Method and system for methylation sequencing

The method addresses signal loss and bias in standard methylation sequencing by converting unmethylated cytosine to uracil in single-stranded nucleic acid molecules, enhancing the accuracy and quality of methylation sequencing for disease detection and classification.

KR1020260162276APending Publication Date: 2026-11-10FREENOME HOLDINGS INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
KR1020267011646
Authority / Receiving Office
KR · KR
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-05-08
Filing Date
2024-09-13
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Standard methylation sequencing methods suffer from signal loss and bias, leading to reduced accuracy and quality in nucleic acid methylation analysis, which affects disease detection and classification.

Method used

A method for preparing a sequencing library that involves obtaining single-stranded nucleic acid molecules, ligating a conversion-resistant nucleic acid adapter, and converting unmethylated cytosine to uracil, followed by amplification and sequencing to generate a methylation profile.

Benefits of technology

This approach minimizes signal loss and bias, improving the accuracy and quality of methylation sequencing, enabling better disease detection and classification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure PCT00002_ABST
    Figure PCT00002_ABST
Patent Text Reader

Abstract

The methods and systems provided herein address the limitations of nucleic acid methylation sequencing and disease detection using it by improving quality, sensitivity, and accuracy. The methods provided herein further include minimally destructive methylation switching methods and nucleic acid library preparation techniques that improve methylation sequencing by minimizing signal loss and bias. Additionally, the present invention provides higher accuracy and more complete methylation state information, which enables the generation of higher quality features for use in machine learning models and classifier generation.
Need to check novelty before this filing date? Find Prior Art

Description

Background Technology

[0001] Cross-reference

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 582,948 filed September 15, 2023 and U.S. Provisional Application No. 63 / 644,275 filed May 8, 2024, each of which is incorporated herein by reference in its entirety for all purposes.

[0003] background

[0004] Nucleic acid methylation can indicate tumor characteristics and phenotypic states, and thus holds high potential for utilization in personalized medicine as well as in early disease detection and / or diagnosis. For example, DNA methylation abnormalities can be associated with various stages of cancer, ranging from tumor initiation to cancer progression and metastasis. Abnormal DNA methylation patterns can occur early in the pathogenesis of cancer and thus can provide a mechanism for early cancer detection. These characteristics enable the use of DNA methylation patterns for cancer diagnosis.

[0005] The method and system for preparing a nucleic acid library for methylation sequencing provided herein address the limitations of standard methylation sequencing methods by minimizing signal loss and reducing bias introduced in standard library preparation methods. These methods and systems thereby improve the quality and accuracy of nucleic acid methylation sequencing and its application, for example, in disease detection. More accurate and complete information regarding methylation status enables the generation of higher-quality features for use in machine learning models and classifier generation.

[0006] In one aspect, the present disclosure provides a method for preparing a sequencing library for methylation sequencing of one or more nucleic acid molecules of a biological sample or a derivative thereof, said method

[0007] (a) a step of obtaining a nucleic acid composition, wherein the nucleic acid composition comprises a plurality of single-stranded nucleic acid molecules obtained from or derived from a biological sample;

[0008] (b) a step of ligating a nucleic acid adapter to a single-stranded nucleic acid molecule among a plurality of single-stranded nucleic acid molecules to produce an adapter-ligated nucleic acid molecule, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method; and

[0009] (c) A step of exposing an adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation enrichment method to produce a converted adapter-ligated nucleic acid molecule.

[0010] In some embodiments, the ligation in (b) further comprises treatment with a deoxyribonucleic acid (DNA) ligase. In some embodiments, the ligation in (b) further comprises treatment with a polynucleotide kinase.

[0011] In some embodiments, the nucleic acid adapter comprises a double-stranded oligonucleotide comprising an adapter sequence and an overhang sequence. In some embodiments, the overhang sequence is a 3' overhang sequence. In some embodiments, the overhang sequence comprises a random sequence of oligonucleotides.

[0012] In some embodiments, the nucleic acid adapter comprises one or more methylated cytosine bases. In some embodiments, the nucleic acid adapter does not comprise a methylated cytosine base or an unmethylated cytosine base.

[0013] In some embodiments, the nucleic acid adapter includes a unique molecular identifier. In some embodiments, the unique molecular identifier is configured to enable the measurement of the enrichment efficiency of the methylation conversion method.

[0014] In some embodiments, the method further comprises the step of (a) previously processing a biological sample or a derivative thereof to produce a plurality of single-stranded nucleic acid molecules. In some embodiments, the processing comprises denaturing double-stranded nucleic acid molecules within the biological sample or a derivative thereof. In some embodiments, the denaturation further comprises applying heat to the double-stranded nucleic acid molecules. In some embodiments, the denaturation further comprises applying heat to the double-stranded nucleic acid molecules and then rapidly cooling the denatured single-stranded nucleic acid molecules.

[0015] In some embodiments, the method further comprises the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the potential for forming a nucleic acid duplex. In some embodiments, the method does not comprise the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the potential for forming a nucleic acid duplex. In some embodiments, the binder is a single-stranded nucleic acid binding protein (SSB).

[0016] In some embodiments, the method further comprises the step of amplifying the converted adapter-ligated nucleic acid molecule. In some embodiments, the amplification comprises a polymerase chain reaction (PCR).

[0017] In some embodiments, the method further comprises the step of contacting a converted adapter-ligated nucleic acid molecule or a derivative thereof with a nucleic acid probe to produce a concentrated nucleic acid molecule, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel. In some embodiments, the nucleic acid probe comprises an unmethylated nucleic acid probe. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to methylated cytosine bases at CpG loci from CpG or CH loci of a reference panel.

[0018] In some embodiments, the method further includes the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof.

[0019] In some embodiments, the method further comprises the step of sequencing an enriched nucleic acid molecule or a derivative thereof to generate sequencing data. In some embodiments, the method further comprises the step of analyzing the sequencing data to generate a methylation profile of the nucleic acid molecule of the biological sample or its derivative. In some embodiments, the methylation profile includes hypermethylation and / or hypomethylation analysis. In other embodiments, the methylation profile includes hypermethylation analysis. In yet another embodiment, the methylation profile includes hypomethylation analysis. In some embodiments, the analysis further comprises comparing the sequencing data with a reference sequence.

[0020] In some embodiments, the method further comprises the step of, after (b), performing an extension reaction on the adapter-ligated nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. In some embodiments, the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter.

[0021] In some embodiments, the nucleic acid molecule is deoxyribonucleic acid (DNA). In some embodiments, the DNA is cell-free DNA.

[0022] In some embodiments, the biological sample is a cell-free biological sample. In some embodiments, the cell-free biological sample is a plasma sample.

[0023] In some embodiments, the methylation concentration method includes treatment by one or more enzymes. In some embodiments, the methylation concentration method includes treatment by a ten-eleven translocation (TET) enzyme. In some embodiments, the methylation concentration method does not include treatment with non-sulfite.

[0024] In one aspect, the present disclosure provides a method for preparing a sequencing library for methylation sequencing of nucleic acid molecules of a biological sample or a derivative thereof, said method

[0025] (a) a step of obtaining a nucleic acid composition, wherein the nucleic acid composition comprises a plurality of single-stranded nucleic acid molecules obtained from or derived from a biological sample;

[0026] (b) a step of exposing a single-stranded nucleic acid molecule among a plurality of single-stranded nucleic acid molecules to conditions sufficient to convert unmethylated cytosine to uracil using a methylation concentration method to produce a converted single-stranded nucleic acid molecule; and

[0027] (c) a step of ligating a nucleic acid adapter to a converted single-stranded nucleic acid molecule to produce an adapter-ligated converted nucleic acid molecule, wherein the nucleic acid adapter comprises a nucleic acid that is resistant to base conversion by a methylation enrichment method.

[0028] In some embodiments, the ligation in (c) further comprises treatment with a deoxyribonucleic acid (DNA) ligase. In some embodiments, the ligation in (c) further comprises treatment with a polynucleotide kinase.

[0029] In some embodiments, the nucleic acid adapter comprises a double-stranded oligonucleotide comprising an adapter sequence and an overhang sequence. In some embodiments, the overhang sequence is a 3' overhang sequence. In some embodiments, the overhang sequence comprises a random sequence of oligonucleotides.

[0030] In some embodiments, the nucleic acid adapter comprises one or more methylated cytosine bases. In some embodiments, the nucleic acid adapter does not comprise a methylated cytosine base or an unmethylated cytosine base.

[0031] In some embodiments, the nucleic acid adapter includes a unique molecular identifier. In some embodiments, the unique molecular identifier is configured to enable the measurement of the concentration efficiency of the methylation concentration method.

[0032] In some embodiments, the method further comprises the step of (a) previously treating a sample or a derivative thereof to produce a plurality of single-stranded nucleic acid molecules. In some embodiments, the treatment comprises denaturing double-stranded nucleic acid molecules within the biological sample or a derivative thereof. In some embodiments, the denaturation comprises applying heat to the double-stranded nucleic acid molecules. In some embodiments, the denaturation comprises applying heat to the double-stranded nucleic acid molecules to produce a plurality of single-stranded nucleic acid molecules, and then performing rapid cooling on the plurality of single-stranded nucleic acid molecules.

[0033] In some embodiments, the method further comprises the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the potential for forming a nucleic acid duplex. In some embodiments, the method does not comprise the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the potential for forming a nucleic acid duplex. In some embodiments, the binder is a single-stranded nucleic acid binding protein (SSB).

[0034] In some embodiments, the method further includes the step of amplifying the adapter-ligated converted nucleic acid molecule. In some embodiments, the amplification includes a polymerase chain reaction (PCR).

[0035] In some embodiments, the method further comprises the step of contacting an adapter-ligated converted nucleic acid molecule or a derivative thereof with a nucleic acid probe to produce a concentrated nucleic acid molecule, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel. In some embodiments, the nucleic acid probe comprises an unmethylated nucleic acid probe. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to methylated cytosine bases at CpG loci from CpG or CH loci of a reference panel.

[0036] In some embodiments, the method further comprises the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof. In some embodiments, the method further comprises the step of sequencing the enriched nucleic acid molecule or a derivative thereof to generate sequencing data. In some embodiments, the method further comprises the step of analyzing the sequencing data to generate a methylation profile of the nucleic acid molecule of the biological sample or a derivative thereof. In some embodiments, the methylation profile includes hypermethylation and / or hypomethylation analysis. In other embodiments, the methylation profile includes hypermethylation analysis. In yet another embodiment, the methylation profile includes hypomethylation analysis. In some embodiments, the method for analysis further comprises the step of comparing the sequencing data with a reference sequence.

[0037] In some embodiments, the method further comprises the step of, after (b), performing an extension reaction on the adapter-ligated converted nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. In some embodiments, the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter.

[0038] In some embodiments, the nucleic acid molecule is deoxyribonucleic acid (DNA). In some embodiments, the DNA is cell-free DNA.

[0039] In some embodiments, the biological sample is a cell-free biological sample. In some embodiments, the cell-free biological sample is a plasma sample.

[0040] In some embodiments, the methylation concentration method includes treatment with one or more enzymes. In some embodiments, the methylation concentration method includes treatment with a ten-eleven translocation (TET) enzyme. In some embodiments, the methylation concentration method does not include treatment with non-sulfite.

[0041] In one aspect, the present disclosure is

[0042] (a) a step of producing an adapter-ligated nucleic acid molecule by ligating a nucleic acid adapter to a single-stranded nucleic acid molecule obtained from or derived from a biological sample of a subject, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method;

[0043] (b) a step of exposing an adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation concentration method to produce a converted adapter-ligated nucleic acid molecule;

[0044] (c) a step of amplifying the converted adapter-ligated nucleic acid molecule to produce an amplified nucleic acid molecule;

[0045] (d) a step of generating an enriched nucleic acid molecule by contacting an amplified nucleic acid molecule or a derivative thereof with a nucleic acid probe, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel;

[0046] (e) a step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof;

[0047] (f) a step of comparing the nucleic acid sequence of an enriched nucleic acid molecule or its derivative with a reference nucleic acid sequence; and

[0048] (g) a method comprising the step of training a machine learning model to generate a classifier that distinguishes between subjects having cancer and subjects not having cancer, wherein the machine learning model is trained with a methylation profile generated from (i) a first set of nucleic acid samples from subjects having cancer and (ii) a second set of nucleic acid samples from subjects not having cancer.

[0049] In some embodiments, the methylation profile includes hypermethylation and / or hypomethylation analysis.

[0050] In another embodiment, the methylation profile includes a hypermethylation analysis.

[0051] In another embodiment, the methylation profile includes a hypomethylation analysis.

[0052] In some embodiments, amplification includes a polymerase chain reaction (PCR).

[0053] In some embodiments, the method further includes the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof to a depth greater than 100x.

[0054] In some embodiments, the nucleic acid probe comprises unmethylated nucleic acid. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. In some embodiments, the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to methylated cytosine bases at CpG loci from CpG or CH loci of a reference panel.

[0055] In some embodiments, the method further comprises the step of sequencing an enriched nucleic acid molecule or a derivative thereof to generate sequencing data. In some embodiments, the method further comprises the step of analyzing the sequencing data to generate a methylation profile of the nucleic acid molecule of the biological sample or its derivative. In some embodiments, the methylation profile includes a hypermethylation and / or hypomethylation analysis. In other embodiments, the methylation profile includes a hypermethylation analysis. In yet another embodiment, the methylation profile includes a hypomethylation analysis.

[0056] In some embodiments, the method further comprises the step of (a) subsequently performing an extension reaction on the adapter-ligated converted nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. In some embodiments, the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter.

[0057] In some embodiments, the single-stranded nucleic acid molecule is deoxyribonucleic acid (DNA). In some embodiments, the DNA is cell-free DNA.

[0058] In some embodiments, the biological sample is a cell-free biological sample. In some embodiments, the cell-free biological sample is a plasma sample.

[0059] In some embodiments, the methylation concentration method includes treatment with one or more enzymes. In some embodiments, the methylation concentration method includes treatment with a ten-eleven translocation (TET) enzyme. In some embodiments, the methylation concentration method does not include treatment with non-sulfite.

[0060] In some embodiments, the reference panel includes a CpG or CH locus associated with the transcription start site.

[0061] In some embodiments, the method further includes the step of identifying the origin organization for the nucleic acid molecule.

[0062] In some embodiments, the method further includes the step of identifying the genomic location and fragment length for a nucleic acid molecule.

[0063] In some embodiments, the machine learning model is trained using feature inputs selected from the group consisting of base-wise methylation percentage for CpG, base-wise methylation percentage for CHG, base-wise methylation percentage for CHH, the number or ratio of fragments with different numbers or ratios of methylated CpGs observed within a region, conversion efficiency, low-methylation blocks, methylation levels for CpG, methylation levels for CHH, methylation levels for CHG, fragment length, fragment midpoint, methylation levels for chrM, methylation levels for LINE1, methylation levels for ALU, dinucleotide coverage, uniformity of coverage, global average CpG coverage, CpG islands, CGI shelf, and average coverage on the CGI shore.

[0064] In some embodiments, a classifier for distinguishing between subjects with cancer and subjects without cancer includes a set of measurements representing methylation profiles from methylation sequencing data from subjects with cancer and subjects without cancer, the measurements are used to generate a set of features corresponding to characteristics of the methylation profiles, the set of features is processed by a machine learning or statistical model, and the machine learning or statistical model provides a feature vector useful as a classifier for distinguishing between a group of subjects with cancer and a group of subjects without cancer.

[0065] In some embodiments, the method

[0066] (a) A step of obtaining a methylation profile of a biological sample by assaying using a methylation enrichment method;

[0067] (b) a step of classifying the methylation profile of a biological sample by a trained machine learning algorithm as indicating the presence of cancer in the subject; and

[0068] (c) If the trained machine learning algorithm classifies the biological sample as negative for cancer at a specified confidence level, it further includes a step of outputting a report identifying the biological sample as negative for cancer.

[0069] In some embodiments, the methylation profile includes hypermethylation and / or hypomethylation analysis.

[0070] In another embodiment, the methylation profile includes a hypermethylation analysis.

[0071] In another embodiment, the methylation profile includes a hypomethylation analysis.

[0072] In some embodiments, the method

[0073] (a) A step of determining the baseline methylation profile of a biological sample of a subject in a baseline methylation state;

[0074] (b) determining the test methylation profile of a biological sample of the subject at one or more time points after the baseline methylation status; and

[0075] (c) A step of determining a change in the test methylation profile compared to the baseline methylation profile, wherein the change indicates a change in the subject's minimum residual disease state, further comprising the step of determining the change in the test methylation profile compared to the baseline methylation profile.

[0076] In some embodiments, the methylation profile includes hypermethylation and / or hypomethylation analysis.

[0077] In another embodiment, the methylation profile includes a hypermethylation analysis.

[0078] In another embodiment, the methylation profile includes a hypomethylation analysis.

[0079] In some embodiments, cancer includes two or more of colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer.

[0080] In another embodiment, the cancer is colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer.

[0081] In one embodiment, the cancer is colorectal cancer. In another embodiment, the cancer is lung cancer. In another embodiment, the cancer is pancreatic cancer. In another embodiment, the cancer is liver cancer.

[0082] In some embodiments, the least residual disease state is selected from the group consisting of response to treatment, tumor load, residual tumor after surgery, recurrence, secondary screening, primary screening, and cancer progression.

[0083] In one aspect, the present disclosure comprises the steps of: ligating a nucleic acid adapter to a single-stranded nucleic acid molecule obtained from or derived from a biological sample of an object to produce an adapter-ligated nucleic acid molecule, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method; exposing the adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation enrichment method to produce a converted adapter-ligated nucleic acid molecule; amplifying the converted adapter-ligated nucleic acid molecule to produce an amplified nucleic acid molecule; contacting the amplified nucleic acid molecule or a derivative thereof with a nucleic acid probe to produce an enriched nucleic acid molecule, wherein the nucleic acid probe comprises a nucleic acid sequence at least partially complementary to a CpG or CH locus of a reference panel; determining the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof; and comparing the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof with a reference nucleic acid sequence. The present invention provides a method comprising the step of training a machine learning model to generate a classifier that distinguishes between subjects having indications and subjects not having indications, wherein the machine learning model is trained with a methylation profile generated from (i) a first set of nucleic acid samples from subjects having indications and (ii) a second set of nucleic acid samples from subjects not having indications.

[0084] In some embodiments, the method further comprises the steps of: obtaining a methylation profile of a biological sample by testing by a methylation enrichment method; classifying the methylation profile of the biological sample as indicating the presence of an indication in a subject by a trained machine learning algorithm; and outputting a report identifying the biological sample as negative for an indication when the trained machine learning algorithm classifies the biological sample as negative for an indication at a specified confidence level.

[0085] In some embodiments, the method further comprises the steps of: determining a baseline methylation profile of a biological sample of a subject in a baseline methylation state; determining a test methylation profile of a biological sample of a subject at one or more time points after the baseline methylation state; and determining a change in the test methylation profile compared to the baseline methylation profile, wherein the change indicates a change in the minimum residual disease state of the indication in the subject.

[0086] In some embodiments, the least residual disease state is selected from a group consisting of response to treatment, relapse, secondary screening, primary screening, and progression of indications.

[0087] In some embodiments, the methylation profile includes a hypermethylation analysis and / or a hypomethylation analysis.

[0088] In some embodiments, indications include intestinal-related diseases, immune-mediated inflammatory diseases, neurological diseases, kidney diseases, prenatal diseases, or metabolic diseases.

[0089] In one aspect, the present disclosure is

[0090] (a) A computer-readable media product comprising a classifier of the present disclosure,

[0091] A classifier comprises a set of measurements representing methylation profiles from methylation sequencing data from subjects with cancer and subjects without cancer, wherein the set of measurements is used to generate a set of features corresponding to characteristics of the methylation profiles from subjects with cancer and subjects without cancer, the set of features is processed by a machine learning or statistical model, and the machine learning or statistical model provides a feature vector useful for distinguishing subjects with cancer and subjects without cancer, a computer-readable medium product; and

[0092] (b) one or more processors for executing instructions stored on a computer-readable media product

[0093] Provides a system that includes

[0094] In some embodiments, the classifier is selected from a group consisting of a linear discriminant analysis (LDA) classifier, a quadratic discriminant analysis (QDA) classifier, a support vector machine (SVM) classifier, a random forest (RF) classifier, a linear kernel support vector machine classifier, a first-order polynomial kernel support vector machine classifier, a second-order polynomial kernel support vector machine classifier, a ridge regression classifier, an elastic net algorithm classifier, a sequential minimum optimization algorithm classifier, a naive Bayes algorithm classifier, and a non-negative matrix decomposition (NMF) prediction algorithm classifier.

[0095] In some embodiments, the system is further configured to perform any of the above methods.

[0096] In some embodiments, the system includes one or more processors configured to perform any of the above methods.

[0097] In some embodiments, the system includes modules that each perform the operation of any of the methods.

[0098] In one aspect, the present disclosure provides a cancer detection kit comprising a reagent for performing any of the above methods and instructions for detecting a cancer signal.

[0099] In some embodiments, the reagent is selected from the group consisting of primer sets, PCR reaction components, sequencing reagents, methylation enrichment reagents and library preparation reagents.

[0100] In some embodiments, a machine learning model is trained using training data obtained from training biological samples, wherein a first subset of training biological samples is identified as corresponding to subjects having a cell proliferative disorder, and a second subset of training biological samples is identified as corresponding to subjects not having a cell proliferative disorder.

[0101] In some embodiments, a classifier is provided in a system for detecting cell proliferative disorders, and the system comprises a) a computer-readable medium comprising a classifier operable to classify objects based on a methylation signature panel; and b) one or more processors for executing instructions stored in the computer-readable medium.

[0102] In some embodiments, the system comprises a classification circuit composed of a machine learning classifier selected from a group consisting of a deep learning classifier, a neural network classifier, a linear discriminant analysis (LDA) classifier, a quadratic discriminant analysis (QDA) classifier, a support vector machine (SVM) classifier, a random forest (RF) classifier, a K-nearest neighbors classifier, a linear kernel support vector machine classifier, a first-order or second-order polynomial kernel support vector machine classifier, a ridge regression classifier, an elastic net algorithm classifier, a sequential minimum optimization algorithm classifier, a naive Bayes algorithm classifier, and a principal component analysis classifier.

[0103] In some embodiments, the method further includes the step of presenting a report or a graphical user interface to a user's electronic device. In some embodiments, the user is an object, an entity, or a patient.

[0104] In some embodiments, the method further includes a step of determining the presence or susceptibility of a cell proliferative disorder in a subject, individual, or patient.

