US 11149306

US Patent 11149306, titled "Methods and systems for detecting genetic variants," was issued on October 19, 2021, from an application filed on July 31, 2020 (Application number US16/945,124). The patent is currently assigned to Guardant Health Inc., and the inventors are AmirAli Talasaz, Helmy Eltoukhy, and Stefanie Ann Ward Mortimer.

Abstract:
The patent describes methods and systems for identifying genetic variants, such as copy number variations (CNVs), in a polynucleotide sample. This involves tagging double-stranded polynucleotides with duplex tags, sequencing the polynucleotides, and then estimating the total number of polynucleotides at specific genetic locations. This estimation accounts for molecules that were tagged but not sequenced, by considering both polynucleotides where both complementary strands were detected ("Pairs") and those where only one strand was detected ("Singlets").

Plain-Language Overview of Independent Claims:

• Claim 1: Describes a method for detecting and/or quantifying rare DNA (less than 1% concentration) in a sample containing various DNA fragments with high accuracy (over 99.9% specificity). This method involves tagging the DNA fragments in a single reaction with different tags, each having a molecular barcode, ensuring that more than 30% of fragments are tagged at both ends. The tagged fragments are then amplified, sequenced (including the barcodes and DNA fragments), and processed to create "consensus reads" representing single strands of the original DNA. These consensus reads are then quantified to detect or measure the rare DNA.

• Claim 15: Defines a specific type of "library adaptor" (tags). These adaptors are short polynucleotide molecules (up to 80 bases) with molecular barcodes (at least 4 bases long). The barcodes must be distinct from each other (minimum edit distance of 1), positioned away from the adaptor's ends, can optionally have identical terminal bases across all adaptors, and do not contain a complete "sequencer motif" (a sequence that fully enables sequencing by a machine).

• Claim 23: Outlines a method for processing and/or analyzing a nucleic acid sample from a subject. It involves exposing DNA fragments from the sample to the "library adaptors" as defined in Claim 15 to create tagged fragments. These tagged fragments are then subjected to amplification reactions to produce more copies.

• Claim 32: Describes a method using a computer processor to analyze DNA sequence data. It starts by generating sequence reads from DNA molecules covering specific genomic regions (loci) across a defined list of genes (e.g., ALK, BRAF, TP53). A computer then groups these reads into "families," where each family comes from a single original DNA molecule. Within each family, the reads are combined to generate a "consensus sequence." The computer then "calls" (identifies) the genetic sequence at specific loci based on these consensus sequences and detects genetic variants, their frequencies, and total counts.

• Claim 41: Details a method for preparing DNA for sequencing, performed entirely within a single reaction vessel. It involves providing original DNA molecules and specific library adaptors (with different molecular barcodes, but without complete sequencer motifs). These adaptors are attached to the original DNA at an efficiency of at least 10%, creating tagged DNA. This tagged DNA is then amplified and subsequently sequenced.

• Claim 50: Presents a system for analyzing a subject's DNA. This system includes a communication interface to receive DNA sequence data, computer memory to store this data, and a computer processor. The processor is programmed to perform several steps: (i) group sequence reads into families (each from a single original DNA molecule), (ii) merge reads within each family to create a consensus sequence, (iii) identify the sequence at specific genomic locations (loci) based on the consensus, and (iv) detect genetic variants, their frequencies, and total counts at those loci. The genomic loci for this system correspond to the same specific list of genes as in Claim 32.

• Claim 51: Defines a set of oligonucleotide molecules designed to specifically bind (hybridize) to at least 5 genes from a broad list of cancer-related genes (the same list as in Claims 32 and 50).

• Claim 56: Describes a kit containing two containers. The first container holds a plurality of library adaptors, each with a different molecular barcode, conforming to the characteristics described in Claim 15 (less than or equal to 80 bases, molecular barcodes at least 4 bases, etc.). The second container holds sequencing adaptors, which include part of a sequencer motif and optionally a sample barcode.

• Claim 59: Claims a method for detecting sequence variants in a cell-free DNA sample with high sensitivity and specificity: detecting rare DNA (concentration less than 1%) with greater than 99.9% specificity.

• Claim 60: Claims a method for detecting genetic variants in a DNA sample with a detection limit of at least 1% and a specificity greater than 99.9%.

