Individual and cohort pharmacological phenotype prediction platform

• US 10,553,318 B2
• Filed: 02/05/2019
• Issued: 02/04/2020
• Est. Priority Date: 05/12/2017
• Status: Active Grant

• Current Assignee: Board of Regents of the University of Michigan (University of Michigan)
• Original Assignee: Board of Regents of the University of Michigan (University of Michigan)
• Inventors: Athey, Brian D., Allyn-Feuer, Ari, Higgins, Gerald A., Burns, James S., Kalinin, Alexandr, Pauls, Brian, Ade, Alex, Reamaroon, Narathip
• Primary Examiner(s): Pope, Daryl C
• Application Number:
  US16/267,546
• Publication Number:
  US 20190172584A1
• Time in Patent Office: 364 Days
• Field of Search: 340506, 340 31, 3405391, 34053911, 34053913, 34053912, 3405731
• CPC Class Codes: A61K 31/37 Coumarins, e.g. psoralen
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 20/20 Allele or variant detection, e.g. single nucleotide polymorphism [SNP] detection
• CPC Class Codes: G16B 30/00 ICT specially adapted for sequence analysis involving nucleotides or amino acids
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/20 Supervised data analysis
• CPC Class Codes: G16B 40/30 Unsupervised data analysis
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

For patients who exhibit or may exhibit primary or comorbid disease, pharmacological phenotypes may be predicted through the collection of panomic data over a period of time. A machine learning engine may generate a statistical model based on training data from training patients to predict pharmacological phenotypes, including drug response and dosing, drug adverse events, disease and comorbid disease risk, drug-gene, drug-drug, and polypharmacy interactions. Then the model may be applied to data for new patients to predict their pharmacological phenotypes, and enable decision making in clinical and research contexts, including drug selection and dosage, changes in drug regimens, polypharmacy optimization, monitoring, etc., to benefit from additional predictive power, resulting in adverse event and substance abuse avoidance, improved drug response, better patient outcomes, lower treatment costs, public health benefits, and increases in the effectiveness of research in pharmacology and other biomedical fields.

Claims

1. A computer-implemented method for identifying pharmacological phenotypes, the method executed by one or more processors programmed to perform the method, the method comprising:

identifying a plurality of single nucleotide polymorphisms (SNPs) correlated with a particular set of pharmacological phenotypes;

comparing the plurality of SNPs to a database of SNPs to identify additional SNPs that are linked to the plurality of SNPs, wherein the plurality of SNPs and additional SNPs are included in a set of permissive candidate variants;

performing, by one or more processors, a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants based on at least one of;

regulatory function, variant dependence, a presence of target gene relationships for the permissive candidate variants, or whether the permissive candidate variants are non-synonymous coding variants with a minor allele frequency;

performing, by the one or more processors, a network analysis to filter the subset of intermediate candidate variants into candidate variants of SNPs, genes associated with the SNPs, or networks associated with the SNPs which are causally related to the particular set of pharmacological phenotypes;

obtaining a biological sample of a patient;

comparing the biological sample to the SNPs which are causally related to the particular set of pharmacological phenotypes; and

providing, by the one or more processors, an indication of one or more pharmacological phenotypes for the patient for display to a health care provider based on the comparison.

2. The method of claim 1, wherein performing a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants includes:

evaluating, by the one or more processors, genomic regions around the set of permissive candidate variants for regulatory function.

3. The method of claim 2, wherein genomic regions around the set of permissive candidate variants are evaluated for regulatory function to at least one of:

(1) determine whether sequence contexts for the set of permissive candidate variants influence the regulatory function, or (2) determine target genes for the set of permissive candidate variants.

4. The method of claim 3, wherein target genes for the set of permissive candidate variants are determined using machine learning techniques.

5. The method of claim 1, wherein performing a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants includes:

scoring, by the one or more processors, each of the permissive candidate variants based on at least one of;

regulatory function, variant dependence, a presence of target gene relationships for the permissive candidate variants, or whether the permissive candidate variants are non-synonymous coding variants with a minor allele frequency; and

assigning, by the one or more processors, one or more of the permissive candidate variants that are scored above a threshold score to the subset of intermediate candidate variants.

6. The method of claim 1, wherein the particular set of pharmacological phenotypes includes a set of warfarin phenotypes, and the identified plurality of SNPs which are causally related to the set of warfarin phenotypes include a warfarin response pathway.

7. The method of claim 1, wherein providing the one or more pharmacological phenotypes for the patient for display to a health care provider includes:

generating, by the one or more processors, a risk analysis display for the patient including at least one of;

a predicted response by the patient to each of one or more drugs, a risk of one or more illnesses, or a likelihood of substance abuse.

8. The method of claim 1, wherein the one or more pharmacological phenotypes include at least one of:

a predicted response by second patient to one or more drugs, a risk of one or more illnesses, an adverse drug event or adverse drug reaction to the one or more drugs, or a likelihood of substance abuse.