[0105] In some embodiments, the trained algorithm (e.g., machine learning model or classifier) ​​includes a supervised or semi-supervised machine learning algorithm. In some embodiments, the supervised machine learning algorithm includes a deep learning algorithm, a support vector machine (SVM), a neural network, or a random forest.

[0106] In some embodiments, the method further comprises the step of providing a subject with a second diagnostic test or procedure, such as a colonoscopy, CT scan, MRI, ultrasound, or other non-invasive detection test or procedure, including but not limited to the methylation signature profile or analysis described herein, at least partially.

[0107] In some embodiments, the method further comprises the step of providing a therapeutic intervention to a subject, such as a therapeutic intervention (e.g., chemotherapy, radiotherapy, immunotherapy, or surgery) to treat a patient with a proliferative disorder, or administering a therapeutic agent to the subject, based at least partially on the methylation signature profile or analysis described herein.

[0108] In some embodiments, the method further comprises the step of monitoring the presence or susceptibility to a cell proliferative disorder, wherein the monitoring comprises evaluating the presence or susceptibility to a cell proliferative disorder of a subject at a plurality of time points, and at least the evaluation is based on the presence or susceptibility to a cell proliferative disorder determined at each of the plurality of time points.

[0109] In some embodiments, differences in the assessment of the presence or susceptibility of a subject to a cell proliferative disorder between multiple time points represent one or more clinical indications selected from the group consisting of (i) diagnosis of the presence or susceptibility of the subject to a cell proliferative disorder, (ii) prognosis of the presence or susceptibility of the subject to a cell proliferative disorder, and (iii) efficacy or non-efficacy of a course of treatment to treat the presence or susceptibility of the subject to a cell proliferative disorder.

[0110] In some embodiments, the method further comprises the step of stratifying the cell proliferative disorder of a subject by determining the subtype of the subject's cell proliferative disorder among a plurality of distinct subtypes or stages of cell proliferative disorder using a trained algorithm.

[0111] Another aspect of the present disclosure provides a non-transient computer-readable medium comprising machine-executable code that implements any of the methods described above or elsewhere in the present invention when executed by one or more computer processors.

[0112] Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto. The computer memory comprises machine-executable code that implements any of the methods described above or elsewhere in the present invention when executed by one or more computer processors.

[0113] Another aspect of the present disclosure provides a computer-readable medium comprising a) a classifier for distinguishing between a population of subjects having a cell proliferative disorder and subjects not having a cell proliferative disorder based on a methylation signature panel using a machine learning model; and b) a system comprising one or more processors for executing instructions stored in the computer-readable medium.

[0114] Additional aspects and benefits of the present disclosure will be readily understood by those skilled in the art from the following detailed description, and only exemplary embodiments of the present disclosure are shown and described. As will be apparent, other different embodiments of the present disclosure are possible, and various details thereof may be modified in various obvious respects without departing from the scope of the present disclosure. Accordingly, the drawings and description should be considered as illustrative in nature and not limiting.

[0115] Inclusion by reference

[0116] All disclosures, patents, and patent applications mentioned in this specification are incorporated herein by reference to the same extent as each individual disclosure, patent, or patent application is specifically and individually indicated as being incorporated by reference. Where any disclosure or patent or patent application incorporated by reference contradicts the disclosures contained herein, this specification is intended to take precedence over and / or precede any such contradictory material. Brief explanation of the drawing

[0117] Fig. 1 The present invention provides a schematic diagram of an exemplary single-stranded DNA (ssDNA) library preparation method as described herein, compared to a standard double-stranded DNA (dsDNA) library preparation method. Fig. 2 It provides a schematic diagram of an exemplary ssDNA library manufacturing workflow. Fig. 3 It provides a comparison of examples of methylation signals according to distance from the 3' end of hypermethylated DNA fragments using the ssDNA library preparation method versus the dsDNA library preparation method. Fig. 4 It provides a comparison of examples of captured reads according to the length of DNA fragments detected using the ssDNA library preparation method versus the dsDNA library preparation method. Fig. 5This provides a comparison of examples of hypermethylation scores of cell-free DNA (cfDNA) extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 6 This provides a comparison of examples of hypermethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 7 The present invention provides a comparison of examples of hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 8 The present invention provides a comparison of examples of hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 9a and 9b The present invention provides schematic diagrams of examples of two methods in which an ssDNA library generated by the method disclosed herein can be converted into a dsDNA library after ligation. After a cleanup operation to remove excess adapters after adapter ligation, a primer extension reaction may be initiated to generate a strand complementary to the adapter-ligated nucleic acid by annealing a single-strand-specific (SSS) primer to the 3' end of the adapter-ligated nucleic acid ( Fig. 9aAlternatively, after ligating an incompletely overlapping splint adapter to ssDNA, an SSS primer can be annealed at the 3' end of the adapter-ligated nucleic acid to initiate a primer extension reaction and remove the excess adapter ( Fig. 9b This method enables nucleic acid amplification without an initial cleanup operation to remove excess adapters. Fig. 10 After converting an ssDNA library into a dsDNA library, the document provides a comparison of examples of methylated cytosine (mC) protection rates in ssDNA using an ssDNA library workflow with second strand synthesis (SSS) and an ssDNA library workflow without second strand synthesis (SOP). Fig. 11 It shows an example of a computer system programmed or otherwise configured to implement the method provided herein. Specific details for implementing the invention

[0118] The present invention provides methods and systems related to the preparation of a library of methylated regions and sequencing for the methylation profiling of nucleic acids, such as cell-free deoxyribonucleic acid (cfDNA). The methods and systems support practical applications as described herein by addressing the limitations of existing library preparation and subsequent methylation sequencing and profiling of nucleic acids derived from biological samples by minimizing signal loss, reducing bias, and improving coverage, uniformity of coverage, resolution, and accuracy of methylation data. The sequencing data generated from the methods provided herein may be useful for applications using methylation profiling data to classify or stratify populations. Such classification or stratification of populations may include, for example, identifying and / or detecting individuals with a disease, dividing stages of disease progression (including detection of minimal residual disease), or determining the response of individuals to specific treatments for the disease.

[0119] Methylation analysis can be coupled with DNA sequencing to determine the likelihood that a sample is normal, tumor-derived, or disease-beneficial. For example, the relative abundance of methylated or unmethylated DNA fragments mapped to or aligned with specific genomic regions can be used to detect or determine disease potential. Such sequencing methods may involve the preparation of standard DNA libraries, for instance, in which enzymatic end repair processes are utilized before attaching sequencing adapters to double-stranded DNA (dsDNA) fragments. However, these processes may fail to preserve fragment ends and may fail to capture nicked or single-stranded DNA, which can lead to a reduction in the overall methylation signal and a bias in the fragment length distribution. Consequently, end repair can result in a relative undercounting of methylated molecules and an overcounting of unmethylated molecules, both of which can lead to an underestimation of the signal contribution from tumor-derived DNA. Consequently, the sensitivity of these methods for cancer detection may be reduced.

[0120] The method and system of the present invention may include the steps of preparing a single-stranded DNA (ssDNA) sequencing library and applying a sample from the ssDNA library to a methylation question treatment method so that unmethylated cytosine bases within the ssDNA are converted to uracil bases. The method may include the steps of denaturing double-stranded DNA (dsDNA) into ssDNA and attaching a conversion-resistant nucleic acid adapter to the ssDNA, which enables intramolecular methylation evaluation in a manner that does not require end repair and yields improved and more accurate methylation analysis results that are not hindered by the undercounting of methylated molecules occurring in the dsDNA library preparation method. After attaching the conversion-resistant nucleic acid adapter to the ssDNA, the adapter-ligated ssDNA may be applied to an enzymatic methylation (EM) conversion or non-sulfite treatment method to provide base-level resolution of DNA methylation. In one embodiment, adapter ligation is performed prior to methylation base conversion. In another embodiment, the methylation question treatment method is performed before adapter ligation. In certain cases, performing adapter ligation before methylation base conversion may be preferable to performing adapter ligation after methylation base conversion. Additionally, since chemical treatment can result in more extensive DNA damage and molecular loss than enzymatic methods, enzymatic conversion methods may be preferable to chemical conversion methods (e.g., non-sulfite treatment).

[0121] In the case of a method in which ssDNA library preparation is coupled with EM conversion, the adapter-ligated sequence can be amplified and sequenced after EM conversion. However, standard adapters may not be suitable for this method. Therefore, the library adapter sequence described in the method of the present invention may differ from the adapter sequence used in standard double-strand sequencing library preparation in terms of cytosine residue content and modification so that the adapter residues are resistant to conversion by chemical or enzymatic conversion methods. Additionally, in certain embodiments, the library adapter sequence may include an additional unique molecular identifier (UMI) sequence that can be used to monitor the efficiency of the EM conversion reaction.

[0122] For example, various ssDNA library preparation methods, including adaptase-based ssDNA and terminal deoxyribonucleotidyl transferase (TdT) assisted adenylate connector-mediated ssDNA (TACS) library preparation, may be used in combination with methylation question processing methods. Additional ssDNA library preparation approaches may include the use of T4-ligase-mediated single-strand nucleic acid ligation using splinted oligonucleotides. These approaches may include splinted adapter tagging (SPLAT) and single-reaction single-strand library preparation (SRSLY) or single-strand adapter library preparation (SALP), which can perform adapter ligation by utilizing ssDNA binding proteins (SSB) and removing terminal repairs.

[0123] Adaptase-based, TACS, and SPLAT methods for ssDNA library preparation may involve additional cleanup operations that are not present in splint adapter-based ssDNA library preparation methods. This cleanup can be associated with DNA loss, particularly the loss of shorter DNA fragments that carry methylation signals critical for disease detection. Additional signal loss may occur if these methods are coupled with methylation analysis methods, such as EM conversion or bisulfite treatment, that may be performed prior to library preparation. Purification and recovery processes for EM conversion and bisulfite treatment may result in greater DNA loss when processing shorter DNA without adapters than when processing longer DNA ligated with adapters. Furthermore, as EM conversion or bisulfite treatment converts unmethylated cytosines within the adapter sequence, numerous additional design considerations may be required to ensure these adapters are suitable for downstream processes (e.g., amplification and sequencing). These additional design constraints may also apply to adapters used in splint ligation processes such as SPLAT or SRSLY, which may not be inherently suitable for enzymatic conversion downstream of library preparation. Furthermore, for example, adding SSB processing in an SRSLY workflow may be unnecessary for cfDNA library preparation and may impede the performance of the methylation conversion process.

[0124] The method described herein may include a method for preparing an ssDNA library suitable for a methylation conversion workflow without SSB processing. Such a method may be useful for determining the unique methylation status of cfDNA via sequencing by minimizing molecular loss and eliminating bias introduced in standard library preparation methods. After denaturing dsDNA into ssDNA, the ssDNA may be ligated to an adapter designed not to be affected by subsequent enzymatic or chemical processing used for DNA methylation detection. Such adapter sequences may also include a module for evaluating the performance or efficiency level of such enzymatic or chemical processing methods. After processing, the ligated and processed DNA may be amplified and sequenced. Information obtained from such sequencing, including but not limited to the identity, quantity, size, precise end, and relative methylation status of the DNA fragment, may be used to determine whether the cfDNA is of cancer origin.

[0125] Compared to conventional library preparation methods coupled with methyl conversion methods, the method described herein offers various advantages. The method preserves methylation signals more faithfully than conventional methods from DNA extraction to library preparation, methyl conversion, and sequencing. For example, the end repair process in dsDNA library preparation involves excising and inserting DNA from 3' and 5' overhangs, respectively. As a result of these modifications, actual DNA methylation is underestimated in the sequencing of dsDNA EM-converted libraries. Since DNA fragment methylation status alone is sufficient to identify cfDNA derived from cancer patients, this reduced fidelity decreases the sensitivity of these cancer detection methods. Similarly, reduced methylation reads in sequenced dsDNA EM-converted libraries diminish the utility of low or absent DNA methylation as a cancer detection signal. All these problems are overcome by utilizing ssDNA library preparation and EM conversion.

[0126] The above method can enable the sequencing of DNA that is typically lost in dsDNA library preparation, including single-stranded fragments from short ssDNA and nicked dsDNA. The identity and relative amount of these fragments can serve as a tool to distinguish DNA from cancer patients from DNA from healthy patients.

[0127] The library preparation method of the present invention preserves the ends of DNA fragments that may be lost or altered during the end repair process required in traditional dsDNA library preparation methods. Fragment-end data can serve as a source of signals used to identify tumor-derived DNA.

[0128] Conventional ssDNA library preparation methods require capturing DNA fragments after non-sulfite or EM conversion to preserve adapter sequences for downstream processes (e.g., sequencing). Since these DNA samples must undergo a cleanup process between conversion and adapter ligation, they may suffer more extensive DNA loss than the method of the present invention, which first cleans up larger adapter-ligated DNA molecules after ligation. This process improves molecular recovery rates and, consequently, enhances the ability of these assays, particularly for smaller fractions of tumor-derived signals within cell-free DNA populations. The method of the present invention can also eliminate unnecessary reagents, such as SSB processing, which can impair downstream assay performance.

[0129] Fig. 1 The present invention provides a schematic diagram of an example of dsDNA library preparation as opposed to ssDNA library preparation as described in the method of the present invention. In this example, the input cfDNA may vary in strand state (e.g., single strand or double strand), presence of nicks, and presence and length of 5' and 3' overhangs.

[0130] Fig. 1As illustrated in [Figure], the ssDNA library preparation method preserves fragment ends and can capture DNA that is not captured using the standard dsDNA library preparation method. dsDNA library preparation workflow ( Fig. 1 In the bottom left, a terminal repair enzyme is used to remove or extend the ends of the dsDNA fragment (pink open box) before the Y-shaped adapter is ligated onto the dsDNA. This terminal repair process can result in the exclusion of methylation signals from these fragment ends. ssDNA library manufacturing workflow ( Fig. 1 In the bottom right, a conversion-tolerant ssDNA fit adapter is directly ligated to the ssDNA fragment. Since the end repair process is excluded from the ssDNA library manufacturing workflow, the fragment ends and the methylation signals arising therefrom are preserved. As described herein, the ssDNA library manufacturing method can capture essentially all input DNA (after denaturation of dsDNA), including dsDNA end repair and DNA fragments and regions (pink lines) lost during library manufacturing. Furthermore, the ssDNA library manufacturing method described herein captures intrinsic DNA methylation while excluding end repair artifacts originating from the dsDNA library manufacturing method.

[0131] Fig. 2 This provides a schematic diagram of an exemplary ssDNA library manufacturing workflow. After nucleic acid extraction, ssDNA is processed by a method for manufacturing an ssDNA library using a conversion-resistant splinted adapter, followed by EM conversion, amplification, target capture, and sequencing.

[0132] Fig. 3This provides a comparison of examples of methylation signals as a function of distance from the 3' fragment end using ssDNA library preparation versus dsDNA library preparation. When measuring CpG methylation in fragments overlapping with hypermethylated genomic regions, fragments from the dsDNA library exhibit a loss of methylation signal near the 3' end of the fragment due to end repair (gray line (dsDNA)). In contrast, the methylation signal is uniformly high across the length of the fragments within the ssDNA library (green line (ssDNA)).

[0133] Fig. 4 This provides a comparison of examples of captured reads according to fragment length using ssDNA library preparation versus dsDNA library preparation. As shown, the ssDNA library preparation method can capture shorter cfDNA fragments corresponding to short ssDNA that are not captured by the dsDNA library preparation method.

[0134] Fig. 5This provides a comparison of examples of hypermethylation model scores for cfDNA extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) patient samples detected using the ssDNA library preparation described herein against other dsDNA library preparation methods. The hypermethylation rates of cfDNA derived from selected genomic regions distinguished AA and CRC patient-derived cfDNA from cancer-negative cfDNA when processed using both the ssDNA library preparation described herein and other dsDNA library preparation methods. This experiment demonstrates that the ssDNA library preparation method described herein yields methylation analysis (e.g., hypermethylation scores) equivalent to other dsDNA library preparation methods. This workflow is equivalent to end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, when using DNA hypermethylation analysis.

[0135] Fig. 6This provides a comparison of examples of hypermethylation model scores for cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) patient samples detected using the ssDNA library preparation method described herein against other dsDNA library preparation methods. The hypermethylation rates of cfDNA derived from selected genomic regions distinguished cfDNA derived from liver cancer, lung cancer, and pancreatic cancer patients from cancer-negative cfDNA when processed using both the ssDNA library preparation method described herein and other dsDNA library preparation methods. This experiment demonstrates that the ssDNA library preparation method described herein yields methylation analysis (e.g., hypermethylation scores) equivalent to other dsDNA library preparation methods. Therefore, the ssDNA library preparation workflow described herein is equivalent to end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, when using DNA hypermethylation analysis.

[0136] Fig. 7This provides a comparison of examples of hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) patient samples detected using the ssDNA library preparation described herein versus the dsDNA library preparation. The hypomethylation rates of cfDNA derived from selected genomic regions distinguished AA and CRC patient-derived cfDNA from cancer-negative cfDNA when processed using the ssDNA library preparation, but failed to distinguish them when processed using the dsDNA library preparation. This experiment demonstrates that the ssDNA library preparation method described herein optimizes recovery, methylation analysis, quantification, and targeted capture. This workflow can demonstrate superior performance compared to end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, using DNA hypomethylation analysis.

[0137] Fig. 8 The present invention provides a comparison of examples of hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 8 As illustrated in, The hypomethylation rate of cfDNA derived from selected genomic regions distinguished cfDNA derived from liver, lung, and pancreatic cancer patients from cancer-negative cfDNA when processed using ssDNA library preparation, but failed to distinguish them when processed using dsDNA library preparation. This experiment provides additional evidence that the ssDNA library preparation method described herein optimizes recovery, methylation analysis, quantification, and targeted capture across multiple cancer types, and also demonstrates that this workflow can outperform end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, using DNA hypomethylation analysis.

[0138] Fig. 9a and Fig. 9b This exemplifies a method in which an ssDNA library can be converted into a dsDNA library after ligation. The ssDNA library may undergo a cleanup operation after ligation and is subsequently added to a polymerization reaction comprising a reaction buffer, nucleotides, an oligonucleotide primer complementary to a portion of the 3' adapter sequence, and a polymerase ( Fig. 9a Alternatively, an ssDNA library can be generated using a splint adapter sequence so that it can be amplified without a cleanup operation to remove excess adapters ( Fig. 9b ).

[0139] Fig. 10After converting an ssDNA library to a dsDNA library, a comparison is provided for examples of protection rates for methylated cytosine (mC) in ssDNA using an ssDNA library workflow with second strand synthesis (SSS) and an ssDNA library workflow without second strand synthesis (SOP). The protection rate depends on the identity of the base following the methylated cytosine (e.g., CpA, CpG, CpC, or CpT). cfDNA_SOP is a cell-free DNA sample using the ssDNA library workflow without second strand synthesis; ct_SOP is a manipulated DNA sample using the ssDNA library workflow without second strand synthesis; cfDNA_SSS is a cell-free DNA sample using an ssDNA library containing second strand synthesis; and ct_SSS is a manipulated DNA sample using an ssDNA library containing second strand synthesis.

[0140] definition

[0141] As used herein, singular terms include both singular and plural referents unless the context clearly indicates otherwise.

[0142] As used herein, the terms "plasma cell-free DNA," "circulating free DNA," "cell-free DNA," or "cfDNA" generally refer to DNA molecules circulating in the non-cellular portion of the blood. Circulating nucleic acids in the blood may originate from necrotic or apoptotic cells that indicate diseases such as cancer. In cancer, circulating DNA exhibits characteristic signs of the disease, including mutations in oncogenes and microsatellite mutations. Such circulating DNA may be referred to as circulating tumor DNA (ctDNA). Viral genome sequences, DNA, or RNA in plasma are potential biomarkers for disease.

[0143] In some embodiments, the cell-free fraction of blood is preferably serum or plasma. As used herein, the term “cell-free fraction” of a biological sample generally refers to a fraction of a biological sample that is substantially free of cells. As used herein, the term “substantially free of cells” may refer to a product from a biological sample containing fewer than about 20,000 cells per mL, fewer than about 2,000 cells per mL, fewer than about 200 cells per mL, or fewer than about 20 cells per mL. Genomic DNA (gDNA) refers to unfragmented DNA released from leukocytes that contaminates the cell-free fraction of blood. To mitigate gDNA contamination of the sample, a highly controlled sample processing workflow may be implemented, and the sample may be screened for the presence of gDNA.

[0144] As used herein, the terms “to detect,” “detecting,” or “detecting” a condition or result generally include detecting the presence of an indication (e.g., cancer), detecting a condition or result, or detecting a predisposition to a condition or result.

[0145] As used herein, the term “diagnose” or “diagnosis” of a condition or outcome generally includes predicting or diagnosing a condition or outcome, determining predisposing factors for a condition or outcome, monitoring a patient’s treatment, diagnosing a patient’s response to treatment, and the prognosis, progression, and response to a specific treatment of a condition or outcome.

[0146] As used herein, the term “position” generally refers to a nucleotide position within an identified strand of a nucleic acid molecule.

[0147] As used herein, the term “nucleic acid” generally refers to DNA, RNA, DNA / RNA chimera, or hybrid that may be single-stranded (ss) or double-stranded (ds). Nucleic acid may be genomic or derived from the genome of a eukaryotic or prokaryotic cell, or may be synthesized, cloned, amplified, or reverse transcribed. In certain embodiments of the method and composition, nucleic acid preferably refers to genomic DNA as required by the context.

[0148] As used herein, unless otherwise specified, the term “modified cytosine” generally refers to 5-methylcytosine (5mC), 5-hydroxymethylcytosine (5hmC), formyl-modified cytosine, carboxy-modified cytosine, 5-carboxylcytosine (5caC), or cytosine modified by any other chemical group.

[0149] As used herein, the terms “methylcytosine deoxygenase,” “deoxygenase,” or “oxygenase” generally refer to an enzyme that converts 5mC to 5hmC. Non-limiting examples of methylcytosine deoxygenase include, for example, ten-eleven translocation (TET) enzymes, e.g., TET1, TET2, TET3, Naegleria TET, and genetically engineered or non-genetically engineered variants thereof. TET2 is an example of a methylcytosine deoxygenase that oxidizes at least 90%, at least 92%, at least 94%, at least 96%, at least 98%, or at least 99% of all 5mC.

[0150] As used herein, the terms “methylation conversion method,” “methylation enrichment method,” or “methylation treatment method” generally refer to a method of exposing a nucleic acid molecule to conditions sufficient to convert unmethylated cytosine within the nucleic acid molecule into uracil. The above method is useful for distinguishing between methylated cytosine and unmethylated cytosine in a nucleic acid molecule.

