencing (Soni and Meller, 2007, Clin. Chem.53, 1996-2001), semiconductor sequencing (Ion Torrent, Personal Genome Machine), DNA nanoball sequencing, technology from Dover Systems (Polonator), and nanopore-based strategies (e.g., Oxford Nanopore, Genia Technologies, and Nabsys), which do not require amplification of native DNA prior to sequencing or are otherwise non-transforming. These systems allow for high order multiplexing in parallel to sequence many nucleic acid molecules isolated from a specimen. Each of these platforms allows for clonally propagated or unamplified single molecule sequencing of nucleic acid fragments. Specific platforms include, for example, (i) sequencing by ligation of dye-modified probes (including circular ligation and cleavage), (ii) pyrosequencing, and (iii) single molecule sequencing.

[0060] In some embodiments, the analyzing comprises quantitatively detecting the methylation status of the amplification product. In some cases, the detecting comprises real-time quantitative probe-based PCR or digital probe-based PCR. In some cases, the detecting comprises real-time quantitative probe-based PCR. In other cases, the detecting comprises digital probe-based PCR, optionally digital droplet PCR.

[0061] II. Methylation Profile In certain aspects, the methods provided herein include a multimodal epigenetic signature comprising one or more features derived from a methylation profile. As described herein, the methylation profile is based on the presence and / or absence of methylation at one or more methylation sites. In some embodiments, the methylation profile comprises a qualitative feature of one or more methylation sites, such as the presence or absence of methylation at the methylation site. In some embodiments, the methylation profile comprises a quantitative feature of one or more methylation sites, such as obtained via beta values ​​and / or cell heterogeneity adjusted cLonal methylation (CHALM).

[0062] DNA methylation is the attachment of a methyl group at the C5 position of the nucleotide bases cytosine and at the N6 position of adenine. Adenine methylation occurs primarily in prokaryotes, while cytosine methylation occurs in both prokaryotes and eukaryotes. In some embodiments, cytosine methylation occurs at CpG dinucleotide motifs. In some embodiments, cytosine methylation occurs, for example, at CHG and CHH motifs, where H is adenine, cytosine, or thymine. In some embodiments, one or more CpG dinucleotide motifs or CpG sites form a CpG island, a short DNA sequence rich in CpG dinucleotides. In some embodiments, CpG islands are present in the 5' region of about half of all human genes. The length of a CpG island is usually, but not necessarily, about 0.2 to about 1 kb. Cytosine methylation further includes 5-methylcytosine (5-mCyt) and 5-hydroxymethylcytosine. CpG (cytosine-phosphate-guanine) or CG motif refers to a region of a DNA molecule in which a cytosine nucleotide occurs next to a guanine nucleotide in a linear strand. In some embodiments, the cytosine in the CpG dinucleotide is methylated to form a 5-methylcytosine. In some embodiments, the cytosine in the CpG dinucleotide is methylated to form a 5-hydroxymethylcytosine.

[0063] In some embodiments, one or more DNA regions, such as methylation sites, are hypermethylated. In such cases, hypermethylation refers to an increase in methylation events in a region compared to a reference region (such as another region of DNA or the same region in a control sample). In some cases, hypermethylation is observed in one or more cancer types and is useful, for example, as a diagnostic and / or prognostic marker. In some embodiments, one or more DNA regions are hypomethylated. In some embodiments, hypomethylation refers to a loss of a methyl group at a 5-methylcytosine nucleotide in a first region compared to a reference region (such as another region of DNA or the same region in a control sample). In some embodiments, hypomethylation is observed in one or more cancer types and is useful, for example, as a diagnostic and / or prognostic marker.

[0064] In some embodiments, methylation at one or more methylation sites is assessed using a non-destructive methylation sequencing technique, such as EM-seq, as described herein. In some embodiments, the methylation assessment methods encompassed herein include using a probe to assess methylation at a methylation site. In some embodiments, the methylation assessment methods encompassed herein include using a panel of probes to assess methylation at multiple methylation sites.

[0065] In some embodiments, one or more methylation sites of the methylation profile are any of those provided in the Illumina Infinium HumanMethylation450 BeadChip (450K) available at the time of filing of this application. The methylation markers included in the Illumina Infinium HumanMethylation450 BeadChip (450K) are known. See, for example, Wang et al., BMC Bioinformatics, 19, 2018, which is incorporated by reference in its entirety herein. In some embodiments, one or more methylation sites of the methylation profile are any of those provided in the Twist Methylome panel available at the time of filing of this application.

[0066] In some embodiments, one or more methylation sites in the methylation profile are from a selected CpG methylation site. In some embodiments, one or more methylation sites in the methylation profile are from 1, 5, 10, 15, 20, 25, 30, 35, 40, 50, 75, or 100, 150, 200, 250, 300, 400, 500, 750, 1000, 2000, 2500, 3000, 4000, 5000, 7500, 10000, 20000, 25000, 30000, 40 ...60000, 70000, 80000, 90000, 10000, 110000, 120000, 130000, 14000, 15000, 16000, 17000, 18000, 19000, 210000, 220000, 230000, 240000, 25000, 260000, 27000, 28000, 310000, 320000, 330000, 34000, 35000, 36 At least 1, 5, 10, 15, 20, 25, 30, 35, 40, 50, 75, or more than 100, 150, 200, 250, 300, 400, 500, 750, 1000, 2000, 2500, 30 ...500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 3000, 3500, 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2200, 2300, 24 000, 5000, 7500, 10000, 20000, 25000, 30000, 40000, 50000, 75000, 100000, 200000, 300000, 400000, 500000, 600000, and 700000, or about 1, 5, 10, 15, 20, 25, 30, 35, 40, 50, 75, or 100, 150, 200, 250, 3 00, 400, 500, 750, 1000, 2000, 2500, 3000, 4000, 5000, 7500, 10000, 20000, 25000, 30000, 40000, 50000, 75000, 100000, 200000, 300000, 400000, 500000, 600000, and 700000 selected CpG methylation sites. In some embodiments, one or more methylation sites in the methylation profile are from about 1 to about 500,000 selected CpG methylation sites. In some embodiments, one or more methylation sites of the methylation profile are selected from the group consisting of: cg18081940, cg23089825, cg16395183, cg19811148, cg07790615, cg20996351, cg04977528, cg24465685, cg20428713, cg13678973, cg25339566, cg16596317, cg23786625, cg11328303, cg19578660, cg02272851, cg10298052,cg13585930、cg23575688、cg12394201、cg08149193、cg18854419、cg07603330、cg10658542、cg13099890、cg22302985、cg13596497、cg14507533、cg25366582、cg22396555、cg10566012、cg05168229、cg10795666、cg25078444、cg16038120、cg23883632、cg18380808、cg13615592、cg00250422、cg19691260、cg16558770、cg15681853、cg03397724、cg10514097、cg06674117、cg16047279、cg12127472、cg08843809、cg08697732、cg06384763、cg04203646、cg17112426、cg08278741、cg14587524、cg26087117、cg18320766、cg08063125、cg10004780、cg18921980、cg02514318、cg20002504、cg18897632、cg15313459、cg19370054、cg16564824、cg02631468、cg01471196、cg23770904、cg18412834、cg24080247、cg11549874、cg13155421、cg19442495、cg22536150、cg05413061、cg23346462、cg09477895、cg13605674、cg13314965、cg09417547、cg00181669、cg23967169、cg10237419、cg21077559、cg27600205、cg19755714、cg18797590、cg00699993、cg06485940、cg27661394、cg00939495、cg11036833、cg23915769、cg07224726、cg02022733、cg03640756、cg15361590、cg04598517、cg06782035、cg13954457、cg25482900、cg20952257、cg14062050、cg01881524、cg11538641、cg11387340、cg05389236、cg19419054、cg10575547, cg17240815, cg24772267, cg00920327, cg00772257, cg26253500, cg23244488, cg22 778435, cg26065247, cg02088996, cg19868631, cg22280038, cg07803375, cg20230721, cg0333333 0, cg21517947, cg10406295, cg05166490, cg07739205, cg20980783, cg06617456, cg01568998, cg13407456, cg23758305, cg20675505, cg07585876, cg03734437, cg13410764, or any combination thereof.

