MULTIMODAL ANALYSIS OF CIRCULATING TUMOR NUCLEIC ACID MOLECULES

Information

  • Patent Application
  • 20230212690
  • Publication Number
    20230212690
  • Date Filed
    December 16, 2022
    a year ago
  • Date Published
    July 06, 2023
    11 months ago
Abstract
In an aspect, there is provided a method of detecting the presence of ctDNA from cancer cells in a subject comprising: (a) providing a sample of cell-free DNA from a subject; (b) subjecting the sample to library preparation to permit subsequent sequencing of the cell-free methylated DNA; (c) optionally adding a first amount of filler DNA to the sample, wherein at least a portion of the filler DNA is methylated, then further optionally denaturing the sample; (d) capturing cell-free methylated DNA using a binder selective for methylated polynucleotides; (e) sequencing the captured cell-free methylated DNA; (f) comparing the sequences of the captured cell-free methylated DNA to control cell-free methylated DNAs sequences from healthy and cancerous individuals; (g) identifying the presence of DNA from cancer cells if there is a statistically significant similarity between one or more sequences of the captured cell-free methylated DNA and cell-free methylated DNAs sequences from cancerous individuals; wherein in at least one of the capturing step, the comparing step or the identifying step, the subject cell-free methylated DNA is limited to a sub-population according to a fragment length metric.
Description
BACKGROUND

Circulating tumor DNA (ctDNA) has increasingly demonstrated potential as a non-invasive, tumor-specific biomarker for routine clinical use. ctDNA is derived from tumor cells predominately undergoing cell-death and released into circulation of various bodily fluids including blood. In most cancer patients, the majority of blood-derived cell-free DNA originates from peripheral blood leukocytes (PBLs); therefore, identification of tumor-derived genetic and epigenetic alterations are required for ctDNA detection and quantification. In addition, the fraction of ctDNA observed may range from <0.1% to 90% of total cell-free DNA at diagnosis depending on several factors including primary site of the tumor and disease burden. ctDNAs has been providing non-invasive access to the tumor's molecular landscape and disease burden. Methods for detecting ctDNA with increased sensitivity especially in subjects with lower abundance of ctDNA are needed.


INCORPORATION BY REFERENCE

All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publica-tion, patent, or patent application was specifically and indi-vidually indicated to be incorporated by reference.


SUMMARY

In an aspect, there is provided a method of detecting the presence of ctDNA from cancer cells in a subject comprising:

    • (a) providing a sample of cell-free DNA from a subject;
    • (b) subjecting the sample to library preparation to permit subsequent sequencing of the cell-free methylated DNA;
    • (c) optionally adding a first amount of filler DNA to the sample, wherein at least a portion of the filler DNA is methylated, then further optionally denaturing the sample;
    • (d) capturing cell-free methylated DNA using a binder selective for methylated polynucleotides;
    • (e) sequencing the captured cell-free methylated DNA;
    • (f) comparing the sequences of the captured cell-free methylated DNA to control cell-free methylated DNAs sequences from healthy and cancerous individuals;
    • (g) identifying the presence of DNA from cancer cells if there is a statistically significant similarity between one or more sequences of the captured cell-free methylated DNA and cell-free methylated DNAs sequences from cancerous individuals;
    • wherein in at least one of the capturing step, the comparing step or the identifying step, the subject cell-free methylated DNA is limited to a sub-population according to a fragment length metric.


In as aspect, the present disclosure provides methods for determining whether a subject has or is at risk of having a disease. The methods comprise: subjecting a plurality of nucleic acid molecules derived from a cell-free nucleic acid sample obtained from said subject to sequencing to generate at least one profile selected from the group consisting of (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and processing said at least one profile to determine whether said subject has or is at risk of said disease at a sensitivity of at least 80% or at a specificity of at least about 90%, wherein said cell-free nucleic acid sample comprises less than 30 nanograms (ng)/milliliter (ml) of said plurality of nucleic acid molecules.


In some embodiments, the cell-free nucleic acid sample comprises less than 10 ng/ml of said plurality of nucleic acid molecules. In some embodiments, the cell-free nucleic acid sample comprises less than 5 ng/ml of said plurality of nucleic acid molecules. In some embodiments, the cell-free nucleic acid sample comprises less than 1 ng/ml of said plurality of nucleic acid molecules. In some embodiments, the subjecting of (a) generates at least two profiles selected from the group consisting of (i), (ii) and (iii). In some embodiments, the at least two profiles comprise said methylation profile and said fragment length profile.


In some embodiments, the at least two profiles comprise said mutation profile and said fragment length profile. In some embodiments, the at least two profiles comprise said methylation profile and said mutation profile. In some embodiments, the subjecting of (a) generates said methylation profile, said mutation profile, and said fragment length profile.


In another aspect, the present disclosure provides methods for processing a cell-free nucleic acid sample of a subject to determine whether said subject has or is at risk of having a disease. The methods comprise providing said cell-free nucleic acid sample comprising a plurality of nucleic acid molecules; subjecting said plurality of nucleic acid molecules or derivatives thereof to sequencing to generate a plurality of sequencing reads; computer processing said plurality of sequencing reads to identify, for said plurality of nucleic acid molecules, (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and using at least said methylation profile, said mutation profile and said fragment length profile to determine whether said subject has or is at risk of having said disease.


In some embodiments, the disease comprises a cancer. In some embodiments, the cancer is selected from the group consisting of the cancer is selected from the group consisting of adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain/cns tumors, breast cancer, castleman disease, cervical cancer, colon/rectum cancer, endometrial cancer, esophagus cancer, ewing family of tumors, eye cancer, gallbladder cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal tumor (gist), gestational trophoblastic disease, hodgkin disease, kaposi sarcoma, kidney cancer, laryngeal and hypopharyngeal cancer, leukemia (acute lymphocytic, acute myeloid, chronic lymphocytic, chronic myeloid, chronic myelomonocytic), liver cancer, lung cancer (non-small cell, small cell, lung carcinoid tumor), lymphoma, lymphoma of the skin, malignant mesothelioma, multiple myeloma, myelodysplastic syndrome, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, non-hodgkin lymphoma, oral cavity and oropharyngeal cancer, osteosarcoma, ovarian cancer, penile cancer, pituitary tumors, prostate cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, sarcoma—adult soft tissue cancer, skin cancer (basal and squamous cell, melanoma, merkel cell), small intestine cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, uterine sarcoma, vaginal cancer, vulvar cancer, waldenstrom macroglobulinemia, wilms tumor, squamous cell carcinoma, and head and neck squamous cell carcinoma. In some embodiments, the cancer is squamous cell carcinoma. In some embodiments, the cancer is head and neck squamous cell carcinoma.


In some embodiments, the plurality of cell-free nucleic acid molecules comprises circulating tumor nucleic acid molecules. In some embodiments, the circulating tumor nucleic acid comprises circulating tumor DNA. In some embodiments, the circulating tumor nucleic acid comprises circulating tumor RNA. In some embodiments, the methylation profile comprises a plurality of Differentially Methylated Regions (DMRs). In some embodiments, the plurality of DMRs is ctDNA derived. In some embodiments, a plurality of DMRs derived from peripheral blood leukocytes is removed from said methylation profile. In some embodiments, the plurality of DMRs comprises at least about 56 genomic regions with hypo-methylation levels compared to corresponding genomic regions from a normal healthy subject. In some embodiments, the plurality of DMRs comprises at least about 941 genomic regions with hyper-methylation levels compared to corresponding genomic regions from a normal healthy subject. In some embodiments, a DMR comprises a size of at least about 300 bp. In some embodiments, a DMR comprises a size of at least about 100 bp to at least about 200 bp. In some embodiments, a DMR comprises a size of at least about 100 bp to at least about 150 bp. In some embodiments, a DMR comprises at least 8 CpG genomic islands. In some embodiments, the normal healthy subject comprises a same set of risk factors as said subject.


In some embodiments, the mutation profile comprises a missense variant, a nonsense variant, a deletion variant, an insertion variant, a duplication variant, an inversion variant, a frameshift variant, or a repeat expansion variant. In some embodiments, any variant that is present in a genomic DNA sample obtained from a plurality of peripheral blood leukocytes, wherein said plurality of peripheral blood leukocytes is obtained from said subject, is removed from the mutation profile. In some embodiments, any variant that is derived from clonal hematopoiesis is removed from said mutation profile. In some embodiments, the mutation profile does not comprise a variant of gene DNMT3A, TET2, or ASXL1. In some embodiments, the mutation profile does not comprise a canonical cancer driver gene. In some embodiments, the mutation profile comprises non-canonical cancer driver gene, where said non-canonical gene is GRIN3A or MYC.


In some embodiments, the fragment length profile comprises selecting cell free nucleic acid molecules based on a range of fragment length of about at least 80 bp to 170 bp. In some embodiments, the fragment length profile comprises selecting cell free nucleic acid molecules based on a range of fragment length of about at least 100 bp to 150 bp. In some embodiments, the circulating tumor nucleic acid molecules are enriched.


In some embodiments, the methods further comprise mixing said cell free nucleic acid sample with a filler DNA molecules to yield a DNA mixture. In some embodiments, the filler DNA molecules comprise a length of about 50 bp to 800 bp. In some embodiments, the filler DNA molecules comprise a length of about 100 bp to 600 bp. In some embodiments, the filler DNA molecules comprises at least about 5% methylated filler DNA molecules. In some embodiments, the filler DNA molecules comprises at least about 20% methylated filler DNA. In some embodiments, the filler DNA molecules comprises at least about 30% methylated filler DNA. In some embodiments, the filler DNA molecules comprises at least about 50% methylated filler DNA.


In some embodiments, the methods further comprise incubating said DNA mixture with a binder that is configured to bind methylated nucleotides to generate an enriched sample. In some embodiments, the binder comprises a protein comprising a methyl-CpG-binding domain. In some embodiments, the protein is a MBD2 protein. In some embodiments, the binder comprises an antibody. In some embodiments, the antibody is a 5-MeC antibody. In some embodiments, the antibody is a 5-hydroxymethyl cytosine antibody. In some embodiments, the sequencing does not comprise bisulfite sequencing. In some embodiments, the cell-free nucleic acid sample comprises a blood sample. In some embodiments, the blood sample comprises a plasma sample. In some embodiments, the methods further comprise detecting an origin of cancer tissue.


In some embodiments, the methods further comprise generating a report comprising a prognosis of said subject's survival rate. In some embodiments, the methods further comprise providing a treatment to said subject. In some embodiments, subsequent to treatment of said disease, the methods further comprise providing a second report indicating whether said treatment is effective.


In another aspect, the present disclosure provides methods for determining whether a subject has or is at risk of having a condition, comprising: assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject; detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 5; and comparing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 5.


In some embodiments, the cell-free nucleic acid molecule comprises ctDNA. In some embodiments, the methods comprise performing the sequence analysis, and wherein said sequencing analysis comprises a cell-free methylated DNA immunoprecipitation (cfMeDIP) sequencing. In some embodiments, the detecting comprises measuring a methylation level of at least a portion of said nucleic acid molecule comprised in: six or more, ten or more, fifteen or more, twenty or more, thirty or more, forty or more, fifty or more, sixty or more, seventy or more, eighty or more, ninety or more, or one hundred or more DMRs listed in Table 5.


In another aspect, the present disclosure provides methods method for determining whether a subject has a higher survival rate after receiving a treatment for a disease, comprising: assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject; detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 6; and processing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 6.


In some embodiments, the cell-free nucleic acid molecule comprises ctDNA. In some embodiments, the detecting comprises providing a composite methylation score (CMS). In some embodiments, the CMS comprises a sum of beta-values of DMRs listed in Table 6. In some embodiments, a higher CMS indicates an inferior survival for said subject. In some embodiments, the CMS is not dependent on an abundance of ctDNA. In some embodiments, the disease is squamous cell carcinoma. In some embodiments, the cancer is head and neck squamous cell carcinoma.


In another aspect, the present disclosure provides systems for determining whether a subject has or is at risk of having a disease, comprising one or more computer processors that are individually or collectively programmed to implement a process comprising: subjecting a plurality of nucleic acid molecules derived from a cell-free nucleic acid sample obtained from said subject to sequencing to generate at least one profile of (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and processing said at least one profile to determine whether said subject has or is at risk of said disease at a sensitivity of at least 80% or at a specificity of at least about 90%, wherein said cell-free nucleic acid sample comprises less than 30 ng/ml of said plurality of nucleic acid molecules.


In another aspect, the present disclosure provides systems for processing a cell-free nucleic acid sample of a subject to determine whether said subject has or is at risk of having a disease, comprising one or more computer processors that are individually or collectively programmed to implement a process comprising: providing said cell-free nucleic acid sample comprising a plurality of nucleic acid molecules; subjecting said plurality of nucleic acid molecules or derivatives thereof to sequencing to generate a plurality of sequencing reads; computer processing said plurality of sequencing reads to identify, for said plurality of nucleic acid molecules, (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and using at least said methylation profile, said mutation profile and said fragment length profile to determine whether said subject has or is at risk of having said disease.





BRIEF DESCRIPTION OF FIGURES

The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. These and other features of the preferred embodiments of the invention will become more apparent in the following detailed description in which reference is made to the appended drawings wherein:



FIG. 1. Utilization of PBL-filtering for detection of ctDNA by CAPP-Seq. A) Mutant allele fraction of candidate SNVs identified in matched patient plasma and/or PBLs. Pearson's correlation was performed on SNVs strictly found in both matched patient plasma and PBLs. Candidate SNVs found only in patient plasma are denoted within the dashed red box. B) Oncoprint of candidate SNVs identified in both matched patient plasma and PBLs. The top histogram denotes the number of SNVs per patient whereas the right histogram denotes the number of patients with a specified gene mutated. C) Mean MAF of candidate SNVs across HNSCC patient cfDNA (red circle) and PBL (blue circle) before and after removal of PBL-associated SNVs. Patients with SNVs absent after PBL filtering are indictive of false positive detection of ctDNA. E) Oncoprint of selected PBL-filtered SNVs identified in 20/32 HNSCC patients. The top and right histograms denote that as previously described in (B). F) Mean mutant allele percentage of PBL-filtered SNVs across all HNSCC patients. For each SNV per patient, the mutant allele percentage was calculated by the fraction of reads containing the SNV of interest, compared to reads that contained the native sequence overlapping the SNV base-pair position



FIG. 2. Utilization of PBL-filtering for detection of ctDNA by CAPP-Seq. B) Mutant allele fraction of candidate SNVs identified in matched patient plasma and/or PBLs. Pearson's correlation was performed on SNVs strictly found in both matched patient plasma and PBLs. Candidate SNVs found only in patient plasma are denoted within the dashed red box. C) Oncoprint of candidate SNVs identified in both matched patient plasma and PBLs. The top histogram denotes the number of SNVs per patient whereas the right histogram denotes the number of patients with a specified gene mutated. D) Mean MAF of candidate SNVs across HNSCC patient cfDNA (red circle) and PBL (blue circle) before and after removal of PBL-associated SNVs. Patients with SNVs absent after PBL filtering are indictive of false positive detection of ctDNA. E) Oncoprint of selected PBL-filtered SNVs identified in 20/32 HNSCC patients. The top and right histograms denote that as previously described in (B). F) Mean mutant allele percentage of PBL-filtered SNVs across all HNSCC patients. For each SNV per patient, the mutant allele percentage was calculated by the fraction of reads containing the SNV of interest, compared to reads that contained the native sequence overlapping the SNV base-pair position.



FIG. 3. Identification of informative regions for detection of ctDNA by cfMeDIP-seq. B) Pearson's correlation of 300-bp non-overlapping windows with >=8 CpGs from patient and healthy donor cfDNA cfMeDIP-seq profiles (n=52) against FaDu genomic DNA (gDNA) [1×1×52 comparisons], unmatched PBL gDNA [1×51×52 comparisons], and matched PBL gDNA [1×1×52 comparisons] MeDIP-seq profiles. C) Performance of in-silico PBL-depletion in healthy donor (right) and HNSCC (left) PBL MeDIP-seq profiles. Absolute methylation scores were calculated from MeDIP-seq counts via MeDEStrand (Methods). 300-bp non-overlapping windows before PBL-depletion (blue) correspond with all windows from chromosome 1-22 with >=8 CpGs (n=702,488). 300-bp non-overlapping windows after PBL-depletion (red) include an additional filter where the median absolute methylation across healthy donor PBLs is <0.1 (n=99,997). D) Workflow of ctDNA detection by differential methylation analysis of HNSCC and healthy donor cfMeDIP-seq profiles. cfMeDIP-seq profiles from HNSCC patients with detectable SNVs by CAPP-Seq (i.e. CAPP-Seq positive, n=20) were compared to healthy donors (n=20) within PBL-depleted windows to identify HNSCC-associated cfDNA methylation. Hyper- and hypo-methylated regions are denoted as regions with higher or lower methylation in the HNSCC cohort compared to healthy donors at an FDR<10%. E) Permutation analysis of hyper-methylated regions annotated by CpG site (n=10,000 total permutations). Significant enrichment/depletion is denoted as observed z-scores with a p-value less than 0.05. F, Permutation analysis of hyper-methylated regions within tumor-specific methylated cytosines from TCGA (n=1000 permutations total). Significant enrichment/depletion is denoted as observed z-scores with a p-value less than 0.05.



FIG. 4. Concordance of ctDNA detection and abundance between CAPP-Seq and cfMeDIP-seq profiles. A) Median fragment length of detected SNVs across HNSCC patients by CAPP-seq. For each patient, the median fragment length of each SNV and matched reference allele was measured. The distribution of median fragment length for each mutation or matched reference allele is shown per patient. Extremes of boxes and centerlines define upper and lower quartiles and medians, respectively. In cases with a single SNV, the coloured line denotes the median length of fragments containing the SNV or matched reference allele, respectively. B) Fragment length distributions within HNSCC hyper-methylated regions by cfMeDIP-seq. Fragment lengths from healthy donors were pooled prior to analysis, where each subsequent box denotes an individual HNSCC cfMeDIP-seq profile. Extremes of boxes and centerlines define upper and lower quartiles and medians, respectively. Individual HNSCC samples are ordered based on increasing mean methylation (RPKM) within the hyper-methylated regions. Dashed blue line defines the median fragment length across all healthy donors. C) Ratio of enrichment for hyper-DMR regions by fragments between 100-150 bp compared to enrichment for hyper-DMR regions by fragments between 100-220 bp. Ratios were converted to percent increase/decrease for ease of interpretation. D) Ratio of enrichment for hyper-DMR regions by fragments between 100-150 bp compared to enrichment for hyper-DMR regions by fragments between 100-220 bp.+ symbols denote HNSCC patients with detectable ctDNA by CAPP-Seq (CAPP-Seq positive). E) Supervised hierarchal classification of cfMeDIP-seq profiles limited to 100-150 bp, by log-transformed RPKM values across HNSCC hyper-methylated regions. RPKM values for each cfMeDIP-seq profile was log 2-transformed prior to Euclidean transformation and clustered using Ward's method. Methylation clusters were defined at a threshold of k=4. F), Relationship of mean mutant allele frequency and mean RPKM from identified SNVs and hyper-methylated regions by CAPP-seq and cfMeDIP-seq (limited to 100-150 bp), respectively. Points denote individual samples from HNSCC or healthy donor plasma. Solid red line and shaded grey area denotes the fitted linear regression model and associated 95% confidence interval, respectively. G) AUROC analysis based on methylation values (limited to 100-150 bp) within HNSCC hyper-methylated regions, comparing HNSCC to healthy donor cfMeDIP-seq profiles. Detection of ctDNA was defined as instances where mean methylation was above the max value across healthy donors. H) Kaplan-Meier curve analysis for overall survival of patients within methylation cluster 1+2+3, compared to methylation cluster 4. I+J) Comparison of median fragment lengths from CAPP-Seq and cfMeDIP-seq profiles (I) and median fragment length from CAPP-Seq and 100-150:151-220 bp ratio from cfMeDIP-seq profiles (J). Points defined individual HNSCC samples within methylation cluster 1 and 2. Solid red line and shaded grey area denotes the fitted linear regression model and 95% confidence interval, respectively.



FIG. 5. Prognostic utility of specific methylated regions within ctDNA detected by cfMeDIP-seq. A) Relationship of mean mutant allele fraction and mean RPKM from identified mutations and hyper-methylated regions by CAPP-seq and cfMeDIP-seq (limited to 100-150 bp), respectively. Points denote individual samples from HNSCC or healthy control plasma. Solid red line: fitted linear regression model. Grey boundaries: 95% confidence interval. B) Kaplan-Meier analysis depicting overall survival of patients with detectable ctDNA both by CAPP-Seq and cfMeDIP-seq (mean methylation above healthy controls within hyper-DMRs) C) Identification of prognostic regions based on disease-specific survival by multivariate Cox Proportional Hazard regression analysis across HNSCC primary tumors provided by the TCGA (n=520). Regions were defined as 300-bp windows as previously described. HumanMethylation450K data was obtained from the TCGA and beta-values from probe IDs overlapping with each region were averaged. Candidate regions for prognostic analysis was selected based on elevated methylation across primary tumors (n=520) compared to solid adjacent normal tissue (n=50) (Wilcoxon's test, adjusted p value <0.05, log 2FC>1). G-H) Spearman's correlation from methylation of a particular 300-bp region (boxes) to the RNA expression of a particular transcript. Regions with an absolute R value >=0.3 (denoted by dashed grey lines) were labeled as significant associations. Methylated regions which were prognostic for disease-specific survival of HNSCC patients provided by the TCGA (n=520) are denoted with a red outline. Prognostic regions which were further associated with RNA expression are denoted as solid red. Example prognostic methylated regions associated with RNA expression; (G) OSR1, (H) LINC01391 are provided. E) Kaplan-Meier curve of overall survival for HNSCC-TCGA patients based on total methylation across five regions affecting expression of ZNF323/ZSCAN1, LINC01391, GATA-AS1, OSR1, and STK3/MST2 respectively. Patients were stratified based on either being below (Blw med. blue) or above (Abv med. red) the median total methylation of the five regions previously identified in (D) across all primary tumors. F) Kaplan-Meier curve of overall survival as described in (E) for HNSCC plasma cohort with detectable ctDNA by CAPP-Seq. To calculate total methylation across the five genes with prognostic association, RPKM values were scaled accordingly across all hyper-DMR regions previously identified prior to survival analysis.



FIG. 6. Clinical utility of ctDNA detection by cfMeDIP-seq for longitudinal monitoring. A) ctDNA kinetics typically observed across patients throughout treatment. Complete clearance was defined as a change from detected ctDNA at diagnosis to a decrease in ctDNA abundance below the threshold of detection (i.e. 0.2%) at first available mid-/post-treatment timepoint. Partial clearance was defined as a change from detected ctDNA at diagnosis to a decrease (>=90%) in ctDNA abundance above the threshold of detection at first available mid-/post-treatment timepoint. No clearance was defined as an increase in ctDNA abundance in mid-/post-treatment samples compared to at diagnosis. lastFU=sample collection at last follow-up, RT=radiotherapy. B) Changes in ctDNA abundance at diagnosis to first available mid-/post-treatment timepoint across HNSCC patients (n=30). Red lines denote patients that demonstrated kinetics of no-clearance, whereas grey lines denote patients with kinetics of clearance/partial-clearance. C, Kaplan-Meier curve of recurrence-free survival. Patients were stratified based on kinetics of clearance (i.e. no clearance vs. clearance/partial clearance).



FIG. 7. Comparison of cfMeDIP-seq analysis performed on all or ctDNA-enriched fragments. ctDNA-enriched fragments are defined as fragments ranging from 100-150 bp in length. A) Mutant allele frequency of mutations identified by CAPP-Seq vs. mean RPKM values of previously identified HNSCC hyperDMRs in cfMeDIP-seq profiles containing all fragments (left) or ctDNA-enriched fragments (right). B) Area under the curve analysis (AUROC) for ctDNA detection in HNSCC cfMeDIP-seq profiles (CAPP-Seq positive only: red, CAPP-Seq positive and negative: blue) versus healthy donors. Results of cross-validation analysis using CAPP-Seq positive patients is also shown (replicates=50). Analysis is shown for cfMeDIP-seq profiles with all fragments (left) or ctDNA-enriched fragments (right). C) Kaplan-Meier analysis for recurrence-free survival based on longitudinal cfMeDIP-seq profiling with all fragments (left) or ctDNA-enriched fragments. Patients were classified as being positive for post-treatment ctDNA if they demonstrated methylation abundance within the previously identified hyperDMRs greater than 0.2% ctDNA.



FIG. 8. shows a computer system that is programmed or otherwise configured to implement methods provided herein FIG. 9. Sample characteristics of isolated cell-free DNA from HNSCC and healthy donors. A) Schematic defining timepoints of blood isolation. B) cfDNA yields (normalized to per mL of plasma) across timepoints for HNSCC patients as well as healthy donors (i.e. “Normal”).



FIG. 10. Analysis of the number of SNVs per HNSCC patient covered by the CAPP-Seq selector assessed either among all 364 patients in the HNSC TCGA cohort (blue diamonds) or using leave-one-out cross-validation (LOOCV; red squares).



FIG. 11. Oncoprint of all PBL-filtered SNVs identified in 20/32 HNSCC patients (Related to FIG. 2E).



FIG. 12. Related figures for identification of informative regions (related to FIGS. 3B and C). A) Median RPKM values of genome-wide (chromosomes 1-22) 300-bp non-overlapping bins based on >=n CpGs. B) Differential methylation analysis between HNSCC and healthy donor PBLs within PBL-depleted windows as described in FIG. 2B and Methods. Hypomethylated regions (i.e. regions with elevated methylation in healthy donor PBLs) are denoted in blue.



FIG. 13. Related figures to results of differential methylation analysis between HNSCC and healthy donor cfDNA samples within PBL-depleted windows (FIG. 2D). A) DMRs were defined based on the original 300-bp non-overlapping windows used for the initial analysis. DMRs immediately adjacent to each other were binned into their respective widths (i.e. two 300-bp windows are each independently defined as having a length of 600-bp). B) Permutation analysis of CpG features as defined in FIG. 2E, based on hypo-methylated regions.



FIG. 14. Supervised hierarchical clustering of TCGA primary tumors based on identified of cancer-specific differentially methylated cytosines. Cancer_type (column) refers to the classification of each primary tumor or PBL sample, whereas cancer_DMCs (row) refers to cancer-specific differentially methylated cytosines identified for each cancer type (PBLs excluded).



FIG. 15. Related figures to FIG. 4. A) Median fragment length of identified SNVS by CAPP-Seq per patient compared to mean mutant allele fraction. B) Median fragment length within hyper-DMRS by cfMeDIP-seq per patient compared to mean RPKM of hyper-DMRs.



FIG. 16. Related figures to CAPP-Seq and cfMeDIP-seq concordance analysis (FIG. 4E). A) Area under the curve values obtained from cross-validation analysis (n=50) of differentially methylated region calling between CAPP-Seq positive HNSCC cfDNA samples and healthy donors. B) Kaplan-Meir analysis for overall survival of HNSCC patients based on the detection of ctDNA by CAPP-Seq. C) and D) mean RPKM and mean mutant allele fraction of HNSCC patient samples stratified based on methylation cluster (FIG. 4D).



FIG. 17. Identification of regions of potential clinical utility (related to FIG. 6). A) Genome-track of genes currently used in commercially available liquid biopsy tests with overlap to HNSCC primary tumors within the TCGA as well as plasma-derived hyper-DMRs from our HNSCC cohort. Bottom dark blue bar with arrows denotes the direction of transcription for the specified gene. Red bars indicate location of 300-bp windows overlapping with hyper-DMRs from plasma of our HNSCC cohort as well as primary tumors from the TCGA. B-D) Spearman's correlation from methylation of a particular 300-bp region (boxes) to the RNA expression of a particular transcript. Regions with an absolute R value >=0.3 (denoted by dashed grey lines) were labeled as significant associations. Methylated regions which were prognostic for disease-specific survival of HNSCC patients provided by the TCGA (n=520) are denoted with a red outline. Prognostic regions which were further associated with RNA expression are denoted as solid red. Figures were generated for all five genes contained prognostic methylated regions associated with RNA expression; (B) GATA2-AS1, (C) ZNF323, (D), STK3.



FIG. 18. Extension of FIG. 6A, displaying changes in ctDNA abundance by cfMeDIP-seq throughout treatment for all HNSCC patients (n=32)





DETAILED DESCRIPTION

In the following description, numerous specific details are set forth to provide a thorough understanding of the invention. However, it is understood that the invention may be practiced without these specific details.


The present disclosure provides methods, systems, and kits for multimodal analysis of ctDNA in determining a likelihood of a subject having cancer with high sensitivity and/or high specificity. Further, the present disclosure provides methods, systems, and kits for detecting minimal residual disease (MRD) after a cancer treatment, and for evaluating whether such cancer treatment is therapeutically effective.


Identification of specific molecular features from ctDNA prior to treatment may inform prognosis and/or be predictive response to therapy, whereas detection of ctDNA after treatment may aid in identification of MRD and aid in identifying patients at high risk of recurrence and/or death. To achieve robust sensitivity, most clinical studies utilize ctDNA detection methods interrogating few regions, matched tumor profiling, and/or cases of high ctDNA abundance. However, for cancers that harbor low levels of ctDNA or lack common/known aberrations across patients, additional strategies may be utilized to achieve similar degrees of sensitivity. Genome-wide profiling techniques may help improve sensitivity by covering considerably more regions; however, the amount of cell-free DNA and sequencing depth required to achieve detection below a fraction of 1% has been cost-prohibitive.


Two tailored genome-wide profiling techniques capable of highly sensitive ctDNA detection have been described. The first, CAncer Personalized Profiling by deep Sequencing (CAPP-Seq), utilizes a broad panel of hybrid-capture probes targeting over 100 genes to identify low allele frequency mutations. The second, cell-free Methylated DNA ImmunoPrecipitation sequencing (cfMeDIP-seq), enriches for methylated cfDNA fragments through use of an anti-5-methylcytosine (anti-5mC) antibody. The identification of mutations or hypermethylation events by these respective methods have their respective advantages. Mutations may distinguish ctDNA from healthy sources of cell-free DNA due to their irreversible disposition, provided that appropriate error suppression tools are employed and any contribution of mutations from clonal hematopoiesis is taken into account. DNA hypermethylation events potentially affect a larger number of recurrent genomic regions in cancer, contributing to their ability to inform the tumor-of-origin through cell-free DNA analysis. Moreover, hypermethylation events in the vicinity of cancer driver genes may influence their expression, thereby potentially reflecting cancer behavior and providing prognostic value. To date no study has utilized the combination of both mutation- and methylation-based methods for improved tumor-naïve detection and characterization of ctDNA in localized cancers.


Utilization of fluid-based biomarkers for prognostication, risk stratification, and disease surveillance may improve patient outcomes by guiding treatment decisions without the need for invasive tumor sampling. Although circulating tumor (ct)DNA in particular has shown promise as a liquid biopsy tool, in patients with low disease burden such as those with localized non-metastatic cancer, paired tumor profiling is often required. We hypothesized that multimodal analysis of genetic and epigenetic features from plasma cell-free DNA may enable broad applications of tumor-naïve ctDNA profiling. Mutation- and methylation-based profiling identified ctDNA in 65% of localized head and neck cancer patients. Results from both approaches were quantitative and strongly correlated, and their combined analysis revealed common features of tumor-derived DNA fragments. Moreover, ctDNA methylomes revealed tumor histology, putative prognostic biomarkers, and dynamic patterns of treatment response. These findings will aid future non-invasive biomarker discovery efforts and will inform clinical implementation of ctDNA for localized cancers.


Certain methods of capturing cell-free methylated DNA are described in Applicant's WO 2017/190215 and WO 2019/010564, both of which are incorporated by reference.


Specifically, we utilize both CAPP-Seq and cfMeDIP-seq to perform tumor-naïve ctDNA detection within a cohort of localized head and neck squamous cell carcinoma (HNSCC) patients. HNSCC is a clinically heterogenous disease that frequently recurs after definitive treatment and may benefit greatly from ctDNA detection to better inform treatment decisions and disease management33. We demonstrate that utilization of both methods in parallel, as well as matched PBL-profiling, may achieve high-confidence tumor-naïve ctDNA detection. Furthermore, we show that the combined analysis reveals common molecular features of tumor-derived DNA fragments. Finally, we show that ctDNA methylomes revealed tumor histology, putative prognostic biomarkers, and dynamic patterns of treatment response, providing a blueprint for future biomarker studies in other disease settings


In an aspect, there is provided a method of detecting the presence of ctDNA from cancer cells in a subject comprising:

    • (a) providing a sample of cell-free DNA from a subject;
    • (b) subjecting the sample to library preparation to permit subsequent sequencing of the cell-free methylated DNA;
    • (c) optionally adding a first amount of filler DNA to the sample, wherein at least a portion of the filler DNA is methylated, then further optionally denaturing the sample;
    • (d) capturing cell-free methylated DNA using a binder selective for methylated polynucleotides;
    • (e) sequencing the captured cell-free methylated DNA;
    • (f) comparing the sequences of the captured cell-free methylated DNA to control cell-free methylated DNAs sequences from healthy and cancerous individuals;
    • (g) identifying the presence of DNA from cancer cells if there is a statistically significant similarity between one or more sequences of the captured cell-free methylated DNA and cell-free methylated DNAs sequences from cancerous individuals;
    • wherein in at least one of the capturing step, the comparing step or the identifying step, the subject cell-free methylated DNA is limited to a sub-population according to a fragment length metric.


