Molecular counting of methylated cell-free DNA for treatment monitoring

Abstract

Embodiments of the present disclosure include a method of quantifying methylation in a DNA sample, the method including treating the sample to encode the presence or absence of DNA methylation, adding a set of synthetic molecules (e.g., quality control template (QCT) molecules) to the sample, creating a co-amplification mixture, sequencing the co-amplification mixture, and determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules.

Description

[Technical field]

[0001] 1. CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of and priority to U.S. Provisional Patent Application No. 63 / 322,014, filed March 21, 2022, and U.S. Provisional Patent Application No. 63 / 439,492, filed January 17, 2023, each of which is incorporated by reference in its entirety herein.

[0002] 2. Sequence Listing This application contains a Sequence Listing having two sequences submitted via the Patent Center, which is incorporated herein by reference in its entirety. The XML copy, created on Mar. 18, 2023, is named 38227-55032-SequenceListing-WO.xml and is 2,905 bytes in size. [Background technology]

[0003] 3.Background Although new cancer treatments continue to be developed at unprecedented speed, assessing whether a cancer treatment is effective for an individual patient remains relatively cumbersome and qualitative. Imaging is the gold standard for measuring cancer status, but PET / CT and CT scans expose patients to harmful radiation doses, and these imaging facilities may be far from where the patient lives. Imaging sessions are scheduled months apart, which limits how quickly oncologists can recognize how well the cancer is responding to treatment. Qualitative metrics such as physical examinations and patient-reported symptoms are often confounded by treatment side effects and therefore not accurate enough to evaluate treatment effectiveness. Nevertheless, oncologists want to assess the degree to which a patient's cancer is responding to treatment to inform treatment plans, such as whether the patient should continue their existing treatment or switch to a new treatment plan.

[0004] Non-invasive liquid biopsies that assay cell-free DNA (cfDNA) have been developed to quantify levels of circulating tumor DNA (ctDNA) from blood samples. Multiple studies have found that levels of ctDNA are useful and quantifiable information for longitudinal ctDNA measurements that accurately track tumor progression. Many assays specifically quantify the presence of somatic mutations (e.g., single nucleotide variants, copy number changes, insertions, deletions) using variant allele frequency (VAF) to track these VAFs over time. However, assays that rely on quantifying somatic mutations have limitations. Because VAFs can be extremely small (approximately 0.1-0.5%), there can be significant and unavoidable molecular sampling limitations. For variants present at low VAF, false negatives can occur, and a sample with an average of 2 variant molecules may actually have 0 variant molecules in 13.5% of opportunities simply due to Poisson sampling. Even if a variant is detected, the measurement is noisy due to the small number of variant molecules. For example, 9 variant molecules in a sample would have a standard deviation of 3 variant molecules due to sampling noise, resulting in a coefficient of variation (CV) of 33%. A small number of molecules can increase the noise of a VAF-based time series measurement, making it difficult to interpret. Moreover, understanding the degree of molecular sampling limitation requires an estimate of the number of variant molecules, information that many assays do not provide.

[0005] Therefore, absolute quantification to determine the number of variant molecules present is required. Droplet digital PCR has been shown to be a sensitive approach to detect ctDNA. However, due to the limited amount of initial sample, at most a few genomic locations can be probed simultaneously.

[0006] Given the molecular sampling challenges associated with assaying somatic mutations, methylation was explored as an alternative biomarker for ctDNA. Methylation has long been shown to be a strong, consistent, and genome-wide biomarker for cancer. However, while methylation signals are significantly more abundant compared to signals from somatic mutations, accurately and precisely quantifying the amount of methylation is a challenge. Methylation-specific qPCR is a commonly used method, but due to the exponential nature of the assay, qPCR Ct measurements have high CV (typically within a few tens of percent). Given the limited amount of sample, improving accuracy with multiple replicates is a challenge. In addition, non-tumor cfDNA may contribute to background methylation signals, complicating the task of quantifying signals belonging to ctDNA. Furthermore, although sampling multiple locations for methylation would help improve the performance of the assay, interrogating multiple loci using qPCR is difficult and time-consuming. Summary of the Invention

[0007] 4. Overview Provided herein is a methylation-based approach for more accurate and precise evaluation of treatment monitoring, for example for cancer patients.The methylation-based approach is a pan-cancer assay that provides longitudinal determination of cancer burden through methylation evaluation.This evaluation is based in part on nucleotide (DNA) methylation at specific positions in cancer, and can identify multiple cancer methylation signatures.

[0008] Accurate counting of methylated DNA molecules in blood samples can inform the cancer status of patients. Because some methylated molecules are rare (i.e., have low concentrations in the sample), they must be amplified in order to be detected. However, amplification can further exacerbate the challenge of detecting methylated molecules, as detection of methylated molecules can be noisy and methylated molecules amplify at different rates. Here, applicants have developed a methylation-based approach that first determines the number of methylated molecules present in a sample by adding a quantitative counting template (QCT) before amplification, providing a measurement tool that allows quantification after the amplification step.

[0009] Methylation-based approaches are distinct from other assays that rely on single nucleotide variant (SNV)-based ctDNA monitoring, because methylated ctDNA is a global and additive marker that allows for a more robust and cumulative measurement of ctDNA. While typical SNV-based ctDNA monitoring assays, such as tumor-informed MRD assays, look at an average of 9 SNVs, the methods described herein can typically quantify an average of 90 methylation loci, resulting in a 10-fold increase in signal.

[0010] The methylation-based approach described is a next-generation sequencing (NGS)-based test designed to measure changes in methylated tumor molecules in cancer patients from blood samples. In particular, the method quantifies methylated ctDNA (circulating tumor DNA) molecules isolated from cell-free DNA (cfDNA) at loci known to be hypermethylated in tumors compared to healthy tissues.

[0011] In one embodiment, plasma and buffy coat are isolated from whole blood collected from a patient. Cell-free DNA (cfDNA) is extracted from the plasma and DNA (e.g., genomic DNA) is extracted from the buffy coat. The number of methylated molecules is quantified in both cfDNA and DNA (from the buffy coat) using QCT (e.g., as described in U.S. Patent Application Publication Nos. 2020 / 0040380A1, 2019 / 0095577A1, 2019 / 0114389A1, and 2019 / 0211395A1, which are incorporated by reference) at over 500 positions in the genome known to be hypermethylated in cancer compared to non-cancerous tissues and blood. To remove background from the ctDNA signal, the methylation measured in DNA (e.g., methylated DNA in the buffy coat) is subtracted from the cfDNA methylation (e.g., methylated molecules from plasma). To quantify DNA methylation in a sample (e.g., to calculate a Tumor Methylation Score), the remaining cfDNA methylation molecules are summed across all hypermethylated positions.

[0012] The quantified DNA methylation in the sample from this time point collection (e.g., Tumor Methylation Score in the sample) can be compared to the most recently reported quantified DNA methylation (e.g., Tumor Methylation Score) to determine a determination of increase, decrease, or no change. If the change in quantified DNA methylation reaches a significant threshold, an increase or decrease can be reported. In baseline testing without any prior collections, no interpretive call (e.g., increase or decrease) of the change in quantified DNA methylation is made. The limit of detection (LOD) allows for detection of a 0.2 percentage point change in tumor rate and 3 standard deviations of separation.

[0013] The present disclosure features a method for quantifying the number of methylated molecules present in a sample using a quantitative counting template (QCT). For example, a sample containing DNA sequences (e.g., a mixture of methylated and unmethylated DNA sequences) is treated to convert unmethylated cytosines in the DNA sequences to uracil (i.e., to code the presence or absence of DNA methylation in the DNA sequences). In some cases, the sample will have rare DNA sequences containing methylated cytosines, which require amplification to detect and quantify these DNA sequences. As previously mentioned, amplification impairs the ability to quantify methylated molecules. The addition and co-amplification of QCT alleviates these problems. Thus, QCT is added to the sample and co-amplifies with the treated DNA sequences to generate a co-amplified mixture. The co-amplified mixture is then sequenced. Sequencing and subsequent analysis determine the number of QCT molecules, the number of methylated cytosines, and the number of unmethylated cytosines. Alternatively, sequencing and subsequent analysis determines the number of QCT molecules and the number of methylated sequencing reads (i.e., sequencing reads that contain uracil). The number of methylated molecules in a sample is quantified based on the number of methylated reads and the number of reads from QCT molecules. The quantified methylated molecules can then be used to facilitate diagnosis, treatment, or further evaluation of a subject.

[0014] This methylation-based approach can also be used to determine whether different assays performed on the same sample are reliable. For example, if a sample is subjected to two assays: (1) an assay to determine the prevalence of somatic mutations (e.g., variant allele frequency (VAF) measurement), and (2) an assay to quantify DNA methylation using QCT molecules as described herein, the latter can be used to inform the reliability of the former. In particular, the quantification of the number of methylated molecules (based on QCT molecules) as an indicator of the presence of cancer in a sample can be used to determine whether the variant allele frequency measurement from the sample (another indicator of the presence of cancer) is reliable. This is possible because the QCT molecules in the methylation assay provide an accurate and reliable way to determine the number of methylated molecules in a sample and avoid false negatives or false positives. This essentially means that if the VAF measurement and methylation analysis are performed on material from the same sample, the accuracy and reliability of the methylation measurement can be imparted to the VAF measurement to give the VAF a confidence score (e.g., a "true call" (i.e., if the VAF should be trusted) or a "no call" (i.e., if the VAF should not be trusted). For example, if the VAF is small (about 0.1-0.5%), this may be reported as "negative" for the presence of cancer. If the corresponding quantification of methylated DNA (using the methods described herein) indicates the presence of cancer (e.g., if the number of methylated DNA sequences exceeds a predefined threshold), then the VAF is a "false negative" or "no call" and the VAF measurement should not be trusted. Importantly, these results may suggest the need for additional evaluation of the subject, such as new treatment selection assays, genomic profiling, and / or changes in treatment regimen.

[0015] This methylation-based approach can also be used to determine changes in a methylation profile in a subject over time. A change in a methylation profile in a subject may indicate the presence of cancer or a change in an existing cancer. The appearance of new methylated loci at the next time point, an increase in the contribution of a methylated locus to the total methylation signal at the next time point, or a combination thereof, is indicative of a change in the methylation profile in the subject. In particular, the methylation-based approach described herein allows for precise determination of the methylation profile by the use of QCT molecules and their ability to reliably and accurately quantify the number of methylated molecules at each time point. This is a functionality that allows comparison between time points and allows the determination of the methylation profile in a subject over time. Other ctDNA assays lack the accuracy and reliability between measurements performed at different time points, and therefore these assays cannot measure the methylation profile in a subject over time. Therefore, the ability to measure changes in the methylation profile that subsequently inform the oncologist's strategy regarding diagnosis, treatment, and further evaluation of the subject is unique to the methylation-based approach described herein.

[0016] In summary, the present disclosure features a method for quantifying the number of methylated molecules present in a sample using quantitative counting templates (QCT) in a multiplexed manner. Background subtraction using methylation signals from the corresponding buffy coat of the sample not only reduces noise in the assay but also improves robustness against genomic DNA (gDNA) contamination. Background levels of methylation in healthy subjects vary from day to day, suggesting the importance of background subtraction for these time series measurements. This approach can be adapted to measure methylation signals in multiple cancer types utilizing a single assay chemistry.

[0017] In one aspect, the disclosure features a method of quantifying DNA methylation in a sample containing a DNA sequence, the method including: treating the sample to encode the presence or absence of DNA methylation in the DNA sequence, the sample including at least 10 target loci from the DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules including a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule that includes at least one of the target loci, and a mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule; creating a co-amplification mixture including the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; sequencing the co-amplification mixture to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; and quantifying the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules.

[0018] In some embodiments, the sample is taken from a blood draw that includes plasma and buffy coat, where the plasma includes cfDNA and the buffy coat includes genomic DNA (gDNA) sequences.

[0019] In some embodiments, the method further includes extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample prior to treating the sample to encode the presence or absence of DNA methylation.

[0020] In some embodiments, the method further comprises quantifying DNA methylation in a sample containing gDNA sequences from the buffy coat.

[0021] In some embodiments, the method includes treating a sample containing DNA sequences from a buffy coat to encode the presence or absence of DNA methylation within the DNA sequences, the sample comprising at least 10 target loci from the DNA sequences; adding a set of synthetic molecules to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule that comprises at least one of the target loci, and a mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule; creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequences; sequencing the co-amplification mixture to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; and determining a number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules, thereby quantifying DNA methylation within the gDNA sequences from the buffy coat.

[0022] In some embodiments, the method further comprises adding a spike-in of known sequence and amount to the sample prior to the treating step. In some embodiments, wherein treating the sample to encode the presence or absence of DNA methylation comprises bisulfite conversion or enzymatic conversion. In some embodiments, wherein the set of synthetic molecules is a set of quantitative counting templates (QCT). In some embodiments, each locus of the at least 10 target loci is selected based on a predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue.

[0023] In some embodiments, at least 100 target loci are amplified, or at least 500 target loci are amplified.

[0024] In some embodiments, the co-amplified mixture is sequenced at a read depth of at least 1 read per molecule.

[0025] In some embodiments, the co-amplification mixture is sequenced at a read depth of 10 reads per molecule or more, 100 reads per molecule or more, or 1000 reads per molecule or more.

[0026] In some embodiments, the method further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the sample.

[0027] In some embodiments, the counting comprises counting target loci from at least 10 target loci that contain less than a threshold amount of methylated molecules in the buffy coat sample.

[0028] In some embodiments, the sample further comprises at least one normalization locus predicted to have hypermethylation in cfDNA in both cancerous and non-cancerous tissue.

[0029] In some embodiments, the co-amplification mixture comprises an amplified set of synthetic molecules, an amplified set of all or a portion of at least 10 target loci, and at least one amplified normalization locus.

[0030] In some embodiments, the method further comprises normalizing the tabulated number of methylated molecules from at least two of the at least ten target loci by a methylated molecule for at least one normalization locus.

[0031] In some embodiments, the method further comprises subtracting background methylation.

[0032] In some embodiments, the method further comprises removing selected hypermethylated target loci.

[0033] In some embodiments, the method further comprises subtracting background methylation and removing selected hypermethylated loci.

[0034] In some embodiments, subtracting background methylation comprises subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA.

[0035] In some embodiments, subtracting background methylation comprises subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA on a locus basis.

[0036] In some embodiments, removing selected hypermethylated target loci includes removing target loci that have a hypermethylated cfDNA signal in non-cancerous tissue.

[0037] In some embodiments, removing selected hypermethylated target loci comprises removing target loci having a total number of methylated molecules in the buffy coat that exceeds a threshold. In some embodiments, the threshold is sample specific. In some embodiments, the threshold is subject specific. In some embodiments, the threshold comprises a predetermined mean tumor methylation quantitative equivalent (QE), a predetermined maximum tumor methylation QE, or a combination thereof. In some embodiments, the threshold comprises a predetermined mean tumor methylation QE.

[0038] In some embodiments, removing selected hypermethylated target loci is performed prior to determining the number of methylated molecules in the sample.

[0039] In some embodiments, subtracting background methylation is performed prior to determining the number of methylated molecules in the sample.

[0040] In some embodiments, removing hypermethylated loci and subtracting background methylation is performed prior to quantifying the number of methylated molecules in a sample.

[0041] In another aspect, the disclosure features a method of determining a DNA methylation profile in a subject, the method including: treating a sample isolated from the subject to encode the presence or absence of DNA methylation in a DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules including a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule that includes at least one of the target loci, and a mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule; creating a co-amplification mixture including the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; sequencing the co-amplification mixture at a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; quantifying the number of methylated molecules in the sample based on the number of methylated sequence reads in the sample and the number of reads from the set of synthetic molecules; and determining a methylation profile in the subject based on the number of methylated molecules in the sample.

[0042] In another aspect, the disclosure features a method of determining a DNA methylation profile in a subject over time, the method including, at a first timepoint: i) treating a sample isolated from the subject to encode the presence or absence of DNA methylation in the DNA sequence, the sample comprising at least 10 target loci from the DNA sequence; ii) adding to the sample a set of synthetic molecules, the set of molecules comprising target associated regions having nucleotide sequences that match target sequence regions of an endogenous target molecule comprising at least one of the target loci and mutant regions having nucleotide sequences that do not match the target sequence regions of the endogenous target molecule; iii) combining the amplified set of synthetic molecules with at least 10 target loci from the DNA sequence. and an amplified set of 10 target loci; iv) sequencing the co-amplification mixture at a read depth of at least one sequence read per molecule in the sample to generate sequence reads; v) determining the number of sequence reads that are methylated sequence reads; vi) quantifying the number of methylated molecules in the sample based on the number of methylated sequence reads in the sample and the number of reads from the set of synthetic molecules; repeating steps i) to vi) at a second timepoint; and determining a DNA methylation profile in the subject based on the number of methylated molecules in the sample at the first timepoint and the number of methylated molecules in the sample at the second timepoint.

[0043] In some embodiments, the number of methylated molecules in the sample from each target locus is quantified based on the number of methylated sequence reads from each target locus in the sample and the number of reads from the set of synthetic molecules.

[0044] In some embodiments, determining a methylation profile in the subject at a first timepoint and a second timepoint identifies a change in the methylation profile between the first timepoint and the second timepoint.

[0045] In some embodiments, the method further includes repeating steps i)-vi) at a third timepoint; and determining a methylation profile in the subject based on the number of methylated molecules for each target locus in the sample at the third timepoint.

[0046] In some embodiments, determining a methylation profile in the subject at the first timepoint, the second timepoint, and the third timepoint identifies a change in the methylation profile between the first timepoint and the second timepoint, between the second timepoint and the third timepoint, between the first timepoint and the third timepoint, or a combination thereof.

[0047] In some embodiments, the change in the methylation profile is indicative of an alteration in the tumor in the subject. In some embodiments, the alteration in the tumor includes a change in size of the tumor, a change in the prevalence of somatic mutations associated with the tumor, the presence of new somatic mutations associated with the tumor, a chromosomal abnormality associated with the tumor, or the tumor being resistant to treatment. In some embodiments, the change in the methylation profile is incorporated into a clinical recommendation for the subject.

[0048] In some embodiments, the method further comprises assigning the change in the methylation profile to an increased, decreased, or unchanged metric based on a comparison to a significance threshold, which in some embodiments is predetermined or dynamically calculated.

[0049] In some embodiments, the sample is taken from a blood draw that includes plasma and buffy coat, where the plasma includes cfDNA and the buffy coat includes genomic DNA (gDNA) sequences.

[0050] In some embodiments, the method further includes extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample prior to treating the sample to encode the presence or absence of DNA methylation.

[0051] In some embodiments, the method further comprises quantifying DNA methylation in a sample comprising gDNA sequences from the buffy coat.

[0052] In some embodiments, the method includes treating a sample containing DNA sequences from a buffy coat to encode the presence or absence of DNA methylation in the DNA sequences, the sample comprising at least 10 target loci from the DNA sequences; adding a set of synthetic molecules to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule comprising at least one of the target loci and a mutant region having a nucleotide sequence that does not match the sequence region of the endogenous target molecule; generating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequences; sequencing the co-amplification mixture to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; quantifying the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules; and counting the number of methylated molecules across at least two target loci to quantify DNA methylation in a sample containing DNA sequences from a buffy coat, thereby quantifying DNA methylation in the buffy coat.

[0053] In some embodiments, the method further comprises adding a spike-in of known sequence and amount to the sample prior to the treatment step.

[0054] In some embodiments, the set of synthetic molecules is a set of quantitative counting templates (QCT).

[0055] In some embodiments, each locus of the at least 10 target loci is selected based on a predicted increase in DNA methylation at that locus in cancerous tissue compared to non-cancerous tissue.

[0056] In some embodiments, at least 100 target loci are amplified, or at least 500 target loci are amplified.

[0057] In some embodiments, the sample further comprises at least one normalization locus predicted to have hypermethylation in cfDNA across both cancerous and non-cancerous tissues.

[0058] In some embodiments, the co-amplification mixture comprises an amplified set of synthetic molecules, an amplified set of all or a portion of at least 10 target loci, and at least one amplified normalization locus.

[0059] In some embodiments, the method further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantitate DNA methylation in the sample.

[0060] In some embodiments, the counting comprises counting target loci from at least 10 target loci that contain less than a threshold amount of methylated molecules in the buffy coat sample.

[0061] In some embodiments, the method further comprises normalizing the tabulated number of methylated molecules from at least two of the at least ten target loci by a methylated molecule for at least one normalization locus.

[0062] In some embodiments, the method further comprises subtracting background methylation.

[0063] In some embodiments, the method further comprises filtering the hypermethylated target loci.

[0064] In some embodiments, the method further comprises subtracting background methylation and removing selected hypermethylated loci.

[0065] In some embodiments, subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat (e.g., as quantified herein) from the number of methylated molecules in the cfDNA.

[0066] In some embodiments, subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat (e.g., as quantified herein) from the number of methylated molecules in the cfDNA on a locus basis.

[0067] In some embodiments, removing selected hypermethylated loci, subtracting background methylation, or both, is performed prior to quantifying the number of methylated molecules in a sample.

[0068] In some embodiments, removing selected hypermethylated target loci comprises removing target loci having a total number of methylated molecules in the buffy coat that exceeds a threshold. In some embodiments, the threshold is sample specific. In some embodiments, the threshold is subject specific. In some embodiments, the threshold comprises a predetermined mean tumor methylation QE, a predetermined maximum tumor methylation QE, or a combination thereof. In some embodiments, the threshold comprises a predetermined mean tumor methylation QE.

[0069] In some embodiments, the method further comprises performing a treatment selection assay on the subject upon detecting an alteration in the DNA methylation profile, in some embodiments, the treatment selection assay comprises genomic profiling to detect novel somatic mutations, prevalence of somatic mutations, or both.