[0067] In some embodiments, one or more methylation sites of the methylation profile are selected from the group consisting of cg18081940, cg23089825, cg16395183, cg19811148, cg07790615, cg20996351, cg04977528, cg24465685, cg20428713, cg13678973, cg25339566, cg16596317, cg23786625, cg11328303, cg19578660, cg02272851, cg10298052, cg13585930, cg23575688, cg1 2394201, cg08149193, cg18854419, cg07603330, cg10658542, cg13099890 , cg22302985, cg13596497, cg14507533, cg25366582, cg22396555, cg1056 6012, cg05168229, cg10795666, cg25078444, cg16038120, cg23883632, cg 18380808, cg13615592, cg00250422, cg19691260, cg16558770, cg1568185 3, cg03397724, cg10514097, cg06674117, cg16047279, cg12127472, cg088 43809, cg08697732, cg06384763, cg04203646, cg17112426, cg08278741, c g14587524, cg26087117, cg18320766, cg08063125, cg10004780, cg189219 80, cg02514318, cg20002504, cg18897632, cg15313459, cg19370054, cg16 564824, cg02631468, cg01471196, cg23770904, cg18412834, cg24080247, cg11549874, cg13155421, cg19442495, cg22536150, cg05413061, cg23346 462, cg09477895, cg13605674, cg13314965, cg09417547, cg00181669, cg2 3967169, cg10237419, cg21077559, cg27600205, cg19755714, cg18797590,cg00699993, cg06485940, cg27661394, cg00939495, cg11036833, cg23915769, cg 07224726, cg02022733, cg03640756, cg15361590, cg04598517, cg06782035, cg13 954457, cg25482900, cg20952257, cg14062050, cg01881524, cg11538641, cg1138 7340, cg05389236, cg19419054, cg10575547, cg17240815, cg24772267, cg009203 27, cg00772257, cg26253500, cg23244488, cg22778435, cg26065247, cg02088996 , cg19868631, cg22280038, cg07803375, cg20230721, cg03333330, cg21517947, c g10406295, cg05166490, cg07739205, cg20980783, cg06617456, cg01568998, cg13407456, cg23758305, cg20675505, cg07585876, cg03734437, and cg13410764.

[0068] In some embodiments, the one or more methylation sites of the methylation profile comprise methylation sites in one or more gene promoter regions.

[0069] In some embodiments, the methylation profile comprises quantitative information from at least one of the one or more methylation sites.

[0070] In some embodiments, the quantitative information is based on a beta value from at least one methylation site. In some embodiments, the methylation profile includes a quantitative value threshold to indicate when the level of methylation at a methylation site meets a condition, such as hypermethylation above a particular beta value or hypomethylation below a particular beta value.

[0071] In some embodiments, the quantitative information is based on the cellular heterogeneity-adjusted cLonal methylation (CHALM) ratio from at least one methylation site. See, for example, Xuet et al., Nat Commun, 12, 2021. CHALM is a method for quantifying cellular heterogeneity-adjusted average methylation. CHALM quantifies promoter methylation as the ratio of methylated reads (at ≧1 mCpG) to all reads mapped to a given promoter region.

[0072] In certain embodiments, the information obtained from the methylation sites is mathematically combined, and the combined value correlates to a fundamental therapeutic question, such as a diagnostic question. In some embodiments, the information from multiple methylation sites is combined by any suitable state-of-the-art mathematical method. Well-known mathematical methods for correlating a combination of biomarkers, for example, to a disease state, include discriminant analysis (DA) (e.g., linear, quadratic, regularized DA), discriminant function analysis (DFA), kernel methods (e.g., SVM), multidimensional scaling (MDS), non-parametric methods (e.g., k-nearest neighbor classifiers), PLS (partial least squares), tree-based methods (e.g., logistic regression, CART, random forests, boosting / backing methods), generalized linear models (e.g., logistic regression), principal component-based methods (e.g., SIMCA), generalized additive models, fuzzy logic-based methods, neural networks, and genetic algorithm-based methods. In some embodiments, the mathematical model includes a p-value test or a t-value test or an F-test. The evaluated (best, i.e., lowest p- or t-value) methylation sites can then be subsequently selected and added to the methylation panel until a particular value, such as a diagnostic value with a desired confidence level, is reached. Such methods include the probability difference t-test (Wright GW and Simon R., Bioinformatics 19:2448-2455, 2003). The skilled artisan will have no problem in selecting an appropriate method for evaluating the methylation sites and combinations described herein.For more details on these statistical methods, see the following references: Ruczinski et al., 12 J. of Computational and Graphical Statistics 475-511 (2003); Friedman, JH, 84 J. of the American Statistical Association 165-75 (1989); Histy, Trevor; Tibshirani, Robert; Friedman, Jerome, The Elements of Statistical Learning, Springer Series in Statistics (2001); Breiman, L., Friedman, JH, Olshen, RA, Stone, CJ Classification and regression trees, California: Wadsworth (1984); Breiman, L., 45 Machine Learning 5-32 (2001); Pepe, MS, The Statistical Evaluation of Medical Tests for Classification and Prediction, Oxford Statistical Science Series, 28 (2003), and Duda, RO, Hart, PE, Stork, DO, Pattern Classification, Wiley Interscience, 2nd Edition (2001).

[0073] In some embodiments, the methods provided herein include models for prediction. These models can be based on multi-covariate predictors (Radmacher et al., J of Computational Biology 9:505-511, 2002), diagonal linear discriminant analysis (Dudoit et al., Journal of the American Statistical Association 97:77-87, 2002), nearest neighbor classification (also Dudoit et al.), and support vector machines with linear kernels (Ramaswamy et al., PNAS USA 98:15149-54, 2001). Another classification method is the greedy-pairs method described by Bo and Jonassen (Genome Biology 3(4):research0017.1-0017.11, 2002). The greedy-pairs approach starts by ranking all markers based on their individual t-scores on the training set. This method attempts to select pairs of markers that work well together to distinguish classes. In addition, a binary tree classifier to utilize methylation profiles is optionally used to predict the class of future samples. The first node of the tree incorporated a binary classifier that distinguished two subsets of the total set of classes. Each binary classifier was based on a "support vector machine" that incorporated markers that were differentially expressed between markers at a significance level (e.g., 0.01, 0.05, or 0.1) as assessed by a probability difference t-test (Wright GW and Simon R., Bioinformatics 19:2448-2455, 2003). The classifiers for all possible binary partitions are evaluated, and the partition selected is the one with the smallest cross-validated prediction error. The process is then repeated successively for the two subsets of classes determined by the previous binary partition. The prediction error of the binary tree classifier can be estimated by cross-validating the entire tree-building process.This global cross-validation involves reselecting the optimal partition at each node and reselecting the markers used for each cross-validated training set, as described by Simon et al. (Simon et al., Journal of the National Cancer Institute 95:14-18, 2003). A portion of the samples is withheld, a binary tree is developed with the remaining samples, and then a multi-fold cross-validation predicts the class membership of the withheld samples. This is repeated several times, each time a different proportion of samples is retained. Samples are randomly partitioned into fractional test sets (Simon R and Lam A. BRB-ArrayTools User's Guide, version 3.2. Biological Research Branch, National Cancer Institute).