Various sequencing techniques are known to the person skilled in the art, such as polymerase chain reaction (PCR) followed by Sanger sequencing. Also available are next-generation sequencing (NGS) techniques, also known as high-throughput sequencing, which includes various sequencing technologies including: Illumina (Solexa) sequencing, Roche 454 sequencing, Ion torrent: Proton/PGM sequencing, SOLiD sequencing, long reads sequencing (Oxford Nanopore and Pactbio). NGS allow for the sequencing of DNA and RNA much more quickly and cheaply than the previously used Sanger sequencing. In some embodiments, said sequencing is optimized for short read sequencing.


The term “subject” as used herein refers to any member of the animal kingdom. Thus, the methods and described herein are applicable to both human and veterinary disease and animal models. Preferred subjects are “patients,” i.e., living humans that are being investigated to determine whether treatment or medical care is needed for a disease or condition; or that are receiving medical care for a disease or condition (e.g., cancer).


The term “genome,” as used herein, generally refers to genomic information from a subject, which may be, for example, at least a portion or an entirety of a subject's hereditary information. A genome can be encoded either in DNA or in RNA. A genome can comprise coding regions (e.g., that code for proteins) as well as non-coding regions. A genome can include the sequence of all chromosomes together in an organism. For example, the human genome ordinarily has a total of 46 chromosomes. The sequence of all of these together may constitute a human genome.


The term “nucleic acid” used herein refers to a polynucleotide comprising two or more nucleotides, i.e., a polymeric form of nucleotides of any length, either deoxyribonucleotides (dNTPs) or ribonucleotides (rNTPs), or analogs thereof. Non-limiting examples of nucleic acids include deoxyribonucleic (DNA), ribonucleic acid (RNA), coding or non-coding regions of a gene or gene fragment, loci (locus) defined from linkage analysis, exons, introns, messenger RNA (mRNA), transfer RNA, ribosomal RNA, short interfering RNA (siRNA), short-hairpin RNA (shRNA), micro-RNA (miRNA), ribozymes, cDNA, recombinant nucleic acids, branched nucleic acids, plasmids, vectors, isolated DNA of any sequence, isolated RNA of any sequence, nucleic acid probes, and primers. A nucleic acid may comprise one or more modified nucleotides, such as methylated nucleotides and nucleotide analogs. If present, modifications to the nucleotide structure may be made before or after assembly of the nucleic acid. The sequence of nucleotides of a nucleic acid may be interrupted by non-nucleotide components. A nucleic acid may be further modified after polymerization, such as by conjugation or binding with a reporter agent. A “variant” nucleic acid is a polynucleotide having a nucleotide sequence identical to that of its original nucleic acid except having at least one nucleotide modified, for example, deleted, inserted, or replaced, respectively. The variant may have a nucleotide sequence at least about 80%, 90%, 95%, or 99%, identity to the nucleotide sequence of the original nucleic acid.


Cell-free methylated DNA is DNA that is circulating freely in the blood stream, and are methylated at various regions of the DNA. Samples, for example, plasma samples may be taken to analyze cell-free methylated DNA. Studies reveal that much of the circulating nucleic acids in blood arise from necrotic or apoptotic cells and greatly elevated levels of nucleic acids from apoptosis is observed in diseases such as cancer. Particularly for cancer, where the circulating DNA bears hallmark signs of the disease including mutations in oncogenes, microsatellite alterations, and, for certain cancers, viral genomic sequences, DNA or RNA in plasma has become increasingly studied as a potential biomarker for disease. For example, a quantitative assay for low levels of circulating tumor DNA in total circulating DNA may serve as a better marker for detecting the relapse of colorectal cancer compared with carcinoembryonic antigen, the standard biomarker used clinically. The circulating cfDNA may comprise circulating tumor DNA (ctDNA).


As used herein, “library preparation” includes list end-repair, A-tailing, adapter ligation, or any other preparation performed on the cell free DNA to permit subsequent sequencing of DNA.


As used herein, “filler DNA” may be noncoding DNA or it may consist of amplicons.


In some embodiments, the fragment length metric is fragment length. In some preferable embodiments, the subject cell-free methylated DNA is limited to fragments having a length of <170 bp, <165 bp, <160 bp, <155 bp, <150 bp, <145 bp, <140 bp, <135 bp, <130 bp, <125 bp, <120 bp, <115 bp, <110 bp, <105 bp, or <100 bp. In other preferable embodiments, the subject cell-free methylated DNA is limited to fragments having a length of between about 100-about 150 bp, 110-140 bp, or 120-130 bp.


In some embodiments, the fragment length metric is the fragment length distribution of the subject cell-free methylated DNA. In some preferable embodiments, the subject cell-free methylated DNA is limited to fragments within the bottom 50th, 45th, 40th, 35th, 30th, 25th, 20th, 15th or 10th percentile based on length.


In some embodiments, the subject cell-free methylated DNA is further limited to fragments within Differentially Methylated Regions (DMRs).


In some embodiments, the limiting of the subject cell-free methylated DNA is during the capturing step.


In some embodiments, the limiting of the subject cell-free methylated DNA is during the comparing step.


In some embodiments, the limiting of the subject cell-free methylated DNA is during the identifying step.


In some embodiments, the comparison step is based on fit using a statistical classifier. Statistical classifiers using DNA methylation data may be used for assigning a sample to a particular disease state, such as cancer type or subtype. For the purpose of cancer type or subtype classification, a classifier would consist of one or more DNA methylation variables (i.e., features) within a statistical model, and the output of the statistical model would have one or more threshold values to distinguish between distinct disease states. The particular feature(s) and threshold value(s) that are used in the statistical classifier may be derived from prior knowledge of the cancer types or subtypes, from prior knowledge of the features that are likely to be most informative, from machine learning, or from a combination of two or more of these approaches.


In some embodiments, the classifier is machine learning-derived. Preferably, the classifier is an elastic net classifier, lasso, support vector machine, random forest, or neural network.


The genomic space that is analyzed may be genome-wide, or preferably restricted to regulatory regions (i.e., FANTOM5 enhancers, CpG Islands, CpG shores and CpG Shelves).


Preferably, the percentage of spike-in methylated DNA recovered is included as a covariate to control for pulldown efficiency variation.


For a classifier capable of distinguishing multiple cancer types (or subtypes) from one another, the classifier would preferably consist of differentially methylated regions from pairwise comparisons of each type (or subtype) of interest.


In some embodiments, the control cell-free methylated DNAs sequences from healthy and cancerous individuals are comprised in a database of Differentially Methylated Regions (DMRs) between healthy and cancerous individuals.


In some embodiments, the control cell-free methylated DNA sequences from healthy and cancerous individuals are limited to those control cell-free methylated DNA sequences which are differentially methylated as between healthy and cancerous individuals in DNA derived from cell-free DNA from bodily fluids, such as from blood serum, cerebral spinal fluid, urine stool, sputum, pleural fluid, ascites, tears, sweat, pap smear fluid, endoscopy brushings fluid, . . . etc., preferably from blood plasma.


Samples


A sample can be any biological sample isolated from a subject. For example, a sample may comprise, without limitation, bodily fluid, whole blood, platelets, serum, plasma, stool, red blood cells, white blood cells or leukocytes, endothelial cells, tissue biopsies, synovial fluid, lymphatic fluid, ascites fluid, interstitial or extracellular fluid, the fluid in spaces between cells, including gingival crevicular fluid, bone marrow, cerebrospinal fluid, saliva, mucous, sputum, semen, sweat, urine, fluid from nasal brushings, fluid from a pap smear, or any other bodily fluids. A bodily fluid may include saliva, blood, or serum. A sample may also be a tumor sample, which may be obtained from a subject by various approaches, including, but not limited to, venipuncture, excretion, ejaculation, massage, biopsy, needle aspirate, lavage, scraping, surgical incision, or intervention or other approaches. A sample may be a cell-free sample (e.g., substantially free of cells). DNA samples may be denatured, for example, using sufficient heat.


In some embodiments, the present disclosure provides a system, method, or kit that includes or uses one or more biological samples. The one or more samples used herein may comprise any substance containing or presumed to contain nucleic acids. A sample may include a biological sample obtained from a subject. In some embodiments, a biological sample is a liquid sample.


In some embodiments, the sample comprises less than about 100 ng, 90 ng, 80 ng, 75 ng, 70 ng, 60 ng, 50 ng, 40 ng, 30 ng, 20 ng, 10 ng, 5 ng, 1 ng or any amount in between the numbers of cell-free nucleic acid molecules. Further, in some embodiments, the sample comprises less than about 1 pg, less than about 5 pg, less than about 10 pg, less than about 20 pg, less than about 30 pg, less than about 40 pg, less than about 50 pg, less than about 100 pg, less than about 200 pg, less than about 500 pg, less than about 1 ng, less than about 5 ng, less than about 10 ng, less than about 20 ng, less than about 30 ng, less than about 40 ng, less than about 50 ng, less than about 100 ng, less than about 200 ng, less than about 500 ng, less than about 1000 ng, or any amount in between the numbers of cell-free nucleic acid molecules.


In some embodiments, the present disclosure comprises methods and systems for filling in the sample with a amount of filler DNA to generate a mixture sample, wherein the mixture sample comprises at least about 50 ng, 55 ng, 60 ng, 65 ng, 70 ng, 75 ng, 80 ng, 85 ng, 90 ng, 95 ng, 100 ng, 120 ng, 140 ng, 160 ng, 180 ng, 200 ng, or any amount in between the numbers of the total amount of the nucleic acid mixture. In some embodiments, the filler DNA comprises at least about 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% methylated filler DNA with remainder being unmethylated filler DNA, and preferably between 5% and 50%, between 10%-40%, or between 15%-30% methylated filler DNA. In some embodiments, the mixture sample comprise an amount of filler DNA from 20 ng to 100 ng, preferably 30 ng to 100 ng, more preferably 50 ng to 100 ng. In some embodiments, the cell-free DNA from the sample and the first amount of filler DNA together comprises at least 50 ng of total DNA, preferably at least 100 ng of total DNA.


In some embodiments, the filler DNA is 50 bp to 800 bp long, preferably 100 bp to 600 bp long, and more preferably 200 bp to 600 bp long. In some embodiments, the filler DNA is double stranded. The filler DNA is double stranded. For example, the filler DNA can be junk DNA. The filler DNA may also be endogenous or exogenous DNA. For example, the filler DNA is non-human DNA, and in preferred embodiments, λ DNA. As used herein, “λ DNA” refers to Enterobacteria phage λ DNA. In some embodiments, the filler DNA has no alignment to human DNA.


In some embodiments, the sample may be taken before and/or after treatment of a subject with a disease or disorder. Samples may be obtained from a subject during a treatment or a treatment regime. Multiple samples may be obtained from a subject to monitor the effects of the treatment over time. The sample may be taken from a subject known or suspected of having a disease or disorder for which a definitive positive or negative diagnosis is not available via clinical tests. The sample may be taken from a subject suspected of having a disease or disorder. The sample may be taken from a subject experiencing unexplained symptoms, such as fatigue, nausea, weight loss, aches and pains, weakness, or bleeding. The sample may be taken from a subject having explained symptoms. The sample may be taken from a subject at risk of developing a disease or disorder due to factors such as familial history, age, hypertension or pre-hypertension, diabetes or pre-diabetes, overweight or obesity, environmental exposure, lifestyle risk factors (e.g., smoking, alcohol consumption, or drug use), or presence of other risk factors.


In some embodiments, a sample may be taken at a first time point and sequenced, and then another sample may be taken at a subsequent time point and sequenced. Such methods may be used, for example, for longitudinal monitoring purposes to track the development or progression of a disease. In some embodiments, the progression of a disease may be tracked before treatment, after treatment, or during the course of treatment, to determine the treatment's effectiveness. For example, a method as described herein may be performed on a subject prior to, and after, a medical treatment to measure the disease's progression or regression in response to the medical treatment.


After obtaining a sample from the subject, the sample may be processed to generate datasets indicative of a disease or disorder of the subject. For example, a presence, absence, or quantitative assessment of cell-free nucleic acid molecules (e.g., ctDNA molecules) of the sample at a panel of cancer-associated genomic loci or microbiome-associated loci may be indicative of a cancer of the subject. Processing the sample obtained from the subject may comprise (i) subjecting the sample to conditions that are sufficient to isolate, enrich, or extract a plurality of cell-free nucleic acid molecules, and (ii) assaying the plurality of cell-free nucleic acid molecules to generate the dataset (e.g., nucleic acid sequences). In some embodiments, a plurality of cell-free nucleic acid molecules is extracted from the sample and subjected to sequencing to generate a plurality of sequencing reads.


In some embodiments, the cell-free nucleic acid molecules may comprise cell-free ribonucleic acid (cfRNA) or cell-free deoxyribonucleic acid (cfDNA). The cell-free nucleic acid molecules (e.g., cfRNA or cfDNA) may be extracted from the sample by a variety of methods. The cell-free nucleic acid molecule may be enriched by a plurality of probes configured to enrich nucleic acid (e.g., RNA or DNA) molecules corresponding to a panel of cancer-associated genomic loci. The probes may have sequence complementarity with nucleic acid sequences from one or more of the panel of cancer-associated genomic loci. The panel of cancer-associated genomic loci may comprise at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least about 25, at least about 30, at least about 35, at least about 40, at least about 45, at least about 50, at least about 55, at least about 60, at least about 65, at least about 70, at least about 75, at least about 80, at least about 85, at least about 90, at least about 95, at least about 100, or more distinct cancer-associated genomic loci. The probes may be nucleic acid molecules (e.g., RNA or DNA) having sequence complementarity with nucleic acid sequences (e.g., RNA or DNA) of the one or more genomic loci (e.g., cancer-associated genomic loci). These nucleic acid molecules may be primers or enrichment sequences. The assaying of the sample using probes that are selective for the one or more genomic loci (e.g., cancer-associated genomic loci or microbiome-associated loci) may comprise use of array hybridization, polymerase chain reaction (PCR), or nucleic acid sequencing (e.g., RNA sequencing or DNA sequencing).


Nucleic Acid Molecules Sequencing


The present disclosure provides methods and technologies for determining the sequence of nucleotide bases in one or more polynucleotides. The polynucleotides may be, for example, nucleic acid molecules such as deoxyribonucleic acid (DNA) or ribonucleic acid (RNA), including variants or derivatives thereof (e.g., single stranded DNA). Sequencing may be performed by various systems currently available, such as, without limitation, a sequencing system by Illumina®, Pacific Biosciences (PacBio®), Oxford Nanopore®, or Life Technologies (Ion Torrent®). Further, any sequencing methods that provides fragment length such as pair-end sequencing may be utilized. Alternatively or in addition, sequencing may be performed using nucleic acid amplification, polymerase chain reaction (PCR) (e.g., digital PCR, quantitative PCR, or real time PCR), or isothermal amplification. Such systems may provide a plurality of raw genetic data corresponding to the genetic information of a subject (e.g., human), as generated by the systems from a sample provided by the subject. In some examples, such systems provide sequencing reads (also “reads” herein). A read may include a string of nucleic acid bases corresponding to a sequence of a nucleic acid molecule that has been sequenced. In some situations, systems and methods provided herein may be used with proteomic information.


In some embodiments, the sequencing reads are obtained via a next-generation sequencing method or a next-next-generation sequencing method. In some embodiments, the sequencing methods comprises CAncer Personalized Profiling by deep Sequencing (CAPP-Seq), which is a next-generation sequencing based method used to quantify circulating DNA in cancer (ctDNA). This method may be generalized for any cancer type that is known to have recurrent mutations and may detect one molecule of mutant DNA in 10,000 molecules of healthy DNA. In some embodiments, the sequencing methods comprise cfMeDIP sequencing as described by Shen et al., sensitive tumor detection and classification using plasma cell-free DNA methylomes, (2018) Nature, which is incorporated herein in its entirety. In some embodiments, the sequencing comprises bisulfite sequencing.


In some embodiments, sequencing comprises modification of a nucleic acid molecule or fragment thereof, for example, by ligating a barcode, a unique molecular identifier (UMI), or anothertag to the nucleic acid molecule or fragment thereof. Ligating a barcode, UMI, or tag to one end of a nucleic acid molecule or fragment thereof may facilitate analysis of the nucleic acid molecule or fragment thereof following sequencing. In some embodiments, a barcode is a unique barcode (e.g., a UMI). In some embodiments, a barcode is non-unique, and barcode sequences may be used in connection with endogenous sequence information such as the start and stop sequences of a target nucleic acid (e.g., the target nucleic acid is flanked by the barcode and the barcode sequences, in connection with the sequences at the beginning and end of the target nucleic acid, creates a uniquely tagged molecule). A barcode, UMI, or tag may be a known sequence used to associate a polynucleotide or fragment thereof with an input or target nucleic acid molecule or fragment thereof. A barcode, UMI, or tag may comprise natural nucleotides or non-natural (e.g., modified) nucleotides (e.g., as described herein). A barcode sequence may be contained within an adapter sequence such that the barcode sequence may be contained within a sequencing read. A barcode sequence may comprise at least 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, or more nucleotides in length. In some cases, a barcode sequence may be of sufficient length and may be sufficiently different from another barcode sequence to allow the identification of a sample based on a barcode sequence with which it is associated. A barcode sequence, or a combination of barcode sequences, may be used to tag and subsequently identify an “original” nucleic acid molecule or fragment thereof (e.g., a nucleic acid molecule or fragment thereof present in a sample from a subject). In some cases, a barcode sequence, or a combination of barcode sequences, is used in conjunction with endogenous sequence information to identify an original nucleic acid molecule or fragment thereof. For example, a barcode sequence, or a combination of barcode sequences, may be used with endogenous sequences adjacent to a barcode, UMI, or tag (e.g., the beginning and end of the endogenous sequences).


Processing a nucleic acid molecule or fragment thereof may comprise performing nucleic acid amplification. For example, any type of nucleic acid amplification reaction may be used to amplify a target nucleic acid molecule or fragment thereof and generate an amplified product. Non-limiting examples of nucleic acid amplification methods include reverse transcription, primer extension, polymerase chain reaction (PCR), ligase chain reaction, asymmetric amplification, rolling circle amplification, and multiple displacement amplification (MDA). Examples of PCR include, but are not limited to, quantitative PCR, real-time PCR, digital PCR, emulsion PCR, hot start PCR, multiplex PCR, asymmetric PCR, nested PCR, and assembly PCR. Nucleic acid amplification may involve one or more reagents such as one or more primers, probes, polymerases, buffers, enzymes, and deoxyribonucleotides. Nucleic acid amplification may be isothermal or may comprise thermal cycling. and/or with the length of the endogenous sequence.


Methylation Profile


The present disclosure provides methods, systems, and kits for producing a methylation profile of a subject that has a disease/condition or is suspected of having such disease/condition, wherein the methylation profile may be used to determine whether the subject has the disease/condition or is at risk of having the disease/condition. Before using cfMeDIP-seq, the samples disclosed herein are subjected to library preparation. In short, after end-repair and A-tailing, the samples are ligated to nucleic acid adapters and digested using enzymes. As described above under the sample section, the prepared libraries may be combined with filler nucleic acids (e.g., filler λ DNAs) to minimize the effect of low abundance ctDNA in the prepared libraries and generate mixed samples. In some embodiments, when the disease/condition is a locoregionally (non-metastatic) cancer, the amount of ctDNA is low and may not be easily and accurately measured and quantified. The mixed samples are brought to at least about 50 ng, 80 ng, 100 ng, 120 ng, 150 ng, or 200 ng and are subjected to further enrichment.


The methods, system, and kits described herein are applicable to a wide variety of cancers, including but not limited to adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain/cns tumors, breast cancer, castleman disease, cervical cancer, colon/rectum cancer, endometrial cancer, esophagus cancer, ewing family of tumors, eye cancer, gallbladder cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal tumor (gist), gestational trophoblastic disease, hodgkin disease, kaposi sarcoma, kidney cancer, laryngeal and hypopharyngeal cancer, leukemia (acute lymphocytic, acute myeloid, chronic lymphocytic, chronic myeloid, chronic myelomonocytic), liver cancer, lung cancer (non-small cell, small cell, lung carcinoid tumor), lymphoma, lymphoma of the skin, malignant mesothelioma, multiple myeloma, myelodysplastic syndrome, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, non-hodgkin lymphoma, oral cavity and oropharyngeal cancer, osteosarcoma, ovarian cancer, penile cancer, pituitary tumors, prostate cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, sarcoma—adult soft tissue cancer, skin cancer (basal and squamous cell, melanoma, merkel cell), small intestine cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, uterine sarcoma, vaginal cancer, vulvar cancer, waldenstrom macroglobulinemia, wilms tumor. In an embodiment, the cancer is head and neck squamous cell carcinoma.


A binder may be used to enrich the mixed samples. In some embodiments, the binder is a protein comprising a Methyl-CpG-binding domain. One such exemplary protein is MBD2 protein. As used herein, “Methyl-CpG-binding domain (MBD)” refers to certain domains of proteins and enzymes that is approximately 70 residues long and binds to DNA that contains one or more symmetrically methylated CpGs. The MBD of MeCP2, MBD1, MBD2, MBD4 and BAZ2 mediates binding to DNA, and in cases of MeCP2, MBD1 and MBD2, preferentially to methylated CpG. Human proteins MECP2, MBD1, MBD2, MBD3, and MBD4 comprise a family of nuclear proteins related by the presence in each of a methyl-CpG-binding domain (MBD). Each of these proteins, with the exception of MBD3, is capable of binding specifically to methylated DNA.


In other embodiments, the binder is an antibody and capturing cell-free methylated DNA comprises immunoprecipitating the cell-free methylated DNA using the antibody. As used herein, “immunoprecipitation” refers a technique of precipitating an antigen (such as polypeptides and nucleotides) out of solution using an antibody that specifically binds to that particular antigen. This process may be used to isolate and concentrate a particular protein or DNA from a sample and requires that the antibody be coupled to a solid substrate at some point in the procedure. The solid substrate includes for examples beads, such as magnetic beads. Other types of beads and solid substrates may be used.


One exemplary antibody is 5-MeC antibody. For the immunoprecipitation procedure, in some embodiments at least 0.05 μg of the antibody is added to the sample; while in more preferred embodiments at least 0.16 μg of the antibody is added to the sample. To confirm the immunoprecipitation reaction, in some embodiments the method described herein further comprises the step of adding a second amount of control DNA to the sample.


The enriched samples are further amplified, purified, and sequenced to generate a plurality of sequence reads. The plurality of sequence reads is analyzed to identify a plurality of Differentially Methylated Regions (DMRs). In some embodiments, the plurality of DMRs comprises DMRs derived from cell free nucleic acid molecules that are derived from peripheral blood leukocytes (PBLs). In some embodiments, the plurality of DMRs comprises at least about 750,000 non-overlapping about 300-bp nucleic acid fragment window. These fragments comprise greater than or equal to 8 CpG islands. In some embodiments, DMRs are identified from comparing sequence reads generated from samples obtained from patients with the disease/condition to sequence reads generated from samples obtained from healthy controls. In some embodiments, the healthy controls comprise a same set of risk factors for developing the disease/condition. In some embodiments, the plurality of DMRs comprises at least about 997 DMRs: about 941 hypermethylated in HNSCC and 56 hypomethylated in HNSCC (Table 5). Using the same disclosed approach here, hypermethylated DMRs may be detected for a different cancer (e.g., lung cancer, pancreatic cancer, colorectal cancer) and hypomethylated DMRs may be detected for the different cancer.