[0070] In another aspect, the disclosure features a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample containing a DNA sequence, the method including the steps of: determining the presence or absence of a somatic mutation in the sample; treating the sample to encode the presence or absence of DNA methylation in the DNA sequence, the sample comprising at least 10 target loci from the DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules comprising target associated regions having nucleotide sequences that match target sequence regions of endogenous target molecules that comprise at least one of the target loci, and mutation regions having nucleotide sequences that do not match sequence regions of the endogenous target molecules; creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; and analyzing the sequences of the somatic mutations in the sample to generate sequence reads. Quantifying DNA methylation in the sample, comprising the steps of: sequencing the co-amplification mixture at a read depth of at least one read sequence per child; determining a number of sequence reads that are methylated sequence reads; and quantifying the number of methylated molecules in the sample based on the number of methylated sequence reads from the sample and the number of reads from the set of synthetic molecules; and returning an indication of confidence as a true call result for the presence or absence of a somatic mutation if the number of methylated molecules in the sample exceeds a predetermined threshold or a dynamically calculated threshold, or returning an indication of confidence as a no call result for the presence or absence of a somatic mutation if the number of methylated molecules in the sample is below a predetermined threshold or a dynamically calculated threshold, whereby the true call identifies confidence in the determination of the presence or absence of a somatic mutation.

[0071] In some embodiments, determining the presence or absence of a somatic mutation in the sample comprises determining a mutant allele frequency for the somatic mutation.

[0072] In some embodiments, the sample is taken from a blood draw that includes plasma and buffy coat, where the plasma includes cfDNA and the buffy coat includes genomic DNA (gDNA) sequences.

[0073] In some embodiments, the method further includes extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample prior to treating the sample to encode the presence or absence of DNA methylation.

[0074] In some embodiments, the method further comprises quantifying DNA methylation in a sample containing DNA sequences from the buffy coat.

[0075] In some embodiments, the method includes treating a sample containing DNA sequences from a buffy coat to encode the presence or absence of DNA methylation within the DNA sequences, the sample comprising at least 10 target loci from the DNA sequences; adding a set of synthetic molecules to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule that comprises at least one of the target loci, and a mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule; creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequences; sequencing the co-amplification mixture to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; and quantifying the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules, thereby quantifying DNA methylation within the gDNA sequences from the buffy coat.

[0076] In some embodiments, the method further comprises adding a spike-in of known sequence and amount to the sample prior to the treatment step.

[0077] In some embodiments, the set of synthetic molecules is a set of quantitative counting templates (QCT).

[0078] In some embodiments, each locus of the at least 10 target loci is selected based on a predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue.

[0079] In some embodiments, at least 100 target loci are amplified, or at least 500 target loci are amplified.

[0080] In some embodiments, the sample further comprises at least one normalization locus predicted to have hypermethylation in cfDNA across both cancerous and non-cancerous tissues.

[0081] In some embodiments, the co-amplification mixture comprises an amplified set of synthetic molecules, an amplified set of all or a portion of at least 10 target loci, and at least one amplified normalization locus.

[0082] In some embodiments, the method further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the sample.

[0083] In some embodiments, the counting comprises counting target loci from at least 10 target loci that contain less than a threshold amount of methylated molecules in the buffy coat sample.

[0084] In some embodiments, the method further comprises normalizing the tabulated number of methylated molecules from at least two of the at least ten target loci by a methylated molecule for at least one normalization locus.

[0085] In some embodiments, the method further comprises subtracting background methylation.

[0086] In some embodiments, the method further comprises removing selected hypermethylated loci.

[0087] In some embodiments, the method further comprises subtracting background methylation and removing selected hypermethylated loci.

[0088] In some embodiments, subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat (e.g., as quantified herein) from the number of methylated molecules in the cfDNA.

[0089] In some embodiments, subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat (e.g., as quantified herein) from the number of methylated molecules in the cfDNA on a locus basis.

[0090] In some embodiments, removing selected hypermethylated loci, subtracting background methylation, or both is performed prior to quantifying the number of methylated molecules in a sample.

[0091] In some embodiments, removing selected hypermethylated target loci comprises removing target loci having a total number of methylated molecules in the buffy coat that exceeds a threshold. In some embodiments, the threshold is sample specific. In some embodiments, the threshold is subject specific. In some embodiments, the threshold comprises a predetermined mean tumor methylation QE, a predetermined maximum tumor methylation QE, or a combination thereof. In some embodiments, the threshold comprises a predetermined mean tumor methylation QE.

[0092] In some embodiments, the method further comprises performing or repeating a treatment selection assay for the subject if the method returns a true call and if the prevalence of somatic mutations indicates the presence of cancer. In some embodiments, the treatment selection assay comprises genomic profiling to detect novel somatic mutations, prevalence of somatic mutations, or both.

[0093] In some embodiments, the method further comprises, upon returning a no call, repeating the methods described herein with a different sample taken from the subject.

[0094] In some embodiments, the method further comprises incorporating an indication of confidence into a clinical recommendation to the subject. [Brief description of the drawings]

[0095] 5. Brief explanation of multiple views of the drawing [Figure 1A] FIG. 1A shows a flowchart representation of a non-limiting embodiment of a method for quantifying DNA methylation in a sample. [Figure 1B] FIG. 1B shows a flowchart representation of a non-limiting embodiment of a method for quantifying DNA methylation in a sample. [Figure 2A] Figure 2A shows data from a lung cancer detection assay. Figure 2A shows total methylation molecules in technical replicates of lung cancer-detected samples with various tumor rates (0% tumor, 0.5% tumor, 1% tumor, 2.5% tumor, and 5% tumor) using a lung cancer assay. 0.5% universal methylation and 5% universal methylation were controls. [Figure 2B] FIG. 2B shows data from the lung cancer detection assay in design samples from two different lung cancer subjects. FIG. 2B shows the total (sum) methylation QE across target loci (n=113). Design samples were taken from subjects 18-0901 and 20-0407. Design samples were assayed at 0% tumor rate and 5% tumor rate using the PCR volumes and thermal cycler protocol shown in FIG. 2B. [Diagram 3]FIG. 3 shows the agreement values ​​of total QE for each of the samples analyzed in FIG. 2A. [Figure 4] Figure 4 shows methylation in lung cancer hypothesis samples with masking and background subtraction for each of the samples analyzed in Figure 3. Figure 4 shows the total normalized methylation molecules (y-axis) in lung cancer hypothesis samples and controls with various tumor rates (0% tumor, 0.5% tumor, 1% tumor, 2.5% tumor, and 5% tumor): 0.5% universal methylation and 5% universal methylation were controls. Data is shown with background subtraction and masking of target loci with high background methylation. [Diagram 5] FIG. 5 shows data from a simulation of cfDNA tumor signal quantified using somatic mutation and methylation approaches. The top panel shows mutant allele frequency (y-axis) versus fold gDNA contamination (x-axis) quantified using somatic mutation. The middle panel shows methylated molecules (y-axis) versus fold gDNA contamination (x-axis) quantified using methylation without buffy coat background methylation. The bottom panel shows methylated molecules (y-axis) versus fold gDNA contamination (x-axis) quantified using methylation with buffy coat background methylation. [Figure 6] Figure 6 shows the methylation profile of buffy coats (i.e., 5000 genome equivalents (ge) of buffy coats) isolated from different tubes, different tube types, different days, and different subjects from the same blood draw. Data shown as methylated molecules (y-axis) in different buffy coats (x-axis). [Figure 7] Figure 7 shows hierarchical clustering of methylation profiles of buffy coats isolated from different tubes of the same blood draw, different tube types, different days, and different subjects (i.e., 5000 genome equivalents (ge) of buffy coats). Cluster distances were calculated using the L1 norm. [Figure 8]Figure 8 shows data on the performance of the pan-cancer assay in design samples. Data is presented as normalized methylation QE (y-axis) for each design sample from different cancer types. Design samples were prepared at 0% tumor rate and 5% tumor rate. [Figure 9] Figures 9A-9F show data for non-limiting examples of loci used in the "personalized blacklist". Data are presented as normalized tumor QE (y-axis: "qe_norm") for each of the 12 samples shown on the x-axis in cfDNA and buffy coat. Figure 9A shows tumor QE (normalized) for intron cg03134157_77. Figure 9B shows tumor QE (normalized) for intron cg20907051_11. Figure 9C shows tumor QE (normalized) for intron_cg12880300_0. Figure 9D shows tumor QE (normalized) for locus IRX4_cg13974394_0. Figure 9E shows tumor QE (normalized) for locus EMID2_cg25290307_0. Figure 9F shows tumor QE (normalized) for intron_cg11453719_0. [Figure 10] FIG. 10 shows the normalized total tumor molecules of the specimen (y-axis) against the x-axis based on the global blacklist thresholds described herein (e.g., (1) maximum tumor QE in any non-cancer specimen >15; and (2) average tumor QE (including 0) >2). [Figure 11] Figure 11 shows the normalized total tumor molecules (y-axis) for the new blacklist compared to the normalized total tumor molecules for the old blacklist. Each blacklist was applied to the methylation data of the 30 subjects listed in Figure 11 and plotted on a graph. [Figure 12]Figures 12A-12D show data for two subjects (out of 30) who had slightly different methylation results after applying the new blacklist: subject 8272 and subject 6885. Data are presented as normalized total tumor molecules (y-axis) between days 0, 63, and 154 from first collection. Figure 12A shows data for subject 8272 using the old blacklist. Figure 12B shows data for subject 8272 using the new blacklist. Figure 12C shows data for subject 6885 using the old blacklist. Figure 12D shows data for subject 6885 using the new blacklist. [Figure 13] 13A-13E show data for five subjects created using the new global blacklist: BTO-1 (FIG. 13A), BTO-2 (FIG. 13B), BTO-3 (FIG. 13C), BTO-4 (FIG. 13D), and BTO-5 (FIG. 13E). [Figure 14] FIG. 14 shows the Response Scores measured as triplicates at each tumor rate for each tumor type. [Figure 15] FIG. 15 shows the response scores measured as triplicates at each tumor rate for each tumor type. [Figure 16] Figure 16 shows response scores measured in 8 cancer subjects, each with samples from two time points. The 8 cancer subjects included 3 cancer types (lung=4; pancreatic=3, endometrial=1). The colors in the legend indicate the calls made based on the change in response score. [Figure 17] Figure 17 shows the response scores measured in 12 healthy subjects, each with samples from two time points. The cfDNA sample at time point 1 of subject HV006 failed the sample QC criteria, so no call was made for that subject. Unchanged or indeterminate calls were made for the remaining 11 subjects. [Figure 18]Figure 18 shows the sensitivity for each tumor rate plotted with a 95% confidence interval. The dotted line indicates the 95% sensitivity, i.e., the sensitivity threshold for the limit of detection. For each tumor rate, sensitivity was calculated by comparing each of the 16 replicates to each of the 160% tumor rate replicates (256 comparisons in total). [Figure 19] Figure 19 shows box plots for the median and interquartile range of response scores at each tumor rate. The standard deviation and mean of each tumor rate were used to calculate the CV. [Figure 20] FIG. 20 shows the mock response scores for the 1% and 2% tumor DNA samples. [Figure 21] Figure 21 shows a methylation analysis for a 56 year old male subject with pancreatic ductal adenocarcinoma stage IV at the time of first collection. The subject was treated with folfiri, followed by folfiri maintenance, followed by folox. Figure 21 shows tumor QE (y-axis) versus days since last treatment (x-axis). Triangles indicate clinician staging of the cancer using imaging. [Figure 22] Figure 22 shows a methylation analysis for a 50 year old male subject with pancreatic ductal adenocarcinoma stage IV at the time of first collection. The subject was treated with fluorouracil 5000mg continued for 46 hours + leucovorin 800mg + oxaliplatin 180mg + irinotecan 300mg. Figure 22 shows tumor QE (y-axis) versus days since last treatment (x-axis). Triangles indicate clinician-mediated cancers using contrast imaging. [Diagram 23] Figure 23 shows a methylation analysis for an 87-year-old male subject with colon adenocarcinoma stage IV at the time of first collection. The subject was treated with 6 cycles of fluorouracil 1100mg + leucovorin 1100mg + bevacizumab 500mg. Figure 23 shows tumor QE (y-axis) versus days since last treatment (x-axis). Triangles indicate clinician-mediated cancers using contrast imaging. [Figure 24]Figure 24 shows methylation analysis for a 56-year-old female subject with stage IV lung squamous cell carcinoma at the time of first collection. The subject was treated with 6 doses of docetaxel. Figure 24 shows tumor QE (y-axis) versus days since last treatment (x-axis). Triangles indicate that the subject entered hospice care. [Diagram 25] Figure 25 shows the results of treatment selection assays performed on aliquots of plasma collected at the same time points as the methylation assays for 40 different samples. For each sample, Figure 25 shows the maximum variant allele frequency (VAF) (y-axis label: maximum measured VAF in hybrid capture) versus normalized total tumor QE (x-axis). [Figure 26] Figure 26 shows the methylation profile of subject 6885 measured at various time points over the course of approximately 294 days. In particular, Figure 26 shows the normalized total tumor methylator (y-axis) separated into loci by color. Methylation profiles were measured on days 0, 63, 145, 166, 229, and 294. Numbers in the legend indicate the percentage of total methylation at each locus at each individual time point. [Figure 27] Figure 27 shows the methylation profile of subject 5458 measured at various time points over the course of approximately 250 days. In particular, Figure 27 shows the normalized total tumor methylator (y-axis) separated into loci by color. Methylation profiles were measured on days 0, 35, 63, 119, 201, and 250. Numbers in the legend indicate the percentage of total methylation at each locus at each individual time point. [Figure 28] Figure 28 shows the Tumor Methylation Score expressed as the normalized sum of methylated molecules at over 500 loci that were hypermethylated in circulating tumor DNA (ctDNA) on each harvest date. The number of days since first harvest is indicated in parentheses. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0096] 6. Detailed Description In one embodiment, the present disclosure provides a method for quantifying DNA methylation in a sample containing a DNA sequence. The method may include: treating the sample to encode the presence or absence of DNA methylation in the DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules including a target-associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule and a mutated region having a nucleotide sequence that does not match a sequence region of the endogenous target molecule; creating a co-amplification mixture including an amplified set of synthetic molecules and an amplified set of at least 10 target loci from the DNA sequence; sequencing the co-amplification mixture at a read depth of at least 1 read sequence per molecule in the sample to generate sequence reads; determining the number of sequence reads that are methylated sequence reads; quantifying (determining) the absolute number of methylated molecules in the sample based on the number of methylated sequence reads from the sample and the number of reads from the set of synthetic molecules; and optionally counting the number of molecules across at least two loci.

[0097] As shown in FIG. 1A , an embodiment of a method for quantifying DNA methylation in a sample may include treating the sample to encode the presence or absence of DNA methylation 110; adding a set of synthetic molecules (e.g., quality control template (QCT) molecules) to the sample, each synthetic molecule comprising a target associated region and a mutated region 112; creating a co-amplification mixture comprising the amplified set of synthetic molecules and an amplified set of at least 10 target loci from a DNA sequence 114; sequencing the co-amplification mixture at a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads 116; determining a number of sequence reads that are methylated sequence reads 118; quantifying (determining) the number (e.g., absolute number) of methylated molecules in the sample based on the number of methylated sequence reads from the sample and the number of reads from the set of synthetic molecules 120; and facilitating a diagnosis 122 and / or facilitating a treatment of one or more diseases based on the amount of methylated molecules 124.

[0098] In some embodiments, a spike-in of known sequence and amount is added to the sample prior to treating the sample to encode for the presence or absence of DNA methylation.

[0099] In some embodiments, treating the sample to encode the presence or absence of DNA methylation is performed by bisulfite conversion or enzymatic conversion, or any other suitable methodology of treatment that separately identifies methylated DNA molecules.

[0100] In some embodiments, the set of co-amplified synthetic molecules is a set of quantitative counting templates (QCT).

[0101] In some embodiments, target loci are selected due to increased DNA methylation in cancerous tissue compared to normal tissue.

[0102] In some embodiments, at least 100 target loci are amplified.

[0103] In some embodiments, at least 800 target loci are amplified.

[0104] In some embodiments, the co-amplified mixture is sequenced at a read depth of at least 1 read per molecule.

[0105] In some embodiments, the co-amplification mixture is sequenced at a read depth of 10 reads per molecule or more, 100 reads per molecule or more, or 1000 reads per molecule or more.

[0106] In some embodiments, the co-amplification mixture is sequenced at a read depth of at least 1000 sequencing reads per genomic location.

[0107] In some embodiments, the method further comprises counting the number of molecules across loci, wherein the loci are cancer-specific loci.

[0108] In some embodiments, quantification of DNA methylation is performed in a cell-free DNA sample.

[0109] In some embodiments, cell-free DNA from plasma and cell-free DNA from buffy coat are extracted from the same patient prior to any spike-in step and processing the samples to encode the presence or absence of DNA methylation.

[0110] In some embodiments, the method further comprises counting the number of molecules across loci in the cell-free DNA sample, the loci containing fewer methylated molecules than a threshold amount in the buffy coat sample.

[0111] In some embodiments, the co-amplification mixture generated includes an amplified set of spike-in molecules, an amplified set of synthetic molecules, and an amplified set of at least 10 target loci from a plurality of DNA sequences and at least one locus that demonstrates hypermethylation in cell-free DNA.

[0112] In some embodiments, the method further comprises counting the number of methylated molecules across the target locus and at least one locus that demonstrates hypermethylation in the cell-free DNA.

[0113] In some embodiments, the method further comprises normalizing the count of target methylated molecules by the count of molecules with hypermethylation in cell-free DNA.

[0114] In some embodiments, the method further comprises the step of background subtraction.

[0115] In some embodiments, the method further comprises filtering hypermethylated loci.

[0116] In some embodiments, the method further comprises the steps of background subtraction and filtering hypermethylated loci.

[0117] In some embodiments, the background subtraction step comprises subtracting the number of methylated molecules (or methylation signal) measured in the buffy coat from the number of methylated molecules (or methylation signal) in the cfDNA.

[0118] In some embodiments, further comprising subtracting the number of methylated molecules (or methylation signals) measured in the buffy coat from the number of methylated molecules (or methylation signals) in the cfDNA on a locus basis.

[0119] In some embodiments, the step of filtering hypermethylated loci comprises filtering target loci with high background methylation prior to quantifying the absolute number of methylated molecules in the sample.

[0120] In some embodiments, the filtered target loci include target loci that have a number of methylated molecules (or methylation signals) that have buffy coat methylation above a threshold.

[0121] In some embodiments, the threshold is sample specific. In some embodiments, the threshold is subject specific. In such instances, this is referred to as an "individualized" blacklist. In some embodiments, where the threshold is sample or subject specific, the threshold is determined by evaluating buffy coat methylation signals of loci with hypermethylated cfDNA signals in non-cancer subjects.

[0122] In some embodiments, the threshold comprises a predetermined mean tumor methylation QE, a predetermined maximum tumor methylation QE, or a combination thereof.

[0123] In some embodiments, the predetermined mean tumor methylation QE is 0.5 or more, 1.0 or more, 1.5 or more, 2.0 or more, 2.5 or more, 3.0 or more, 3.0 or more, 3.5 or more, 4.0 or more, 4.5 or more, or 5.0 or more.

[0124] In some embodiments, the predetermined maximum tumor methylation QE in any non-cancerous specimen is 5 or more, 10 or more, 15 or more, 20 or more, 25 or more, 30 or more, 35 or more, 40 or more, 45 or more, or 50 or more.

[0125] In one embodiment, the thresholds include: (1) maximum tumor QE in any non-cancer specimen >15; and (2) average tumor QE (including 0) >2.

[0126] In some embodiments, target loci are selected that are differentially methylated across at least two cancer tissue sources.

[0127] In some embodiments, at least five loci per cancer tissue of origin are targeted.

[0128] In some embodiments, the cancer tissue of origin is determined based on the abundance of methylated molecules across target loci.

[0129] One aspect of the disclosure provides a method of quantifying the amount of tumor DNA in a sample containing a DNA sequence, the method comprising: i. adding a set of synthetic molecules to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule and a mutant region having a nucleotide sequence that does not match a sequence region of the endogenous target molecule; ii. creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; iii. sequencing the co-amplification mixture to a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads; iv. quantifying the absolute number of molecules containing somatic mutations in the sample based on the number of reads from the sample and the number of reads from the set of synthetic molecules; and v. determining a lower limit of mutant allele frequency based on the total number of molecules present in the sample at each locus prior to amplification.

[0130] One aspect of the disclosure provides a method of quantifying methylation in a DNA sample, the method comprising: adding a spike-in of known sequence to the sample; treating the sample to encode the presence or absence of methylation into the DNA sequence itself; adding a set of QCT or spike-in molecules to the sample, the set of molecules including target associated regions having nucleotide sequences that match target sequence regions of an endogenous target molecule and mutant regions having nucleotide sequences that do not match sequence regions of the endogenous target molecule; simultaneously amplifying the mixture of sample and QCT / spike-in molecules at 10 or more different loci; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample); determining the number of methylated reads based on the read sequences; computationally determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the QCT / spike-in at each locus; and tabulating the number of molecules across the loci.

[0131] In some embodiments, the method includes adding a fixed amount of spike-in to each sample prior to conversion so that the efficiency of the methylation conversion process can be monitored.

[0132] In some embodiments, treating the sample to encode the presence or absence of methylation on the DNA sequence itself includes performing bisulfite conversion or enzymatic conversion. Bisulfite conversion is described in U.S. Patent No. 8,257,950, which is incorporated by reference in its entirety. Bisulfite conversion and enzymatic conversion convert unmethylated cytosine to uracil. As a result, by determining the number of methylated reads based on the read sequence, it can be determined whether each read originates from a methylated molecule.

[0133] In some embodiments, QCT may be used to quantify each locus, or in other embodiments, spike molecules may be used to quantify each locus.

[0134] In some embodiments, the sample and QCT / spike-in mixture is simultaneously amplified at 10 or more different loci. In some embodiments, these loci are selected based on being more methylated in tumors compared to normal tissue of the same tissue type. In some embodiments, more than 100 positions are amplified. In some embodiments, more than 800 positions are amplified. In some embodiments, these loci have increased methylation in cancer.

[0135] In some embodiments, the method includes sequencing the amplified mixture (e.g., at a read depth of at least 1 read per molecule in the sample). The read depth qualifier can be optional. In some embodiments, the read depth qualifier is at least 1000 reads per locus (e.g., based on an expectation of 1000-2000 sample molecules at each genomic location per blood tube).

[0136] In some embodiments, treating the sample to encode the presence or absence of methylation into the DNA sequence itself comprises bisulfite conversion or enzymatic conversion.

[0137] In some embodiments, the method comprises computationally determining the number of methylated molecules in the sample at each locus based on the number of methylated reads from the sample and the number of reads from the QCT / spike-in. For spike-ins, the method comprises dividing the number of sample reads by the number of spike-in reads multiplied by the number of spike-ins added to the sample.

