Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12217830-B2
Publication Date
2025-02-04
Expiration Date
Abstract
The disclosure provides methods for estimating tumor purity from tumor samples without use of matched-normal controls. A set of genomic regions are identified based on a nucleic acid sequence data that is aligned to a reference genome. Each genomic region of the set of genomic regions includes one or more nucleotide-sequence variants relative to a corresponding genomic region of the reference genome. A B-allele frequency distribution for the biological sample is determined based on a B-allele frequency determined for each genomic region of the set of genomic regions. The B-allele frequency distribution is processed using a trained machine-learning model to estimate a metric identifying tumor purity in the biological sample.
Core Innovation
The invention determines tumor purity of a biological sample of a subject for informing a cancer feature and evaluating a treatment efficacy. The method obtains nucleic acid sequence data from one or more sequencers that represent a plurality of nucleic acid molecules of the biological sample, aligns the sequence data to a reference genome, and identifies a set of genomic regions containing one or more nucleotide-sequence variants relative to the reference genome. For each genomic region, the method determines a B-allele frequency and determines a B-allele frequency distribution for the biological sample.
The invention processes the B-allele frequency distribution using a trained machine-learning model to estimate a probability of a true tumor purity as a function of a predicted tumor purity in the biological sample. The trained machine-learning model is trained on a training dataset generated from nucleic acid sequence data derived from one or more tumor cells diluted into normal cells. Based on the estimated probability, the method generates a report to inform the cancer feature and evaluate the treatment efficacy for the subject.
Claims Coverage
The document includes two independent claims, a method and a system. Both claims share three inventive features: sequencing alignment and B-allele frequency distribution formation, machine-learning-based tumor purity probability estimation, and report generation for cancer feature and treatment efficacy evaluation.
Determining tumor purity using aligned sequencing variants and a B-allele frequency distribution
Obtaining nucleic acid sequence data; aligning the nucleic acid sequence data to a reference genome; identifying genomic regions with nucleotide-sequence variants; determining a B-allele frequency for each genomic region; and determining a B-allele frequency distribution for the biological sample.
Estimating true tumor purity probability using a machine-learning model trained on tumor-cell diluted training data
Processing the B-allele frequency distribution using a trained machine-learning model to estimate a probability of a true tumor purity as a function of a predicted tumor purity, wherein the model is trained on a training dataset generated from nucleic acid sequence data derived from one or more tumor cells diluted into normal cells.
Generating a report for cancer feature and treatment efficacy evaluation
Generating a report to inform the cancer feature and evaluate the treatment efficacy for the subject based on the estimated probability of a true tumor purity as a function of a predicted tumor purity in the biological sample.
A system performing sequencing alignment, B-allele distribution formation, ML-based tumor purity probability estimation, and report generation
A system comprising data processors and instructions to perform obtaining nucleic acid sequence data, aligning to a reference genome, identifying genomic regions with nucleotide-sequence variants, determining B-allele frequencies, determining a B-allele frequency distribution, processing the distribution using a trained machine-learning model trained on tumor-cell diluted into normal cells data to estimate a probability of true tumor purity as a function of predicted tumor purity, and generating a report to inform a cancer feature and evaluate treatment efficacy.
Across the independent claims, the core coverage is the combination of aligning nucleic-acid sequencing data to a reference genome, forming a B-allele frequency distribution from variant-containing genomic regions, using a trained machine-learning model trained on tumor-cell diluted into normal cells to estimate a probability of true tumor purity as a function of predicted tumor purity, and generating a report for cancer feature and treatment efficacy evaluation.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?