TABLE 5







A list of ctDNA derived DMRs











DMR


windowPos (Genomic
ensemblId Gene ID (a
methylation


position of each DMR)
DMR related to a gene)
level





chr1.50881501.50881800
ENSG00000142700
hyper


chr1.50881801.50882100
ENSG00000142700
hyper


chr1.63786301.63786600
ENSG00000230798
hyper


chr1.119527501.119527800
ENSG00000092607
hyper


chr1.119550601.119550900
ENSG00000239216
hyper


chr1.148603801.148604100
ENSG00000207205
hyper


chr1.149155501.149155800
ENSG00000202167
hyper


chr1.149223301.149223600
ENSG00000206737
hyper


chr1.149223601.149223900
ENSG00000206737
hyper


chr1.17216101.17216400
ENSG00000058453
hyper


chr1.91182301.91182600
ENSG00000143032
hyper


chr1.98511601.98511900
ENSG00000225206
hyper


chr1.99470101.99470400
ENSG00000117598
hyper


chr1.145944901.145945200
ENSG00000201105
hyper


chr1.147486601.147486900
ENSG00000206791
hyper


chr1.148598101.148598400
ENSG00000237253
hyper


chr1.148760401.148760700
ENSG00000237343
hyper


chr1.149223901.149224200
ENSG00000206737
hyper


chr1.149224201.149224500
ENSG00000206737
hyper


chr1.17215801.17216100
ENSG00000058453
hyper


chr1.20810101.20810400
ENSG00000162545
hyper


chr1.26551801.26552100
ENSG00000236155
hyper


chr1.50893501.50893800
ENSG00000142700
hyper


chr1.57888301.57888600
ENSG00000173406
hyper


chr1.63785401.63785700
ENSG00000230798
hyper


chr1.63786001.63786300
ENSG00000230798
hyper


chr1.66258301.66258600
ENSG00000184588
hyper


chr1.75595801.75596100
ENSG00000224127
hyper


chr1.77334601.77334900
ENSG00000117069
hyper


chr1.91182601.91182900
ENSG00000143032
hyper


chr1.91183801.91184100
ENSG00000143032
hyper


chr1.92948101.92948400
ENSG00000162676
hyper


chr1.98511301.98511600
ENSG00000225206
hyper


chr1.99469801.99470100
ENSG00000117598
hyper


chr1.110612401.110612700
ENSG00000143093
hyper


chr1.111216901.111217200
ENSG00000177272
hyper


chr1.111506101.111506400
ENSG00000121931
hyper


chr1.119526601.119526900
ENSG00000092607
hyper


chr1.119526901.119527200
ENSG00000092607
hyper


chr1.119527201.119527500
ENSG00000092607
hyper


chr1.119532601.119532900
ENSG00000092607
hyper


chr1.119536201.119536500
ENSG00000092607
hyper


chr1.119543101.119543400
ENSG00000226172
hyper


chr1.119550901.119551200
ENSG00000239216
hyper


chr1.119551201.119551500
ENSG00000239216
hyper


chr1.145944601.145944900
ENSG00000201105
hyper


chr1.145963501.145963800
ENSG00000207418
hyper


chr1.145979401.145979700
ENSG00000207418
hyper


chr1.145990801.145991100
ENSG00000229828
hyper


chr1.147486301.147486600
ENSG00000206791
hyper


chr1.147505201.147505500
ENSG00000206585
hyper


chr1.147521101.147521400
ENSG00000206585
hyper


chr1.147752701.147753000
ENSG00000234283
hyper


chr1.147753001.147753300
ENSG00000234283
hyper


chr1.147775201.147775500
ENSG00000238107
hyper


chr1.147790501.147790800
ENSG00000235988
hyper


chr1.149156101.149156400
ENSG00000202167
hyper


chr1.149156401.149156700
ENSG00000202167
hyper


chr1.149224501.149224800
ENSG00000206737
hyper


chr1.149400001.149400300
ENSG00000273213
hyper


chr1.149719501.149719800
ENSG00000234232
hyper


chr1.242687101.242687400
ENSG00000180287
hyper


chr1.165323701.165324000
ENSG00000162761
hyper


chr1.177140401.177140700
ENSG00000198797
hyper


chr1.207999301.207999600
ENSG00000203709
hyper


chr1.217311301.217311600
ENSG00000196482
hyper


chr1.234041101.234041400
ENSG00000183780
hyper


chr1.237204901.237205200
ENSG00000198626
hyper


chr1.240255001.240255300
ENSG00000155816
hyper


chr2.19558501.19558800
ENSG00000143867
hyper


chr1.161039401.161039700
ENSG00000186517
hyper


chr1.165321601.165321900
ENSG00000162761
hyper


chr1.165323401.165323700
ENSG00000162761
hyper


chr1.165324301.165324600
ENSG00000162761
hyper


chr1.165324601.165324900
ENSG00000162761
hyper


chr1.167090701.167091000
ENSG00000198842
hyper


chr1.167682601.167682900
ENSG00000198771
hyper


chr1.169396501.169396800
ENSG00000117477
hyper


chr1.169396801.169397100
ENSG00000117477
hyper


chr1.170630101.170630400
ENSG00000116132
hyper


chr1.173638801.173639100
ENSG00000183831
hyper


chr1.180203701.180204000
ENSG00000121454
hyper


chr1.180204001.180204300
ENSG00000121454
hyper


chr1.180204301.180204600
ENSG00000121454
hyper


chr1.200010001.200010300
ENSG00000116833
hyper


chr1.200011201.200011500
ENSG00000116833
hyper


chr1.214159501.214159800
ENSG00000230461
hyper


chr1.217307401.217307700
ENSG00000196482
hyper


chr1.217307701.217308000
ENSG00000196482
hyper


chr1.217308001.217308300
ENSG00000196482
hyper


chr1.217309501.217309800
ENSG00000196482
hyper


chr1.217309801.217310100
ENSG00000196482
hyper


chr1.217310101.217310400
ENSG00000196482
hyper


chr1.217311601.217311900
ENSG00000196482
hyper


chr1.217313101.217313400
ENSG00000196482
hyper


chr1.217313401.217313700
ENSG00000196482
hyper


chr1.220959601.220959900
ENSG00000186205
hyper


chr1.224804401.224804700
ENSG00000143786
hyper


chr1.224804701.224805000
ENSG00000143786
hyper


chr1.228652201.228652500
ENSG00000181201
hyper


chr1.235814101.235814400
ENSG00000168243
hyper


chr1.239550601.239550900
ENSG00000133019
hyper


chr1.239550901.239551200
ENSG00000133019
hyper


chr1.239551201.239551500
ENSG00000133019
hyper


chr1.242686801.242687100
ENSG00000180287
hyper


chr2.1746901.1747200
ENSG00000130508
hyper


chr2.5830801.5831100
ENSG00000224128
hyper


chr2.5831101.5831400
ENSG00000224128
hyper


chr2.19555801.19556100
ENSG00000143867
hyper


chr2.45155101.45155400
ENSG00000259439
hyper


chr2.45159301.45159600
ENSG00000259439
hyper


chr2.45160201.45160500
ENSG00000259439
hyper


chr2.45170101.45170400
ENSG00000138083
hyper


chr2.45171301.45171600
ENSG00000138083
hyper


chr2.45228301.45228600
ENSG00000170577
hyper


chr2.45228601.45228900
ENSG00000170577
hyper


chr2.45231301.45231600
ENSG00000170577
hyper


chr2.45231901.45232200
ENSG00000170577
hyper


chr2.45233401.45233700
ENSG00000170577
hyper


chr2.50574301.50574600
ENSG00000179915
hyper


chr2.85107301.85107600
ENSG00000186854
hyper


chr2.119600401.119600700
ENSG00000163064
hyper


chr2.119607601.119607900
ENSG00000163064
hyper


chr2.176933401.176933700
ENSG00000174279
hyper


chr2.63280801.63281100
ENSG00000115507
hyper


chr2.80531701.80532000
ENSG00000066032
hyper


chr2.115920601.115920900
ENSG00000175497
hyper


chr2.131721901.131722200
ENSG00000136002
hyper


chr2.177030001.177030300
ENSG00000128652
hyper


chr2.63279001.63279300
ENSG00000115507
hyper


chr2.63279901.63280200
ENSG00000115507
hyper


chr2.63280201.63280500
ENSG00000115507
hyper


chr2.63280501.63280800
ENSG00000115507
hyper


chr2.63281101.63281400
ENSG00000115507
hyper


chr2.63281401.63281700
ENSG00000115507
hyper


chr2.63285301.63285600
ENSG00000115507
hyper


chr2.63285601.63285900
ENSG00000115507
hyper


chr2.71017201.71017500
ENSG00000183733
hyper


chr2.73147201.73147500
ENSG00000135638
hyper


chr2.73519501.73519800
ENSG00000135625
hyper


chr2.80529901.80530200
ENSG00000066032
hyper


chr2.80530201.80530500
ENSG00000066032
hyper


chr2.84743401.84743700
ENSG00000115423
hyper


chr2.85107001.85107300
ENSG00000186854
hyper


chr2.111876901.111877200
ENSG00000153094
hyper


chr2.119600701.119601000
ENSG00000163064
hyper


chr2.119614501.119614800
ENSG00000163064
hyper


chr2.119614801.119615100
ENSG00000163064
hyper


chr2.119616301.119616600
ENSG00000163064
hyper


chr2.119616601.119616900
ENSG00000163064
hyper


chr2.124782301.124782600
ENSG00000228400
hyper


chr2.139537201.139537500
ENSG00000144227
hyper


chr2.149645701.149646000
ENSG00000231079
hyper


chr2.157176601.157176900
ENSG00000153234
hyper


chr2.168150001.168150300
ENSG00000228222
hyper


chr2.172946101.172946400
ENSG00000172878
hyper


chr2.172952401.172952700
ENSG00000144355
hyper


chr2.173099701.173100000
ENSG00000232555
hyper


chr2.173100001.173100300
ENSG00000232555
hyper


chr2.175191901.175192200
ENSG00000231453
hyper


chr2.175193401.175193700
ENSG00000231453
hyper


chr2.175193701.175194000
ENSG00000231453
hyper


chr2.175205701.175206000
ENSG00000217236
hyper


chr2.176931601.176931900
ENSG00000174279
hyper


chr2.176931901.176932200
ENSG00000174279
hyper


chr2.176933101.176933400
ENSG00000174279
hyper


chr2.176936401.176936700
ENSG00000174279
hyper


chr2.176943301.176943600
ENSG00000174279
hyper


chr2.176946601.176946900
ENSG00000174279
hyper


chr2.176947201.176947500
ENSG00000174279
hyper


chr2.176948101.176948400
ENSG00000174279
hyper


chr2.176964901.176965200
ENSG00000170178
hyper


chr2.176965201.176965500
ENSG00000170178
hyper


chr2.176976901.176977200
ENSG00000128710
hyper


chr2.176981101.176981400
ENSG00000128710
hyper


chr2.177054301.177054600
ENSG00000128645
hyper


chr2.177054601.177054900
ENSG00000128645
hyper


chr2.182322601.182322900
ENSG00000115232
hyper


chr2.200333701.200334000
ENSG00000119042
hyper


chr2.200334001.200334300
ENSG00000119042
hyper


chr3.27770401.27770700
ENSG00000163508
hyper


chr3.62353801.62354100
ENSG00000241472
hyper


chr2.223161901.223162200
ENSG00000135903
hyper


chr2.223162801.223163100
ENSG00000135903
hyper


chr2.223166401.223166700
ENSG00000163081
hyper


chr2.223176301.223176600
ENSG00000267034
hyper


chr2.229046101.229046400
ENSG00000153820
hyper


chr2.237072601.237072900
ENSG00000168505
hyper


chr2.237082201.237082500
ENSG00000233611
hyper


chr3.27765001.27765300
ENSG00000163508
hyper


chr3.27765301.27765600
ENSG00000163508
hyper


chr3.62353501.62353800
ENSG00000241472
hyper


chr3.192126001.192126300
ENSG00000114279
hyper


chr4.4868101.4868400
ENSG00000163132
hyper


chr4.13532401.13532700
ENSG00000109705
hyper


chr4.13532701.13533000
ENSG00000109705
hyper


chr3.169377901.169378200
ENSG00000085276
hyper


chr3.170137201.170137500
ENSG00000013297
hyper


chr3.194409001.194409300
ENSG00000185112
hyper


chr3.128210701.128211000
ENSG00000179348
hyper


chr3.129693901.129694200
ENSG00000170893
hyper


chr3.129694201.129694500
ENSG00000170893
hyper


chr3.137480101.137480400
ENSG00000168875
hyper


chr3.138657301.138657600
ENSG00000244578
hyper


chr3.138657901.138658200
ENSG00000244578
hyper


chr3.147077401.147077700
ENSG00000243620
hyper


chr3.147105901.147106200
ENSG00000174963
hyper


chr3.147109501.147109800
ENSG00000174963
hyper


chr3.147109801.147110100
ENSG00000174963
hyper


chr3.147110101.147110400
ENSG00000174963
hyper


chr3.147114301.147114600
ENSG00000174963
hyper


chr3.147124201.147124500
ENSG00000174963
hyper


chr3.157812601.157812900
ENSG00000168779
hyper


chr3.157821301.157821600
ENSG00000168779
hyper


chr3.159944401.159944700
ENSG00000180044
hyper


chr3.170136901.170137200
ENSG00000013297
hyper


chr3.173302801.173303100
ENSG00000169760
hyper


chr3.181422001.181422300
ENSG00000242808
hyper


chr3.181441501.181441800
ENSG00000242808
hyper


chr3.192126301.192126600
ENSG00000114279
hyper


chr3.192231901.192232200
ENSG00000114279
hyper


chr4.4856401.4856700
ENSG00000273396
hyper


chr4.9178201.9178500
ENSG00000229924
hyper


chr4.13533001.13533300
ENSG00000109705
hyper


chr4.20255701.20256000
ENSG00000145147
hyper


chr4.20256001.20256300
ENSG00000145147
hyper


chr4.37245601.37245900
ENSG00000174145
hyper


chr4.37245901.37246200
ENSG00000174145
hyper


chr4.41749501.41749800
ENSG00000109132
hyper


chr4.41875501.41875800
ENSG00000245870
hyper


chr4.42398701.42399000
ENSG00000178343
hyper


chr4.44449501.44449800
ENSG00000183783
hyper


chr4.54969901.54970200
ENSG00000145216
hyper


chr4.85402801.85403100
ENSG00000163623
hyper


chr5.2743201.2743500
ENSG00000170561
hyper


chr4.134071801.134072100
ENSG00000138650
hyper


chr4.174450001.174450300
ENSG00000164107
hyper


chr4.190938901.190939200
ENSG00000201145
hyper


chr4.81187501.81187800
ENSG00000138675
hyper


chr4.85414501.85414800
ENSG00000163623
hyper


chr4.85414801.85415100
ENSG00000163623
hyper


chr4.85417801.85418100
ENSG00000163623
hyper


chr4.85418101.85418400
ENSG00000163623
hyper


chr4.85418401.85418700
ENSG00000163623
hyper


chr4.104640901.104641200
ENSG00000169836
hyper


chr4.107956501.107956800
ENSG00000155011
hyper


chr4.110223601.110223900
ENSG00000188517
hyper


chr4.111533101.111533400
ENSG00000250103
hyper


chr4.111555301.111555600
ENSG00000164093
hyper


chr4.111562501.111562800
ENSG00000164093
hyper


chr4.121992301.121992600
ENSG00000173376
hyper


chr4.122686201.122686500
ENSG00000164112
hyper


chr4.134069401.134069700
ENSG00000250241
hyper


chr4.134071501.134071800
ENSG00000138650
hyper


chr4.134072101.134072400
ENSG00000138650
hyper


chr4.134072401.134072700
ENSG00000138650
hyper


chr4.134072701.134073000
ENSG00000138650
hyper


chr4.134073901.134074200
ENSG00000138650
hyper


chr4.144621301.144621600
ENSG00000183090
hyper


chr4.147561601.147561900
ENSG00000151615
hyper


chr4.158143201.158143500
ENSG00000120251
hyper


chr4.158143501.158143800
ENSG00000120251
hyper


chr4.172733701.172734000
ENSG00000174473
hyper


chr4.172734601.172734900
ENSG00000174473
hyper


chr4.174422101.174422400
ENSG00000164107
hyper


chr4.174427801.174428100
ENSG00000164107
hyper


chr4.174429601.174429900
ENSG00000164107
hyper


chr4.174430201.174430500
ENSG00000164107
hyper


chr4.174448501.174448800
ENSG00000164107
hyper


chr5.2754901.2755200
ENSG00000186493
hyper


chr5.3104701.3105000
ENSG00000249808
hyper


chr5.3116701.3117000
ENSG00000249808
hyper


chr5.3590701.3591000
ENSG00000170549
hyper


chr5.3599401.3599700
ENSG00000170549
hyper


chr5.3600601.3600900
ENSG00000170549
hyper


chr5.3602101.3602400
ENSG00000170549
hyper


chr5.54518701.54519000
ENSG00000234602
hyper


chr5.54519001.54519300
ENSG00000234602
hyper


chr5.122422501.122422800
ENSG00000223652
hyper


chr5.32712601.32712900
ENSG00000113389
hyper


chr5.40680901.40681200
ENSG00000171522
hyper


chr5.42994801.42995100
ENSG00000271788
hyper


chr5.42995101.42995400
ENSG00000271788
hyper


chr5.54519301.54519600
ENSG00000234602
hyper


chr5.57878101.57878400
ENSG00000152932
hyper


chr5.63257401.63257700
ENSG00000248285
hyper


chr5.72528901.72529200
ENSG00000249743
hyper


chr5.72529201.72529500
ENSG00000249743
hyper


chr5.72596701.72597000
ENSG00000249743
hyper


chr5.72740101.72740400
ENSG00000251493
hyper


chr5.72740401.72740700
ENSG00000251493
hyper


chr5.80256601.80256900
ENSG00000251450
hyper


chr5.94955701.94956000
ENSG00000178015
hyper


chr5.95768101.95768400
ENSG00000251314
hyper


chr5.95768701.95769000
ENSG00000251314
hyper


chr5.115152001.115152300
ENSG00000129596
hyper


chr5.115152301.115152600
ENSG00000129596
hyper


chr5.122423401.122423700
ENSG00000223652
hyper


chr5.134376301.134376600
ENSG00000224186
hyper


chr5.134825101.134825400
ENSG00000249639
hyper


chr5.134825401.134825700
ENSG00000249639
hyper


chr5.134826001.134826300
ENSG00000249639
hyper


chr5.140012101.140012400
ENSG00000170458
hyper


chr5.140012401.140012700
ENSG00000170458
hyper


chr5.140346601.140346900
ENSG00000204970
hyper


chr5.154026901.154027200
ENSG00000221552
hyper


chr5.172672201.172672500
ENSG00000183072
hyper


chr6.1378501.1378800
ENSG00000261730
hyper


chr6.10421701.10422000
ENSG00000228478
hyper


chr6.26721001.26721300
ENSG00000261584
hyper


chr6.26722501.26722800
ENSG00000261584
hyper


chr6.26722801.26723100
ENSG00000261584
hyper


chr6.26745301.26745600
ENSG00000261584
hyper


chr6.26778601.26778900
ENSG00000241549
hyper


chr6.26778901.26779200
ENSG00000241549
hyper


chr6.26779201.26779500
ENSG00000241549
hyper


chr6.27258301.27258600
ENSG00000158553
hyper


chr6.27462901.27463200
ENSG00000270666
hyper


chr6.27533701.27534000
ENSG00000219738
hyper


chr6.27534001.27534300
ENSG00000219738
hyper


chr6.27648601.27648900
ENSG00000216676
hyper


chr6.27648901.27649200
ENSG00000216676
hyper


chr6.28740901.28741200
ENSG00000221191
hyper


chr6.39281101.39281400
ENSG00000124780
hyper


chr6.58147501.58147800
ENSG00000272541
hyper


chr5.178978501.178978800
ENSG00000176783
hyper


chr6.28411201.28411500
ENSG00000187987
hyper


chr5.170742001.170742300
ENSG00000164438
hyper


chr5.172665301.172665600
ENSG00000183072
hyper


chr5.174158701.174159000
ENSG00000120149
hyper


chr5.174159001.174159300
ENSG00000120149
hyper


chr5.174159301.174159600
ENSG00000120149
hyper


chr5.174486901.174487200
ENSG00000204754
hyper


chr5.177666601.177666900
ENSG00000050767
hyper


chr5.178368001.178368300
ENSG00000178187
hyper


chr6.5026501.5026800
ENSG00000272142
hyper


chr6.6004201.6004500
ENSG00000124785
hyper


chr6.10382101.10382400
ENSG00000137203
hyper


chr6.26614201.26614500
ENSG00000271071
hyper


chr6.26614501.26614800
ENSG00000271071
hyper


chr6.26614801.26615100
ENSG00000271071
hyper


chr6.26721301.26721600
ENSG00000261584
hyper


chr6.26723101.26723400
ENSG00000261584
hyper


chr6.27279901.27280200
ENSG00000158553
hyper


chr6.27280201.27280500
ENSG00000158553
hyper


chr6.27463201.27463500
ENSG00000270666
hyper


chr6.28303801.28304100
ENSG00000235109
hyper


chr6.28367101.28367400
ENSG00000158691
hyper


chr6.28367401.28367700
ENSG00000158691
hyper


chr6.28414801.28415100
ENSG00000231162
hyper


chr6.28554901.28555200
ENSG00000232040
hyper


chr6.28602601.28602900
ENSG00000271440
hyper


chr6.28753801.28754100
ENSG00000265764
hyper


chr6.28778101.28778400
ENSG00000265764
hyper


chr6.32977201.32977500
ENSG00000263756
hyper


chr6.41341501.41341800
ENSG00000238867
hyper


chr6.50818801.50819100
ENSG00000008196
hyper


chr6.56716201.56716500
ENSG00000151914
hyper


chr6.58147201.58147500
ENSG00000272541
hyper


chr6.58147801.58148100
ENSG00000272541
hyper


chr6.58148401.58148700
ENSG00000272541
hyper


chr6.58148701.58149000
ENSG00000272541
hyper


chr6.62995501.62995800
ENSG00000112232
hyper


chr6.74024401.74024700
ENSG00000135314
hyper


chr6.75794701.75795000
ENSG00000111799
hyper


chr6.78172201.78172500
ENSG00000135312
hyper


chr6.78172501.78172800
ENSG00000135312
hyper


chr6.78173101.78173400
ENSG00000135312
hyper


chr6.85473001.85473300
ENSG00000112837
hyper


chr6.99291301.99291600
ENSG00000184486
hyper


chr6.100056001.100056300
ENSG00000112238
hyper


chr6.100441801.100442100
ENSG00000152034
hyper


chr6.100912501.100912800
ENSG00000112246
hyper


chr6.101847001.101847300
ENSG00000164418
hyper


chr6.106433701.106434000
ENSG00000200198
hyper


chr6.108440101.108440400
ENSG00000081087
hyper


chr6.108488401.108488700
ENSG00000112333
hyper


chr6.108488701.108489000
ENSG00000112333
hyper


chr6.108489301.108489600
ENSG00000112333
hyper


chr6.117086401.117086700
ENSG00000183807
hyper


chr6.117591301.117591600
ENSG00000170162
hyper


chr6.133562401.133562700
ENSG00000112319
hyper


chr6.133562701.133563000
ENSG00000112319
hyper


chr6.134214001.134214300
ENSG00000118526
hyper


chr6.137810401.137810700
ENSG00000177468
hyper


chr7.27260101.27260400
ENSG00000243766
hyper


chr7.35301001.35301300
ENSG00000226063
hyper


chr7.1959601.1959900
ENSG00000002822
hyper


chr7.8474701.8475000
ENSG00000122584
hyper


chr7.19184701.19185000
ENSG00000229533
hyper


chr6.137808901.137809200
ENSG00000177468
hyper


chr6.137816701.137817000
ENSG00000177468
hyper


chr6.151562401.151562700
ENSG00000131016
hyper


chr6.159654901.159655200
ENSG00000164694
hyper


chr6.166074601.166074900
ENSG00000112541
hyper


chr6.166580401.166580700
ENSG00000164458
hyper


chr6.166582801.166583100
ENSG00000164458
hyper


chr6.166583101.166583400
ENSG00000164458
hyper


chr7.1270801.1271100
ENSG00000164853
hyper


chr7.8475001.8475300
ENSG00000122584
hyper


chr7.8475301.8475600
ENSG00000122584
hyper


chr7.8481301.8481600
ENSG00000122584
hyper


chr7.8482501.8482800
ENSG00000122584
hyper


chr7.8482801.8483100
ENSG00000122584
hyper


chr7.15726601.15726900
ENSG00000106511
hyper


chr7.19146001.19146300
ENSG00000122691
hyper


chr7.19146301.19146600
ENSG00000122691
hyper


chr7.19146901.19147200
ENSG00000122691
hyper


chr7.19147201.19147500
ENSG00000122691
hyper


chr7.19152001.19152300
ENSG00000122691
hyper


chr7.19158001.19158300
ENSG00000122691
hyper


chr7.19158601.19158900
ENSG00000236536
hyper


chr7.19184401.19184700
ENSG00000229533
hyper


chr7.19185001.19185300
ENSG00000229533
hyper


chr7.22589401.22589700
ENSG00000105889
hyper


chr7.23507401.23507700
ENSG00000136231
hyper


chr7.24324301.24324600
ENSG00000122585
hyper


chr7.24324601.24324900
ENSG00000122585
hyper


chr7.27192301.27192600
ENSG00000254369
hyper


chr7.27196501.27196800
ENSG00000122592
hyper


chr7.27204301.27204600
ENSG00000078399
hyper


chr7.27204601.27204900
ENSG00000078399
hyper


chr7.27205201.27205500
ENSG00000078399
hyper


chr7.27205501.27205800
ENSG00000078399
hyper


chr7.27205801.27206100
ENSG00000078399
hyper


chr7.27206101.27206400
ENSG00000078399
hyper


chr7.27225001.27225300
ENSG00000240990
hyper


chr7.27244501.27244800
ENSG00000243766
hyper


chr7.27244801.27245100
ENSG00000243766
hyper


chr7.27252601.27252900
ENSG00000243766
hyper


chr7.27284701.27285000
ENSG00000253405
hyper


chr7.27291301.27291600
ENSG00000106038
hyper


chr7.27291601.27291900
ENSG00000106038
hyper


chr7.27291901.27292200
ENSG00000106038
hyper


chr7.30721201.30721500
ENSG00000106113
hyper


chr7.31092601.31092900
ENSG00000078549
hyper


chr7.35293201.35293500
ENSG00000164532
hyper


chr7.35297401.35297700
ENSG00000226063
hyper


chr7.35301301.35301600
ENSG00000226063
hyper


chr7.37955701.37956000
ENSG00000086289
hyper


chr7.52156201.52156500
ENSG00000233960
hyper


chr7.54609601.54609900
ENSG00000170419
hyper


chr7.64349101.64349400
ENSG00000198039
hyper


chr7.64349401.64349700
ENSG00000198039
hyper


chr7.71800801.71801100
ENSG00000183166
hyper


chr7.79083601.79083900
ENSG00000234456
hyper


chr7.88388101.88388400
ENSG00000182348
hyper


chr7.93203701.93204000
ENSG00000004948
hyper


chr7.93519301.93519600
ENSG00000127928
hyper


chr7.93519601.93519900
ENSG00000127928
hyper


chr7.94284901.94285200
ENSG00000127990
hyper


chr7.96647401.96647700
ENSG00000105880
hyper


chr7.96650701.96651000
ENSG00000105880
hyper


chr7.96651001.96651300
ENSG00000105880
hyper


chr7.97362301.97362600
ENSG00000006128
hyper


chr7.97362601.97362900
ENSG00000006128
hyper


chr7.97362901.97363200
ENSG00000006128
hyper


chr7.97363201.97363500
ENSG00000006128
hyper


chr7.107641801.107642100
ENSG00000091136
hyper


chr7.107642101.107642400
ENSG00000091136
hyper


chr7.113722801.113723100
ENSG00000128573
hyper


chr7.113723101.113723400
ENSG00000128573
hyper


chr8.99951301.99951600
ENSG00000104375
hyper


chr8.99951601.99951900
ENSG00000104375
hyper


chr8.99951901.99952200
ENSG00000104375
hyper


chr7.123173101.123173400
ENSG00000164675
hyper


chr8.38008201.38008500
ENSG00000147465
hyper


chr8.55372201.55372500
ENSG00000164736
hyper


chr8.60032401.60032700
ENSG00000167912
hyper


chr8.99960601.99960900
ENSG00000164920
hyper


chr7.117119401.117119700
ENSG00000001626
hyper


chr7.121956901.121957200
ENSG00000081803
hyper


chr7.123172801.123173100
ENSG00000164675
hyper


chr7.136554001.136554300
ENSG00000234352
hyper


chr7.136554301.136554600
ENSG00000234352
hyper


chr7.136554601.136554900
ENSG00000234352
hyper


chr7.136554901.136555200
ENSG00000234352
hyper


chr7.137532001.137532300
ENSG00000157680
hyper


chr7.137532301.137532600
ENSG00000157680
hyper


chr7.155241901.155242200
ENSG00000236544
hyper


chr7.155242801.155243100
ENSG00000236544
hyper


chr7.155243701.155244000
ENSG00000236544
hyper


chr7.155259301.155259600
ENSG00000164778
hyper


chr7.155259601.155259900
ENSG00000164778
hyper


chr7.155301601.155301900
ENSG00000146910
hyper


chr7.156795601.156795900
ENSG00000130675
hyper


chr7.156797101.156797400
ENSG00000130675
hyper


chr7.156797401.156797700
ENSG00000130675
hyper


chr7.156810901.156811200
ENSG00000243479
hyper


chr7.156811201.156811500
ENSG00000243479
hyper


chr7.157482001.157482300
ENSG00000155093
hyper


chr7.157482301.157482600
ENSG00000155093
hyper


chr8.4849501.4849800
ENSG00000183117
hyper


chr8.4849801.4850100
ENSG00000183117
hyper


chr8.21996601.21996900
ENSG00000168476
hyper


chr8.23563801.23564100
ENSG00000180053
hyper


chr8.23564101.23564400
ENSG00000180053
hyper


chr8.23564401.23564700
ENSG00000253471
hyper


chr8.24858901.24859200
ENSG00000253832
hyper


chr8.25905001.25905300
ENSG00000221818
hyper


chr8.33372001.33372300
ENSG00000129696
hyper


chr8.33372301.33372600
ENSG00000129696
hyper


chr8.37655701.37656000
ENSG00000020181
hyper


chr8.55366201.55366500
ENSG00000164736
hyper


chr8.55367101.55367400
ENSG00000164736
hyper


chr8.55367401.55367700
ENSG00000164736
hyper


chr8.57026101.57026400
ENSG00000172680
hyper


chr8.65283301.65283600
ENSG00000253554
hyper


chr8.65290801.65291100
ENSG00000254377
hyper


chr8.65499601.65499900
ENSG00000172817
hyper


chr8.67873501.67873800
ENSG00000261787
hyper


chr8.70981801.70982100
ENSG00000147596
hyper


chr8.70983901.70984200
ENSG00000147596
hyper


chr8.70984201.70984500
ENSG00000147596
hyper


chr8.72470401.72470700
ENSG00000253379
hyper


chr8.72471001.72471300
ENSG00000253379
hyper


chr8.72754501.72754800
ENSG00000235531
hyper


chr8.72754801.72755100
ENSG00000235531
hyper


chr8.72917101.72917400
ENSG00000235531
hyper


chr8.72917401.72917700
ENSG00000235531
hyper


chr8.76316701.76317000
ENSG00000164749
hyper


chr8.76317001.76317300
ENSG00000164749
hyper


chr8.85094401.85094700
ENSG00000184672
hyper


chr8.85094701.85095000
ENSG00000184672
hyper


chr8.93114001.93114300
ENSG00000079102
hyper


chr8.97167001.97167300
ENSG00000156466
hyper


chr8.97170001.97170300
ENSG00000156466
hyper


chr8.97170301.97170600
ENSG00000156466
hyper


chr8.97170601.97170900
ENSG00000156466
hyper


chr8.99952201.99952500
ENSG00000104375
hyper


chr8.99960301.99960600
ENSG00000164920
hyper


chr8.99960901.99961200
ENSG00000164920
hyper


chr8.99961201.99961500
ENSG00000164920
hyper


chr8.99986101.99986400
ENSG00000229625
hyper


chr9.970801.971100
ENSG00000137090
hyper


chr9.1045201.1045500
ENSG00000173253
hyper


chr9.1045801.1046100
ENSG00000173253
hyper


chr9.41454901.41455200
ENSG00000237625
hyper


chr9.79629301.79629600
ENSG00000204612
hyper


chr8.132053701.132054000
ENSG00000155897
hyper


chr8.109094701.109095000
ENSG00000147655
hyper


chr8.114444601.114444900
ENSG00000164796
hyper


chr8.114444901.114445200
ENSG00000164796
hyper


chr8.114447001.114447300
ENSG00000164796
hyper


chr8.132053401.132053700
ENSG00000155897
hyper


chr8.132054001.132054300
ENSG00000155897
hyper


chr9.117001.117300
ENSG00000170122
hyper


chr9.117301.117600
ENSG00000170122
hyper


chr9.117601.117900
ENSG00000170122
hyper


chr9.117901.118200
ENSG00000170122
hyper


chr9.843001.843300
ENSG00000137090
hyper


chr9.843301.843600
ENSG00000137090
hyper


chr9.973501.973800
ENSG00000064218
hyper


chr9.1042501.1042800
ENSG00000173253
hyper


chr9.1045501.1045800
ENSG00000173253
hyper


chr9.17907001.17907300
ENSG00000107295
hyper


chr9.19788301.19788600
ENSG00000155886
hyper


chr9.19788601.19788900
ENSG00000155886
hyper


chr9.34809601.34809900
ENSG00000257198
hyper


chr9.36739501.36739800
ENSG00000165304
hyper


chr9.36739801.36740100
ENSG00000165304
hyper


chr9.69201001.69201300
ENSG00000204793
hyper


chr9.79628701.79629000
ENSG00000204612
hyper


chr9.79629001.79629300
ENSG00000204612
hyper


chr9.79630501.79630800
ENSG00000204612
hyper


chr9.79631401.79631700
ENSG00000204612
hyper


chr9.79636801.79637100
ENSG00000204612
hyper


chr9.90114001.90114300
ENSG00000196730
hyper


chr9.96713401.96713700
ENSG00000131668
hyper


chr9.96715201.96715500
ENSG00000131668
hyper


chr9.100610701.100611000
ENSG00000178919
hyper


chr9.100611301.100611600
ENSG00000178919
hyper


chr9.133537201.133537500
ENSG00000130711
hyper


chr10.102996001.102996300
ENSG00000227128
hyper


chr10.102997201.102997500
ENSG00000227128
hyper


chr9.124414501.124414800
ENSG00000136848
hyper


chr9.126775201.126775500
ENSG00000106689
hyper


chr9.126777301.126777600
ENSG00000106689
hyper


chr9.127212901.127213200
ENSG00000180264
hyper


chr9.129380401.129380700
ENSG00000136944
hyper


chr9.129386101.129386400
ENSG00000136944
hyper


chr10.8076901.8077200
ENSG00000197308
hyper


chr10.8077201.8077500
ENSG00000197308
hyper


chr10.21783301.21783600
ENSG00000204682
hyper


chr10.22765501.22765800
ENSG00000077327
hyper


chr10.23462101.23462400
ENSG00000168267
hyper


chr10.23480401.23480700
ENSG00000168267
hyper


chr10.28035001.28035300
ENSG00000230500
hyper


chr10.44879101.44879400
ENSG00000107562
hyper


chr10.50605501.50605800
ENSG00000165606
hyper


chr10.63212401.63212700
ENSG00000196932
hyper


chr10.71337601.71337900
ENSG00000236154
hyper


chr10.94828201.94828500
ENSG00000187553
hyper


chr10.94833901.94834200
ENSG00000095596
hyper


chr10.102894901.102895200
ENSG00000107807
hyper


chr10.102996301.102996600
ENSG00000227128
hyper


chr10.106400401.106400700
ENSG00000156395
hyper


chr10.110671801.110672100
ENSG00000222436
hyper


chr10.118031101.118031400
ENSG00000151892
hyper


chr10.118031401.118031700
ENSG00000151892
hyper


chr10.118033801.118034100
ENSG00000151892
hyper


chr10.118891501.118891800
ENSG00000148704
hyper


chr10.118892401.118892700
ENSG00000148704
hyper


chr10.119301301.119301600
ENSG00000229847
hyper


chr10.119304901.119305200
ENSG00000170370
hyper


chr10.119305201.119305500
ENSG00000170370
hyper


chr10.119494201.119494500
ENSG00000234952
hyper


chr10.119494501.119494800
ENSG00000234952
hyper


chr11.18813901.18814200
ENSG00000110786
hyper


chr11.31826401.31826700
ENSG00000007372
hyper


chr11.69832501.69832800
ENSG00000202070
hyper


chr10.122709601.122709900
ENSG00000227307
hyper


chr10.124896601.124896900
ENSG00000188620
hyper


chr10.124905601.124905900
ENSG00000188816
hyper


chr10.124908901.124909200
ENSG00000188816
hyper


chr10.131761201.131761500
ENSG00000108001
hyper


chr10.131767801.131768100
ENSG00000108001
hyper


chr11.7041601.7041900
ENSG00000158077
hyper


chr11.14995501.14995800
ENSG00000175868
hyper


chr11.22363201.22363500
ENSG00000091664
hyper


chr11.31825801.31826100
ENSG00000007372
hyper


chr11.31826101.31826400
ENSG00000007372
hyper


chr11.31827001.31827300
ENSG00000007372
hyper


chr11.32454601.32454900
ENSG00000184937
hyper


chr11.32455801.32456100
ENSG00000184937
hyper


chr11.32459401.32459700
ENSG00000183242
hyper


chr11.32459701.32460000
ENSG00000183242
hyper


chr11.35641801.35642100
ENSG00000179431
hyper


chr11.43602901.43603200
ENSG00000149084
hyper


chr11.62693701.62694000
ENSG00000168539
hyper


chr11.66188701.66189000
ENSG00000174576
hyper


chr11.69451501.69451800
ENSG00000110092
hyper


chr11.69451801.69452100
ENSG00000110092
hyper


chr11.69452101.69452400
ENSG00000110092
hyper


chr11.69452401.69452700
ENSG00000110092
hyper


chr11.69517501.69517800
ENSG00000162344
hyper


chr11.69517801.69518100
ENSG00000162344
hyper


chr11.69831901.69832200
ENSG00000202070
hyper


chr11.69832201.69832500
ENSG00000202070
hyper


chr11.70211401.70211700
ENSG00000131626
hyper


chr11.91958401.91958700
ENSG00000242248
hyper


chr11.100999201.100999500
ENSG00000082175
hyper


chr11.100999501.100999800
ENSG00000082175
hyper


chr11.101453101.101453400
ENSG00000137672
hyper


chr11.101453401.101453700
ENSG00000137672
hyper


chr11.122848201.122848500
ENSG00000188909
hyper


chr11.123066601.123066900
ENSG00000254710
hyper


chr12.54424501.54424800
ENSG00000273049
hyper


chr12.106974901.106975200
ENSG00000257545
hyper


chr12.115173301.115173600
ENSG00000257817
hyper


chr12.128752501.128752800
ENSG00000181234
hyper


chr12.6184501.6184800
ENSG00000110799
hyper


chr12.14134201.14134500
ENSG00000273079
hyper


chr12.16500601.16500900
ENSG00000008394
hyper


chr12.22093801.22094100
ENSG00000069431
hyper


chr12.22094701.22095000
ENSG00000069431
hyper


chr12.25056301.25056600
ENSG00000060982
hyper


chr12.30323101.30323400
ENSG00000257262
hyper


chr12.43944901.43945200
ENSG00000173157
hyper


chr12.48397201.48397500
ENSG00000139219
hyper


chr12.54321301.54321600
ENSG00000249641
hyper


chr12.54329701.54330000
ENSG00000249641
hyper


chr12.54338701.54339000
ENSG00000123364
hyper


chr12.54339001.54339300
ENSG00000123364
hyper


chr12.54339301.54339600
ENSG00000123364
hyper


chr12.54345901.54346200
ENSG00000123407
hyper


chr12.54354601.54354900
ENSG00000228630
hyper


chr12.54408301.54408600
ENSG00000273049
hyper


chr12.54408601.54408900
ENSG00000273049
hyper


chr12.54423301.54423600
ENSG00000273049
hyper


chr12.54424801.54425100
ENSG00000273049
hyper


chr12.54441001.54441300
ENSG00000198353
hyper


chr12.58021801.58022100
ENSG00000135454
hyper


chr12.81471601.81471900
ENSG00000111058
hyper


chr12.85673101.85673400
ENSG00000180318
hyper


chr12.85673401.85673700
ENSG00000180318
hyper


chr12.85674301.85674600
ENSG00000180318
hyper


chr12.95941801.95942100
ENSG00000136014
hyper


chr12.103344301.103344600
ENSG00000171759
hyper


chr12.106979401.106979700
ENSG00000257545
hyper


chr12.114838201.114838500
ENSG00000089225
hyper


chr12.114845701.114846000
ENSG00000089225
hyper


chr12.114846301.114846600
ENSG00000255399
hyper


chr12.114846601.114846900
ENSG00000255399
hyper


chr12.114847501.114847800
ENSG00000255399
hyper


chr12.114878101.114878400
ENSG00000255399
hyper


chr12.114878401.114878700
ENSG00000255399
hyper


chr12.114878701.114879000
ENSG00000255399
hyper


chr12.115107301.115107600
ENSG00000135111
hyper


chr12.115109401.115109700
ENSG00000135111
hyper


chr12.115173601.115173900
ENSG00000257817
hyper


chr12.128752201.128752500
ENSG00000181234
hyper


chr12.133484701.133485000
ENSG00000072609
hyper


chr12.133485001.133485300
ENSG00000072609
hyper


chr12.133485301.133485600
ENSG00000072609
hyper


chr13.58203601.58203900
ENSG00000118946
hyper


chr13.112716601.112716900
ENSG00000182968
hyper


chr13.78493201.78493500
ENSG00000136160
hyper


chr14.38724601.38724900
ENSG00000176435
hyper


chr14.38724901.38725200
ENSG00000176435
hyper


chr14.42077401.42077700
ENSG00000165379
hyper


chr13.23500201.23500500
ENSG00000262198
hyper


chr13.25320301.25320600
ENSG00000231417
hyper


chr13.25320601.25320900
ENSG00000231417
hyper


chr13.28492201.28492500
ENSG00000247381
hyper


chr13.28552801.28553100
ENSG00000183463
hyper


chr13.28674001.28674300
ENSG00000122025
hyper


chr13.58203901.58204200
ENSG00000118946
hyper


chr13.58206001.58206300
ENSG00000118946
hyper


chr13.78492901.78493200
ENSG00000136160
hyper


chr13.79170601.79170900
ENSG00000234377
hyper


chr13.95354701.95355000
ENSG00000238230
hyper


chr13.100608601.100608900
ENSG00000139800
hyper


chr13.100620301.100620600
ENSG00000139800
hyper


chr13.100641301.100641600
ENSG00000043355
hyper


chr13.100641601.100641900
ENSG00000043355
hyper


chr13.100641901.100642200
ENSG00000043355
hyper


chr13.108520201.108520500
ENSG00000204442
hyper


chr13.108520501.108520800
ENSG00000204442
hyper


chr13.108520801.108521100
ENSG00000204442
hyper


chr13.109147501.109147800
ENSG00000232087
hyper


chr13.109148401.109148700
ENSG00000232087
hyper


chr13.109148701.109149000
ENSG00000232087
hyper


chr13.112708501.112708800
ENSG00000200072
hyper


chr13.112712401.112712700
ENSG00000200072
hyper


chr14.29234701.29235000
ENSG00000176165
hyper


chr14.29235001.29235300
ENSG00000176165
hyper


chr14.29254501.29254800
ENSG00000186960
hyper


chr14.36979801.36980100
ENSG00000257520
hyper


chr14.36982201.36982500
ENSG00000257520
hyper


chr14.36982501.36982800
ENSG00000257520
hyper


chr14.36983401.36983700
ENSG00000257520
hyper


chr14.36983701.36984000
ENSG00000257520
hyper


chr14.36991801.36992100
ENSG00000253563
hyper


chr14.37116301.37116600
ENSG00000258661
hyper


chr14.37123501.37123800
ENSG00000258661
hyper


chr14.37128601.37128900
ENSG00000198807
hyper


chr14.38724301.38724600
ENSG00000176435
hyper


chr14.42074401.42074700
ENSG00000258636
hyper


chr14.52781701.52782000
ENSG00000125384
hyper


