Enhanced Error Suppression for Accurate Detection of Low-Frequency Variants

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Low-frequency variant detection is hindered by high error rates in next-generation sequencing (NGS).
• Current methods like unique molecular identifiers (UMIs) often lead to significant read waste and cost inefficiency due to redundant sequencing requirements.
• Optimizing data utilization and error suppression in NGS is crucial for accurate variant identification.

Purpose of the Study:

• To introduce a novel approach, the enhanced error suppression strategy (EES), for improving low-frequency variant detection in NGS.
• To address the challenges of read waste and cost inefficiency associated with traditional error suppression techniques.
• To enhance the accuracy and data utilization of NGS by minimizing sequencing errors.

Main Methods:

• Developed an enhanced error suppression strategy (EES) incorporating single-read correction and Bayes' theorem.
• Optimized data utilization by reserving and complementing single reads and single-strand consensus sequences (SSCSs).
• Applied Bayesian inference to enhance the accuracy of next-generation sequencing data.

Main Results:

• Achieved a significantly low background error rate of less than 4.4 × 10^-5 per base pair.
• Demonstrated a 22.9-fold increase in duplex consensus sequence (DCS) recovery compared to traditional methods.
• Showcased superior error suppression across various base substitution types, improving variant detection accuracy.

Conclusions:

• EES offers a significant advancement in detecting low-frequency variants by enhancing data utilization and reducing sequencing errors.
• The strategy improves the sensitivity and accuracy of NGS applications.
• EES is highly valuable for clinical and research settings requiring precise variant detection.