[0138] In some embodiments, the method includes counting the number of molecules across loci, hi certain embodiments, a portion of loci are counted (e.g., only lung cancer-specific loci are counted).

[0139] One aspect of the disclosure includes a method of quantifying methylation in cell-free DNA, the method including: extracting cell-free DNA from plasma and DNA from buffy coat from the same individual; adding a spike-in of known sequence to each DNA sample; treating each DNA sample to code for the presence or absence of methylation in the DNA sequence itself; adding a set of QCT or spike-in molecules to each DNA sample, the set of molecules including target associated regions having nucleotide sequences that match target sequence regions of the endogenous target and mutant regions having nucleotide sequences that do not match sequence regions of the endogenous target molecule; simultaneously amplifying the mixture of sample and QCT / spike-in molecules at 10 or more different loci; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample); determining the number of methylated reads based on the read sequences; and computationally determining the number of methylated molecules in the sample based on the number of reads from the sample and the number of reads from the QCT / spike-in at each locus; and tabulating.

[0140] In some embodiments, the method includes counting the number of molecules across loci in the cell-free DNA sample while excluding loci that have a significant amount of methylated molecules in the buffy coat sample. To determine significance, if any, the method includes determining a threshold number of molecules.

[0141] For example, cfDNA has different background methylation signals from different patients. Because most cell-free DNA originates from white blood cells, one aspect of the method can use the methylation signal from buffy coat as a baseline. In some embodiments, the method includes minimizing cfDNA background signal by ignoring any loci with significant methylation in buffy coat (e.g., useful for minimal residual disease (MRD) detection applications). In other embodiments, the method avoids performing background subtraction by selecting loci with low methylation background in most people.

[0142] One aspect of the disclosure includes a method of quantifying methylation in a DNA sample, comprising: adding a spike-in of known sequence to the sample; treating the sample to encode the presence or absence of methylation into the DNA sequence itself; adding a QCT or a set of spike-in molecules to the sample, the set of molecules including a) a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target, and b) a mutant region having a nucleotide sequence that does not match a sequence region of the endogenous target molecule; analyzing the mixture of sample and QCT / spike-in molecules simultaneously at at least 10 different target loci, and cfDNA. The method includes amplifying at least one locus known to be highly methylated in the sample; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample); determining the number of methylated reads based on the read sequence; computationally determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the QCT / spike-in at each locus to tabulate the number of methylated molecules across the target genes and hypermethylated genes; and normalizing the tabulated number of target methylated molecules by the tabulated number of hypermethylated genes. For example, normalizing the tabulated number of target methylated molecules by the tabulated number of hypermethylated loci may include 100 methylated molecules for 5000 input molecules.

[0143] For example, in some embodiments, the method can monitor the amount of ctDNA over time.As a result, the measurement of methylation molecules may need to be constant with respect to the total amount of DNA input.For example, if a patient's cancer is stable, but twice as much cfDNA is collected at a second time point compared to a first time point, the cancer may appear to have doubled in size even though it remains the same size.In one aspect of the method of the present invention, the method of the present invention can look at hypermethylated loci in cfDNA, regardless of whether cfDNA originates from a tumor, and can normalize target loci to these hypermethylated control loci.

[0144] One aspect of the disclosure includes a method of quantifying methylation in a DNA sample, comprising treating the sample to encode the presence or absence of methylation into the DNA sequence itself; measuring the mixture of the sample and QCT / spike-in molecules simultaneously at 10 or more different target loci, and at least one locus known to be highly methylated in cfDNA (e.g., at least one locus, at least two loci, at least three loci, at least four loci, at least five loci, at least six loci, at least seven loci, at least eight loci, at least The method includes amplifying a mixture of methylated reads at at least 9 loci, at least 10 loci, and the like; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample); determining the number of methylated reads based on the read sequences; computationally determining the average number of methylated reads within the target loci and the average number of methylated reads at highly methylated loci; tabulating the number of methylated reads across the target loci and the hypermethylated loci; and normalizing the tabulated number of target methylated reads by the tabulated number of methylated reads at hypermethylated genes.

[0145] One embodiment of the present disclosure includes a method of quantifying methylation in a DNA sample, the method including adding a spike-in of known sequence to the sample; treating the sample to encode the presence or absence of methylation into the DNA sequence itself; adding a set of QCT or spike-in molecules to the sample, the set of molecules including a) target associated regions having nucleotide sequences that match target sequence regions of the endogenous target and b) mutant regions having nucleotide sequences that do not match sequence regions of the endogenous target molecule; co-amplifying the mixture of sample and QCT / spike-in molecules at 10 or more different target loci along with loci selected to be differentially methylated across at least two cancer tissues of origin; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample) to determine the number of methylated reads based on the read sequences; computationally determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the QCT / spike-in at each locus; and determining the tissue of origin based on the abundance of methylated molecules across the loci.

[0146] For example, methylation patterns vary depending on the tumor tissue of origin. In one embodiment of the present invention, the method includes determining the tumor tissue of origin from the cfDNA sample based on which positions have methylated molecules.

[0147] In some embodiments, the mixture of sample and QCT / spike-in molecules is amplified at 5 or more loci (eg, one locus per cancer tissue of origin).

[0148] In some embodiments, determining the tissue of origin for the prevalence of methylated molecules across loci comprises assigning a score to each tissue type.

[0149] In one aspect of the disclosure, provided herein is a method of quantifying the amount of tumor DNA in a DNA sample, the method comprising: adding a set of QCT or spike-in molecules to the sample, the set of molecules comprising: a) a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target; and b) a mutation region having a nucleotide sequence that does not match a sequence region of an endogenous target molecule; simultaneously amplifying the mixture of sample and QCT / spike-in molecules at 10 or more different target loci; sequencing the amplified mixture (at a read depth of at least 1 read per molecule in the sample); computationally determining the number of molecules containing somatic mutations in the sample based on the number of reads from the sample and the number of reads from the QCT / spike-in at each locus; and determining a lower limit of mutant allele frequency based on the total number of molecules initially present in the sample at each locus.

[0150] In another aspect, the disclosure features a method of determining a DNA methylation profile in a subject, the method including: treating a sample isolated from the subject to encode the presence or absence of DNA methylation in a DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules including a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target and a mutant region having a nucleotide sequence that does not match a sequence region of an endogenous target molecule; creating a co-amplification mixture including the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; sequencing the co-amplification mixture at a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads; determining a number of sequence reads that are methylated sequence reads; quantifying (determining) a number of methylated molecules for each target locus in the sample based on the number of methylated sequence reads from each target locus in the sample and the number of reads from the set of synthetic molecules; and determining a methylation pattern in the subject based on the number of methylated molecules for the target loci in the sample.

[0151] In another aspect, the disclosure features a method of determining a DNA methylation profile in a subject over time, the method including, at a first timepoint: i) treating a sample isolated from the subject to encode the presence or absence of DNA methylation in a DNA sequence; ii) adding a set of synthetic molecules to the sample, the set of molecules including target associated regions having nucleotide sequences that match nucleotide sequences corresponding to endogenous targets and mutant regions having nucleotide sequences that do not match sequence regions of the endogenous target molecules; iii) creating a co-amplification mixture including the amplified set of synthetic molecules and the amplified set of at least ten target loci from the DNA sequence; iv) extracting the sequence linkages from the DNA sequence; the co-amplified mixture at a read depth of at least one read sequence per molecule in the sample to generate a methylation pattern in the subject based on the number of methylated sequence reads from each target locus in the sample and the number of reads from the set of synthetic molecules; repeating steps i)-vi) at a second timepoint; and determining a methylation pattern in the subject based on the number of methylated molecules for each target locus in the sample at the first timepoint and the number of methylated molecules for each target locus in the sample at the second timepoint.

[0152] In some embodiments of determining a DNA methylation profile in a subject over time, the method includes repeating steps i)-vi) at a third timepoint, a fourth timepoint, or a fifth timepoint, and determining a methylation pattern in the subject based on the number of methylated molecules for each target locus in the sample at the third timepoint, the fourth timepoint, or the fifth timepoint.

[0153] In some embodiments of determining a DNA methylation profile in a subject, the method performs a treatment selection assay on the subject upon detecting a change in the DNA methylation profile. In some embodiments, the treatment selection assay comprises genomic profiling to detect novel somatic mutations, the prevalence of somatic mutations, or both.

[0154] In some embodiments, the methods detect methylation signals at each locus, thereby generating a methylation profile at each time point that includes a summary of the methylation signals at each locus (see, e.g., Example 10). The methylation profile can then change over time with each time point and can be expressed as a summary of the methylation signals at each locus (see, e.g., Example 10).

[0155] Quantifying the methylation profile over time can allow monitoring of changes in the methylation profile that are indicative of changes in disease progression. For example, a large increase in the percentage that one or more loci contribute to the methylation profile from one time point to another (i.e., a large increase in the composition of the methylation pattern associated with one or more loci) can indicate increased disease progression (e.g., increased tumor size).

[0156] In another aspect, the disclosure features a method for quantifying the prevalence of somatic mutations in a sample containing a DNA sequence, the method including the steps of: determining the prevalence of somatic mutations in the sample; treating the sample to encode the presence or absence of DNA methylation in the DNA sequence, the sample including at least 10 target loci from the DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules including target associated regions having nucleotide sequences that match target sequence regions of an endogenous target and mutation regions having nucleotide sequences that do not match sequence regions of an endogenous target molecule; and creating a co-amplification mixture including the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence. quantitating DNA methylation in a sample comprising the steps of: sequencing the co-amplified mixture at a read depth of at least one sequence read per molecule in the sample to generate sequence reads; determining the number of sequence reads that are methylated sequence reads; and quantifying (determining) the absolute number of methylated molecules in the sample based on the number of methylated sequence reads from the sample and the number of reads from the set of synthetic molecules; and returning a true call result for the prevalence of somatic mutations if the number of methylated molecules is above a predetermined threshold or a dynamically calculated threshold, or returning a no call result for the prevalence of somatic mutations if the number of methylated molecules is below a predetermined threshold or a dynamically calculated threshold.

[0157] In some embodiments, the somatic mutation is selected from one or more of a single nucleotide variant, a copy number change, an insertion, and a deletion.

[0158] In some embodiments, determining the prevalence of somatic mutations comprises using variant allele frequency (VAF).

[0159] In some embodiments, the prevalence of somatic mutations indicates the presence of cancer if the prevalence of somatic mutations is equal to or greater than a predetermined threshold.

[0160] In some embodiments, the method includes performing a treatment selection assay on the subject if the method returns a true call and if the prevalence of somatic mutations indicates the presence of cancer, in some embodiments, the treatment selection assay includes genomic profiling to detect novel somatic mutations, prevalence of somatic mutations, or both.

[0161] In some embodiments, the method also includes repeating the method of determining the prevalence of somatic mutations and the method of quantifying DNA methylation in different samples taken from the subject upon returning a no call.

[0162] Method embodiments described herein, for example as shown in Figure 1A, may be implemented in any suitable manner. Method embodiments as shown in Figure 1A may include various combinations and permutations of the various system components and various method steps, including any variations (e.g., embodiments, variations, examples, specific examples, diagrams, etc.), and some of the method and / or process embodiments described herein may be performed asynchronously (e.g., serially), simultaneously (e.g., in parallel), or in any other suitable order by and / or with one or more examples, elements, components, and / or other aspects of system 200 and / or other objects described herein.

[0163] Any of the variations (e.g., embodiments, variations, examples, specific examples, diagrams, etc.) described herein, and / or any portion of the variations described herein, may additionally or alternatively be combined, aggregated, excluded, used, performed in series, performed in parallel, and / or applied in other ways.

[0164] Some of the embodiments of the method as described in FIG. 1A may be embodied and / or performed at least in part as a machine configured to receive a computer readable medium storing computer readable instructions. The instructions may be executed by a computer executable component that may be integrated with the system. The computer readable medium may be stored in any suitable computer readable medium such as a RAM, a ROM, a flash memory, an EEPROM, an optical device (CD or DVD), a hard drive, a floppy drive, or any suitable device. The computer executable component may be a general or application specific processor, although any suitable dedicated hardware or combination hardware / firmware device may alternatively or additionally execute the instructions.

[0165] Those skilled in the art will recognize from the above detailed description, and from the figures and claims, that improvements and modifications can be made to the embodiments of the method (e.g., as shown in FIG. 1A) without departing from the scope defined in the claims.

[0166] 6.1. Making synthetic molecules and adding synthetic molecules to samples In some embodiments, a method of quantifying DNA methylation in a sample includes adding to a treated sample a set of synthetic molecules (e.g., QCT molecules), the set of synthetic molecules including target-associated regions (e.g., associated with a genetic disorder, etc.) having sequence similarity to a target sequence region of an endogenous target molecule, and mutant regions (e.g., including an embedded molecular identifier (EMI) region including a set of variable "N" bases, where each "N" base is selected from any one of an "A" base, a "G" base, a "T" base, and a "C" base) having sequence dissimilarity to the sequence region of the endogenous target molecule.

[0167] In some embodiments, the method may include generating a set of QCT molecules that can function to generate molecules to be used (e.g., added, processed, sequenced, etc.) in one or more stages (e.g., steps, phases, periods, periods, etc.) of at least one of sequencing library preparation and sequencing (e.g., high throughput sequencing, etc.), such as to facilitate downstream computations (e.g., determining the number of QCT sequence reads to facilitate quantification of the number (or percentage) of methylated molecules).

[0168] In some embodiments, the synthetic molecule (e.g., a QCT molecule) comprises a target-associated region (e.g., one or more target-associated regions per QCT molecule, etc.). As shown in FIG. 1B, the target-associated region comprises sequence similarity (e.g., total sequence similarity; sequence similarity that meets a threshold condition; sequence similarity of a specified number of bases, etc.) to one or more target sequences of one or more target molecules (e.g., endogenous target molecules; corresponding to one or more biological targets; etc.), but may additionally or alternatively comprise any suitable association with any suitable component of one or more target molecules. The target-associated region preferably allows for co-amplification of a nucleic acid molecule (e.g., a nucleic acid, a nucleic acid fragment, etc.) that comprises the target sequence region with the corresponding QCT molecule (e.g., including a target-associated region, etc.), which may facilitate improved accuracy in molecule counting (e.g., in determining molecule counting parameters; by accounting for amplification bias; etc.), but may additionally or alternatively allow for any suitable step associated with sequencing library preparation, sequencing, and / or some of the method embodiments described herein. In one example, the methods described herein may include co-amplifying a set of QCT molecules and a nucleic acid molecule comprising a biological target based on sequence similarity between a target associated region and a target sequence region of the biological target, where quantifying DNA methylation in the sample may include determining a target molecule count (e.g., a methylated molecule count) that describes the number of methylated molecules associated with sequencing based on a set of synthetic molecules (e.g., QCT sequence read clusters).

[0169] In some embodiments, the synthetic molecule (e.g., QCT molecule) may omit the target-associated region. For example, the QCT molecule may be used with a component of a sample that includes a biological target without target association (e.g., without a predetermined similarity to the target sequence region of the biological target) and / or without corresponding co-amplification with the sample component (e.g., a nucleic acid molecule that includes the target sequence region, etc.). In some examples, the QCT molecule may be pre-processed to be compatible with sequencing, such as when the pre-processed QCT molecule can be added to a processed sample suitable for sequencing (e.g., to improve user convenience) so that it can be co-sequenced without the need for co-amplification. The QCT molecule with the target-associated region removed can be used to facilitate contamination parameter determination, preferably, but may additionally or alternatively be used to facilitate any suitable sequencing-related parameter determination. In a specific example, the set of QCT molecules can be adapted for subsequent sequencing (e.g., high-throughput sequencing such as NGS, etc.), where generating the set of QCT molecules can include amplifying a first portion of the QCT molecules of the set of QCT molecules (e.g., each including a first shared QCT identifier region, etc.); and amplifying a second portion of the QCT molecules of the set of QCT molecules (e.g., each including a second shared QCT identifier region, etc.), where QCT molecule sequencing reads are derived from sequencing corresponding to a QCT mixture generated based on the first portion of the QCT molecules and a sample including a biological target (e.g., including a first target molecule corresponding to the biological target, etc.), and an additional QCT mixture generated based on the second portion of the QCT molecules and an additional sample including the biological target (e.g., including a second target molecule corresponding to the biological target, etc.), where the sample and the additional sample correspond to a first sample compartment and a second sample compartment, respectively, of the sample compartment. However, the target associated region and / or the QCT molecule excluding the target associated region may be constructed in any suitable manner.

[0170] In some embodiments, the synthetic molecule (e.g., a QCT molecule) comprises one or more mutation regions (e.g., one or more mutation regions per QCT molecule; adjacent mutation regions; separate mutation regions; etc.). As shown in FIG. 1B, the mutation regions preferably comprise sequence dissimilarity (e.g., complete sequence dissimilarity; dissimilarity of a specified number of bases; partial sequence dissimilarity; etc.) to one or more sequence regions of the target molecule (e.g., distinct from the target sequence region; etc.). The mutation regions may additionally or alternatively comprise one or more EMI regions. In a variation, the EMI regions may comprise a set of variable "N" bases (e.g., one or more variable "N" bases, etc.), where each "N" base is selected from any one of "A" bases, "G" bases, "T" bases, and "C" bases (e.g., randomly selected; selected according to a predetermined statistical distribution and / or probability; etc.). In variations, the EMI regions may include synthetic regions (e.g., on microarrays; using silicon-based synthesis; etc.) that include one or more designated bases (e.g., designed and synthesized bases; etc.), such as synthetic regions designed to facilitate QCT sequence read cluster determination (e.g., by maximizing pairwise Hamming distance between EMI regions). In variations, the QCT molecule may additionally or alternatively include multiple EMI regions (e.g., multiple EMI regions; adjacent EMI regions; separate EMI regions: EMI regions that include variable "N" bases; mutation regions, including EMI regions that include synthetic regions; etc.).For example, each mutation region of the set of QCT molecules can include an embedded molecular identifier region that includes a set of variable "N" bases, where each "N" base is selected from any one of an "A" base, a "G" base, a "T" base, and a "C" base; each QCT molecule of the set of QCT molecules can further include an additional EMI region that includes an additional set of variable "N" bases, where the additional EMI region is separated from the EMI region by a sequence region of the QCT molecule; the set of variable "N" bases and the additional set of variable "N" bases can each include a determined (e.g., predetermined) number of "N" bases (e.g., more than three "N" bases, more than any suitable number of "N" bases, an exact number of "N" bases; etc.); and determining a sequencing-related parameter (e.g., a contamination parameter) can be based on a QCT sequence read cluster derived based on the EMI region and the additional EMI region of the set of QCT molecules (e.g., based on distinctly different EMI sequence reads corresponding to pairs of the EMI region and the additional EMI region, etc.).

[0171] In some embodiments, as shown in FIG. 1B, the QCT molecules can include a QCT identifier region that identifies the QCT molecule (and / or other suitable QCT molecules), such as a shared QCT identifier region (e.g., a shared sequence region that has dissimilarity to one or more sequence regions of the target molecule, etc.) that identifies the QCT molecule as belonging to a set of QCT molecules (e.g., where different QCT identifier regions are unique to different sets of QCT molecules, etc.). In one example, the mutation region of each QCT molecule of the first set of QCT molecules can include a first EMI region separated from the second EMI region by at least the first QCT identifier region, where each additional QCT molecule of the second set of QCT molecules can include a first additional EMI region separated from the second additional EMI region by at least the second QCT identifier region. In one example, the first EMI region, the second EMI region, the first additional EMI region, and the second additional EMI region can comprise a set of variable "N" bases, where each "N" base is selected from any one of an "A" base, a "G" base, a "T" base, and a "C" base, and where computationally determining the set of QCT sequence read clusters can include determining the set of QCT sequence read clusters based on the first and second QCT identifier regions, and based on the first EMI region, the second EMI region, the first additional EMI region, and the second additional EMI region. In one example, for each QCT molecule of the first set of QCT molecules, the corresponding QCT molecule sequence is characterized by overall sequence similarity to the first sequence template of the biological target except for the first QCT identifier region, the first EMI region, and the second EMI region, and wherein for each additional QCT molecule of the second set of QCT molecules, the corresponding additional QCT molecule sequence is characterized by overall sequence similarity to the second sequence template except for the second QCT identifier region, the first additional EMI region, and the second additional EMI region. In a specific example, the QCT molecule sequence can be identical to the target molecule sequence (e.g., one or more regions of the target molecule sequence, etc.) except for two separate sections of the 5N sequence interrupted by distinct, previously determined QCT identifier regions (e.g., unique identifier sequences, etc.).In specific examples, QCT identifier regions (e.g., unique QCT ID sequences, as shown in FIG. 1B) can be used to allow for the use of multiple QCT libraries that can be added in one step for internal control or at different steps to track loss of input biological targets or other user errors. Additionally or alternatively, QCT identifier regions can be configured in any suitable manner. However, QCT molecules can include any suitable combination of regions of any suitable type (e.g., where different QCT molecules include the same or different types and / or numbers of regions; with any suitable sequence similarity and / or dissimilarity to sequence regions of the target molecule, etc.).

[0172] In some embodiments, the method may additionally or alternatively include generating one or more QCT libraries (e.g., each QCT library containing a QCT molecule, etc.), where the QCT library may include multiple sets of QCT molecules, etc., each set of QCT molecules may be distinguished by a different QCT identifier region, etc. In one example, generating a QCT library may include amplifying different sets of QCT molecules (e.g., in preparation for sequencing, the QCT molecules are amplified prior to addition to one or more components of the sample to generate a QCT mixture, etc.). In one example, generating a QCT library may include determining the number of QCT molecules contained in the QCT library. In a specific example, a solution to the birthday problem can be used to determine the maximum number of unique QCT molecules to be included in each sample given the individual diversity of the QCT molecules: for 4^10 sequences that can be created with 10 variable N bases in the QCT molecules, up to 1200 QCT molecules can be used with a probability of about 0.5 of a single effective EMI collision (exp(-1200 *1199 / 2 / 4^10)~0.5), with 200 QCT molecules, the probability of a single effective collision is about 2%, etc. In a specific example, creating a QCT library can include creating a QCT library adapted for deployment (e.g., in at least one single step of sequencing library preparation and high-throughput sequencing, etc.) of less than 0.00001 nanograms (and / or other suitable amount) of amplifiable QCT molecules for each sample in a set of samples. However, determining the number of QCT molecules included in the QCT library and creating the QCT library can be performed in any suitable manner.