chr15.45996601.45996900
ENSG00000259200
hyper


chr15.75251401.75251700
ENSG00000198794
hyper


chr15.79383001.79383300
ENSG00000058335
hyper


chr14.52534801.52535100
ENSG00000087303
hyper


chr14.52535101.52535400
ENSG00000087303
hyper


chr14.52535401.52535700
ENSG00000087303
hyper


chr14.52536001.52536300
ENSG00000087303
hyper


chr14.52536301.52536600
ENSG00000087303
hyper


chr14.52734901.52735200
ENSG00000168229
hyper


chr14.52735501.52735800
ENSG00000168229
hyper


chr14.57261901.57262200
ENSG00000270163
hyper


chr14.57274801.57275100
ENSG00000165588
hyper


chr14.57275101.57275400
ENSG00000165588
hyper


chr14.57275401.57275700
ENSG00000165588
hyper


chr14.57276301.57276600
ENSG00000165588
hyper


chr14.57278701.57279000
ENSG00000248550
hyper


chr14.57279001.57279300
ENSG00000248550
hyper


chr14.60386401.60386700
ENSG00000261120
hyper


chr14.60975301.60975600
ENSG00000179008
hyper


chr14.60976201.60976500
ENSG00000179008
hyper


chr14.60976501.60976800
ENSG00000179008
hyper


chr14.61104601.61104900
ENSG00000258952
hyper


chr14.61109101.61109400
ENSG00000258952
hyper


chr14.61109701.61110000
ENSG00000126778
hyper


chr14.61110001.61110300
ENSG00000126778
hyper


chr14.61110601.61110900
ENSG00000126778
hyper


chr14.95234701.95235000
ENSG00000133937
hyper


chr14.95237701.95238000
ENSG00000133937
hyper


chr14.95238001.95238300
ENSG00000133937
hyper


chr14.99713101.99713400
ENSG00000127152
hyper


chr15.53075701.53076000
ENSG00000169856
hyper


chr15.53076601.53076900
ENSG00000169856
hyper


chr15.53080801.53081100
ENSG00000169856
hyper


chr15.76632001.76632300
ENSG00000159556
hyper


chr15.76633201.76633500
ENSG00000159556
hyper


chr15.79382701.79383000
ENSG00000058335
hyper


chr15.81410701.81411000
ENSG00000156206
hyper


chr15.88800601.88800900
ENSG00000260305
hyper


chr15.89903401.89903700
ENSG00000255571
hyper


chr15.89949301.89949600
ENSG00000255571
hyper


chr15.89949601.89949900
ENSG00000255571
hyper


chr15.89949901.89950200
ENSG00000255571
hyper


chr15.89952001.89952300
ENSG00000255571
hyper


chr16.51189001.51189300
ENSG00000103449
hyper


chr16.54324001.54324300
ENSG00000177508
hyper


chr17.46796401.46796700
ENSG00000159182
hyper


chr17.46832101.46832400
ENSG00000170703
hyper


chr17.48042901.48043200
ENSG00000199492
hyper


chr15.95388301.95388600
ENSG00000260521
hyper


chr15.95388601.95388900
ENSG00000260521
hyper


chr16.51184801.51185100
ENSG00000103449
hyper


chr16.89268601.89268900
ENSG00000259803
hyper


chr17.5974201.5974500
ENSG00000179314
hyper


chr15.96911401.96911700
ENSG00000259275
hyper


chr15.96959401.96959700
ENSG00000259542
hyper


chr16.3220501.3220800
ENSG00000262521
hyper


chr16.12994501.12994800
ENSG00000237515
hyper


chr16.12994801.12995100
ENSG00000237515
hyper


chr16.29086201.29086500
ENSG00000260908
hyper


chr16.49314601.49314900
ENSG00000102924
hyper


chr16.49314901.49315200
ENSG00000102924
hyper


chr16.51190201.51190500
ENSG00000103449
hyper


chr16.54318001.54318300
ENSG00000177508
hyper


chr16.54322201.54322500
ENSG00000177508
hyper


chr16.54970501.54970800
ENSG00000259711
hyper


chr16.54971401.54971700
ENSG00000259711
hyper


chr16.54971701.54972000
ENSG00000259711
hyper


chr16.54972301.54972600
ENSG00000259711
hyper


chr16.55362901.55363200
ENSG00000259283
hyper


chr16.55363201.55363500
ENSG00000259283
hyper


chr16.55364701.55365000
ENSG00000259283
hyper


chr16.55365001.55365300
ENSG00000259283
hyper


chr16.55365301.55365600
ENSG00000259283
hyper


chr16.56672101.56672400
ENSG00000205362
hyper


chr16.86529901.86530200
ENSG00000268388
hyper


chr17.7976101.7976400
ENSG00000179477
hyper


chr17.8868601.8868900
ENSG00000141506
hyper


chr17.8907601.8907900
ENSG00000065320
hyper


chr17.26554801.26555100
ENSG00000237575
hyper


chr17.27942301.27942600
ENSG00000264031
hyper


chr17.35285401.35285700
ENSG00000255509
hyper


chr17.36103201.36103500
ENSG00000108753
hyper


chr17.36103501.36103800
ENSG00000108753
hyper


chr17.36103801.36104100
ENSG00000108753
hyper


chr17.36104101.36104400
ENSG00000108753
hyper


chr17.37321501.37321800
ENSG00000141748
hyper


chr17.43974601.43974900
ENSG00000186868
hyper


chr17.46673701.46674000
ENSG00000120093
hyper


chr17.46796101.46796400
ENSG00000159182
hyper


chr17.46811701.46812000
ENSG00000242407
hyper


chr17.46824901.46825200
ENSG00000242407
hyper


chr17.47301901.47302200
ENSG00000173868
hyper


chr17.48042301.48042600
ENSG00000199492
hyper


chr17.48042601.48042900
ENSG00000199492
hyper


chr18.19746901.19747200
ENSG00000266010
hyper


chr19.2488501.2488800
ENSG00000099860
hyper


chr17.70113301.70113600
ENSG00000234899
hyper


chr18.907201.907500
ENSG00000265671
hyper


chr18.22929301.22929600
ENSG00000198795
hyper


chr18.44335801.44336100
ENSG00000101638
hyper


chr18.49868401.49868700
ENSG00000187323
hyper


chr19.20606101.20606400
ENSG00000231205
hyper


chr19.20606401.20606700
ENSG00000231205
hyper


chr19.37287901.37288200
ENSG00000267254
hyper


chr19.37288201.37288500
ENSG00000267254
hyper


chr19.38042401.38042700
ENSG00000267470
hyper


chr17.59534701.59535000
ENSG00000121075
hyper


chr17.62075101.62075400
ENSG00000264954
hyper


chr17.70112401.70112700
ENSG00000234899
hyper


chr17.70112701.70113000
ENSG00000234899
hyper


chr17.75370201.75370500
ENSG00000184640
hyper


chr18.905101.905400
ENSG00000265671
hyper


chr18.906601.906900
ENSG00000265671
hyper


chr18.906901.907200
ENSG00000265671
hyper


chr18.10032601.10032900
ENSG00000263630
hyper


chr18.12307501.12307800
ENSG00000176014
hyper


chr18.19745701.19746000
ENSG00000266010
hyper


chr18.19747501.19747800
ENSG00000266010
hyper


chr18.22929001.22929300
ENSG00000198795
hyper


chr18.31803001.31803300
ENSG00000101746
hyper


chr18.44790001.44790300
ENSG00000215474
hyper


chr18.55105801.55106100
ENSG00000119547
hyper


chr18.63418501.63418800
ENSG00000081138
hyper


chr18.63418801.63419100
ENSG00000081138
hyper


chr18.67068601.67068900
ENSG00000206052
hyper


chr18.67068901.67069200
ENSG00000206052
hyper


chr18.74961601.74961900
ENSG00000166573
hyper


chr18.76734601.76734900
ENSG00000263146
hyper


chr19.9608701.9609000
ENSG00000198028
hyper


chr19.12306001.12306300
ENSG00000234773
hyper


chr19.16479901.16480200
ENSG00000127527
hyper


chr19.20844001.20844300
ENSG00000269110
hyper


chr19.21182701.21183000
ENSG00000268326
hyper


chr19.22715101.22715400
ENSG00000197360
hyper


chr19.38182801.38183100
ENSG00000120784
hyper


chr19.38183101.38183400
ENSG00000120784
hyper


chr20.55500001.55500300
ENSG00000251772
hyper


chr19.46907401.46907700
ENSG00000169515
hyper


chr19.46993201.46993500
ENSG00000230510
hyper


chr19.48918001.48918300
ENSG00000105464
hyper


chr19.48918301.48918600
ENSG00000105464
hyper


chr19.58238401.58238700
ENSG00000269026
hyper


chr21.38082901.38083200
ENSG00000159263
hyper


chr21.46360201.46360500
ENSG00000160256
hyper


chr19.44952301.44952600
ENSG00000267188
hyper


chr19.44952601.44952900
ENSG00000267188
hyper


chr19.46929901.46930200
ENSG00000169515
hyper


chr19.52839301.52839600
ENSG00000269535
hyper


chr19.52839601.52839900
ENSG00000269535
hyper


chr19.52873201.52873500
ENSG00000221923
hyper


chr19.53073301.53073600
ENSG00000167562
hyper


chr19.54401701.54402000
ENSG00000126583
hyper


chr19.56879701.56880000
ENSG00000131848
hyper


chr19.56904901.56905200
ENSG00000018869
hyper


chr19.56989201.56989500
ENSG00000198046
hyper


chr19.56989501.56989800
ENSG00000166770
hyper


chr19.58095001.58095300
ENSG00000171649
hyper


chr19.58220401.58220700
ENSG00000204519
hyper


chr19.58238701.58239000
ENSG00000269026
hyper


chr19.58400101.58400400
ENSG00000204514
hyper


chr19.58520701.58521000
ENSG00000176593
hyper


chr19.58873201.58873500
ENSG00000268230
hyper


chr19.58951201.58951500
ENSG00000131849
hyper


chr19.58951501.58951800
ENSG00000131849
hyper


chr20.291001.291300
ENSG00000225377
hyper


chr20.865801.866100
ENSG00000101280
hyper


chr20.5296201.5296500
ENSG00000101292
hyper


chr20.5296501.5296800
ENSG00000101292
hyper


chr20.5296801.5297100
ENSG00000101292
hyper


chr20.5297101.5297400
ENSG00000101292
hyper


chr20.9489301.9489600
ENSG00000225988
hyper


chr20.9495301.9495600
ENSG00000225988
hyper


chr20.10198501.10198800
ENSG00000227906
hyper


chr20.21501901.21502200
ENSG00000125820
hyper


chr20.21681901.21682200
ENSG00000125813
hyper


chr20.21687301.21687600
ENSG00000125813
hyper


chr20.21694201.21694500
ENSG00000125813
hyper


chr20.21694501.21694800
ENSG00000125813
hyper


chr20.21694801.21695100
ENSG00000125813
hyper


chr20.22548901.22549200
ENSG00000259974
hyper


chr20.22549201.22549500
ENSG00000259974
hyper


chr20.22558201.22558500
ENSG00000259974
hyper


chr20.22563601.22563900
ENSG00000125798
hyper


chr20.37356301.37356600
ENSG00000101438
hyper


chr20.37357501.37357800
ENSG00000101438
hyper


chr20.45087601.45087900
ENSG00000215452
hyper


chr20.45087901.45088200
ENSG00000215452
hyper


chr20.55500901.55501200
ENSG00000251772
hyper


chr21.22369801.22370100
ENSG00000154654
hyper


chr21.34398301.34398600
ENSG00000227757
hyper


chr21.34398601.34398900
ENSG00000227757
hyper


chr21.34443301.34443600
ENSG00000227757
hyper


chr21.34444201.34444500
ENSG00000184221
hyper


chr21.38066701.38067000
ENSG00000224269
hyper


chr21.38069401.38069700
ENSG00000224269
hyper


chr21.38069701.38070000
ENSG00000224269
hyper


chr21.38077201.38077500
ENSG00000159263
hyper


chr21.38077501.38077800
ENSG00000159263
hyper


chr22.48963001.48963300
ENSG00000219438
hyper


chr22.50629801.50630100
ENSG00000170638
hyper


chr22.51042001.51042300
ENSG00000008735
hyper


chr1.147736201.147736500
ENSG00000199879
hypo


chr1.27319201.27319500
ENSG00000253368
hypo


chr1.50489401.50489700
ENSG00000186094
hypo


chr1.11097601.11097900
ENSG00000009724
hypo


chr2.26593501.26593800
ENSG00000138018
hypo


chr2.39004801.39005100
ENSG00000152147
hypo


chr1.155826901.155827200
ENSG00000116580
hypo


chr2.44314201.44314500
ENSG00000219391
hypo


chr2.96810301.96810600
ENSG00000158050
hypo


chr2.96970501.96970800
ENSG00000144028
hypo


chr2.239685301.239685600
ENSG00000226992
hypo


chr4.181317301.181317600
ENSG00000251025
hypo


chr4.159644701.159645000
ENSG00000171497
hypo


chr5.391201.391500
ENSG00000063438
hypo


chr5.5886901.5887200
ENSG00000261037
hypo


chr5.34306501.34306800
ENSG00000215158
hypo


chr5.125937001.125937300
ENSG00000164902
hypo


chr6.35181601.35181900
ENSG00000146197
hypo


chr6.114180901.114181200
ENSG00000155130
hypo


chr5.170745301.170745600
ENSG00000164438
hypo


chr6.37070101.37070400
ENSG00000216412
hypo


chr7.6387901.6388200
ENSG00000178397
hypo


chr7.5553601.5553900
ENSG00000155034
hypo


chr7.57484501.57484800
ENSG00000270957
hypo


chr7.130275301.130275600
ENSG00000239021
hypo


chr8.17770201.17770500
ENSG00000104760
hypo


chr9.1009201.1009500
ENSG00000228783
hypo


chr9.132388201.132388500
ENSG00000148335
hypo


chr9.136890301.136890600
ENSG00000235106
hypo


chr10.71905201.71905500
ENSG00000156521
hypo


chr10.1584301.1584600
ENSG00000185736
hypo


chr11.1404601.1404900
ENSG00000174672
hypo


chr12.52317301.52317600
ENSG00000139567
hypo


chr12.122459101.122459400
ENSG00000110987
hypo


chr12.122687701.122688000
ENSG00000158113
hypo


chr12.130527001.130527300
ENSG00000261650
hypo


chr12.133050001.133050300
ENSG00000269676
hypo


chr14.50540401.50540700
ENSG00000273065
hypo


chr14.103995001.103995300
ENSG00000260285
hypo


chr17.17739301.17739600
ENSG00000072310
hypo


chr16.67562401.67562700
ENSG00000039523
hypo


chr16.84545101.84545400
ENSG00000140950
hypo


chr16.89939401.89939700
ENSG00000141002
hypo


chr17.46184401.46184700
ENSG00000002919
hypo


chr16.3209101.3209400
ENSG00000261889
hypo


chr19.17457001.17457300
ENSG00000130299
hypo


chr19.30363901.30364200
ENSG00000267433
hypo


chr17.75468301.75468600
ENSG00000184640
hypo


chr17.81082801.81083100
ENSG00000262898
hypo


chr19.8408101.8408400
ENSG00000186994
hypo


chr19.14332201.14332500
ENSG00000240803
hypo


chr20.60717001.60717300
ENSG00000101182
hypo


chr21.9438301.9438600
ENSG00000238411
hypo


chr19.45004801.45005100
ENSG00000167384
hypo


chr19.50880001.50880300
ENSG00000131408
hypo


chr20.45439801.45440100
ENSG00000266136
hypo









Genomic Mutation Profile


The present disclosure provides methods, systems, and kits for producing a mutation profile of a subject that has a disease/condition or is suspected of having such disease/condition, wherein the methylation profile may be used to determine whether the subject has the disease/condition or is at risk of having the disease/condition. The samples disclosed herein are subjected to library preparation and next generation deep sequencing (e.g., CAPP-Seq). A plurality of sequencing reads is generated and analyzed. In some embodiments, deep sequencing may be configured to maximize identifying genomic mutations associated with the disease/condition. For example, not meant to be limiting, for head and neck squamous cell carcinoma (HNSCC), a panel of canonical HNSCC driver genes may be included in the selector for CAPP-seq. Further, for lung cancer, a panel of lung cancer drive genes may be included in the selector for CAPP-seq. Moreover, for pancreatic cancer, a panel of pancreatic cancer drive genes may be included in the selector for CAPP-seq. In some embodiments, including genes without known driver effects in a particular cancer type in the selector for CAPP-seq may increase the sensitivity of ctDNA detection.


In some embodiments, the relative measure of ctDNA abundance is calculate from the mean mutant allele fractions (MAFs). In some embodiments, the mean MAF of mutations identified a subject and comprised in his/her mutation profile ranges from at least about 0.01% to at least about 10%. The ctDNA fraction of a sample disclosed herein is about at least 0.01%, 0.02%, 0.03%, 0.04%, 0.05%, 0.06%, 0.07%, 0.08%, 0.09%, 0.1%, 0.15%, 0.2%, 0.5%, 1%, 1.5%, 2%, 2.5%, 3%, 3.5%, 4%, 4.5%, 5%, 5.5%, 6%, 6.5%, 7%, 7.5%, 8%, 8.5%, 9%, 9.5%, 10%, or any percentage in between.


In some embodiments, the generated mutation profile of a subject does not include mutation variants derived from cell-free nucleic acid molecules derived from PBLs. In some embodiments, the mutation profile comprises genetic polymorphisms, such as missense variant, a nonsense variant, a deletion variant, an insertion variant, a duplication variant, an inversion variant, a frameshift variant, or a repeat expansion variant. In some embodiments, the mutation profile may comprise mutation variant derived from a fraction of cell-free nucleic acid molecules of a specific size range.


Fragment Length Profile


In some embodiment, the length of ctDNA fragments is shorter than cell-free nucleic acid molecules derived from a healthy subject. In some embodiments, the length of ctDNA comprising at least one mutation is shorter than the length of cell free nucleic acid molecule containing a corresponding reference allele. In some embodiments, a length of a ctDNA fragment containing at least one DMR is shorter than a cell-free nucleic acid molecule fragment containing the corresponding genomic region.


In some embodiments, the sequencing does not utilize bisulfite sequence because it causes degradation of ctDNA fragments and prevents the preservation of the length distribution of ctDNAs. In some embodiments, the fragment length of ctDNA is at least from 60 to 500 bp, 80 to 300 bp, 90 to 250 bp, 80 to 170 bp, or 100 to 150 bp. In some embodiments, the present disclosure provides an enrichment of the cell free nucleic acid samples based on selecting cell free molecules of a certain size. In some embodiments, the multimodal analysis comprises utilizing the mutation profile described herein and the fragment length profile by selectively including a plurality of nucleic acid molecules in the mutation profile based on their fragment length. In some embodiments, the multimodal analysis comprises utilizing the methylation profile described herein and the fragment length profile by selectively including a plurality of nucleic acid molecules in the methylation profile based on their fragment length. In some embodiments, the multimodal analysis comprises utilizing the mutation profile, methylation profile, and the fragment length profile together by selectively including a plurality of nucleic acid molecules in the mutation profile based on their fragment length and by selectively including a plurality of nucleic acid molecules in the methylation profile based on their fragment length respectively.


Methods and Systems for Detecting Cancer, Determining Tissue of Origin for Tumor, and Providing Prognosis


The present disclosure provides methods and systems for determining whether a subject has or is at risk of having a disease, wherein the methods and systems comprises subjecting a plurality of nucleic acid molecules derived from a cell-free nucleic acid sample obtained from said subject to sequencing to generate at least one profile of (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and processing said at least one profile to determine whether said subject has or is at risk of said disease at a sensitivity of at least 80% or at a specificity of at least about 90%, wherein said cell-free nucleic acid sample comprises less than 30 ng/ml of said plurality of nucleic acid molecules. In some embodiments, the sensitivity is at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers. In some embodiments, the specificity is at least about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers.


In some embodiments, the methods and systems comprises subjecting a plurality of nucleic acid molecules derived from a cell-free nucleic acid sample obtained from said subject to sequencing to generate at least two profiles of (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile. The methods provide a sensitivity of at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers. In some embodiments, the sensitivity when using two profiles is increased by at least about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, or percentage in between any of the numbers compared to the sensitivity when using one profile. In some embodiments, the sensitivity when using three profiles is increased by at least about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, or percentage in between any of the numbers compared to the sensitivity when using two profile.


Further, the methods provide a specificity of at least about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers. In some embodiments, the specificity when using two profiles is increased by at least about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, or percentage in between any of the numbers compared to the specificity when using one profile. In some embodiments, the specificity when using three profiles is increased by at least about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, or percentage in between any of the numbers compared to the specificity when using two profile.


The present disclosure provides methods and systems for processing a cell-free nucleic acid sample of a subject to determine whether said subject has or is at risk of having a disease, the methods and systems comprise providing said cell-free nucleic acid sample comprising a plurality of nucleic acid molecules; subjecting said plurality of nucleic acid molecules or derivatives thereof to sequencing to generate a plurality of sequencing reads; computer processing said plurality of sequencing reads to identify, for said plurality of nucleic acid molecules, (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and using at least said methylation profile, said mutation profile and said fragment length profile to determine whether said subject has or is at risk of having said disease. In some embodiments, the methods provide a sensitivity of at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers. The methods provide a specificity of at least about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers.


The present disclosure provides methods and systems for determining a tissue origin of a tumor, comprising identifying a plurality of Differentially Methylated Regions (DMRs), wherein the plurality of DMRs is specific for a particular cancer (e.g., breast cancer, colon cancer, prostate cancer, HSNCC) and derived from a fraction of cell-free nucleic acid molecules. In some embodiments, the fraction of the cell-free nucleic acid molecules is derived from ctDNA. In some embodiments, the methods provides a sensitivity of at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers. The methods provide a specificity of at least about 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, or any percentage in between the numbers.


The present disclosure describes methods and systems for providing a prognosis to a subject after receiving a treatment for a disease/condition. For example, the treatment comprises a surgical removal of a tumor, a chemotherapy designed for a specific type of cancer, a radio therapy, or an immune therapy (e.g., TCR, CAR, etc.). In some embodiments, the methods or systems comprise subjecting a plurality of nucleic acid molecules derived from a cell-free nucleic acid sample obtained from said subject to sequencing to generate at least one profile of (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and monitoring or detecting minimal residual disease (MRD) based at least based on the at least one profile.


The present disclosure provides methods and systems for determining whether a subject has a disease/condition by assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject; detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 5; and comparing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 5. In some embodiments, the methylation level of at least about six or more, ten or more, fifteen or more, twenty or more, thirty or more, forty or more, fifty or more, sixty or more, seventy or more, eighty or more, ninety or more, or one hundred or more, two hundred or more, three hundred or more, four hundred or more, five hundred or more, six hundred or more, or seven hundred or more DMRs listed in Table 5 is measured and compared to the methylation level of the corresponding DMRs in a healthy subject as discussed herein.


Once a subject is accurately diagnosed and receives a treatment to treat the cancer, such as surgical removal, chemotherapy, radio therapy, etc., it is important to monitor the effectiveness of the treatment and predict the patient's survival rate. Further, it is important to detect minimal residual disease of cancer cells. The present disclosure provides methods and systems for determining whether a subject has a higher survival rate after receiving a treatment for a disease, the methods and systems comprise assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject; detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 6; and comparing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 6. In some embodiments, the DMRs listed in Table 6 represent regions associated with genes ZSCAN31, LINC01391, GATA2-AS1, STK3, and OSR1.












Table 6 ctDNA derived DMR










ensemblId - DMR



windowPos - DMR genomic region
associated gene ID
DMR





chr2.19555801.19556100
ENSG00000143867
hyper


chr3.128210701.128211000
ENSG00000179348
hyper


chr3.138657301.138657600
ENSG00000244578
hyper


chr6.28303801.28304100
ENSG00000235109
hyper


chr8.99951901.99952200
ENSG00000104375
hyper









In some embodiments, the method further comprises the step of adding a second amount of control DNA to the sample for confirming the immunoprecipitation reaction.


As used herein, the “control” may comprise both positive and negative control, or at least a positive control.


In some embodiments, the method further comprises the step of adding a second amount of control DNA to the sample for confirming the capture of cell-free methylated DNA.


In some embodiments, identifying the presence of DNA from cancer cells further includes identifying the cancer cell tissue of origin.


In some instances, tumor tissue sampling may be challenging or carry significant risks, in which case diagnosing and/or subtyping the cancer without the need for tumor tissue sampling may be desired. For example, lung tumor tissue sampling may require invasive procedures such as mediastinoscopy, thoracotomy, or percutaneous needle biopsy; these procedures may result in a need for hospitalization, chest tube, mechanical ventilation, antibiotics, or other medical interventions. Some individuals may not undergo the invasive procedures needed for tumor tissue sampling either because of medical comorbidities or due to preference. In some instances, the actual procedure for tumor tissue procurement may depend on the suspected cancer subtype. In other instances, cancer subtype may evolve over time within the same individual; serial assessment with invasive tumor tissue sampling procedures is often impractical and not well tolerated by patients. Thus, non-invasive cancer subtyping via blood test may have many advantageous applications in the practice of clinical oncology.


Accordingly, in some embodiments, identifying the cancer cell tissue of origin further includes identifying a cancer subtype. Preferably, the cancer subtype differentiates the cancer based on stage (e.g., early stage lung cancer treated with surgery vs late stage lung cancer treated with chemotherapy), histology (e.g., small cell carcinoma vs adenocarcinoma vs squamous cell carcinoma in lung cancer), gene expression pattern or transcription factor activity (e.g., ER status in breast cancer), copy number aberrations (e.g., HER2 status in breast cancer), specific rearrangements (e.g., FLT3 in AML), specific gene point mutational status (e.g., IDH gene point mutations), and DNA methylation patterns (e.g., MGMT gene promoter methylation in brain cancer).


In some embodiments, comparison in step (f) is carried out genome-wide.


In other embodiments, the comparison in step (f) is restricted from genome-wide to specific regulatory regions, such as, but not limited to, FANTOM5 enhancers, CpG Islands, CpG shores, CpG Shelves, or any combination of the foregoing.


In some embodiments, the methods herein are for use in the detection of the cancer.


In some embodiments, the methods herein are for use in monitoring therapy of the cancer.


Data Analysis Systems and Methods


The methods and systems disclosed herein may comprises algorithms or uses thereof. The one or more algorithms may be used to classify one or more samples from one or more subjects. The one or more algorithms may be applied to data from one or more samples. The data may comprise biomarker expression data. The methods disclosed herein may comprise assigning a classification to one or more samples from one or more subjects. Assigning the classification to the sample may comprise applying an algorithm to the methylation profile, mutation profile, and fragment length profile. In some cases, the at least one profile is inputted to a data analysis system comprising a trained algorithm for classifying the sample as obtained from a subject has a disease or minor injuries.


A data analysis system may be a trained algorithm. The algorithm may comprise a linear classifier. In some instances, the linear classifier comprises one or more of linear discriminant analysis, Fisher's linear discriminant, Naïve Bayes classifier, Logistic regression, Perceptron, Support vector machine, or a combination thereof. The linear classifier may be a support vector machine (SVM) algorithm. The algorithm may comprise a two-way classifier. The two-way classifier may comprise one or more decision tree, random forest, Bayesian network, support vector machine, neural network, or logistic regression algorithms.


The algorithm may comprise one or more linear discriminant analysis (LDA), Basic perceptron, Elastic Net, logistic regression, (Kernel) Support Vector Machines (SVM), Diagonal Linear Discriminant Analysis (DLDA), Golub Classifier, Parzen-based, (kernel) Fisher Discriminant Classifier, k-nearest neighbor, Iterative RELIEF, Classification Tree, Maximum Likelihood Classifier, Random Forest, Nearest Centroid, Prediction Analysis of Microarrays (PAM), k-medians clustering, Fuzzy C-Means Clustering, Gaussian mixture models, graded response (GR), Gradient Boosting Method (GBM), Elastic-net logistic regression, logistic regression, or a combination thereof. The algorithm may comprise a Diagonal Linear Discriminant Analysis (DLDA) algorithm. The algorithm may comprise a Nearest Centroid algorithm. The algorithm may comprise a Random Forest algorithm. In some embodiments, for discrimination of preeclampsia and non-preeclampsia, the performance of logistic regression, random forest, and gradient boosting method (GBM) is superior to that of linear discriminant analysis (LDA), neural network, and support vector machine (SVM).


Kits


The present disclosure provides kits for identifying or monitoring a disease or disorder (e.g., cancer) of a subject. A kit may comprise probes for identifying a quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of a panel of cancer-associated genomic loci in a sample of the subject. A quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of a panel of cancer-associated genomic loci in the sample may be indicative of the disease or disorder (e.g., cancer) of the subject. The probes may be selective for the sequences at the panel of cancer-associated genomic loci (e.g., DMR listed in Tables 3, 5 and 6) in the sample. A kit may comprise instructions for using the probes to process the sample to generate datasets indicative of a quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of the panel of cancer-associated genomic loci in a sample of the subject.


The probes in the kit may be selective for the sequences at the panel of cancer-associated genomic loci in the sample. The probes in the kit may be configured to selectively enrich nucleic acid (e.g., RNA or DNA) molecules corresponding to the panel of cancer-associated genomic loci. The probes in the kit may be nucleic acid primers. The probes in the kit may have sequence complementarity with nucleic acid sequences from one or more of the panel of cancer-associated genomic loci or genomic regions. The panel of cancer-associated genomic loci or microbiome-associated genomic loci or genomic regions may comprise at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, or more distinct panel of cancer-associated genomic loci or genomic regions.


The instructions in the kit may comprise instructions to assay the sample using the probes that are selective for the sequences at the panel of cancer-associated genomic loci in the cell-free biological sample. These probes may be nucleic acid molecules (e.g., RNA or DNA) having sequence complementarity with nucleic acid sequences (e.g., RNA or DNA) from one or more of the plurality of panel of cancer-associated genomic loci. These nucleic acid molecules may be primers or enrichment sequences. The instructions to assay the cell-free biological sample may comprise introductions to perform array hybridization, polymerase chain reaction (PCR), or nucleic acid sequencing (e.g., DNA sequencing or RNA sequencing) to process the sample to generate datasets indicative of a quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of the panel of cancer-associated genomic loci in the sample. A quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of a panel of cancer-associated genomic loci in the sample may be indicative of a disease or disorder (e.g., cancer).


The instructions in the kit may comprise instructions to measure and interpret assay readouts, which may be quantified at one or more of the panel of cancer-associated genomic loci to generate the datasets indicative of a quantitative measure (e.g., indicative of a presence, absence, or relative amount) of sequences at each of the panel of cancer-associated genomic loci in the sample. For example, quantification of array hybridization or polymerase chain reaction (PCR) corresponding to the panel of cancer-associated genomic loci may generate the datasets indicative of a quantitative measure (e.g., indicative of apresence, absence, or relative amount) of sequences at each of the panel of cancer-associated genomic loci in the sample. Assay readouts may comprise quantitative PCR (qPCR) values, digital PCR (dPCR) values, digital droplet PCR (ddPCR) values, fluorescence values, etc., or normalized values thereof.


Computer System


In some embodiments, certain steps are carried out by a computer processor. The present system and method may be practiced in various embodiments. A suitably configured computer device, and associated communications networks, devices, software and firmware may provide a platform for enabling one or more embodiments as described above. By way of example, FIG. 8 shows a generic computer device 100 that may include a central processing unit (“CPU”) 102 connected to a storage unit 104 and to a random access memory 106. The CPU 102 may process an operating system 101, application program 103, and data 123. The operating system 101, application program 103, and data 123 may be stored in storage unit 104 and loaded into memory 106, as may be required. Computer device 100 may further include a graphics processing unit (GPU) 122 which is operatively connected to CPU 102 and to memory 106 to offload intensive image processing calculations from CPU 102 and run these calculations in parallel with CPU 102. An operator 107 may interact with the computer device 100 using a video display 108 connected by a video interface 105, and various input/output devices such as a keyboard 115, mouse 112, and disk drive or solid state drive 114 connected by an I/O interface 109. The mouse 112 may be configured to control movement of a cursor in the video display 108, and to operate various graphical user interface (GUI) controls appearing in the video display 108 with a mouse button. The disk drive or solid state drive 114 may be configured to accept computer readable media 116. The computer device 100 may form part of a network via a network interface 111, allowing the computer device 100 to communicate with other suitably configured data processing systems (not shown). One or more different types of sensors 135 may be used to receive input from various sources.


The present system and method may be practiced on virtually any manner of computer device including a desktop computer, laptop computer, tablet computer or wireless handheld. The present system and method may also be implemented as a computer-readable/useable medium that includes computer program code to enable one or more computer devices to implement each of the various process steps in a method in accordance with the present invention. In case of more than computer devices performing the entire operation, the computer devices are networked to distribute the various steps of the operation. It is understood that the terms computer-readable medium or computer useable medium comprises one or more of any type of physical embodiment of the program code. In particular, the computer-readable/useable medium may comprise program code embodied on one or more portable storage articles of manufacture (e.g. an optical disc, a magnetic disk, a tape, etc.), on one or more data storage portioned of a computing device, such as memory associated with a computer and/or a storage system.