[0074] III. Nucleosome dynamics profile In certain aspects, the methods provided herein include a multimodal epigenetic signature comprising one or more features obtained from a nucleosome dynamics profile.As described herein, a nucleosome dynamics profile comprises information based on the location of one or more nucleosomes ascertained from sequencing data, such as obtained from a cfDNA sample.For example, in some embodiments, a nucleosome dynamics profile comprises information based on the presence and / or absence of nucleosomes at loci on genomic DNA, including genomic DNA in systemic circulation as cfDNA.

[0075] In some embodiments, the nucleosome dynamics profile includes nucleosome position information, for example, represented by window protection score (WPS). WPS is determined by analyzing the number of sequenced DNA fragments that completely span a window, for example, a 120 bp window, centered on a given genomic coordinate, minus the number of fragments that have an endpoint within the same window, and correlates with the location of nucleosomes. In some embodiments, the nucleosome position information is based on WPS. In some embodiments, WPS is the average WPS. Methods for determining WPS are known in the art, see, for example, Snyder et al., Cell, 164, 2016, which is incorporated herein by reference in its entirety.

[0076] In some embodiments, the nucleosome dynamics profile includes nucleosome occupancy information. Nucleosome occupancy reflects the frequency with which nucleosomes occupy a nucleosome position. In some embodiments, nucleosome occupancy is based on the frequency with which nucleosomes occupy a genomic region. In some embodiments, a sliding window within the target site is used to assess nucleosome occupancy, such as a window of 250-2000 base pairs, sliding in 10 base pair steps across the target region. In some embodiments, nucleosome occupancy is obtained by normalized read coverage, which is measured in counts per million. Tools for determining normalized read coverage, including bamCoverage from deepTools, are known in the art.

[0077] In some embodiments, the nucleosome dynamics profile comprises nucleosome fuzziness information. Nucleosome fuzziness reflects the deviation of measured nucleosome positions. In some embodiments, nucleosome fuzziness is based on the deviation of nucleosome positions from preferred nucleosome positions.

[0078] In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome position information and nucleosome occupancy. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome position information and nucleosome fuzziness. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome occupancy and nucleosome fuzziness. In some embodiments, the nucleosome dynamics profile comprises information derived from nucleosome position information, nucleosome occupancy, and nucleosome fuzziness.

[0079] In some embodiments, nucleosome dynamic information is obtained through Dynamic Analysis of Nucleosome Position and Occupancy by Sequencing (DANPOS). See, e.g., Chen et al., Genome Res, 23, 2013, which is incorporated herein by reference in its entirety. DANPOS and its further versions, such as DANPOS 2, are comprehensive bioinformatics pipelines designed for dynamic nucleosome analysis at single nucleotide resolution. In some embodiments, the nucleosome dynamics profile comprises a locus-specific nucleosome score based on any one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness.

[0080] IV. Fragmentation Profile In certain aspects, the methods provided herein include a multimodal epigenetic signature that comprises one or more features obtained from a fragmentation profile.As described herein, the fragmentation profile is based on the proportion of nucleic acid fragments in one or more nucleobase length windows.As described in more detail below, the proportion of nucleic acid fragments can be based on a desired population of nucleic acid fragments, such as all sequencing reads from an assay or a subset thereof.In some embodiments, the population of nucleic acid fragments used to evaluate the fragmentation profile includes fragments associated with a target location, such as a target chromosome or one or more loci.

[0081] In some embodiments, the fragmentation profile includes one or more nucleobase length windows occupying a range of about 30 to about 250 bases, such as about 60 to about 200 bases, about 80 to about 200 bases, about 120 to about 220 bases, about 120 to about 180 bases, or about 140 to about 200 bases. In some embodiments, the fragmentation profile includes one or more nucleobase length windows occupying nucleic acids having a length of about 250 bases or less, such as about 240 bases or less, 230 bases or less, 220 bases or less, 210 bases or less, 200 bases or less, 190 bases or less, 180 bases or less, 170 bases or less, 160 bases or less, 150 bases or less, 140 bases or less, 130 bases or less, or 120 bases or less. In some embodiments, the fragmentation profile comprises one or more nucleobase length windows encompassing nucleobase lengths of about 147 bases and / or about 167 bases.

[0082] In some embodiments, the fragmentation profile comprises a window of one or more nucleobases from about 80 bases to about 150 bases, about 80 bases to about 155 bases, about 80 bases to about 160 bases, about 80 bases to about 165 bases, about 125 bases to about 155 bases, about 170 bases to about 175 bases, about 170 bases to about 200 bases, about 175 bases to about 200 bases, about 151 bases to about 200 bases, about 156 bases to about 200 bases, about 161 bases to about 200 bases, or about 166 bases to about 100 bases. In some embodiments, the fragmentation profile comprises nucleobase length windows of about 80 bases to about 150 bases, about 80 bases to about 155 bases, about 80 bases to about 160 bases, about 80 bases to about 165 bases, about 125 bases to about 155 bases, about 170 bases to about 175 bases, about 170 bases to about 200 bases, and about 175 bases to about 200 bases.

[0083] In some embodiments, the fragmentation profile comprises one or more ratios of a first nucleobase length window to a second nucleobase length window, in some embodiments, the ratio is one or more of about 80 to about 150 bases relative to about 151 to about 200 bases, about 80 to about 155 bases relative to about 156 to about 200 bases, about 80 to about 160 bases relative to about 161 to about 200 bases, or about 80 to about 165 bases relative to about 166 to about 200 bases. In some embodiments, the fragmentation profile comprises a ratio of about 80 bases to about 150 bases to about 151 bases to about 200 bases, about 80 bases to about 155 bases to about 156 bases to about 200 bases, about 80 bases to about 160 bases to about 161 bases to about 200 bases, and about 80 bases to about 165 bases to about 166 bases to about 200 bases.

[0084] In some embodiments, the nucleobase length window is from about 5 to about 150 bases long, such as from about 10 to about 120 bases long, from about 40 to about 100 bases long, or from about 60 to about 80 bases long. In some embodiments, the nucleobase length window is at least about 5 bases long, such as from about 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, or 150 bases long. In some embodiments, the nucleobase length window is about 150 bases or less in length, e.g., 145 bases or less, 140 bases or less, 135 bases or less, 130 bases or less, 125 bases or less, 120 bases or less, 115 bases or less, 110 bases or less, 105 bases or less, 100 bases or less, 95 bases or less, 90 bases or less, 85 bases or less, 80 bases or less, 75 bases or less, 70 bases or less, 65 bases or less, 60 bases or less, 55 bases or less, 50 bases or less, 45 bases or less, 40 bases or less, 35 bases or less, 30 bases or less, 25 bases or less, 20 bases or less, 15 bases or less, or 10 bases or less in length. In some embodiments, the nucleobase length window is about any of 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 65, 70, 75, 80, 85, 90, 95, 100, 105, 110, 115, 120, 125, 130, 135, 140, 145, or 150 bases long.