[0173] In one embodiment, a QCT library can be generated by synthesizing a complementary strand to a single-stranded oligonucleotide sequence containing a variable "N" sequence. In a specific example, a double-stranded QCT library can be generated by resuspending and annealing QCT Ultramer with a complementary primer sequence, extending the sequence with Klenow fragment (exo-), and treating with exonuclease I. The final product can be purified to remove unused single-stranded DNA molecules, and the QCT library can be quantified using a fluorescent assay such as Qubit HS assay, and then the number of QCT molecules added to each sample can be calculated by using the predicted molecular weight of the double-stranded QCT molecules. However, generating QCT molecules can be performed in any suitable manner.

[0174] In some embodiments, a synthetic molecule library (e.g., a QCT library) can be added at different sequencing library preparation steps (e.g., sample preparation steps) and / or sequencing steps to track sample loss. In one embodiment, a first set of QCT molecules (e.g., QCT1 molecules; a first QCT molecule containing a first shared QCT identifier region; etc.) is dispensed at the time of sample collection, and an equal amount of a second set of QCT molecules (e.g., QCT2 molecules; a second QCT molecule containing a second shared QCT identifier region; etc.) is dispensed after purification of the sample (or after treating the sample to encode the presence or absence of DNA methylation), and the purification yield (or treatment efficiency yield) can be evaluated via a comparison of the molecular counts of the first set of QCT molecules and the second set of QCT molecules (e.g., QCT1 molecule counts vs. QCT2 molecule counts, etc.). In one example, a synthetic molecule library can be used to track the efficiency of treating a sample to encode the presence or absence of DNA methylation (i.e., measuring bisulfite conversion efficiency).

[0175] In some embodiments, the set of synthetic molecules includes a target-associated region having a nucleotide sequence that matches a target sequence region of an endogenous target and a mutated region having a nucleotide sequence that does not match a target sequence region of an endogenous target molecule. As used herein, the term "matched" refers to sufficient complementarity between two nucleotide sequences that the two nucleotide sequences bind to form the basis for a nucleic acid extension reaction (e.g., amplification). Sufficient complementarity can refer to a match that includes up to 5 mismatches between two nucleotide sequences (e.g., a target-associated region on a synthetic molecule and a target sequence region on an endogenous target). As used herein, the phrase "not matched" refers to insufficient complementarity between two nucleotide sequences such that sufficient Watson-Crick base pairing is not achieved to allow binding of the two nucleotide sequences, and little or no amplification occurs when performing an amplification reaction. Insufficient complementarity can refer to a match that includes more than 5 mismatches between two nucleotide sequences (e.g., a target-associated region on a synthetic molecule and a target sequence region on an endogenous target).

[0176] 6.2. Co-amplification and Sequencing of Co-Amplified Mixtures In some embodiments, creating an amplification mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target loci from a DNA sequence comprises amplifying synthetic molecules (e.g., QCT molecules) comprising target-associated regions and nucleic acid molecules (e.g., nucleic acids encoding the presence or absence of DNA methylation). For example, the target-associated regions preferably allow for co-amplification of corresponding QCT molecules (e.g., comprising target-associated regions, etc.) and nucleic acid molecules (e.g., nucleic acids, nucleic acid fragments, etc.) comprising target sequence regions, which may facilitate improved accuracy in molecule counting (e.g., in determining molecule counting parameters; by taking into account amplification bias, etc.), but may additionally or alternatively allow for any suitable step associated with sequencing library preparation, sequencing, and / or part of the method embodiment. In one example, co-amplification of a set of QCT molecules and nucleic acid molecules comprising methylated DNA is based on sequence similarity between the target-associated regions and the target sequence regions of the methylated DNA.

[0177] In some embodiments, the sample comprises DNA sequences having loci predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue. Loci predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue are selected based on, for example, but not limited to, a population-based survey of methylated loci, patient-specific metrics (e.g., methylation patterns from tumors), and / or literature-based studies identifying methylated loci.

[0178] In some embodiments, loci predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue are amplified and analyzed for the presence of DNA methylation. In some embodiments, loci predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue include at least 10, 20, 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000 or more loci, each predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue. In some embodiments, loci predicted to have increased DNA methylation in cancerous tissue compared to non-cancerous tissue also include loci whose DNA sequences are methylated (e.g., highly methylated) in both cancerous and non-cancerous tissue. In such examples, generating a co-amplification mixture results in a co-amplification mixture including an amplified set of synthetic molecules, an amplified set of sets (e.g., at least 10 target loci having a predicted increase in DNA methylation in cancerous tissue compared to non-cancerous tissue), and an amplified set of at least one normalization locus (e.g., a locus that is methylated (e.g., highly methylated) in both cancerous and non-cancerous tissue).

[0179] In some embodiments, sequencing associated with one or more embodiments of method 100 (e.g., associated with 116) preferably includes NGS, NGS-related technologies, massively parallel signature sequencing, polony sequencing, 454 pyrosequencing, Illumina sequencing, SOLiD sequencing, Ion Torrent semiconductor sequencing, DNA nanoball sequencing, Heliscope single molecule sequencing, single molecule real-time (SMRT) sequencing, nanopore DNA sequencing, any generation of sequencing technologies (e.g., second generation sequencing technologies, third generation sequencing technologies, fourth generation sequencing technologies, etc.), amplicon-related sequencing (e.g., targeted amplicon sequencing), metagenomics-related sequencing, sequencing-by-synthesis, tunneling current sequencing, sequencing-by-hybridization, mass spectrometry sequencing, microscopy-based technologies, and / or any suitable technology related to high-throughput sequencing. In some embodiments, sequencing can include any suitable sequencing technique (e.g., Sanger sequencing, capillary sequencing, etc.).

[0180] 6.3. Determining and Quantifying Methylated Molecules In some embodiments, determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the QCT / spike-in at each locus includes one or more of: target molecule count (e.g., absolute molecule count of methylated molecules, such as in the original sample; absolute count of endogenous target molecules, such as in the original sample, etc.); reference molecule count (e.g., absolute count of endogenous reference molecules, such as in the original sample, etc.); QCT molecule count (e.g., corresponding to the number of valid QCT sequence read clusters; corresponding to the number of distinct QCT molecules added to the components of the sample, etc.); association ratio (e.g., a correction factor; the ratio of molecule count to the associated number of sequence reads; etc.); and / or any other suitable parameter related to the methylated molecule count, field.

[0181] As shown in FIG. 1A , the methylated molecule count is preferably used in facilitating one or more diagnoses 122, but may additionally or alternatively be used in (e.g., as an input thereto) any suitable portion of an embodiment of method 100, including facilitating one or more disease treatments 124.

[0182] In some embodiments, determining a methylated molecule count parameter (e.g., methylated molecule count, etc.) may be based on a ratio of a correction factor determined based on the QCT molecule count (e.g., corresponding to the number of QCT sequence read clusters, such as the number of valid QCT sequence read clusters, etc.) and the QCT molecule sequence reads (e.g., the number of QCT molecule sequence reads corresponding to QCT sequence read clusters, etc.), such as by multiplying the number of methylated molecule sequence reads by the ratio of the correction factor. In particular examples, the number of valid non-contaminating QCT sequence read clusters (e.g., remaining QCT sequence read clusters after discarding QCT sequence read clusters having 2 or less reads and / or any suitable number or less reads; etc.) may indicate a QCT molecule count (e.g., the number of QCT molecules for individual sample partitions; individual samples; individual sample discriminators, etc.). In a specific example, the correction factor can be found by dividing the QCT molecule count by the sequencing reads obtained from the corresponding QCT molecule, such that the correction factor multiplied by the sequencing reads belonging to the target molecule (e.g., in each sample section; from each sample; associated with each sample identification factor, etc.) results in a target molecule count (e.g., the absolute number of the first biological target molecule available by the assay for amplification, etc.). In one example, the average QCT sequencing depth used in determining the absolute count of the endogenous target molecule and the absolute count of the endogenous reference molecule is determined separately from the corresponding QCT.

[0183] Alternatively, in one embodiment, the read depth threshold for discarding QCT sequence read clusters (e.g., for determining molecular count parameters and / or appropriate sequencing-related parameters, etc.) can be adaptively determined based on features of the QCT molecular sequence read (e.g., EMI sequence read) depth distribution. For example, the threshold can be set for each indexed sample by computing the average EMI read depth in each sample, computing the square root of this average read depth, and discarding QCT sequence read clusters having a read depth less than the square root of the average read depth. Additionally or alternatively, the read depth threshold for discarding QCT sequence read clusters can be computed in any suitable manner. However, determining methylated molecular counts can be performed in any suitable manner.

[0184] In some embodiments, variant calling involves determining the variant allele frequency (VAF) of a DNA variant. As used herein, "variant allele frequency" or "variant allele rate" refers to the percentage of sequence reads observed to match a specific DNA variant divided by the total coverage of that locus.

[0185] In some embodiments (e.g., of quantifying DNA methylation), the method of FIG. 1A may be used to: a) determine parameters for use in an algorithm for determining a diagnostic outcome of an assay; b) track loss of input methylated DNA (e.g., at different stages of an assay); c) return a no call result if the number of methylated molecules is too low (e.g., to determine if the assay is unreliable, etc.); d) return a no call result if the corresponding quantification of the prevalence of somatic mutation(s) in the sample is too low (e.g., to determine if the assay is unreliable, etc.); e) design assays for detecting changes in methylation profiles; and / or f) support therapeutic and clinical decision making based on the results of a diagnostic assay.

[0186] Embodiments may additionally or alternatively determine the portion of biological material available to the assay, such as by quantification of biological targets (e.g., methylated molecules) based on using QCT molecules, which may be improved by measuring the total genomic material available and calculating the predicted biological target (e.g., methylated molecule) concentration, since not all targets are available to the assay. In specific examples, this may be due to shearing the DNA into a short size distribution, such as in the example of circulating free DNA being assayed in liquid biopsy applications when circulating tumor DNA is assayed.

[0187] Some embodiments of the methods provided herein (see, e.g., FIG. 1A ) may be used in connection with one or more diseases (e.g., in connection with characterizing, diagnosing, treating, and / or performing a process related to one or more diseases), where the diseases may include cancer and / or may otherwise be associated with cancer (e.g., through analysis related to any suitable oncogene, cancer biomarker, and / or other cancer-related target; through analysis related to liquid biopsy), and / or any other suitable disease. In one example, the method (of FIG. 1A and / or as described herein) may include determining a methylated molecule count (e.g., corresponding to the number of methylated molecules in a sample; based on the use of QCT molecules, etc.) to facilitate diagnosis related to liquid biopsy.

[0188] In some embodiments, the number of methylated molecules for cfDNA and the number of methylated molecules for gDNA are quantified in the same workflow. For example, the methylated molecules for cfDNA and the methylated molecules for gDNA can be indexed (e.g., using a defined index sequence) to distinguish between sources of methylated molecules (e.g., cfDNA vs. gDNA). This allows cfDNA and gDNA to be amplified and / or sequenced in the same reaction (e.g., multiplexing).

[0189] 6.4. Processing methylation sequencing reads Some embodiments of the methods provided herein (see, e.g., FIG. 1A ) include a step of processing the methylated sequence reads. In some embodiments, processing the methylated sequence reads includes removing selected hypermethylated target loci; and / or subtracting background methylation. The processing step (e.g., removing selected hypermethylated target loci and / or subtracting background methylation) may be performed prior to determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of sequence reads from the set of synthetic molecules. The processing step (e.g., removing selected hypermethylated target loci and / or subtracting background methylation) may be performed prior to determining the number of molecules among the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of sequence reads from the set of synthetic molecules, but prior to the counting step.

[0190] In some embodiments, the step of subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat (i.e., gDNA) from the number of methylated molecules in the cfDNA. In such examples, methylated molecules are quantified for both cfDNA and gDNA. For example, methylation measured in gDNA is then subtracted from cfDNA methylation to remove background from the cfDNA methylation signal. In some embodiments, the step of background subtraction includes subtracting the number of methylated molecules measured in the buffy coat (i.e., gDNA) from the number of methylated molecules in cfDNA on a locus basis. For example, methylation measured in gDNA for an individual target locus is subtracted from cfDNA methylation measured for the same target locus to remove background from the cfDNA methylation signal on a locus basis.

[0191] In some embodiments, removing selected hypermethylated target loci includes removing target loci that have a total number of methylated molecules in the buffy coat that exceeds a threshold (e.g., a predetermined number of quantified methylated molecules). In some embodiments, removing selected hypermethylated target loci may include removing target loci with high background methylation prior to quantifying the number of methylated molecules in the sample. In some embodiments, the selected hypermethylated target loci that are removed include target loci that have high methylated cfDNA signals in non-cancerous subjects. In some embodiments, the selected hypermethylated target loci are selected from a global blacklist (see, e.g., Example 6). In some embodiments, the selected hypermethylated target loci are selected from an individual-specific blacklist (see, e.g., Example 6).

[0192] In some embodiments, the selected hypermethylated target loci that are removed include target loci that have a number (e.g., total number) of methylated molecules in the buffy coat that exceeds a threshold. In some embodiments, the threshold is sample specific (e.g., the number of methylated molecules in the buffy coat is determined for each sample). In some embodiments, the threshold is subject specific (e.g., the number of methylated molecules in the buffy coat is determined for each subject). In some embodiments, the threshold includes a predetermined average tumor methylation quantitative equivalent (QE), a predetermined maximum tumor methylation QE, or a combination thereof. The quantitative equivalent (QE) is an estimate of the number of genome equivalents of a locus based on QCT analysis. In some embodiments, the threshold includes a maximum tumor QE>15 in any non-cancer specimen, where hypermethylated target loci with a maximum tumor QE greater than 15 in any non-cancer specimen are removed. In some embodiments, the threshold includes an average tumor QE (including 0)>2, where hypermethylated target loci with an average tumor QE greater than 2 are removed.

[0193] 6.5. Aggregation and normalization Some embodiments of the methods provided herein (see, e.g., FIG. 1A ) include counting the number of methylated molecules across all or a portion of the target loci (e.g., a portion of at least 10 target loci) to quantify DNA methylation in the sample. In some embodiments, the counting includes counting the number of methylated molecules across at least two target loci (e.g., at least two of the at least 10 target loci) to quantify DNA methylation in the sample. In some embodiments, the method includes counting methylated molecules in a cfDNA (plasma) sample across at least two target loci (e.g., at least two of the at least 10 target loci). In some embodiments, the method includes counting methylated molecules in a buffy coat (gDNA) sample across at least two target loci (e.g., at least two of the at least 10 target loci).

[0194] In some embodiments, counting the number of methylated molecules across all or a portion of the target loci (e.g., all or a portion of at least 10 target loci) includes counting target loci that contain less than a threshold amount of methylated molecules in the buffy coat sample. The threshold amount of methylated molecules in the buffy coat sample can be predetermined or dynamically calculated. A dynamically calculated threshold is a threshold that is not fixed but varies based on certain conditions or inputs. In the case of a threshold amount of methylated molecules, the threshold can be dynamically calculated based on factors such as the size of the system being tested, the concentration of the molecules present, or the sensitivity of the detection method used to measure the molecules. For example, in a sample (e.g., a blood sample), the threshold amount of methylated molecules may vary depending on the type of cancer or stage of cancer.

[0195] Some embodiments of the methods provided herein (see, e.g., FIG. 1A) include normalizing the aggregate number of methylated molecules to aid in quantifying DNA methylation in a sample. In one embodiment, the method includes normalizing the aggregate number of target methylated molecules by the aggregate number of methylated molecules in a normalization locus (e.g., a methylated molecule for at least one normalization locus). In some embodiments, the method includes normalizing the aggregate number of target methylated molecules by the aggregate number of methylated molecules in a normalization locus. In such an example, the method includes aggregate number of methylated molecules across at least two normalization loci. In some embodiments, the method includes normalizing the aggregate number of methylated molecules from at least two of the at least ten target loci by the methylated molecule for at least one normalization locus.

[0196] The counting and / or normalization step allows for quantification of a Tumor Methylation Score, which represents the normalized sum of methylated molecules that are hypermethylated in a sample (eg, at at least 10 target loci).

[0197] Some embodiments of the methods provided herein (see, e.g., FIG. 1A) can be used to quantitate the presence of tumor DNA in a sample (see, e.g., Section 6.8).

[0198] Some embodiments of the methods provided herein (see, e.g., FIG. 1A) can be used to determine a DNA methylation profile in a subject (see, e.g., Section 6.9).

[0199] Some embodiments of the methods provided herein (see, e.g., FIG. 1A) can be used to quantitate the presence of somatic mutations in a sample (see, e.g., Section 6.10).

[0200] Additionally or alternatively, the data described herein (e.g., sequencing-related parameters, identifiers, read depths, sequence reads, sequence region determinations, QCT molecular designs, primer designs, etc.) may be associated with any suitable temporal indicator (e.g., seconds, minutes, hours, days, weeks, time periods, timepoints, timestamps, etc.), including one or more of: one or more temporal indicators indicating the time the data was collected, determined, transmitted, received, and / or otherwise processed; temporal indicators that provide context to the content described by the data, such as temporal indicators indicating sequences of sequencing library preparation and / or sequencing stages; changes in temporal indicators (e.g., data over time; changes in data; data patterns; data trends; data extrapolation and / or other predictors, etc.); and / or any other suitable indicator related to time.

[0201] Additionally or alternatively, the parameters, metrics, inputs, outputs, and / or other suitable data described herein may relate to types of values, including any one or more of scores, binary values, classifications, confidence intervals, identifiers (e.g., sample identifiers, QCT molecular identifiers, etc.), values ​​along ranges, and / or any other suitable types of values. Any suitable type of data described herein may be used as inputs, produced as outputs, and / or manipulated in any suitable manner for any suitable component associated with an embodiment of method 100 and / or system 200.

[0202] Optionally, the embodiment described in FIG. 1B may additionally or alternatively include a sample handling network configured to process the biological sample to generate molecules (e.g., QCT molecules; QCT libraries; etc.), and / or perform other suitable steps; a sequencing system configured to sequence the genetic material processed from the mixture generated based on the biological sample and the QCT molecules; a computer system (e.g., a remote computing system; a local computing system, etc.) configured to analyze sequence reads, determine QCT sequence read clusters, determine sequencing-related parameters, facilitate diagnosis, facilitate treatment, and / or perform other suitable steps (e.g., computational steps); and / or any other suitable components. The components of the system in FIG. 1B may be physically and / or logically integrated in any manner (e.g., using any suitable distribution of functionality of any components, such as those involved in some of the method embodiments in FIG. 1A, etc.). However, method 1A and system 1B may be configured in any suitable manner.

[0203] 6.6. Facilitating diagnosis, treatment, or further evaluation In one aspect, the disclosure features a method (FIG. 1A) that may additionally or alternatively include facilitating diagnosis 122, which may function to assist in, determine, provide, and / or otherwise facilitate one or more diagnoses of one or more diseases.

[0204] Facilitating one or more diagnoses may include any one or more of: determining one or more diagnoses (e.g., based on the number of methylated molecules, etc.); providing one or more diagnoses (e.g., to one or more users, such as for use by one or more healthcare providers in providing a medical diagnosis to a patient; to one or more healthcare providers, etc.); assisting one or more diagnoses (e.g., providing one or more sequencing-related parameters and / or other suitable parameters to one or more healthcare providers and / or other suitable entities for use in determining a diagnosis, such as in combination with other data), and / or any suitable steps associated with a diagnosis. For example, assisting a diagnosis may include providing (e.g., to a user; to a healthcare provider, etc.) a quantification of DNA methylation in a sample from a patient adapted for use in determining a diagnostic result for an assay associated with liquid biopsy. In one example, quantifying DNA methylation may include quantifying the number of methylated molecules in a sample (e.g., the absolute number of methylated molecules in a sample) to facilitate a diagnosis associated with liquid biopsy.

[0205] In some embodiments, facilitating a diagnosis (e.g., cancer diagnosis) can include facilitating a diagnosis based on the amount of methylated molecules in the sample (see 122 in FIG. 1A). For example, determining the absolute number of methylated molecules can include determining the absolute number of methylated molecules associated with a DNA sequence based on dividing the total read count for the methylated molecules by the average QCT sequencing depth, where determining the absolute number of methylated molecules can include determining the absolute number of methylated molecules associated with a second DNA sequence not predicted to have methylation based on dividing the total read count for the methylated molecules by the average QCT sequencing depth, where facilitating a diagnosis (e.g., cancer diagnosis, etc.) can include facilitating a diagnosis (e.g., cancer diagnosis, etc.) of the methylated molecules based on the comparison.

[0206] In some embodiments, facilitating a diagnosis (e.g., cancer diagnosis) can include facilitating a diagnosis based on the amount of methylated molecules in the sample (see 122 in FIG. 1A). For example, determining the absolute number of methylated molecules can include determining the absolute number of methylated molecules based on dividing the total read count for the methylated molecules by the average QCT sequencing depth, where determining the absolute count of methylated molecules can include determining the absolute count of methylated molecules in a sample that is not expected to have a methylated sequence at the same corresponding DNA sequence based on dividing the total read count for the methylated molecules by the average QCT sequencing depth, where facilitating a diagnosis (e.g., cancer diagnosis) can include facilitating a diagnosis (e.g., cancer diagnosis) of the methylated molecules based on the comparison.

[0207] In one embodiment, the disclosure features a method (FIG. 1A) that may additionally or alternatively include facilitating treatment 124. For example, the methods described herein, including quantifying DNA methylation, determining a DNA methylation profile, and / or quantifying the prevalence of somatic mutations, may be used to assess whether a cancer treatment has been effective for an individual patient. Results provided by the methods described herein may help inform an oncologist's future treatment decisions. In one example, if a method of quantifying DNA methylation in a sample results in the discovery of increased DNA methylation in a subject, the subject's oncologist may use the quantified increase in DNA methylation, in part, as a basis for adapting the subject's treatment regimen. In another example, if a method of determining a methylation profile in a sample results in the discovery of an alteration in the DNA methylation profile indicative of an alteration in the subject's cancer, the subject's oncologist may use the alteration in the DNA methylation profile, in part, as a basis for adapting the subject's treatment regimen. In another example, if a method for quantifying the prevalence of somatic mutations in a sample results in the discovery of somatic mutations indicative of cancer and / or indicative of cancerous changes, the subject's oncologist may use the quantification of the prevalence of somatic mutations, in part, as a basis for adapting the subject's treatment regimen.

