Methods and systems for detecting genetic variants

What is this patent about?

’307 is related to the field of genetic analysis, specifically the detection and quantification of rare genetic variants such as copy number variations (CNV) and single nucleotide variants (SNV) within cell-free DNA (cfDNA) samples. The technology addresses the inherent limitations of liquid biopsies, where the low concentration of tumor-derived DNA and the inefficiencies of sequencing workflows often lead to inaccurate molecule counts and reduced diagnostic sensitivity.

The underlying idea behind ’307 is that the true number of DNA fragments in a sample can be mathematically inferred by tracking the recovery of individual strands from double-stranded molecules. By using a non-unique tagging strategy—where a limited set of barcodes is combined with endogenous fragment sequences—the system can distinguish between molecules where both strands were sequenced (pairs) and those where only one was captured (singlets). This statistical distribution allows the system to calculate the number of unseen molecules that were present in the original sample but lost during the library preparation or sequencing process.

The claims of ’307 focus on a method for determining the number of cfDNA molecules by employing a high-efficiency ligation step using more than a 10× excess of adapters relative to the DNA population. The independent claims specify that adapters containing molecular barcodes are ligated to both ends of the fragments, achieving at least a 20% conversion efficiency. The method then utilizes the resulting sequence reads and barcodes to map fragments to a reference genome and quantify the original molecule population at specific loci based on the detection of one or both strands.

In practice, the invention works by tagging double-stranded polynucleotides with duplex adapters that allow the bioinformatics pipeline to differentiate between the Watson and Crick strands. After amplification and sequencing, the system groups reads into families based on their barcodes and genomic start/stop positions. By analyzing the ratio of paired reads to unpaired reads, the system applies a probabilistic model (such as a binomial distribution) to correct for sampling bias and library loss, providing a more accurate representation of the genetic landscape than simple read counting.

This approach differentiates itself from prior methods by moving beyond simple error correction to address the problem of stochastic sampling loss. While traditional molecular barcoding focuses on eliminating polymerase-induced artifacts, ’307 provides a mechanism to account for the molecules that never make it to the sequencer at all. This statistical inference is critical for detecting minute changes in copy number and rare somatic mutations, ensuring that clinical decisions are based on the actual molecular burden in the patient's blood rather than sequencing artifacts.