# US 12,009,061 B2 · Computational filtering of methylated sequence data for predictive modeling > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,009,061 B2 - Document ID: US12009061B2 - Status: Granted - Kind: B2 - Application number: 16/846,238 - Filed: 2020-04-10 - Published: 2020-10-15 - Granted: 2024-06-11 - Earliest claimed: 2019-04-10 - Assignees: University of Pittsburgh—Of the Commonwealth System of Higher Education - Applicants: University of Pittsburgh—Of the Commonwealth System of Higher Education - Inventors: David Gerard Peters; Tianjiao Chu; Lisa Ann Pan; David N. Finegold - Web page: https://patentgrep.com/patent/US12009061B2 ## CPC - G16B 30/00: ICT specially adapted for sequence analysis involving nucleotides or amino acids - G16B 5/00: ICT specially adapted for modelling or simulations in systems biology, e.g. gene-regulatory networks, protein interaction networks or metabolic networks - G16B 20/00: ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations - G16B 40/20: Supervised data analysis - G16B 50/20: Heterogeneous data integration - G16B 50/30: Data warehousing; Computing architectures - G16H 10/40: for data related to laboratory analysis, e.g. patient specimen analysis - G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems - G16H 50/30: for calculating health indices; for individual health risk assessment - G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY - G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA ## Abstract Computational techniques are disclosed for using methylation profiles to classify the medication condition of a person. Initial sequence data is obtained containing sequences of an initial set of nucleic acids from a biological sample of a person. The initial sequence data is filtered to generate filtered sequence data that describes sequences of a filtered subset of nucleic acids from the biological sample. A methylation profile is determined for the filtered subset of nucleic acids from the biological sample. The methylation profile can be processed to determine a likelihood that the person has the specified medical condition. The system outputs an indication of the likelihood that the person has the specified medical condition. ## Claims ### Claim 1 A method, comprising: obtaining initial sequence data that describes sequences of an initial set of nucleic acids from a biological sample of a person, the initial set of nucleic acids including nucleic acids originating from a plurality of different tissues of the person; filtering the initial sequence data to identify a first subset of nucleic acids whose sequences correspond to sequences in a first set of genomic regions and a second subset of nucleic acids whose sequences correspond to sequences in a second set of genomic regions; determining an observed methylation profile for the first subset of nucleic acids of the person and an observed methylation profile for the second subset of nucleic acids of the person; determining a predicted methylation profile for the second subset of nucleic acids of the person based on the observed methylation profile for the first subset of nucleic acids of the person; determining a likelihood that the person has a specified medical condition, including (i) determining a difference between the predicted methylation profile for the second subset of nucleic acids of the person and the observed methylation profile for the second subset of nucleic acids of the person and (ii) determining whether the difference matches a predefined pattern of deviation for the specified medical condition; and initiating a treatment of the person for the specified medical condition based on the determined likelihood that the person has the specified medical condition, the treatment including administering a therapeutic agent to the person at a dosage level or a dosage interval that is set based on a methylation level of a set of genomic loci associated with the specified medical condition, wherein (i) the specified medical condition is a cancer and the therapeutic agent comprises at least one of an immunotherapy or a chemotherapy, (ii) the specified medical condition is endometriosis and the therapeutic agent is a hormone therapy, (iii) the specified medical condition is necrotizing enterocolitis and the therapeutic agent is an antibiotic therapy, or (iv) the specified medical condition is preeclampsia and the therapeutic agent is an anti-hypertensive medication. ### Claim 2 (depends on 1) The method of claim 1, wherein filtering the initial sequence data to identify the first subset of nucleic acids comprises: comparing nucleic acid sequences from the initial set of nucleic acids to the sequences in the first set of genomic regions; discarding nucleic acid sequences from the initial sequence data that are not determined to match the sequences in the first set of genomic regions; and retaining nucleic acid sequences from the initial sequence data that are determined to match the sequences in the first set of genomic regions. ### Claim 3 (depends on 2) The method of claim 2, wherein at least a first subset of the first set of genomic regions are defined based on the regions in the first subset exhibiting at least a minimum level of stability with respect to at least one of the methylation characteristic or the copy number characteristic in a population of individuals. ### Claim 4 (depends on 3) The method of claim 3, wherein the second set of genomic regions are defined based on the regions in the second set exhibiting at least a minimum