[0208] In one embodiment, the present disclosure features a method that may additionally or alternatively include recommending further evaluation of a subject based on what may function to assist, determine, provide, and / or otherwise facilitate diagnosis or treatment. For example, the methods described herein, including quantifying DNA methylation, determining a DNA methylation profile, and / or quantifying the prevalence of somatic mutations, may be used as a basis for performing or recommending further evaluation of a subject. In some examples, performing further evaluation includes subjecting the subject to a treatment selection assay. In some examples, performing further evaluation includes subjecting the subject to a genomic profiling assay (i.e., an assay that evaluates the mutation profile of the subject).

[0209] In one embodiment, the quantified DNA methylation in a sample collected from a subject, quantified according to the methods described herein, can be incorporated into a clinical recommendation for the subject. The clinical recommendation can include a plan for further testing or treatment. For example, based on the quantified DNA methylation in the sample and the clinical significance of the DNA methylation, a plan for further testing or treatment can be developed. In some cases, this can include additional testing to confirm the diagnosis, monitoring for disease progression or recurrence, or prescribing a targeted therapy specific to the subject's DNA methylation profile.

[0210] 6.7. Quantifying DNA methylation in samples The disclosure also features a method for quantifying DNA methylation in a sample that contains a DNA sequence.

[0211] In some embodiments, the method includes treating the sample to encode the presence or absence of DNA methylation in the DNA sequence, in other embodiments, the sample has already been treated to encode the presence or absence of DNA methylation in the DNA sequence.

[0212] In some embodiments, the sample (e.g., a sample containing a mixture of methylated and unmethylated DNA sequences) may include at least 10 target loci (e.g., loci to be amplified and evaluated to quantify methylation) from the DNA sequences. In some examples, each locus of the at least 10 target loci is selected to predict increased DNA methylation of each locus in cancerous tissue compared to normal tissue. The sample may also include normalization loci (e.g., loci that are highly methylated in both cancerous tissue and normal tissue (e.g., non-cancerous tissue)). In some examples, the normalization loci are included in the target loci (e.g., the at least 10 target loci). The method also includes adding to the sample a set of synthetic molecules that include a target-associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule (the endogenous target molecule includes at least one of the target loci) that has at least one of the target loci, and a mutant region having a sequence that does not match the sequence region of the endogenous target molecule. The method further comprises creating a co-amplification mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target loci from DNA sequences. In some embodiments, the co-amplification mixture comprises at least one normalization locus predicted to have hypermethylation in cfDNA across both cancerous and non-cancerous tissues. The method comprises sequencing the co-amplification mixture to generate sequence reads, and determining the number of sequence reads that are methylated sequence reads. The method additionally comprises quantifying (determining) the number (e.g., absolute number) of methylated molecules (e.g., for each locus) in the sample based on the number of methylated reads (e.g., from each locus) from the sample and the number of reads from the set of synthetic molecules. In one embodiment, the method comprises quantifying (determining) the number (e.g., absolute number) of methylated molecules in the sample for each target locus based on the number of methylated reads from the sample for each target locus and the number of reads from the set of synthetic molecules.In some embodiments, the method further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the sample.

[0213] In another embodiment, the present disclosure also features a method for quantifying DNA methylation in a sample containing cell-free DNA (cfDNA) sequences. In some embodiments, the quantified DNA methylation is a Tumor Methylation Score. The method includes treating a sample (e.g., a sample taken from a drawn blood sample, including plasma with cfDNA and buffy coat with gDNA) to encode the presence or absence of DNA methylation in a DNA sequence (e.g., a cfDNA and / or genomic DNA (gDNA) sequence). The sample (e.g., a sample containing a mixture of methylated and unmethylated DNA sequences) may include at least 10 target loci (e.g., loci to be amplified and evaluated for the presence of methylation) from the DNA sequence. In some examples, each locus of the at least 10 target loci is selected because it predicts increased DNA methylation at each locus in cancerous tissue compared to normal tissue (e.g., non-cancerous tissue). The sample may also include normalization loci (e.g., loci that are highly methylated in both cancerous and normal (e.g., non-cancerous) tissues). A set of synthetic molecules is added to the sample, where the set of molecules includes a target-associated region having a nucleotide sequence that matches a target sequence region of an endogenous target and a mutated region having a nucleotide sequence that does not match a sequence region of an endogenous target molecule. The sample containing the synthetic molecules is amplified, thereby creating a co-amplification mixture that includes an amplified set of synthetic molecules and an amplified set of at least 10 target loci from a DNA sequence, and optionally includes normalization loci. The method includes sequencing the co-amplification mixture at a read depth of at least one read sequence per molecule in the sample to generate sequence reads, and determining the number of sequence reads that are methylated sequence reads.The method also includes processing the methylated sequence reads using one or more of the following: removing selected hypermethylated target loci (e.g., removing target loci with a total number of methylated molecules in the buffy coat above a threshold); and / or subtracting background methylation (e.g., subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA (e.g., on a locus basis). The method further includes quantifying (determining) the number (e.g., absolute number) of methylated molecules in the sample based on the number of methylated reads from the sample and the number of sequence reads from the set of synthetic molecules. In one embodiment, the method includes quantifying (determining) the number (e.g., absolute number) of methylated molecules in the sample for each target locus based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules for each target locus. In some embodiments, the method also includes counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the DNA sample. In some embodiments, the sample is taken from a blood draw that includes plasma and a buffy coat, the plasma including cfDNA sequences and the buffy coat including gDNA.

[0214] In one embodiment, the method includes determining (quantifying) the number of methylated molecules in cfDNA and / or gDNA using QCT molecule counting technology (e.g., as described in U.S. Patent Application Publication No. 2019 / 0211395A1, which is incorporated by reference) at at least 10 target loci, where each locus of the at least 10 target loci is selected because of a predicted increase in DNA methylation at each locus in cancerous tissue compared to normal (e.g., non-cancerous) tissue. In such embodiments, the number of methylated molecules for cfDNA and the number of methylated molecules for gDNA are each quantified (e.g., in separate workflows) by: treating a sample to encode the presence or absence of DNA methylation in the DNA sequence, the sample comprising at least 10 target loci from the DNA sequence; adding a set of synthetic molecules to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target and a mutant region having a nucleotide sequence that does not match a sequence region of an endogenous target molecule; creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci (or a portion thereof) from the DNA sequence; sequencing the co-amplification mixture to generate sequence reads (e.g., at a read depth of at least 1 sequence read per molecule in the sample); determining the number of sequence reads that are methylated sequence reads; and quantifying (determining) the number (e.g., absolute number) of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules.

[0215] In some embodiments, quantifying DNA methylation in a sample includes quantifying the number of methylated molecules for cfDNA and the number of methylated molecules for gDNA, which may be quantified in the same workflow. In one embodiment, the methylated molecules for cfDNA and the methylated molecules for gDNA may be indexed (e.g., with defined index sequences) to allow discrimination between sources of methylated molecules (e.g., cfDNA vs. gDNA), thereby allowing cfDNA and gDNA to be amplified and / or sequenced in the same reaction (e.g., multiplexing).

[0216] In one embodiment, quantifying DNA methylation in a sample containing gDNA sequences from a buffy coat comprises treating the sample containing methylated and unmethylated gDNA sequences from a buffy coat to code the presence or absence of DNA methylation in the gDNA sequences. In such an example, the sample comprises at least 10 target loci from the gDNA sequences, where each locus of the at least 10 target loci is selected to predict increased DNA methylation of each locus in cancerous tissue compared to normal tissue, and is the same target locus analyzed in cfDNA extracted from plasma. The method comprises adding a set of synthetic molecules to the sample, the set of molecules comprising a target-associated region having a nucleotide sequence that matches a target sequence region of an endogenous target, and a mutated region having a nucleotide sequence that does not match a sequence region of an endogenous target molecule. The method also comprises creating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from a DNA sequence; and sequencing the co-amplification mixture to create sequence reads. The method further includes determining the number of sequence reads that are methylated sequence reads; and quantifying (determining) the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules. Optionally, the method includes tabulating the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in a sample containing DNA sequences from the buffy coat.

[0217] In some embodiments, the method for quantifying DNA methylation in a sample includes the addition of a spike-in of known sequence and amount to the sample prior to a treatment step. The spike-in includes a known sequence with unmethylated cytosine that is converted to uracil when subjected to a treatment step. The initial number of spike-in molecules and the percentage of cytosine bases converted to thymine / uracil bases can be calculated. This allows the calculation of bisulfite conversion yield and conversion efficiency, thereby determining bisulfite conversion QC metrics and therefore determining whether a sample fails the bisulfite conversion step (i.e., the step of treating a sample containing a DNA sequence to code for the presence or absence of DNA methylation in the DNA sequence).

[0218] In some embodiments, the method for quantifying DNA methylation in a sample includes processing the methylated sequence reads by one or more of the following: removing selected hypermethylated target loci (e.g., removing target loci with a total number of methylated molecules in the buffy coat above a threshold); or subtracting background methylation (e.g., subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in cfDNA, which may be performed on a locus-by-loci basis). In some embodiments, the processing is performed prior to quantifying the number (e.g., absolute number) of methylated molecules in the sample. In some embodiments, filtering out hypermethylated target loci, subtracting background methylation, or both, is performed prior to quantifying the number of methylated molecules in the sample.

[0219] In some embodiments, the method for quantifying DNA methylation in a sample further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the sample. The counting may include counting target loci that contain fewer than a threshold amount of methylated molecules in the buffy coat sample.

[0220] In some embodiments, a method for quantifying DNA methylation in a sample comprises tabulating the number of methylated molecules across at least 10 target loci (or a portion thereof, e.g., at least two target loci); and normalizing the tabulated number of methylated molecules from all or a portion of the at least 10 target loci by the methylated molecules for at least one normalization locus. In some embodiments, this results in quantification of DNA methylation (e.g., Tumor Methylation Score) that represents the normalized sum of methylated molecules (e.g., at at least 10 target loci that are hypermethylated in the sample).

[0221] In some embodiments, counting the number of methylated molecules across at least 10 target loci (or a portion thereof, e.g., at least two target loci) comprises counting at least one target locus that demonstrates hypermethylation in the sample containing the DNA sequences.

[0222] In one embodiment, the method further comprises determining the cancer tissue of origin by quantifying methylation in the DNA sample, where determining the cancer tissue of origin comprises determining the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules at each locus, and determining the tissue of origin based on the abundance of methylated molecules across the loci.

[0223] 6.8. Quantifying tumor DNA in samples The present disclosure also features a method for quantifying the amount of tumor DNA in a sample containing DNA sequences. The method includes adding a set of synthetic molecules to a sample (e.g., the sample is collected from a blood sample containing plasma with cfDNA and buffy coat with genomic DNA (gDNA)), and the set of molecules includes a target-associated region with a nucleotide sequence that matches the target sequence region of an endogenous target (e.g., containing at least one target locus) and a mutation region with a nucleotide sequence that does not match the sequence region of an endogenous target molecule. The method includes: creating a co-amplification mixture that includes an amplified set of synthetic molecules, an amplified set of at least 10 target loci from a DNA sequence, and optionally one or more amplified normalization loci; sequencing the co-amplification mixture at a read depth of at least 1 read sequence per molecule in the sample to generate sequence reads; determining the number of molecules containing somatic mutations in the sample based on the number of reads from the sample containing somatic mutations and the number of reads from the set of synthetic molecules.

[0224] 6.9. Determining methylation profiles in subjects over time The disclosure also features a method for determining a DNA methylation profile in a subject over time. Determining a methylation profile in a subject over time using serial quantification of DNA methylation allows for an early indication of cancer response to treatment or progression (or recurrence). In light of what has been described herein, the methods of the present invention uniquely allow for this type of serial quantification of DNA methylation.

[0225] In one embodiment, a method for determining a DNA methylation profile in a subject over time includes quantifying DNA methylation at a first time point and a second time point; and determining whether methylation (which is a surrogate tumor measurement) has changed over time (e.g., increased or decreased compared to the measurement at the first time point). In some examples, the change in DNA methylation (i.e., Tumor Methylation Score) exceeds a significant threshold and is reported as an increase or decrease. In some embodiments, a method for determining a DNA methylation profile in a subject over time can distinguish a 0.2 percentage point change (or lower (e.g., 0.1, 0.05, 0.01, 0.05, 0.001 percent change) in DNA methylation (e.g., number of methylated molecules) with a separation of 3 standard deviations.

[0226] In some embodiments, a method for determining a DNA methylation profile in a subject over time includes: i) treating a sample isolated from the subject to encode the presence or absence of DNA methylation in a DNA sequence, the sample comprising at least 10 target loci from the DNA sequence, each locus of the at least 10 target loci being selected based on a predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue, the sample optionally comprising a normalization locus (e.g., a locus that is highly methylated in both cancerous tissue and normal tissue (e.g., non-cancerous tissue)); ii) adding a set of synthetic molecules (e.g., QCT molecules) to the sample, the set of molecules comprising a target associated region having a nucleotide sequence that matches a target sequence region of an endogenous target molecule comprising at least one of the target loci, and a mutant region having a nucleotide sequence that does not match the sequence region of the endogenous target molecule. iii) generating a co-amplification mixture comprising the amplified set of synthetic molecules and the amplified set of at least 10 target loci from the DNA sequence; iv) sequencing the co-amplification mixture at a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads; v) determining the number of sequence reads that are methylated sequence reads; vi) quantifying (determining) the number of methylated molecules (e.g., for each target locus) in the sample based on the number of methylated sequence reads (e.g., from each target locus) in the sample and the number of reads from the set of synthetic molecules; repeating steps i)-vi) at a second timepoint; and determining a methylation pattern in the subject based on the number of methylated molecules (e.g., for each target locus) in the sample at the first timepoint and the number of methylated molecules (e.g., for each target locus) in the sample at the second timepoint. In some embodiments, the sample has already been treated to encode the presence or absence of DNA methylation in the DNA sequence, and therefore step i) is not performed.

[0227] In some embodiments, determining a methylation profile in the subject at the first time point and the second time point identifies a change in the methylation profile.

[0228] In one embodiment, the method for determining a DNA methylation profile in a subject over time comprises repeating steps i)-vi) at a third timepoint; and determining the methylation profile in the subject based on the number of methylated molecules (e.g., for each target locus) in the sample at the third timepoint.

[0229] In some embodiments, determining a methylation profile in a subject at a first timepoint, a second timepoint, and a third timepoint identifies a change in the methylation profile between the first timepoint and the second timepoint, between the second timepoint and the third timepoint, between the first timepoint and the third timepoint, or a combination thereof.

[0230] In one embodiment, a method for determining a DNA methylation profile in a subject over time comprises quantifying the number of methylated molecules in both cfDNA (plasma) and gDNA (buffy coat) utilizing QCT molecular counting technology (e.g., as described in U.S. Patent Application Publication No. 2019 / 0211395A1, which is incorporated by reference) at at least 10 target loci, where each locus of the at least 10 target loci is selected because of a predicted increase in DNA methylation at each locus in cancerous tissue compared to normal (e.g., non-cancerous) tissue. In such an embodiment, the number of methylated molecules for cfDNA is quantified as described herein (see Section 6.7) and the number of methylated molecules for gDNA is quantified as described herein (see Section 6.7), respectively.

[0231] In some embodiments, a method for determining a DNA methylation profile in a subject over time includes the addition of a spike-in of known sequence and amount to a sample prior to a treatment step at each time point. The spike-in includes a known sequence having an unmethylated cytosine that is converted to uracil upon being subjected to the treatment step, allowing for calculation of bisulfite conversion yield and conversion efficiency, thereby establishing bisulfite conversion QC metrics that can be used to determine if a sample will fail the bisulfite conversion step.

[0232] In some embodiments, the method for determining a DNA methylation profile in a subject over time includes processing the methylated sequence reads by one or more of the following: removing selected hypermethylated target loci (e.g., removing target loci with a total number of methylated molecules in the buffy coat above a threshold); or subtracting background methylation (e.g., subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA, which may be performed on a locus-by-loci basis). In some embodiments, the processing is performed prior to quantifying the number (e.g., absolute number) of methylated molecules in the sample. In some embodiments, filtering out hypermethylated target loci, subtracting background methylation, or both, is performed prior to quantifying the number of methylated molecules in the sample.

[0233] In some embodiments, the method for quantifying DNA methylation in a sample comprises counting the number of methylated molecules across at least two of at least ten target loci to quantify DNA methylation in the sample. The counting may comprise counting target loci that contain less than a threshold amount of methylated molecules in the buffy coat sample.

[0234] In some embodiments, a method for determining a DNA methylation profile in a subject includes tabulating the number of methylated molecules across at least 10 target loci (or a portion thereof, e.g., at least two target loci); and normalizing the tabulated number of methylated molecules from all or a portion of the at least 10 target loci by the methylated molecules for at least one normalization locus. In some embodiments, the tabulating and normalizing results in a quantification of DNA methylation (e.g., Tumor Methylation Score) that represents a normalized sum of methylated molecules (e.g., at at least 10 target loci that are hypermethylated in a sample at a given time point). Comparison between time points allows for the determination of a DNA methylation profile (over time) in the subject. In some examples, the DNA methylation profile identifies a change in the methylation profile from a first time point to a second time point, from a second time point to a third time point, or from a first time point to a third time point.

[0235] In some embodiments, the method for determining a DNA methylation profile in a subject over time includes assigning a change in the methylation profile to a metric of increase, decrease, or no change based on a comparison to a significance threshold. The significance threshold can be predetermined or dynamically calculated. For example, but not limited to, the method for determining a DNA methylation profile in a subject over time indicates a change because the DNA methylation profile (e.g., Tumor Methylation Score™) exceeds a predetermined significance threshold of about 15% compared to a previous time point, and the change is an increase. For example, but not limited to, the method for determining a DNA methylation profile in a subject over time indicates a change because the DNA methylation profile (e.g., Tumor Methylation Score™) exceeds a predetermined significance threshold of about 15% compared to a previous time point, and the change is a decrease. In some embodiments, the predetermined significance threshold that must be exceeded for a change in DNA methylation profile (e.g., Tumor Methylation Score) to be considered an increase or decrease is 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2% or 1% compared to a previous timepoint.

[0236] In some embodiments, the change in the methylation pattern indicates a tumor change in the subject. The tumor change includes a change in tumor size, a change in the prevalence of somatic mutations associated with the tumor, the presence of new somatic mutations associated with the tumor, a chromosomal abnormality associated with the tumor, or the tumor being resistant to treatment. In some embodiments, an increase in the DNA methylation profile (e.g., Tumor Methylation Score™) of at least 1.5-fold (e.g., 2-fold, 3-fold, 4-fold, or 5-fold) compared to the DNA methylation profile at a previous time point indicates a tumor change. In some embodiments, an increase in the DNA methylation profile (e.g., Tumor Methylation Score™) of at least 2-fold compared to the DNA methylation profile at a previous time point indicates a tumor change. In some embodiments, the DNA methylation profile identifies an alteration in a tumor (e.g., a tumor associated with a sample) if the DNA methylation in the sample includes at least 10% (e.g., at least 15%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, or at least 90%) of the DNA methylation as occurring from newly methylated target loci (e.g., target loci not previously found to be methylated at a prior timepoint). In some embodiments, the DNA methylation profile identifies an alteration in a tumor (e.g., a tumor associated with a sample) if the DNA methylation in the sample includes at least 40% of the DNA methylation in the sample (e.g., methylation at target loci in cfDNA) as occurring from newly methylated target loci (e.g., target loci not previously found to be methylated at a prior timepoint).In some embodiments, an increase in the DNA methylation profile (e.g., Tumor Methylation Score™) of at least 2-fold compared to the DNA methylation profile at the previous timepoint and at least 40% of the DNA methylation in the sample (e.g., 40% of the Tumor Methylation Score™) is from newly methylated target loci (e.g., target loci not previously found to be methylated at the previous timepoint) indicates an alteration in the tumor (e.g., a tumor associated with the sample).

[0237] In one embodiment, the DNA methylation profile determined using the methods described herein may be incorporated into a clinical recommendation. The clinical recommendation may include a plan for further testing or treatment. For example, based on the quantified DNA methylation in the sample and the clinical significance of the DNA methylation, a plan for further testing or treatment may be developed. In some cases, this may include additional testing to confirm the diagnosis, monitoring for disease progression or recurrence, or prescribing targeted therapy specific to the subject's DNA methylation profile.

[0238] 6.10. Quantifying the Prevalence of Somatic Mutations in a Sample The disclosure also features a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample, where the indication of confidence in the presence or absence of a somatic mutation in a sample is assigned a "true call" or a "no call" based in part on quantification of DNA methylation in the same sample in which the somatic mutation was determined to be present or absent. In one example, concordance between the presence or absence of a somatic mutation and DNA methylation indicates that the methylation (i.e., the methylation analyzed and identified according to the methods described herein) can be used to inform a confidence level that the somatic mutation is present or absent, which can then be used to inform clinical recommendations to the subject.

[0239] In one embodiment, a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample containing a DNA sequence comprises one or more of the following steps: In some embodiments, the method comprises a step for determining the presence or absence of a somatic mutation in the sample. In other embodiments, the method does not comprise a step for determining the presence or absence of a somatic mutation in the sample, since the treatment step has been performed prior to utilizing the methods provided herein. In such an example, the method comprises quantifying DNA methylation in the sample used to determine the presence or absence of a somatic mutation, but the determination of the presence or absence of a somatic mutation is performed prior to utilizing the methods described herein.

[0240] In some embodiments, a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample containing a DNA sequence comprises quantifying DNA methylation in the sample comprising steps as described herein (see, e.g., Section 6.7). In some examples, the method optionally comprises treating the sample to encode the presence or absence of DNA methylation in the DNA sequence, the sample comprising at least 10 target loci from the DNA sequence, each locus of the at least 10 target loci being selected based on a predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue, and the sample optionally comprising normalization loci (e.g., loci highly methylated in both cancerous tissue and normal (e.g., non-cancerous) tissue). In some embodiments, the sample is treated to encode the presence or absence of DNA methylation in the DNA sequence prior to performing the methods described herein. In some embodiments, the method includes adding a set of synthetic molecules to a sample, the set of molecules including a target-associated region having a nucleotide sequence that matches a nucleotide sequence corresponding to an endogenous target and a mutated region having a nucleotide sequence that does not match a sequence region of the endogenous target molecule. The method also includes creating a co-amplification mixture including an amplified set of synthetic molecules and an amplified set of at least 10 target loci from a DNA sequence; and sequencing the co-amplification mixture (e.g., at a read depth of at least 1 sequence read per molecule in the sample to generate sequence reads). The method further includes determining the number of sequence reads that are methylated sequence reads; and quantifying the number of methylated molecules in the sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules.