[0085] In some embodiments, the fragmentation profile comprises two or more windows. The two windows can further be used to calculate a ratio, for example, [number of fragments in the smaller window] / [number of fragments in the larger window]. In some embodiments, the two or more windows of the fragmentation profile are of uniform base length size. In some embodiments, the two or more windows of the fragmentation profile comprise a first window and a second window having different base length sizes. In some embodiments, the two or more windows of the fragmentation profile have some overlap in base length size.

[0086] In some embodiments, the fragments used to construct the fragmentation profile include fragments from whole genome sequencing analysis. In some embodiments, the fragments used to construct the fragmentation profile include fragments from one or more specific locations, such as one or more chromosomes, or one or more target loci or regions.

[0087] V. Machine Learning Technology In certain aspects, the methods described herein include the use of machine learning techniques. In some embodiments, the machine learning techniques include models configured to identify, such as discover, multimodal epigenetic signatures. In some embodiments, the machine learning techniques include models configured to assess the presence of multimodal epigenetic signatures.

[0088] In some embodiments, provided herein is a method of generating an epigenetic signature from a sample obtained from an individual, the method comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; extracting features from the sequencing data, the features comprising information from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate an epigenetic signature based on the plurality of features; and outputting the generated epigenetic signature.

[0089] In some embodiments, a method is provided for identifying a disease epigenetic signature indicative of an individual having a disease, the method comprising receiving sequencing data from a plurality of individuals with the disease and a plurality of individuals without the disease, the sequencing data being obtained from a non-destructive methylation sequencing technique performed on samples obtained from the individuals; and extracting features from the sequencing data, the features being a methylation profile comprising information derived from one or more of the following profiles: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome location information. The method includes extracting features from two or more of: a nucleosome dynamics profile comprising information derived from one or more of a nucleosome fuzziness, or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model, wherein the extracted features from each of a plurality of individuals are populated with an associated classification of the individual as having or not having the disease; training the machine learning model to identify a disease epigenetic signature using the extracted features; and outputting the disease epigenetic signature.

[0090] Machine learning models are known in the art, including those described in other sections. In some embodiments, the machine learning technique includes a support vector machine model. In some embodiments, the machine learning technique includes a random forest machine model. In some embodiments, the machine learning technique includes a logistic regression machine model. In some embodiments, the input of the machine learning technique is a methylation profile, such as based on one or more qualitative and / or quantitative measures associated with one or more methylation sites. In some embodiments, the input of the machine learning technique is a nucleosome dynamics profile, such as based on a locus-specific nucleosome score based on any one or more of (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness. In some embodiments, the input of the machine learning technique is a fragmentation profile, such as based on the proportion of nucleic acid fragments within one or more nucleobase length windows. In some embodiments, the input is two or more of a methylation profile, a nucleosome dynamics profile, or a fragmentation profile.

[0091] In some embodiments, the methods provided herein include training a machine learning model. In some embodiments, the machine learning technique includes a trained model. In some embodiments, the machine learning model is trained by inputting information obtained from any of a methylation profile, a nucleosome dynamics profile, and a fragmentation profile, the information being associated with a sample having a known biological state, such as associated with a diseased or non-disease state. In some embodiments, the machine learning model is trained using single modality data, i.e., each of a methylation profile, a nucleosome dynamics profile, and a fragmentation profile. In some embodiments, the machine learning model is trained using multimodal data, i.e., any combination of two or more of a methylation profile, a nucleosome dynamics profile, and a fragmentation profile. In some embodiments, the training using multimodal data follows a concatenation-based strategy that combines multiple types of features from each sample into a single dataset for model training.

[0092] In some embodiments, the machine learning training involves the use of data from a population of individuals, e.g., a population of individuals with a disease, e.g., cancer, or a population of individuals without a disease, e.g., healthy individuals. In some embodiments, the population is two or more individuals, including any of 5 or more individuals, 50 or more individuals, 100 or more individuals, 500 or more individuals, or 1,000 or more individuals. In some embodiments, the individuals have a confirmed biological condition, e.g., a diagnosis of a disease, using techniques conventional in the art. In some embodiments, the method further comprises a cross-validation procedure, e.g., to reveal the presence of a multimodal epigenetic signature and / or the presence thereof.

[0093] C. Diseases, Individuals, and Samples The disclosure provided herein is useful for multimodal epigenetic signatures for a wide variety of individuals and / or diseases and / or sample types.

[0094] In some embodiments, the individual is a human. In some embodiments, the individual is male. In some embodiments, the individual is female. In some embodiments, the individual is suspected of having a disease. In some embodiments, the individual is suspected of not having a disease. In some embodiments, the individual is a healthy individual, such as an individual not having a disease.

[0095] In some embodiments, the disease is cancer, hi some embodiments, the cancer is selected from the group consisting of prostate cancer, lung cancer, bronchial cancer, colon cancer, rectal cancer, colorectal cancer, bladder cancer, melanoma, kidney cancer, renal pelvis cancer, non-Hodgkin's lymphoma, oral cancer, pharyngeal cancer, leukemia, liver cancer, intrahepatic bile duct cancer, breast cancer, uterine cancer, thyroid cancer, splenic cancer, esophageal cancer, ovarian cancer, brain cancer, and cancer of the nervous system.

[0096] In some embodiments, the cancer comprises a primary tumor. In some embodiments, the cancer comprises one or more metastatic tumors. For example, in some embodiments, the individual has or is suspected to have colon cancer with one or more metastases to any location, such as the liver, lung, bone, or brain.

[0097] In some embodiments, the disease is a benign inflammatory disease. In some embodiments, the disease is diverticulitis. In some embodiments, the disease is ulcerative colitis. In some embodiments, the disease is Crohn's disease. In some embodiments, the disease is infectious colitis. In some embodiments, the disease is non-infectious colitis.

[0098] In some embodiments, the sample is a cell-free DNA (cfDNA) sample. In some embodiments, the sample is a blood sample or a derivative thereof, such as plasma. In some embodiments, the sample is a blood sample, such as a whole blood sample. In some embodiments, the sample is a plasma sample. In some embodiments, the sample is a serum sample. In some embodiments, the sample is a tissue sample. In some embodiments, the sample comprises nucleic acid derived from a tissue within an individual, such as from a diseased tissue. In some embodiments, the method further comprises obtaining a sample, such as a cfDNA sample, such as via blood drawing. In some embodiments, the method further comprises processing the sample, such as to obtain a cfDNA sample. In some embodiments, processing the sample comprises one or more steps to separate blood components from cfDNA.

[0099] In some embodiments, the sample is obtained from a liquid biopsy specimen. In some embodiments, the liquid sample is blood or other liquid samples of biological origin (peripheral blood, serum, plasma, ascites, urine, cerebrospinal fluid (CSF), sputum, saliva, bone marrow, synovial fluid, aqueous humor, amniotic fluid, earwax, breast milk, bronchoalveolar lavage fluid, semen, prostatic fluid, Cowper's or pre-ejaculatory fluid, female ejaculate, sweat, tears, cyst fluid, pleural and peritoneal fluid, pericardial fluid, peritoneal fluid, lymph, chyme, chyle, bile, interstitial fluid, menstrual fluid, pus, sebum, vomit, vaginal secretions / flushes, synovial fluid, mucosal secretions, fecal water, pancreatic juice, sinus lavage, bronchopulmonary aspirate, blastocyst cavity fluid, or the like). In some embodiments, the biological fluid includes, but is not limited to, blood, serum, or umbilical cord blood. In some embodiments, the biological fluid is blood, a blood product, or a blood fraction, such as serum or plasma. In some embodiments, the sample comprises a blood sample. In another embodiment, a serum sample is used. In another embodiment, the sample comprises urine. In some embodiments, a liquid sample also encompasses samples that have been manipulated in any way after their procurement, such as by centrifugation, filtration, precipitation, dialysis, chromatography, treatment with reagents, washing, or enrichment of specific cell populations.