[0151] As used herein, the terms “enzymatic methylation,” “enzymatic methyl,” “EM conversion,” or “EM-seq” generally refer to a method of exposing a nucleic acid molecule to conditions sufficient to convert unmethylated cytosine within the nucleic acid molecule to uracil by treating it with one or more enzymes, e.g., TET enzymes. In some cases, the method does not involve treatment with non-sulfite (e.g., chemical treatment).

[0152] As used herein, the terms “transition-resistant adapter,” “transition-resistant primer,” “transition-allowing adapter,” or “transition-allowing primer” generally refer to a nucleic acid molecule used as an adapter or primer, respectively. Such nucleic acid molecules are resistant to methylation conversion or alteration by methylation enrichment methods (e.g., “deamination-resistant modified cytosine” or methylation-converting agent-resistant nucleotide).

[0153] As used herein, the term “deamination-resistant modified cytosine” refers to one or more modified cytosine nucleotides in a conversion-resistant adapter that are not chemically or enzymatically altered to alter the base pairing specificity of nucleotide bases when treated with a methylation-converting agent. As a non-limiting example, propynyl-C and pyrrolo-C are deamination-resistant modified cytosines that are conversion-resistant nucleotides and may be included in the conversion-resistant adapter used in the library preparation method disclosed herein. These deamination-resistant modified cytosines in the conversion-resistant adapter are not deaminated when exposed to sodium bisulfite or a methylation-converting enzyme (e.g., APOBEC-like enzyme).

[0154] As used herein, the term “methylation convertor-resistant nucleotide” generally refers to a nucleotide containing a nucleic acid base that is not chemically modified to alter the base pairing specificity of the nucleotide base when treated with a methylation convertor. Methylation convertor-resistant nucleotides can be incorporated by nick translation enzymes in primer extension reactions. For example, 5-methylcytosine is a conversion-resistant nucleotide that can be used with sodium bisulfite or enzymatic methylation conversion. Thus, 5-methylcytosine is not deaminated when exposed to sodium bisulfite or methylation convertor enzymes. Instead of incorporating modified nucleotide bases to prevent base conversion, conversion-allowing adapters or conversion-allowing primers incorporate only unmodified bases to enable total base conversion during the conversion reaction for methylation sequencing. In the adapter or primer DNA sequence, “unmodified bases” refer to the usual guanine, adenine, cytosine, and thymine.

[0155] As used herein, the term “cytidine deaminase” generally refers to an enzyme that deaminates cytosine (C) to form uracil (U). Non-limiting examples of cytidine deaminases include apolipoprotein B mRNA-editing enzymes, such as APOBEC3A, and cytidine deaminases of the catalytic polypeptide (APOBEC) family. In any embodiment, the cytidine deaminase described herein may have an amino acid sequence that is at least 90% identical (e.g., at least 95% identical) to the amino acid sequence of GenBank registry number AKE33285.1, which is the sequence of human APOBEC3A. In some embodiments, the cytidine deaminase described herein converts unmodified cytosine to uracil with an efficiency of at least 95%, at least 96%, at least 97%, at least 98%, or at least 99%, preferably at least 99%.

[0156] As used herein, the terms "glucosyltransferase" or "GT" generally refer to an enzyme that catalyzes the transfer of a beta-D-glucosyl or alpha-D-glucosyl residue from UDP-glucose to a 5hmC residue to form 5hmC. APOBEC can convert 5hmC to U at a lower rate compared to converting C or 5mC to U. An example of GT is T4-betaGT (βGT). In one example, GT may be used in combination with deoxygenase. This combination ensures that the deamination of 5hmC is blocked, so that less than 5%, less than 3%, or less than 1% of 5hmC is converted to uracil by deaminase. In another example, GT can be used in combination with deoxygenase along with DNA in the same reaction mixture, so that deoxygenase converts 5mC to 5hmC and 5caC, and GT converts any remaining 5hmC to 5ghmC, ensuring that only cytosine is deaminated.

[0157] As used herein, "part" of a nucleic acid sample and "aliquot" of a nucleic acid sample are intended to generally have the same meaning and may be used interchangeably.

[0158] As used herein, the term “comparing” generally refers to analyzing two or more sequences against each other. In some cases, comparison may be performed by aligning two or more sequences against each other such that the nucleotides at corresponding positions are aligned with each other.

[0159] As used herein, the term "reference sequence" generally refers to the sequence of the fragment being analyzed. The reference sequence may be obtained from a public database or sequenced separately as part of an experiment. In some cases, the reference sequence may be virtual, and the reference sequence may be deaminated (e.g., changing C to U or T, etc.) through computer analysis to enable sequence comparison.

[0160] As used herein, the terms “G”, “A”, “T”, “U”, “C”, “5mC”, “5fC”, “5caC”, “5hmC”, and “5ghmC” generally refer to nucleotides comprising guanine (G), adenine (A), thymine (T), uracil (U), cytosine (C), 5-methylcytosine, 5-formylcytosine, 5-carboxycytosine (5caC), 5-hydroxymethylcytosine, and 5-glucosylhydroxymethylcytosine, respectively. For clarity, C, 5fC, 5caC, 5mC, and 5ghmC are each different moiety.

[0161] As used herein, the term "minimal residual disease" or "MRD" generally refers to a small number of cancer cells remaining in the body after cancer treatment. MRD testing may be performed to determine whether cancer treatment is effective and to guide future treatment plans. Various metrics, including but not limited to response to treatment, tumor load, postoperative residual tumor, recurrence, secondary screening, primary screening, and cancer progression, may be used to evaluate MRD.

[0162] As used herein, the term "next-generation sequencing" or "NGS" generally applies to sequencing a library of genome fragments smaller than 1 kb.

[0163] As used herein, the terms “healthy” or “normal” generally refer to subjects or samples derived therefrom who are free from disease. While health is a dynamic state, the term may refer to a pathological state of a subject who is free from a referenced disease state (e.g., cancer). In one example, when referring to a methylation profile classifying subjects with cancer, the term “healthy” refers to an individual free from cancer, such as CRC. Although other diseases or states of health may be present in the subject, the term “healthy” may indicate the absence of a specified disease for the purpose of comparison or classification between subjects with a disease state and subjects without a disease state and samples derived therefrom.

[0164] As used herein, the term “threshold” generally refers to a value selected to differentiate, separate, or distinguish two groups of subjects. In some embodiments, the threshold distinguishes the methylation status between a diseased (e.g., malignant) state and a non-diseased (e.g., healthy) state. In some embodiments, the threshold distinguishes the stage of the disease (e.g., stage 1, stage 2, stage 3, or stage 4). The threshold may be set according to the disease, based on previous analysis, e.g., previous analysis of a training set, or determined through computer analysis on an input set having known features (e.g., healthy, diseased, or disease stage). The threshold may also be set for a gene region based on predicted methylation values ​​at a specific site. The threshold may differ for each methylation site, and data from multiple sites may be combined in the final analysis.

[0165] Any method and material similar or equivalent to that described herein may also be used to carry out or test the teachings of this document, but some exemplary methods and materials are described herein.

[0166] Any citation of any publications relates to disclosures prior to the filing date and should not be construed as an acknowledgment that the claims are not entitled to be dated prior to such publications by the prior art. Additionally, the provided publication dates may differ from the actual publication dates, which can be independently verified.

[0167] As will be apparent to those skilled in the art upon reading this disclosure, each individual embodiment described and illustrated herein has distinct components and features that can be easily separated from or combined with any of the features of any of the other embodiments without departing from the scope or spirit of this teaching. Any cited method may be performed in the order of the cited events or in any other logically possible order.

[0168] Cell-free DNA (cfDNA) sequencing can be a useful tool for cancer detection. DNA methylation analysis, coupled with sequencing, can determine whether a portion of the cfDNA is likely precancerous or tumor-derived. Many standard DNA library preparation methods utilize enzymatic end repair processes to attach sequencing adapters. However, these processes can remove unique methylation signals from DNA fragments, fail to faithfully preserve molecular ends, and cannot capture nicked or single-stranded DNA. These factors reduce the sensitivity of these methods for cancer detection. Fig. 1 ).

[0169] DNA methylation is a covalent modification of DNA and serves as a stable genetic marker that can play a crucial role in suppressing gene expression and regulating chromatin structure. In humans, DNA methylation primarily occurs at cytosine residues within CpG dinucleotides. Unlike other dinucleotides, CpGs are not evenly distributed throughout the genome and can be concentrated in short, CpG-rich DNA regions known as CpG islands. Generally, the majority of CpG sites within the genome are methylated by approximately 70 to 75%. However, methylation patterns vary by cell type, reflecting their role in regulating cell-type-specific gene expression. In this way, cellular methylation can program the cell's final differentiation state, such as neurons, muscle cells, or immune cells.

[0170] Furthermore, various cell subtypes within a tissue may exhibit different methylation patterns. In cancer cells, CpG methylation can be disregulated, and abnormalities in methylation patterns are some of the earliest events occurring during oncogenesis. The methylation profile in a given cancer type most closely resembles the methylation profile of the tissue of origin. Therefore, abnormal methylation markers on cfDNA fragments can be used to distinguish cancer cells from normal cells and to determine tissue type origin. Generally, overall CpG methylation levels are reduced in cancer cells; however, at specific loci, the average methylation level (or methylation rate) at specific CpG sites in cancer cells may differ compared to matched normal cells. Profiling differentially methylated CpGs (DMCs; single sites) or differentially methylated regions (DMRs; more than one site within a local area) between normal and diseased cells can identify biomarkers of the disease. This approach led to the development of the SEPT9 gene methylation assay (Epi proColon), the first FDA-approved blood-based diagnostic method for colorectal cancer (CRC).

[0171] Methylation profiles as used herein may include both hypomethylation and hypermethylation of DNA analysis, or either one independently. Both hypomethylation and hypermethylation are relative terms indicating less methylation (low) or more methylation (high) than reference standard DNA. When specifically applied to cancer epigenetics, the reference standard may be DNA isolated from normal tissue (e.g., non-cancerous).

[0172] Bisulfite conversion or bisulfite sequencing can be used for DNA methylation analysis. Bisulfite sequencing is a convenient and effective method for mapping DNA methylation to individual bases. Unfortunately, bisulfite conversion is a harsh and destructive process for cfDNA, resulting in the degradation of more than 90% of the sample DNA. Two main approaches to building a bisulfite sequencing library are (1) bisulfite conversion of DNA prior to library construction, which requires ssDNA library construction; and (2) bisulfite conversion of DNA after dsDNA adapter ligation. In either case, severe DNA degradation is involved, which can be problematic, especially for cfDNA, which is present at very low concentrations in plasma and is a limited resource in liquid biopsy applications.

[0173] Alternatively, enzymatic methylation conversion may be used for DNA methylation analysis and sequencing. In one embodiment, methylation conversion is mediated by a non-destructive enzymatic reaction using, for example, a 1011 translocation (TET) enzyme and a cytosine deaminase (e.g., apolipoprotein B mRNA-editing enzyme, cytidine deaminase of the catalytic polypeptide (APOBEC) family) to convert unmethylated cytosine (except methylated cytosine) to uracil. Other embodiments, such as TET-assisted pyridine borane sequencing (TAPS), combine an enzymatic reaction such as TET treatment with a chemical treatment (e.g., the use of pyridine borane).

[0174] The emergence of next-generation DNA sequencing provides advancements in clinical medicine and basic research. However, while this technology has the capability to generate hundreds of billions of DNA sequence nucleotides in a single experiment, it results in hundreds of millions of sequencing errors due to an error rate of about 1%. While these errors may be acceptable in some applications, they pose a significant problem for the "deep sequencing" of genetically heterogeneous mixtures, such as tumors or mixed microbial populations.

[0175] Existing methods require two different sequencing assays and two different cfDNA pools to analyze variants and methylation status within cfDNA. This can be excessively costly in terms of plasma / cfDNA input volume and associated expenses. Furthermore, DNA degradation caused by bisulfite can reduce the sensitivity of variant retrieval methods applicable to bisulfite-converted DNA sequencing data (compared to enzymatic conversion). Therefore, an improved method for cfDNA methylation analysis is needed to preserve the integrity of sample nucleic acids and enhance the accuracy of methylation status analysis at the whole-genome or targeted level.

[0176] As used herein, the terms “including,” “include,” “having,” “with,” or variations thereof are used in the detailed description and / or claims, and are intended to be interpreted as inclusive in a manner similar to the term “comprising.”

[0177] Where a range is used in this disclosure, the range may be expressed herein as "about" from one specific value and / or "about" another specific value. When such a range is expressed in this way, another embodiment includes from one specific value and / or another specific value. In an embodiment where a numerical value is expressed as an approximation, for example through the use of the preceding word "about," the specific value will be understood to form another embodiment. Additionally, it will be understood that the endpoint of each range is important not only in relation to other endpoints but also independently of other endpoints.

[0178] Various aspects of the method and system for manufacturing an ssDNA library of the present disclosure are described above with reference to exemplary applications for illustrative purposes. It should be understood that numerous specific details, relationships, and methods are described to provide a complete understanding of the disclosed method and system. However, those skilled in the art will readily recognize that the method and system may be practiced without one or more specific details or by using other methods. The present disclosure is not limited by a specific order of acts or operations, as some acts may occur in a different order and / or simultaneously with other acts or operations. Furthermore, not all specified acts or operations are required to implement the method according to the present disclosure.

[0179] I. Library Preparation for Enzymatic Methylation Sequencing

[0180] In a first aspect, a method for preparing a sequencing library is provided. The method described herein provides an ssDNA library acceptable for both next-generation non-methylation and methylation next-generation sequencing applications, thereby providing sequencing data for both applications from a single sample. The generated raw sequencing data can be used for conventional cfDNA analyses such as copy number alteration, germline variant detection, somatic variant detection, nucleosome localization, transcription factor profiling, and chromatin immunoprecipitation, as well as for methylation status analysis.

[0181] Adapter ligation for targeted sequencing application

[0182] In one aspect, the present method preserves the integrity and information of nucleic acid sequences for methylation profiling. In one example, combining ssDNA adapter ligation prior to enzymatic conversion increases library complexity for targeted enrichment (or directly whole-genome sequencing) while preserving fragment endpoint information, thereby providing greater sensitivity for detecting rare events such as methylated ctDNA. A comparison of the advantages of ssDNA adapter ligation and adapter ligation prior to methylation conversion Fig. 1 It appears in.

[0183] In one example, a nucleic acid adapter is ligated to the 5' and 3' ends of a collection of nucleic acid fragments within a biological sample to generate a sequencing library. In another example, a collection of nucleic acid adapters is ligated to the nucleic acid fragments within the sample. The nucleic acid adapters described herein may optionally be conversion-allowing nucleic acid adapters. In other examples, the nucleic acid adapter may optionally be a conversion-resistant adapter comprising one or more deamination-resistant modified cytosines, including but not limited to propynyl-C and pyrrolo-C. The adapter collection may contain unique molecular identifier (UMI) sequences of 4 base pairs (bp), 5 bp, and 6 bp in equal proportions. The UMIs may be located adjacent to the library insert nucleic acid. During sequencing, the UMIs are also sequenced as part of the read at the 5' end. The adapter collection may contain a single-length core UMI, which may result in reduced sequencing quality by reducing sequencing complexity at positions corresponding to invariant thymidines. The first 4 bp of each UMI may contain a set of 4 bp (e.g., single-length) core UMI sequences with an edit distance of 2 or greater, nucleotide- and color-balanced. Using a single-length core UMI in the presence of variable-length UMI sequences can facilitate the use of bioinformatics tools built for single-length UMIs for UMI extraction and deduplication. Thus, the 4 bp core sequence can function as a recognition sequence instructing bioinformatics tools to trim 5, 6, or 7 bp sequences, thereby maintaining precise cfDNA endpoint information. A schematic diagram illustrating a staggered adapter Fig. 2It is shown in [figure]. The use of UMIs can enable single-strand error correction as well as read deduplication. In another example, a Unique Dual Index (UDI) is an additional sequence that can be added to a UMI-containing adapter during library preparation to provide sample barcoding and de-multiplexing of samples after sequencing. In various examples, the length of the UDI sequence is 4 bp, 5 bp, 6 bp, 7 bp, 8 bp, or 12 bp or longer. In various examples, the length of the UDI sequence is 12 bp, 8 bp, 7 bp, 6 bp, 5 bp, or 4 bp or less.

[0184] In various embodiments, the nucleic acid adapter may include a UMI having a length of 4 bp to 6 bp. The UMI may be designed not to be unique (e.g., extracted from a specific limited set of sequences).

[0185] In one embodiment, some UMIs contain one or more methylcytosine bases. The efficiency of enzymatic methylation conversion reactions (including TET oxidation and APOBEC deamination) can be evaluated based on the fraction of UMIs that do not match a specific limited set of UMI sequences designed by the UMI mismatch rate. The UMI mismatch rate can be used as an embedded quality control metric to evaluate sequencing library quality. Additionally, where perfect UMI matching is required in a bioinformatics pipeline, the UMI mismatch rate can be used as a filter to remove individual reads that may be of low quality due to incomplete conversion.

[0186] In various embodiments, the UMI mismatch rate is 6% or less, 5% or less, 4% or less, 3% or less, or 2% or less.

[0187] In another embodiment, the UMI contains one or more cytosines containing modifications that can be used to monitor enzyme activity. Non-limiting examples of such modified bases include 5-methylcytosine, 5-hydroxymethylcytosine, 5-formylcytosine, and 5-carboxycytosine.

[0188] The efficiency of subsequent processes, such as DNA purification or enzymatic methylation, can be enhanced by converting the ssDNA library into a dsDNA library. In some embodiments, the conversion from ssDNA to dsDNA may be performed using DNA polymerase and a primer complementary to the 3' end of the adapter sequence. The primer is annealed to the adapter-ligated nucleic acid at the 3' end of the adapter sequence to initiate an extension reaction in the presence of polymerase and deoxynucleotide triphosphate (dNTP). In some embodiments, the conversion from ssDNA to dsDNA may be performed after a purification ("cleanup") operation between ligation and DNA polymerization to remove free excess adapter and exchange reaction buffers. In other embodiments, the reaction components for the conversion from ssDNA to dsDNA (including dNTPs, polymerase, and primers) are added directly to the ligation mixture without a purification operation. In some embodiments, the splint component of the 3' adapter is truncated so as to incompletely overlap with the primer annealing site for primer extension. In this way, a primer complementary to the 3' end of the adapter sequence can be annealed at the primer annealing site. The second strand synthesis (SSS) primer may or may not have end modifiers at the 3' end, the 5' end, or both.

[0189] II. Targeted Methylation Sequencing

[0190] In a targeted methylation sequencing approach, a target region within a biological sample, such as cfDNA, is analyzed to determine the methylation status of a target gene sequence. In some embodiments, the target region includes adjacent nucleotides of the target region of interest, for example, at least about 16 adjacent nucleotides of the target region of interest, or is hybridized with them under strict conditions. In various examples, targeted sequencing can be achieved using hybridization capture and amplicon sequencing approaches.

[0191] A. Hybrid Capture

[0192] The hybridization method provided herein can be used for various forms of nucleic acid hybridization, such as hybridization in solution and hybridization on solid supports (e.g., Northern, Southern, and in-situ hybridization on membranes, microarrays, and cell / tissue slides). In particular, the method is suitable for hybrid capture in solution for the targeted enrichment of specific types of genomic DNA sequences (e.g., exons) used in targeted next-generation sequencing. For the hybrid capture approach, the cell-free nucleic acid sample undergoes library preparation. As used herein, "library preparation" includes adapter ligation of ssDNA as described herein, or any other preparation performed on cell-free DNA to enable subsequent sequencing of DNA. In certain examples, the prepared cell-free nucleic acid library sequence includes an adapter, sequence tag, and index barcode ligated to the cell-free nucleic acid sample molecule. Various commercial kits are available to facilitate library preparation for next-generation sequencing approaches. The construction of next-generation sequencing libraries may involve preparing nucleic acid targets using a series of coordinated enzymatic reactions to generate a random collection of DNA fragments of a specific size for high-throughput sequencing. Advancements and developments in various library preparation technologies have extended the application of next-generation sequencing to fields such as transcriptomics and epigenetics.

[0193] Improvements in sequencing technology have led to changes and improvements in library manufacturing. Next-generation sequencing library manufacturing kits used in this institution include kits developed by companies such as Agilent, Bioo Scientific, Claret Bioscience, Kapa Biosystems, New England Biolabs, Illumina, Life Technologies, Pacific Biosciences, and Roche.

[0194] In various examples for targeted capture gene panels, various library manufacturing kits may be selected from Nextera Flex (Illumina), IonAmpliseq (Thermo Fisher Scientific), Genexus (Thermo Fisher Scientific), Agilent ClearSeq (Illumina), Agilent SureSelect Capture (Illumina), Archer FusionPlex (Illumina), BiooScientific NEXTflex (Illumina), IDT xGen (Illumina), Illumina TruSight (Illumina), Nimblegene SeqCap (Illumina), and Qiagen GeneRead (Illumina).

[0195] In some embodiments, the hybrid capture method is performed on a library sequence prepared using a specific probe. As used herein, the term “specific probe” may refer to a probe specific to a known methylation site. In some embodiments, the specific probe is designed using the human genome as a reference sequence and a specific genomic region known to have a methylation site as a target sequence. Specifically, the genomic region known to have a methylation site may include at least one of a promoter region, a CpG island region, a CGI Shore region, and an imprinted gene region. Thus, when performing hybrid capture using the specific probe of some embodiments, a sequence within the sample genome that is complementary to the target sequence, for example, a region within the sample genome known to have a methylation site (also referred to herein as a “specific genomic region”), can be efficiently captured.

[0196] For example, the methylation regions described herein may be used to design specific probes. In some embodiments, specific probes are designed using commercially available methods, such as eArray systems, for example. The length of the probe may be sufficient to hybridize with sufficient specificity to the methylation region of interest. In various examples, the probe is 10-mer, 11-mer, 12-mer, 13-mer, 14-mer, 15-mer, 16-mer, 17-mer, 18-mer, 19-mer, or 20-mer. In other embodiments, the length of the probe is 50 to 200 nucleotides. In certain embodiments, the length of the probe is 80 to 150 nucleotides. In yet another embodiment, the length of the probe is 100 to 130 nucleotides. In some embodiments, the length of the probe is 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, or 200 nucleotides or more. In some embodiments, the length of the probe is 200, 190, 180, 170, 160, 150, 140, 130, 120, 110, 100, 90, 80, 70, 60, or 50 nucleotides or less.