As used herein, “processor” may be any type of processor, such as, for example, any type of general-purpose microprocessor or microcontroller (e.g., an Intel™ x86, PowerPC™, ARM™ processor, or the like), a digital signal processing (DSP) processor, an integrated circuit, a field programmable gate array (FPGA), or any combination thereof.


As used herein “memory” may include a suitable combination of any type of computer memory that is located either internally or externally such as, for example, random-access memory (RAM), read-only memory (ROM), compact disc read-only memory (CDROM), electro-optical memory, magneto-optical memory, erasable programmable read-only memory (EPROM), and electrically-erasable programmable read-only memory (EEPROM), or the like. Portions of memory 102 may be organized using a conventional filesystem, controlled and administered by an operating system governing overall operation of a device.


As used herein, “computer readable storage medium” (also referred to as a machine-readable medium, a processor-readable medium, or a computer usable medium having a computer-readable program code embodied therein) is a medium capable of storing data in a format readable by a computer or machine. The machine-readable medium may be any suitable tangible, non-transitory medium, including magnetic, optical, or electrical storage medium including a diskette, compact disk read only memory (CD-ROM), memory device (volatile or non-volatile), or similar storage mechanism. The computer readable storage medium may contain various sets of instructions, code sequences, configuration information, or other data, which, when executed, cause a processor to perform steps in a method according to an embodiment of the disclosure. Those of ordinary skill in the art will appreciate that other instructions and operations necessary to implement the described implementations may also be stored on the computer readable storage medium. The instructions stored on the computer readable storage medium may be executed by a processor or other suitable processing device, and may interface with circuitry to perform the described tasks.


As used herein, “data structure” a particular way of organizing data in a computer so that it may be used efficiently. Data structures may implement one or more particular abstract data types (ADT), which specify the operations that may be performed on a data structure and the computational complexity of those operations. In comparison, a data structure is a concrete implementation of the specification provided by an ADT.


The advantages of the present invention are further illustrated by the following examples. The examples and their particular details set forth herein are presented for illustration only and should not be construed as a limitation on the claims of the present invention.


Examples

Materials & Methods


HNSCC and Healthy Donor Peripheral Blood Leukocyte (PBL) and Plasma Acquisition


Patients diagnosed with HNSCC between 2014-2016 were identified from a prospective Anthology of Clinical Outcomes (Wong K. et al. 2010). All studies were approved by the Research Ethics Board at University Health Network. HNSCC patient samples were obtained from the Princess Margaret Cancer Centre's HNC Translational Research program based on the following criteria: 1) presentation of localized disease at diagnosis, 2) collection of blood at diagnosis and at least one timepoint post-treatment, 3) minimum follow-up time of 2 years after diagnosis. All patients received curative-intent treatment consisting of surgery with or without adjuvant radiotherapy. Healthy donors matched by age, gender, and current smoking status were identified from a prospective lung cancer screening program. 5-10 mL of blood was collected in Ethylene-Diamine-Tetraacetic Acid (EDTA) tubes. For HNSCC patients, blood was collected at diagnosis (baseline, BL) as well as three months after primary surgery (3M). Where applicable, additional blood was collected prior to adjuvant radiotherapy (PreRT), mid adjuvant radiotherapy (MidRT), and/or 12 months after primary surgery (12M). Plasma was isolated from blood within 1 hour of collection and stored at −80° C. until further processing. From the same blood collection for HNSCC patients at diagnosis or healthy donors, peripheral blood leukocytes were also isolated.


Cell Culture


The HPV-negative HNSCC cell line, FaDu, was kindly provided by Dr. Bradly Wouters (Princess Margaret Cancer Center) and cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. FaDu cell cultures were incubated in a humidified atmosphere containing 5% CO2 at 37° C. The identity of FaDu cells was confirmed by STR profiling. Cells were subjected to mycoplasma testing (e-MycoTMVALiD Mycoplasma PCR Detection Kit, Intron Bio) prior to use.


Isolation of Cell-Free DNA (cfDNA) and PBL Genomic DNA (gDNA)


cfDNA was isolated from total plasma using the QIAamp Circulating Nucleic Acid Kit (Qiagen) following manufacturer's instructions. Genomic DNA was isolated from PBLs, sheared to 150-200 base-pairs using the Covaris M220 Focused-ultrasonicator, and size-selected by AMPure XP magnetic beads (Beckman Coulter) to remove fragments above 300 base-pairs. Isolated cfDNA and sheared PBL genomic DNA were quantified by Qubit prior to library generation (FIGS. 9A and 9B).


Sequencing Library Preparation


5-10 or 10-20 ng of DNA was used as input for cfMeDIP-seq or CAPP-seq respectively. Input DNA was prepared for library generation using the KAPA HyperPrep Kit (KAPA Biosystems) with some modifications. Library adapters were utilized which incorporate a random 2-bp sequence followed by a constant 1-bp T sequence 5′ adjacent to both strands of input DNA upon ligation. To minimize adapter dimerization during ligation, library adapters were added at a 100:1 adapter:DNA molar ratio (˜0.07 uM per 10 ng of cfDNA) and incubated at 4° C. for 17 hours overnight. After post-ligation cleanup, input DNA was eluted in 40 uL of elution buffer (EB, 10 mM Tris-HCl, pH 8.0-8.5) prior to library generation.


Generation of CAPP-Seq Libraries


Generation of CAPP-seq libraries were performed as described from Newman et al. 2014 with some modification. Libraries were PCR amplified at 10 cycles and up to 12 indexed amplified libraries were pooled together at 500-1000 ng. After the addition of COT DNA and blocking oligos, pooled libraries underwent SpeedVac treatment to evaporate all liquids and were resuspended in 13 uL resuspension mix (8.5 uL 2× Hybridization buffer, 3.4 uL Hybridization Component A, 1.1 uL nuclease-free water). 4 uL of hybridization probes (i.e. HNSCC selector) was added to the resuspension mix for a total of 17 uL prior to hybridization. After hybridization and PCR amplification/cleanup, libraries were eluted in 30 uL of IDTE pH 8.0 (1×TE solution). Multiplexed libraries were sequenced at 2×75/100/125 paired runs on the Illumina NextSeq/NovaSeq/HiSeq4000 respectively. Design of the HNSCC selector incorporated frequently recurrent genomic alterations in HNSCC from the COSMIC database as well as the E6 and E7 region of the HPV-16 genome (FIG. 11).


Alignment and Quality Control of CAPP-Seq Libraries


The first two base-pairs on each 5′ end of unaligned paired reads, corresponding to the incorporated random molecular barcodes, were extracted and collated to generate a 4-bp molecular identifier (UMI). The third T base-pair spacer was also removed prior to alignment. Paired reads were aligned to the human genome (genome assembly GRCh37/hg19) by BWA-mem, sorted and indexed by SAMtools (v 1.3.1) and recalibrated for base quality score using the Genome Analysis ToolKit (GATK) BaseRecalibrator (v 3.8) according to best practices (reference). Duplicated sequences from BAM files were collapsed based on their UMIs and labeled as Singletons, Single-Strand Consensus Sequences (SSCS) or Duplex Consensus Sequences (DCS) by ConsensusCruncher44. Quality control of each library was assessed by various metrics obtained form FastQC (Babraham Bioinformatics), as well as various scripts to obtain capture efficiency (CollectHsMetrics, Picard 2.10.9), depth of coverage (DepthOfCoverage, GATK 3.8), and base-pair position error rate (ides-bgreport.pl, Newman et al. 2016).


Detection of Somatic Nucleotide Variants (SNVs) and Quantification of ctDNA


Removal of potential sequencing errors was performed by integrated Digital Error Suppression (iDES) as described by Newman et al. 2016. Background polishing was performed by utilization of our 20 healthy donor cfDNA samples as the training cohort (FIG. 12). To prevent the influence of outliers on downstream analysis, candidate SNVs within the lower 15th or upper 85th percentile of sequencing depth (<=1500×, >=5000×) across HNSCC cfDNA or PBL gDNA samples as well as genes with an average sequencing depth <=500× were excluded from analysis. To account for clonal hematopoiesis, non-germline mutations were defined as having a mutant allele fractions below 10% in plasma. Candidate SNVs in HNSCC cfDNA samples were identified based on the criteria of >=3 supporting reads with duplex support and complete absence in matched PBL gDNA samples. The mutant allele fraction (MAF) of identified SNVs was calculated by the number of reads corresponding to the alternative allele, divided by the sum of reads corresponding to the alternative and reference allele. For each HNSCC cfDNA sample with identifiable SNVs, the mean MAF across SNVs was calculated and used as a measure of ctDNA abundance. In cfDNA samples with only one identifiable SNV, the calculated MAF was used. Many of the detectable cancer-derived mutations may not be homozygous and may not be clonal within the tumor, and for these reasons the mean MAF may be an underestimate of the true ctDNA abundance within cell-free DNA


Generation of cfMeDIP-seq Libraries


The cfMeDIP-seq protocol was performed as described by Shen et al. 2019 with modifications to the library preparation step as described in “Sequencing Library Preparation”. Multiplexed libraries were sequenced at 2×75/100/125 paired runs on the Illumina NextSeq/NovaSeq/HiSeq4000 respectively. For generalizability, cfMeDIP-seq libraries are described as any MeDIP-seq preparation method utilizing 5-10 ng of input DNA regardless of source (i.e. cfDNA, gDNA).


Alignment and Quality Control of cfMeDIP-Seq Libraries


Unaligned paired reads were processed, aligned, sorted and indexed as previously described in Alignment and Quality Control of CAPP-seq Libraries. Duplicated sequences from BAM files were collapsed by SAMtools. Quality control of each library was assessed by various metrics obtained form FastQC (Babraham Bioinformatics), as well as various metrics obtained from the R package MEDIPS (reference) including CpG coverage (MEDIPS.seqCoverage) and enrichment (MEDIPS.CpGenrich).


Selection of Informative Regions in cfMeDIP-Seq Profiles


Fragments generated from paired reads of cfMeDIP-seq libraries were counted within non-overlapping 300 base-pair windows by MEDIPS (MEDIPS.createSet), scaled by Reads Per Kilobase per Million (RPKM), and exported as WIG format (MEDIPS.exportWIG). WIG files from each sample were imported by R and collated as a matrix. Analysis was limited to cfDNA and PBL samples from our 20 healthy donor samples to enable applications within a non-disease context. Informative regions were based on the criteria of CpG density and correlation of RPKM values between cfDNA and matched PBLs. Employing a sliding window based on CpG density (>=n CpGs), a minimum threshold of >=8 CpGs was selected.


Calculation of Absolute Methylation from cfMeDIP-Seq Libraries


Fragments from paired reads of cfMeDIP-seq libraries were counted as previously described in Selection of Informative Regions in cfMeDIP-seq Profiles and scaled to absolute methylation levels by the MeDEStrand R package. To calculate absolute methylation from counts, a logistic regression model was used to estimate bias of DNA pulldown based on CpG density (i.e. CpG density bias) (MeDEStrand.calibrationCurve). Based on the estimated CpG density bias, methylation within each window was corrected for fragments from the positive and negative DNA strand. Windows with corrected fragments were log transformed and scaled to values between 0 and 1 to describe absolute methylation (MeDEStrand.binMethyl). Absolute methylation levels from each cfMeDIP-seq sample was exported as a WIG-like file (i.e. WIG file format without a track-line).


Design of In-Silico PBL Depletion and Evaluation of Performance


To enrich for windows within the disease setting, methylation from PBLs was removed by a process termed “in-silico PBL depletion”. Analysis was limited to PBL samples from our cohort of 20 healthy donor samples to enable applications within a non-cancer specific context. Our strategy for the in-silico PBL depletion was performed as followed:

    • 1. For each informative window as described in Selection of Informative Regions in cfMeDIP-seq Profiles, calculate the median absolute methylation value across healthy donor PBL samples.
    • 2. Define PBL-depleted windows based on the criteria of a median absolute methylation value <0.1.
    • 3. Restrict analysis of cfDNA samples within PBL-depleted windows.


Performance of the in-silico PBL depletion strategy was evaluated by comparing absolute methylation distributions in PBL samples before and after depletion from the healthy donor cohort used as the training set, to the HNSCC cohort used as the validation set.


Differential Methylation Analysis


To enable robust detection of HNSCC-associated differentially methylated regions (DMRs), analysis was limited to HNSCC patients with detectable SNVs in plasma by CAPP-seq (n=20/32). Differential methylation analysis was limited to informative regions after in-silico PBL depletion. A collated matrix of binned fragment counts from HNSCC and healthy donor cfDNA samples, generated as previously described in Selection of Informative Regions in cfMeDIP-seq Profiles, were utilized for identification of DMRs by the DESeq2 R package. Pre-filtering was performed by removal of regions with <10 counts across all cfDNA samples. A single factor defined as condition (HNSCC vs. healthy donor) was used for contrast during differential methylation analysis. Briefly, differential methylation analysis was performed by scaling samples based on size factors and dispersion estimates, followed by fitting of a negative binomial general linear model. For each window, a P-value was calculated between the HNSCC and healthy donor conditions by Wald Test. P-values within regions above the default Cook's distance cut-off were omitted from adjusted P-value calculation (Benjamini-Hochberg). Significant hypermethylated or hypomethylated regions (hyper-/hypo-DMRs) in HNSCC cfDNA samples are defined as windows with an adjusted P-value <0.1.


Enrichment of CpG Features within HNSCC cfDNA Hypermethylated Regions


CpG features such as islands, shores, shelves, and open sea (interCGI) are defined as per the AnnotationHub R package (reference) (hg19_cpgs annotation). ID coordinates of each hypermethylated window (i.e. “chr.start.end”) within PBL-depleted regions were labeled with an overlapping CpG feature using an inhouse R package that utilizes the “annotatr” and “GenomicRanges” R packages (FIG. 13).


To determine the probability of enrichment for an observed overlap of features versus a null distribution, 1000 random samplings was performed. For each sampling, an equal number of bins were chosen based on the number hypermethylated windows, while maintaining an identical distribution of CpGs. The observed number of overlaps for each CpG feature across samplings were used to generate their respective null distributions, which were subsequently transformed onto a z-score scale. The observed overlap of hypermethylated regions for each CpG feature were also z-scored transformed, deriving summary statistics from the null distribution. The estimated P-value of the observed overlap from hypermethylated windows was calculated as the number of random samplings with overlap equal or greater/lesser than the observed overlap of the null distribution.


Enrichment of HNSCC cfDNA Hypermethylated Regions with Cancer-Specific Hypermethylated Cytosines from the Tumor Cancer Genome Atlas (TCGA)


File information from publicly available hm450k profiles of all primary tumors from breast (BRCA), colorectal (COAD), head and neck (HNSC), prostate (PRAD), pancreatic (PAAD), lung adeno (LUAD), and lung squamous (LUSC) were downloaded from the TCGA. Due to the majority of our HNSCC cohort presenting with tumors of the oral cavity, files from the HNSC group were limited to patients with primary site at the “floor of mouth” (n=55). An equal number of hm450k files were randomly selected from each of the remaining cancer types, as well as from a separate database of healthy PBLs (GEO series GSE67393). A manifest of downloaded files is provided in the (FIG. 14).


To generate “tumor-specific” hyper-methylated cytosines, differential methylation analysis by limma was performed for each cancer type, with individual comparisons to each other cancer type as well as PBLs (i.e. contrast). For a given contrast, a linear model is fitted for each probed cytosine incorporating the residual variance and sample beta value, the P-value of observed difference between contrasts is then calculated by the empirical Bayes smoothing. Hypermethylated cytosines with elevated methylation in a given cancer type versus an individual comparison was defined by a log foldchange >=0.25 and an adjusted P-value (Benjamini-Hochberg)<0.01. Hypermethylated cytosines unique to an individual cancer type were designated as “tumor-specific”. For the cases of LUSC, LUAD, and PAAD, either no or very little tumor-specific hypermethylated cytosines were identified (0, 15, 18) and therefore were omitted from subsequent analysis. For comparison with cfMeDIP-seq libraries, base-pair positions from tumor-specific hypermethylated cytosines were overlapped with informative windows after in-silico PBL depletion as described in Design of In-silico PBL Depletion and Evaluation of Performance.


The enrichment of overlap for HNSCC cfDNA hypermethylated regions with tumor-specific regions from TCGA was evaluated by 10,000 random samplings using the same methods described in Enrichment of CpG Features with HNSCC cfDNA Hypermethylated Regions.


Sensitivity and Specificity of ctDNA Detection by cfMeDIP-Seq


For cfMeDIP-seq libraries from our cohort of 32 HNSCC and 20 healthy donor cfDNA samples, ctDNA detection was defined based on the observation of a mean RPKM value across HNSCC cfDNA hypermethylated regions within an individual HNSCC cfDNA sample greater than the max mean RPKM value across healthy donor cfDNA samples. The sensitivity and specificity of ctDNA detection based on this definition was evaluated by Receiver Operating Characteristic (ROC) curve analysis. To minimize any confounding results due to the potential lack of ctDNA release in a subset of patients, ROC curve analysis was also performed in only 20 of the 32 HNSCC cfDNA samples with detectable ctDNA by CAPP-seq. Cross validation to assess the accuracy of ctDNA detection by DMR analysis was performed. Briefly, CAPP-Seq positive patients and healthy donors were randomly assigned to training (60%, n=24) and validation sets (40%, n=16) while maintaining similar ctDNA abundance (as determined by CAPP-Seq) between both sets. Hyper-DMRs were identified by differential methylation analysis between HNSCC and healthy donor samples within the training set. The sensitivity of ctDNA detection within these hyper-DMRs were assessed as previously described (FIG. 2C) within the validation set to obtain an AUROC value. A total of 50 random samplings were performed.


Fragment Length Analysis of ctDNA Detected by CAPP-Seq and cfMeDIP-Seq


For each HNSCC cfDNA CAPP-seq library, the median fragment length from all supporting paired reads of a specified SNV (i.e. singletons, SCSs, DCSs) as well as for paired reads containing the reference allele was measured. In cases where the median fragment length was reported for patients with >1 SNV, the median value across the median fragment length from each SNV was calculated. For each HNSCC cfDNA cfMeDIP-seq library, the median fragment length from all fragments mapping to the previously determined HNSCC cfDNA hypermethylated regions was calculated. Due to the relative absence of methylation within our cohort of 20 healthy donors, the fragment length of each healthy donor cfMeDIP-seq library was collated prior to any calculations. In both types of libraries, fragment length analysis was limited to cfDNA within the 1st peak (i.e. <220 base-pairs).


Enrichment of fragments (100-150 bp or 100-220 bp) within hyper-DMRs was calculated as followed. A null distribution of expected counts was generated from random 300-bp bins within our previously designed PBL-depleted windows at identical number and CpG density distribution, from a total of 30 samplings. Observed counts for each sample were determined based on read counts across hyper-DMRs. For each sample, enrichment was calculated based on the mean observed count divided by the mean expected count.


Supervised Hierarchal Clustering


Prior to clustering, a pseudocount of 0.1 was added to all RPKM values of cfMeDIP-seq libraries to enable log 2 transformation. Values were scaled by Euclidean transformation and clustered by Ward's method. An arbitrary number of three distinct clusters were selected (k=3), designated as methylation clusters 1-3, and used in subsequent analysis.


Metrics of ctDNA Detection and Quantification on HNSCC Patient Clinical Outcomes


The potential clinical utility of ctDNA detection was evaluated by three metrics: 1) detection of SNVs by CAPP-seq, 2) detection of increased mean RPKM in hypermethylated regions by cfMeDIP-seq. For comparative analysis, patients were stratified based on the following criteria: 1) presence or absence of SNVs, 2) methylation cluster 1 vs. methylation cluster 2+3. Patient characteristics are described in Table 1.























sampleID
pathology
smoking_status
smoking_patext missing or illegible when filed yeartext missing or illegible when filed
age_text missing or illegible when filed
gender
dx_site
subsite
t_stage
n_stage





1
HNSCC
Current
37
76
Male
Lip & Oral
Tongue
T1
N0








Cavity


2
HNSCC
Ex-smoker
20
81
Male
Paranasal
Maxillary
T3
N0








sinus
Sinus


3
HNSCC
Current
15
54
Femaltext missing or illegible when filed
Lip & Oral
Tongue
T2
N2b








Cavity


4
HNSCC
Ex-smoker
20
63
Male
Lip & Oral
Retromolar
T4a
N2b








Cavity
Trigone


5
HNSCC
Current
30
47
Male
Lip & Oral
Tongue
T4a
N2b








Cavity


6
HNSCC
Current
 2
22
Male
Lip & Oral
Tongue
T2
N1








Cavity


7
HNSCC
Ex-smoker
40
69
Male
Lip & Oral
Floor of
T4a
N2c








Cavity
Mouth


8
HNSCC
Ex-smoker
10
80
Male
Lip & Oral
Lower
T4a
N2b








Cavity
Alveolus &


9
HNSCC
Current
50
62
Male
Hypopharynx
Post-cricoid
T4a
N2c


10
HNSCC
Current
50
63
Femaltext missing or illegible when filed
Lip & Oral
Floor of
T3
N2c








Cavity
Mouth


11
HNSCC
Non-smoker
NA
68
Male
Lip & Oral
Lower
T4a
N2b








Cavity
Alveolus &


12
HNSCC
Ex-smoker
15
78
Male
Lip & Oral
Floor of
T1
N1








Cavity
Mouth


13
HNSCC
Non-smoker
NA
53
Male
Lip & Oral
Tongue
T2
N2b








Cavity


14
HNSCC
Ex-smoker
 5
59
Femaltext missing or illegible when filed
Lip & Oral
Floor of
T1
N0








Cavity
Mouth


15
HNSCC
Ex-smoker
25
79
Male
Larynx
Supraglottis
T4a
N1


16
HNSCC
Current
55
74
Male
Lip & Oral
Floor of
T2
N0








Cavity
Mouth


17
HNSCC
Current
40
64
Male
Lip & Oral
Lower
T4a
N1








Cavity
Alveolus &


18
HNSCC
Current
35
65
Femaltext missing or illegible when filed
Lip & Oral
Floor of
T4a
N0








Cavity
Mouth


19
HNSCC
Current
30
52
Femaltext missing or illegible when filed
Lip & Oral
Tongue
T2
N2c








Cavity


20
HNSCC
Current
30
46
Male
Larynx
Glottis
T4a
N2c


21
HNSCC
Non-smoker
NA
46
Male
Larynx
Supraglottis
T4a
N2b


22
HNSCC
Current
75
74
Male
Hypopharynx
Pyriform
T4a
N2c


23
HNSCC
Current
35
56
Male
Lip & Oral
Tongue
T1
N2c








Cavity


24
HNSCC
Non-smoker
NA
33
Male
Lip & Oral
Tongue
T1
N0








Cavity


25
HNSCC
Non-smoker
NA
60
Femaltext missing or illegible when filed
Lip & Oral
Tongue
T1
N1








Cavity


26
HNSCC
Current
40
65
Male
Hypopharynx
Pyriform
T3
N0


27
HNSCC
Current
15
49
Male
Lip & Oral
Floor of
T4a
N0








Cavity
Mouth


28
HNSCC
Current
30
54
Male
Lip & Oral
Floor of
T4a
N2c








Cavity
Mouth


29
HNSCC
Non-smoker
NA
54
Male
Lip & Oral
Tongue
T1
N0








Cavity


30
HNSCC
Current
30
56
Male
Lip & Oral
Floor of
T2
N0








Cavity
Mouth


1
Healthy donotext missing or illegible when filed
Ex-smoker
35
69
Male
NA
NA
NA
NA


2
Healthy donotext missing or illegible when filed
Current
84
74
Male
NA
NA
NA
NA


3
Healthy donotext missing or illegible when filed
Ex-smoker
40
77
Male
NA
NA
NA
NA


4
Healthy donotext missing or illegible when filed
Non-smoker
NA
82
Male
NA
NA
NA
NA


5
Healthy donotext missing or illegible when filed
Ex-smoker
14
61
Male
NA
NA
NA
NA


6
Healthy donotext missing or illegible when filed
Current
55
71
Male
NA
NA
NA
NA


7
Healthy donotext missing or illegible when filed
Current
50
65
Male
NA
NA
NA
NA


8
Healthy donotext missing or illegible when filed
Ex-smoker
30
69
Male
NA
NA
NA
NA


9
Healthy donotext missing or illegible when filed
Ex-smoker
41
57
Femaltext missing or illegible when filed
NA
NA
NA
NA


10
Healthy donotext missing or illegible when filed
Current
10
81
Male
NA
NA
NA
NA


11
Healthy donotext missing or illegible when filed
Current
39
64
Femaltext missing or illegible when filed
NA
NA
NA
NA


12
Healthy donotext missing or illegible when filed
Ex-smoker
30
65
Femaltext missing or illegible when filed
NA
NA
NA
NA


13
Healthy donotext missing or illegible when filed
Current
  17.5
64
Male
NA
NA
NA
NA


14
Healthy donotext missing or illegible when filed
Ex-smoker
50
77
Male
NA
NA
NA
NA


15
Healthy donotext missing or illegible when filed
Ex-smoker
10
59
Femaltext missing or illegible when filed
NA
NA
NA
NA


16
Healthy donotext missing or illegible when filed
Non-smoker
NA
64
Male
NA
NA
NA
NA


17
Healthy donotext missing or illegible when filed
Ex-smoker
20
66
Male
NA
NA
NA
NA


18
Healthy donotext missing or illegible when filed
Ex-smoker
  24.75
60
Femaltext missing or illegible when filed
NA
NA
NA
NA


19
Healthy donotext missing or illegible when filed
Ex-smoker
15
56
Male
NA
NA
NA
NA


20
Healthy donotext missing or illegible when filed
Non-smoker
NA
83
Male
NA
NA
NA
NA




















sampleID
m_stage
clinical_stage
hpv_status
chemotherapy
treatment
vital_status
cause_of_death
relapse







1
M0
I
NA
No
Sx only
Alive
NA
No



2
M0
III
Negative
No
post-op
Dead
Cancer
Yes



3
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



4
M0
IVA
NA
No
post-op
Alive
NA
No



5
M0
IVA
Negative
No
post-op
Dead
Cancer
Yes



6
M0
III
NA
Yes
post-op Ctext missing or illegible when filed
Dead
Cancer
Yes



7
M0
IVA
NA
No
post-op
Alive
NA
No



8
M0
IVA
NA
No
post-op
Dead
Cancer
Yes



9
M0
IVA
Negative
No
post-op
Dead
Cancer
Yes



10
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Dead
Index Cancer
Yes



11
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Dead
Cancer
Yes



12
M0
III
NA
No
Sx only
Alive
NA
No



13
M0
IVA
Negative
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



14
M0
I
NA
No
Sx only
Alive
NA
No



15
M0
IVA
Negative
No
post-op
Alive
NA
No



16
M0
II
NA
No
Sx only
Dead
Unknown
No



17
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



18
M0
IVA
NA
No
post-op
Alive
NA
No



19
M0
IVA
Negative
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



20
M0
IVA
Negative
Yes
post-op Ctext missing or illegible when filed
Alive
NA
Yes



21
M0
IVA
Negative
No
post-op
Alive
NA
No



22
M0
IVA
NA
No
post-op
Dead
Cancer
Yes



23
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



24
M0
I
NA
No
Sx only
Alive
NA
No



25
M0
III
NA
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



26
M0
III
NA
No
post-op
Dead
Unknown
No



27
M0
IVA
NA
No
post-op
Alive
NA
No



28
M0
IVA
NA
Yes
post-op Ctext missing or illegible when filed
Alive
NA
No



29
M0
I
Negative
No
post-op
Alive
NA
No



30
M0
II
NA
No
post-op
Alive
NA
No



1
NA
NA
NA
NA
NA
NA
NA
NA



2
NA
NA
NA
NA
NA
NA
NA
NA



3
NA
NA
NA
NA
NA
NA
NA
NA



4
NA
NA
NA
NA
NA
NA
NA
NA



5
NA
NA
NA
NA
NA
NA
NA
NA



6
NA
NA
NA
NA
NA
NA
NA
NA



7
NA
NA
NA
NA
NA
NA
NA
NA



8
NA
NA
NA
NA
NA
NA
NA
NA



9
NA
NA
NA
NA
NA
NA
NA
NA



10
NA
NA
NA
NA
NA
NA
NA
NA



11
NA
NA
NA
NA
NA
NA
NA
NA



12
NA
NA
NA
NA
NA
NA
NA
NA



13
NA
NA
NA
NA
NA
NA
NA
NA



14
NA
NA
NA
NA
NA
NA
NA
NA



15
NA
NA
NA
NA
NA
NA
NA
NA



16
NA
NA
NA
NA
NA
NA
NA
NA



17
NA
NA
NA
NA
NA
NA
NA
NA



18
NA
NA
NA
NA
NA
NA
NA
NA



19
NA
NA
NA
NA
NA
NA
NA
NA



20
NA
NA
NA
NA
NA
NA
NA
NA








text missing or illegible when filed indicates data missing or illegible when filed







Cross-Validation of ctDNA-Derived Methylation by cfMeDIP-Seq Analysis


To evaluate the robustness of cfMeDIP-seq for identifying ctDNA-derived methylation, Receiver Operating Characteristics (ROC) curve analysis was performed. To minimize confounding results due to low/absent ctDNA, analysis was limited to HNSCC patients with detectable ctDNA by CAPP-seq. Patient and healthy control cfMeDIP-seq profiles were split into a training set (HNSCC: n=12/20; healthy control: n=12/20) and testing set (HNSCC: n=8/20; healthy control: n=8/20). Training and testing sets were balanced for ctDNA abundance as determined by CAPP-Seq analysis. A total of 50 splits were performed with ROC curve analysis performed on each iteration.


Identification of Prognostic Regions in HNSCC by TCGA Analysis


All available HNSCC cases from TCGA with matched legacy hm450k and RNA expression data were selected (n=520). Survival data was obtained from Jianfang et al. With regards to the hm450k data, methylation was summarized to 300-bp regions as described previously by calculating the mean beta-value between probe IDs within a particular region. To identify regions hypermethylated in HNSCC primary tumors compared to adjacent normal tissue, independent Wilcoxon tests were performed for each region. Regions with an adjusted p-value <0.05 (Holms method) as well as a log-fold change >=1 in primary tumors compared to adjacent normal tissue, were selected for subsequent analysis. To identify hypermethylated regions associated with prognosis, multivariate Cox Regression was performed, considering age, gender, and clinical stage, selecting regions with p-values <0.05. Survival analysis was limited to a maximum follow-up time of 5 years post-diagnosis, reflecting what was observed within the HNSCC cfDNA cohort. To further identify prognostic regions associated with changes in gene expression, Spearman's correlation was calculated for hm450k primary tumor profiles for each region, to matched RNA expression profiles for transcripts within a 2-Kb window. Regions with absolute Rho values >0.3 and a false discovery rate <0.05 were selected, resulting in the final identification of 5 prognostic regions associated with ZNF323/ZSCAN31, LINC01395, GATA2-AS1, OSR1, and STK3/MST2 expression. For TCGA patient profiles, the Composite Methylation Score (CMS) was obtained by calculating the sum of beta-values across all 5 prognostic regions. For cfMeDIP-seq profiles, RPKM values across all 943 hyper-DMRs were scaled to a total sum of 1 and the CMS was obtained by calculating the sum of these scaled RPKM values across all 5 prognostic regions.


Longitudinal Monitoring of Post-Treatment Plasma Samples by cfMeDIP-Seq


cfMeDIP-seq libraries were successfully generated for 30/32 patients (FIGS. 17A-17D). For the remaining two patients, insufficient material was isolated from plasma and/or did not pass quality metrics. ctDNA quantification of post-treatment cfMeDIP-seq libraries was performed as previously described, calculating the mean RPKM values across identified hypermethylated regions by differential methylation analysis. For ease on interpretation, both pre-treatment and post-treatment cfMeDIP-seq libraries were converted to percent DNA values based on linear regression against mean MAF calculated by matched CAPP-Seq profiles. To achieve high confidence detection of residual disease, a minimum ctDNA fraction of 0.2% was required in post-treatment samples, corresponding to the maximum of mean RPKM values observed across all healthy controls.


Results & Discussion


Multimodal Profiling of Cell-Free DNA in Localized HNSCC


To examine the ability of multimodal profiling to characterize ctDNA in the setting of localized cancer, we recruited 32 HNSCC patients into a prospective observational study in which peripheral blood samples were collected at serial timepoints (FIG. 9A; Table 1). All patients were treated with surgery, with a subset receiving adjuvant radiotherapy (n=14) or chemoradiotherapy (n=11). With a median follow up of 43.2 months, 9/32 patients (28%) developed recurrence (actuarial 2-year recurrence-free survival: 88%).


As the majority of patients exhibited a heavy smoking history, which is well-described to alter the genomic/epigenomic landscape of somatic tissue and contribute to premalignant lesions, we also analyzed blood samples from 20 risk-matched healthy donors previously enrolled in a lung cancer screening program34-37. Cell-free DNA from plasma as well as genomic DNA (gDNA) from PBLs were co-isolated from blood and subjected to quantification and analysis (Supplementary FIG. 1A). In contrast to other studies that have demonstrated significantly elevated levels of total plasma cell-free DNA in metastatic disease compared to healthy controls38-41, no significant difference was observed between our HNSCC cohort and healthy donors (Supplementary FIG. 1B).