• Claim 61: Describes a method for quantifying the total number of individual double-stranded DNA molecules in a sample. This involves: (a) providing a sample of double-stranded DNA, (b) tagging each complementary strand with a unique duplex tag, (c) sequencing the tagged strands, (d) reducing redundancy in the sequence data, (e) sorting reads into "paired reads" (both complementary strands detected) and "unpaired reads" (only one strand detected), (f) quantifying these paired and unpaired reads at specific genetic locations, and (g) using a computer to estimate the total number of original double-stranded DNA molecules at each location based on the counts of paired and unpaired reads.

• Claim 70: Describes a system with a computer-readable medium that, when executed by a computer processor, performs a method. This method includes: (a) receiving sequence reads of DNA tagged with duplex tags, (b) reducing redundancy in these reads, (c) sorting reads into paired and unpaired categories (as defined in Claim 61), (d) quantifying these paired and unpaired reads at specific genetic locations, and (e) estimating the total number of double-stranded DNA molecules at each location based on these quantitative measures.

• Claim 71: Details another method for analyzing double-stranded polynucleotide molecules. Similar to Claim 61, it involves tagging complementary strands with duplex tags, sequencing, reducing redundancy, and sorting reads into paired and unpaired categories. However, it specifically focuses on (f) determining quantitative measures of at least two of: paired reads, unpaired reads at genetic loci, read depth of paired reads, and read depth of unpaired reads.

• Claim 74: Describes a method for determining copy number variation (CNV) using control and test DNA. It involves tagging control DNA with a "control tag" and identifying tags, and tagging test DNA with a distinguishable "test tag" and identifying tags. The tagged control and test DNA are then mixed, amplified, and sequenced. A computer groups reads from the same original DNA molecule and classifies them as either control or test based on their tags. Finally, it quantifies control and test DNA at specific locations and determines CNV in the test DNA based on their relative quantities.

• Claim 77: Outlines a computer-implemented method for detecting genomic alterations. It involves generating sequence reads from DNA molecules at specific genomic locations, grouping these reads into families (each from a single original molecule), "calling" (identifying) the base or sequence at that location for each family, and then detecting genomic alterations, their frequencies, and total counts among these calls.

• Claim 79: Describes a method for determining the actual number of individual double-stranded DNA fragments in a sample. This is done by: (a) measuring the number of molecules where both DNA strands are detected, (b) measuring the number of molecules where only one DNA strand is detected, (c) inferring the number of molecules where neither strand was detected from (a) and (b), and (d) using all three (a)-(c) to determine the total number of double-stranded DNA fragments.

• Claim 85: Claims a method for reducing distortion in a sequencing assay using internal controls. It involves tagging control DNA with a first set of tags and test DNA with a second, distinguishable set of tags. These are mixed, and their quantities are determined. The quantities of the tagged control DNA are then used to correct for distortion in the quantities of the tagged test DNA.

• Claim 89: Describes a method for analyzing double-stranded DNA. It involves ligating molecular barcode adaptors to double-stranded DNA in a single reaction vessel, creating a tagged library with many different tags. Sequence reads are generated for each DNA molecule, grouped into families (from a single original DNA molecule) based on tag and fragment end information, and then bases at each position are "called" based on the family members.

• Claim 94: Outlines a method for detecting disease cell heterogeneity (differences within a disease population) from a sample containing both healthy and disease cell DNA. It involves: quantifying DNA with sequence variants at multiple genetic locations, determining the copy number variation (CNV) at these locations (indicating gene dosage in disease cells), using a computer to calculate the relative amount of variant DNA per gene dosage at each location, and then comparing these relative measures. Different relative measures indicate tumor heterogeneity.

• Claim 95: Describes a method for subjecting a patient to "pulsed therapy cycles." Each cycle has two periods: a first period with a high drug dose (when tumor burden is above a first clinical level) and a second period with a reduced drug dose (when tumor burden is below a second clinical level).

Litigation:
US Patent 11149306 has been involved in several litigation actions:

• An IPR case (IPR2025-01434) was filed at the PTAB, which was "Not Instituted - Procedural."
• There is first worldwide family litigation filed.
• A US case was filed in the Court of Appeals for the Federal Circuit (case 24-1626).
• Two US cases were filed in the Delaware District Court (cases 1:22-cv-00334 and 1:24-cv-00687).
• Another IPR case (IPR2022-01400) was filed at the PTAB, which resulted in a "Final Written Decision."