[0197] Target regions for methylation analysis can be screened using database resources (e.g., gene ontology). Based on the principle of complementary base pairing, a single-strand capture probe can be combined complementarily with a single-strand target sequence, thereby successfully capturing the target region. In some embodiments, the designed probe can be designed as a solid capture chip (where the probe is immobilized on a solid support) or a liquid capture chip (where the probe is free in a liquid). However, due to limiting factors such as probe length, probe density, and high cost, solid capture chips are rarely used, whereas liquid capture chips are used more frequently.

[0198] In some embodiments, compared to normal sequences (where the average content of A, T, C, and G bases is 25% each), GC-rich sequences (where the average content of GC bases is higher than 60%) may result in a decrease in capture efficiency due to the molecular structure of C and G bases. For major research areas, e.g., CGI regions (CpG islands), a larger amount of probes may be required to obtain sufficient and accurate CGI data.

[0199] B. Amplicon-based sequencing

[0200] The converted ssDNA fragment can be amplified. In some embodiments, amplification is performed using primers designed to anneale to a methylated converted target sequence having at least one methylated site within it. The methylation sequencing conversion results in the conversion of unmethylated cytosine to uracil, while 5-methylcytosine is unaffected. The "converted target sequence" may refer to a sequence in which the cytosine known as the methylated site is fixed to "C" (cytosine), while the cytosine known as unmethylated is fixed to "U" (uracil), which can be treated as "T" (thymine) for primer design purposes.

[0201] In various examples, the source of DNA is cell-free DNA obtained from whole blood, plasma, or serum, or genomic DNA extracted from cells or tissues. In some embodiments, the size of the amplified fragment is about 100 to 200 base pairs in length. In some embodiments, the DNA source is extracted from a cellular source (e.g., tissue, biopsy, cell line), and the length of the amplified fragment is about 100 to 350 base pairs. In some embodiments, the amplified fragment comprises at least one 20-base-pair sequence containing at least one, at least two, at least three, or more than three CpG dinucleotides. Amplification may be performed using a set of primer oligonucleotides according to the present disclosure, and a heat-stable polymerase may be used. Amplification of multiple DNA segments may be performed simultaneously in a single and identical reaction vessel. In some embodiments, two or more fragments are amplified simultaneously. For example, amplification may be performed using a polymerase chain reaction (PCR).

[0202] Primers designed to target such sequences may exhibit a certain bias toward the converted methylation sequence. In some embodiments, PCR primers are designed to be methylation-specific for targeted methylation sequencing applications. Methylation-specific primers can enable higher sensitivity in some applications. For example, primers may be designed to include a discriminant nucleotide (specific to the methylation sequence after the non-sulfite conversion) positioned to achieve optimal discrimination, such as in PCR applications. The discriminant may be located at the 3' end or the second-to-end position.

[0203] In some embodiments, primers are designed to amplify DNA fragments of length ranging from 75 to 350 base pairs (bp), which is the typical size range of circulating DNA. Optimizing primer design with the target size in mind can increase the sensitivity of the method described herein. Primers may be designed to amplify regions of length from about 50 to 200, about 75 to 150, or about 100 or 125 bp.

[0204] In one embodiment, the amplification operation includes using primers containing a unique double index (UDI) sequence.

[0205] In one embodiment, the length of the UDI sequence is 4 base pairs (bp), 5 bp, 6 bp, 7 bp, 8 bp, 9 bp, 10 bp, 11 bp, or 12 bp or more. In one embodiment, the length of the UDI sequence is 12 bp, 11 bp, 10 bp, 9 bp, 8 bp, 7 bp, 6 bp, 5 bp, or 4 bp or less.

[0206] In some embodiments, the methylation status of a pre-selected CpG site within a nucleic acid sequence is detected by an amplicon-based approach using methylation-specific PCR (MSP) primer oligonucleotides. Using methylation-specific primers for the amplification of converted methylated ssDNA enables the distinction between methylated and unmethylated nucleic acids. An MSP primer pair comprises at least one primer that hybridizes to a converted CpG dinucleotide. Thus, the sequence of said primer comprises at least one CpG, TpG, or CpA dinucleotide. An MSP primer specific to unmethylated DNA comprises a "T" at the 3' position of the C site within the CpG. Thus, the base sequence of such primer may comprise a sequence having a length of at least 18 nucleotides that hybridizes to the pre-treated nucleic acid sequence and a sequence complementary thereto, and the base sequence comprises at least one CpG, TpG, or CpA dinucleotide. In some embodiments of the above method, the MSP primer has two to five CpG, TpG, or CpA dinucleotides. In some embodiments, the dinucleotides are located within the 3' half of the primer; for example, in the case of a primer with a length of 18 bases, the specified dinucleotides are located within the first 9 bases from the 3' end of the sequence. In addition to CpG, TpG, or CpA dinucleotides, the primer may further include several methylated bases (e.g., cytosine converted to thymine, or guanine converted to adenosine in the hybridization strand). In some embodiments, the primer is designed to have two or fewer cytosine and / or guanine bases.

[0207] In some embodiments, each region is amplified section by section using multiple primer pairs. In some embodiments, these sections do not overlap. The sections may be immediately adjacent or separated from each other (e.g., separated by up to 10 base pairs (bp), 20 bp, 30 bp, 40 bp, or 50 bp). Since target regions (including CpG islands, CpG shores, and / or CpG shelves) are typically longer than 75 to 150 bp, this example allows for the evaluation of the methylation status of a region spanning more (or all) of a given target region.

[0208] Primers can be designed for target regions using appropriate tools such as Primer3, Primer3Plus, and Primer-BLAST. As discussed, methylation enrichment conversion results in the conversion of unmethylated cytosine to uracil and methylated cytosine to thymine. Therefore, depending on the required degree of methylation specificity, the converted sequences can be utilized for primer localization or targeting.

[0209] C. Enzymatic conversion for DNA methylation sequencing application

[0210] Non-sulfite conversion can damage input DNA and lead to overall yield loss, fragmentation, and biased sequencing data. As an alternative, enzymatic methyl conversion can be used in methylation sequencing workflows. Examples of enzymatic methyl conversion workflows include enzymatic methyl-seq (EM-seq) and TET-assisted pyridine borane sequencing (TAPS).

[0211] EM-seq is a minimally destructive conversion methylation sequencing method for converting cytosine to uracil within nucleic acids. This non-sulfite-free method preserves the length of nucleic acid molecules while achieving conversion rates similar to non-sulfite sequencing. Additionally, EM-seq can result in higher sequencing quality scores for cytosine and guanine base pairs and provide more uniform coverage for various genomic features, such as CpG islands. EM-seq consists of two sets of enzymatic reactions. In the initiation reaction, ten-eleven translocation (TET) enzymes (e.g., TET1, TET2, TET3, Naegleria TET and genetically engineered versions and / or variants thereof) and β-glucosyltransferase (e.g., T4 BGT) convert 5mC and 5hmC into a product that cannot be deaminated by a cytosine deaminase (e.g., APOBEC) or is resistant to deamination. In the second reaction, a cytosine deaminase (e.g., APOBEC) deaminates unmodified (e.g., unmethylated) cytosine by converting it into uracil.

[0212] In another embodiment, TAPS can be used in an enzymatic methylation sequencing workflow. TAPS is a minimally destructive conversion methylation sequencing method for converting cytosine within nucleic acids to uracil. This non-sulfite-free method minimizes DNA degradation and thus preserves the length of nucleic acid molecules while achieving conversion rates similar to sodium bisulfite sequencing. TAPS can result in higher sequencing quality scores for cytosine and guanine base pairs and can provide more uniform coverage for various genomic features, such as CpG islands.

[0213] In TAPS, a 10-11 transposase (e.g., TET1) is used to oxidize both 5mC and 5hmC to 5caC. Pyridine borane is used to reduce 5caC to dihydrouracil, which is converted to thymine as a uracil derivative after PCR. TAPS can be performed in two different ways: TAPSβ and chemically assisted pyridine borane sequencing (CAPS). In TAPSβ, β-glucosyltransferase is used to label 5hmC with glucose to protect it from oxidation and reduction reactions, enabling specific detection of 5mC. In CAPS, potassium perruthenate acts as a chemical substitute for TET1 and specifically oxidizes 5hmC, enabling direct detection.

[0214] In one example of enzymatic methyl conversion, a combination of enzymatic conversion converting unmodified C to U and staggering UMI adapters to match library inserts is useful for targeted sequencing of methylation libraries. For shallow-depth sequencing applications, this combination can reduce the volume input of plasma or the mass input of cfDNA compared to non-sulfite conversion sequencing because the sample cfDNA is not degraded to the same extent.

[0215] In the case of high-depth sequencing application, higher depth sequencing can be obtained from a similar amount of plasma or cfDNA input compared to non-sulfite conversion sequencing because cfDNA is not degraded to the same degree.

[0216] In one example, to prevent the conversion of C to T within the adapter, the cytosine present in the adapter nucleic acid is modified into a 5-methyl group or a 5-hydroxymethyl group.

[0217] One advantage of this approach is that performing adapter ligation before conversion preserves fragment endpoint and length information compared to the approach of performing non-sulfite conversion followed by ssDNA adapter ligation. If nucleic acids are significantly degraded prior to adapter ligation, useful fragment endpoint and length information may be lost.

[0218] Enzymatic conversion of unmodified C to U is less harsh on sample nucleic acid fragments and can provide more complete and uniform coverage compared to non-sulfite conversion methods. Non-sulfite degradation of DNA is non-uniform, so some sequences, including CG dinucleotides—the very sites targeted for investigation in methylation sequencing—are degraded preferentially over others. Therefore, enzymatic approaches provide higher coverage for CpG sites than non-sulfite conversion methods when using the same number of unique reads, and offer higher uniformity of captured reads in targeted enrichment applications. Furthermore, non-bisulfite methods (e.g., enzymatic and TAPS-like chemical conversions) enhance the resolution of biological signals, specifically providing the ability to distinguish between 5mC and 5hmC methylation in nucleic acid sequences. This information and additional resolution can provide valuable insights for computational analysis approaches and other methods.

[0219] In some examples, exposing DNA or barcode-labeled DNA to an enzymatic reaction that converts cytosine nucleotides of DNA or barcode-labeled DNA to uracil nucleotides includes “performing enzymatic conversion.”

[0220] In various examples, glucosylation and oxidation reactions overcome the inherent deamination of 5hmC and 5mC observed by deaminases. Deaminases convert 5mC and unmodified C to U, but do not convert 5ghmC and 5caC. Non-limiting examples of deaminases include APOBEC (apolipoprotein B mRNA editing enzyme, catalytic polypeptide analog). The embodiments described herein utilize enzymes that are substantially sequence-free in the glucosylation, oxidation, and deamination of cytosine. Furthermore, these embodiments do not cause substantially non-specific damage to DNA during the glucosylation, oxidation, and deamination reactions.

[0221] In some embodiments, glucosyltransferase (GT), e.g., beta-glucosyltransferase (βGT), is utilized to covalently bond glucose to 5hmC to protect this modified base from deamination. Other enzymatic or chemical reactions that modify 5hmC may be used to achieve the same effect.

[0222] Generally and in one aspect, the method provided herein comprises: (a) treating an aliquot (part) of a nucleic acid sample in a reaction mixture with a deoxygenase, e.g., TET2, and βGT, to produce a reaction product in which substantially all modified cytosine (C) is oxidized or, in the case of 5hmC, glucosylated; and (b) treating this reaction product with cytidine deaminase to convert substantially all unmodified C to U. Throughout these examples and embodiments, the term “modified” cytosine refers to one or more of 5mC, 5hmC, 5ghmC, 5fC, and 5caC, and 5caC is produced when 5mC, 5hmC, and 5fC are completely oxidized. βGT reacts only with 5hmC. However, some 5hmC may be converted to 5fC and then to 5caC by a deoxygenase before glucosylation occurs. In the presence of deoxygenase, 5mC is mostly completely oxidized to 5caC, but some residual 5hmC may be produced. However, residual 5hmC can be glucosylated by βGT, which prevents a low deamination rate of 5hmC that would otherwise reduce the accuracy of methylation sequencing.

[0223] Therefore, the described method largely distinguishes between unmodified and modified cytosines by treating nucleic acids with deoxygenase before deamination. However, the amount of naturally occurring 5mC in genomic DNA can substantially exceed the amount of 5hmC, which in turn can exceed the amounts of naturally occurring 5fC and 5caC. Therefore, the amount of naturally occurring modified cytosines is generally considered to be an approximation of the amount of naturally occurring 5mC.

[0224] In one example, the above method can be adapted to perform 5hmC sequencing. The 5hmC sequencing method may further include the step of treating an aliquot of a nucleic acid sample with βGT in the absence of deoxygenase, followed by treatment with cytidine deaminase, to produce a reaction product in which substantially all 5hmC within the aliquot is glucosylated and substantially all unmodified C and 5mC are converted to U. After PCR amplification, U is converted to T, so cytosine and 5mC become indistinguishable during sequencing. The resulting reaction product can be sequenced and compared with a reference sequence to distinguish 5hmC from C and 5mC. This distinction of moiety makes it possible to map these modified nucleotides to a reference sequence, for example, a reference sequence from a database or an independently determined reference sequence.

[0225] In some embodiments, by sequencing the reaction product of deoxygenase and deaminase to which βGT has been added, or the amplified product thereof, it may be determined which C is methylated (which may include a small fraction of 5hmC) and which C is not modified. In some embodiments, by sequencing the reaction product of βGT and deaminase to which βGT has not been added, or the amplified product thereof, it may be determined which C is hydroxymethylated and which C is not hydroxymethylated. In some embodiments, by sequencing the reaction product of βGT and deaminase to which βGT has not been added, or the amplified product thereof, it may be determined which C is hydroxymethylated and which C is not modified. Reference DNA may be generated by sequencing the reaction product produced without reacting a nucleic acid sample with any of deoxygenase, βGT, and deaminase. Alternatively, the reference sequence may be, for example, a known reference sequence from a sequence database.

[0226] In one embodiment, the sequence of the reaction product of a deoxygenase and a deaminase having βGT can be compared with a reference sequence. Optionally, by comparing this with the sequence of the reaction product of a βGT and a deaminase (without the addition of deoxygenase), it can be determined whether any cytosine in the nucleic acid sample has been modified by a methyl group or by a hydroxymethyl group.

[0227] In one aspect, a method for performing targeted methylation sequencing of a cell-free DNA (cfDNA) sample from a subject is provided, said method

[0228] a) a step of ligating a conversion-allowable nucleic acid adapter to a single-stranded nucleic acid molecule of a cfDNA sample, wherein the single-stranded nucleic acid molecule comprises unconverted nucleic acid;

[0229] b) a step of enzymatically converting unmethylated cytosine within a single-stranded nucleic acid molecule to uracil to produce converted nucleic acid;

[0230] c) a step of amplifying the nucleic acid converted by the polymerase chain reaction;

[0231] d) a step of probing a nucleic acid converted into a nucleic acid probe complementary to a pre-identified CpG or CH locus panel to enrich a sequence corresponding to a pre-identified CpG or CH locus panel;

[0232] e) a step of determining the nucleic acid sequence of the converted nucleic acid with a depth of >100x; and

[0233] f) To determine the methylation profile of a cfDNA sample from a subject, the step of comparing the nucleic acid sequence of the converted nucleic acid with a reference nucleic acid sequence of a pre-identified CpG or CH locus panel is included.

[0234] In some embodiments, the conversion-allowing nucleic acid adapter is a conversion-resistant nucleic acid adapter comprising one or more deamination-resistant modified cytosines, including but not limited to propynyl-C and pyrrolo-C. In one embodiment, the conversion-resistant adapter may include one or more propynyl-C residues. In another embodiment, the conversion-resistant adapter may include one or more pyrrolo-C residues. In yet another embodiment, the conversion-resistant adapter may include a combination of propynyl-C and pyrrolo-C residues. If the test conversion nucleic acid sequence at the specified CpG locus is T corresponding to reference C, C was in a non-methylated state in the original test nucleic acid fragment. Conversely, if both the test conversion nucleic acid sequence and the reference sequence at the specified CpG locus are C, C was in a methylated state in the original test nucleic acid fragment.

[0235] In one example, the nucleic acid sequence of the converted nucleic acid molecule is sequenced at a depth of about 50 to 500x, about 25 to 1000x, about 50 to 500x, about 250 to 750x, about 500 to 200x, about 750 to 1500x, or about 100 to 2000x. In some embodiments, the nucleic acid sequence is sequenced at a depth of more than 100x or more than 500x.

[0236] In one example, the nucleic acid sequence of the converted nucleic acid molecule is sequenced at a great depth of about 500x, about 1000x, about 2000x, about 3000x, about 4000x, about 5000x, about 6000x, about 7000x, about 8000x, about 9000x, about 10000x, or more than 5000x.

[0237] In one example, the nucleic acid sequence of the converted nucleic acid molecule is sequenced at a depth of about 300x unique, about 400x unique, about 500x unique, about 600x unique, about 700x unique, about 800x unique, about 900x unique, or about 1000x unique, or greater than 500x unique.

[0238] D. Application of Targeted Enrichment Sequencing

[0239] A method for enriching methylation regions of interest in targeted capture applications during sequencing is further provided. A potential issue when applying targeted enrichment capture panels to DNA methylation libraries is a low on-target read rate and a high off-target DNA fragment capture rate. For all regions within the panel, probes can be designed to target DNA derived from methylated CpGs or DNA derived from unmethylated CpGs. In either probe type, all CpG sites present along the region are considered unmethylated or methylated to suit the probe type. The probes can be hybridized to library molecules after non-sulfite / enzymatic conversion and PCR amplification. Subsequently, only the library molecules captured by the probes are sequenced. This method has the advantage of reducing sequencing costs because only a small fraction of the genome is sequenced. In one example, approximately 0.1% of the genome is sequenced. In another example, approximately 0.3% of the genome is sequenced. In yet another example, approximately 0.5% of the genome is sequenced. In one example, about 0.7% of the genome is sequenced. In one example, about 1% of the genome is sequenced. In another example, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, or more than about 10% of the genome is sequenced. In another example, about 10%, about 9%, about 8%, about 7%, about 6%, about 5%, about 4%, about 3%, or less than about 2% of the genome is sequenced.

[0240] Significant off-target capture rates may occur when using target capture enrichment approaches in both bisulfite and enzymatic conversion libraries. These off-target capture rates arise because, in both types of probes that hybridize to partially methylated CpG-derived DNA, all cytosines not located in the CpG site are converted from C to T. The reduction in cytosine content within the probe leads to a decrease in sequence complexity, which consequently lowers the specificity of the probe hybridizing to the target library molecule.

[0241] As used herein, the terms “methylated probe” and “unmethylated probe” refer to probes used to hybridize to methylated CpG and unmethylated CpG in the nucleic acid sequence after conversion, respectively. The probes may be designed to recognize the nucleic acid sequence after conversion. In the post-conversion methylated CpG probe, C remains as C after conversion. In the post-conversion unmethylated CpG probe, C is converted to T after conversion. In both the post-conversion methylated probe and the post-conversion unmethylated probe, all C within the non-CpG dinucleotide is converted to T after conversion.

[0242] Methylated probes retain some cytosines (e.g., cytosines within the CpG site). In contrast, in unmethylated probes, all cytosines are converted to thymine. Unmethylated probes are less complex than methylated probes and have the potential to contribute preferentially to the off-target capture rate. In one example, a probe that hybridizes to DNA derived from a methylated CpG is used in a target enrichment method. In another example, a probe having a sequence substantially complementary to the target, which hybridizes to DNA derived from a methylated CpG, is used in a target enrichment method.

[0243] Probes that hybridize to DNA derived from methylated CpGs for target enrichment can be selected to achieve various aspects. The target capture hybridization reaction occurs at a single temperature. However, the optimal melting temperature (Tm) of probes that hybridize to DNA derived from methylated CpGs is, on average, higher than the Tm of probes not designed to hybridize to DNA derived from methylated CpGs.

[0244] Cytosine base pairing involves three hydrogen bonds, whereas thymine base pairing involves two hydrogen bonds. When cytosine within a probe is converted to thymine, the probe's Tm decreases due to the reduction in hydrogen bonds. Probes designed to collect DNA derived from methylated fragments may contain more cytosine than a set of probes designed to collect corresponding unmethylated fragments in the same genomic region. As a result, probes targeting methylated fragments may have an elevated Tm compared to their corresponding unmethylated probes. As the number of CpG sites within a region increases, the difference in melting temperatures between methylated and unmethylated probes also increases. Probes with higher melting temperatures may hybridize more efficiently to target DNA fragments than probes with lower melting temperatures. Hybridization temperatures are generally selected to be relatively high to facilitate on-target capture. However, at many hybridization temperatures, methylated probes may hybridize more efficiently than unmethylated probes due to the higher melting temperature resulting from the retention of some cytosine. Higher melting temperatures can lead to a bias in which the percentage of CpG methylation levels measured by the target capture hybridization approach appears higher compared to levels measured by sequencing of the library before capture.

[0245] In one example, only a single probe type—either methylated or unmethylated—is used in the hybridization reaction to enrich hypermethylated library molecules or hypomethylated library molecules, respectively. Using a single type of methylated or unmethylated probe avoids the issue of different melting temperatures between probe types. Using a single probe type may also facilitate more efficient capture (or enrichment) of the same DNA fragment type. In one example, the use of only methylated probes provides preferential binding to hypermethylated regions of interest (ROIs) over hypomethylated regions of interest. In another example, the use of only unmethylated probes provides enrichment of unmethylated ROIs.

[0246] Using only a single probe type also allows for the use of higher hybridization temperatures to reduce off-target capture without affecting the relative balance of methylated ROI capture versus non-methylated ROI capture. Thus, probe panels can be designed based on the objective of enriching either hypermethylated DNA fragments or hypomethylated DNA fragments. In one example, where quantification of both hypermethylated and hypomethylated DNA fragments is required, two parallel but separate hybridization reactions are used for both methylation states.

[0247] E. Methylation Analysis

[0248] In various examples, results from enzymatic methylation sequencing are used to analyze the methylation status of nucleic acids in biological samples. In one example, whole-genome enzymatic methyl sequencing ("WG EM-seq") provides high-resolution sequencing by characterizing DNA methylation of almost all cytidine nucleotides within the genome. Other targeting methods, such as targeted enzymatic methyl sequencing ("TEM-seq"), can also be useful for methylation analysis.

[0249] In other examples, assays that can be used for non-sulfite conversion may be used in minimally destructive conversion methods such as enzymatic conversion, TAPS, and CAPS. In various examples, assays used for methylation analysis may be mass spectrometry, methylation-specific PCR (MSP), reduced representative non-sulfite sequencing (RRBS), HELP assay, GLAD-PCR assay, ChIP-on-chip assay, restriction landmark genome scanning, methylated DNA immunoprecipitation (MeDIP), pyrosequencing of non-sulfite-treated DNA, molecular break light assay, methyl-sensitive Southern blotting, high-resolution melting assay (HRM or HRMA), ancient DNA methylation reconstruction, or methylation-sensitive single nucleotide primer extension assay (msSNuPE).

