Back to results
US 12,567,492 B1Granted
Genetic variant detection based on merged and unmerged reads
AssigneeGuardant Health, Inc.
InventorsDarya Chudova, Mohammad Reza Mokhtari
Life of this patent
Estimating the term…
Abstract
Methods and systems for improving identification of nucleic acid variants by identifying genetic sequence reads having identical molecular barcodes and sequences among sequence reads from a nucleic acid sequencer, grouping the genetic reads into a family, and processing families comprising split reads to detect the variants in a sample of polynucleotide molecules.
Claims19
A method for treating a subject having cancer associated with an ALK rearrangement, the method comprising:
- (a) selecting the subject having cancer associated with ALK rearrangement, comprising:
- generating a plurality of paired-end reads from cell-free nucleic acid molecules from a sample of the subject;
- determining a plurality of paired-end reads from among the plurality of paired-end reads based on an overlap criterion;
- determining a plurality of merged reads based on the plurality of paired-end reads that satisfy an overlapping criterion, wherein a merged read comprises a sequence based on respective sequences of paired-end reads;
- determining a plurality of unmerged reads from among the plurality of paired-end reads that do not satisfy the overlap criterion, wherein a given unmerged read comprises a paired-end read that does not satisfy the overlap criterion with a mated paired-end read;
- aligning the plurality of merged reads and the plurality of unmerged reads to a human reference genome to generate a plurality of aligned reads;
- determining a plurality of split reads from among the plurality of aligned reads, wherein a given split read includes a first subsequence portion that aligns to a first nucleic acid locus of the reference sequence and a second subsequence portion that aligns to a second nucleic acid locus of the reference sequence, wherein the second nucleic acid locus is different from the first nucleic acid locus;
- determining for a given split read, a breakpoint comprising a pair of genomic locations corresponding to the first nucleic acid locus and the second nucleic acid locus;
- clustering the plurality of split reads based on their respective breakpoints to generate a plurality of variant clusters;
- and
- determining that one or more variant clusters meeting a predetermined criterion are indicative of an ALK rearrangement, thereby detecting an ALK rearrangement in the sample from the subject;
- and
- (b) administering a treatment comprising lorlatinib or crizotinib to the subject to treat the cancer based on the detected ALK rearrangement.
10 more elements
A method for treating a subject having cancer associated with an ALK rearrangement, the method comprising:
- (a) selecting the subject having cancer associated with ALK rearrangement, comprising:
- generating a plurality of paired-end reads generated from cell-free nucleic acid molecules from a sample of the subject;
- determining a plurality of merged reads based on respective mated pairs of paired-end reads that satisfy an overlap criterion, wherein a merged read comprises a sequence based on respective sequences of paired-end reads;
- determining a plurality of unmerged reads from among the plurality of paired-end reads, wherein a given unmerged read comprises a paired-end read that does not satisfy an overlap criterion with a mated paired-end read;
- generating a plurality of concatenated unmerged reads based on the plurality of unmerged reads;
- aligning the plurality of merged reads and the plurality of concatenated unmerged reads to a human reference genome to generate a plurality of aligned reads;
- determining a plurality of split reads from among the plurality of aligned reads, wherein a given split read from among the plurality of split reads includes a first subsequence portion that aligns to a first nucleic acid locus of the reference sequence and a second subsequence portion that aligns to a second nucleic acid locus of the reference sequence, the second nucleic acid locus different from the first nucleic acid locus;
- and
- determining that a plurality of split reads meeting a predetermined criterion comprise an ALK rearrangement, thereby detecting an ALK rearrangement in the sample from the subject;
- and
- (b) administering a treatment comprising a tyrosine kinase inhibitor to the subject based on the detected ALK rearrangement.
8 more elements
Dependencies are inferred from the claim text.