Multimodal profiling of cell-free DNA and PBL gDNA from patients and healthy controls were conducted (FIG. 1). By subjecting the same samples to both mutation and methylome profiling, we were able to evaluate their contributions to tumor-naïve detection and characterization of ctDNA. Mutations and methylation were independently profiled using CAncer Personalized Profiling by deep Sequencing (CAPP-Seq) and cell-free Methylated DNA ImmunoPrecipitation and high-throughput sequencing (cfMeDIP-seq), respectively. In addition, paired-end sequencing was utilized for both methodologies in order to obtain the lengths of sequenced cell-free DNA fragments.


Tumor-Naïve Detection of Mutation-Based ctDNA from Pre-Treatment Plasma


We first evaluated approaches to improve our confidence of mutation-based ctDNA detection without confirmation within matched tumor samples. Recent studies have illustrated that genes frequently targeted for ctDNA detection, such as TP53, can harbor mutations derived from clonally expanded PBLs. Additionally, as ctDNA contains both genetic and epigenetic features of the tumor, we reasoned that orthogonal analysis of both features in patient cell-free DNA may provide increased confidence of ctDNA detection. Therefore, to achieve tumor-naïve detection of low-abundance ctDNA with high confidence, mutations and methylation were independently profiled by CAPP-Seq and cfMeDIP-seq, respectively, for both cfDNA and matched PBLs.


To evaluate the sensitivity of ctDNA detection in HPV-negative THNSCC without prior knowledge from the tumor, we first measured the abundance of mutations in baseline plasma samples (FIG. 2A). CAPP-Seq was conducted with a sequencing panel designed to maximize the number of HNSCC-associated mutations (Table 3 and FIG. 10). We also employed established error suppression methodologies to remove background base substitution errors.