9. A computing device for identifying pharmacological phenotypes using statistical modeling and machine learning techniques, the computing device comprising:

a communication network,

one or more processors; and

a non-transitory computer-readable memory coupled to the one or more processors and storing thereon instructions that, when executed by the one or more processors, cause the computing device to;

identify a plurality of single nucleotide polymorphisms (SNPs) correlated with a particular set of pharmacological phenotypes;

compare the plurality of SNPs to a database of SNPs to identify additional SNPs that are linked to the plurality of SNPs, wherein the plurality of SNPs and additional SNPs are included in a set of permissive candidate variants;

perform a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants based on at least one of;

regulatory function, variant dependence, a presence of target gene relationships for the permissive candidate variants, or whether the permissive candidate variants are non-synonymous coding variants with a minor allele frequency;

perform a network analysis to filter the subset of intermediate candidate variants into candidate variants of SNPs, genes associated with the SNPs, or networks associated with the SNPs which are causally related to the particular set of pharmacological phenotypes;

obtain a biological sample of a patient;

compare the biological sample to the SNPs which are causally related to the particular set of pharmacological phenotypes; and

provide, via the communication network, an indication of one or more pharmacological phenotypes for the patient for display to a health care provider based on the comparison.

10. The computing device of claim 9, wherein to perform a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants, the instructions cause the computing device to:

evaluate genomic regions around the set of permissive candidate variants for regulatory function.

11. The computing device of claim 10, wherein genomic regions around the set of permissive candidate variants are evaluated for regulatory function to at least one of:

(1) determine whether sequence contexts for the set of permissive candidate variants influence the regulatory function, or (2) determine target genes for the set of permissive candidate variants.

12. The computing device of claim 11, wherein target genes for the set of permissive candidate variants are determined using machine learning techniques.

13. The computing device method of claim 9, wherein to perform a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants, the instructions cause the computing device to:

score each of the permissive candidate variants based on at least one of;

regulatory function, variant dependence, a presence of target gene relationships for the permissive candidate variants, or whether the permissive candidate variants are non-synonymous coding variants with a minor allele frequency; and

assign one or more of the permissive candidate variants that are scored above a threshold score to the subset of intermediate candidate variants.

14. The computing device of claim 9, wherein the particular set of pharmacological phenotypes includes a set of warfarin phenotypes, and the identified plurality of SNPs which are causally related to the set of warfarin phenotypes include a warfarin response pathway.

15. The computing device of claim 9, wherein to provide the one or more pharmacological phenotypes for the patient for display to a health care provider, the instructions cause the computing device to:

generate a risk analysis display for the patient including at least one of;

a predicted response by the patient to each of one or more drugs, a risk of one or more illnesses, or a likelihood of substance abuse.

16. A method for treating a patient comprising:

obtaining a biological sample of a patient;

comparing or having compared the biological sample to a set of single nucleotide polymorphisms (SNPs) within a warfarin response pathway having a plurality of SNPs each of which are causally related to warfarin response phenotypes;

determining a dosage of warfarin for administering to the patient based on the comparison; and

administering the determined dosage of warfarin to the patient.

17. The method of claim 16, wherein the one or more SNPs in the warfarin response pathway are identified by:

identifying a plurality of SNPs correlated with warfarin response phenotypes;

comparing the plurality of SNPs to a database of SNPs to identify additional SNPs that are linked to the plurality of SNPs, wherein the plurality of SNPs and additional SNPs are included in a set of permissive candidate variants;

performing a bioinformatics analysis to filter the set of permissive candidate variants into a subset of intermediate candidate variants based on at least one of;

regulatory function, variant dependence, a presence of target gene relationships for the permissive candidate variants, or whether the permissive candidate variants are non-synonymous coding variants with a minor allele frequency; and

performing a pathway analysis to filter the subset of intermediate candidate variants into candidate variants of SNPs, genes associated with the SNPs, or pathways associated with the SNPs which are causally related to the warfarin response phenotypes.

18. The method of claim 16, wherein the warfarin response phenotypes include at least one of:

a warfarin efficacy phenotype, an adverse warfarin response phenotype, a warfarin dose phenotype, hemostatic factors and hematological phenotypes, hemorrhaging end-stage coagulation, or thrombin generation potential phenotypes.

19. The method of claim 16, further comprising:

comparing or having compared SNPs in the biological sample to the set of SNPs within the warfarin response pathway; and

assigning scores to the SNPs in the biological sample that match with the set of SNPs within the warfarin response pathway; and

combining the assigned scores to determine the dosage for administering to the patient.

20. The method of claim 16, wherein the warfarin response pathway includes one or more of:

complement factor H related 5 (CFHR5) gene, fibrinogen alpha chain gene (FGA) gene, flavin containing monoxygenase 5 (FMO5) gene, histidine rich glycoprotein (HRG) gene, kininogen 1 (KNG1) gene, surfeit 4 (SURF4) gene, alpha 1-3-N-acetylgalactosaminyltransferase and alpha 1-3-galactosyltransferase (ABO) gene, lysozyme (LYZ) gene, polycomb group ring finger 3 (PCGF3) gene, serine protease 8 (PRSS8) gene, transient receptor potential cation channel subfamily C member 4 associated pattern (TRPC4AP) gene, solute carrier family 44 member 2 (SLC44A2) gene, sphingosine kinase 1 (SPHK1) gene, or ubiquitin specific peptidase 7 (USP7) gene.