[0100] D. Exemplary Methods In some aspects, provided herein is a method of determining an epigenetic signature from a sample obtained from an individual, the method comprising analyzing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual to determine an epigenetic signature, the epigenetic signature comprising features from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites, a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness, or a fragmentation profile comprising information derived from read distribution within one or more base length windows. In some embodiments, the epigenetic signature comprises features from the methylation profile and the nucleosome dynamics profile, e.g., nucleosome occupancy information.

[0101] In another aspect, provided herein is a method of generating an epigenetic signature from a sample obtained from an individual, the method comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; extracting features from the sequencing data, the features comprising information from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information, (b) nucleosome occupancy, or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution within one or more base length windows; inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate an epigenetic signature based on the plurality of features; and outputting the generated epigenetic signature. In some embodiments, the epigenetic signature comprises features from a methylation profile and a nucleosome dynamics profile, such as, for example, nucleosome occupancy information.

[0102] In another aspect, provided herein is a method of diagnosing a disease in an individual, the method comprising: determining an epigenetic signature from data obtained from a non-destructive methylation sequencing technique performed on a sample obtained from the individual, the epigenetic signature comprising features from two or more of the following profiles: a methylation profile comprising information derived from one or more methylation sites; a nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from read distribution in one or more base length windows; and diagnosing the disease in the individual based on the epigenetic signature compared to the disease epigenetic signature. In some embodiments, the method further comprises diagnosing the disease in the individual based on the epigenetic signature compared to the disease epigenetic signature. In some embodiments, the epigenetic signature comprises features from a methylation profile and a nucleosome dynamics profile, such as, for example, nucleosome occupancy information.

[0103] In another aspect, provided herein is a method of treating a disease in an individual, the method comprising diagnosing the individual as having the disease according to any method provided herein and administering an agent to treat the disease in the individual. In some embodiments, the epigenetic signature comprises features from a methylation profile and a nucleosome dynamics profile, e.g., nucleosome occupancy information.

[0104] In another aspect, provided herein is a method of identifying a disease epigenetic signature indicative of an individual having a disease, the method comprising receiving sequencing data from a plurality of individuals having the disease and a plurality of individuals not having the disease, the sequencing data being obtained from a non-destructive methylation sequencing technique performed on samples obtained from the individuals; and extracting features from the sequencing data, the features being a methylation profile comprising information derived from one or more of the following profiles: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome location information. The method includes: extracting features from two or more of a nucleosome dynamics profile including information derived from one or more of a nucleosome fuzziness or a fragmentation profile including information derived from a read distribution in one or more base length windows; inputting the extracted features into a machine learning model, where the extracted features from each of a plurality of individuals are populated with an associated classification of the individual as having or not having the disease; training the machine learning model to identify a disease epigenetic signature using the extracted features; and outputting the disease epigenetic signature. In some embodiments, the epigenetic signature includes features from the methylation profile and the nucleosome dynamics profile, such as, for example, nucleosome occupancy information.

[0105] E. Systems, Kits, and Components In certain aspects, contemplated herein are systems, kits, and components useful for carrying out the methods described herein.

[0106] In some embodiments, a system is provided herein, such as a computer system, for analyzing data according to the description provided herein. In some embodiments, the system is configured to receive sequencing data. In some embodiments, the system is configured to extract features from the sequencing data, such as corresponding to any one or combination of a methylation profile, a nucleosome profile, or a fragmentation profile. In some embodiments, the system is configured to input the features extracted from the sequencing data into a machine learning model. In some embodiments, the system is configured to train the machine learning model. In some embodiments, the system is configured to output an epigenetic signature. In some embodiments, a system is provided herein that is configured to perform a computer-implemented method described herein. In some embodiments, the system includes one or more processors and a memory that stores one or more programs, the one or more programs configured to be executed by the one or more processors, the one or more programs including instructions for performing the method described herein. In some embodiments, the system includes a machine learning model.

[0107] In some embodiments, provided herein are kits and / or kit components for performing aspects of the methods described herein. For example, in some embodiments, provided herein are kits and / or kit components for obtaining and / or processing samples from individuals, such as for obtaining cfDNA samples. In some embodiments, provided herein are kits and / or kit components for performing the non-destructive methylation sequencing techniques described herein.

[0108] The scope of the present invention is not intended to be limited to the specific disclosed embodiments, which are provided, for example, to illustrate various aspects of the present invention. Various modifications to the compositions and methods described herein will become apparent from the descriptions and teachings herein. Such variations can be made without departing from the true scope and spirit of the present disclosure and are intended to be included within the scope of the present disclosure.

[0109] Exemplary embodiments The following exemplary embodiments are provided herein:

[0110] Embodiment 1. A method of determining an epigenetic signature from a sample obtained from an individual, comprising analyzing data obtained from a non-destructive methylation sequencing technique performed on said sample obtained from said individual to determine said epigenetic signature, The epigenetic signature is characterized in that it is a profile of: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; A fragmentation profile that contains information derived from the distribution of reads in one or more base length windows The method includes features obtained from two or more of the above.

[0111] Embodiment 2. A method for generating an epigenetic signature from a sample obtained from an individual, comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; Extracting features from the sequencing data, The features are characterized by the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; A fragmentation profile that contains information derived from the distribution of reads in one or more base length windows said extracting including information from two or more of: inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate the epigenetic signature based on a plurality of the features; Outputting the generated epigenetic signature. The method comprising:

[0112] Embodiment 3. A method for diagnosing a disease in an individual, comprising: determining an epigenetic signature obtained from a non-destructive methylation sequencing technique performed on a sample obtained from said individual, The epigenetic signature is characterized in that it is a profile of: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; A fragmentation profile that contains information derived from the distribution of reads in one or more base length windows and comparing the epigenetic signature to a disease epigenetic signature and diagnosing the disease in the individual based on the epigenetic signature; The method comprising:

[0113] Embodiment 4. The method of embodiment 1 or 2, further comprising diagnosing disease in said individual based on said epigenetic signature compared to a disease epigenetic signature.

[0114] Embodiment 5. A method of treating a disease in an individual, comprising: diagnosing said individual as having the disease according to embodiment 3 or 4; and administering an agent to treat said disease in said individual.