[0250] The methylation profile of cfDNA can be identified by applying a sequence alignment method to map methyl-seq reads obtained from targeted methyl sequencing of the whole genome or the human reference genome to the human reference genome. Non-limiting examples of sequence alignment methods include bwa-meth, bismark, Last, GSNAP, BSMAP, NovoAlign, Bison, Metagenomic Phylogenetic Analysis (e.g., MetaPhlAn2), BLAT, Burrows-Wheeler Aligner (BWA), Bowtie, Bowtie2, Bfast, BioScope, CLC bio, Cloudburst, Eland / Eland2, GenomeMapper, GnuMap, Karma, MAQ, MOM, Mosaik, MrFAST / MrsFAST, PASS, PerM, RazerS, RMAP, SSAHA2, Segemehl, SeqMap, SHRiMP, Slider / SliderII, Srprism, Stampy, vmatch, ZOOM, and SOAP / SOAP2 alignment tools.

[0251] F. Identification of Somatic Variants

[0252] In various examples, enzymatically converted DNA is used to infer the methylation status of C residues within the genome. However, because enzymatic conversion of DNA converts unmethylated C residues into U residues and does not introduce other chemical changes to the DNA, somatic variants that do not correspond to C or T bases in reference or query sequences can also be identified in converted DNA. These somatic variants can be identified using existing methods designed for unconverted DNA.

[0253] G. Nucleosome localization inference

[0254] Cytosine methylation at CpG sites can be significantly enriched in nucleosome-spanning DNA compared to flanking DNA. Therefore, CpG methylation patterns can also be used to infer nucleosome localization using machine learning approaches. EM-seq datasets can be analyzed according to the same methods used in WGS to generate features for processing by machine learning methods and models, regardless of whether methylation is switched. Consequently, 5mC patterns can be used to predict nucleosome localization, which can aid in gene expression inference and / or the classification of diseases and cancers. In another example, features can be obtained from a combination of methylation status and nucleosome localization information.

[0255] Metrics used for methylation analysis include, but are not limited to, M-bias (base-specific methylation % for CpG, CHG, CHH), conversion efficiency (e.g., average methylation % for 100-CHH), hypomethylation blocks, methylation levels (e.g., global average methylation for CpG, CHH, CHG, chrM, LINE1, or ALU), dinucleotide coverage (normalized coverage of dinucleotides), coverage uniformity (e.g., unique CpG sites in 1x and 10x average genomic coverage), global average CpG coverage (depth), and average coverage in CpG islands, CGI shelves, and CGI shores. In one example, fragment endpoint and length information are used as feature inputs for analysis. These metrics can be used as feature inputs for machine learning methods and models.

[0256] In another aspect, the present disclosure provides a method comprising: (a) providing a biological sample containing cfDNA from a subject; (b) exposing the cfDNA to conditions sufficient for selective enrichment of methylated cfDNA in the sample; (c) enzymatically converting unmethylated cytosine nucleobases of the cfDNA to uracil nucleobases; (d) sequencing the cfDNA to generate sequence reads; (e) processing the sequence reads by computer to (i) determine the degree of methylation of the cfDNA based on the presence of uracil nucleobases, and (ii) generate degradation parameters by modeling at least partial degradation of the cfDNA; and (f) determining gene sequence features using the degradation parameters and the degree of methylation.

[0257] In some examples, sequencing of cfDNA involves determining the degree of DNA methylation based on the ratio of unconverted cytosine nucleotides to converted cytosine nucleotides. In some examples, converted cytosine nucleotides are detected as uracil nucleotides. In some examples, uracil nucleotides are observed as thymine nucleotides in sequence reads.

[0258] In some examples, the generation of degradation parameters involves using Bayesian models. In some examples, Bayesian models are based on strand bias, enzymatic conversion, or overconversion. In some examples, the computer processing of sequence reads involves using degradation parameters under the framework of paired HMMs or naive Bayesian models.

[0259] H. Differential Methylation Region (DMR) Analysis

[0260] In one example, the methylation analysis is a Differential Methylation Region (DMR) analysis. DMR is used to quantify CpG methylation across genomic regions. Regions are dynamically assigned through exploration. Multiple samples from different classes may be analyzed, and the most differentially methylated regions among the different classes can be identified. Regions from a differentially methylated subset may be selected and used for classification. The number of CpGs captured in a region may be used for analysis. The size of the region may be variable. In one example, a preliminary exploration process is performed to group multiple CpG sites into a single region. In one example, DMR is used as an input feature for processing by machine learning methods and models.

[0261] I. Methylated Haplotype Blocks and Methylated Haplotype Loads

[0262] In one example, a haplotype block assay is applied to the sample. The identification of methylated haplotype blocks aids in the deconvolution of heterogeneous tissue samples and the mapping of tumor-origin tissues from plasma DNA. Tightly coupled CpG sites, referred to as methylated haplotype blocks (MHBs), can be identified in WGBS data. A metric called Methylated Haplotype Load (MHL) is used to perform tissue-specific methylation analysis at the block level. This method provides informative blocks useful for the deconvolution of heterogeneous samples. This method is useful for the quantitative estimation of tumor load from circulating cfDNA and for mapping of origin tissues. In one example, haplotype blocks are used as input features for processing by machine learning methods and models.

[0263] J. Targeted Methylation Calling Analysis for Identification of Origin Cell Types

[0264] In one aspect, a targeted methylation calling method is used to identify the origin cell type of a cfDNA molecule based on methylation patterns. The method provides a stochastic model of the joint methylation status of multiple adjacent CpG sites on individual sequencing reads to utilize the broad characteristics of DNA methylation for signal amplification. The model calculates the probabilities of the sequencing reads for each cell type, then constructs a mixed model for all cell types and fits it to the model.

[0265] Traditional DNA methylation analysis focuses on the methylation rate (β value) of individual CpG sites within a cell population to represent the proportion of cells with methylated CpG sites. These population-average measurements are often not sensitive enough to capture abnormal methylation signals that affect only a small portion of cfDNA. However, based on the broad nature of DNA methylation, disease-specific cfDNA reads can be distinguished from normal cfDNA reads through computer analysis.

[0266] Furthermore, considering the extensive nature of DNA methylation, the joint methylation status of multiple adjacent CpG sites can be used to easily distinguish cancer-specific cfDNA reads from normal cfDNA reads. The mean methylation value (indicated as the α-value) of all CpG sites within a given read provides the difference (0 and 1) between abnormally methylated cfDNA and normal cfDNA (α-tumor = 0% and α-normal = 100%). The methylation α-value is used to estimate whether the joint probability of all CpG sites within a read follows the DNA methylation signature of the disease. This method can sensitively identify cfDNA of multiple origin cell types among all cfDNA in plasma.

[0267] In various examples, alignment tools are used to align reads to a reference genome and to call methylated cytosines. PCR duplicates are removed, and the number of methylated and unmethylated cytosines for each CpG site is quantified. The methylation level of a CpG cluster is calculated as the ratio between the number of methylated cytosines and the total number of cytosines within the cluster. This WGBS data processing procedure calculates the average methylation level of CpG clusters in normal plasma samples used for methylation marker identification. When plasma cfDNA samples are used as test data, the joint methylation status for all CpG sites of individual sequencing reads aligned to the marker panel region is extracted and then processed by a machine learning model. In this approach, CpG methylation calls are used as input features for methylation status analysis and feature generation. To improve the input data quality of cfDNA methylation data with high coverage, reads containing fewer than 2, fewer than 3, or fewer than 4 CpG sites can be filtered out and removed.

[0268] The methylation sequencing method described herein improves sequencing read quality by, for example, reducing PCR errors and biases and reducing DNA degradation that occurs during non-sulfite conversion. In one example, methylation sequencing data is used to model overlapping regions. In one example, machine learning modeling can determine the origin cell type for identified methylated DNA regions.

[0269] In various examples, the model can classify more than two origin cell types. In other examples, the model can classify sequences into 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 50, 75, 100, or more than 100 different cell types.

[0270] K. DNA hydroxymethylation analysis

[0271] In one aspect of the concept of the present invention, 5hmC sequencing can be achieved by substituting hydroxymethylation within the adapter nucleic acid during the adapter ligation process, and then conjugating 5mC and 5hmC using deoxygenase and βGT, or by conjugating glucose to 5hmC residues within the test nucleic acid library insert using only βGT. When comparing the generated sequencing data with a reference genome, all C positions within the reference representing the corresponding C in the test sequence are interpreted as hydroxymethylated C, and all C within the reference representing T in the test sequence are interpreted as either unmodified C or methylated C. Thus, the interpretation of data for hydroxymethylation analysis is the same as for methylation analysis.

[0272] In one aspect of the concept of the present invention, methylation and hydroxymethylation sequencing libraries can be compared to determine the level of each cytosine modification (e.g., 5m or 5mC) with single nucleotide resolution.

[0273] In one aspect of the concept of the present invention, since the hydroxymethylation status reading is identical to the methylation status, all analysis methods used with methylation sequencing data can also be applied to hydroxymethylation sequencing data.

[0274] III. Computer Systems and Machine Learning Methods

[0275] A. Sample Features

[0276] As used herein, the term "feature" is associated with machine learning and pattern recognition, and may refer to individual measurable characteristics or features of observed phenomena. Features are usually numeric, but structural features such as strings and graphs are used in syntactic pattern recognition. The concept of "feature" is related to the concept of explanatory variables used in statistical techniques such as linear regression.

[0277] In one embodiment, the features are processed into a feature matrix for machine learning analysis.

[0278] For multiple tests, the system uses a machine learning model to identify the set of features to be processed. The system performs tests for each molecular class and forms feature vectors from the measured values. The system uses a machine learning model to process the feature vectors and obtains an output classification regarding whether the biological sample possesses specific characteristics.

[0279] In one embodiment, the machine learning model outputs a classifier that distinguishes between two groups or classes of objects or features within a group of objects or features. In one embodiment, the classifier is a trained machine learning classifier.

[0280] In one embodiment, an informative locus or feature of a biomarker within the cancer tissue is tested to form a profile. A receiver operation feature (ROC) curve is useful for plotting the performance of a specific feature (e.g., any biomarker described herein and / or any additional biomedical information item) in distinguishing between two groups (e.g., individuals responding to the therapeutic agent and individuals not responding to it). In some embodiments, feature data across the entire population (e.g., case group and control group) are sorted in ascending order based on the value of a single feature.

[0281] In some embodiments, the condition is an advanced adenoma (AA), colorectal cancer (CRC), colorectal carcinoma, or inflammatory bowel disease.

[0282] The terms “input feature” or “feature” refer to variables used in a model to predict the output classification (label) of a sample, e.g., pathology, sequence content (e.g., mutation), proposed data collection task, or proposed treatment. The values ​​of the variables may be determined for the sample and may be used to determine the classification. Examples of input features of genetic data include alignment variables related to the alignment of sequence data (e.g., sequence reads) to the genome, and non-alignment variables related, e.g., the sequence content of sequence reads, measurements of proteins or autoantibodies, or average methylation levels in genomic regions.

[0283] The values ​​of variables can be determined for a sample and can be used to determine classification. Examples of input features of genetic data include alignment variables related to the alignment of sequence data (e.g., sequence reads) to the genome, and non-alignment variables related, for example, the sequence content of sequence reads, measurements of proteins or autoantibodies, or average methylation levels in genomic regions. In various examples, genetic features such as V-plot measurements, FREE-C, cfDNA measurements at transcription start sites, and DNA methylation levels across cfDNA fragments are used as input features for machine learning methods and models.

[0284] In one example, sequencing information includes, but is not limited to, information regarding multiple genetic features such as transcription start sites, transcription factor binding sites, open and closed states of chromatin, nucleosome localization or occupancy.

[0285] B. Data Analysis

[0286] In some embodiments, the present disclosure provides a system, method, or kit having data analysis implemented as a software application, computing hardware, or both. In various embodiments, the analysis application or system comprises at least a data receiving module, a data preprocessing module, a data analysis module (which may operate on one or more types of genomic data), a data interpretation module, or a data visualization module. In some embodiments, the data receiving module may include a computer system that connects laboratory hardware or instruments to a computer system that processes laboratory data. In some embodiments, the data preprocessing module may include a hardware system or computer software that performs operations on the data to prepare it for analysis. Examples of operations that may be applied to the data in the preprocessing module include affine transformation, denoising, data cleaning, reformatting, or subsampling. A data analysis module that may be specialized for the analysis of genomic data from one or more genomic materials may identify abnormal patterns associated with diseases, pathologies, conditions, risks, pathological states, or phenotypes by, for example, receiving assembled genomic sequences and performing probabilistic and statistical analysis. The data interpretation module may utilize analytical methods derived, for example from statistics, mathematics, or biology, to support an understanding of the relationships between identified abnormal patterns and health status, functional status, prognosis, or risk. The data visualization module may facilitate the understanding or interpretation of results by generating visual representations of the data using mathematical modeling, computer graphics, or rendering methods.

[0287] In various embodiments, a machine learning method is applied to distinguish samples within a sample group. In one embodiment, a machine learning method is applied to distinguish samples between healthy samples and samples of progressive adenoma.

[0288] In one embodiment, one or more machine learning tasks used to train a methylation-based prediction engine include one or more of a generalized linear model, a generalized additive model, a nonparametric regression task, a random forest classifier, a spatial regression task, a Bayesian regression model, a time series analysis, a Bayesian network, a Gaussian network, a decision tree learning task, an artificial neural network, a recurrent neural network, a reinforcement learning task, a linear / non-linear regression task, a support vector machine, a clustering task, and a genetic algorithm task.

[0289] In various embodiments, the computer processing method is selected from logistic regression, multiple linear regression (MLR), dimensionality reduction, partial least squares (PLS) regression, principal component regression, autoencoder, variant autoencoder, singular value decomposition, Fourier base, wavelet, discriminant analysis, support vector machine, decision tree, classification and regression tree (CART), tree-based method, random forest, gradient boost tree, logistic regression, matrix decomposition, multidimensional scaling (MDS), dimensionality reduction method, t-distribution stochastic neighborhood embedding (t-SNE), multilayer perceptron (MLP), network clustering, neuro-fuzzy, and artificial neural networks.

[0290] In some embodiments, the method disclosed herein may include computer analysis of nucleic acid sequencing data of samples from an individual or a number of individuals. By identifying variants inferred from sequence data through analysis, sequence variants may be identified based on probabilistic modeling, statistical modeling, mechanical modeling, network modeling, or statistical inference. Non-limiting examples of analysis methods include principal component analysis, autoencoders, singular value decomposition, Fourier bases, wavelets, discriminant analysis, regression, support vector machines, tree-based methods, networks, matrix decomposition, and clustering. Non-limiting examples of variants include germline variants or somatic mutations. In some embodiments, variants may refer to already known variants. Already known variants may be scientifically identified or reported in the literature. In some embodiments, variants may refer to presumed variants associated with biological changes. Biological changes may be known or unknown. In some embodiments, presumed variants may be reported in the literature but not yet biologically identified.

[0291] Alternatively, presumed variants may be inferred based on computer analysis disclosed herein, although they are not reported in the literature. In some embodiments, germline variants may refer to nucleic acids that induce natural or normal variations.

[0292] Natural or normal variations may include, for example, skin color, hair color, and normal body weight. In some embodiments, somatic mutations may refer to nucleic acids that induce acquired or abnormal variations. Acquired or abnormal variations may include, for example, cancer, obesity, pathology, symptoms, diseases, and disorders. In some embodiments, the analysis may include distinguishing between germline variants. Germline variants may include, for example, individual variants and somatic mutations. In some embodiments, identified variants may be used by clinicians or other healthcare professionals to improve medical methodologies, diagnostic accuracy, and cost reduction.

[0293] Additionally, the present invention provides an improved method and a computing system or software medium capable of distinguishing sequencing errors, somatic mutations, and germline variants within nucleic acids introduced through amplification and / or sequencing techniques. The provided method may include the step of simultaneously calling up and scoring variants from aligned sequencing data of all samples obtained from a patient.

[0294] Samples obtained from subjects other than patients may also be used. Other samples may also be collected from subjects previously analyzed by sequencing assays or targeted sequencing assays (e.g., targeted resequencing assays). The methods, computing systems, or software media disclosed herein can improve the identification and accuracy of variants or mutations (e.g., germline or somatic variants or mutations including copy number variants, single nucleotide variants, indels, and gene fusions), and can lower the detection lower limit by reducing the number of false positive and false negative identifications.

[0295] C. Create Classifier

[0296] In one aspect, the present system and method provide a classifier generated based on feature information derived from methylation sequence analysis from a biological sample of cfDNA prepared by the ssDNA library preparation method described herein. The classifier forms part of a prediction engine for distinguishing groups within a population based on methylation sequence features identified in a biological sample such as cfDNA.

[0297] In some embodiments, the classifier normalizes methylation information by formatting similar parts of the methylation information into a combined format and a combined scale; stores the normalized methylation information in a columnar database; trains a methylation prediction engine by applying one or more machine learning tasks to the stored normalized methylation information, and the methylation prediction engine maps combinations of one or more features for specific groups; applies the methylation prediction engine to accessed field information to identify methylation associated with the groups; and is generated by classifying objects into groups.

[0298] Specificity can be defined as the probability of obtaining a negative test result among those without disease. Specificity is equal to the number of disease-free individuals who received a negative result divided by the total number of disease-free individuals.

[0299] In various embodiments, the model, classifier, or prediction test has a specificity of at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 97%, at least 98%, or at least 99%.

[0300] Sensitivity can be defined as the probability of a positive test result among individuals with the disease. Sensitivity is equal to the number of individuals with the disease who tested positive divided by the total number of individuals with the disease.

[0301] In various embodiments, the model, classifier, or predictive test has a sensitivity of at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 97%, at least 98%, or at least 99%.

[0302] In some embodiments, the group is selected from healthy state (asymptomatic), cancer, intestinal-related disease, immune-mediated inflammatory disease, neurological disease, kidney disease, prenatal disease, and metabolic disease.

[0303] D. Digital Processing Unit

[0304] In some embodiments, the subject matter described herein may include a digital processing device or a use thereof. In some embodiments, the digital processing device may include one or more hardware central processing units (CPUs), graphics processing units (GPUs), or tensor processing units (TPUs) that perform the functions of the device. In some embodiments, the digital processing device may include an operating system configured to execute executable instructions. In some embodiments, the digital processing device may optionally be connected to a computer network. In some embodiments, the digital processing device may optionally be connected to the Internet to access the World Wide Web. In some embodiments, the digital processing device may optionally be connected to a cloud computing infrastructure. In some embodiments, the digital processing device may optionally be connected to an intranet. In some embodiments, the digital processing device may optionally be connected to a data storage device.

[0305] Non-limiting examples of suitable digital processing devices include server computers, desktop computers, notebook computers, subnotebook computers, netbook computers, netpad computers, set-top computers, handheld computers, internet appliances, mobile smartphones, and tablet computers. Suitable tablet computers may include, for example, computers in booklet, slate, and convertible configurations.

[0306] In some embodiments, the digital processing device may include an operating system configured to execute executable instructions. For example, the operating system may include software containing programs and data that manage the device's hardware and provide services for the execution of applications. Non-limiting examples of operating systems include Ubuntu, FreeBSD, OpenBSD, NetBSD®, Linux, Apple® Mac OS X Server®, Oracle® Solaris®, Windows Server®, and Novell® NetWare®. Non-limiting examples of suitable personal computer operating systems include Microsoft® Windows®, Apple® Mac OS X®, UNIX®, and UNIX-family operating systems, such as GNU / Linux®. In some embodiments, the operating system may be provided by cloud computing, and cloud computing resources may be provided by one or more service providers.

[0307] In some embodiments, the device may include a storage and / or memory device. The storage and / or memory device may be one or more physical devices used to store data or programs temporarily or permanently. In some embodiments, the device may be volatile memory and requires power to retain stored information. In some embodiments, the device may be non-volatile memory and retains stored information even when power is not supplied to the digital processing device. In some embodiments, the non-volatile memory may include flash memory. In some embodiments, the non-volatile memory may include dynamic random access memory (DRAM). In some embodiments, the non-volatile memory may include ferroelectric random access memory (FRAM). In some embodiments, the non-volatile memory may include phase change random access memory (PRAM). In some embodiments, the device may be a storage device including, for example, a CD-ROM, a DVD, a flash memory device, a magnetic disk drive, a magnetic tape drive, an optical disk drive, and a cloud computing-based storage. In some embodiments, the storage and / or memory device may be a combination of devices such as those disclosed herein.

[0308] In some embodiments, the digital processing device may include a display for sending visual information to a user. In some embodiments, the display may be a cathode ray tube (CRT). In some embodiments, the display may be a liquid crystal display (LCD). In some embodiments, the display may be a thin-film transistor liquid crystal display (TFT-LCD). In some embodiments, the display may be an organic light-emitting diode (OLED) display. In some embodiments, the OLED display may be a passive matrix OLED (PMOLED) or active matrix OLED (AMOLED) display. In some embodiments, the display may be a plasma display. In some embodiments, the display may be a video projector. In some embodiments, the display may be a combination of devices such as those disclosed herein.

[0309] In some embodiments, the digital processing device may include an input device for receiving information from a user. In some embodiments, the input device may be a keyboard. In some embodiments, the input device may be a pointing device including, for example, a mouse, a trackball, a trackpad, a joystick, a game controller, or a stylus. In some embodiments, the input device may be a touch screen or a multi-touch screen. In some embodiments, the input device may be a microphone for capturing voice or other sound input. In some embodiments, the input device may be a video camera for capturing motion or visual input. In some embodiments, the input device may be a combination of devices such as those disclosed herein.

[0310] E. Non-transient computer-readable storage media

[0311] In some embodiments, the subject matter disclosed herein may comprise one or more non-transient computer-readable storage media coded with a program containing instructions executable by an operating system of a digital processing device optionally connected to a network. In some embodiments, the computer-readable storage media may be a tangible component of the digital processing device. In some embodiments, the computer-readable storage media may optionally be detachable from the digital processing device. In some embodiments, the computer-readable storage media may include, for example, a CD-ROM, a DVD, a flash memory device, a solid-state memory (SSD), a magnetic disk drive, a magnetic tape drive, an optical disk drive, a cloud computing system and service, etc. In some embodiments, the program and instructions may be coded on the media permanently, substantially permanently, semi-permanently, or non-transiently.