difference with respect to the methylation characteristic or the copy number characteristic between individuals who have the specified medical condition and individuals who do not have the specified medical condition. ### Claim 5 (depends on 1) The method of claim 1, wherein the biological sample comprises plasma, and the initial set of nucleic acids comprises cell-free DNA in the plasma. ### Claim 6 (depends on 1) The method of claim 1, comprising: identifying a set of reference component methylomes in the initial set of nucleic acids and a filtered subset of nucleic acids; determining a proportion of the reference component methylomes at a reference set of CpG sites in the initial set or the filtered subset of nucleic acids; generating predictions of methylation levels at a target set of CpG sites in the initial set or the filtered subset of nucleic acids; comparing the predictions of methylation levels at the target set of CpG sites to observed methylation levels; and determining whether the person likely has or does not have the specified medical condition based on the comparison. ### Claim 7 (depends on 1) The method of claim 1, wherein the biological sample comprises a stool sample or cerebrospinal fluid. ### Claim 8 (depends on 1) The method of claim 1, further comprising: determining a copy number profile for a filtered subset of nucleic acids from the biological sample; and processing the copy number profile a methylation profile for the filtered subset of nucleic acids to determine the likelihood that the person has the specified medical condition. ### Claim 9 (depends on 1) The method of claim 1, wherein the initial set of nucleic acids were treated to facilitate detection of methylated sites before sequencing. ### Claim 10 (depends on 1) The method of claim 1, wherein the observed methylation profile for the first or second subsets of sequences indicates, for each of a plurality of genomic loci, a methylation level of the locus. ### Claim 11 (depends on 1) The method of claim 1, wherein the genomic loci is a CpG site, CpG island, differentially methylated region (DMR), promoter region, enhancer region, or CpG island shore. ### Claim 12 (depends on 1) The method of claim 1, wherein determining the likelihood that the person has the specified medical condition comprises determining a probability that the person has the specified medical condition. ### Claim 13 (depends on 1) The method of claim 1, wherein determining the likelihood that the person has the specified medical condition comprises generating a binary indication that the person either likely has the specified medical condition or likely does not have the specified medical condition. ### Claim 14 (depends on 1) The method of claim 1, comprising providing data representing the observed methylation profile for the second subset of nucleic acids as input to a machine-learning model. ### Claim 15 (depends on 14) The method of claim 14, wherein the machine-learning model comprises at least one of a classifier, an artificial neural network, a support vector machine, a decision tree, or a regression model. ### Claim 16 (depends on 14) The method of claim 14, wherein the machine-learning model is used to derive the predicted methylation profile for the second subset of nucleic acids is based on the observed methylation profile for the first subset of nucleic acids, wherein the machine-learning model comprises at least one of an artificial neural network, a regression tree, or a regression model. ### Claim 17 (depends on 1) The method of claim 1, wherein correlating the difference between the predicted methylation profile for the second subset of nucleic acids and the observed methylation profile for the second subset of nucleic acids with the likelihood that the person has the specified medical condition comprises comparing the difference to a minimum difference criterion. ### Claim 18 (depends on 1) The method of claim 1, wherein the predefined pattern of deviation differs for different specified medical conditions. ### Claim 19 (depends on 1) The method of claim 1, wherein administering the therapeutic agent to the person at the dosage level or the dosage interval that is set based on the methylation level of the set of genomic loci associated with the specified medical condition comprises: administering the therapeutic agent to the person at a first dosage level or a first dosage interval; monitoring a change in the methylation level of the set of genomic loci associated with the specified medical condition after administration of the therapeutic agent at the first dosage level or the first dosage interval; and adjusting the administration of the therapeutic agent from the first dosage level or the first dosage interval to a second dosage level or a second dosage interval based on the change in the methylation level of the set of genomic loci associated with the specified medical condition. ### Claim 20 (depends on 1) The method of claim 1, wherein the dosage level is set based on the methylation level of the set of genomic loci associated with the specified medical condition. ### Claim 21 (depends on 1) The method of claim 1, wherein the specified medical condition is the cancer. ### Claim 22 (depends on 21) The method of claim 21, wherein the cancer is ovarian cancer. ### Claim 23 (depends on 1) The method of claim 1, wherein the specified medical condition is endometriosis. ### Claim 24 (depends on 1) The method of claim 1, wherein the specified medical condition is necrotizing enterocolitis. ### Claim 