Enhancing Next-Generation Sequencing Sensitivity with High-Recovery Adapter Ligation and Cas9-Mediated Dimer Elimination

BACKGROUND: Accurate detection of ultra-low-frequency variants is a major challenge in clinical liquid biopsy. In early cancer detection and minimal residual disease monitoring, Circulating tumor (ctDNA) may fall below 0.1% variant allele frequency, making sensitivity highly dependent on molecular recovery during library preparation. Losses at early steps, especially adapter ligation, permanently reduce analyzable molecules and cannot be rescued by deeper sequencing or bioinformatic refinement. METHODS: We developed Powerful Recovery and Improved Dimer Elimination (PRIDE) next-generation sequencing NGS, a library preparation strategy that increases adapter ligation efficiency and removes adapter dimers via sequence-specific Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) cleanup. PRIDE NGS is compatible with standard clinical work flows and requires no added sequencing depth or changes to downstream bioinformatic pipelines. Performance was assessed by targeted sequencing of cell-free (cfDNA) reference standards and clinical plasma samples. RESULTS: PRIDE NGS improved recovery and detection of low-frequency variants vs conventional preparation. In reference standards, it detected more variants at low allele frequencies, particularly below 0.1%. In clinical plasma samples, it similarly increased detection, including variants predicted to have moderate or high functional impact. These gains occurred at comparable or lower sequencing depth, indicating sensitivity improvements driven by enhanced molecular recovery. CONCLUSIONS: By overcoming a key bottleneck in library preparation, PRIDE NGS lowers the practical detection threshold for ultra-low-frequency variants in liquid biopsy. This clinically applicable approach improves analytical sensitivity by lowering the detection limit without increasing the sequencing burden, supporting routine testing and longitudinal monitoring.