[0312] F. Computer Systems

[0313] The present disclosure provides a computer system programmed to implement the method of the present disclosure. Fig. 11 The invention illustrates a computer system (1101) programmed or otherwise configured to store, process, identify, or interpret patient data, biological data, biological sequences, or reference sequences. The computer system (1101) may process various aspects of the patient data, biological data, biological sequences, or reference sequences of the present disclosure. The computer system (1101) may be a user's electronic device or a computer system located remotely from the electronic device. The electronic device may be a mobile electronic device.

[0314] The computer system (1101) includes a central processing unit (CPU, also referred to herein as "processor" and "computer processor") (1105), which may be a single-core or multi-core processor or a plurality of processors for parallel processing. The computer system (1101) also includes memory or memory location (1110) (e.g., random access memory, read-only memory, flash memory), electronic storage unit (1115) (e.g., hard disk), communication interface (1120) for communicating with one or more other systems (e.g., network adapter), and peripheral devices (1125) such as a cache, other memory, data storage and / or electronic display adapter. The memory (1110), storage unit (1115), interface (1120), and peripheral devices (1125) communicate with the CPU (1105) via a communication bus (solid line), such as a motherboard. The storage unit (1115) may be a data storage unit (or data repository) for storing data. A computer system (1101) can be operably coupled to a computer network (“Network”) (1130) with the help of a communication interface (1120). The network (1130) may be the Internet, an intranet and / or extranet, or an intranet and / or extranet communicating with the Internet. In some embodiments, the network (1130) is a communication and / or data network. The network (1130) may include one or more computer servers, which may enable distributed computing such as cloud computing. In some embodiments, the network (1130) may implement a peer-to-peer network with the help of the computer system (1101), which allows a device coupled to the computer system (1101) to act as a client or server.

[0315] The CPU (1105) may execute a series of machine-readable instructions that may be implemented as a program or software. The instructions may be stored in a memory location such as memory (1110). The instructions may be passed to the CPU (1105), and the CPU (1105) may then be programmed or otherwise configured to implement the method of the present disclosure. Examples of operations performed by the CPU (1105) may include fetching, decoding, execution, and writeback.

[0316] The CPU (1105) may be part of a circuit such as an integrated circuit. One or more other components of the system (1101) may be included in the circuit. In some embodiments, the circuit is an application-specific integrated circuit (ASIC).

[0317] The storage unit (1115) can store files such as drivers, libraries, and stored programs. The storage unit (1115) can store user data, for example, user preferences and user programs. In some embodiments, the computer system (1101) may include one or more additional data storage units located outside the computer system (1101), such as being located on a remote server that communicates with the computer system (1101) via an intranet or the Internet.

[0318] The computer system (1101) can communicate with one or more remote computer systems through a network (1130). For example, the computer system (1101) can communicate with a user's remote computer system. Examples of remote computer systems include a personal computer (e.g., a portable PC), a slate or tablet PC (e.g., an Apple® iPad, a Samsung® Galaxy Tab), a telephone, a smartphone (e.g., an Apple® iPhone, an Android-enabled device, a Blackberry®), or a personal digital assistant. The user can access the computer system (1101) through the network (1130).

[0319] The method described herein may be implemented by machine-executable code (e.g., computer processor) stored in an electronic storage location of a computer system (1101), e.g., memory (1110) or an electronic storage unit (1115). Machine-executable or machine-readable code may be provided in the form of software. During use, the code may be executed by the processor (1105). In some embodiments, the code may be retrieved from the storage unit (1115) and stored in memory (1110) so that the processor (1105) can access it immediately. In some embodiments, the electronic storage unit (1115) may be omitted, and machine-executable instructions are stored in memory (1110).

[0320] The code may be pre-compiled and configured for use on a machine having a processor suited to execute the code, or it may be interpreted or compiled at runtime. The code may be provided in a programming language that can be selected to execute it in a pre-compiled, interpreted, or compiled manner.

[0321] Aspects of the systems and methods provided herein, such as the computer system (1101), may be implemented through programming. Various aspects of the technology may be considered as “products” or “manufactured articles” in the form of machine (or processor) executable code and / or related data typically contained in or implemented on a type of machine-readable medium. Machine-executable code may be stored in memory (e.g., read-only memory, random access memory, flash memory) or electronic storage devices such as hard disks. A medium of the “storage” type may include some or all of the tangible memory of a computer, processor, etc., or related modules thereof, such as various semiconductor memory, tape drives, disk drives, etc., which can provide non-transient storage at any time for software programming. All or part of the software may be communicated from time to time via the Internet or various other communication networks. Such communication enables the loading of software from one computer or processor to another computer or processor, for example, from a management server or host computer to a computer platform of an application server. Accordingly, another type of medium capable of containing software elements includes optical, electric, and electromagnetic waves used across physical interfaces between local devices, or through wired and optical wired networks and various wireless links. Physical elements that transmit these waves, such as wired or wireless links, optical links, etc., may also be considered as media containing software. As used herein, terms such as computer or machine "readable medium" refer to any medium that participates in providing instructions to a processor for execution, unless limited to non-transient and tangible "storage" media.

[0322] Accordingly, machine-readable media, such as computer-executable code, may take various forms including, but not limited to, tangible storage media, carrier media, or physical transmission media. Non-volatile storage media include, for example, optical or magnetic disks, such as any storage device in any computer(s), which may be used to implement the database, etc., illustrated in the drawings. Volatile storage media include dynamic memory, such as the main memory of such computer platforms. Tangible transmission media include coaxial cables, copper wires, and optical fibers, and include wiring that constitutes a bus within a computer system. Carrier transmission media may take the form of electrical or electromagnetic signals, or acoustic or light waves, such as those generated during radio frequency (RF) and infrared (IR) data communication. Accordingly, common forms of computer-readable media include, for example, floppy disks, flexible disks, hard disks, magnetic tapes, any other magnetic media, CD-ROMs, DVDs or DVD-ROMs, any other optical media, punch cards, paper tapes, any other physical storage media with a punch pattern, RAM, ROM, PROM and EPROM, FLASH-EPROM, any other memory chips or cartridges, carriers carrying data or instructions, cables or links carrying such carriers, or any other media from which a computer can read programming code and / or data. Many of these forms of computer-readable media may be involved in delivering one or more sequences of one or more instructions to a processor for execution.

[0323] The computer system (1101) may include or communicate with an electronic display (1135) that includes, for example, a nucleic acid sequence, an enriched nucleic acid sample, an expression profile, and an analysis of the expression profile, a user interface (UI) (1140). Examples of the UI include, but are not limited to, a graphical user interface (GUI) and a web-based user interface.

[0324] The method and system of the present disclosure may be implemented by one or more algorithms. The algorithm may be implemented by software when executed by a central processing unit (1105). The algorithm may, for example, probe a plurality of regulatory elements, sequence a nucleic acid sample, concentrate a nucleic acid sample, determine an expression profile of a nucleic acid sample, analyze an expression profile of a nucleic acid sample, and store or propagate the results of the expression profile analysis.

[0325] In some embodiments, the subject matter disclosed herein may include at least one computer program or use thereof. A computer program may be a series of instructions written to perform a specified task and executable on a CPU, GPU, or TPU of a digital processing unit. Computer-readable instructions may be implemented as program modules, such as functions, objects, application programming interfaces (APIs), data structures, etc., that perform a specific task or implement a specific abstract data type. A computer program may be written in various versions of various languages.

[0326] The functions of computer-readable instructions can be combined or distributed as desired in various environments. In some embodiments, a computer program may include a single sequence of instructions. In some embodiments, a computer program may include multiple sequences of instructions. In some embodiments, a computer program may be provided from a single location. In some embodiments, a computer program may be provided from multiple locations. In some embodiments, a computer program may include one or more software modules. In some embodiments, a computer program may include, in part or wholly, one or more web applications, one or more mobile applications, one or more standalone applications, one or more web browser plugins, extensions, add-ins or add-ons, or a combination thereof.

[0327] In some embodiments, computer processing may be a method of statistics, mathematics, biology, or any combination thereof. In some embodiments, computer processing methods include, for example, logistic regression, dimensionality reduction, principal component analysis, autoencoders, singular value decomposition, Fourier basis, wavelets, discriminant analysis, support vector machines, tree-based methods, random forests, gradient boost trees, matrix decomposition, network clustering, and neural networks.

[0328] In some embodiments, the computer processing method is a supervised machine learning method including, for example, regression, support vector machines, tree-based methods, and networks.

[0329] In some embodiments, the computer processing method is an unsupervised machine learning method including, for example, clustering, networking, principal component analysis, and matrix decomposition.

[0330] G. Database

[0331] In some embodiments, the subject matter disclosed herein may include one or more databases or be used to store patient data, biological data, biological sequences, or reference sequences. Reference sequences may be derived from databases. Various databases may be suitable for storing and retrieving sequence information. In some embodiments, suitable databases may include, for example, relational databases, non-relational databases, object-oriented databases, object databases, entity-relationship model databases, association databases, and XML databases. In some embodiments, the database may be Internet-based. In some embodiments, the database may be Web-based. In some embodiments, the database may be cloud computing-based. In some embodiments, the database may be based on one or more local computer storage devices.

[0332] IV. Cancer Detection and Diagnosis

[0333] The trained machine learning methods, models, and discriminative classifiers described herein are useful for various medical applications, including cancer detection, diagnosis, and treatment responsiveness. Since the models are trained with individual metadata and analyte-derived features, applications can be customized to stratify entities within a population and guide treatment decisions accordingly.

[0334] A. Detection / Diagnosis

[0335] The method and system provided herein can generate detection and / or diagnosis results of a subject having cancer (e.g., CRC) or other indications by analyzing data obtained from a subject (patient) by performing predictive analysis using an artificial intelligence-based approach. For example, the present application can provide a diagnosis that a subject has cancer by detecting cancer through the application of a predictive algorithm to the obtained data. The predictive algorithm may include an artificial intelligence-based predictor, such as a machine learning-based predictor configured to process the obtained data and diagnose a subject having cancer.

[0336] A machine learning predictor can be trained using a dataset from one or more cohort sets of patients with cancer as input, for example, a dataset generated by performing multiple analyte tests on biological samples of the subjects, and using the subjects' known diagnostic results (e.g., staging and / or tumor fraction) as output to the machine learning predictor.

[0337] It may be generated from one or more sets of subjects. The training dataset may include feature sets and labels corresponding to features related to diagnosis. Features may include characteristics such as specific ranges or categories of cfDNA assay measures, such as the number of cfDNA fragments in biological samples obtained from healthy and diseased samples that overlap with or belong to each set of bins (genome windows) of the reference genome. For example, a feature set collected from a given subject at a given time point may collectively serve as a diagnostic signature, which may indicate the identified cancer of the subject at that time point. Features may also include labels indicating the diagnostic results of the subject, such as in the case of one or more cancers.

[0338] The label may include results such as the subject's known diagnoses (e.g., staging and / or tumor fraction). The results may include features associated with the subject's cancer. For example, features may indicate that the subject has one or more cancers.

[0339] A training set (e.g., a training dataset) may be selected by random sampling of datasets corresponding to one or more subject sets (e.g., a retrospective and / or prospective cohort of patients with or without one or more cancers). Alternatively, a training set (e.g., a training dataset) may be selected by proportional sampling of datasets corresponding to one or more subject sets (e.g., a retrospective and / or prospective cohort of patients with or without one or more cancers). The training set may be balanced among datasets corresponding to one or more subject sets (e.g., patients from different clinical institutions or clinical trials). A machine learning predictor may be trained until specific predetermined conditions for accuracy or performance are met, such as having a minimum requirement value corresponding to a diagnostic accuracy measure. For example, the diagnostic accuracy measure may correspond to the diagnosis of one or more cancers in the subject, staging, or prediction of the tumor fraction.

[0340] Examples of detection and diagnostic accuracy measures may include sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and area under the ROC curve (AUC) corresponding to diagnostic accuracy regarding the detection or prediction of cancer (e.g., colorectal cancer).

[0341] In another aspect, the present disclosure provides a method for detecting or identifying cancer in a subject, the method comprising: (a) providing a biological sample comprising ssDNA molecules of a cell-free DNA sample derived from the subject; (b) methylating and sequencing the ssDNA molecules from the subject to generate a plurality of sequencing reads; (c) aligning the sequencing reads to a reference genome; (d) generating quantitative measurements of the sequencing reads in each of a first plurality of genomic regions of the reference genome to generate a first feature set, wherein the first plurality of genomic regions of the reference genome comprises at least about 10 distinct regions; and (e) applying a trained algorithm to the first feature set to generate the likelihood that the subject has the cancer.

[0342] For example, these predetermined conditions may include values ​​for sensitivity to predict cancer (e.g., colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer) such as at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%.

[0343] As another example, these predetermined conditions may include values ​​of specificity for predicting cancer (e.g., colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer) such as, for example, at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%.

[0344] As another example, these predetermined conditions may include a positive payoff value (PPV) for predicting cancer (e.g., colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer) of, for example, at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%.

[0345] As another example, these predetermined conditions may include a negative predictive value (NPV) for predicting cancer (e.g., colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer) of, for example, at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%.

[0346] As another example, these predetermined conditions may include an AUC of an ROC curve predicting cancer (e.g., colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer) of about 0.50, at least about 0.55, at least about 0.60, at least about 0.65, at least about 0.70, at least about 0.75, at least about 0.80, at least about 0.85, at least about 0.90, at least about 0.91, at least about 0.92, at least about 0.93, at least about 0.94, at least about 0.95, at least about 0.96, at least about 0.97, at least about 0.98, or at least about 0.99.

[0347] In some examples of any of the aforementioned aspects, the method further includes a step of monitoring the progression of the disease in the subject, and the monitoring is based at least partially on genetic sequence features. In some examples, the disease is cancer.

[0348] In some examples of any of the aforementioned aspects, the method further includes a step of determining the tissue origin of the cancer of the subject, and the determination is based at least partially on genetic sequence features.

[0349] In some examples of any of the aforementioned aspects, the method further includes a step of estimating the tumor burden of the subject, and the estimation is based at least partially on genetic sequence features.

[0350] B. Treatment Responsiveness

[0351] The predictive classifiers, systems, and methods described herein are useful for classifying populations for a number of clinical applications (e.g., based on performing multiple analyte tests on biological samples of individuals). Examples of such clinical applications include early-stage cancer detection, cancer diagnosis, classification of cancer into specific stages of disease, or determination of responsiveness or resistance to therapeutic agents for cancer treatment.

[0352] The methods and systems described herein are applicable to various types of cancer, similar to grades and stages, and are therefore not limited to a single type of cancer disease. Accordingly, the combination of analytes and assays in the systems and methods described herein can be used to predict the responsiveness of cancer therapeutic agents across different types of cancer in different tissues and to classify individuals based on therapeutic responsiveness. In one example, the classifier described herein stratifies groups of individuals into therapeutic responders and non-responders.

[0353] The present disclosure also provides a method for determining a drug target for a condition or disease of interest (e.g., a gene related to a specific class or important gene), the method comprising the steps of: evaluating a sample obtained from an individual for a gene expression level for at least one gene; and identifying one or more drug targets related to the classification by determining the gene related to the classification of the sample using a neighborhood analysis method.

[0354] The present disclosure also provides a method for determining the efficacy of a drug designed to treat a class of disease, the method comprising: obtaining a sample from an individual having a class of disease; exposing the sample to a drug; evaluating the level of gene expression for at least one gene of the sample exposed to the drug; and classifying the sample exposed to the drug into a predetermined class of disease as a function of the relative level of gene expression of the sample relative to the level of gene expression of the model, using a computer model constructed by a weighted voting method.

[0355] The present disclosure also provides a method for determining the efficacy of a drug designed to treat a class of diseases, wherein an individual is exposed to the drug, and the method comprises the steps of: obtaining a sample from an individual exposed to the drug; evaluating the gene expression level of the sample for at least one gene; and classifying the sample into a predetermined class of diseases, including evaluating the gene expression level of the sample in comparison to the gene expression level of the model using a model constructed by a weighted voting method.

[0356] Another application is a method for determining whether an individual belongs to a phenotypic class (e.g., intelligence, response to treatment, lifespan, susceptibility to viral infection, or obesity), said method comprising the steps of: obtaining a sample from an individual; evaluating the sample for a gene expression level for at least one gene; and classifying the sample into a predetermined class of disease, including evaluating the gene expression level of the sample in comparison to the gene expression level of the model using a model constructed by a weighted voting method.

[0357] Biomarkers can be useful for predicting the prognosis of colorectal cancer patients. The ability to classify patients into high-risk (poor prognosis) or low-risk (good prognosis) groups enables the selection of appropriate therapies for these patients. For example, high-risk patients are likely to benefit from aggressive therapies, whereas the therapy may not offer significant benefits to low-risk patients.

[0358] Predictive biomarkers can guide treatment decisions by identifying a subset of patients who may be "exceptional responders" to a specific cancer therapy or individuals who can benefit from alternative treatment modalities.

[0359] In one aspect, the systems and methods described herein relating to classifying populations based on therapeutic responsiveness refer to, but are not limited to, cancers treated with chemotherapy agents of the class of DNA damaging agents, DNA repair targeted therapies, DNA damage signaling inhibitors, DNA damage-induced cell cycle arrest inhibitors, and inhibitors of processes that indirectly cause DNA damage. Each of these chemotherapy agents may be considered a "DNA damage treatment."

[0360] Patient analysis data are classified into high-risk and low-risk patient groups, such as patients at high or low risk of clinical recurrence, and the results can be used to determine the course of treatment. For example, patients determined to be at high risk may be treated with adjuvant chemotherapy after surgery. For patients considered to be at low risk, adjuvant chemotherapy after surgery may be withheld. Accordingly, the present disclosure provides a method for constructing a gene expression profile of a colon cancer tumor indicating a risk of recurrence in certain aspects.

[0361] In various examples, the classifier described herein stratifies populations between responders and non-responders to treatment.

[0362] In various examples, therapeutic agents are selected from alkylating agents, plant alkaloids, antitumor antibiotics, antimetabolites, topoisomerase inhibitors, retinoids, checkpoint inhibitor therapies, and VEGF inhibitors.

[0363] Examples of therapeutic agents for which populations can be stratified into responders and non-responders include, but are not limited to: sorafenib, regorafenib, imatinib, eribulin, gemcitabine, capecitabine, pazopanib, lapatinib, dabrafenib, sunitinib, crizotinib, everolimus, torisirolimus, sirolimus, axitinib, gefitinib, anastrozole, bicalutamide, fulvestrant, raltitrexed, pemetrexed, goserelin acetate, erlotinib, vemurafenib, bismodegib, tamoxifen citrate, paclitaxel, docetaxel, cabazitaxel, oxaliplatin, ziv-aflibercept, bevacizumab, trastuzumab, pertuzumab, panitumumab, taxane, bleomycin, melphalan, Plumbagin, Camptosar, Mitomycin-C, Mithoxantrone, Poly(styrene-maleate)-conjugated Neocarzinostatin (SMANCS), Doxorubicin, Pegylated Doxorubicin, FOLFORI, 5-Fluoruracil, Temozolomide, Fasireotide, Tegafur, Gimeracil, Oteracil, Itraconazole, Bortezomib, Lenalidomide, Irinotecan, Epirubicin, Romidepsin, Resminostat, Tasquinimod, Refametinib, Lapatinib, Tyverb®, Arenegyr, NGR-TNF, Fasireotide, Signifor®, Ticilimumab, Tremelimumab, Lansoprazole, PrevOnco®, ABT-869, Linifanib, Borolanib, Tivantinib, Tarceva®, Erlotinib, Chemotherapy agents including Stivarga®, regorafenib, fluoro-sorafenib, brivanib, liposomal doxorubicin, lenvatinib, ramucirumab, peretinoin, mufarfostat, Teysuno®, tegafur, Gimeracil, oteracil, and orantinib; and antibody therapies including, but not limited to, alemtuzumab, atezolizumab, ipilimumab, nivolumab, ofatumumab, pembrolizumab, or rituximab.

[0364] In another example, the population can be stratified into responders and non-responders to checkpoint inhibitor therapy, such as compounds that bind to PD-1 or CTLA4.

[0365] In another example, the population can be stratified into responders and non-responders to anti-VEGF therapy that binds to VEGF pathway targets.

[0366] V. Indications

[0367] In some examples, the biological condition may include a disease. In some examples, the biological condition may be the stage of a disease. In some examples, the biological condition may be a gradual change in biological state. In some examples, the biological condition may be a therapeutic effect. In some examples, the biological condition may be a drug effect. In some examples, the biological condition may be a surgical effect. In some examples, the biological condition may be a biological state following a lifestyle change. Non-limiting examples of lifestyle changes include dietary changes, smoking changes, and sleep pattern changes. In some examples, the biological condition is unknown. The analysis described herein may include machine learning to infer or interpret unknown biological conditions.

[0368] In one example, the present system and method are particularly useful for applications related to colon cancer (cancer that forms in the tissue of the colon (the longest part of the large intestine)). Most colon cancers are adenocarcinomas (cancers that originate from cells that form the inner wall of an internal organ and have glandular characteristics). The progression of cancer is characterized by the stage, that is, the extent of the cancer within the body. Staging is generally based on the size of the tumor, whether the cancer is contained in lymph nodes, and whether the cancer has spread from its original location to other parts of the body. The stages of colon cancer include stages I, II, III, and IV. Unless otherwise specified, the term “colon cancer” refers to stage 0, stage I, stage II (including stages IIA or IIB), stage III (including stages IIIA, IIB, or IIIC), or stage IV colon cancer. In some examples of the present invention, colon cancer is from any stage. In one example, the colon cancer is stage I colorectal cancer. In one example, the colon cancer is stage II colorectal cancer. In one example, the colon cancer is stage III colorectal cancer. In another example, the colon cancer is stage IV colorectal cancer.

[0369] Pathological conditions that can be inferred by the disclosed method include, for example, cancer, intestinal-related diseases, immune-mediated inflammatory diseases, neurological diseases, kidney diseases, prenatal diseases, and metabolic diseases.

[0370] In some examples, the method of the present disclosure may be used to diagnose cancer. Non-limiting examples of cancer include adenoma (adenomatous polyp), sessile adenoma (SSA), progressive adenoma, colorectal dysplasia, colorectal adenoma, colorectal cancer, colon cancer, rectal cancer, colorectal carcinoma, colorectal adenocarcinoma, carcinoid tumor, gastrointestinal carcinoid tumor, gastrointestinal stromal tumor (GIST), lymphoma, and sarcoma.