25 (depends on 1) The method of claim 1, wherein the specified medical condition is preeclampsia. ### Claim 26 A computing system, comprising: one or more processors; and one or more non-transitory computer-readable media having instructions stored thereon that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: obtaining initial sequence data that describes sequences of an initial set of nucleic acids from a biological sample of a person, the initial set of nucleic acids including nucleic acids originating from a plurality of different tissues of the person; filtering the initial sequence data to identify a first subset of nucleic acids whose sequences correspond to sequences in a first set of genomic regions and a second subset of nucleic acids whose sequences correspond to sequences in a second set of genomic regions; determining an observed methylation profile for the first subset of nucleic acids of the person and an observed methylation profile for the second subset of nucleic acids of the person; determining a predicted methylation profile for the second subset of nucleic acids of the person based on the observed methylation profile for the first subset of nucleic acids of the person; determining a likelihood that the person has a specified medical condition, including (i) determining a difference between the predicted methylation profile for the second subset of nucleic acids of the person and the observed methylation profile for the second subset of nucleic acids of the person and (ii) determining whether the difference matches a predefined pattern of deviation for the specified medical condition wherein a treatment of the person for the specified medical condition is initiated based on the determined likelihood that the person has the specified medical condition, the treatment including administering a therapeutic agent to the person at a dosage level or a dosage interval that is set based on a methylation level of a set of genomic loci associated with the specified medical condition, wherein (i) the specified medical condition is a cancer and the therapeutic agent comprises at least one of an immunotherapy or a chemotherapy, (ii) the specified medical condition is endometriosis and the therapeutic agent is a hormone therapy, (iii) the specified medical condition is necrotizing enterocolitis and the therapeutic agent is an antibiotic therapy, or (iv) the specified medical condition is preeclampsia and the therapeutic agent is an anti-hypertensive medication. ### Claim 27 One or more non-transitory computer-readable media having instructions stored thereon that, when executed by one or more processors, cause performance of operations comprising: obtaining initial sequence data that describes sequences of an initial set of nucleic acids from a biological sample of a person, the initial set of nucleic acids including nucleic acids originating from a plurality of different tissues of the person; filtering the initial sequence data to identify a first subset of nucleic acids whose sequences correspond to sequences in a first set of genomic regions and a second subset of nucleic acids whose sequences correspond to sequences in a second set of genomic regions; determining an observed methylation profile for the first subset of nucleic acids of the person and an observed methylation profile for the second subset of nucleic acids of the person; determining a predicted methylation profile for the second subset of nucleic acids of the person based on the observed methylation profile for the first subset of nucleic acids of the person; determining a likelihood that the person has a specified medical condition, including (i) determining a difference between the predicted methylation profile for the second subset of nucleic acids of the person and the observed methylation profile for the second subset of nucleic acids of the person and (ii) determining whether the difference matches a predefined pattern of deviation for the specified medical condition; initiating a treatment of the person for the specified medical condition based on the determined likelihood that the person has the specified medical condition, the treatment including administering a therapeutic agent to the person at a dosage level or a dosage interval that is set based on a methylation level of a set of genomic loci associated with the specified medical condition, wherein (i) the specified medical condition is a cancer and the therapeutic agent comprises at least one of an immunotherapy or a chemotherapy, (ii) the specified medical condition is endometriosis and the therapeutic agent is a hormone therapy, (iii) the specified medical condition is necrotizing enterocolitis and the therapeutic agent is an antibiotic therapy, or (iv) the specified medical condition is preeclampsia and the therapeutic agent is an anti-hypertensive medication. ### Claim 28 A method, comprising: obtaining initial sequence data that describes sequences of an initial set of nucleic acids from a biological sample of a person, the initial set of nucleic acids including nucleic acids originating from a plurality of different tissues of the person; filtering the initial sequence data to identify a first subset of sequences from the initial sequence data that correspond to a first pre-defined set of genomic regions; filtering the initial sequence data to identify a second subset of sequences from the initial sequence data that correspond to a second pre-defined set of genomic regions; processing data that includes an