TABLE 3







Targeted genomic regions of HNSCC CAPP-Seq selector














Chr
Start
End
Length
Exon
Strand
Gene_canonical
Transcript

















chr1
16464346
16464592
247
5

EPHA2
NM_004431.4


chr1
16475056
16475543
488
3

EPHA2
NM_004431.4


chr1
22838527
22838627
101
11
+
ZBTB40
NM_014870.3


chr1
38227190
38227732
543
3

EPHA10
NM_001099439.1


chr1
57257787
57258322
536
2

C1orf168
NM_001004303.4


chr1
57480678
57480885
208
14

DAB1
NM_021080.3


chr1
74575077
74575237
161
5

LRRIQ3
NM_001105659.1


chr1
89616097
89616258
162
6

GBP7
NM_207398.2


chr1
97547897
97547997
101
22

DPYD
NM_000110.3


chr1
97770870
97770970
101
18

DPYD
NM_000110.3


chr1
98165041
98165141
101
6

DPYD
NM_000110.3


chr1
99771257
99772532
1276
7
+
LPPR4
NM_014839.4


chr1
103471370
103471470
101
18

COL11A1
NM_001854.3


chr1
103480058
103480158
101
13

COL11A1
NM_001854.3


chr1
115251151
115251275
125
5

NRAS
NM_002524.4


chr1
115252189
115252349
161
4

NRAS
NM_002524.4


chr1
115256420
115256599
180
3

NRAS
NM_002524.4


chr1
115258670
115258798
129
2

NRAS
NM_002524.4


chr1
119575732
119575894
163
6

WARS2
NM_015836.3


chr1
149859079
149859463
385
1

HIST2H2AB
NM_175065.2


chr1
149905298
149905423
126
10

MTMR11
NM_001145862.1


chr1
154898829
154898959
131
4

PMVK
NM_006556.3


chr1
157555965
157556231
267
6

FCRL4
NM_031282.2


chr1
157557066
157557332
267
5

FCRL4
NM_031282.2


chr1
158152697
158152797
101
5
+
CD1D
NM_001766.3


chr1
158326517
158326686
170
6
+
CD1E
NM_030893.3


chr1
158817522
158817653
132
6
+
MNDA
NM_002432.1


chr1
161479695
161479795
101
4
+
FCGR2A
NM_001136219.1


chr1
161514492
161514592
101
4

FCGR3A
NM_000569.7


chr1
169390627
169391473
847
3

CCDC181
NM_021179.2


chr1
169578748
169578892
145
8

SELP
NM_003005.3


chr1
190067262
190068200
939
8

FAM5C
NM_199051.2


chr1
196309568
196309668
101
16

KCNT2
NM_198503.3


chr1
197390164
197391037
874
6
+
CRB1
NM_201253.2


chr1
198713182
198713332
151
26
+
PTPRC
NM_002838.4


chr1
205038983
205039124
142
18
+
CNTN2
NM_005076.4


chr1
205779302
205779509
208
2

SLC41A1
NM_173854.5


chr1
207010005
207010151
147
2
+
IL19
NM_153758.2


chr1
215847663
215848873
1211
63

USH2A
NM_206933.2


chr1
216850475
216850747
273
2

ESRRG
NM_001438.3


chr1
248028042
248028156
115
3
+
TRIM58
NM_015431.3


chr2
1459847
1460008
162
7
+
TPO
NM_000547.5


chr2
15415719
15415924
206
44

NBAS
NM_015909.3


chr2
27121416
27121557
142
2
+
DPYSL5
NM_020134.3


chr2
28634819
28634952
134
4
+
FOSL2
NM_005253.3


chr2
51254660
51255257
598
2

NRXN1
NM_004801.5


chr2
65540892
65541048
157
6

SPRED2
NM_181784.2


chr2
75720459
75720730
272
4

EVA1A
NM_001135032.1


chr2
79349113
79349251
139
4
+
REG1A
NM_002909.4


chr2
80136776
80136880
105
7
+
CTNNA2
NM_004389.3


chr2
80529667
80530938
1272

+
CTNNA2
NM_004389.3


chr2
90249111
90249392
282

+
abParts


chr2
99012565
99013683
1119
8
+
CNGA3
NM_001298.2


chr2
106497875
106498430
556
3
+
NCK2
NM_001004720.2


chr2
133402677
133403123
447
2
+
GPR39
NM_001508.2


chr2
138169278
138169409
132
14
+
THSD7B
NM_001316349.1


chr2
141081496
141081596
101
81

LRP1B
NM_018557.2


chr2
141359045
141359175
131
42

LRP1B
NM_018557.2


chr2
141806555
141806673
119
11

LRP1B
NM_018557.2


chr2
160206240
160206346
107
28

BAZ2B
NM_013450.3


chr2
164466119
164468153
2035
3

FIGN
NM_018086.3


chr2
166003407
166003510
104
12

SCN3A
NM_006922.3


chr2
176958040
176958203
164
1
+
HOXD13
NM_000523.3


chr2
178095512
178096736
1225
5

NFE2L2
NM_006164.4


chr2
178097119
178097311
193
4

NFE2L2
NM_006164.4


chr2
178097972
178098072
101
3

NFE2L2
NM_006164.4


chr2
178098732
178098999
268
2

NFE2L2
NM_006164.4


chr2
178129232
178129332
101
1

NFE2L2
NM_006164.4


chr2
198948735
198950751
2017
2
+
PLCL1
NM_006226.3


chr2
202122944
202123115
172
1
+
CASP8
NM_001080125.1


chr2
202131173
202131524
352
2
+
CASP8
NM_001080125.1


chr2
202134222
202134338
117

+
CASP8
NM_001080125.1


chr2
202136214
202136354
141
3
+
CASP8
NM_001080125.1


chr2
202137349
202137509
161
4
+
CASP8
NM_001080125.1


chr2
202137593
202137693
101
5
+
CASP8
NM_001080125.1


chr2
202139594
202139694
101
6
+
CASP8
NM_001080125.1


chr2
202140209
202140309
101

+
CASP8
NM_001080125.1


chr2
202141539
202141702
164
7
+
CASP8
NM_001080125.1


chr2
202142434
202142538
105

+
CASP8
NM_001080125.1


chr2
202149528
202150050
523
8
+
CASP8
NM_001080125.1


chr2
202151171
202151327
157
9
+
CASP8
NM_001080125.1


chr2
202152123
202152223
101
9
+
CASP8
NM_001080125.1


chr2
213878514
213878658
145
7

IKZF2
NM_016260.2


chr2
217124225
217124379
155
4

MARCH4
NM_020814.2


chr2
225338951
225339103
153
16

CUL3
NM_003590.4


chr2
225342906
225343072
167
15

CUL3
NM_003590.4


chr2
225346598
225346805
208
14

CUL3
NM_003590.4


chr2
225360538
225360693
156
13

CUL3
NM_003590.4


chr2
225362459
225362576
118
12

CUL3
NM_003590.4


chr2
225365069
225365214
146
11

CUL3
NM_003590.4


chr2
225367671
225367799
129
10

CUL3
NM_003590.4


chr2
225368358
225368549
192
9

CUL3
NM_003590.4


chr2
225370662
225370859
198
8

CUL3
NM_003590.4


chr2
225371564
225371730
167
7

CUL3
NM_003590.4


chr2
225376060
225376309
250
6

CUL3
NM_003590.4


chr2
225378230
225378365
136
5

CUL3
NM_003590.4


chr2
225379318
225379499
182
4

CUL3
NM_003590.4


chr2
225400234
225400368
135
3

CUL3
NM_003590.4


chr2
225422365
225422583
219
2

CUL3
NM_003590.4


chr2
225427900
225428047
148


CUL3
NM_003590.4


chr2
225431632
225431732
101


CUL3
NM_003590.4


chr2
225434395
225434499
105


CUL3
NM_003590.4


chr2
225449643
225449743
101
1

CUL3
NM_003590.4


chr2
227924129
227924320
192
28

COL4A4
NM_000092.4


chr3
6903212
6903557
346
1
+
GRM7
NM_181874.2


chr3
12046075
12046175
101
1
+
SYN2
NM_133625.4


chr3
30691783
30691948
166
4
+
TGFBR2
NM_001024847.2


chr3
30732941
30733073
133
8
+
TGFBR2
NM_001024847.2


chr3
49412866
49413022
157
2

RHOA
NM_001664.3


chr3
59737949
59738049
101
9

FHIT
NM_001166243.2


chr3
59908056
59908156
101
8

FHIT
NM_001166243.2


chr3
59997061
59997161
101
7

FHIT
NM_001166243.2


chr3
59999732
59999878
147
6

FHIT
NM_001166243.2


chr3
60522592
60522695
104
5

FHIT
NM_001166243.2


chr3
109028016
109028177
162
4

DPPA2
NM_138815.3


chr3
109049478
109049642
165
5

DPPA4
NM_018189.3


chr3
115395072
115395172
101
3
+
GAP43
NM_001130064.1


chr3
124418715
124418880
166
56
+
KALRN
NM_001024660.4


chr3
126707543
126708597
1055
1
+
PLXNA1
NM_032242.3


chr3
129370517
129370617
101
6

TMCC1
NM_001017395.3


chr3
132435600
132435753
154
4

NPHP3
NM_153240.4


chr3
140167411
140167511
101
6
+
CLSTN2
NM_022131.2


chr3
145912910
145913069
160
8

PLSCR4
NM_020353.2


chr3
147108778
147109008
231
4

ZIC4
NM_001168378.1


chr3
147113699
147114241
543
3

ZIC4
NM_001168378.1


chr3
147127953
147128847
895
1
+
ZIC1
NM_003412.3


chr3
148458870
148459830
961
3
+
AGTR1
NM_004835.4


chr3
150908512
150908706
195
13
+
MED12L
NM_053002.5


chr3
155198904
155200710
1807
23

PLCH1
NM_001130960.1


chr3
157146110
157146277
168
5

VEPH1
NM_001167912.1


chr3
164727066
164727183
118
35

SI
NM_001041.3


chr3
168838843
168839000
158
7

MECOM
NM_004991.3


chr3
169540059
169540508
450
1
+
LRRIQ4
NM_001080460.1


chr3
170819255
170819398
144
22

TNIK
NM_015028.3


chr3
178916613
178916965
353
2
+
PIK3CA
NM_006218.3


chr3
178917477
178917687
211
3
+
PIK3CA
NM_006218.3


chr3
178919077
178919328
252
4
+
PIK3CA
NM_006218.3


chr3
178921331
178921577
247
5
+
PIK3CA
NM_006218.3


chr3
178922283
178922383
101
6
+
PIK3CA
NM_006218.3


chr3
178927382
178927488
107
7
+
PIK3CA
NM_006218.3


chr3
178927973
178928126
154
8
+
PIK3CA
NM_006218.3


chr3
178928218
178928353
136
9
+
PIK3CA
NM_006218.3


chr3
178935997
178936122
126
10
+
PIK3CA
NM_006218.3


chr3
178936974
178937074
101
11
+
PIK3CA
NM_006218.3


chr3
178937358
178937523
166
12
+
PIK3CA
NM_006218.3


chr3
178937736
178937840
105
13
+
PIK3CA
NM_006218.3


chr3
178938773
178938945
173
14
+
PIK3CA
NM_006218.3


chr3
178941868
178941975
108
15
+
PIK3CA
NM_006218.3


chr3
178942487
178942609
123
16
+
PIK3CA
NM_006218.3


chr3
178943739
178943839
101
17
+
PIK3CA
NM_006218.3


chr3
178947059
178947230
172
18
+
PIK3CA
NM_006218.3


chr3
178947791
178947909
119
19
+
PIK3CA
NM_006218.3


chr3
178948012
178948164
153
20
+
PIK3CA
NM_006218.3


chr3
178951881
178952152
272
21
+
PIK3CA
NM_006218.3


chr3
181430148
181431102
955
1
+
SOX2
NM_003106.3


chr3
189349285
189349385
101
1
+
TP63
NM_003722.4


chr3
189455528
189455657
130
2
+
TP63
NM_003722.4


chr3
189456430
189456563
134
3
+
TP63
NM_003722.4


chr3
189526060
189526315
256
4
+
TP63
NM_003722.4


chr3
189582019
189582207
189
5
+
TP63
NM_003722.4


chr3
189584470
189584586
117
6
+
TP63
NM_003722.4


chr3
189585621
189585731
111
7
+
TP63
NM_003722.4


chr3
189586368
189586505
138
8
+
TP63
NM_003722.4


chr3
189587104
189587204
101
9
+
TP63
NM_003722.4


chr3
189590647
189590784
138
10
+
TP63
NM_003722.4


chr3
189604182
189604340
159
11
+
TP63
NM_003722.4


chr3
189607128
189607273
146
12
+
TP63
NM_003722.4


chr3
189608574
189608674
101
13
+
TP63
NM_003722.4


chr3
189611994
189612291
298
14
+
TP63
NM_003722.4


chr3
192516294
192517141
848
2

MB21D2
NM_178496.3


chr4
1808555
1810599
2045
17, 18
+
FGFR3
NM_001163213.1


chr4
3768690
3768968
279
1
+
ADRA2C
NM_000683.3


chr4
9783961
9785062
1102
1
+
DRD5
NM_000798.4


chr4
20619060
20619186
127
36
+
SLIT2
NM_004787.3


chr4
22749400
22749650
251
3
+
GBA3
NR_102355.1


chr4
41984000
41984281
282
1
+
DCAF4L1
NM_001029955.3


chr4
44176944
44177258
315
2

KCTD8
NM_198353.2


chr4
57896425
57896548
124
24
+
POLR2B
NM_000938.2


chr4
73012693
73013517
825
4
+
NPFFR2
NM_004885.2


chr4
94376876
94377125
250
11
+
GRID2
NM_001510.3


chr4
110914377
110914477
101
19
+
EGF
NM_001963.5


chr4
115891585
115891740
156
4

NDST4
NM_022569.2


chr4
153247160
153247380
221
10

FBXW7
NM_033632.3


chr4
153249384
153249527
144
9

FBXW7
NM_033632.3


chr4
162306901
162307560
660
16

FSTL5
NM_020116.4


chr4
175896742
175898747
2006
2
+
ADAM29
NM_001278127.1


chr4
187509745
187510374
630
27

FAT1
NM_005245.3


chr4
187516842
187516980
139
26

FAT1
NM_005245.3


chr4
187517693
187518325
633
25

FAT1
NM_005245.3


chr4
187518835
187518946
112
24

FAT1
NM_005245.3


chr4
187519125
187519279
155
23

FAT1
NM_005245.3


chr4
187521051
187521514
464
22

FAT1
NM_005245.3


chr4
187522422
187522580
159
21

FAT1
NM_005245.3


chr4
187524056
187524188
133
20

FAT1
NM_005245.3


chr4
187524329
187525131
803
19

FAT1
NM_005245.3


chr4
187525530
187525728
199
18

FAT1
NM_005245.3


chr4
187527223
187527367
145
17

FAT1
NM_005245.3


chr4
187530336
187530474
139
16

FAT1
NM_005245.3


chr4
187530954
187531169
216
15

FAT1
NM_005245.3


chr4
187532539
187532929
391
14

FAT1
NM_005245.3


chr4
187534262
187534496
235
13

FAT1
NM_005245.3


chr4
187535344
187535498
155
12

FAT1
NM_005245.3


chr4
187538158
187538355
198
11

FAT1
NM_005245.3


chr4
187538861
187542929
4069
10

FAT1
NM_005245.3


chr4
187549307
187549518
212
9

FAT1
NM_005245.3


chr4
187549641
187549917
277
8

FAT1
NM_005245.3


chr4
187554837
187554977
141
7

FAT1
NM_005245.3


chr4
187557178
187557389
212
6

FAT1
NM_005245.3


chr4
187557738
187558068
331
5

FAT1
NM_005245.3


chr4
187560856
187560956
101
4

FAT1
NM_005245.3


chr4
187584452
187584767
316
3

FAT1
NM_005245.3


chr4
187627716
187630981
3266
2

FAT1
NM_005245.3


chr5
1295105
1295250
146
1

TERT
NM_198253.2


chr5
11022897
11023091
195
17

CTNND2
NM_001332.3


chr5
11082807
11082958
152
16

CTNND2
NM_001332.3


chr5
11346483
11346705
223
9

CTNND2
NM_001332.3


chr5
11364834
11364947
114
8

CTNND2
NM_001332.3


chr5
11397156
11397256
101
6

CTNND2
NM_001332.3


chr5
13719059
13719207
149
72

DNAH5
NM_001369.2


chr5
15936690
15937234
545
4
+
FBXL7
NM_012304.4


chr5
23522412
23522514
103
7
+
PRDM9
NM_020227.3


chr5
26881404
26881715
312
12

CDH9
NM_016279.3


chr5
26885756
26885968
213
11

CDH9
NM_016279.3


chr5
26903761
26903931
171
6

CDH9
NM_016279.3


chr5
41159226
41159326
101
12

C6
NM_000065.3


chr5
45262057
45262790
734
8

HCN1
NM_021072.3


chr5
45353201
45353348
148
5

HCN1
NM_021072.3


chr5
63256300
63257482
1183
1

HTR1A
NM_000524.3


chr5
89943366
89943472
107
17
+
GPR98
NM_032119.3


chr5
90040933
90041033
101
51
+
GPR98
NM_032119.3


chr5
101834431
101834544
114
1

SLCO6A1
NM_173488.4


chr5
113698588
113698688
101
1
+
KCNN2
NM_021614.3


chr5
139192995
139193155
161
3
+
PSD2
NM_032289.2


chr5
140165874
140168242
2369
1
+
PCDHA1
NM_031410.2


chr5
140201500
140203558
2059

+
PCDHA1
NM_031411.2


chr5
140235927
140237983
2057

+
PCDHA1
NM_031411.2


chr5
140261864
140264052
2189

+
PCDHA12
NM_018903.3


chr5
142513531
142513670
140
19
+
ARHGAP26
NM_015071.4


chr5
153026493
153026601
109
3
+
GRIA1
NM_001258022.1


chr5
158139993
158140095
103
13

EBF1
NM_024007.4


chr5
161524717
161524860
144
4
+
GABRG2
NM_198903.2


chr5
176636699
176639007
2309
5
+
NSD1
NM_022455.4


chr5
176675181
176675325
145
11
+
NSD1
NM_022455.4


chr5
176687014
176687153
140
14
+
NSD1
NM_022455.4


chr5
176709465
176709582
118
19
+
NSD1
NM_022455.4


chr6
348126
348270
145
6
+
DUSP22
NM_020185.4


chr6
26027282
26027418
137
1

HIST1H4B
NM_003544.2


chr6
26031884
26032208
325
1

HIST1H3B
NM_003537.3


chr6
26045648
26046020
373
1
+
HIST1H3C
NM_003531.2


chr6
26056116
26056553
438
1

HIST1H1C
NM_005319.3


chr6
26204884
26205157
274
1
+
HIST1H4E
NM_003545.3


chr6
26216704
26216848
145
1

HIST1H2BG
NM_003518.3


chr6
26217209
26217595
387
1
+
HIST1H2AE
NM_021052.2


chr6
26225487
26225675
189
1
+
HIST1H3E
NM_003532.2


chr6
26271336
26271610
275
1

HIST1H3G
NM_003534.2


chr6
27777864
27778028
165
1
+
HIST1H3H
NM_003536.2


chr6
27805924
27806024
101
1

HIST1H2AK
NM_003510.2


chr6
27834626
27835200
575
1

HIST1H1B
NM_005322.2


chr6
27839691
27840063
373
1

HIST1H3I
NM_003533.2


chr6
28554274
28554444
171
1

SCAND3
NM_052923.1


chr6
29910586
29910798
213
2
+
HLA-A
NM_002116.7


chr6
29911084
29911184
101
3
+
HLA-A
NM_002116.7


chr6
31323135
31323361
227
4

HLA-B
NM_005514.7


chr6
31939824
31939958
135
1
+
STK19
NM_032454.1


chr6
32725549
32725660
112


HLA-DQB2
NM_001198858.1


chr6
33282954
33284459
1506
2

ZBTB22
NM_001145338.1


chr6
35773512
35773626
115
1
+
LHFPL5
NM_182548.3


chr6
46107689
46108044
356
2
+
ENPP4
NM_014936.4


chr6
52883079
52883179
101
7

ICK
NM_014920.3


chr6
55113473
55113573
101
3
+
HCRTR2
NM_001526.4


chr6
66204658
66205296
639
4

EYS
NM_001142800.1


chr6
87725250
87726088
839
2
+
HTR1E
NM_000865.2


chr6
96651055
96652047
993
3
+
FUT9
NM_006581.3


chr6
100382273
100382393
121
5

MCHR2
NM_032503.2


chr6
100838250
100838940
691
11

SIM1
NM_005068.2


chr6
105606462
105606600
139
4

POPDC3
NM_022361.4


chr6
112671162
112671571
410
3
+
RFPL4B
NM_001013734.2


chr6
116937940
116938344
405
1
+
RSPH4A
NM_001010892.2


chr6
119337959
119338094
136
5

FAM184A
NM_024581.5


chr6
119341141
119341266
126
4

FAM184A
NM_024581.5


chr6
127796883
127797498
616
6

SOGA3
NM_001012279.2


chr6
130761862
130762868
1007
2
+
TMEM200A
NM_001258276.1


chr6
134210543
134210975
433
1
+
TCF21
NM_003206.3


chr6
146480484
146480608
125
3
+
GRM1
NM_001278065.1


chr6
146720035
146720767
733
8
+
GRM1
NM_001278065.1


chr6
152655142
152655243
102
77

SYNE1
NM_182961.3


chr6
152763220
152763323
104
31

SYNE1
NM_182961.3


chr6
158449876
158450036
161
3
+
SYNJ2
NM_003898.3


chr6
165715111
165715618
508
2

C6orf118
NM_144980.3


chr7
1586590
1586690
101
9

TMEM184A
NM_001097620.1


chr7
8790657
8791330
674
3
+
NXPH1
NM_152745.2


chr7
11468613
11468713
101
14

THSD7A
NM_015204.2


chr7
11630120
11630220
101
4

THSD7A
NM_015204.2


chr7
11675890
11676535
646
2

THSD7A
NM_015204.2


chr7
21639523
21639717
195
15
+
DNAH11
NM_001277115.1


chr7
34118610
34118795
186
13
+
BMPER
NM_133468.4


chr7
34125372
34125526
155
14
+
BMPER
NM_133468.4


chr7
36895203
36895310
108
22

ELMO1
NM_014800.10


chr7
37955723
37956086
364
1

SFRP4
NM_003014.3


chr7
37988473
37988621
149
2
+
EPDR1
NM_017549.4


chr7
38398137
38398237
101


TRGV3


chr7
45614686
45614786
101
1
+
ADCY1
NM_021116.2


chr7
50611605
50611705
101
2

DDC
NM_000790.3


chr7
53103444
53104199
756
1
+
POM121L12
NM_182595.3


chr7
55086964
55087064
101
1
+
EGFR
NM_005228.4


chr7
55209978
55210130
153
2
+
EGFR
NM_005228.4


chr7
55210997
55211181
185
3
+
EGFR
NM_005228.4


chr7
55214298
55214433
136
4
+
EGFR
NM_005228.4


chr7
55218971
55219071
101
5
+
EGFR
NM_005228.4


chr7
55220238
55220357
120
6
+
EGFR
NM_005228.4


chr7
55221703
55221845
143
7
+
EGFR
NM_005228.4


chr7
55223522
55223639
118
8
+
EGFR
NM_005228.4


chr7
55224225
55224352
128
9
+
EGFR
NM_005228.4


chr7
55224438
55224538
101
10
+
EGFR
NM_005228.4


chr7
55225351
55225451
101
11
+
EGFR
NM_005228.4


chr7
55227831
55228031
201
12
+
EGFR
NM_005228.4


chr7
55229191
55229324
134
13
+
EGFR
NM_005228.4


chr7
55231421
55231521
101
14
+
EGFR
NM_005228.4


chr7
55232972
55233130
159
15
+
EGFR
NM_005228.4


chr7
55238837
55238937
101
16
+
EGFR
NM_005228.4


chr7
55240675
55240817
143
17
+
EGFR
NM_005228.4


chr7
55241613
55241736
124
18
+
EGFR
NM_005228.4


chr7
55242414
55242514
101
19
+
EGFR
NM_005228.4


chr7
55248985
55249171
187
20
+
EGFR
NM_005228.4


chr7
55259411
55259567
157
21
+
EGFR
NM_005228.4


chr7
55260446
55260546
101
22
+
EGFR
NM_005228.4


chr7
55266409
55266556
148
23
+
EGFR
NM_005228.4


chr7
55268007
55268107
101
24
+
EGFR
NM_005228.4


chr7
55268880
55269048
169
25
+
EGFR
NM_005228.4


chr7
55269401
55269501
101
26
+
EGFR
NM_005228.4


chr7
55270209
55270318
110
27
+
EGFR
NM_005228.4


chr7
55272948
55273310
363
28
+
EGFR
NM_005228.4


chr7
82581206
82586179
4974
5

PCLO
NM_033026.5


chr7
86394555
86394750
196
2
+
GRM3
NM_000840.2


chr7
86415580
86416335
756
3
+
GRM3
NM_000840.2


chr7
86493597
86493712
116
6
+
GRM3
NM_000840.2


chr7
95157167
95157582
416
3
+
ASB4
NM_016116.2


chr7
98256517
98256642
126
4
+
NPTX2
NM_002523.2


chr7
104377174
104377335
162
2
+
LHFPL3
NM_199000.2


chr7
106508168
106509929
1762
2
+
PIK3CG
NM_002649.3


chr7
111368415
111368577
163
52

DOCK4
NM_014705.3


chr7
117351679
117351779
101
23

CTTNBP2
NM_033427.2


chr7
119914693
119915730
1038
1
+
KCND2
NM_012281.2


chr7
121943706
121944308
603
1

FEZF1
NM_001024613.3


chr7
140453074
140453193
120
15

BRAF
NM_004333.4


chr7
142206549
142206746
198


TCRBV12S3


chr7
142498863
142499043
181

+
TCRVB


chr7
146829380
146829579
200
8
+
CNTNAP2
NM_014141.5


chr7
154561126
154561281
156
9
+
DPP6
NM_130797.3


chr7
154862704
154863349
646
1
+
HTR5A
NM_024012.3


chr8
2820009
2820155
147
61

CSMD1
NM_033225.5


chr8
4494929
4495029
101
2

CSMD1
NM_033225.5


chr8
38271145
38271322
178
19

FGFR1
NM_001174067.1


chr8
38271435
38271541
107
18

FGFR1
NM_001174067.1


chr8
38271669
38271807
139
17

FGFR1
NM_001174067.1


chr8
38272062
38272162
101
16

FGFR1
NM_001174067.1


chr8
38272296
38272419
124
15

FGFR1
NM_001174067.1


chr8
38273387
38273578
192
14

FGFR1
NM_001174067.1


chr8
38274823
38274934
112
13

FGFR1
NM_001174067.1


chr8
38275387
38275509
123
12

FGFR1
NM_001174067.1


chr8
38275745
38275891
147
11

FGFR1
NM_001174067.1


chr8
38277050
38277253
204
10

FGFR1
NM_001174067.1


chr8
38279314
38279459
146
9

FGFR1
NM_001174067.1


chr8
38282026
38282217
192
8

FGFR1
NM_001174067.1


chr8
38283639
38283763
125
7

FGFR1
NM_001174067.1


chr8
38285438
38285611
174
6

FGFR1
NM_001174067.1


chr8
38285861
38285961
101
5

FGFR1
NM_001174067.1


chr8
38287199
38287466
268
4

FGFR1
NM_001174067.1


chr8
38314873
38315052
180
3

FGFR1
NM_001174067.1


chr8
41166589
41166689
101
1

SFRP1
NM_003012.4


chr8
55533681
55534104
424
2
+
RP1
NM_006269.1


chr8
56015331
56015802
472
1
+
XKR4
NM_052898.1


chr8
73479973
73480509
537
2
+
KCNB2
NM_004770.2


chr8
73848397
73850116
1720
3
+
KCNB2
NM_004770.2


chr8
74922235
74922364
130
3
+
LY96
NM_015364.4


chr8
88885017
88886181
1165
1

DCAF4L2
NM_152418.3


chr8
92972524
92972728
205
11

RUNX1T1
NM_001198634.1


chr8
98289087
98290056
970
1

TSPYL5
NM_033512.2


chr8
107782022
107782413
392
1

ABRA
NM_139166.4


chr8
110980373
110980810
438
4

KCNV1
NM_014379.3


chr8
113256622
113256777
156
65

CSMD3
NM_198123.1


chr8
113259248
113259360
113
64

CSMD3
NM_198123.1


chr8
113347557
113347703
147
45

CSMD3
NM_198123.1


chr8
113585729
113585886
158
24

CSMD3
NM_198123.1


chr8
113988121
113988225
105
7

CSMD3
NM_198123.1


chr8
128748804
128748904
101
1
+
MYC
NM_002467.4


chr8
128750493
128751265
773
2
+
MYC
NM_002467.4


chr8
128752641
128753204
564
3
+
MYC
NM_002467.4


chr8
133925337
133925510
174
20
+
TG
NM_003235.4


chr8
139163459
139165437
1979
13

FAM135B
NM_015912.3


chr8
139601514
139601679
166
65

COF22A1
NM_152888.2


chr8
140630758
140631261
504
2

KCNK9
NR_104210.1


chr8
144990492
144996546
6055
32

PFEC
NM_201380.3


chr8
145770918
145771163
246
5

ARHGAP39
NM_025251.2


chr9
14088266
14088366
101
11

NFIB
NM_001190737.1


chr9
14112989
14113089
101
10

NFIB
NM_001190737.1


chr9
14116206
14116345
140
9

NFIB
NM_001190737.1


chr9
14120438
14120623
186
8

NFIB
NM_001190737.1


chr9
14125630
14125765
136
7

NFIB
NM_001190737.1


chr9
14146687
14146806
120
6

NFIB
NM_001190737.1


chr9
14150143
14150264
122
5

NFIB
NM_001190737.1


chr9
14155808
14155908
101
4

NFIB
NM_001190737.1


chr9
14179702
14179802
101
3

NFIB
NM_001190737.1


chr9
14306987
14307519
533
2

NFIB
NM_001190737.1


chr9
14313445
14313545
101
1

NFIB
NM_001190737.1


chr9
21968184
21968284
101
3

CDKN2A
NM_058195.3


chr9
21970899
21971208
310
2

CDKN2A
NM_058195.3


chr9
21974676
21974826
151


CDKN2A
NM_058195.3


chr9
21994137
21994330
194
1

CDKN2A
NM_058195.3


chr9
27949107
27950605
1499
7

LINGO2
NM_001258282.1


chr9
33385564
33385664
101
7

AQP7
NM_001170.2


chr9
37014992
37015149
158
3

PAX5
NM_016734.2


chr9
37486583
37486778
196
2
+
POLR1E
NM_022490.2


chr9
93650796
93650909
114
13
+
SYK
NM_003177.6


chr9
104499571
104499920
350
1

GRIN3A
NM_133445.2


chr9
111745475
111745575
101
6

CTNNAL1
NM_003798.3


chr9
112898500
112900745
2246
8
+
PALM2-AKAP2
NM_007203.4


chr9
113703900
113704085
186
3

LPAR1
NM_001401.3


chr9
119976687
119976992
306
3

ASTN2
NM_014010.4


chr9
121929388
121930287
900
8

DBC1
NM_014618.2


chr9
131348113
131348216
104
19
+
SPTAN1
NM_001130438.2


chr9
139390522
139392010
1489
34

NOTCH1
NM_017617.4


chr9
139393349
139393449
101
33

NOTCH1
NM_017617.4


chr9
139393563
139393711
149
32

NOTCH1
NM_017617.4


chr9
139395003
139395299
297
31

NOTCH1
NM_017617.4


chr9
139396199
139396365
167
30

NOTCH1
NM_017617.4


chr9
139396446
139396546
101
29

NOTCH1
NM_017617.4


chr9
139396723
139396940
218
28

NOTCH1
NM_017617.4


chr9
139397633
139397782
150
27

NOTCH1
NM_017617.4


chr9
139399124
139399556
433
26

NOTCH1
NM_017617.4


chr9
139399761
139400333
573
25

NOTCH1
NM_017617.4


chr9
139400978
139401091
114
24

NOTCH1
NM_017617.4


chr9
139401167
139401425
259
23

NOTCH1
NM_017617.4


chr9
139401756
139401889
134
22

NOTCH1
NM_017617.4


chr9
139402406
139402591
186
21

NOTCH1
NM_017617.4


chr9
139402683
139402837
155
20

NOTCH1
NM_017617.4


chr9
139403321
139403523
203
19

NOTCH1
NM_017617.4


chr9
139404184
139404413
230
18

NOTCH1
NM_017617.4


chr9
139405104
139405257
154
17

NOTCH1
NM_017617.4


chr9
139405603
139405723
121
16

NOTCH1
NM_017617.4


chr9
139407472
139407586
115
15

NOTCH1
NM_017617.4


chr9
139407843
139407989
147
14

NOTCH1
NM_017617.4


chr9
139408961
139409154
194
13

NOTCH1
NM_017617.4


chr9
139409741
139409852
112
12

NOTCH1
NM_017617.4


chr9
139409934
139410168
235
11

NOTCH1
NM_017617.4


chr9
139410432
139410546
115
10

NOTCH1
NM_017617.4


chr9
139411723
139411837
115
9

NOTCH1
NM_017617.4


chr9
139412203
139412389
187
8

NOTCH1
NM_017617.4


chr9
139412588
139412744
157
7

NOTCH1
NM_017617.4


chr9
139413042
139413276
235
6

NOTCH1
NM_017617.4


chr9
139413894
139414017
124
5

NOTCH1
NM_017617.4


chr9
139417301
139417640
340
4

NOTCH1
NM_017617.4


chr9
139418168
139418431
264
3

NOTCH1
NM_017617.4


chr9
139438465
139438565
101
2

NOTCH1
NM_017617.4


chr9
139440158
139440258
101
1

NOTCH1
NM_017617.4


chr10
7214456
7214623
168
18

SFMBT2
NM_001018039.1


chr10
18276392
18276492
101
7
+
SLC39A12
NM_001145195.1


chr10
43882405
43882523
119
4

HNRNPF
NM_001098206.1


chr10
45941018
45941118
101
14
+
ALOX5
NM_000698.4


chr10
50819199
50820314
1116
1
+
SLC18A3
NM_003055.2


chr10
55955479
55955595
117
12

PCDH15
NM_001142763.1


chr10
56138541
56138702
162
5

PCDH15
NM_001142763.1


chr10
62647976
62648791
816
6

RHOBTB1
NM_014836.4


chr10
68686714
68688016
1303


CTNNA3
NM_013266.3


chr10
81070746
81070846
101
24
+
ZMIZ1
NM_020338.3


chr10
84745206
84745340
135
9
+
NRG3
NM_001010848.3


chr10
87614252
87614372
121
8

GRID1
NM_017551.2


chr10
89624216
89624316
101
1
+
PTEN
NM_000314.6


chr10
89653774
89653874
101
2
+
PTEN
NM_000314.6


chr10
89685242
89685342
101
3
+
PTEN
NM_000314.6


chr10
89690774
89690874
101
4
+
PTEN
NM_000314.6


chr10
89692769
89693008
240
5
+
PTEN
NM_000314.6


chr10
89711874
89712016
143
6
+
PTEN
NM_000314.6


chr10
89717609
89717776
168
7
+
PTEN
NM_000314.6


chr10
89720650
89720875
226
8
+
PTEN
NM_000314.6


chr10
89725043
89725229
187
9
+
PTEN
NM_000314.6


chr10
117884807
117885065
259
6

GFRA1
NM_005264.5


chr11
532635
532755
121
5

HRAS
NM_001130442.2


chr11
533296
533612
317
4

HRAS
NM_001130442.2


chr11
533766
533945
180
3

HRAS
NM_001130442.2


chr11
534211
534322
112
2

HRAS
NM_001130442.2


chr11
5529367
5530481
1115
2

UBQLN3
NM_017481.3


chr11
21592314
21592463
150
18
+
NELL1
NM_006157.4


chr11
66617733
66617886
154
17

PC
NM_022172.2


chr11
67209118
67209218
101
4

CORO1B
NM_020441.2


chr11
68696699
68696799
101
8
+
IGHMBP2
NM_002180.2


chr11
69456081
69456279
199
1
+
CCND1
NM_053056.2


chr11
69457798
69458014
217
2
+
CCND1
NM_053056.2


chr11
69458599
69458759
161
3
+
CCND1
NM_053056.2


chr11
69462761
69462910
150
4
+
CCND1
NM_053056.2


chr11
69465885
69466050
166
5
+
CCND1
NM_053056.2


chr11
70049565
70049851
287
1
+
FADD
NM_003824.3


chr11
70052238
70052579
342
2
+
FADD
NM_003824.3


chr11
70253397
70253497
101
3
+
CTTN
NM_001184740.1


chr11
70253610
70253710
101
4
+
CTTN
NM_001184740.1


chr11
70255936
70256066
131
5
+
CTTN
NM_001184740.1


chr11
70260647
70260758
112
6
+
CTTN
NM_001184740.1


chr11
70261746
70261846
101
7
+
CTTN
NM_001184740.1


chr11
70263118
70263229
112
8
+
CTTN
NM_001184740.1


chr11
70265851
70265962
112
9
+
CTTN
NM_001184740.1


chr11
70266505
70266616
112
10
+
CTTN
NM_001184740.1


chr11
70269023
70269123
101
11
+
CTTN
NM_001184740.1


chr11
70271422
70271522
101
12
+
CTTN
NM_001184740.1


chr11
70275156
70275305
150
13
+
CTTN
NM_001184740.1


chr11
70277291
70277391
101
14
+
CTTN
NM_001184740.1


chr11
70279206
70279384
179
15
+
CTTN
NM_001184740.1


chr11
70279738
70279838
101
16
+
CTTN
NM_001184740.1


chr11
70281128
70281228
101
17
+
CTTN
NM_001184740.1


chr11
70281571
70281850
280
18
+
CTTN
NM_001184740.1


chr11
70282387
70282514
128
19
+
CTTN
NM_001184740.1


chr11
84027931
84028145
215


DLG2
NM_001142699.1


chr11
101981579
101981900
322
1
+
YAP1
NM_001130145.2


chr11
101984874
101985125
252
2
+
YAP1
NM_001130145.2


chr11
102033186
102033302
117
3
+
YAP1
NM_001130145.2


chr11
102056748
102056862
115
4
+
YAP1
NM_001130145.2


chr11
102076623
102076805
183
5
+
YAP1
NM_001130145.2


chr11
102080221
102080321
101
6
+
YAP1
NM_001130145.2


chr11
102094352
102094483
132
7
+
YAP1
NM_001130145.2


chr11
102098199
102098312
114
8
+
YAP1
NM_001130145.2


chr11
102100432
102100671
240
9
+
YAP1
NM_001130145.2


chr11
102738638
102738799
162
5, 6

MMP12
NM_002426.5


chr11
122848357
122848580
224
3

BSX
NM_001098169.1


chr11
132016210
132016342
133
2
+
NTM
NM_001144058.1


chr11
132527002
132527155
154
2

OPCML
NM_002545.4


chr12
4479555
4479838
284
3

FGF23
NM_020638.2


chr12
7635997
7636248
252
12

CD163
NM_004244.5


chr12
20832976
20833115
140
16
+
PDE3A
NM_000921.4


chr12
24398208
24398319
112


SOX5
NM_152989.4


chr12
25360174
25360274
101
6

KRAS
NM_033360.3


chr12
25362729
25362846
118
6

KRAS
NM_033360.3


chr12
25368375
25368495
121
5

KRAS
NM_033360.3


chr12
25378548
25378707
160
4

KRAS
NM_033360.3


chr12
41966189
41967569
1381
10
+
PDZRN4
NM_001164595.1


chr12
48526695
48526800
106
7
+
PFKM
NM_001166686.1


chr12
50367072
50367303
232
1
+
AQP6
NM_001652.3


chr12
54396247
54396387
141
2
+
HOXC9
NM_006897.2


chr12
56221127
56222202
1076
2

DNAJC14
NM_032364.5


chr12
63541188
63541364
177
2

AVPR1A
NM_000706.4


chr12
69202214
69202314
101
1
+
MDM2
NM_002392.5


chr12
69202980
69203080
101
2
+
MDM2
NM_002392.5


chr12
69207321
69207421
101
3
+
MDM2
NM_002392.5


chr12
69210591
69210725
135
4
+
MDM2
NM_002392.5


chr12
69214079
69214179
101
5
+
MDM2
NM_002392.5


chr12
69218126
69218226
101
6
+
MDM2
NM_002392.5


chr12
69218333
69218433
101
7
+
MDM2
NM_002392.5


chr12
69222550
69222711
162
8
+
MDM2
NM_002392.5


chr12
69229608
69229764
157
9
+
MDM2
NM_002392.5


chr12
69230440
69230540
101
10
+
MDM2
NM_002392.5


chr12
69233053
69233629
577
11
+
MDM2
NM_002392.5


chr12
70003992
70004471
480
1

LRRC10
NM_201550.3


chr12
72056923
72057344
422
1

ZFC3H1
NM_144982.4


chr12
78400201
78400964
764
8
+
NAV3
NM_014903.5


chr12
81471975
81472120
146
1
+
ACSS3
NM_024560.3


chr12
109972412
109972571
160
28
+
UBE3B
NM_183415.2


chr12
112182634
112182734
101
14
+
ACAD10
NM_001136538.1


chr12
113515302
113515402
101
2
+
DTX1
NM_004416.2


chr12
113704009
113704109
101
5
+
TPCN1
NM_001143819.2


chr12
122812641
122812741
101
17

CLIP1
NM_001247997.1


chr12
130184385
130185232
848
2

TMEM132D
NM_133448.2


chr13
36700036
36700223
188
2

DCLK1
NM_004734.4


chr13
67799495
67802562
3068
2

PCDH9
NM_203487.2


chr13
70549770
70549923
154
2

KLHL1
NM_020866.2


chr13
88327662
88329861
2200
2
+
SLITRK5
NM_015567.1


chr13
108518048
108518794
747
1

FAM155A
NM_001080396.2


chr14
23346341
23346654
314
7
+
LRP10
NM_014045.4


chr14
23444182
23444313
132
5

AJUBA
NM_032876.5


chr14
23447552
23447654
103
2

AJUBA
NM_032876.5


chr14
23450499
23451384
886
1

AJUBA
NM_032876.5


chr14
23887494
23887594
101
30

MYH7
NM_000257.3


chr14
24788947
24789094
148
22

ADCY4
NM_001198592.1


chr14
30046451
30046551
101
18

PRKD1
NM_002742.2


chr14
42360586
42361026
441
4
+
LRFN5
NM_152447.4


chr14
47530495
47530773
279
7

MDGA2
NM_001113498.2


chr14
52520338
52520463
126
5

NID2
NM_007361.3


chr14
59112195
59113679
1485
4
+
DACT1
NM_016651.5


chr14
70633590
70634947
1358
2

SLC8A3
NM_183002.2


chr14
80328090
80328224
135
17
+
NRXN3
NM_004796.5


chr14
95921754
95922001
248
5

SYNE3
NM_152592.4


chr14
102508971
102509085
115
69
+
DYNC1H1
NM_001376.4


chr14
105246425
105246553
129
3

AKT1
NM_001014431.1


chr14
106054571
106054674
104


DKFZp686O16217


chr15
23811018
23812439
1422
1
+
MKRN3
NM_005664.3


chr15
24921050
24924416
3367
1
+
NPAP1
NM_018958.2


chr15
26806093
26806284
192
8

GABRB3
NM_000814.5


chr15
28326815
28326986
172
2

OCA2
NM_000275.2


chr15
45007630
45007843
214
2
+
B2M
NM_004048.2


chr15
48500014
48500322
309
2
+
SLC12A1
NM_001184832.1


chr15
74883637
74883737
101
6
+
ARID3B
NM_006465.3


chr15
75641257
75641451
195
2
+
NEIL1
NM_001256552.1


chr15
79292108
79292224
117
18

RASGRF1
NM_002891.4


chr15
84581961
84582061
101
16
+
ADAMTSL3
NM_207517.2


chr15
89424689
89424837
149
3

HAPFN3
NM_178232.3


chr15
91835658
91835758
101
14
+
SV2B
NM_014848.6


chr15
96875607
96875744
138
1
+
NR2F2
NM_021005.3


chr16
3293433
3293684
252
10

MEFV
NM_000243.2


chr16
3452110
3452372
263
1
+
ZNF174
NM_003450.2


chr16
3788559
3788673
115
26

CREBBP
NM_004380.2


chr16
50811735
50811852
118
7
+
CYLD
NM_001042355.1


chr16
56973961
56974110
150
6
+
HERPUD1
NM_014685.3


chr16
64984669
64984857
189
12

CDH11
NM_001797.3


chr16
65032503
65032725
223
4

CDH11
NM_001797.3


chr16
67650647
67650781
135
5
+
CTCF
NM_006565.3


chr16
72188111
72188258
148
4

PMFBP1
NM_031293.2


chr16
77465304
77465454
151
3

ADAMTS18
NM_199355.3


chr16
84132677
84132849
173
3

MBTPS1
NM_003791.3


chr16
89986155
89986384
230
1
+
MC1R
NM_002386.3


chr16
90161901
90162305
405

+
TUBB4Q


chr17
7572918
7573018
101
11

TP53
NM_001276760.1


chr17
7573926
7574033
108
10

TP53
NM_001276760.1


chr17
7576525
7576658
134


TP53
NM_001276760.1


chr17
7576839
7576939
101
9

TP53
NM_001276760.1


chr17
7577018
7577155
138
8

TP53
NM_001276760.1


chr17
7577498
7577608
111
7

TP53
NM_001276760.1


chr17
7578176
7578289
114
6

TP53
NM_001276760.1


chr17
7578369
7578554
186
5

TP53
NM_001276760.1


chr17
7579310
7579580
271
4

TP53
NM_001276760.1


chr17
7579660
7579760
101
3

TP53
NM_001276760.1


chr17
7579826
7579926
101
2

TP53
NM_001276760.1


chr17
10303757
10304049
293
27

MYH8
NM_002472.2


chr17
10369589
10369733
145
4

MYH4
NM_017533.2


chr17
21318730
21319867
1138

+
KCNJ12
NM_001194958.2


chr17
26684313
26684473
161
1, 2

POLDIP2
NM_015584.4


chr17
34077206
34077306
101
2

GAS2L2
NM_139285.3


chr17
37855776
37855876
101

+
ERBB2
NM_001005862.2


chr17
37856478
37856578
101
1
+
ERBB2
NM_004448.3


chr17
37863232
37863452
221
2
+
ERBB2
NM_004448.3


chr17
37864563
37864797
235
3
+
ERBB2
NM_004448.3


chr17
37865560
37865715
156
4
+
ERBB2
NM_004448.3


chr17
37866050
37866150
101
5
+
ERBB2
NM_004448.3


chr17
37866328
37866464
137
6
+
ERBB2
NM_004448.3


chr17
37866582
37866744
163
7
+
ERBB2
NM_004448.3


chr17
37868170
37868310
141
8
+
ERBB2
NM_004448.3


chr17
37868564
37868711
148
9
+
ERBB2
NM_004448.3


chr17
37869395
37869532
138

+
ERBB2
NM_004448.3


chr17
37871525
37871625
101
10
+
ERBB2
NM_004448.3


chr17
37871688
37871799
112
11
+
ERBB2
NM_004448.3


chr17
37871982
37872202
221
12
+
ERBB2
NM_004448.3


chr17
37872543
37872696
154
13
+
ERBB2
NM_004448.3


chr17
37872757
37872868
112
14
+
ERBB2
NM_004448.3


chr17
37873562
37873747
186
15
+
ERBB2
NM_004448.3


chr17
37876013
37876113
101
16
+
ERBB2
NM_004448.3


chr17
37879561
37879720
160
17
+
ERBB2
NM_004448.3


chr17
37879780
37879923
144
18
+
ERBB2
NM_004448.3


chr17
37880154
37880273
120
19
+
ERBB2
NM_004448.3


chr17
37880968
37881174
207
20
+
ERBB2
NM_004448.3


chr17
37881291
37881467
177
21
+
ERBB2
NM_004448.3


chr17
37881567
37881667
101
22
+
ERBB2
NM_004448.3


chr17
37881949
37882116
168
23
+
ERBB2
NM_004448.3


chr17
37882804
37882922
119
24
+
ERBB2
NM_004448.3


chr17
37883057
37883266
210
25
+
ERBB2
NM_004448.3


chr17
37883537
37883810
274
26
+
ERBB2
NM_004448.3


chr17
37883931
37884307
377
27
+
ERBB2
NM_004448.3


chr17
42248166
42248420
255
1
+
ASB16
NM_080863.4


chr17
65026607
65026939
333
4
+
CACNG4
NM_014405.3


chr17
80788943
80790329
1387

+
TBCD
NM_005993.4


chr18
580456
580887
432
1
+
CETN1
NM_004066.2


chr18
5397092
5397423
332
18

EPB41L3
NM_012307.3


chr18
5415858
5416180
323
13

EPB41L3
NM_012307.3


chr18
13825967
13826657
691
1
+
MC5R
NM_005913.2


chr18
13884632
13885468
837
2

MC2R
NM_000529.2


chr18
22804523
22807584
3062
4

ZNF521
NM_015461.2


chr18
63547682
63547974
293
12
+
CDH7
NM_033646.2


chr18
64172066
64172442
377
12

CDH19
NM_021153.3


chr18
67406200
67406339
140
6
+
DOK6
NM_152721.5


chr19
2121161
2121310
150
13

AP3D1
NM_001261826.1


chr19
3964691
3964914
224
3

DAPK3
NM_001348.2


chr19
5455842
5456254
413
1
+
ZNRF4
NM_181710.3


chr19
10597317
10597504
188
6

KEAP1
NM_203500.1


chr19
10599857
10600054
198
5

KEAP1
NM_203500.1


chr19
10600313
10600539
227
4

KEAP1
NM_203500.1


chr19
10602242
10602948
707
3

KEAP1
NM_203500.1


chr19
10610060
10610719
660
2

KEAP1
NM_203500.1


chr19
20728545
20728771
227
4

ZNF737
NM_001159293.1


chr19
36218401
36218501
101
16
+
KMT2B
NM_014727.2


chr19
37440566
37440780
215
7
+
ZNF568
NM_198539.3


chr19
42260627
42260788
162
2
+
CEACAM6
NM_002483.6


chr19
49933848
49933948
101
12

SLC17A7
NM_020309.3


chr19
52619657
52620045
389
4

ZNF616
NM_178523.4


chr19
54313207
54314440
1234
3

NLRP12
NM_001277126.1


chr19
54466452
54466611
160
1
+
CACNG8
NM_031895.5


chr19
54677829
54678114
286
8

MBOAT7
NM_024298.4


chr19
55377999
55378186
188
9
+
KIR3DL2
NM_006737.3


chr19
57640090
57642406
2317
4
+
USP29
NM_020903.2


chr20
29625872
29625984
113
4
+
FRG1B
NR_003579.1


chr20
32264537
32264785
249
7

E2F1
NM_005225.2


chr20
32264910
32265136
227
6

E2F1
NM_005225.2


chr20
32265231
32265346
116
5

E2F1
NM_005225.2


chr20
32266006
32266159
154
4

E2F1
NM_005225.2


chr20
32267560
32267780
221
3

E2F1
NM_005225.2


chr20
32268127
32268227
101
2

E2F1
NM_005225.2


chr20
32273809
32274070
262
1

E2F1
NM_005225.2


chr20
33033101
33033228
128
12
+
ITCH
NM_001257137.2


chr20
36850850
36850999
150
10

KIAA1755
NM_001029864.1


chr20
50139651
50140541
891
2

NFATC2
NM_173091.3


chr20
61488772
61488905
134
4

TCFL5
NM_006602.3


chr21
41142929
41143079
151
4
+
IGSF5
NM_001080444.1


chr21
47545907
47546046
140
26
+
COL6A2
NM_001849.3


chr22
22127161
22127271
111
7

MAPK1
NM_002745.4


chr22
30722668
30722864
197
1

TBC1D10A
NM_001204240.1


chr22
32554979
32555104
126
1

C22orf42
NM_001010859.1


chr22
36708122
36708258
137
14

MYH9
NM_002473.5


chr22
37603210
37603433
224
2

SSTR3
NM_001051.4


chr22
41565506
41565620
115
26
+
EP300
NM_001429.3


chr22
42070953
42071075
123
3

NHP2L1
NM_001003796.1


chr22
42538728
42538881
154
3

CYP2D7P1
NR_002570.3


chrX
12734345
12734914
570
15
+
FRMPD4
NM_014728.3


chrX
30268766
30269598
833
2
+
MAGEB1
NM_177404.2


chrX
32328251
32328385
135
42

DMD
NM_004006.2


chrX
34148022
34150317
2296
1

FAM47A
NM_203408.3


chrX
41000545
41000684
140
9
+
USP9X
NM_001039590.2


chrX
53560971
53561071
101
83

HUWE1
NM_031407.6


chrX
74494188
74494382
195
1
+
UPRT
NM_145052.3


chrX
78616825
78616976
152
5

ITM2A
NM_004867.4


chrX
79281104
79281236
133
4
+
TBX22
NM_016954.2


chrX
92926977
92928298
1322
1

NAP1L3
NM_004538.5


chrX
102337167
102337278
112
9

NXF3
NM_022052.1


chrX
107976933
107979425
2493
1

IRS4
NM_003604.2


chrX
110644216
110644425
210
3

DCX
NM_000555.3


chrX
114540769
114540930
162
4
+
LUZP4
NM_016383.4


chrX
123514614
123515060
447
32

TENM1
NM_001163278.1


chrX
134483098
134483239
142
3
+
ZNF449
NM_152695.5


chrX
142716465
142718808
2344
2

SLITRK4
NM_001184749.2


chrX
142795425
142795525
101
2

SPANXN2
NM_001009615.2









Plasma and PBL samples from HNSCC patients at diagnosis and healthy donors by CAPP-Seq, utilizing 10-30 ng of input DNA were profiled. To achieve sensitive detection of ctDNA at low abundance, we applied a CAPP-Seq selector optimized to maximize the number of detected mutations in HNSCC (Table 2 and FIG. 10). We further improved our analytical sensitivity through integrated Digital Error Suppression (iDES), incorporating custom molecular barcodes and removing background base substitution errors as identified within healthy donor plasma samples (Methods).









TABLE 2







Reported yields of cell-free DNA


normalized to total plasma volume











sampleID
dsDNApermLPlasma
timepoint















1
10.69473684
Normal



2
19.6137931
Normal



3
11.2
Normal



4
9.76
Normal



5
11.57
Normal



6
7.72
Normal



7
15.83283582
Normal



1
5.09
Diagnosis



1
9.6
Post-surgery



2
12.34
Diagnosis



2
16.65
Mid-radiotherapy



3
5.55
Diagnosis



3
6.443076923
Mid-radiotherapy



3
5.659701493
Post-treatment-1



3
8.516129032
Post-treatment-2



4
13.65
Diagnosis



4
13.18
Mid-radiotherapy



5
11.76
Diagnosis



5
8.66
Mid-radiotherapy



6
6.75
Diagnosis



6
9.6
Mid-radiotherapy



6
13.23
Post-treatment-1



6
8.68
Post-treatment-2



7
10.28571429
Diagnosis



7
15.08571429
Mid-radiotherapy



7
4.96875
Post-surgery



7
6.941538462
Post-treatment-1



8
16.68
Diagnosis



8
12.93
Mid-radiotherapy



8
23.21
Post-surgery



9
20.01509434
Diagnosis



9
20.05970149
Mid-radiotherapy



10
12.18
Diagnosis



10
14.32
Mid-radiotherapy



10
8.93
Post-surgery



11
27.04
Diagnosis



11
20.06
Mid-radiotherapy



11
26.68
Post-surgery



11
9.07
Post-treatment-1



12
7.2
Diagnosis



12
6.93
Post-surgery



13
8.87
Diagnosis



13
7.69
Post-surgery



14
5.73
Diagnosis



14
9.28
Post-surgery



15
17.31940299
Diagnosis



15
19.63636364
Mid-radiotherapy



16
21.75
Diagnosis



16
30.28
Post-surgery



17
14.02
Diagnosis



17
15.65
Post-surgery



18
8.076
Diagnosis



18
8.671
Mid-radiotherapy



18
7.504
Post-surgery



18
10.386
Post-treatment-1



19
5.16
Diagnosis



19
11.41333333
Mid-radiotherapy



19
17.6
Post-surgery



20
52.58181818
Diagnosis



20
9.523809524
Mid-radiotherapy



20
24.38709677
Post-surgery



20
55.8
Post-treatment-1



21
8.903225806
Diagnosis



21
10.28571429
Mid-radiotherapy



21
14.55
Post-surgery



21
9.68
Post-treatment-1



22
69.96
Diagnosis



22
10.25
Mid-radiotherapy



22
26.71
Post-treatment-1



23
8.023880597
Diagnosis



23
6.889655172
Mid-radiotherapy



23
13.73333333
Post-surgery



24
4.34
Diagnosis



24
11.78
Post-surgery



25
13.76
Diagnosis



25
10
Post-surgery



26
31.16
Diagnosis



26
24
Mid-radiotherapy



26
16.8
Post-treatment-1



27
7.219047619
Diagnosis



27
6.978461538
Mid-radiotherapy



27
6.95625
Post-surgery



28
27.78
Diagnosis



28
7.1
Mid-radiotherapy



28
8.62
Post-surgery



29
14.86451613
Diagnosis



29
12.16
Mid-radiotherapy



29
8.828571429
Post-treatment-1



30
10.575
Diagnosis



30
12.75
Post-surgery



30
14.55
Post-treatment-1



4
14.42033898
Normal



4
8.66
Normal



4
6.92
Normal



4
12.51764706
Normal



4
11.70526316
Normal



4
13.99148936
Normal



4
7.670588235
Normal



4
11.328
Normal



4
8.465454545
Normal



4
8.27
Normal



4
6.498461538
Normal



4
12.72
Normal



4
21.63
Normal










After selecting for candidate somatic single nucleotide variants (SNVs) based on plasma profiling and removal of likely germline mutations, we characterized potential false-positives due to clonal hematopoiesis (CH) by comparison with matched PBL profiles. Of the 24 patients with identifiable candidate SNVs, 10 demonstrated identical SNVs within their matched PBL profile with highly correlated mutant allele fractions (MAFs) (R=0.94, p=1.392e−07, FIG. 2B). With the exception of PIK3CA, genes harboring these SNVs were unique to each patient (FIG. 2C). As genes that are commonly affected by CH, such as DNMT3A, TET2, and ASXL1, were not included within the CAPP-Seq selector, our findings of patient-unique SNVs within matched cfDNA and PBL samples further emphasizes the benefit of this approach over gene level filtering. Plasma samples from 4 patients were strictly positive for SNVs derived from CH (FIG. 2D), suggesting that matched PBL profiling may greatly minimize false-positive detection of ctDNA at low abundance.


After removing candidate SNVs potentially reflective of CH, ctDNA was detected within plasma of 20 patients (median [range]: 3 [1-10] SNVs per patient). To evaluate the plausibility of these SNVs, we compared our results to whole-exome sequencing data from 279 HNSCC tumors published by The Cancer Genome Atlas (TCGA)45, observing similarities in frequently mutated genes including TP53 (65% vs. 72%), PIK3CA (20% vs. 21%), FAT1 (15% vs. 23%), and NOTCH1 (10% vs. 19%) (FIG. 2E). Interestingly, two patients presented with single SNVs not found within these genes (GRIN3A and MYC, FIG. 11), demonstrating the added utility of profiling genes with unknown/non-driver effects to increase detection sensitivity OF ctDNA.


Calculating ctDNA abundance based on the mean MAF of SNVs, ctDNA levels ranged from 0.14% to 4.83% (FIG. 2F). This lower limit of detection is similar to that previously described by others utilizing tumor-naïve CAPP-Seq analysis, estimated at ˜0.14%. Including patients with undetectable ctDNA, the median ctDNA abundance across our HNSCC cohort was 0.49%-similar to what has been observed in localized NSCLC by CAPP-Seq.


Tumor-Naive Detection of Methylation-Based ctDNA from Baseline Plasma


Next, we sought to define ctDNA-associated methylation patterns in the HNSCC and healthy control samples. As the CAPP-Seq results illustrated the impact of false positive mutations arising from PBLs, we reasoned that a reduction of false positive ctDNA-associated methylation may be achieved by removal of PBL-derived DNA methylation signals. Therefore, we used matched PBL MeDIP-seq profiles from the HNSCC and healthy control samples to suppress their contribution to the cell-free DNA methylation signal (FIG. 3A) we evaluated whether matched PBL analysis may also enable methylation-based ctDNA detection (FIG. 3A). Pre-treatment HNSCC and healthy donor plasma as well as PBLs were profiled by cfMeDIP-seq, utilizing 5-10 ng of input DNA. As previously described, methylation abundance was defined within nonoverlapping 300 bp windows across chromosomes 1-22 (n=9,603,454 windows) with read counts normalized to reads per kilobase per million (RPKM) (Methods).


As the anti-5mC antibody utilized for methylation pulldown preferentially binds to DNA fragments at increasing CpG densities, including CpG islands, we first characterized this interaction to identify regions likely to be highly represented within cfMeDIP-seq data. We also applied MeDIP-seq to the HNSCC cell-line FaDu to assess the preferential binding of cancer-derived methylated DNA fragments. Comparing DNA fragment pulldown abundance (median RPKM) across windows with varying numbers of CpGs, we observed increasing enrichment up to ≥8 CpGs for both PBLs and FaDu (FIGS. 12A and 12B). FaDu demonstrated greater enrichment compared to PBLs at ≥8 CpGs per 300 bp window. This result is consistent with the established phenomenon of CpG island hypermethylation in cancer cells including FaDu. Based on these observations, we determined that windows with ≥8 CpGs (n=702,488) may be most informative for ctDNA detection and were therefore utilized for all subsequent analysis.


For patients with localized cancer, the vast majority of plasma cell-free DNA originates from PBLs. Therefore, we sought to exploit PBL MeDIP-seq profiles to bioinformatically suppress this contribution to the cell-free DNA signal. We compared RPKM values for each window within cfMeDIP-seq profiles generated from HNSCC and healthy donor cfDNA, to MeDIP-seq profiles generated from FaDu (1-by-1 comparison), unpaired PBLs (1-by-51 comparison), or paired PBLs (1-by-1 comparison). In accordance with PBLs being the main contributor of plasma cell-free DNA, genome-wide methylation profiles were highly correlated between plasma cell-free DNA and either paired or unpaired PBLs (modal R=0.92 and R=0.91, respectively). The strengths of these correlations likely reflect the known outsize contribution of PBLs to plasma cfDNA. In contrast, correlations were weaker between plasma cell-free DNA and FaDu (modal R=0.78) (FIG. 3B).


To select a threshold of decreased methylation across PBLs while considering preferential pulldown, we scaled and normalized PBL cfMeDIP-seq profiles to absolute methylation levels (0-1) based on logistic regression modelling via the MeDEStrand R package (Methods). We selected 99,997 windows that demonstrated median absolute methylation values <0.1 across healthy donor PBLs. When these windows were applied to left-out HNSCC PBLs we observed similar distributions of absolute methylation to that of the utilized healthy donor PBLs (FIG. 3B), demonstrating generalizability of this approach. Likewise, none of these windows individually showed significantly higher methylation across HNSCC PBLs compared to healthy donor PBLs (FIG. 3C and FIG. 12B), limiting any source of HNSCC-specific PBL methylation that may confound ctDNA detection. In other words, these results confirm that the main source of cfDNA methylation in both control and locoregionally confined HPV-negative HNSCC plasma are derived from PBLs and that bioinformatic removal of PBL-derived methylation may limit signals that confound ctDNA quantification.


Tumor-Naïve Detection of Pre-Treatment Methylation-Based ctDNA


To identify common ctDNA-derived hypermethylated regions within our HNSCC cohort, we performed differential methylation analysis comparing HNSCC patients with detectable ctDNA by CAPP-Seq (n=20) to healthy donors. Utilizing the 99,994 300-bp windows depleted for methylation in PBLs, we identified ctDNA-derived differentially methylated regions (DMRs) by comparing the 20 HNSCC patients with CAPP-Seq-detectable ctDNA to the 20 healthy controls. In total we identified 997 differentially methylated regions (DMRs) (hypermethylated: 941, hypomethylated: 56) across HNSCC samples (FIG. 3C). Approximately half of hypermethylated regions (hyper-DMRs) were found to be immediately adjacent to one another, with blocks of hypermethylation extending up to 1800 base-pairs in length (FIG. 13A). These data suggest the presence of CpG islands within the identified hyper-DMRs. Conversely, no adjacent hypomethylated regions (hypo-DMRs) were observed. Of the 300-bp hyper-DMRs, 47.5% resided in contiguous blocks of hypermethylation signals extending up to 1800 bp in length (FIG. 13A), indicative of CpG islands that typically span 300-3000-bp in length. Indeed, CpG islands were significantly enriched for hyper-DMRs (FIG. 3E). In contrast, CpG islands were significantly depleted for hypo-DMRs (FIG. 13B).


To determine whether these hyper-DMRs were indeed enriched for CpG islands, we next assessed the enrichment of hyper-DMRs for CpG islands, shores, shelves, and open seas by permutation analysis (Methods). As expected, a significant enrichment of CpG islands as well as a significant depletion of shores and open sea was observed within the hyper-DMRs (FIG. 3E). In contrast, the hypo-DMRs were significantly enriched for open sea and depleted for CpG islands (Supplementary FIG. 5B), in accordance with hypomethylation of CpG-sparse regions frequently observed across cancers.