[0115] Embodiment 6. A method of identifying a disease epigenetic signature indicative of an individual having a disease, comprising: receiving sequencing data from a plurality of individuals with the disease and a plurality of individuals without the disease; the receiving, wherein the sequencing data is obtained from a non-destructive methylation sequencing technique performed on a sample obtained from the individual; and Extracting features from the sequencing data, The features are characterized by the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; A fragmentation profile that contains information derived from the distribution of reads in one or more base length windows said extracting including information from two or more of: inputting the extracted features into a machine learning model, inputting the extracted features from each of the plurality of individuals populated with an associated classification of the individual as having or not having the disease; training the machine learning model to identify the disease epigenetic signature using the extracted features; Outputting the disease epigenetic signature. The method comprising:

[0116]

[0023] Embodiment 7. Each of the one or more methylation sites of the methylation profile is selected from the group consisting of cg18081940, cg23089825, cg16395183, cg19811148, cg07790615, cg20996351, cg04977528, cg24465685, cg20428713, cg13678973, cg25339566, cg16596317, cg23786625, cg11328303, cg19578660, cg02272851, cg10298052, cg13585930, cg23575688, c g12394201, cg08149193, cg18854419, cg07603330, cg10658542, cg130998 90, cg22302985, cg13596497, cg14507533, cg25366582, cg22396555, cg10 566012, cg05168229, cg10795666, cg25078444, cg16038120, cg23883632, cg18380808, cg13615592, cg00250422, cg19691260, cg16558770, cg156818 53, cg03397724, cg10514097, cg06674117, cg16047279, cg12127472, cg08 843809, cg08697732, cg06384763, cg04203646, cg17112426, cg08278741, cg14587524, cg26087117, cg18320766, cg08063125, cg10004780, cg18921 980, cg02514318, cg20002504, cg18897632, cg15313459, cg19370054, cg16 564824, cg02631468, cg01471196, cg23770904, cg18412834, cg24080247, cg11549874, cg13155421, cg19442495, cg22536150, cg05413061, cg23346 462, cg09477895, cg13605674, cg13314965, cg09417547, cg00181669, cg2 3967169, cg10237419, cg21077559, cg27600205, cg19755714, cg18797590,cg00699993, cg06485940, cg27661394, cg00939495, cg11036833, cg23915769, cg0722 4726, cg02022733, cg03640756, cg15361590, cg04598517, cg06782035, cg13954457, c g25482900, cg20952257, cg14062050, cg01881524, cg11538641, cg11387340, cg05389 236, cg19419054, cg10575547, cg17240815, cg24772267, cg00920327, cg00772257, cg2 6253500, cg23244488, cg22778435, cg26065247, cg02088996, cg19868631, cg2228003 8, CG07803375, CG20230721, CG03333330, CG21517947, CG10406295, CG05166490, CG077 39205, cg20980783, cg06617456, cg01568998, cg13407456, cg23758305, cg20675505, cg07585876, cg03734437, and cg13410764.

[0117] Embodiment 8. The method of any one of embodiments 1 to 7, wherein the one or more methylation sites of the methylation profile include methylation sites in one or more gene promoter regions.

[0118] Embodiment 9. The method of any one of embodiments 1 to 8, wherein the methylation profile comprises quantitative information from at least one of the one or more methylation sites.

[0119] Embodiment 10. The method of embodiment 9, wherein said quantitative information is based on a β value from said at least one methylation site.

[0120] Embodiment 11. The method of embodiment 9, wherein said quantitative information is based on a CHALM ratio from said at least one methylation site.

[0121] Embodiment 12. The method of any one of embodiments 1 to 11, wherein the nucleosome dynamics information is based on nucleosomes at genomic loci.

[0122] Embodiment 13. The method of any one of embodiments 1 to 12, wherein the nucleosome position information is based on a window protection score (WPS).

[0123] Embodiment 14. The method of embodiment 13, wherein the WPS is an average WPS.

[0124] Embodiment 15. The method of any one of embodiments 1 to 14, wherein the nucleosome occupancy is based on the frequency with which nucleosomes occupy a genomic region.

[0125] Embodiment 16 The method of embodiment 15, wherein said nucleosome occupancy is obtained via normalized read coverage measured in counts per million.

[0126] Embodiment 17. The method of any one of embodiments 1 to 16, wherein the nucleosome fuzziness is based on the deviation of nucleosome positions from preferred nucleosome positions.

[0127] Embodiment 18. The method of any one of embodiments 1 to 18, wherein the fragmentation profile is based on one or more base length windows occupying a range of 30 to 250 base lengths.

[0128] Embodiment 19. The method of embodiment 19, wherein the base length window is at least 10 bases long.

[0129] Embodiment 20. The method according to any one of embodiments 12 to 17, wherein the nucleosome dynamics information is obtained via DANPOS.

[0130] Embodiment 21 The method of any one of embodiments 1 to 20, wherein the epigenetic signature indicates whether the individual has a disease.

[0131] Embodiment 22. The method of any one of embodiments 1 to 21, wherein the epigenetic signature comprises features from the methylation profile and the nucleosome dynamics profile.

[0132] Embodiment 23. The method of any one of embodiments 1 to 21, wherein the epigenetic signature comprises features from the methylation profile and the fragmentation profile.

[0133] Embodiment 24. The method of any one of embodiments 1 to 21, wherein the epigenetic signature comprises features from the nucleosome dynamics profile and the fragmentation profile.

[0134] Embodiment 25. The method of any one of embodiments 1 to 24, wherein the epigenetic signature comprises features from the methylation profile, the nucleosome dynamics profile, and the fragmentation profile.

[0135] Embodiment 26 The method of any one of embodiments 22, 24, or 25, wherein the nucleosome dynamics profile comprises information derived from nucleosome position information.

[0136] Embodiment 27. The method of any one of embodiments 22 or 24 to 26, wherein the nucleosome dynamics profile comprises information derived from nucleosome occupancy.

[0137] Embodiment 28. The method of any one of embodiments 22 or 24 to 27, wherein the nucleosome dynamics profile comprises information derived from nucleosome fuzziness.

[0138] Embodiment 29. The method according to any one of embodiments 1 to 28, wherein the non-destructive methylation sequencing technique is the EM-seq technique.

[0139] Embodiment 30. The method according to any one of embodiments 1 to 29, wherein the non-destructive methylation sequencing technique is performed based on a target gene position.

[0140] Embodiment 31 The method of any one of embodiments 1 to 30, further comprising performing said non-destructive methylation sequencing technique.

[0141] Embodiment 32. The method of any one of embodiments 1 to 31, wherein the data obtained from the non-destructive methylation sequencing technique comprises a plurality of sequence reads.

[0142] Embodiment 33. The method of embodiment 32, further comprising processing the plurality of sequence reads to remove low quality reads and / or remove adapter contamination and / or filter based on the size of the sequence reads.

[0143]

[0036] Embodiment 34. The method of embodiment 32 or 33, further comprising aligning the plurality of sequence reads to a reference genome.

[0144] Embodiment 35. The method of any one of embodiments 2 or 6 to 34, wherein the machine learning model comprises a support vector machine model, a random forest machine model, or a logistic regression machine model.

[0145] Embodiment 36. The method of embodiment 35, further comprising a cross-validation procedure.

[0146] Embodiment 37 The method of any one of embodiments 1 to 22, wherein the sample is a cell-free DNA sample.

[0147] Embodiment 38. The method of any one of embodiments 1 to 23, further comprising obtaining the sample.

[0148] Embodiment 39. The method of any one of embodiments 1 to 24, wherein the disease is cancer.

[0149] Embodiment 40 The method of embodiment 39, wherein the cancer is colorectal cancer.

[0150] Embodiment 41 The method of any one of embodiments 1 to 40, wherein the individual is a human.

[0151] Embodiment 42 The method of any one of embodiments 1 to 41, wherein the individual is suspected of having a disease. EXAMPLES

[0152] Working Example Example 1: Methodology of Multimodal Epigenetic Sequencing Assay (MESA) for Accurate Detection of Human Cancer This example describes a method of multimodal epigenetic sequencing assay (MESA) for accurate detection of human cancer. The method demonstrated herein is a flexible and sensitive method capable of combining at least two profiles (e.g., selected from methylation profile, nucleosome dynamics profile, and fragmentation profile) within a single assay using non-destructive enzymatic methylation sequencing and innovative bioinformatics algorithms.

[0153] Plasma cell-free DNA (cfDNA) is degraded DNA fragments released into the bloodstream. In healthy individuals, plasma cfDNA is mainly derived from apoptosis of normal hematopoietic cells, with minimal contribution from other tissues. In individuals with certain physiological or pathological conditions, the proportion of cfDNA may have different origins, such as diseased tissues, when compared with healthy conditions.