[0241] Following determination of the presence or absence of a somatic mutation and the amount of DNA methylation in the sample, the method returns an indication of confidence (i.e., confidence in determining the presence or absence of a somatic mutation) as a result of a true call regarding the presence or absence of a somatic mutation if the number of methylated molecules in the sample exceeds a predetermined or dynamically calculated threshold, or as a result of a no call regarding the presence or absence of a somatic mutation if the number of methylated molecules in the sample is equal to or less than a predetermined or dynamically calculated threshold, whereby the true call identifies confidence in determining the presence or absence of a somatic mutation.

[0242] In one embodiment, a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample comprises quantifying the number of methylated molecules in both cfDNA (plasma) and gDNA (buffy coat) utilizing QCT molecular counting technology (e.g., as described in U.S. Patent Application Publication No. 2019 / 0211395A1, which is incorporated by reference) at at least 10 target loci, each locus of the at least 10 target loci being selected for predicting increased DNA methylation at each locus in cancerous tissue compared to normal (e.g., non-cancerous) tissue. In such an embodiment, the number of methylated molecules for cfDNA is quantified as described herein (see Section 6.7), and the number of methylated molecules for gDNA is quantified as described herein (see Section 6.7), respectively. Quantifying the number of methylated molecules provides a mechanism for assigning a determination of the presence or absence of a somatic mutation to a true call (e.g., a correct call) or a no call (e.g., insufficient DNA to assign sufficient confidence to the abundance measurement). For example, quantifying DNA methylation in the same sample used to determine the presence or absence of a somatic mutation returns an indication of confidence for the somatic mutation determination. If the number of methylated molecules in the sample exceeds a predetermined or dynamically calculated threshold, a true call result for the presence or absence of a somatic mutation is returned. If the number of methylated molecules in the sample is equal to or less than a predetermined or dynamically calculated threshold, a no call result for the presence or absence of a somatic mutation is returned. A no call indicates that the determination of the presence or prevalence of a somatic mutation cannot be assigned sufficient confidence to say that the measurement is correct.

[0243] In some embodiments, a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample includes the addition of a spike-in of known sequence and amount to the sample prior to a treatment step. The spike-in includes a known sequence having an unmethylated cytosine that is converted to uracil upon being subjected to the treatment step. As previously discussed, this allows for the calculation of bisulfite conversion yield and conversion efficiency, thereby establishing bisulfite conversion QC metrics that can be used to determine whether a sample will fail the bisulfite conversion step.

[0244] In some embodiments, the method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample includes processing the methylated sequence reads by one or more of the following: removing selected hypermethylated target loci (e.g., removing target loci with a total number of methylated molecules in the buffy coat above a threshold); or subtracting background methylation (e.g., subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA, which may be performed on a locus-by-loci basis). In some embodiments, the processing is performed prior to quantifying the number (e.g., absolute number) of methylated molecules in the sample. In such examples, filtering hypermethylated loci, subtracting background methylation, or both are performed prior to quantifying the absolute number of methylated molecules in the sample.

[0245] In some embodiments, if a true call is returned and the presence of somatic mutations indicates the presence of cancer, the method includes performing or repeating a treatment selection assay on the subject (e.g., the treatment selection assay includes genomic profiling to detect novel somatic mutations, the prevalence of somatic mutations, or both). In some embodiments, if a no call is returned, the method is repeated with a different sample, at least in part, taken from the same subject.

[0246] In some embodiments, the method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample further comprises counting the number of methylated molecules across at least two of the at least ten target loci to quantify DNA methylation in the sample. The counting may include counting target loci that contain less than a threshold amount of methylated molecules in the buffy coat.

[0247] In some embodiments, a method for determining an indication of confidence in the presence or absence of a somatic mutation in a sample includes tabulating a number of methylated molecules across at least 10 target loci (or a portion thereof, e.g., at least two target loci); and normalizing the tabulated number of methylated molecules from all or a portion of the at least 10 target loci by a methylated molecule for at least one normalization locus.

[0248] 6.11.Kit In another aspect, the disclosure features a kit for quantifying DNA methylation in a sample utilizing the methods described herein. In some embodiments, a kit for quantifying DNA methylation in a sample utilizing the methods described herein. In some embodiments, the kit may include one or more of the following: conversion reagents required to encode the presence or absence of DNA methylation in a DNA sequence, synthetic molecules (e.g., QCT molecules) that are added to the sample at various points in sample processing, reagents required for amplification of the DNA sample, and reagents required to prepare a sequencing library, such as those required for indexing PCR.

[0249] 7. Working Example 7.1. Materials and Methods for Examples 1 to 10 7.1.1. Selection of Tumor Hypermethylation Targets The Cancer Genome Atlas (TCGA) was searched for subjects with human methylation data for both tumor and normal tissues of the same tissue type. Data from TCGA were retrieved using the Infinium HumanMethylation450 BeadChip (Illumina, San Diego, CA) to provide beta values ​​representing methylation rates at specific CpG positions in the genome. In addition, methylation data from leukocytes retrieved from patients of similar age (mean=63.9 years, SD=13.3 years) to the cancer patients were obtained from GEO accession GSE40279 (see Hannum, G. et al., (2013). Molecular Cell, 49(2), 359-367. doi.org / 10.1016 / j.molcel.2012.10.016).

[0250] Tumor hypermethylation was calculated by subtracting normal tissue beta from tumor beta at each CpG site. To avoid selecting CpG sites with spurious hypermethylation, the average hypermethylation was calculated for each CpG island, and CpG islands were ranked according to the level of hypermethylation. In addition, to minimize background signal from buffy coat contribution to cfDNA, CpG islands were filtered with an average leukocyte beta < 0.2. From each selected CpG island, multiple CpG positions were selected for primer design.

[0251] Lung cancer-specific assays were designed using TCGA data from subjects with lung adenocarcinoma and lung squamous cell carcinoma. Pan-cancer assays were also designed using TCGA data from many cancer types (e.g., lung adenocarcinoma, lung squamous cell carcinoma, colon adenocarcinoma, invasive breast carcinoma, pancreatic adenocarcinoma, hepatocellular carcinoma, bladder urothelial carcinoma, esophageal carcinoma, renal cell carcinoma, papillary renal cell carcinoma, prostate cancer, thyroid cancer, and endometrial cancer).

[0252] 7.1.2. Selection of Highly Methylated Targets In addition to selecting targets with hypermethylation in tumors relative to normal tissues, control loci that were highly methylated in both buffy coat, tumor, and normal tissues were also selected. Methylation data were acquired and analyzed similarly to the hypermethylated targets, but ranked according to the height of the leukocyte methylation signal and filtered for beta > 0.9 for leukocytes, tumor tissues, and normal tissues from six cancer types (lung adenocarcinoma, lung squamous cell carcinoma, colon adenocarcinoma, breast invasive carcinoma, pancreatic adenocarcinoma, and hepatocellular carcinoma).

[0253] 7.1.3. Primer and QCT Design Primer3 was used to design primer pairs targeting the genomic location of the selected CpG positions using the in silico bisulfite converted reference genome. This step was performed assuming full methylation of the reference hg19 human genome, converting all non-CpG Cs to Ts. The ideal annealing temperature was set to 60°C.

[0254] A single stranded QCT was designed for each amplicon. The predicted amplicon sequence for each primer pair was determined using Bowtie and the converted human genome, and 17 bases of flanking genomic sequence were added to both the 5' and 3' ends. Eleven bases in the insertion region of the QCT were replaced with N, allowing a small number of unique QCTs to be added to each PCR reaction.

[0255] 7.1.4. Additional Amplicons In addition to tumor hypermethylation and hypermethylation genomic targets, additional amplicons were included in the assay. Fifteen genotyping amplicons were designed targeting commonly found A>G or G>A SNPs according to NCBI's Single Nucleotide Polymorphism Database (dbSNP). Genotyping at these locations can be used to check sample replacements that may be performed during laboratory processing or in the clinic. Three amplicons targeting genomic locations on the Y chromosome were included to determine the patient's gender.

[0256] Two artificial amplicons were included to check the bisulfite conversion yield and conversion efficiency. These synthetic oligos are spiked in a certain amount into cfDNA and gDNA samples just before bisulfite conversion. These oligos are then amplified in a multiplex reaction together with all other amplicons. During analysis, the initial number of spike-in molecules can be calculated along with the percentage of cytosine bases converted to thymidine bases (oligos were synthesized without any methylation). In the no-template control of the multiplex PCR, the same amount of bisulfite conversion spike-in is added after bisulfite conversion, allowing the estimation of the total bisulfite conversion yield. Calculating the bisulfite conversion yield and conversion efficiency allows the determination of bisulfite conversion QC metrics and therefore allows the determination of whether a sample has failed the bisulfite conversion step of this process.

[0257] 7.1.5. Source of specimens To obtain multiple tumor specimens spanning a range of cancer types, archived snap-frozen tumors and buffy coats from the same subjects were obtained from Spectrum Health (Grand Rapids, MI).

[0258] To test whether the assay could detect methylation changes that coincided with clinical outcomes, samples were collected from cancer patients both retrospectively and prospectively. Retrospective samples consisting of archived plasma and buffy coats were obtained through collaborations with the University of California at San Diego and the University of Florida. Prospective sample collection was performed through contract research organizations and their partner clinics, and patients who had been diagnosed with cancer but had not started treatment were enrolled. Blood was collected in Streck tubes at subsequent time points pre- and post-treatment. Clinical outcomes were provided when available.

[0259] To test how the assay works in healthy individuals, blood was collected from healthy volunteers at various time points.

[0260] 7.1.6. Preparation of specimens Blood collected in EDTA tubes was spun down within 1 hour of collection and plasma and buffy coat were isolated. Blood collected in Streck tubes was allowed to sit overnight before centrifugation and isolation of plasma and buffy coat. Plasma volume was recorded for analytical normalization.

[0261] cfDNA was extracted using the QIAamp Circulating Nucleic Acid Kit (Qiagen), and gDNA was extracted from tumor and buffy coat samples using the DNeasy Blood & Tissue Kit (Qiagen).

[0262] By mixing tumor genomic DNA (gDNA) at various tumor ratios into buffy coat gDNA, a design sample was created that mimicked a cell-free DNA (cfDNA) sample from a cancer patient.

[0263] 7.1.7. Bisulfite conversion and library preparation for next generation sequencing Samples were bisulfite converted using either the Diagenode Premium Bisulfite Kit (Cat. No. C02030030) or the Zymo EZ-96 DNA Methylation-Lightning MagPrep Kit (Cat. No. D5046). If the sample volume of either cfDNA or gDNA was larger than the recommended input volume, the sample was split in half and converted separately, then remixed. Enzymatic conversion was also tested. As a positive control for detecting methylation, universally methylated genomic DNA (Cat. No. S7821, Sigma-Aldrich) was diluted into buffy coat at various tumor ratios.

[0264] Primer mixes were created by pooling all primer pairs and iteratively removing and / or rebalancing the concentration of each primer pair to optimize for balanced coverage across target amplicons. QCT was diluted to 200 molecules per PCR reaction for each amplicon. A total of 113 target amplicons were included for the lung cancer assay and 679 target amplicons were included for the pan-cancer assay.

[0265] Multiplex PCR was performed on the bisulfite converted specimens using Q5U polymerase (NEB, Ipswich, MA). Indexing PCR was then performed using Q5 polymerase (NEB, Ipswich, MA) to sequence multiple samples with the same sequencing run performed on the double index. Pooled libraries were bead washed and loaded onto a NextSeq 2000 (Illumina, San Diego, CA) sequencing instrument using P3 100 cycle reagents for unidirectional sequencing and 5% PhiX.

[0266] 7.1.8. Counting the number of methylated molecules Fastq files were adaptor trimmed at the 3' end using BBDuk and then mapped to a custom genome composed of target hypermethylated and hypermethylated amplicons and QCT sequences using BWA-MEM.

[0267] For each amplicon, the reads that mapped to the target amplicon (e.g., target hypermethylated or hypermethylated amplicon) were binned based on sequence. The number of CpGs contained in each sequence was calculated, and each sequence was classified as belonging to a methylated read if the number of CpGs was equal to or greater than the maximum number of CpGs possible in the amplicon minus 1. The reads mapping to the corresponding QCT sequences were processed and binned separately based on the random N sequence of QCT. The average number of reads per QCT molecule was calculated, assuming that the sequence of each QCT molecule was unique to each reaction. The total number of methylated molecules for that amplicon can then be calculated by dividing the total number of methylated reads by the average reads per QCT molecule.

[0268] When measurements were obtained on paired cfDNA and buffy coat samples, background levels of methylation were reduced by subtracting the buffy coat methylation signal from the cfDNA methylation signal on a locus basis. Approximately 55% of cfDNA is estimated to originate from leukocytes based on methylation patterns (see Moss, J. et al., (2018), Nature Communications, 9(1), doi.org / 10.1038 / s41467-018-07466-6), suggesting that subtracting the entirety of buffy coat methylation is a conservative approach to remove background methylation signals. Any calculated negative numbers of molecules after background subtraction were capped at 0.

[0269] To perform comparable subtraction between buffy coat and cfDNA samples, samples must first be normalized to the input amount of genome equivalents. The average of methylated molecules at hypermethylated loci is used to estimate the total number of genome equivalents for the entire sample. The number of methylated molecules at each locus is normalized by the estimated genome equivalent of that sample. The normalized number of methylated molecules is then used for background subtraction.

[0270] 7.1.9. Filtering hypermethylated loci Based on testing of healthy subjects, certain hypermethylated loci were found to have significant methylation signals even after buffy coat subtraction. To minimize the amount of false cancer signals, certain loci were removed from the analysis. In some cases, this is referred to herein as "blacklisting." Any loci found to either consistently contribute moderate amounts of methylation or occasionally contribute large amounts of methylation in healthy subjects were added to a list of loci to ignore. These loci often contain large amounts of methylation in the buffy coat.

[0271] After filtering out hypermethylated genes, results from healthy subjects were used to establish a noise floor, below which methylation signals are not interpretable as they are equivalent to the amount found in healthy subjects.

[0272] 7.1.10. Calling methylation changes If data for two time points were available, a call was made as to whether there was an increase, decrease, or no change in the amount of methylated molecules. Calls were made by modeling the measurements at each time point as a normal distribution with the mean of the number of methylated molecules measured (normalized to the total number of input molecules). A standard deviation was assigned to the normal distribution based on the total number of methylated molecules, which can be done by imputing based on previous studies where the coefficient of variation was measured from the design sample (see, for example, group 3 in Example 7). The normal distribution from the first time point was subtracted from the second time point and normalized to the mean of the normal distribution of the first time point to create a normalized difference normal distribution. If the mean of this distribution was ≧−15% and the log2 likelihood ratio that the difference was ≧−15% compared to <−15% exceeded 3, an "increase" call was made. Similarly, if the mean of the distribution was ≦-15% and the log2 likelihood ratio for differences ≦-15% compared to >-15% was greater than 3, a "reduced" call was made. If the relative difference was not of a large enough magnitude or the statistical likelihood was not strong enough, a "no change" call was made. If either timepoint was below the noise floor, the mean of that timepoint was set to the noise floor, the standard deviation was still calculated based on the total number of methylated molecules, and a call was still made. If the methylation signals from both timepoints were below the noise floor, an "indeterminate" call was made. Any other appropriate statistical method, including but not limited to mean, median, standard deviation, likelihood ratio, expected maximum, statistical significance test, may be used to make the call.

[0273] 7.1.11. Methylation Profile To take advantage of the fact that the assay measures methylation at hundreds of hypermethylated positions, a methylation profile was determined at each timepoint. One way to construct a methylation profile was to group loci based on the timepoint at which they were first found to have an informative methylation signal, defined as more than two molecules after background subtraction. Grouping loci by the timepoint at which they were first found to have informative methylation signals revealed when and how much the methylation profile changed.

[0274] 7.2. Example 1. Lung cancer assay can detect significant methylation in 0.5% tumor incidence samples The lung cancer assay was evaluated for detection limits using designed samples. Without wishing to be bound by theory, designed samples are prepared to mimic clinical samples as closely as possible. In these experiments, designed samples contained sheared tumor gDNA that was sheared to an average fragment length of approximately 170 bp using an ultrasonicator to mimic the size distribution of cfDNA. A universally methylated designed sample was used as a positive control (e.g., instead of tumor DNA as a positive control). The positive control was sheared in the same manner as the designed lung cancer samples. A total of 1.35E9 sequencing reads were obtained, resulting in an average of 11E6 reads per sample. There was an average of 62 reads per QCT across all samples and amplicons. Each designed sample was generated at 5000 genome equivalents (ge).

[0275] Technical replicates of lung cancer specimens were tested with 0%, 0.5%, 1%, 2.5%, and 5% tumor fragments (see Figure 2A). The assay was able to distinguish 0.5% tumor fraction samples from 0% tumor fraction, suggesting a detection limit of at least 0.5%.

[0276] The lung cancer assay was also used to measure methylation in additional lung tumor specimens (see FIG. 2B), which shows that the assay was used to detect methylation in genomic samples from two additional lung cancer subjects. In these experiments, different PCR volumes and different thermal cycling protocols were utilized (see FIG. 2B).

[0277] Additional analysis utilizing the lung cancer assay included measuring the concordance value (CV) of the samples analyzed in FIG. 1A (FIG. 3). The CV was used to determine whether the amount of ctDNA and, accordingly, the tumor was truly increased or decreased, or whether the increase or decrease was a false positive or false negative. Without any additional post-processing, the CV of the methylation measurement was approximately 10% or less. This suggested that the methylation measurement generated utilizing the methods described herein confidently identifies a 30% signal change within 3 standard deviations. According to the RECIST guidelines, a partial response was defined by at least a 30% decrease in the longest diameter of the target tumor lesion, and progressive disease was defined by at least a 20% increase in the longest diameter of the target tumor lesion. In light of these RECIST guidelines, the assays described herein allow for the assessment of tumor response to treatment in the context of the current standard of care. Indeed, quantification of treatment response according to RECIST guidelines is limited by the selection of a limited number of identifiable tumors on imaging, whereas methylation assays provide a measure of all tumors present in the body that shed DNA into the bloodstream.

[0278] 7.3. Example 2. Background subtraction improves signal-to-background ratio Optimization has been performed to improve signal-to-background ratio.Without wishing to be bound by theory, improving signal-to-background ratio is particularly useful for detecting smaller tumor signals.For example, one medical application is for minimal residual disease (MRD) detection, where the difference between having any tumor signal and having zero signal can be the difference between recurrence and remission.

[0279] Despite filtering for targets with low methylation in buffy coat and normal tissues in the target selection process, some targets may still have a significant amount of background methylation. One approach to reduce the influence of background methylation was to mask the signals from target loci with buffy coat methylation above a certain threshold. Another parallel approach was to subtract the methylation signal measured in buffy coat from the methylation signal measured in cfDNA (plasma) on a locus-by-locus basis.

[0280] Figure 4 shows the total normalized methylation molecules for technical replicates of lung cancer design samples with various tumor rates. In particular, Figure 4 shows data from a lung cancer assay that included background subtraction (i.e., subtracting the methylation signal measured in the buffy coat from the methylation signal measured in the cfDNA) and masking of target genes with high background methylation (i.e., masking signals from target loci with buffy coat methylation above a certain threshold). For example, background methylation signals were calculated by utilizing the buffy coat methylation signal per locus normalized to the average hypermethylated locus methylation molecules in each buffy coat sample (e.g., the average of normalized loci) and averaging across all buffy coat samples. 55% of the background signal was subtracted from each cfDNA design sample methylation signal and normalized to the average hypermethylated locus methylation molecules in each design sample. Hypermethylated target loci with more than 10 methylation molecules per 1000 hypermethylated locus methylation molecules were masked. Any other suitable threshold value may be determined.

[0281] Masking the loci and subtracting the buffy coat methylation signal from the signal in the cfDNA reduced the background signal from an average of 277 methylated molecules to 111 methylated molecules in the 0% tumor rate specimens. More stringent masking of hypermethylated target loci was expected to further reduce the background signal.

[0282] Collectively, this data demonstrated that signal-to-background ratios can be improved by utilizing background subtraction (i.e., subtracting the methylation signal measured in the buffy coat from the methylation signal measured in cfDNA) in parallel with masking of target loci with high background methylation (i.e., masking signals from target loci with buffy coat methylation above a certain threshold).

[0283] 7.4. Example 3. Hypermethylation-based assays are robust to gDNA contamination, unlike somatic mutation-based approaches A common challenge with cfDNA assays is gDNA contamination. This can occur when buffy coat is inadvertently isolated with the plasma fraction during plasma isolation from centrifuged whole blood. gDNA contamination can have a detrimental effect on the accuracy of cfDNA tumor quantification, especially when utilizing mutant allele frequencies of somatic mutations for cfDNA tumor quantification. This is because additional buffy coat contributes to the denominator. As shown in Figure 5 (top panel), 1x gDNA contamination (e.g., contamination of twice the total number of molecules) reduced the mutant allele frequency by half. This can result in a false positive signal. If the positive signal correlates with treatment efficacy, such results have the potential to change medical decision-making.

[0284] This experiment showed that tumor hypermethylation detection was less vulnerable to gDNA contamination than the somatic mutation ctDNA assay, at least because targets were selected with low buffy coat methylation and because background subtraction methods were used. Given that the background methylation signal was zero in the buffy coat, the absolute number of methylated molecules was not affected at all by gDNA contamination (Figure 5; middle panel). Even if the buffy coat contributed a small amount of methylation background signal (Figure 5; lower panel), measuring buffy coat methylation and subtracting that signal from the cfDNA signal kept the total methylation signal within a 20% increase with 1× the amount of gDNA contamination. This minimized the impact of gDNA contamination on tumor methylation measurements and any potential downstream medical decisions made based on tumor methylation measurements.

[0285] In addition to target selection and background subtraction, there have been additional biochemical methods to minimize gDNA contamination, including bead purification to size-select DNA fragments in cfDNA samples, but these methods can also fail, so it is useful to have an orthogonal approach for robustness against gDNA contamination.

