Methods and systems for microsatellite analysis
AssigneeOrbit Genomics, Inc.
InventorHarold GARNER
Life of this patent
Estimating the term…
Abstract
The present disclosure provides methods and systems for classifying microsatellite and minor alleles in a sample. Also, the present disclosure provides methods and systems for generating classifiers for conditions based on microsatellite loci and for performing pan-cancer assays. The methods and systems can involve next-generation sequencing of nucleic acid samples from subjects and genotyping microsatellite loci in the samples.
Claims24
A computer-implemented method for constructing an optimized classifier for a condition, the method comprising ranking subsets of a plurality of microsatellites as classifiers for the condition in a plurality of optimization cycles, wherein the subsets of the plurality of microsatellites comprise microsatellites in an initial population of microsatellites correlated with the condition, thereby identifying an optimized subset of the subsets of the microsatellites as the optimized classifier for the condition,
- wherein the ranking comprises:
- (i) performing a receiver operating characteristic (ROC) analysis with (1) the subsets of the plurality of microsatellites, and (2) microsatellites in reference samples from subjects with the condition and without the condition;
- and
- (ii) adding a sensitivity and a specificity of the classifier for the condition for each subset of the plurality of microsatellites;
- wherein an optimization cycle of the plurality of optimization cycles comprises adding new subsets of the initial population of microsatellites to subsets from a previous optimization cycle of the plurality of optimization cycles, wherein a first portion of the new subsets is generated by splitting and recombining chosen subsets from the previous optimization cycle, and a second portion of the new subsets is generated by selecting microsatellites from the initial population of microsatellites.
(canceled)
(canceled)
(canceled)
16-61. (canceled)
A computer system configured to construct an optimized classifier for a condition, comprising:
- a memory comprising computer-executable instructions;
- and
- one or more processors collectively or individually programmed to execute the computer-executable instructions, wherein the computer-executable instructions comprise:
- ranking subsets of a plurality of microsatellites as classifiers for the condition in a plurality of optimization cycles, wherein the subsets of the plurality of microsatellites comprise microsatellites in an initial population of microsatellites correlated with the condition, thereby identifying an optimized subset of the subsets of the microsatellites as the optimized classifier for the condition,
- wherein the ranking comprises:
- (i) performing an receiver operating characteristic (ROC) analysis with (1) the subsets of microsatellites, and (2) microsatellites in reference samples from subjects with the condition and without the condition;
- and
- (ii) adding a sensitivity and a specificity of the classifier for the condition for each subset of the plurality of microsatellites;
- and
- wherein an optimization cycle of the plurality of optimization cycles comprises adding new subsets of the initial population of microsatellites to subsets from a previous optimization cycle of the plurality of optimization cycles, wherein a first portion of the new subsets are generated by splitting and recombining chosen subsets from the previous optimization cycle, and a second portion of the new subsets are generated by selecting microsatellites from the initial population of microsatellites.
A non-transitory computer-readable medium comprising instruction that, when executed by a processor of a processing system, cause the processing system to perform a method of constructing an optimized classifier for a condition, the method comprising ranking subsets of a plurality of microsatellites as classifiers for the condition in a plurality of optimization cycles, wherein the subsets of the plurality of microsatellites comprise microsatellites in an initial population of microsatellites correlated with the condition, thereby identifying an optimized subset of the subsets of the microsatellites as the optimized classifier for the condition,
- wherein the ranking comprises:
- (i) performing a receiver operating characteristic (ROC) analysis with (1) the subsets of microsatellites, and (2) microsatellites in reference samples from subjects with the condition and without the condition;
- and
- (ii) adding a sensitivity and a specificity of the classifier for the condition for each subset of the plurality of microsatellites;
- and
- wherein an optimization cycle of the plurality of optimization cycles comprises adding new subsets of the initial population of microsatellites to subsets from a previous optimization cycle of the plurality of optimization cycles, wherein a first portion of the new subsets is generated by splitting and recombining chosen subsets from the previous optimization cycle, and a second portion of the new subsets is generated by selecting microsatellites from the initial population of microsatellites.
Dependencies are inferred from the claim text.