observed methylation profile of the first subset of sequences to generate a predicted methylation profile of the second subset of sequences; determining a difference between an observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences; determining whether the difference between the observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences matches a predefined pattern of deviation for a specified medical condition, wherein the person is deemed to have the specified medical condition if the difference between the observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences matches the predefined pattern of deviation for the specified medical condition; and initiating a treatment of the person for the specified medical condition based on the determined likelihood that the person has the specified medical condition, the treatment including administering a therapeutic agent to the person at a dosage level or a dosage interval that is set based on a methylation level of a set of genomic loci associated with the specified medical condition, wherein (i) the specified medical condition is a cancer and the therapeutic agent comprises at least one of an immunotherapy or a chemotherapy, (ii) the specified medical condition is endometriosis and the therapeutic agent is a hormone therapy, (iii) the specified medical condition is necrotizing enterocolitis and the therapeutic agent is an antibiotic therapy, or (iv) the specified medical condition is preeclampsia and the therapeutic agent is an anti-hypertensive medication. ### Claim 29 (depends on 28) The method of claim 28, wherein: the first pre-defined set of genomic regions are regions that exhibit at least a minimum level of stability with respect to at least one of a methylation characteristic or a copy number characteristic in a population of individuals; and the second pre-defined set of genomic regions are regions that exhibit at least a minimum difference with respect to at least one of the methylation characteristic or the copy number characteristic between a first sub-population of individuals who have the specified medical condition and a second sub-population of individuals who do not have the specified medical condition. ### Claim 30 (depends on 28) The method of claim 28, wherein the first pre-defined set of genomic regions is a first reference set of genomic regions, and the second pre-defined set of genomic regions is a first target set of genomic regions; the method further comprising: selecting the first reference set of genomic regions as the first pre-defined set of genomic regions from a database that includes a plurality of reference sets of genomic regions, wherein different ones of the plurality of reference sets of genomic regions correspond to different medical conditions; and selecting the first target set of genomic regions as the second pre-defined set of genomic regions from the database, wherein the database further includes a plurality of target sets of genomic regions, wherein different ones of the plurality of target sets of genomic regions correspond to different medical conditions. ### Claim 31 One or more non-transitory computer-readable media having instructions stored thereon that, when executed by one or more processors, cause performance of operations comprising: obtaining initial sequence data that describes sequences of an initial set of nucleic acids from a biological sample of a person, the initial set of nucleic acids including nucleic acids originating from a plurality of different tissues of the person; filtering the initial sequence data to identify a first subset of sequences from the initial sequence data that correspond to a first pre-defined set of genomic regions; filtering the initial sequence data to identify a second subset of sequences from the initial sequence data that correspond to a second pre-defined set of genomic regions; processing data that includes an observed methylation profile of the first subset of sequences to generate a predicted methylation profile of the second subset of sequences; determining a difference between an observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences; determining whether the difference between the observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences matches a predefined pattern of deviation for a specified medical condition, wherein the person is deemed to have the specified medical condition if the difference between the observed methylation profile of the second subset of sequences and the predicted methylation profile of the second subset of sequences matches the predefined pattern of deviation for the specified medical condition; wherein a treatment of the person for the specified medical condition is initiated based on the determined likelihood that the person has the specified medical condition, the treatment including administering a therapeutic agent to the person at a dosage level or a dosage interval that is set based on a methylation level of a set of genomic loci associated with the specified medical condition, wherein (i) the specified medical condition is a cancer and the therapeutic agent comprises at least one of an immunotherapy or a chemotherapy, (ii) the specified medical condition is endometriosis and the therapeutic agent is a hormone therapy, (iii) the specified medical condition is necrotizing enterocolitis and the therapeutic agent is an antibiotic therapy, or (iv) the specified medical condition is preeclampsia and the therapeutic agent is an anti-hypertensive medication.