Detection of rare single nucleotide variants through synthesis and selection

Abstract

Next-generation sequencing (NGS) has greatly expanded our understanding of the human genome, extending its utility to the detection and analysis of cancer-related mutations. However, detecting mutations with variant allele frequencies (VAF) below 1% remains challenging due to the intrinsic error rates of NGS platforms, which range from 0.1% to 1% per base. Although advanced detection technologies can achieve a limit of detection (LoD) as low as 0.01%, they are complex, time-consuming, and costly. To address these issues, we introduce a novel approach for detecting low-frequency single nucleotide variants (SNVs) through a synthesis and selection mechanism that we developed. The first step of our proposed method involves calling variants from low-depth sequencing data. The initial sequences of the sequencing reads are used as barcodes. To distinguish targets, a combination of nucleotides modified with reversible terminators and irreversible terminators is synthesized within the barcode region. Initially, a reversible terminator complementary to the target barcode and other irreversible terminators are synthesized. Then, a reducing agent is used to remove the reversible terminator, allowing the synthesis process to be repeated. After synthesis, only the targets can elongate as their terminators are removed. This allows the detection of the selectively enriched target at low-depth sequencing. Through synthesis and selection, the target read was amplified 600-fold, and the variant allele frequency increased from 1.16% to 44.90% in standard human genomic DNA samples. This method reduces the NGS read requirement by over 100-fold, significantly lowering the cost. Additionally, our approach does not require the design and synthesis of hybridization probes based on known variants, thereby saving time and cost and enabling the identification of novel cancer variants.