[0371] Non-limiting examples of cancers that can be inferred by the disclosed method and system include acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), adrenocortical carcinoma, Kaposi sarcoma, anal cancer, basal cell carcinoma, biliary tract cancer, bladder cancer, bone cancer, osteosarcoma, malignant fibrous histiocytoma, brainstem glioma, brain cancer, craniopharyngioma, ependymoma, ependymoma, medulloblastoma, medulloblastoma, pineal parenchymal tumor, breast cancer, bronchial tumor, Burkitt lymphoma, non-Hodgkin lymphoma, carcinoid tumor, cervical cancer, chordoma, chronic lymphoblastic leukemia (CLL), chronic myeloid leukemia (CML), colon cancer, colorectal cancer, cutaneous T-cell lymphoma, tubular carcinoma in situ, endometrial cancer, esophageal cancer, Ewing sarcoma, eye cancer, intraocular melanoma, retinoblastoma, fibrous histiocytoma, gallbladder cancer, gastric cancer, Glioma, blastocyst leukemia, head and neck cancer, heart cancer, hepatocellular carcinoma, Hodgkin lymphoma, hypopharyngeal cancer, kidney cancer, laryngeal cancer, lip cancer, oral cavity cancer, lung cancer, non-small cell carcinoma, small cell carcinoma, melanoma, mouth cancer, myelodysplastic syndrome, multiple myeloma, medulloblastoma, nasal cavity cancer, sinus cancer, neuroblastoma, nasopharyngeal cancer, oral cancer, oropharyngeal cancer, osteosarcoma, ovarian cancer, pancreatic cancer, papillomatosis, paraganglioma, parathyroid cancer, penile cancer, pharyngeal cancer, pituitary tumor, plasma cell neoplasm, prostate cancer, rectal cancer, renal cell carcinoma, rhabdomyosarcoma, salivary gland cancer, Sézary syndrome, skin cancer, small intestine cancer, soft tissue sarcoma, squamous cell carcinoma, testicular cancer, throat cancer, thymoma, thyroid cancer, urethral cancer, uterine cancer, uterine sarcoma, vaginal cancer, It includes vulvar cancer, Waldenström macroglobulinemia, and Wilms tumor.

[0372] Non-limiting examples of bowel-related diseases that can be inferred by the disclosed method and system include Crohn's disease, colitis, ulcerative colitis (UC), inflammatory bowel disease (IBD), irritable bowel syndrome (IBS), and celiac disease. In some examples, the diseases are inflammatory bowel disease, colitis, ulcerative colitis, Crohn's disease, microscopic colitis, collagenous colitis, lymphocytic colitis, bypass colitis, Behcet's disease, and unclassified colitis.

[0373] Non-limiting examples of immune-mediated inflammatory diseases that can be inferred by the disclosed method and system include psoriasis, sarcoidosis, rheumatoid arthritis, asthma, rhinitis (hay fever), food allergy, eczema, lupus, multiple sclerosis, fibromyalgia, type 1 diabetes, and Lyme disease. Non-limiting examples of neurological diseases that can be inferred by the disclosed method and system include Parkinson's disease, Huntington's disease, multiple sclerosis, Alzheimer's disease, stroke, epilepsy, neurodegeneration, and neuropathy.

[0374] Non-limiting examples of renal diseases that can be inferred by the disclosed method and system include interstitial nephritis, acute renal failure, and nephropathy. Non-limiting examples of prenatal diseases that can be inferred by the disclosed method and system include Down syndrome, aneuploidy, spina bifida, trisomy, Edwards syndrome, teratomas, sacrococcygeal teratomas (SCT), ventriculectasis, renal agenesis, cystic fibrosis, and fetal hydrops. Non-limiting examples of metabolic diseases that can be inferred by the disclosed method and system include cystinosis, Fabry disease, Gaucher disease, Lesch-Nyhann syndrome, Niemann-Pick disease, phenylketonuria, Pompe disease, and Tay-Sachs disease.

[0375] Specific details of specific examples may be combined in any suitable manner without departing from the spirit and scope of the disclosed examples of the concept of the invention. However, other examples of the concept of the invention may relate to specific examples or specific combinations of such individual aspects. All patents, patent applications, publications, and descriptions mentioned herein are incorporated by reference in their entirety for all purposes.

[0376] VI. Kit

[0377] The present disclosure provides a kit for identifying or monitoring cancer in a subject. The kit may include a probe for identifying quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of a plurality of cancer-associated genomic loci in a cell-free biological sample of a subject. Quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of a plurality of cancer-associated genomic loci in a cell-free biological sample may indicate one or more cancers. The probe may be selective for sequences at a plurality of cancer-associated genomic loci in a cell-free biological sample. The kit may include instructions for processing the cell-free biological sample using the probe to generate a dataset indicating quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of a plurality of cancer-associated genomic loci in a cell-free biological sample of a subject. In one embodiment, the kit includes a primer set, PCR reaction components, sequencing reagents, minimally destructive conversion reagents, and library preparation reagents.

[0378] The probe in the kit may be selective for sequences at multiple cancer-associated genomic loci in a cell-free biological sample. The probe in the kit may be configured to selectively enrich nucleic acid (e.g., RNA or DNA) molecules corresponding to the multiple cancer-associated genomic loci. The probe in the kit may be a nucleic acid primer. The probe in the kit may have sequence complementarity with nucleic acid sequences from one or more of the multiple cancer-associated genomic loci or genomic regions. The multiple cancer-associated genomic loci or genomic regions may include 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 or more distinct cancer-associated genomic loci or genomic regions identified for targeted methylation sequencing. A plurality of cancer-associated genomic loci or genomic regions may include 30, 29, 28, 27, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3 or two or fewer distinct cancer-associated genomic loci or genomic regions identified for targeted methylation sequencing.

[0379] Instructions within the kit may include instructions for testing a cell-free biological sample using selective probes for sequences at multiple cancer-associated genomic loci within the cell-free biological sample. These probes may be nucleic acid molecules (e.g., RNA or DNA) that have sequence complementarity with nucleic acid sequences (e.g., RNA or DNA) from one or more of the multiple cancer-associated genomic loci. These nucleic acid molecules may be primers or enrichment sequences. Instructions for testing the cell-free biological sample may include instructions for generating a dataset indicating quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of the multiple cancer-associated genomic loci within the cell-free sample by processing the cell-free biological sample by performing array hybridization, polymerase chain reaction (PCR), or nucleic acid sequencing (e.g., DNA sequencing or RNA sequencing). Quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of the multiple cancer-associated genomic loci within the cell-free biological sample may indicate one or more cancers.

[0380] Instructions within the kit may include instructions for measuring and interpreting test readings that can generate a dataset representing quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of the multiple cancer-associated genomic loci within a cell-free biological sample by quantifying one or more of the multiple cancer-associated genomic loci. For example, quantification of array hybridization or polymerase chain reaction (PCR) corresponding to the multiple cancer-associated genomic loci may generate a dataset representing quantitative measurements (e.g., indicating presence, absence, or relative amount) of sequences at each of the multiple cancer-associated genomic loci within the cell-free biological sample. Test readings may include quantitative PCR (qPCR) values, digital PCR (dPCR) values, digital droplet PCR (ddPCR) values, fluorescence values, etc., or normalized values ​​thereof.

[0381] Examples

[0382] Example 1

[0383] DNA extraction and library preparation for targeted EM-seq using a conversion-allowable sequencing adapter / primer system.

[0384] A. DNA manufacturing

[0385] Starting material: 2 mL of plasma

[0386] DNA was extracted from 2 mL of plasma using QIAamp circulating nucleic acid DNA purification according to the manufacturer's protocol and eluted to a volume of 30 μL.

[0387] B. Degeneration from dsDNA to ssDNA

[0388] The extracted DNA samples (including dsDNA) were denatured to generate ssDNA by incubating the plates in a thermal cycle at 98°C for about 3 minutes. After removing them from the thermal cycle, the denatured samples were placed on a frozen plate block on ice.

[0389] C. Transition-allowing adapter ligation

[0390] In this example, a functional set of a conversion-allowing adapter, PCR primers, and sequencing primers Example 2 Tested as described in [document]. Sequencing library yields were determined for libraries generated with a conversion-allowable adapter or a 5mC-containing adapter.

[0391] A switch-allowing single-stranded DNA library adapter was prepared comprising both 5' and 3' adapter pairs, referred to as Adapter A and Adapter B, respectively ( Fig. 2 Adapter A consisted of an upper DNA oligonucleotide having a 5' terminal amino modifier containing a 12-carbon spacer. All cytosines in the upper adapter were 5mC. The 3' sequence of each upper adapter contained one of several mCp-containing motifs used to evaluate enzymatic oxidation and deamination performance. Each upper A adapter was annealed to a lower adapter (Adapter A') containing unmethylated cytosines. The lower adapter strand had a sequence completely complementary to the upper adapter strand, except for a random 7-nucleotide 5' overhang sequence. The lower adapter ends were modified with a 5' amino modifier and a 3' amino modifier containing a 6-carbon spacer. The Adapter B pair consisted of an upper DNA oligonucleotide having a 5' phosphate, followed by one of several mCp-containing motifs, a sequence with all cytosines methylated, and 3' dideoxycytosine. Each upper B adapter was annealed to a lower adapter (adapter B') containing unmethylated cytosine. The lower adapter strand had a sequence completely complementary to the upper adapter strand, except for a random 7-nucleotide 3' overhang sequence. All lower B adapter oligonucleotides were blocked with a 5' terminal amino modifier and a 3' amino modifier containing a 12-carbon spacer. Alternatively, a switch-allowing adapter can be created by excluding all cytosine from the upper adapter sequence.

[0392] Approximately 20 μL of cfDNA mass input was divided into 96-well plates. The plates were sealed, vortexed, centrifuged, and spin-down. Subsequently, the samples were placed in a thermal cycler at 98°C (lid temperature 105°C) for 3 minutes. The samples were removed from the thermal cycler and immediately placed on a frozen plate block over ice for 5 minutes. Next, 2 μL each of 5' and 3' single-stranded DNA adapter pairs were added to the ligation mixture, which was then added to each sample and mixed 10 times with a pipette. Next, 26 μL of ligation buffer was added to the samples and mixed 10 times with a pipette. The plates were sealed, centrifuged, spin-down, vortexed, centrifuged again, and spin-down. Subsequently, the plates were incubated in a thermal cycler at 37°C (lid temperature 45°C) for one hour.

[0393] After 1 hour of ligation incubation, the plates were removed from the thermal cycle for bead-based cleanup. First, a diluted bead mixture was prepared: 75 μL of 10 mM Tris-HCl (pH 8.5) was added to 58.5 μL of previously diluted and buffer-exchanged AMPure beads. Approximately 133.5 μL of the bead mixture was added to each sample. Subsequently, the sample-bead mixture was incubated on a workbench at room temperature for about 15 minutes. Then, the plates were placed on a Permagen bar magnet for 5 minutes or until the supernatant appeared clear. The supernatant was then removed without disturbing the magnetized bead pellets. While the plates were still on the magnet rack, 200 μL of freshly prepared 80% ethanol (EtOH) was added to each well. The EtOH was removed after 30 seconds. Next, this washing process was repeated (200 μL of ethanol was added and removed after 30 seconds). Subsequently, the plate was removed from the magnet, sealed, and pulse-spinning was performed using a centrifuge. The plate was placed back on the magnet, and the remaining EtOH was removed with a small pipette. The magnetized beads were dried on a workbench at room temperature for 2 minutes, after which they were removed from the magnet. Without a magnet, 16 μL of 10 mM Tris-HCl was added to the beads and thoroughly mixed with a pipette. The beads were incubated at room temperature for 5 minutes, after which the plate was placed back on the magnet. Once the supernatant became clear, 15 μL of the supernatant was transferred to a new plate.

[0394] D. Second strand synthesis

[0395] In an alternative embodiment, a second strand synthesis (SSS) operation may be added to convert the ssDNA library into a dsDNA library after adapter ligation and before the enzymatic oxidation reaction. Following the preparation of the ssDNA library, a bead purification cleanup was performed. Subsequently, KAPA PCR master mix, Tris-HCl, and 2 μL of 50 μM primers complementary to the 3' adapter were added to the library. The sample was vortexed, centrifuged to spin down, and then a single extension reaction was performed in a thermal cycler. Following this extension, the sample was washed with purification beads and eluted with elution buffer.

[0396] E. Second strand synthesis without cleanup

[0397] After preparing the ssDNA library, no cleanup was performed. Immediately after adapter ligation, the following were added to each sample: 1 μL of 50 μM primer complementary to the 3' adapter, 0.5 μL of dNTPs (from the Fast Start kit), 0.2 μL of Fast Taq polymerase, and water for volume expansion. The samples were vortexed, spun by centrifugation, and then extended in a thermal cycler. Following the extension, the samples were washed with purification beads and eluted with elution buffer.

[0398] F. Second-strand synthesis without cleanup using an alternative adapter

[0399] Single-strand ligation was performed as described above, provided that the lower adapter B sequence was truncated so as not to be completely complementary to the upper adapter B sequence. No cleanup was performed after the preparation of the ssDNA library. Immediately after adapter ligation, the following were added to each sample: 1 μL of 50 μM primer complementary to the 3' adapter, 0.5 μL of dNTP (from the Fast Start kit), 0.2 μL of Bst polymerase, and water to increase volume. The samples were vortexed and spun by centrifugation, followed by one PCR cycle on a thermal cycler. Following extension, the samples were washed with purification beads and eluted with elution buffer.

[0400] Whether single-stranded or double-stranded (regardless of whether a second strand is synthesized), an ssDNA library with a conversion-allowing adapter subsequently Example 2 It was used as a starting material for subsequent methylation conversion and sequencing reactions described in.

[0401] Example 2

[0402] Preparation of targeted EM-seq library.

[0403] A. Oxidation of 5-methylcytosine and 5-hydroxymethylcytosine

[0404] The TET2 reaction buffer was prepared according to the manufacturer's protocol. Subsequently, the TET2 reaction buffer was added to a tube of TET2 reaction buffer supplement and thoroughly mixed. On ice, the TET2 reaction buffer, oxidation supplement, oxidation booster, and TET2 enzyme Example 1It was added directly to the ssDNA library ("sample") prepared according to [the formula]. Subsequently, each sample mixture was thoroughly mixed by vortexing. After centrifuging the mixture briefly, iron solution was added to the mixture. Subsequently, the mixture was thoroughly mixed by vortexing or pipetting up and down, and centrifuged briefly. Then, the mixture was incubated in a thermal cycle at 37°C for 1 hour. Subsequently, the mixture was transferred to ice before treatment with 1 μl of stopping reagent, and the appearance of the mixture was yellow. Subsequently, the mixture was thoroughly mixed by vortexing or pipetting up and down at least 10 times, and centrifuged briefly. Finally, the mixture was incubated in a thermal cycle at 37°C for 30 minutes, followed by incubation at 4°C.

[0405] B. Cleanup of TET2-transformed DNA

[0406] After vortexing the NEBNext® sample purification beads, they were added to each sample and thoroughly mixed by moving them up and down with a pipette. The samples were incubated on a workbench at room temperature for at least 5 minutes. Subsequently, the beads were separated from the supernatant by placing the tubes against a suitable magnetic stand. After 5 minutes (or when the solution became clear), the supernatant was carefully removed and discarded to avoid disturbing the beads containing the DNA target. While the tubes remained on the magnetic stand, freshly prepared 80% ethanol was added to each tube. Before carefully removing and discarding the supernatant, the samples were incubated at room temperature for 30 seconds. The washing process was repeated once for a total of two washes. After the second wash, all visible liquid was removed using a p10 pipette tip. Subsequently, the tubes were placed on the magnetic stand with the lids open and the beads were air-dried for 2 minutes. Then, the tubes were removed from the magnetic stand. DNA was eluted from the beads using elution buffer. Elution buffer was added to each tube and thoroughly mixed by pipetting up and down 10 times. Subsequently, the samples were incubated at room temperature for at least 1 minute. If necessary, the samples were rapidly centrifuged to collect the liquid from the sides of the tubes before placing the tubes back on the magnetic stand. Then, the tubes were placed back on the magnetic stand. After 3 minutes (or whenever the solution became clear), the DNA eluted from the supernatant was transferred to a new PCR tube.

[0407] C. DNA denaturation

[0408] The DNA samples (including dsDNA) were denatured by incubating the plates in a heat exchanger at 85°C for 10 minutes. After removing them from the heat exchanger, the DNA samples were immediately placed on a frozen plate block on ice.

[0409] D. Deamination of Cytosine

[0410] APOBEC reaction buffer, bovine serum albumin (BSA), and APOBEC were added to the denatured DNA. Subsequently, the mixture was thoroughly mixed by vortexing or pipetting up and down at least 10 times, followed by centrifugation for a short period. The mixture was then incubated according to the following protocol: incubated at 4°C for 10 minutes, followed by a total of 46 increases of 1°C every 2 minutes and 15 seconds until the sample temperature reached 50°C. Subsequently, maintained at 50°C for 10 minutes, followed by maintenance at 4°C in a thermal cycler.

[0411] E. Treatment with heat-unstable proteinase K

[0412] To stop APOBEC activity, heat-unstable proteinase K was added to the samples after deamination, incubated at 37°C for 30 minutes for 15 minutes, and then inactivated at 65°C for 10 minutes. After the first TLPK treatment, the samples were amplified via PCR using indexing primers. Amplification could be performed for any number of cycles. After PCR, a second TLPK treatment was performed. After the second TLPK treatment, a bead purification cleanup was performed.

[0413] F. Cleanup of Deamination DNA

[0414] After vortexing the NEBNext® sample purification beads, they were added to each sample and thoroughly mixed by pipetting up and down at least 10 times. All liquid was carefully ejected from the tip during the final mixing. Subsequently, the samples were incubated on a workbench at room temperature for at least 5 minutes. After 5 minutes (or when the solution became clear), the supernatant was carefully removed and discarded. With the tubes placed on a magnetic stand, freshly prepared 80% ethanol was added to the tubes. Then, the samples were incubated at room temperature for 30 seconds before carefully removing and discarding the supernatant. The washing process was repeated once for a total of two washes. Next, the tubes were placed on a magnetic stand with the lids open, and the beads were air-dried for 90 seconds. Subsequently, the DNA target was eluted from the beads using elution buffer. Elution buffer was added to each tube and thoroughly mixed by pipetting up and down 10 times. The samples were incubated at room temperature for at least 1 minute. If necessary, the sample was rapidly centrifuged to collect the liquid from the side of the tube before placing the tube back on the magnetic stand. Then, the tube was placed back on the magnetic stand. After 3 minutes (or whenever the solution became clear), the eluted DNA target in the supernatant was transferred to a new PCR tube.

[0415] Example 3

[0416] Target capture and multiple amplification.

[0417] After quantification, the samples were pooled and enriched. Subsequently, targeted enrichment was performed on the enzymatically converted library to specifically enrich pre-identified DNA fragments containing target CpG sites using 5'-biotinized capture probes. These probes could be methylated, demethylated, or a combination of both. Hybrid selection was performed using the TWIST rapid hybridization target capture kit. Hybridization could occur at 58°C, 60°C, or other temperatures. Bead washing was performed at a heated temperature, e.g., hybridization temperature + 3°C. Following hybridization, the captured DNA fragments were amplified by PCR. The target capture library was sequenced on an Illumina Novaseq Sequencer for 2 x 150 cycle runs.

[0418] Example 4

[0419] Targeted methylation classification.

[0420] The raw data file was used for alignment and methylation calling to enable targeted methylation analysis of pre-identified regions of the genome. Whole-genome amplification of the enzymatically converted DNA was performed.

[0421] FASTQ files were mapped to a reference genome, and methylation scores were calculated for disease classification. Featured data, including sets of CpG sites associated with health, disease, disease status, and treatment responsiveness, were processed using machine learning models to identify classifiers that stratify individuals within a population based on hypermethylation model scores and / or hypomethylation model scores.

[0422] Example 5

[0423] Comparison of ssDNA library preparation and dsDNA library preparation in methylation studies

[0424] A. Advanced adenoma (AA) and colorectal cancer (CRC)

[0425] Blood samples were obtained from 11 healthy human subjects, 23 human subjects with advanced adenoma (AA), and 30 human subjects with colorectal cancer (CRC). DNA was extracted from the blood samples, and libraries were prepared according to (i) the ssDNA library preparation method described in Example 1 and (ii) another dsDNA library preparation method to produce an ssDNA library and a dsDNA library, respectively (E7120 NEBNext® Enzymatic Methyl-seq Kit E7120).

[0426] Both the ssDNA library (generated by the method disclosed herein) and the dsDNA library underwent the targeted methylation switching sequencing method as described in Examples 2 and 3. Raw data files were used for alignment and methylation calling to enable targeted methylation analysis for pre-identified regions of the genome for both hypermethylation model scoring and hypomethylation model scoring.

[0427] B. Results

[0428] Fig. 5This provides a comparison of hypermethylation model scores for cfDNA extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) samples detected using the ssDNA library preparation described in Example 1 against another dsDNA library preparation method. The hypermethylation rates of cfDNA derived from selected genomic regions distinguished cfDNA from AA and CRC patients from cancer-negative cfDNA when processed using both the ssDNA library preparation and another dsDNA library preparation. This experiment demonstrates that the ssDNA library preparation method described herein can yield methylation analyses (e.g., hypermethylation scores) equivalent to other dsDNA library preparation methods. Thus, this workflow is equivalent to end-repair-dependent methods in distinguishing cfDNA samples from cancer patients from cfDNA samples from healthy patients, for example, when performing DNA hypermethylation analysis.

[0429] Fig. 7This provides a comparison of the hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), advanced adenoma (AA), and colorectal cancer (CRC) samples detected using the ssDNA library preparation described in Example 1 against another dsDNA library preparation method. The hypomethylation rates of cfDNA derived from selected genomic regions distinguished cfDNA derived from AA and CRC patients from cancer-negative cfDNA when processed using the ssDNA library preparation described herein, but failed to distinguish them when processed using the dsDNA library preparation. Thus, this example demonstrates that the ssDNA library preparation method described herein optimizes recovery, methylation analysis, quantification, and targeted capture. Accordingly, in certain embodiments, this workflow may demonstrate superior performance compared to end-repair-dependent methods in distinguishing cfDNA samples derived from cancer patients from cfDNA samples derived from healthy patients, for example, using DNA hypomethylation analysis.

[0430] C. Liver cancer, lung cancer, and pancreatic cancer

[0431] Blood samples were obtained from 11 healthy human subjects, 10 human subjects with liver cancer, 27 human subjects with lung cancer, and 21 human subjects with pancreatic cancer. DNA was extracted from the blood samples, and libraries were prepared according to (i) the ssDNA library preparation method described in Example 1 and (ii) other dsDNA library preparation methods (E7120 NEBNext® Enzymatic Methyl-seq Kit E7120).