Finally, as methylation of certain regions may distinguish tissue-of-origin as previously described using cfMeDIP-seq, we also investigated whether the hyper-DMRs contained regions specific to HNSCC or other cancers. To identify tumor-specific methylated regions, we utilized HumanMethylation450K (hm450k) data generated from primary tumors provided by TCGA (Methods). Comparing primary tumors from breast invasive carcinoma (BRCA), colon adenocarcinoma (COAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD), HNSCC, pancreatic adenocarcinoma (PAAD), and PBLs, we identified sufficient hypermethylated CpGs (≥50) specific for BRCA, COAD, PRAD, and HNSCC (Methods) (FIG. 14). As expected, we observed significant enrichment of the plasma-derived DMRs overlapping with HNSC-specific hypermethylated CpGs, as well as a significant depletion of overlap across BRCA-, COAD-, and PRAD-specific hypermethylated CpGs (FIG. 3F), suggesting that the hyper-DMRs contain regions specific to HNSCC origin when compared to various other cancer types.


Mutation-Based and Methylation-Based ctDNA Detection are Highly Concordant


A growing number of studies have described ctDNA to be associated with decreased fragment length compared to healthy sources of plasma cell-free DNA, providing an additional metric for robust tumor-naïve detection. As targeted sequencing has been previously shown to detect ctDNA at reduced fragment length, we first utilized our CAPP-Seq profiles to determine whether we may observe similar trends within HNSCC patients. For each identified SNV per patient (FIG. 2E), we measured the median length of fragments containing the SNV allele as well as the overlapping reference allele. For cases where multiple SNVs were identified within a patient sample, the median value across all SNVs and their reference alleles was used. In accordance with previous findings, we observed a consistent decrease in ctDNA fragment size compared to healthy cell-free DNA across patients (median [range] Δ=−17.5 [1-58] bp) (FIG. 4A). There was no significant association between the mean MAF of these mutations and fragment length (FIG. 15A).


Unlike bisulfite-based DNA methylation approaches, cfMeDIP-seq does not cause DNA degradation and, therefore, preserves the original fragment size distribution. This provides a novel opportunity to map DNA methylation and fragment lengths concomitantly. The distribution of fragment lengths within the previously identified plasma derived hyper-DMRs for each patient was assessed. Due to the nature of these regions having low methylation across our healthy donors, DNA fragments across donors were combined for comparison. Similar to the mutation-based analysis, we observed a reduction in fragment length from 19/20 CAPP-Seq positive patients compared to grouped healthy controls (median [range] Δ=−7 [1-21] bp) (FIG. 4B). This represented a smaller reduction in fragment lengths compared with the mutation-based analysis, possibly due to partial contribution by healthy tissues of cell-free DNA fragments within the hyper-DMRs. Supporting this notion, the samples with the shortest hyper-DMR fragments displayed higher methylated ctDNA abundance (Pearson r=−0.64, p=0.002) (FIG. 15B). When the ratio of small (100-150 bp) versus large (151-220 bp) fragments were used for our hyper-DMRs, an approach previously described to enrich for ctDNA, we observed a similar trend of ctDNA enrichment across the majority of CAPP-Seq positive HNSCC samples (median [range]=28 [−8 to 63]%) (FIG. 4C).


To assess how the plasma cell-free DNA hyper-DMRs identified in our HNSCC cohort may vary across individuals within these small fragments (100-150 bp), we first performed hierarchical clustering. Four dominant clusters emerged utilizing the ConsensusClusterPlus R package, each with distinct levels of methylation across the hyper-DMRs (FIG. 4E and FIG. 16C). Likewise, the three clusters were defined by distinct ctDNA abundance as determined by CAPP-Seq (FIG. 16D), suggesting a potential relationship between mean hyper-DMR methylation and mutation-based ctDNA abundance.


We next investigated whether fragment lengths were concordant between ctDNA molecules Identified by both CAPP-Seq and cfMeDIP-seq, potentially providing an additional layer of validation towards our multimodal approach. To minimize the possibility of background DNA fragments confounding the calculated fragment length of ctDNA within cfMeDIP-seq profiles, we limited analysis to patients above the median methylation levels across hyper-DMRs (n=10 HNSCC patients). Strikingly, ctDNA fragment length was highly concordant between paired CAPP-Seq and cfMeDIP-seq profiles for each patient (Pearson r=0.86, p=0.0016) (FIG. 4C) despite entirely different genomic regions being represented with these two profiling approaches (CAPP-Seq: 43 distinct mutations, cfMeDIP-seq: 941 hyper-DMRs).


To further characterize the relationship between hyper-DMR methylation levels and mutation-based ctDNA abundance, we compared the mean RPKM values across the 941 hyper-DMRs to the mean MAF values determined by CAPP-Seq for each patient. Similar to the trends we observed between methylation clusters, we observed a significant positive correlation (Pearson correlation, R=0.85, p=5e-10) (FIG. 4F). To evaluate the sensitivity of ctDNA detection within these hyper-DMRs by cfMeDIP-seq, we compared mean RPKM values between our HNSCC cohort and healthy donors. For CAPP-Seq positive patients (n=20), ctDNA detection was highly concordant (AUC=0.998) with a marginal decrease in performance upon incorporation of CAPP-Seq negative patients (n=12) (AUC=0.944) (FIG. 4G). Cross validation (n=50 samplings) across CAPP-Seq positive patients and healthy donors resulted in a median AUC value of 0.984 (FIG. 16A), demonstrating the robustness of the approach disclosed herein.


Based on these observations, we evaluated whether we may enrich ctDNA within cfMeDIP-seq profiles by limiting analysis to cell-free DNA fragments of reduced length. We assessed the proportion of cell-free DNA fragments within hyper-DMRs consisting of small (100 to 150 bp) fragments, as similar methods have been described to enrich for ctDNA using non-methylation-based approaches. Indeed, this resulted in ctDNA enrichment across the majority of CAPP-Seq positive HNSCC samples (median [range]=28 [−8 to 63] %) but not for any of the healthy controls (FIG. 4D). Thus, in silico size selection of cell-free DNA fragments enriches for ctDNA within cfMeDIP-seq libraries and may contribute to tumor-naive multimodal ctDNA analysis.


In patients with localized non-metastatic cancer, detection of ctDNA by CAPP-Seq at diagnosis has previously been described to be associated with poor prognosis. Likewise, ctDNA levels as assessed by methylation of SHOX2 and SEPT9 are associated with poor prognosis in HNSCC. Therefore, we asked whether detection or quantification of ctDNA by CAPP-Seq and cfMeDIP-seq at diagnosis would be associated with clinical outcomes within our HNSCC cohort. Indeed, detection of ctDNA by CAPP-Seq (i.e. CAPP-Seq positive vs. CAPP-Seq negative) (hazard ratio [HR]=7.6, log-rank p=0.026; Supplementary FIG. 8D) as well as increased methylation within our previously identified hyper-DMRs (i.e., methylation cluster 1+2+3 vs. methylation cluster 4) (HR=4.51, p=0.038; FIG. 4G), was correlated with shorter survival times. Consistent with this finding, mean RPKM across the hyper-DMRs correlated with cancer stage (Supplementary FIG. 8E).


We next compared the median fragment length of ctDNA identified by either mutation- or methylation-based profiling. To minimize the possibility of background DNA fragments confounding the calculated fragment length of ctDNA within cfMeDIP-seq profiles, we selected patients with high ctDNA abundance as defined by hierarchical clustering (i.e. methylation clusters 1 and 2, FIG. 4D, Supplemental FIG. 8A-B). With this approach, ctDNA fragment length was highly concordant between paired CAPP-Seq and cfMeDIP-seq profiles for each patient (R=0.83, p=0.0016) (FIG. 4H) despite entirely different genomic regions being represented with these two profiling approaches. In addition, similar to our analysis with fragments of all lengths, we observed the same relationship between small fragment ratio and ctDNA fragment length by CAPP-Seq (R=−0.79, p=0.0038) (FIG. 4I).


These results suggest that the similar decrease in fragment length observed from ctDNA detected by CAPP-Seq and cfMeDIP-seq may be a result of inherent properties of the tumor, rather than by genomic region, and that utilization of shorter fragment lengths may contribute to more specific identification of ctDNA.


Application of Multimodal ctDNA Detection for Prognostication


To evaluate the potential clinical applications of tumor-naive multimodal ctDNA analysis, we compared ctDNA with clinical outcomes in the HNSCC cohort. Fragment-length informed cfMeDIP-seq profiles were strongly associated with MAFs in matched CAPP-Seq profiles (Pearson r=0.85, p=3×10-9), suggesting that methylation intensity within the 941 hyper-DMRs is indeed reflective of ctDNA abundance (FIG. 5C). Importantly, cross-validation analysis confirmed the robustness of these hyper-DMRs for detecting ctDNA (FIG. 16C). Patients with ctDNA detected in baseline plasma by both mutation- and methylation-based methods (n=19) were significantly more likely to have advanced disease (i.e., stage III-IVA) (n=18/19) when compared to patients with no detectable ctDNA (n=8/13) (Fisher's exact test p=0.028) and displayed dramatically worse overall survival (hazard ratio [HR]=7.55, 95% confidence interval [CI]=[0.95 to 59.94], log-rank p=0.025) (FIG. 5G). In comparison, stage alone was unable to predict patients with worse overall survival (HR=2.59, 95% CI=[0.32 to 20.46], log-rank p=0.35) (FIG. 16D), further demonstrating the potential clinical utility of multimodal ctDNA profiling.


Due to the known effects of DNA methylation on gene expression and resultant functional activity of cancer drivers, we reasoned that ctDNA methylation patterns at particular loci might have prognostic significance independent of ctDNA abundance. To evaluate whether our previously identified hyper-DMRs contain specific regions associated with prognosis independent of ctDNA abundance, we interrogated DNA methylation, RNA expression, and clinical outcome data provided by the TCGA for all available HNSCC patients (n=520) (FIG. 5C). First, we calculated mean β-values across all CpGs contained within distinct 300-bp windows from TCGA hm450k methylation array data. Limiting analysis to probed hm450k regions overlapping with our plasma-derived hyper-DMRs (n=764/941), we identified 483 hypermethylated regions in primary tumors (n=520) compared to adjacent normal tissue (n=50) (Wilcoxon test, FDR<0.05, log 2FC>1). We observed that several of these hypermethylated regions overlapped or were located near CpGs within genes that are profiled by commercially available methylation-based ctDNA diagnostic tests, including SEPT9 and SHOX2 which have been previously assessed in HNSCC, as well as TWIST1 and ONECUT2 (FIG. 17A). These results provide further evidence supporting the potential clinical relevance of our plasma derived hyper-DMRs.


To further probe the potential clinical utility of these hypermethylated regions held in common by our HNSCC cohort and TCGA HNSC hm450k profiles, we performed univariate Cox proportional-hazards regression across all TCGA HNSCC patients with available hm450k profiles and disease-specific survival (DSS) outcomes (n=493/520). We identified 33 regions that were significantly associated with DSS (p<0.05). To further select prognostic regions likely to have a functional role in tumorigenesis, we compared the methylation levels of each region (n=33) to the expression of surrounding gene transcripts within 2 kb. Next, we used the TCGA HNSCC cohort to identify a subset of the 483 DMRs that were associated with (1) prognosis in multivariable Cox regression and (2) expression of neighboring gene transcripts. Five regions were identified to satisfy both criteria, with increased methylation of each region resulting in higher expression of ZNF323/ZSCAN31, LINC01391, and GATA2-AS1 (FIG. 5G, FIG. 17A-17C, as well as lower expression of STK3/MST2 and OSR1, respectively (FIG. 5H) (FIG. 5D). The regions associated with decreased and increased expression as a result of methylation were found to reside within the promoter or 1st exon/intron and gene body, respectively. We constructed a composite methylation score (CMS) from these 5 regions (Table 6) and stratified the TCGA HNSCC cohort according to this score (FIG. 5E). A higher CMS was significantly associated with inferior survival outcomes (HR=1.67, 95% CI=[1.25, 2.21], log-rank p=3.4×10−4).


Finally, we evaluated whether the CMS may also provide similar prognostic information when applied to ctDNA. To enrich for ctDNA, analysis of cfMeDIP-seq libraries were limited to fragments between 100-150 bp in length as described above (FIG. 4E). To account for the relative contribution of ctDNA methylation levels provided by the 5 putative prognostic markers, we normalized the cfMeDIP-seq RPKM values from these regions to the entire 941 hyper-DMRs. This produced a similar trend with higher CMS being marginally associated with worse survival (log-rank p=0.1; HR=3.06) (FIG. 5F) suggesting that increased methylation of these putative prognostic regions identified from TCGA may also be informative within cfMeDIP-seq profiles. Moreover, these results highlight how plasma cell-free DNA methylome profiling may be leveraged in combination with existing multi-omic cancer databases for biomarker discovery.


Disease Surveillance after Definitive Treatment by cfMeDIP-Seq


As cfMeDIP-seq achieved sensitive and quantitative ctDNA detection in HNSCC patients, we reasoned that as with CAPP-seq, cfMeDIP-seq may also be capable of monitoring therapy-related changes in ctDNA abundance. To quantify percent ctDNA within posttreatment cfMeDIP-seq profiles, we applied a linear transformation of mean RPKM across the previously identified plasma-derived hyper-DMRs (n=941), limiting fragment size between 100 to 150 bp to further enrich ctDNA. We calculated the detection threshold of 0.2% ctDNA based on the maximum of mean RPKM values observed across all healthy controls. For CAPP-Seq positive HNSCC patients with one or more available post-treatment samples (n=20), cfMeDIP-seq was performed utilizing 10 ng of input cfDNA.


Measuring changes in ctDNA abundance throughout treatment, we observed a variety of kinetics indicative of complete clearance (CC), partial clearance (PC; greater than 90% reduction), or no clearance (NC) (FIG. 6A, Supplementary FIG. 10). Among 18 eligible patients, 5 (28%) demonstrated No Clearance (FIG. 6B). No Clearance patients were more likely to experience disease recurrence compared with those with Complete or Partial Clearance (HR=8.73, 95% CI=[1.5, 50.92], log-rank p=0.0046) (FIG. 6C). Interestingly, all patients with ctDNA abundance greater at last sample collection compared to at diagnosis, demonstrated disease recurrence. In addition, the only patient who did not have documented disease recurrence within this group was lost to follow-up but died within a year after treatment from unknown cause. For the 13 patients with undetectable post-treatment ctDNA by cfMeDIP-seq, 9 remained disease-free with a median of 44.4 months of follow up (min=12.2, max=58.7). Among the other 4 patients, one had persistent disease within regional lymph nodes, and the others experienced relapse 3.5 to 7.7 months (median 7.4 months) after last collection. Of note, these relapses among the patients with undetectable post-treatment ctDNA were considerably more delayed compared to the 4 relapses among the patients with detectable post-treatment ctDNA (median [range]: 3.0 [1.7 to 5.2] months) after last collection. Taken together, these results demonstrate that plasma cell-free DNA methylome profiling by cfMeDIP-seq may be used to assess response to definitive treatment and identify patients at high risk of rapid recurrence.


Discussion

Broad implementation of ctDNA in clinical settings may be accelerated by methods that can be applied across patients and in the absence of tumor material. In the work described, we evaluated the capabilities of multimodal genome-wide cell-free DNA profiling techniques for tumor-naïve detection of ctDNA within an exploratory cohort of low-ctDNA HNSCC patients. We show that incorporation of matched PBLs improves ctDNA detection using both mutations (i.e., CAPP-Seq) as well as DNA methylation (i.e., cfMeDIP-seq). Furthermore, by utilizing CAPP-Seq to stratify patients with detectable and non-detectable ctDNA, we achieved robust identification of ctDNA-derived methylation patterns. We showed for the first time that biophysical properties of plasma cell-free DNA reflective of tumor origin (i.e., reduced fragment length) are conserved across molecular aberrations and detection platforms. Tumor-naïve ctDNA detection and quantification find multiple clinical uses, and the prognostic association of ctDNA abundance and methylation patterns are investigated.


Tumor-naive ctDNA detection currently encounters several limitations due to low ctDNA abundance. Recent studies have profiled paired PBLs and/or healthy control plasma to identify mutations derived from clonal hematopoiesis, a main contributor to false positive detection of ctDNA; however, the incorporation of orthogonal metrics may further improve accuracy and clinical applicability. Here, we evaluated the capabilities of multimodal genome-wide cell-free DNA profiling techniques for tumor-naive ctDNA detection within a cohort of HNSCC patients with low ctDNA abundance. We demonstrated a high degree of concordance between ctDNA metrics (abundance and fragment lengths) detected by mutation-based and methylation-based profiling methods. Moreover, we showed that tumor-naive multimodal ctDNA profiling may provide value by identifying putative prognostic biomarkers independent of ctDNA abundance, as well as by monitoring ctDNA abundance in serial samples.


Tumor-naïve detection of ctDNA has numerous practical advantages in both research and clinical settings. Recent studies have utilized matched tumor profiling for validation of identified ctDNA-derived regions at low abundance in early stage disease to improve sensitivity. However, one limitation of these approaches is the number of informative regions lost due to sampling heterogeneity of the tumor, which may be further exacerbated when applied to post-treatment ctDNA derived from previously unsampled sub-clones. Additionally, the clinical benefit of these tumor-informed detection methods is limited to cancers readily accessible by biopsy, circumventing one of the main strengths of non-invasive liquid biopsies. By utilizing a tumor-naïve multimodal profiling strategy, we achieved similar results in early stage cancers without the disadvantages of tumor-informed methods.


This is the first work to utilize mutation and methylation profiling for comprehensive detection of ctDNA from a cohort of localized cancer patients. Extending this multimodal profiling approach to other cancer types and disease settings will be important to the continued development of liquid biopsies. Additionally, while numerous ctDNA studies in HNSCC have been described utilizing detection methods based on mutation, methylation, or HPV profiling, here we described the first application of genome-wide mutation/methylation profiling methods identifying previously known targets (i.e. TP53 mutations or SEPT9/SHOX2 methylation) in addition to less-/non-investigated targets.


Tumor-naive detection of ctDNA has numerous practical advantages in both research and clinical settings. Although tumor mutational profiling may identify patient-specific markers for ctDNA detection at low abundance, such personalized approaches rely on high purity tumor samples from cancer types with sufficient mutational load. Mutational profiling for personalized assay design may be costly and time consuming, and it rarely accounts for genomic heterogeneity within primary tumors or across metastatic clones. Additionally, ctDNA detection methods that depend on access to tumor tissue diminish a key advantage of non-invasive liquid biopsies. By integrating independent cell-free DNA properties, we achieved sensitive ctDNA detection in early stage cancers without the disadvantages of tumor-informed methods.


In our analysis, we selected patients with detectable ctDNA by CAPP-Seq in order to identify ctDNA-derived methylation patterns using cfMeDIP-seq. This approach provided additional validation of the tumor-derived nature of plasma cell-free DNA in our cohort. The ctDNA methylation patterns were able to quantify ctDNA abundance in a similar manner to ctDNA mutations. In addition, methylation patterns revealed the tumor-of-origin and identified putative prognostic and dynamic biomarkers. The combination of CAPP-Seq and cfMeDIP-seq enabled an in-depth molecular characterization of low-abundance ctDNA. Mutation-based ctDNA quantification contributed to the discovery of HNSCC-specific hyper-DMRs in plasma, some of which were confirmed to be prognostic even after adjusting for ctDNA abundance. Thus, simultaneous profiling of mutations and methylation may complement one another by revealing quantitative, tissue-specific, and prognostic ctDNA biomarkers. Moreover, methylome profiling may prove particularly useful in cancer types with few recurrent or clonal mutations.


Similar to previous studies, we also observed a decreased in ctDNA fragment length compared to healthy donor cell-free DNA using both mutation- and methylation-based approaches. Unlike healthy cell-free DNA, which is consistently at ˜166-167 bp on average, the length of ctDNA between patients may be highly variable. Factors that influence ctDNA fragment length may include position-dependant fragmentation49, metastatic vs. non-metastatic disease73, as well as dysregulated kinetics of various intra/extracellular DNases responsible for healthy cell-free DNA fragmentation74. Interestingly, we observed high concordance between fragment lengths of ctDNA identified by CAPP-Seq and cfMeDIP-seq for eligible patients despite both techniques probing different regions and tumor-derived aberrations. These compelling data provide further evidence regarding the relevance and reproducibility of plasma cell-free DNA fragmentation in cancer patients.


We observed that detectable ctDNA by CAPP-Seq or elevated ctDNA abundance by cfMeDIP-seq, was associated with poor prognosis within our HNSCC cohort. These results are in accordance with previous HNSCC ctDNA studies, where detection of ctDNA by methylation56, as well as increased abundance by copy number aberrations75 or HPV detection76, identified high-risk patients. There was an imperfect association with tumor stage, suggesting that other unmeasured features of tumor biology may contribute to ctDNA abundance.


To our knowledge, no study has previously identified prognostic regions in HNSCC cell-free DNA independent of ctDNA detection/abundance, perhaps in part due to limitation of commonly used ctDNA detection methods. We demonstrated that cell-free DNA methylome profiles may serve as a discovery tool, which in conjunction with TCGA data, identified novel prognostic methylation biomarkers in HNSCC. A composite methylation score comprised of 5 DMRs demonstrated consistent prognostic associations across methylation detection platforms (hm450k and cfMeDIP-seq) and biospecimen types (tumor tissue and plasma cell-free DNA). Although future larger cohorts are needed to validate our findings, this study indicates that genome-wide identification of methylated regions by cfMeDIP-seq may enable discovery of novel prognostic biomarkers.


The performance of cfMeDIP-seq was evaluated in connection with disease prognosis. By applying a stringent threshold greater than ˜0.2% ctDNA post-treatment as detectable disease, we were able to predict disease recurrence for 4 out of 9 patients. For the remaining 5 patients that relapsed (n=4) or had persistent disease (n=1), who failed to have detectable ctDNA post-treatment, we observed typically longer times to recurrence suggesting that the fraction of ctDNA at those timepoints may have been below cfMeDIP-seq's lower limit of detection. In subsequent studies utilizing cfMeDIP-seq for tumor-naïve disease surveillance, more frequent plasma collection post-treatment may help address these limitations.


As we have demonstrated the potential clinical utility of multimodal profiling within localized disease and HNSCC, these methods contribute to future biomarker discovery and ultimately clinal utility for patients with a variety of cancer types. This study makes multiple notable contributions. It is the first to combine analyses of cell-free DNA mutations, methylation, and fragment lengths. Moreover, we methodically profiled plasma samples and paired PBLs from both HNSCC patients and risk-matched healthy controls. These analyses have revealed key insights regarding the optimal handling of multimodal profiling for ctDNA detection and characterization. For instance, our unique approaches to removing the contributing methylation signals from leukocytes and using fragment length characteristics to enrich for tumor-derived methylation will prove useful for future studies.


In conclusion, we demonstrate that tumor-naïve CAPP-Seq profiling of ctDNA enables high-confidence identification of ctDNA-derived methylation by cfMeDIP-seq. Utilizing the strength of epigenetic profiling by cfMeDIP-seq, we further show that these ctDNA-derived methylated regions demonstrate potential as markers of tumor-of-origin, prognosis, and treatment response. The incorporation of several approaches that we have described for improved sensitivity of ctDNA detection by cfMeDIP-seq in HNSCC, such as PBL-depleted windows and restriction of analysis to short fragments, may also be applied to various other localized cancers for clinical benefit. The disclosed framework are widely applicable to other clinical settings where tumor tissue availability may be limited.


Although preferred embodiments of the invention have been described herein, it will be understood by those skilled in the art that variations may be made thereto without departing from the spirit of the invention or the scope of the appended claims. All documents disclosed herein, including those in the following reference list, are incorporated by reference.

Claims
  • 1-54. (canceled)
  • 55. A method for processing a cell-free nucleic acid sample of a subject to determine whether said subject has or is at risk of having a disease, comprising: (a) providing said cell-free nucleic acid sample comprising a plurality of nucleic acid molecules;(b) subjecting said plurality of nucleic acid molecules or derivatives thereof to sequencing to generate a plurality of sequencing reads;(c) computer processing said plurality of sequencing reads to identify, for said plurality of nucleic acid molecules, (i) a methylation profile, (ii) a mutation profile, and (iii) a fragment length profile; and(d) using at least said methylation profile, said mutation profile and said fragment length profile to determine whether said subject has or is at risk of having said disease.
  • 56. The method of claim 55, wherein the disease comprises a cancer.
  • 57. The method of claim 56, wherein the cancer is selected from the group consisting of the cancer is selected from the group consisting of adrenal cancer, anal cancer, bile duct cancer, bladder cancer, bone cancer, brain/cns tumors, breast cancer, castleman disease, cervical cancer, colon/rectum cancer, endometrial cancer, esophagus cancer, ewing family of tumors, eye cancer, gallbladder cancer, gastrointestinal carcinoid tumors, gastrointestinal stromal tumor (gist), gestational trophoblastic disease, hodgkin disease, kaposi sarcoma, kidney cancer, laryngeal and hypopharyngeal cancer, leukemia (acute lymphocytic, acute myeloid, chronic lymphocytic, chronic myeloid, chronic myelomonocytic), liver cancer, lung cancer (non-small cell, small cell, lung carcinoid tumor), lymphoma, lymphoma of the skin, malignant mesothelioma, multiple myeloma, myelodysplastic syndrome, nasal cavity and paranasal sinus cancer, nasopharyngeal cancer, neuroblastoma, non-hodgkin lymphoma, oral cavity and oropharyngeal cancer, osteosarcoma, ovarian cancer, penile cancer, pituitary tumors, prostate cancer, retinoblastoma, rhabdomyosarcoma, salivary gland cancer, sarcoma—adult soft tissue cancer, skin cancer (basal and squamous cell, melanoma, merkel cell), small intestine cancer, stomach cancer, testicular cancer, thymus cancer, thyroid cancer, uterine sarcoma, vaginal cancer, vulvar cancer, waldenstrom macroglobulinemia, wilms tumor, squamous cell carcinoma, and head and neck squamous cell carcinoma.
  • 58. The method of claim 57, wherein the cancer is squamous cell carcinoma.
  • 59. The method of claim 58, wherein the cancer is head and neck squamous cell carcinoma.
  • 60. The method of claim 55, wherein said plurality of cell-free nucleic acid molecules comprises circulating tumor nucleic acid molecules.
  • 61. The method of claim 60, wherein the circulating tumor nucleic acid molecules comprises circulating tumor DNA.
  • 62. The method of claim 60, wherein the circulating tumor nucleic acid molecules comprises circulating tumor RNA.
  • 63. The method of claim 55, wherein said methylation profile comprises a plurality of Differentially Methylated Regions (DMRs).
  • 64. The method of claim 63, wherein said plurality of DMRs is ctDNA derived.
  • 65. The method of claim 63, wherein said plurality of DMRs derived from peripheral blood leukocytes is removed from said methylation profile.
  • 66. The method of claim 63, wherein said plurality of DMRs comprises at least about 56 genomic regions with hypo-methylation levels compared to corresponding genomic regions from a normal healthy subject.
  • 67. The method of claim 63, wherein said plurality of DMRs comprises at least about 941 genomic regions with hyper-methylation levels compared to corresponding genomic regions from a normal healthy subject.
  • 68. The method of claim 63, wherein a DMR of the plurality of DMRs comprises a size of at least about 300 bp.
  • 69. The method of claim 68, wherein said DMR comprises a size of at least about 100 bp to at least about 200 bp.
  • 70. The method of claim 68, wherein said DMR comprises a size of at least about 100 bp to at least about 150 bp.
  • 71. The method of claim 63, wherein a DMR of the plurality of DMRs comprises at least 8 CpG genomic islands.
  • 72. The method of claim 66, wherein said normal healthy subject comprises a same set of risk factors as said subject.
  • 73. The method of claim 55, wherein said mutation profile comprises a missense variant, a nonsense variant, a deletion variant, an insertion variant, a duplication variant, an inversion variant, a frameshift variant, or a repeat expansion variant.
  • 74. The method of claim 55, wherein any variant that is present in a genomic DNA sample obtained from a plurality of peripheral blood leukocytes, wherein said plurality of peripheral blood leukocytes is obtained from said subject, is removed from said mutation profile.
  • 75. The method of claim 55, wherein any variant that is derived from clonal hematopoiesis is removed from said mutation profile.
  • 76. The method of claim 75, wherein said mutation profile does not comprise a variant of gene DNMT3A, TET2, or ASXL1.
  • 77. The method of claim 75, wherein said mutation profile does not comprise a canonical cancer driver gene.
  • 78. The method of claim 75, wherein said mutation profile comprises a non-canonical cancer driver gene, where said non-canonical cancer driver gene is GRIN3A or MYC.
  • 79. The method of claim 55, wherein said fragment length profile comprises selecting cell free nucleic acid molecules based on a range of fragment length of about at least 80 bp to 170 bp.
  • 80. The method of claim 55, wherein said fragment length profile comprises selecting cell free nucleic acid molecules based on a range of fragment length of about at least 100 bp to 150 bp.
  • 81. The method of claim 60, wherein said circulating tumor nucleic acid molecules are enriched.
  • 82. The method of claim 55, further comprising mixing said cell free nucleic acid sample with a filler DNA molecules to yield a DNA mixture.
  • 83. The method of claim 82, wherein said filler DNA molecules comprise a length of about 50 bp to 800 bp.
  • 84. The method of claim 82, wherein said filler DNA molecules comprise a length of about 100 bp to 600 bp.
  • 85. The method of claim 82, wherein said filler DNA molecules comprises at least about 5% methylated filler DNA molecules.
  • 86. The method of claim 82, wherein said filler DNA molecules comprises at least about 20% methylated filler DNA.
  • 87. The method of claim 82, wherein said filler DNA molecules comprises at least about 30% methylated filler DNA.
  • 88. The method of claim 82, wherein said filler DNA molecules comprises at least about 50% methylated filler DNA.
  • 89. The method of claim 55, further comprising incubating said DNA mixture with a binder that is configured to bind methylated nucleotides to generate an enriched sample.
  • 90. The method of claim 89, wherein said binder comprises a protein comprising a methyl-CpG-binding domain.
  • 91. The method of claim 90, wherein said protein is a MBD2 protein.
  • 92. The method of claim 89, wherein said binder comprises an antibody.
  • 93. The method of claim 92, wherein said antibody is a 5-MeC antibody.
  • 94. The method of claim 9, wherein said antibody is a 5-hydroxymethyl cytosine antibody.
  • 95. The method of claim 55, wherein said determining comprises sequencing said plurality of sequencing reads, and wherein said sequencing does not comprise bisulfite sequencing.
  • 96. The method of claim 55, wherein said cell-free nucleic acid sample comprises a blood sample.
  • 97. The method of claim 96, wherein said blood sample comprises a plasma sample.
  • 98. The method of claim 55, further comprising detecting an origin of cancer tissue.
  • 99. The method of claim 55, further comprising generating a report comprising a prognosis of said subject's survival rate.
  • 100. The method of claim 55, further comprising providing a treatment to said subject.
  • 101. The method of claim 100, subsequent to treatment of said disease, further comprising providing a second report indicating whether said treatment is effective.
  • 102. A method for determining whether a subject has or is at risk of having a condition, comprising: (a) assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject;(b) detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 5; and(c) comparing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 5.
  • 103. The method of claim 102, wherein said cell-free nucleic acid molecule comprises ctDNA.
  • 104. The method of claim 102, wherein said assaying comprises performing a sequence analysis, and wherein said sequence analysis comprises cell-free methylated DNA immunoprecipitation (cfMeDIP) sequencing.
  • 105. The method of claim 102, wherein said detecting comprises measuring a methylation level of at least a portion of said nucleic acid molecule comprised in: six or more, ten or more, fifteen or more, twenty or more, thirty or more, forty or more, fifty or more, sixty or more, seventy or more, eighty or more, ninety or more, or one hundred or more DMRs listed in Table 5.
  • 106. A method for determining whether a subject has a higher survival rate after receiving a treatment for a disease, comprising: (a) assaying a cell-free nucleic acid molecule from at least a portion of a sample from said subject;(b) detecting a methylation level of at least a portion of said cell-free nucleic acid molecule comprised in a differentially methylated region (DMR) listed in Table 6; and(c) processing, using at least one computer processor, said methylation level detected in (b) to a methylation level of corresponding portion(s) of said cell-free nucleic acid molecules comprised in said DMR listed in Table 6.
  • 107. The method of claim 106, wherein said cell-free nucleic acid molecule comprises ctDNA.
  • 108. The method of claim 106, wherein said detecting comprises providing a composite methylation score (CMS).
  • 109. The method of claim 107, wherein said CMS comprises a sum of beta-values of DMRs listed in Table 6.
  • 110. The method of claim 107, wherein a higher CMS indicates an inferior survival for said subject.
  • 111. The method of claim 107, wherein said CMS is not dependent on an abundance of ctDNA.
  • 112. The method of claim 106, wherein said disease is squamous cell carcinoma.
  • 113. The method of claim 112, wherein the squamous cell carcinoma is head and neck squamous cell carcinoma.
  • 114. The method of claim 106, further comprising selecting cell free nucleic acid molecules based on a range of fragment length of about at least 80 bp to 170 bp.
  • 115-116. (canceled)
CROSS REFERENCES

This application is a continuation application of International Application No. PCT/CA2021/050842, filed Jun. 8, 2021, which claims the benefit of U.S. provisional patent application No. 63/041,151, filed Jun. 19, 2020, which are each entirely incorporated herein by reference.

Provisional Applications (1)
Number Date Country
63041151 Jun 2020 US
Continuations (1)
Number Date Country
Parent PCT/CA2021/050842 Jun 2021 US
Child 18067661 US