[0286] 7.5. Example 4. Day-to-day variability in buffy coat methylation in healthy subjects was comparable to intersubject variability The degree of day-to-day variation in methylation in healthy subjects was evaluated as part of the assay's ability to perform accurate time series measurements. Since a significant portion of the cfDNA signal originates from the buffy coat, buffy coat methylation profiles were measured from healthy subjects across different tubes, different tube types, different days, and different subjects from the same drawn blood. The data in Figure 6 shows the methylation profiles for buffy coats (i.e., 5000 g.e. of buffy coat) isolated from different tubes, different tube types, different days, and different subjects from the same drawn blood (see Table 1 for sample identifiers in Figure 6). TIFF2025509878000002.tif161170TIFF2025509878000003.tif246170TIFF2025509878000004.tif242170

[0287] For example, the two tubes from subject 5 obtained on the same day with the same tube type were very similar as expected (see Figure 6). The EDTA tube was very similar to both Streck tubes, so tube type did not cause significant differences. Data from the Streck tube collected on day 2 was similar to the data from the Streck tube collected on day 1, with notable differences at a few loci. Subject 10 had multiple methylated genes that were not methylated in any of the tubes from subject 5, so there were notable subject differences.

[0288] Figure 7 shows hierarchical clustering of methylation profiles from buffy coats isolated from different tubes of the same drawn blood, different tube types, different days, and different subjects. Clustering was performed based on the methylated molecules at each target hypermethylated locus and normalized to the amount of methylated molecules measured at the hypermethylated loci. Cluster distances were calculated using the L1 norm.

[0289] Hierarchical clustering revealed that tubes collected on the same day clustered closest to each other regardless of tube type (see FIG. 7). Examples of tubes clustering on the same day as each other included subject 1 day 1, subject 3 day 1, subject 3 day 2, and subject 10 day 1. However, tubes collected on different days from the same subject clustered much further apart as did tubes collected from different subjects. This suggested that background subtraction would be more optimal if performed using methylation profiles from buffy coats collected at each unique timepoint.

[0290] 7.6. Example 5. Pan-cancer assay detects methylation signals in multiple cancer types A pan-cancer assay was designed based on the methods described herein.

[0291] Figure 8 shows data from the methylation signal of the pan-cancer assay in engineered samples of several different cancer types. In these experiments, 5% tumor rate engineered samples were made by mixing 5% tumor gDNA by weight with 95% buffy coat from the same subject. 0% tumor rate engineered samples were pure buffy coat. Specimens were labeled with their cancer type (BRCA: invasive breast carcinoma, COAD: colon adenocarcinoma, LIHC: hepatocellular carcinoma, LUNG: lung carcinoma, PAAD: pancreatic adenocarcinoma). Universally methylated samples were included as positive controls for the methylation assay.

[0292] As shown in FIG. 8, the pan-cancer assay was able to detect methylation signals in 5% tumor rate engineered samples of multiple cancer types using the same chemistry in all samples. There was notable variability in methylation signals between specimens, which could be due to two causes. First, the methylation profile of the tumor may differ between tumors. This profile may or may not match the target positions included in the assay. In the assay described herein, the increased number of target positions may help minimize the impact of biological variability in methylation between tumors. The variability could also be due to technical limitations in engineered samples made with tumor specimens, since the tumors themselves contained an unknown proportion of normal tissue. Therefore, comparisons between specimens of measurements made with engineered samples of the same tumor rate must be made with caution. Despite the unknown exact tumor quality, the assay was still able to distinguish the 5% tumor rate engineered tumor samples from pure buffy coat samples of the same subjects.

[0293] Taken together, this data confirms that the methods described herein have potential for pan-cancer detection.

[0294] 7.7. Conclusions of Examples 1 to 5 This data demonstrated an innovative approach to cancer treatment monitoring by quantifying the absolute number of methylated molecules. Existing ctDNA assays have insufficient accuracy for treatment monitoring due to the very limited number of detectable variants and therefore the very low total number of variant molecules. Whole exome sequencing of tumor biopsies can be performed to identify additional variants and overcome the limitations of molecular sampling, but requires tumor biopsies, which may be difficult to obtain from patients in their entirety or may not have enough material remaining in the clinic. By solving the multiple technical shortcomings of existing treatment monitoring approaches, we believe that this methylation-based tumor native multiplexed quantitative assay will be of great help to oncologists and patients to guide treatment decision-making.

[0295] 7.8. Example 6. Filtering hypermethylated loci The lung cancer assay described in Example 2 utilizes background subtraction, which involves subtracting the methylation signal measured in the buffy coat from the methylation signal measured in cfDNA, in parallel with masking of target loci with high background methylation (i.e., masking the signal from target loci with buffy coat methylation above a certain threshold), to improve the signal-to-background ratio. However, masking of target loci with high background methylation has proven to be an effective but insensitive tool that requires additional optimization. The following description is an attempt to further improve the signal-to-background ratio using masking of target loci.

[0296] An attempt to improve the signal-to-background ratio using masking of target loci involved the analysis of 16 samples from 10 separate healthy subjects with the aim of identifying genes that, when masked (removed), would reduce the signal-to-background ratio.

[0297] The pan-cancer assay(s) described herein measured methylation at loci that are frequently hypermethylated in cancer. One limitation of the assay(s) was that not all methylation signals were cancer signals. This could be due to loci having a non-zero amount of background methylation (e.g., methylation in non-cancer subjects).

[0298] As mentioned above, limiting the amount of background methylation has an impact on accurate treatment monitoring. Without wishing to be bound by theory, randomly presented methylation may introduce spurious signals that may lead to inaccurate clinical interpretation. A large amount of background methylation (e.g., more than 20 molecules) also raises the noise floor (e.g., the total amount of methylation observed in subjects without cancer). A large amount of background methylation also affects the ability to detect relative changes in cancer over time, since background may be present at each time point.

[0299] Previous studies have established an amplicon blacklist that contributes to significant background methylation (see Example 2). This blacklist was generated by analyzing four samples from four healthy subjects.

[0300] Although effective in limiting the amount of background methylation, a more comprehensive analysis of non-cancer samples was still required to better determine the updated amplicon blacklist. As mentioned above, 16 specimens from 10 separate subjects who had previously performed methylation assays were analyzed to determine the updated amplicon blacklist. The updated blacklist was applied to the method described herein, and the samples were evaluated for whether the clinical interpretation changed significantly following analysis with the amplicon blacklist.

[0301] Methods. The training set included 16 specimens from 10 separate subjects as described above. The analysis included 8 specimens from batch 2, which did not have cancer but had undergone liver surgery. These 8 specimens were collected twice from 4 subjects. The subjects were 38, 39, 83 and 84 years old. The analysis also included 8 specimens from 6 healthy subjects, i.e. 2 subjects had 2 time points each and the remaining subjects had 1 time point each.

[0302] Two approaches were used to determine the updated amplicon blacklist: individual amplicon blacklists that are patient or sample specific, and global amplicon blacklists.

[0303] 7.8.1. Individual Blacklists The individualized approach allows for the blacklisting of only those loci determined to be likely contributing to the background methylation signal in an individual sample. One non-limiting example of "individualized blacklisting" or "sample-specific blacklisting" involved evaluating buffy coat methylation signals for loci with hypermethylated cfDNA signals in non-cancer subjects (see Figures 9A-9F).

[0304] Additional factors for identifying loci to be included in the "individual blacklist" included determining the age and sex of the patient from whom the sample was taken.

[0305] FIG. 9A shows an example of a locus used in the individual-specific blacklist: intron cg03134157_77 ("intron0313"). As shown in FIG. 9A, the tumor QE for intron0313 showed significantly greater cfDNA methylation than buffy coat methylation only when significant buffy coat methylation was present. Tumor QE was calculated by subtracting blue from red. Error bars show 1 standard deviation as estimated by molecular counting noise based on the QE of the raw data. Overall, the consistency of methylation at intron0313, or little to no methylation at intron0313, meant that the buffy coat methylation threshold for this locus was about 20 tumor QE. This meant that the buffy coat methylation threshold was about 20 tumor QE to properly exclude this locus. Note that an overly small buffy coat threshold may run the risk of blacklisting too many loci (essentially becoming a global blacklist).

[0306] A second locus was analyzed for use in "individual blacklisting". Figure 9B shows the methylation of intron cg20907051_11 ("intron2090"). In particular, intron2090 showed a non-significant difference in cfDNA methylation in buffy coat methylation, regardless of whether significant buffy coat methylation was present (see Figure 9B). Since a huge number of tested healthy samples have non-zero tumor QE, intron2090 is an example of a locus to blacklist for all samples.

[0307] Additional loci analyzed for inclusion in the "individual blacklist" are shown in Figures 9C-9F. Figure 9C shows that the tumor QE for intron_cg12880300_0 had an average tumor QE of 59, with 9 of 16 specimens having a tumor QE greater than 2. This locus was an example of a locus to blacklist for all samples, as it had consistently high buffy coat methylation and variable cfDNA methylation.

[0308] FIG. 9D shows tumor QE for locus IRX4_cg13974394_0 ("IRX4"). At the IRX4 locus, significant tumor QE (red greater than blue) was observed in only one of 16 specimens. This one sample was not flagged during QC. In fact, the reads per QCT was about 4.9 in this cfDNA sample, lower than the other samples but not significantly lower. This data suggested that IRX4 may be best assessed on a sample-by-sample basis. In addition, the data suggested that high gDNA signal would make any actual tumor QE measurement at this locus relatively imprecise. Moreover, at this locus, methylation appears to correlate with age. BTO-23 and BTO-24 are in their 80s, BTO-21 and BTO-22 are in their late 30s, and healthy volunteers are overwhelmingly on the younger side.

[0309] Figures 9E-9F show the tumor QEs for loci EMID2_cg25290307_0 ("EMID2") and intron_cg11453719_0, respectively. As shown in Figure 9E, a filter of 10QE (dashed line) for gDNA removed BTO-21_2, BTO-23_1, and BTO-23_2, but not BTO-24_2. As shown in Figure 9F, a filter of 10QE (dashed line) for gDNA removed all samples except BTO-24_1 and healthy_009_1. For these two loci, the individual-specific blacklist approach requires a relatively strict threshold of 10QE, but still cannot ignore spurious methylation signals in multiple samples. Therefore, it may be best to include these loci in a global blacklist.

[0310] 7.8.2. Global Blacklist To determine the criteria for including amplicons in the global blacklist, an empirical approach was used with set thresholds for the mean and maximum tumor methylation QE (described below). The thresholds were applied to determine the blacklist and to calculate the new total tumor QE in non-cancer samples. The thresholds were also applied to previous analyses (e.g., UCSD and UF results) to determine whether these results are still clinically valid.

[0311] This analysis was focused in part on the top 20 amplicons based on mean tumor methylation QE (see Table 2). In Table 2, the columns show the number of samples with a tumor QE of greater than 2 for the locus across all 16 samples (n_samples_gt2), the mean tumor QE (mean_tumor_norm) and maximum tumor QE (max_tumor_norm), and the mean buffy coat methylation QE (mean bg_norm). TIFF2025509878000005.tif149170

[0312] Based on an analysis of the top 20 amplicons, some of which are shown in Table 2, thresholds for global blacklisting included: (1) maximum tumor QE in non-cancer samples >15; and (2) average tumor QE (including 0) >2.

[0313] In a non-limiting example, applying the following thresholds: (1) maximum tumor QE in any non-cancer specimen > 15; and (2) average tumor QE (including 0) > 2, resulted in a global blacklist containing 32 amplicons. Compared to the historical blacklist (see Table 3), 16 amplicons were duplicated, meaning that 6 amplicons were removed and 16 amplicons were added. TIFF2025509878000006.tif244170

[0314] A new global blacklist created based on the thresholds described above (i.e., (1) maximum tumor QE in any non-cancer specimen > 15; and (2) average tumor QE (including 0) > 2) was applied to 16 non-cancer specimens to determine the noise floor. As shown in FIG. 10, no sample had methylation greater than a tumor QE of 114. Therefore, the noise floor was set at 120 in the additional analysis. However, one of skill in the art will recognize that the noise floor may vary based on similar types of analyses applied to different samples or different sets of samples.

[0315] The new global blacklist was then compared to the old global blacklist. As shown in Figure 11, comparing the normalized total tumor molecules using either the old or new global blacklist was equal for most subjects, if not fewer tumor molecules. In addition, the methylation results for all 30 subjects were manually reviewed to assess whether the clinical interpretation was changed. Only two of the total 30 subjects had slightly different results, both within the statistical noise. For both subjects 8272 and 6885, the results of the last retrieval changed slightly, but still within the molecular noise error bars (see Figures 12A-12D). Overall, this data indicated that small changes in the blacklist did not significantly affect clinical outcomes.

[0316] The new global blacklist was then compared to the old global blacklist for samples from five patients (see BTO-1-BTO-5 in Figures 13A-13E, respectively) who had various cancer stages but had surgical removal of liver cancer. As shown in Figures 13A-13E, the trends for each patient were consistent with the previous analysis. In addition, each patient had a decrease in tumor methylation, except for BTO-1, a stage 4 liver cancer patient who had high levels of methylation detected and whose sample was provided the day after surgery.

[0317] 7.8.3. Conclusion This data showed the benefit of global versus individual blacklists (either patient-specific or sample-specific). In particular, global blacklists avoided addressing unanswered questions of biological and temporal variability. However, this analysis excluded the use of individual blacklists, and additional optimization may be required to design effective individual blacklists.

[0318] The data also showed that adjusting the blacklist from 22 to 32 amplicons did not significantly change clinical outcomes beyond the measurement noise in two clinical trial datasets. Importantly, the noise floor increased the normalized total tumor QE from 50 to 120.

[0319] 7.9. Example 7. Analytical and clinical validation of the pan-cancer assay The pan-cancer assays provided in this disclosure and described, for example, in Example 5, have undergone analytical validation, including testing for accuracy, precision, reproducibility, sensitivity, and specificity.

[0320] 7.9.1. Introduction and summary of the test group Accurate, rapid, and available treatment monitoring for cancer patients is an unmet medical need. When the outcome of a cancer treatment regimen is uncertain, determining early rather than late whether the treatment is effective for a patient allows switching to a different treatment regimen, potentially resulting in extended lifespan, reduces unnecessary side effects from ineffective treatment, improves quality of life, and improves the overall efficiency of the health care system. Evaluating treatment effectiveness earlier or predicting final treatment outcome can be very important for late-stage cancer patients or for efficient execution of clinical trials for new cancer treatments where time is of the essence.

[0321] Circulating tumor DNA (ctDNA) obtained through liquid biopsies from cancer patients has been shown to reflect the amount of cancer present in the patient's body. In addition, methylation has been shown to be a robust cancer biomarker, and several groups have developed methylation-based assays using ctDNA for various cancer diagnostic applications. Using our patented QCT technology (Tsao et al., 2019), we have developed a novel assay, termed the pan-cancer assay, to quantify the amount of methylation in ctDNA for accurate and precise treatment monitoring applications.

[0322] Validation studies were performed to demonstrate the analytical and clinical validation of the pan-cancer treatment monitoring assay.

[0323] Analytical validation included: Group 1: accuracy, precision, and reproducibility with sheared tumor DNA samples; Group 2: concordance with clinical samples and known clinical outcomes; Group 3: limit of detection using sheared tumor DNA samples; and Group 4: diversity of cancer types.

[0324] 7.9.2. Group 1: Accuracy, Precision, and Reproducibility with Sheared Tumor DNA Samples The objective of Group 1 was to calculate accuracy, precision, and reproducibility based on the sheared tumor DNA samples. Replicates of sheared tumor DNA samples with 1% and 2% tumor fraction were generated from a total of seven different tumors from seven different cancer types. The sheared tumor DNA samples were processed in two different batches and loaded onto two different sequencers. A response score, which accounts for the amount of methylation in the sample, was calculated for each sample, and the response scores of the 1% and 2% tumor fraction samples were compared to each other.

[0325] As summarized and described in further detail below, data from Group 1 demonstrated 100% accuracy (i.e., 40 out of 40 comparisons were called correctly); 100% precision (i.e., no subjects within each batch had discordant results); and 100% reproducibility (i.e., 9 out of 19 comparisons were concordant between batches).

[0326] In both batches, all comparisons of the 1% and 2% tumor rate samples that passed QC were called "increased," consistent with the identity of these samples. Response scores from both batches are shown in Figures 14 and 15.

[0327] Additional calculations were performed for the sensitivity and specificity of the pan-cancer assay. Sensitivity was calculated based on a comparison of 1% and 2% samples as described above, and specificity was calculated by comparing the 1% to themselves and the 2% to themselves within each batch. This analysis revealed a sensitivity = 100% [95% CI: 91.2%, 100%] (i.e., 40 out of 40 comparisons) and specificity = 100% [95% CI: 95.5%, 100%] (i.e., 80 out of 80 comparisons).

[0328] Collectively, the results from Group 1 demonstrate that the pan-cancer assay is accurate, precise, and reproducible across operators and sequencers.

[0329] 7.9.3. Group 2: Matching clinical samples to known clinical outcomes The purpose of Group 2 was to evaluate the accuracy of the pan-cancer assay on clinical samples from a total of 20 subjects. These 20 subjects consisted of 8 cancer subjects with known clinical outcomes and 12 healthy subjects with no known history of cancer. Response scores were measured at two time points for each subject. Calls were made based on the response scores from both time points and the calls were compared to the known clinical outcomes.

[0330] As summarized and described in further detail below, data from Group 2 showed 100% accuracy for cancer subjects (i.e., 6 out of 6 concordant calls); 100% accuracy for healthy subjects (11 out of 11 concordant calls).

[0331] All 6 of the 6 cancer subjects who had a detectable signal had calls consistent with known clinical outcomes. The results for all cancer subjects are plotted in Figure 16, and the known clinical outcomes can be found in Table 4. TIFF2025509878000007.tif80170

[0332] Eleven of the 12 healthy subjects had unchanged or indeterminate calls for cancer, consistent with the health status of these subjects (see FIG. 17).

[0333] In summary, the results from Group 2 demonstrate the efficacy of the pan-cancer assay to accurately evaluate clinical samples from a total of 20 subjects.

[0334] 7.9.4. Group 3: Limit of detection using sheared tumor DNA samples The objective of group 3 was to evaluate the detection limit of the pan-cancer assay by measuring sensitivity under different input conditions.

[0335] In these experiments, sheared tumor DNA samples with 0%, 0.25%, 0.5%, 1%, and 2% tumor fraction were generated from single tumors and their matched buffy coats and analyzed using the pan-cancer assay. For each tumor fraction, response scores were measured in 16 replicates, and these measurements were used to calculate the CV. The sensitivity of the assay was calculated by comparing the response scores of the samples at each tumor fraction to the response score of the 0% sample, and the detection limit was then determined.

[0336] As summarized and described in further detail below, the results showed a detection limit = 0.25% tumor rate.

[0337] The detection limit was defined as the lowest tumor rate with at least 95% sensitivity. In the 0.25% tumor rate samples, the sensitivity was 96.5% [95% CI: 93.4%, 98.4%]. The sensitivity for each tumor rate is shown in Figure 18.

[0338] The response scores for each tumor rate are plotted in Figure 19. As expected, higher tumor rates corresponded to higher response scores. Within each group of technical replicates, the CV of the response scores was calculated at a given tumor rate (see Figure 19). There is a clear separation between 0% and 0.25% tumor rates, consistent with the 0.25% tumor rate detection limit (see Figure 19). The standard deviation and mean of each tumor rate were used to calculate the CV.

[0339] In summary, the results from Group 3 demonstrate that the limit of detection for the pan-cancer assay is at least 0.25% tumor rate.

[0340] 7.9.5. Group 4: Cancer type diversity The purpose of Group 4 was to evaluate the performance of the pan-cancer assay across different cancer types. Tumor and matched buffy coat gDNA samples were processed for each of 54 different cancer patients spanning 10 unique cancer types.

[0341] In these experiments, bioinformatics simulations were used to evaluate the performance of the pan-cancer assay across different cancer types. For example, bioinformatics was used to simulate 1% and 2% tumor DNA samples by scaling down the number of methylated molecules based on the estimated tumor purity. Using a standard analysis workflow, the response scores of all 2% tumor samples were correctly called as an increase relative to the 1% tumor sample response scores (see FIG. 20; Table 5). As shown in Table 5, the sensitivity across cancer types was 100% [95% CI: 93.4%, 100%]. TIFF2025509878000008.tif136170

[0342] In summary, the data confirm that the pan-cancer assay works with high sensitivity across a range of cancer types: in all 54 cancer patients in the validation group, the pan-cancer assay detected an increase in response score when a change in tumor prevalence from 1% to 2% was simulated.

[0343] 7.9.6. Conclusion This analytical and clinical validation report describes the results from a series of experiments conducted to support the validation of the pan-cancer assay. Collectively, the results show that the assay has a high degree of accuracy and precision, a high degree of reproducibility, a high degree of sensitivity and specificity, and a low detection limit. By measuring the amount of ctDNA in cancer patients by liquid biopsy, the pan-cancer assay provides oncologists with additional information that can be used to improve clinical outcomes for cancer patients.

[0344] 7.10. Example 8. Detection of methylation consistent with disease progression A pan-cancer assay was used to assess the concordance between methylation and disease progression. In this analysis, both methylation and CT imaging were used to assess disease progression in four patients. Methylation was monitored using the pan-cancer assay. CT imaging was performed by clinicians and used to stage disease progression. Data shown in Figures 21-24 are presented as methylation ("tumor QE") over time ("days since treatment initiation"), with triangles indicating the stage of disease progression by the clinician.

[0345] FIG. 21 shows a methylation analysis and clinician stage call of disease progression for a 56-year-old male subject with pancreatic ductal adenocarcinoma stage IV at the time of first collection (see triangles in FIG. 21). As recorded in FIG. 21, methylation analysis was performed at the indicated time points and was performed concurrently with a treatment regimen of folfiri followed by folfiri maintenance followed by folox. The methylation levels here were consistent with the clinician stage call. For example, a decrease in methylation corresponded to a decrease in tumor size or stable tumor growth (as indicated by the clinician), and an increase in methylation corresponded to the likelihood of disease progression (as indicated by the clinician).

[0346] FIG. 22 shows methylation analysis and clinician stage call of disease progression for a 50-year-old male subject with pancreatic adenocarcinoma stage IV at the time of first collection (see triangles in FIG. 22). As noted in FIG. 22, methylation analysis was performed at the indicated time points and was concomitant with a treatment regimen of 5000 mg fluorouracil continuous for 46 hours + 800 mg leucovorin + 180 mg oxaliplatin + 300 mg irinotecan. The methylation levels here were consistent with the clinician stage call. For example, the methylation levels increased before day 200 and continued to rise past day 200, so the clinician stage call indicated disease progression.