[0432] Both the ssDNA library (generated by the method disclosed herein) and the dsDNA library underwent the targeted methylation switching sequencing method as described in Examples 2 and 3. Raw data files were used for alignment and methylation calling to enable targeted methylation analysis for pre-identified regions of the genome for both hypermethylation model scoring and hypomethylation model scoring.

[0433] D. Results

[0434] Fig. 6 This provides a comparison of hypermethylation model scores for cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) samples detected using the ssDNA library preparation method described herein against other dsDNA library preparation methods. The hypermethylation rate of cfDNA derived from selected genomic regions distinguished cfDNA derived from liver cancer, lung cancer, and pancreatic cancer patients from cancer-negative cfDNA when processed using both the ssDNA library preparation method described herein and other dsDNA library preparation methods. This example demonstrates that the ssDNA library preparation method described herein yields methylation analysis (e.g., hypermethylation scores) equivalent to other dsDNA library preparation methods. Thus, the ssDNA library preparation workflow described herein is equivalent to end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, when using DNA hypermethylation analysis.

[0435] Fig. 8 The present invention provides a comparison of the hypomethylation scores of cfDNA extracted from healthy / cancer-negative (NEG), liver cancer (liver), lung cancer (lung), and pancreatic cancer (pancreas) patient samples detected using the ssDNA library preparation method as described herein versus the dsDNA library preparation method. Fig. 8 As illustrated in, The hypomethylation rate of cfDNA derived from selected genomic regions distinguished cfDNA derived from liver, lung, and pancreatic cancer patients from cancer-negative cfDNA when processed using ssDNA library preparation, but failed to distinguish them when processed using dsDNA library preparation. This example provides additional evidence that the ssDNA library preparation method described herein optimizes recovery, methylation analysis, quantification, and targeted capture across multiple cancer types, and demonstrates that this workflow can exhibit superior performance compared to end-repair-dependent methods in distinguishing cancer patient-derived cfDNA samples from healthy patient-derived cfDNA samples, for example, when using DNA hypomethylation analysis.

[0436] Although preferred embodiments of the concept of the present invention have been presented and described herein, it will be obvious to those skilled in the art that such embodiments are provided merely as examples. The concept of the present invention is not intended to be limited by the specific examples provided herein.

[0437] Although the concept of the present invention has been described with reference to the foregoing specification, the description and illustration of embodiments thereof are not intended to be interpreted in a limiting sense. Now, to those skilled in the art, numerous modifications, changes, and substitutions will occur without departing from the concept of the present invention. Furthermore, it should be understood that all aspects of the concept of the present invention are not limited to the specific descriptions, configurations, or relative proportions presented herein, which vary depending on various conditions and variables.

[0438] It should be understood that various alternatives to the embodiments of the concept of the present invention described herein may be used in practicing the concept of the present invention. Accordingly, the concept of the present invention is intended to include any such alternatives, modifications, variations, or equivalents. The following claims define the scope of the concept of the present invention, and methods, structures, and equivalents within the scope of these claims are intended to be encompassed thereby.

Claims

Claim 1 A method for preparing a sequencing library for methylation sequencing of one or more nucleic acid molecules of a biological sample or a derivative thereof, comprising: (a) a step of obtaining a nucleic acid composition, wherein the nucleic acid composition comprises a plurality of single-stranded nucleic acid molecules obtained from or derived therefrom a biological sample; (b) a step of ligating a nucleic acid adapter to a single-stranded nucleic acid molecule among the plurality of single-stranded nucleic acid molecules to produce an adapter-ligated nucleic acid molecule, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method; and (c) A manufacturing method comprising the step of producing a converted adapter-ligated nucleic acid molecule by exposing the adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation concentration method. Claim 2 A method according to claim 1, wherein the ligation in (b) further comprises treatment with a deoxyribonucleic acid (DNA) ligase. Claim 3 A method according to claim 1, wherein the ligation in (b) further comprises treatment with a polynucleotide kinase. Claim 4 A method according to claim 1, wherein the nucleic acid adapter comprises a double-stranded oligonucleotide comprising an adapter sequence and an overhang sequence. Claim 5 In paragraph 4, the method in which the overhang sequence is a 3' overhang sequence. Claim 6 A method according to claim 4, wherein the overhang sequence comprises a random sequence of oligonucleotides. Claim 7 A method according to claim 1, wherein the nucleic acid adapter comprises one or more methylated cytosine bases. Claim 8 A method according to claim 1, wherein the nucleic acid adapter does not contain a methylated cytosine base or a non-methylated cytosine base. Claim 9 A method according to claim 1, wherein the nucleic acid adapter comprises a unique molecular identifier. Claim 10 A method according to claim 9, wherein a unique molecular identifier is configured to enable the measurement of the concentration efficiency of a methylation conversion method. Claim 11 A method according to claim 1, further comprising the step of (a) previously processing a biological sample or a derivative thereof to produce a plurality of single-stranded nucleic acid molecules. Claim 12 A method according to claim 11, wherein the treatment comprises denaturing double-stranded nucleic acid molecules within a biological sample or a derivative thereof. Claim 13 A method in paragraph 12, wherein the denaturation further comprises applying heat to a double-stranded nucleic acid molecule. Claim 14 A method according to claim 12, further comprising applying heat to a double-stranded nucleic acid molecule to denature it, and then rapidly cooling the denatured single-stranded nucleic acid molecule. Claim 15 A method according to claim 1, further comprising the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the possibility of forming a nucleic acid duplex. Claim 16 A method according to claim 1, wherein the method does not include the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the possibility of forming a nucleic acid duplex. Claim 17 A method according to claim 15 or 16, wherein the binder is a single-stranded nucleic acid binding protein (SSB). Claim 18 A method according to claim 1, further comprising the step of amplifying a converted adapter-ligated nucleic acid molecule. Claim 19 A method according to paragraph 18 in which the amplification involves a polymerase chain reaction (PCR). Claim 20 The method of claim 1 further comprises the step of contacting a converted adapter-ligated nucleic acid molecule or a derivative thereof with a nucleic acid probe to produce a concentrated nucleic acid molecule, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel. Claim 21 A method according to claim 20, wherein the nucleic acid probe comprises a non-methylated nucleic acid probe. Claim 22 A method according to claim 20, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. Claim 23 A method according to claim 20, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to cytosine bases methylated at CpG loci from CpG or CH loci of a reference panel. Claim 24 A method according to any one of claims 20 to 23, further comprising the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof. Claim 25 A method according to any one of claims 20 to 24, further comprising the step of sequencing an enriched nucleic acid molecule or a derivative thereof to generate sequencing data. Claim 26 A method according to claim 25, further comprising the step of analyzing sequencing data to generate a methylation profile of nucleic acid molecules of a biological sample or its derivative. Claim 27 A method according to paragraph 26 in which the analysis further includes comparing sequencing data with a reference sequence. Claim 28 A method according to claim 1, further comprising the step of (b) subsequently performing an extension reaction on an adapter-ligated nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. Claim 29 A method according to claim 28, wherein the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter. Claim 30 A method in which the nucleic acid molecule in claim 1 is deoxyribonucleic acid (DNA). Claim 31 In paragraph 30, a method in which the DNA is cell-free DNA. Claim 32 In claim 1, the biological sample is a cell-free biological sample. Claim 33 In paragraph 32, a method in which the cell-free biological sample is a plasma sample. Claim 34 A method according to claim 1, wherein the methylation concentration method comprises treatment by one or more enzymes. Claim 35 A method according to claim 1, wherein the methylation concentration method comprises treatment with a ten eleven translocation (TET) enzyme. Claim 36 A method according to claim 1, wherein the methylation concentration method does not include treatment with non-sulfite. Claim 37 A method for preparing a sequencing library for methylation sequencing of nucleic acid molecules of a biological sample or a derivative thereof, comprising: (a) a step of obtaining a nucleic acid composition, wherein the nucleic acid composition comprises a plurality of single-stranded nucleic acid molecules obtained from or derived therefrom a biological sample; (b) a step of exposing a single-stranded nucleic acid molecule among the plurality of single-stranded nucleic acid molecules to conditions sufficient to convert unmethylated cytosine to uracil using a methylation enrichment method to produce a converted single-stranded nucleic acid molecule; and (c) a step of ligating a nucleic acid adapter to the converted single-stranded nucleic acid molecule to produce an adapter-ligated converted nucleic acid molecule, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method. Claim 38 A method according to paragraph 37, wherein the ligation in (c) further comprises treatment with a deoxyribonucleic acid (DNA) ligase. Claim 39 A method according to paragraph 37, wherein the ligation in (c) further comprises treatment with a polynucleotide kinase. Claim 40 A method according to claim 37, wherein the nucleic acid adapter comprises a double-stranded oligonucleotide comprising an adapter sequence and an overhang sequence. Claim 41 In paragraph 40, the method in which the overhang sequence is a 3' overhang sequence. Claim 42 A method according to claim 40 in which the overhang sequence comprises a random sequence of oligonucleotides. Claim 43 A method according to claim 37 in which the nucleic acid adapter comprises one or more methylated cytosine bases. Claim 44 A method according to claim 37 in which the nucleic acid adapter does not contain a methylated cytosine base or a nonmethylated cytosine base. Claim 45 In paragraph 37, a method in which the nucleic acid adapter comprises a unique molecular identifier. Claim 46 A method according to claim 45, wherein a unique molecular identifier is configured to enable the measurement of the concentration efficiency of a methylation concentration method. Claim 47 In claim 37, a method further comprising the step of (a) previously processing a biological sample or a derivative thereof to produce a plurality of single-stranded nucleic acid molecules. Claim 48 A method according to paragraph 47, wherein the treatment comprises denaturing double-stranded nucleic acid molecules within a biological sample or a derivative thereof. Claim 49 A method in paragraph 47 in which the denaturation involves applying heat to a double-stranded nucleic acid molecule. Claim 50 A method according to claim 47, comprising applying heat to a double-stranded nucleic acid molecule to produce a plurality of single-stranded nucleic acid molecules, and then performing rapid cooling on the plurality of single-stranded nucleic acid molecules. Claim 51 A method according to claim 37, further comprising the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the possibility of forming a nucleic acid duplex. Claim 52 A method according to claim 37, wherein the method does not include the step of treating a single-stranded nucleic acid molecule or a derivative thereof with a binder configured to reduce the possibility of forming a nucleic acid duplex. Claim 53 A method according to claim 51 or 52 in which the binder is a single-stranded nucleic acid binding protein (SSB). Claim 54 A method according to claim 37, further comprising the step of amplifying an adapter-ligated converted nucleic acid molecule. Claim 55 In paragraph 54, a method in which the amplification includes a polymerase chain reaction (PCR). Claim 56 The method of claim 37 further comprises the step of contacting an adapter-ligated converted nucleic acid molecule or a derivative thereof with a nucleic acid probe to produce a concentrated nucleic acid molecule, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel. Claim 57 A method according to claim 56 in which the nucleic acid probe comprises a non-methylated nucleic acid probe. Claim 58 A method according to claim 56, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. Claim 59 A method according to claim 56, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to cytosine bases methylated at CpG loci from CpG or CH loci of a reference panel. Claim 60 A method according to any one of claims 56 to 59, further comprising the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof. Claim 61 A method according to any one of claims 56 to 59, further comprising the step of sequencing an enriched nucleic acid molecule or a derivative thereof to generate sequencing data. Claim 62 A method according to claim 61, further comprising the step of analyzing sequencing data to generate a methylation profile of nucleic acid molecules of a biological sample or a derivative thereof. Claim 63 A method according to paragraph 62 in which the analysis further includes comparing sequencing data with a reference sequence. Claim 64 In claim 37, a method further comprising the step of (b) subsequently performing an extension reaction on an adapter-ligated converted nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. Claim 65 A method according to claim 64, wherein the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter. Claim 66 In paragraph 37, the method in which the nucleic acid molecule is deoxyribonucleic acid (DNA). Claim 67 In paragraph 66, the method in which the DNA is cell-free DNA. Claim 68 In paragraph 37, a method in which the biological sample is a cell-free biological sample. Claim 69 In paragraph 68, a method in which the cell-free biological sample is a plasma sample. Claim 70 In paragraph 37, the methylation concentration method comprises treatment by one or more enzymes. Claim 71 In paragraph 37, the methylation concentration method comprises treatment with a ten-eleven translocation (TET) enzyme. Claim 72 In paragraph 37, the methylation concentration method does not involve treatment with non-sulfite. Claim 73 (a) a step of producing an adapter-ligated nucleic acid molecule by ligating a nucleic acid adapter to a single-stranded nucleic acid molecule obtained from or derived from a biological sample of a subject, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method; (b) a step of generating a converted adapter-ligated nucleic acid molecule by exposing the adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation enrichment method; (c) a step of generating an amplified nucleic acid molecule by amplifying the converted adapter-ligated nucleic acid molecule; (d) a step of generating an enriched nucleic acid molecule by contacting the amplified nucleic acid molecule or a derivative thereof with a nucleic acid probe, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel; (e) a step of determining the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof; (f) a step of comparing the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof with a reference nucleic acid sequence; and (g) a step of training a machine learning model to generate a classifier that distinguishes between subjects having cancer and subjects not having cancer, wherein the machine learning model is trained with methylation profiles generated from (i) a first set of nucleic acid samples from subjects having cancer and (ii) a second set of nucleic acid samples from subjects not having cancer. Claim 74 A method according to paragraph 73 in which the amplification involves a polymerase chain reaction (PCR). Claim 75 A method according to claim 73, further comprising the step of determining the nucleic acid sequence of an enriched nucleic acid molecule or a derivative thereof to a depth greater than 100x. Claim 76 In paragraph 73, a method wherein the nucleic acid probe comprises unmethylated nucleic acid. Claim 77 A method according to claim 73, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to unmethylated cytosine bases at CpG loci from CpG or CH loci of a reference panel. Claim 78 A method according to claim 73, wherein the nucleic acid probe is configured to selectively hybridize to one or more target regions of interest corresponding to cytosine bases methylated at CpG loci from CpG or CH loci of a reference panel. Claim 79 A method according to any one of claims 73 to 78, further comprising the step of sequencing an enriched nucleic acid molecule or a derivative thereof to generate sequencing data. Claim 80 A method according to claim 79, further comprising the step of analyzing sequencing data to generate a methylation profile of nucleic acid molecules of a biological sample or a derivative thereof. Claim 81 In paragraph 73, a method further comprising the step of (a) subsequently performing an extension reaction on an adapter-ligated converted nucleic acid molecule to produce a partially double-stranded nucleic acid molecule or a complete double-stranded nucleic acid molecule. Claim 82 A method according to claim 81, wherein the extension reaction is performed in the presence of a polymerase, a plurality of deoxynucleotide triphosphates (dNTPs), and a primer complementary to the 3' end of the nucleic acid adapter. Claim 83 In paragraph 73, a method in which the single-stranded nucleic acid molecule is deoxyribonucleic acid (DNA). Claim 84 In paragraph 83, the method in which the DNA is cell-free DNA. Claim 85 In paragraph 73, a method in which the biological sample is a cell-free biological sample. Claim 86 In paragraph 85, a method in which the cell-free biological sample is a plasma sample. Claim 87 In paragraph 73, the methylation concentration method comprises treatment by one or more enzymes. Claim 88 In paragraph 73, the methylation concentration method comprises treatment with a ten-eleven translocation (TET) enzyme. Claim 89 In paragraph 73, the methylation concentration method does not involve treatment with non-sulfite. Claim 90 In paragraph 73, a method in which the reference panel includes a CpG or CH locus associated with the transcription start site. Claim 91 A method according to paragraph 73, further comprising the step of identifying the tissue origin of nucleic acid molecules. Claim 92 A method according to claim 73, further comprising the step of identifying the genomic location and fragment length of a nucleic acid molecule. Claim 93 A method according to claim 73, wherein a machine learning model is trained using feature inputs selected from the group consisting of base-wise methylation % for CpG, base-wise methylation % for CHG, base-wise methylation % for CHH, the number or ratio of fragments with different numbers or ratios of methylated CpGs observed within a region, conversion efficiency, hypomethylation blocks, methylation levels for CpG, methylation levels for CHH, methylation levels for CHG, fragment length, fragment midpoint, methylation levels for chrM, methylation levels for LINE1, methylation levels for ALU, dinucleotide coverage, uniformity of coverage, global average CpG coverage, CpG islands, CGI shelves, and average coverage on CGI shores. Claim 94 A method according to claim 73, wherein a classifier for distinguishing between subjects with cancer and subjects without cancer comprises a set of measurements representing methylation profiles from methylation sequencing data of subjects with cancer and subjects without cancer, wherein the measurements are used to generate a feature set corresponding to characteristics of the methylation profiles, the feature set is processed by a machine learning or statistical model, and the machine learning or statistical model provides a feature vector useful as a classifier for distinguishing between a group of subjects with cancer and a group of subjects without cancer. Claim 95 A method according to claim 73, further comprising: (a) obtaining a methylation profile of a biological sample by testing by a methylation enrichment method; (b) classifying the methylation profile of the biological sample as indicating the presence of cancer in a subject by a trained machine learning algorithm; and (c) outputting a report identifying the biological sample as negative for cancer when the trained machine learning algorithm classifies the biological sample as negative for cancer at a specified confidence level. Claim 96 A method according to claim 73, further comprising: (a) determining a baseline methylation profile of a biological sample of a subject in a baseline methylation state; (b) determining a test methylation profile of a biological sample of a subject at one or more time points after the baseline methylation state; and (c) determining a change in the test methylation profile in comparison to the baseline methylation profile, wherein the change indicates a change in the subject's minimum residual disease state. Claim 97 A method according to claim 96 in which the minimum residual disease status is selected from a group consisting of response to treatment, tumor load, residual tumor after surgery, recurrence, secondary screening, primary screening, and cancer progression. Claim 98 A method according to claim 95 or 96, wherein the methylation profile comprises a hypermethylation analysis and / or a hypomethylation analysis. Claim 99 A method according to paragraph 73 in which the cancer includes two or more of colorectal cancer, breast cancer, pancreatic cancer, liver cancer, or lung cancer. Claim 100 In paragraph 73, the method in which the cancer is colorectal cancer, breast cancer, pancreatic cancer, liver cancer or lung cancer. Claim 101 In paragraph 100, the method in which the cancer is colorectal cancer. Claim 102 In Clause 100, the method in which the cancer is lung cancer. Claim 103 In Clause 100, the method in which the cancer is liver cancer. Claim 104 In Clause 100, the method in which the cancer is pancreatic cancer. Claim 105 (a) a step of producing an adapter-ligated nucleic acid molecule by ligating a nucleic acid adapter to a single-stranded nucleic acid molecule obtained from or derived from a biological sample of a subject, wherein the nucleic acid adapter comprises a nucleic acid resistant to base conversion by a methylation enrichment method; (b) a step of generating a converted adapter-ligated nucleic acid molecule by exposing the adapter-ligated nucleic acid molecule to conditions sufficient to convert unmethylated cytosine to uracil using a methylation enrichment method; (c) a step of generating an amplified nucleic acid molecule by amplifying the converted adapter-ligated nucleic acid molecule; (d) a step of generating an enriched nucleic acid molecule by contacting the amplified nucleic acid molecule or a derivative thereof with a nucleic acid probe, wherein the nucleic acid probe comprises a nucleic acid sequence that is at least partially complementary to a CpG or CH locus of a reference panel; (e) a step of determining the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof; (f) a step of comparing the nucleic acid sequence of the enriched nucleic acid molecule or a derivative thereof with a reference nucleic acid sequence; and (g) a step of training a machine learning model to generate a classifier that distinguishes between subjects with indications and subjects without indications, wherein the machine learning model is trained with methylation profiles generated from (i) a first set of nucleic acid samples from subjects with indications and (ii) a second set of nucleic acid samples from subjects without indications. Claim 106 A method according to claim 105, further comprising: (h) obtaining a methylation profile of a biological sample by testing by a methylation enrichment method; (i) classifying the methylation profile of the biological sample as indicating the presence of an indication in a subject by a trained machine learning algorithm; and (j) outputting a report identifying the biological sample as negative for an indication when the trained machine learning algorithm classifies the biological sample as negative for an indication at a specified confidence level. Claim 107 A method according to claim 105, further comprising: (h) determining a baseline methylation profile of a biological sample of a subject in a baseline methylation state; (i) determining a test methylation profile of a biological sample of a subject at one or more time points after the baseline methylation state; and (j) determining a change in the test methylation profile in comparison to the baseline methylation profile, wherein the change indicates a change in the minimum residual disease state of the indication in the subject. Claim 108 A method according to claim 107 in which the minimum residual disease status is selected from a group consisting of response to treatment, relapse, secondary screening, primary screening, and progression of indication. Claim 109 A method according to claim 106 or 107, wherein the methylation profile comprises a hypermethylation analysis and / or a hypomethylation analysis. Claim 110 A method according to paragraph 105, wherein the indications include intestinal-related diseases, immune-mediated inflammatory diseases, neurological diseases, kidney diseases, prenatal diseases or metabolic diseases. Claim 111 A system comprising: (a) a computer-readable media product comprising a classifier of claim 73, wherein the classifier comprises a set of measurements representing methylation profiles from methylation sequencing data from objects having cancer and objects not having cancer, wherein the set of measurements is used to generate a set of features corresponding to characteristics of methylation profiles from objects having cancer and objects not having cancer, wherein the set of features is processed by a machine learning or statistical model, and wherein the machine learning or statistical model provides a feature vector useful for distinguishing between objects having cancer and objects not having cancer; and (b) one or more processors for executing instructions stored in the computer-readable media product. Claim 112 A system according to claim 111, wherein the classifier is selected from the group consisting of a Linear Discriminant Analysis (LDA) classifier, a Quantitative Discriminant Analysis (QDA) classifier, a Support Vector Machine (SVM) classifier, a Random Forest (RF) classifier, a Linear Kernel Support Vector Machine classifier, a First-order Polynomial Kernel Support Vector Machine classifier, a Second-order Polynomial Kernel Support Vector Machine classifier, a Ridge Regression classifier, an Elastic Net algorithm classifier, a Sequential Minimization algorithm classifier, a Naive Bayes algorithm classifier, and a Non-Negative Matrix Decomposition (NMF) prediction algorithm classifier. Claim 113 A system further configured to perform any of the above methods in Clause 111. Claim 114 A system according to claim 111, wherein the system comprises one or more processors configured to perform any of the above methods. Claim 115 A system according to claim 111, wherein the system comprises a module that performs the operation of any of the above methods. Claim 116 A kit for detecting cancer, comprising a reagent for performing any of the above methods and instructions for detecting a cancer signal. Claim 117 A kit according to claim 116, wherein the reagent is selected from the group consisting of primer sets, PCR reaction components, sequencing reagents, methylation enrichment reagents, and library preparation reagents.