[0154] A frequently reported epigenetic alteration in cancer cells is DNA methylation, which may occur early in tumorigenesis. Bisulfite genomic sequencing is considered the gold standard technique for DNA methylation detection. However, bisulfite treatment severely damages DNA, resulting in incomplete capture of cifDNA methylome and biasing downstream studies of potential biomarkers.

[0155] In this study, we utilized a recently developed bisulfite-free DNA methylation sequencing method that utilizes non-destructive enzymes. As demonstrated below, we found that the non-destructive nature of enzymatic methylation sequencing also allows for additional epigenetic analysis (e.g., fragmentation profiles, or nucleosome dynamics profiles, including nucleosome position, nucleosome occupancy, and nucleosome fuzziness) simultaneously in cfDNA methylation sequencing analysis. Although nucleosome organization is weakly related to fragmentation profiles, nucleosome information can provide information beyond fragmentation. Nucleosome organization focuses on location-specific cfDNA fragments, whereas fragmentation profiles focus only on cfDNA fragment size overall. Two samples may have the same fragment size distribution but have widely different nucleosome organization in most regions. Furthermore, fragmentation profiles usually require whole genome sequencing, whereas nucleosome organization is amenable to targeted sequencing in small regions (e.g., 2 kb).

[0156] Here, we demonstrate a three-in-one method to measure cfDNA to obtain methylation profiles, nucleosome dynamics profiles, and fragmentation profiles in a single assay using non-destructive enzymatic methylation sequencing and highly innovative bioinformatics algorithms. By integrating and analyzing these multimodal features, accurate detection of colon cancer was significantly improved. Using EM-seq techniques, we designed an enzyme-based targeted cfDNA methylation sequencing panel for 83 colon cancer patients and 83 healthy individuals. Target regions included both the commercially available Twist Methylome panel and a custom nucleosome organization panel including open chromatin ATAC peaks, CpG islands, enhancers, transcription start sites (TSS), RNA splicing sites, and polyadenylation sites (PAS) of cancer genes.

[0157] The raw sequencing data was first trimmed by TrimGalore to remove low-quality reads and potential adapter contamination. The remaining sequencing reads were then aligned to the hg19 human genome reference using BSMAP. The aligned reads were further processed with Samtools and Bedtools to retain only primary mapped reads with fragment sizes between 80bp and 200bp. This final bam file served as the input file for all subsequent processes.

[0158] Using this cfDNA data, we extracted three types of features from a single assay: methylation profile, nucleosome dynamics profile, and fragmentation profile. Specifically, for the methylation profile, we performed conventional average methylation (beta value) of the target methylation site. Methrative.py (BSMAP) was used to extract the methylation rate from the aligned bam file of the target methylation site. In addition, CHALM methylation analysis was performed according to Xu et al. (Nature Communication, 2021). For the nucleosome dynamics profile, three features were evaluated: nucleosome position information (by window protection score, WPS), nucleosome occupancy, and nucleosome fuzziness. The window protection score (WPS) is used to evaluate the position by the concept that cfDNA fragment endpoints should cluster around nucleosome boundaries and be depleted at the nucleosome itself. WPS was calculated as the number of complete fragments minus the number of fragment endpoints within a given window size. We calculated the average WPS for each sliding window described herein. Nucleosome occupancy reflects the frequency with which nucleosomes occupy a given DNA region in a cell population. We divided each target region of 2 kB into sliding windows of 500 or 1000 bp in 10 bp steps. We then calculated nucleosome occupancy features for each sliding window in two ways: (1) normalized read coverage, measured in counts per million (CPM) using the bamCoverage tool in deepTools, and (2) occupancy values ​​reported by DANPOS2. In a cell population, the exact location of nucleosomes within each DNA region may deviate from the most favored location. We define the fuzziness score as the deviation of nucleosome positions within a region in a cell population. For each sliding window described above for nucleosome occupancy, we calculated the average fuzziness score (reported by DANPOS2) of all nucleosomes whose centers are located within the window. The fragmentation profile was defined as the percentage of cfDNA fragments within a particular size range.The features included P(80–150), P(80–155), P(80–160), P(80–165), P(125–155), P(170–175), P(170–200), P(175–200), P(80–150) / P(151–200), P(80–155) / P(156–200), P(80–160) / P(161–200) and P(80-165) / P(166-200), where P(xy) is the proportion of fragments in the size range of x to y bp. Two sets of these features were extracted at different scales. For the first set, the fragmentation profile was calculated by combining all the target regions. For the second set, the fragmentation profile for each chromosome was calculated, respectively.

[0159] Machine learning cancer detection models were trained and tested using either single-modality data, i.e., each of the three types of features, or multimodal data, i.e., any combination of the three types of features. The multimodal data was integrated with a concatenation-based strategy, which combines multiple types of features from each sample into one single dataset for model training. Three machine learning algorithms were used: support vector machine, random forest, and logistic regression. These algorithms were trained and evaluated using the following cross-validation procedure: At each iteration, samples were split into 70%-30% training-test sub-datasets. The area under the receiver operating characteristic curve (AUC-ROC) and sensitivity were then calculated accordingly.

[0160] Here, we present exemplary results from a multimodal model based on a combination of traditional average methylation and nucleosome occupancy reported by DANPOS2. The multimodal model was trained using a random forest algorithm. Median values ​​over 50 iterations are shown in Table 1. The results first show that both DNA methylation and nucleosome occupancy alone are highly sensitive predictors of colon cancer in patient cfDNA. Furthermore, adding methylation to nucleosome occupancy outperforms either single model alone, indicating a significant improvement using a multimodal approach. [Table 1]

[0161] Combining the multimodal approach with our innovative bioinformatics algorithms will provide a major advance in the field of cfDN liquid biopsy cancer detection. In addition to what we demonstrate herein, another advantage of the multimodal assay is its flexibility. Each of the three modalities of epigenomic information can be included or excluded from the predictive model. For example, a cancer type with relatively unchanged nucleosome occupancy may benefit only from integrating the remaining two modalities, methylation and fragmentation. Removing the nucleosome organization in this case would prevent confounding and unnecessary complexity. This multimodal approach allows the development of unbiased combinatorial predictive models. All three modalities are captured simultaneously in a single targeted sequencing assay, providing complete flexibility without the expense of multiplexed assays.

Claims

1. 1. A method for determining an epigenetic signature from a sample obtained from an individual, comprising analysing data obtained from a non-destructive methylation sequencing technique performed on said sample obtained from said individual to determine said epigenetic signature, The epigenetic signature comprises the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from the read distribution of one or more base length windows; The present invention includes features obtained from two or more of the following: The method.

2. 1. A method for generating an epigenetic signature from a sample obtained from an individual, comprising: receiving sequencing data obtained from a non-destructive methylation sequencing technique performed on the sample obtained from the individual; Extracting features from the sequencing data, The features include the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from the read distribution of one or more base length windows; the extracting including information from two or more of: inputting the extracted features into a machine learning model; analyzing the features using the machine learning model to generate the epigenetic signature based on a plurality of the features; outputting the generated epigenetic signature; and The method comprising:

3. 1. A method of diagnosing a disease in an individual, comprising: determining an epigenetic signature obtained from a non-destructive methylation sequencing technique performed on a sample obtained from said individual, The epigenetic signature comprises the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from the read distribution of one or more base length windows; the determining including features obtained from two or more of: comparing the epigenetic signature to a disease epigenetic signature and diagnosing the disease in the individual based on the epigenetic signature; The method comprising:

4. 3. The method of claim 1 or 2, further comprising diagnosing disease in the individual based on the epigenetic signature in comparison to a disease epigenetic signature.

