Systems and methods to detect rare mutations and copy number variation

What is this patent about?

’171 is related to the field of genetic analysis and bioinformatics, specifically focusing on the detection of rare mutations and copy number variations within cell-free polynucleotides. The technology addresses the significant challenge of identifying low-frequency somatic variants, such as those shed by tumors into the bloodstream, which are typically obscured by the inherent noise and distortion produced during PCR amplification and high-throughput sequencing processes.

The underlying idea behind ’171 is to treat the sequencing process as a noisy communication channel and apply a digital signaling approach to recover the original genetic message. By attaching molecular barcodes to both ends of individual DNA fragments before amplification, the system creates a traceable link between progeny reads and their unique parent molecules. This allows the system to distinguish between true biological variants and random technical errors by collapsing groups of reads into high-fidelity consensus sequences.

The claims of ’171 focus on a method for detecting somatic genetic variants by tagging cfDNA molecules with a specific number of molecular barcodes relative to the expected number of duplicate fragments. The independent claims require tagging both ends of the DNA fragments and define a barcode diversity (n) that is scaled to a measure of central tendency (z) of fragments sharing the same start and stop positions. This statistical bottlenecking ensures that duplicate molecules can be uniquely identified and grouped into families for accurate variant calling.

In practice, the invention works by mapping the barcoded sequencing reads to a reference genome and grouping them into families based on their unique barcodes and their genomic coordinates. By analyzing the base calls across all members of a family, the system can filter out errors that appear in only a subset of the progeny. This error-correction mechanism enables the detection of mutations at frequencies as low as 0.1%, which is often below the baseline error rate of standard sequencing platforms.

This approach differs from prior methods by optimizing the conversion efficiency of the library preparation and using the relationship between barcode diversity and fragment redundancy to maximize sensitivity. While traditional sequencing often loses rare signals during sample prep or mistakes them for noise, this method uses probabilistic modeling and family-based collapsing to provide a near-perfect representation of the original sample. This allows for non-invasive monitoring of disease progression and treatment efficacy through a simple blood draw.