Systems and methods to detect rare mutations and copy number variation

What is this patent about?

’063 is related to the field of molecular diagnostics and bioinformatics, specifically focusing on the high-sensitivity detection of genetic aberrations in cell-free DNA (cfDNA). The technology addresses the challenge of identifying rare somatic mutations and copy number variations that are often masked by the inherent noise and distortion of standard next-generation sequencing workflows. By analyzing extracellular polynucleotides found in bodily fluids like blood or plasma, the system provides a non-invasive means of monitoring cancer progression and treatment efficacy.

The underlying idea behind ’063 is to treat the sequencing process as a communication channel where the original DNA molecules are the message and sequencing artifacts are noise. To overcome this noise, the invention utilizes a digital strategy that involves tagging individual DNA fragments with molecular barcodes and using the unique start and stop positions of these fragments to track them. By grouping progeny molecules into families derived from the same original parent molecule, the system can collapse multiple reads into a single consensus sequence, effectively filtering out random errors introduced during PCR amplification or the sequencing run itself.

The claims of ’063 focus on a method for detecting somatic genetic variants by non-uniquely tagging cfDNA molecules with a limited set of molecular barcodes. The process requires that the DNA fragments be flanked on both ends by these barcodes, creating a pool of non-uniquely tagged parent polynucleotides. The method specifically relies on a combination of the barcode sequence and the specific genomic coordinates—the beginning and end base positions—to group sequencing reads into families. This multi-layered identification allows the system to distinguish between true biological variants and technical artifacts across various mutation types, including single nucleotide variants and gene fusions.

In practice, the invention achieves high sensitivity by ensuring that the number of unique barcodes used is significantly smaller than the total number of DNA fragments that map to a specific genomic location. This non-unique tagging approach simplifies the molecular biology required while still providing enough diversity, when combined with fragment length and alignment data, to uniquely identify the original parent molecules. Once the reads are grouped into families, the system applies statistical or probabilistic models to determine the frequency of variants, allowing for the detection of mutations that occur at frequencies as low as 0.1% or less.

This approach differs from prior methods by maximizing the conversion efficiency of the library preparation, ensuring that rare tumor-derived fragments are not lost before they reach the sequencer. While traditional sequencing often requires large amounts of input DNA to overcome losses, this method is optimized for low-input samples typical of liquid biopsies. By shifting the focus from simple read counting to the analysis of molecular families, the technology provides a much clearer picture of the tumor's genetic landscape, enabling real-time monitoring of disease evolution and the emergence of drug resistance.