5. 1. A method of treating a disease in an individual, comprising: diagnosing the individual as having the disease of claim 3 or claim 4; administering a drug to treat the disease in the individual; and The method comprising:

6. 1. A method for identifying a disease epigenetic signature indicative of an individual having a disease, comprising: receiving sequencing data from a plurality of individuals with the disease and a plurality of individuals without the disease; the receiving, wherein the sequencing data is obtained from a non-destructive methylation sequencing technique performed on a sample obtained from the individual; and Extracting features from the sequencing data, The features include the following profile: a methylation profile comprising information derived from one or more methylation sites; A nucleosome dynamics profile comprising information derived from one or more of: (a) nucleosome position information; (b) nucleosome occupancy; or (c) nucleosome fuzziness; or a fragmentation profile comprising information derived from the read distribution of one or more base length windows; the extracting including information from two or more of: inputting the extracted features into a machine learning model, embedded in the extracted features from each of the plurality of individuals is an associated classification of the individual with or without the disease; training the machine learning model to identify the disease epigenetic signature using the extracted features; outputting said disease epigenetic signature; The method comprising:

7. each of the one or more methylation sites of the methylation profile cg18081940、cg23089825、cg16395183、cg19811148、cg07790615、cg20996351、cg04977528、cg24465685、cg20428713、cg13678973、cg25339566、cg16596317、cg23786625、cg11328303、cg19578660、cg02272851、cg10298052、cg13585930、cg23575688、cg12394201、cg08149193、cg18854419、cg07603330、cg10658542、cg13099890、cg22302985、cg13596497、cg14507533、cg25366582、cg22396555、cg10566012、cg05168229、cg10795666、cg25078444、cg16038120、cg23883632、cg18380808、cg13615592、cg00250422、cg19691260、cg16558770、cg15681853、cg03397724、cg10514097、cg06674117、cg16047279、cg12127472、cg08843809、cg08697732、cg06384763、cg04203646、cg17112426、cg08278741、cg14587524、cg26087117、cg18320766、cg08063125、cg10004780、cg18921980、cg02514318、cg20002504、cg18897632、cg15313459、cg19370054、cg16564824、cg02631468、cg01471196、cg23770904、cg18412834、cg24080247、cg11549874、cg13155421、cg19442495、cg22536150、cg05413061、cg23346462、cg09477895、cg13605674、cg13314965、cg09417547、cg00181669、cg23967169、cg10237419、cg21077559、cg27600205、cg19755714、cg18797590、cg00699993、cg06485940、cg27661394、cg00939495、cg11036833, cg23915769, cg07224726, cg02022733, cg03640756, cg15361 590, cg04598517, cg06782035, cg13954457, cg25482900, cg20952257, cg1 4062050, cg01881524, cg11538641, cg11387340, cg05389236, cg19419054 , cg10575547, cg17240815, cg24772267, cg00920327, cg00772257, cg2625 3500, cg23244488, cg22778435, cg26065247, cg02088996, cg19868631, cg 22280038, cg07803375, cg20230721, cg03333330, cg21517947, cg1040629 5, cg05166490, cg07739205, cg20980783, cg06617456, cg01568998, cg13407456, cg23758305, cg20675505, cg07585876, cg03734437, and cg13410764, The method according to any one of claims 1 to 6, wherein the compound is selected from the group consisting of:

8. The method of any one of claims 1 to 7, wherein the one or more methylation sites of the methylation profile comprise methylation sites in one or more gene promoter regions.

9. The method of any one of claims 1 to 8, wherein the methylation profile comprises quantitative information from at least one of the one or more methylation sites.

10. The method of claim 9 , wherein the quantitative information is based on a β value from the at least one methylation site.

11. The method of claim 9 , wherein the quantitative information is based on a CHALM ratio from the at least one methylation site.

12. The method of any one of claims 1 to 11, wherein the nucleosome dynamics information is based on nucleosomes at genomic loci.

13. The method of any one of claims 1 to 12, wherein the nucleosome position information is based on a window protection score (WPS).

14. The method of claim 13 , wherein the WPS is an average WPS.

15. The method according to any one of claims 1 to 14, wherein the nucleosome occupancy is based on the frequency with which a nucleosome occupies a genomic region.

16. 16. The method of claim 15, wherein the nucleosome occupancy is obtained via normalized read coverage measured in counts per million.

17. The method of any one of claims 1 to 16, wherein the nucleosome fuzziness is based on the deviation of nucleosome positions from preferred nucleosome positions.

18. 18. The method of any one of claims 1 to 17, wherein the fragmentation profile is based on one or more base length windows ranging from 30 to 250 base lengths.

19. 20. The method of claim 18, wherein the base length window is at least 10 bases long.

20. The method according to any one of claims 12 to 17, wherein the nucleosome dynamics information is obtained via DANPOS.

21. The method of any one of claims 1 to 20, wherein the epigenetic signature indicates whether the individual has a disease.

22. The method of any one of claims 1 to 21, wherein the epigenetic signature comprises features from the methylation profile and the nucleosome dynamics profile.

23. The method of any one of claims 1 to 21, wherein the epigenetic signature comprises features from the methylation profile and the fragmentation profile.

24. The method of any one of claims 1 to 21, wherein the epigenetic signature comprises features from the nucleosome dynamics profile and the fragmentation profile.

25. The method of any one of claims 1 to 24, wherein the epigenetic signature comprises features from the methylation profile, the nucleosome dynamics profile, and the fragmentation profile.

26. 26. The method of any one of claims 22, 24, or 25, wherein the nucleosome dynamics profile comprises information derived from nucleosome position information.

27. 27. The method of any one of claims 22 or 24 to 26, wherein the nucleosome dynamics profile comprises information derived from nucleosome occupancy.

28. 28. The method of any one of claims 22 or 24 to 27, wherein the nucleosome dynamics profile comprises information derived from nucleosome fuzziness.

29. The method according to any one of claims 1 to 28, wherein the non-destructive methylation sequencing technique is the EM-seq technique.

30. The method according to any one of claims 1 to 29, wherein the non-destructive methylation sequencing technique is performed based on a target gene position.

31. The method of any one of claims 1 to 30, further comprising performing said non-destructive methylation sequencing technique.

32. The method of any one of claims 1 to 31, wherein the data obtained from the non-destructive methylation sequencing technique comprises a plurality of sequence reads.

33. 33. The method of claim 32, further comprising processing the plurality of sequence reads to remove low quality reads and / or remove adapter contamination and / or filter based on size of the sequence reads.

34. 34. The method of Claim 32 or 33, further comprising aligning the plurality of sequence reads to a reference genome.

35. The method of claim 2 or any one of claims 6 to 34, wherein the machine learning model comprises a support vector machine model, a random forest machine model, or a logistic regression machine model.

36. 36. The method of claim 35, further comprising a cross-validation procedure.

37. The method of any one of claims 1 to 22, wherein the sample is a cell-free DNA sample.

38. The method of any one of claims 1 to 23, further comprising obtaining the sample.

39. The method of any one of claims 1 to 24, wherein the disease is cancer.

40. 40. The method of claim 39, wherein the cancer is colorectal cancer.

41. The method of any one of claims 1 to 40, wherein the individual is a human.

42. The method of any one of claims 1 to 41, wherein the individual is suspected of having the disease.

Images