[0347] Figure 23 shows a methylation analysis for an 87-year-old male subject with colon adenocarcinoma stage IV at the time of first collection. Figure 23 also shows the stage of disease progression by clinicians (see triangles in Figure 23). As described in Figure 23, methylation analysis was performed at the indicated time points and was performed concomitantly with a treatment regimen of 6 cycles of fluorouracil 1100mg + leucovorin 1100mg + bevacizumab 500mg. The methylation levels here were consistent with the stage by clinicians. For example, the methylation levels increased before day 200 and continued to rise after day 200, so the stage by clinicians indicated disease progression.

[0348] Figure 24 shows a methylation analysis for a 56-year-old female subject with stage IV squamous cell lung cancer at the time of first collection. Figure 24 also shows the stage of disease progression by clinician (see triangles in Figure 24). As noted in Figure 24, methylation analysis was performed at the indicated time points and was concomitant with a docetaxel treatment regimen. As with the patients analyzed in Figures 21-23, the methylation levels here were consistent with the stage by clinician.

[0349] 7.11. Example 9. Concordance with variant allele frequency Pan-cancer assays were also used to assess the concordance between methylation and mutant allele frequencies.

[0350] In this experiment, 40 samples were run through the treatment selection assay. For each sample, an aliquot of plasma was collected at the same time points as the methylation assay. From the treatment selection assay, the maximum mutant allele frequency (VAF) was calculated and compared against the methylation assay. As shown in Figure 25, there was a correlation between the maximum VAF and the number of methylated molecules. A perfect one-to-one correlation was not expected because somatic mutations may be heterogeneous throughout the tumor and may not reflect the total prevalence of cancer.

[0351] One limitation of the treatment selection assay was false-negative actionable mutations. Given the correlation between VAF and methylation, the methylation results were used to inform whether false-negative actionable mutations identified by VAF were likely false negatives. For example, if an actionable mutation was found but the methylation level was low, this suggested a low tumor rate at this time point and a possibility of a false-negative actionable mutation due to a small number of tumor molecules. On the other hand, if no actionable mutation was detected and the methylation level was high, this suggested a high tumor rate at this time point and a low possibility of a large number of false-negative actionable mutations.

[0352] In summary, the concordance between VAF and methylation demonstrated that methylation (i.e., methylation analyzed and identified according to the methods described herein) can be used to inform the results of VAF analysis and therefore can be used to complement the adjudication of therapeutic selection assays.

[0353] 7.12. Example 10. Changes in methylation profile The pan-cancer assay was also used to assess methylation profiles over time. In these experiments, methylation patterns were analyzed in two subjects (subject 6885 and subject 5458) over 294 and 250 days, respectively. As summarized and described in more detail below, it was observed that some subjects had relatively constant methylation profiles over time, while some patients had large changes in methylation profiles.

[0354] For example, as shown in Figure 26, in subject 6885, each new timepoint had new loci that were methylated. However, with the exception of loci that were present at the day 0 timepoint, these new loci did not contribute significant amounts of methylation at the subsequent timepoints. For example, loci that first appeared on day 63 did not contribute significant amounts of methylation at the subsequent timepoints.

[0355] In a second non-limiting example shown in FIG. 27, subject 5458 had an increase in total methylation at day 119 in conjunction with an interesting methylation pattern. Notably, at day 119, slightly more than half the methylation was from new loci (i.e., methylated loci not detected at day 0, day 35, or day 63), suggesting that the tumor had evolved significantly compared to previous timepoints. In fact, timepoints after day 119 maintained a significant proportion of loci that first appeared at day 119, indicating that the tumor methylation changes were persistent. Based on the large increase in methylation patterns and the rapid compositional change, this subject was flagged for another treatment selection assay to detect any new somatic mutations present in the cancer. Clinically, a pattern such as that observed in subject 5458 may indicate the need to reevaluate treatment options.

[0356] 7.13. Conclusions of Examples 8 to 10 Collectively, this data establishes the utility of the pan-cancer assay for (1) detecting concordance between methylation and disease progression; (2) detecting concordance between methylation and mutant allele frequency; and (3) assessing methylation profiles over time.

[0357] 7.14. Example 11: Methylation patterns in patients with pancreatic ductal adenocarcinoma A pan-cancer assay was used to assess methylation profiles over time in patients with pancreatic ductal adenocarcinoma. The assay was used to detect a decrease in methylated molecules in circulating tumor DNA (ctDNA) compared to earlier measurements. Aberrantly methylated DNA is a known marker of cancer cells (PMID15542813), and changes in methylated ctDNA correspond to changes in tumor incidence. The results suggest that tumor incidence has decreased compared to previous measurements.

[0358] In these experiments, plasma and buffy coat were isolated from whole blood collected in Streck cell-free DNA tubes. Cell-free DNA (cfDNA) was extracted from plasma and genomic DNA (gDNA) was extracted from buffy coat. At over 500 positions in the genome known to be hypermethylated in cancer compared to non-cancerous tissues and blood, the number of methylated molecules was quantified in both cfDNA and gDNA using QCT molecular counting technology (PMID:31591409). To remove background from the ctDNA signal, the methylation measured in gDNA was subtracted from the cfDNA methylation. To calculate the Tumor Methylation Score™, the remaining cfDNA methylated molecules were summed across all hypermethylated positions. Figure 28 shows the Tumor Methylation Score™ expressed as the normalized sum of methylated molecules at over 500 loci hypermethylated in circulating tumor DNA (ctDNA). As shown in Figure 28, a decrease in Tumor Methylation Score™ was detected, which corresponds to a decrease in methylated ctDNA molecules. In particular, Figure 28 shows that the tumor rate was reduced by about 3.1-fold compared to previous measurements.

[0359] The Tumor Methylation Score™ from the current collection was compared to the most recently reported Tumor Methylation Score™ to determine a call of increase, decrease, or unchanged. To be reported as an increase or decrease, the change in Tumor Methylation Score™ must exceed a significant threshold. For baseline testing without any prior collections, no interpretive calls are made for changes in Tumor Methylation Score™. Results should be discussed with a medical professional and interpreted in the context of multiple time points and in conjunction with the patient's complete clinical history.

[0360] In some cases, if the sample contains an insufficient amount of DNA, methylation may not be reported. Results below the limit of detection (LOD) of Tumor Methylation Score™, represented by the cross-hatch shading on the graph, will not be interpreted and will be reported as below the LOD. Performance specifications based on internal validation studies demonstrated that the assay can distinguish a 0.2 percentage point change in tumor rate with a separation of 3 standard deviations. This pan-cancer assay was designed to quantify Tumor Methylation Score™ in patients with solid tumors. Results may be variable or invalid if the patient has recently undergone blood transfusion, stem cell transplant, or other treatment that may significantly affect the composition of cfDNA or buffy coat gDNA.

[0361] 8. Equivalents and Incorporation by Reference While the present invention has been shown and described in detail with reference to preferred and various alternative embodiments, it will be understood by those skilled in the art that various changes in form and detail can be made therein without departing from the spirit and scope of the invention.

Claims

1. A method for quantifying DNA methylation in a sample containing a DNA sequence, wherein the method is The sample is treated to encode the presence or absence of DNA methylation within the DNA sequence, wherein the sample contains at least 10 target loci from the DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule, Including; To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; Sequencing the co-amplified mixture to produce sequence reads; Determining the number of sequence reads that are methylated sequence reads; and A method comprising quantifying the number of methylated molecules in a sample based on the number of methylated reads from the sample and the number of reads from the set of synthetic molecules.

2. The sample is taken from collected blood containing plasma and a buffy coat, wherein the plasma contains cfDNA and the buffy coat contains a genomic DNA (gDNA) sequence. The method further includes extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample before treating the sample to encode the presence or absence of DNA methylation. The method according to claim 1.

3. The sample comprising a DNA sequence from a buffy coat to encode the presence or absence of DNA methylation within the DNA sequence, wherein the sample comprises at least 10 target loci from the DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule, Including; To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; Sequencing the co-amplified mixture to produce sequence reads; Determining the number of sequence reads that are methylated sequence reads; and The number of methylated molecules in the sample is determined based on the number of methylated leads from the sample and the number of leads from the set of synthetic molecules. This allows for the quantification of DNA methylation within the gDNA sequence from the buffy coat. The method according to claim 1, further comprising quantifying DNA methylation in a sample containing a gDNA sequence from a buffy coat.

4. The method according to any one of claims 1 to 3, further comprising adding a known sequence and amount of spike in to the sample prior to the treatment step.

5. The method according to any one of claims 1 to 3, wherein treating the sample to encode the presence or absence of DNA methylation includes bisulfite conversion or enzymatic conversion.

6. The method according to any one of claims 1 to 3, wherein each of the at least 10 target loci is selected based on the predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue.

7. The method according to any one of claims 1 to 3, wherein at least 100 target loci are amplified, or at least 500 target loci are amplified.

8. The method according to any one of claims 1 to 3, wherein the co-amplified mixture is sequenced with a read depth of at least one read per molecule, 10 or more reads per molecule, 100 or more reads per molecule, or 1000 or more reads per molecule.

9. Furthermore, the method according to any one of claims 1 to 3, comprising counting the number of methylated molecules across at least two of the at least 10 target gene loci in order to quantify the DNA methylation in the sample.

10. The method according to claim 9, wherein the aggregation includes aggregating target loci from the at least 10 target loci containing methylated molecules in the buffy coat sample in amounts less than a threshold.

11. The method according to any one of claims 1 to 3, wherein the sample further comprises at least one normalization locus that is predicted to have hypermethylation in cfDNA in both cancerous and non-cancerous tissue, and the co-amplification mixture further comprises the amplified at least one normalization locus.

12. The method according to any one of claims 1 to 3, further comprising normalizing the aggregate number of methylation molecules from at least two of the at least ten target loci by the methylation molecule for the amplified at least one normalization locus.

13. Furthermore, the method according to any one of claims 1 to 3, comprising at least one of subtracting background methylation or removing a selected hypermethylated target locus.

14. The method according to claim 13, wherein the subtraction of background methylation includes subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA.

15. The method according to claim 14, wherein the subtraction of background methylation is performed on a locus basis.

16. The method according to claim 13, wherein removing the selected hypermethylated target locus includes removing a target locus having a hypermethylated cfDNA signal in a non-cancerous tissue.

17. The method according to claim 13, wherein removing the selected hypermethylated target loci includes removing a target locus having a total number of methylated molecules in the buffy coat that exceeds a threshold.

18. The method according to claim 17, wherein the threshold is sample-specific or target-specific.

19. The method according to claim 17, wherein the threshold includes a predetermined mean quantitative equivalent (QE), a predetermined maximum tumor methylation QE, or a combination thereof.

20. The method according to claim 13, wherein at least one of removing a selected hypermethylation target locus or subtracting background methylation is performed before determining the number of methylated molecules in the sample.

21. A method for determining the DNA methylation profile in a subject, wherein the method is Treating a sample isolated from a subject to encode the presence or absence of DNA methylation within the aforementioned DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule, Including; To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; To produce sequence reads, the co-amplified mixture is sequenced at a read depth of at least one read sequence per molecule in the sample; Determining the number of sequence reads that are methylated sequence reads; To quantify the number of methylated molecules in the sample based on the number of methylated sequence reads in the sample and the number of reads from the set of synthetic molecules; and The methylation profile in the subject is determined based on the number of methylated molecules in the sample. Methods that include...

22. A method for determining the DNA methylation profile in a subject over time, wherein the method is At the first time point: i) Treating a sample isolated from the subject to encode the presence or absence of DNA methylation in the DNA sequence, wherein the sample contains at least 10 target gene loci from the DNA sequence; ii) Add the set of synthetic molecules to the sample, and the set of molecules A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule, Including; iii) To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; iv) Sequencing the co-amplified mixture at a read depth of at least one read sequence per molecule in the sample in order to produce sequence reads; v) Determining the number of sequence reads that are methylated sequence reads; vi) Quantifying the number of methylated molecules in the sample based on the number of methylated sequence reads in the sample and the number of reads from the set of synthetic molecules; At the second time point, repeat steps i) to vi); and The DNA methylation profile in the subject is determined based on the number of methylated molecules in the sample at the first time point and the number of methylated molecules in the sample at the second time point. Methods that include...

23. The method according to claim 22, wherein the number of methylated sequence reads from each target gene locus in the sample is quantified based on the number of methylated sequence reads from each target gene locus in the sample and the number of reads from the set of synthetic molecules.

24. The method according to claim 22, wherein determining the methylation profile in the object at the first time point and the second time point identifies the change in the methylation profile between the first time point and the second time point.

25. moreover, Repeat steps i) to vi) at the third time point; and Determining the methylation profile in the subject based on the number of methylated molecules for each target gene locus in the sample at the third time point, The method according to any one of claims 22 to 24, including the method described in any one of claims 22 to 24.

26. The method according to claim 25, wherein determining the methylation profile in the object at the first time point, the second time point, and the third time point identifies changes in the methylation profile between the first time point and the second time point, between the second time point and the third time point, between the first time point and the third time point, or in combination thereof.

27. The method according to any one of claims 22 to 24, wherein the change in the methylation profile indicates a change in the tumor in the subject, and the change in the tumor includes a change in the size of the tumor, a change in the prevalence of somatic mutations associated with the tumor, the presence of new somatic mutations associated with the tumor, a chromosomal abnormality associated with the tumor, or resistance of the tumor to treatment.

28. The method according to claim 24, wherein the change in the methylation profile is incorporated into a clinical recommendation for the subject.

29. The method according to claim 24, further comprising assigning the changes in the methylation profile to metrics that increase, decrease, or remain unchanged based on a comparison to a significant threshold.

30. The method according to claim 29, wherein the significant threshold is predetermined or dynamically calculated.

31. The sample is taken from collected blood containing plasma and a buffy coat, wherein the plasma contains cfDNA and the buffy coat contains a genomic DNA (gDNA) sequence. The method according to any one of claims 22 to 24, further comprising extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample before treating the sample to encode the presence or absence of DNA methylation.

32. The sample comprising a DNA sequence from a buffy coat to encode the presence or absence of DNA methylation within the DNA sequence, wherein the sample comprises at least 10 target loci from the DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the sequence region of the endogenous target molecule, Including; To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; Sequencing the co-amplified mixture to produce sequence reads; Determining the number of sequence reads that are methylated sequence reads; Quantifying the number of methylated molecules in the sample based on the number of methylated leads from the sample and the number of leads from the set of synthetic molecules; and In the sample containing the DNA sequence from the buffy coat, the number of methylated molecules across at least two target gene loci was counted in order to quantify DNA methylation. The method according to claim 31, further comprising quantifying DNA methylation in the buffy coat, and thereby quantifying DNA methylation in a sample containing a gDNA sequence from the buffy coat.

33. The method according to any one of claims 22 to 24, further comprising adding a known sequence and amount of spike in to the sample prior to the treatment step.

34. The method according to any one of claims 22 to 24, wherein each of the at least 10 target loci is selected based on a predicted increase in DNA methylation at that locus in cancerous tissue compared to non-cancerous tissue.

35. The method according to any one of claims 22 to 24, wherein at least 100 target loci are amplified, or at least 500 target loci are amplified.

36. The method according to any one of claims 22 to 24, wherein the sample further comprises at least one normalization locus that is predicted to have hypermethylation in cfDNA across both cancerous and non-cancerous tissues, and the co-amplification mixture further comprises at least one amplified normalization locus.

37. The method according to any one of claims 22 to 24, further comprising counting the number of methylated molecules across at least two of the at least ten target loci in order to quantify the DNA methylation in the sample, wherein the counting includes counting the target loci from the at least ten target loci that contain fewer methylated molecules than a threshold amount in the buffy coat sample.

38. The method according to claim 37, further comprising normalizing the aggregate number of methylated molecules from at least two of the at least ten target loci by the methylated molecule for the at least one normalization locus.

39. Furthermore, the method according to any one of claims 22 to 24, further comprising at least one of subtracting background methylation or filtering out hypermethylated target loci.

40. The method according to claim 39, wherein the subtraction of background methylation includes subtracting the number of methylated molecules measured in the buffy coat from the number of methylated molecules in the cfDNA.

41. The method according to claim 39, wherein the subtraction of background methylation is performed on a locus basis.

42. The method according to claim 39, wherein the removal of selected hypermethylated loci, the subtraction of background methylation, or both is performed before quantifying the number of methylated molecules in the sample.

43. The method according to claim 39, wherein removing the selected hypermethylated target loci includes removing a target locus having a total number of methylated molecules in the buffy coat that exceeds a threshold.

44. The method according to claim 43, wherein the threshold is sample-specific or target-specific.

45. The method according to claim 43, wherein the threshold includes a predetermined mean tumor methylation QE, a predetermined maximum tumor methylation QE, or a combination thereof.

46. Furthermore, the method according to any one of claims 22 to 24, comprising detecting a change in the DNA methylation profile and performing a treatment selection assay on the subject.

47. The method according to claim 46, wherein the treatment selection assay includes genomic profiling for detecting novel somatic mutations, the prevalence of somatic mutations, or both.

48. A method for determining a reliable indication of the presence or absence of somatic mutations in a sample containing a DNA sequence, wherein the method is To determine the presence or absence of somatic mutations in the aforementioned sample; The purpose is to quantify DNA methylation in the aforementioned sample, The steps include: treating the sample to encode the presence or absence of DNA methylation within the DNA sequence, wherein the sample comprises at least 10 target loci from the DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the sequence region of the endogenous target molecule, Steps including; The steps include: preparing a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; To produce sequence reads, the co-amplified mixture is sequenced at a read depth of at least one read sequence per molecule in the sample; The steps include determining the number of sequence reads that are methylated sequence reads; A step of quantifying the number of methylated molecules in the sample based on the number of methylated leads from the sample and the number of leads from the set of synthetic molecules, To quantify DNA methylation in the sample, including; and If the number of methylated molecules in the sample exceeds a predetermined threshold or a dynamically calculated threshold, a confidence indication is returned as a true call result regarding the presence or absence of the somatic mutation, or If the number of methylated molecules in the sample is below a predetermined threshold or below a dynamically calculated threshold, a confidence indication is returned as a no-call result regarding the presence or absence of the somatic mutation. Therefore, true calling identifies confidence in determining the presence or absence of the somatic mutation. Methods that include...

49. The method according to claim 48, wherein determining the presence or absence of the somatic mutation in the sample includes determining the mutation allele frequency for the somatic mutation.

50. The sample is taken from collected blood containing plasma and a buffy coat, wherein the plasma contains cfDNA and the buffy coat contains a genomic DNA (gDNA) sequence. The method further includes extracting cfDNA from the plasma and gDNA sequences from the buffy coat from the sample before treating the sample to encode the presence or absence of DNA methylation. The method according to claim 48 or 49.

51. The sample comprising a DNA sequence from a buffy coat to encode the presence or absence of DNA methylation within the DNA sequence, wherein the sample comprises at least 10 target loci from the DNA sequence; A set of synthetic molecules is added to the sample, and the set of molecules is A target-associated region having a nucleotide sequence that matches the target sequence region of an endogenous target molecule containing at least one of the aforementioned target gene loci, A mutant region having a nucleotide sequence that does not match the target sequence region of the endogenous target molecule, Including; To prepare a co-amplified mixture comprising an amplified set of synthetic molecules and an amplified set of at least 10 target gene loci from the DNA sequence; Sequencing the co-amplified mixture to produce sequence reads; Determining the number of sequence reads that are methylated sequence reads; and The number of methylated molecules in the sample is quantified based on the number of methylated leads from the sample and the number of leads from the set of synthetic molecules. This allows for the quantification of DNA methylation within the gDNA sequence from the buffy coat. The method according to claim 50, further comprising quantifying DNA methylation in a sample containing a DNA sequence from a buffy coat.

52. The method according to claim 48 or 49, further comprising adding a known sequence and amount of spike in to the sample prior to the treatment step.

53. The method according to claim 48 or claim 49, wherein the set of synthetic molecules is a set of quantitative counting templates (QCTs).

54. The method according to claim 48 or claim 49, wherein each of the at least 10 target loci is selected based on a predicted increase in DNA methylation at each locus in cancerous tissue compared to non-cancerous tissue.

55. The method according to claim 48 or claim 49, wherein at least 100 target gene loci are amplified, or at least 500 target gene loci are amplified.

56. The method according to claim 48 or 49, wherein the sample further comprises at least one normalization locus predicted to have hypermethylation in cfDNA across both cancerous and non-cancerous tissues, and the co-amplification mixture further comprises at least one amplified normalization locus.

57. Furthermore, the method according to claim 48 or claim 49, comprising counting the number of methylated molecules across at least two of the at least 10 target gene loci in order to quantify the DNA methylation in the sample.

58. The method according to claim 57, wherein the aggregation includes aggregating the target loci from the at least 10 target loci that contain methylated molecules in less than a threshold amount in the buffy coat sample.

59. The method according to claim 57, further comprising normalizing the aggregate number of methylated molecules from at least two of the at least ten target loci by the methylated molecule for the at least one normalization locus.

60. The method according to claim 48 or claim 49, further comprising subtracting background methylation or removing selected hypermethylated loci.

61. The method according to claim 60, wherein subtracting background methylation includes subtracting the number of methylated molecules measured in the buffy coat as quantified in claim 74 from the number of methylated molecules in the cfDNA.

62. The method according to claim 61, wherein background methylation is subtracted on a locus basis.

63. The method according to claim 60, wherein the removal of selected hypermethylated loci, the subtraction of background methylation, or both is performed before quantifying the number of methylated molecules in the sample.

64. The method according to claim 60, wherein removing selected hypermethylated target loci includes removing target loci having a total number of methylated molecules in the buffy coat that exceeds a threshold.

65. The method according to claim 64, wherein the threshold is sample-specific or target-specific.

66. The method according to claim 64, wherein the threshold includes a predetermined mean tumor methylation QE, a predetermined maximum tumor methylation QE, or a combination thereof.

67. The method according to claim 48 or 49, further comprising returning a true call and, if the prevalence of the somatic mutation indicates the presence of cancer, performing or repeating a treatment selection assay on the subject.

68. The method according to claim 67, wherein the treatment selection assay includes genomic profiling for detecting novel somatic mutations, the prevalence of somatic mutations, or both.

69. The method according to claim 48 or claim 49, further comprising incorporating the trusted instructions into the clinical recommendations for the subject.

Images