Characterization of background noise in capture-based targeted sequencing data.

Area of Science:

• Genomics
• Molecular Biology
• Biotechnology

Background:

• Targeted deep sequencing is vital for detecting low-allelic fraction variants.
• Characterizing sequencing errors is crucial for improving detection limits.
• Baseline noise in sequencing data can impede variant identification.

Purpose of the Study:

• To systematically evaluate errors introduced during specific steps of capture-based targeted sequencing.
• To identify and characterize errors associated with acoustic shearing and hybrid selection.
• To provide insights for improving targeted deep sequencing methodologies.

Main Methods:

• Systematic evaluation of errors in targeted deep sequencing.
• Filtering low-quality bases to analyze background noise.
• Comparative experiments under different shearing conditions.
• Analysis of DNA fragmentation patterns and error localization.

Main Results:

• Acoustic shearing contributes significantly to C:G>A:T and C:G>G:C errors via guanine oxidation.
• Acoustic shearing also causes A>G and A>T substitutions at fragment ends, linked to DNA breakage.
• Hybrid selection contributes to C:G>A:T and C>T errors.

Conclusions:

• This study comprehensively summarizes errors in targeted deep sequencing and their causes.
• Understanding these errors is key to technical improvements in sequencing methods.
• The findings may enhance the future application of targeted deep sequencing for low-allelic fraction variant detection.