SVseq: an approach for detecting exact breakpoints of deletions with low-coverage sequence data

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology

Background:

• Structural variations (SVs), including deletions, are key genetic variations linked to diseases.
• Detecting deletions accurately, especially with low-coverage short sequence reads, remains a significant challenge.
• Existing methods using split reads or discordant insert sizes have limitations with low-coverage data.

Purpose of the Study:

• To develop an efficient and accurate method for detecting deletions from low-coverage short sequence reads.
• To improve upon existing techniques for structural variation analysis in genomic data.

Main Methods:

• Introduced SVseq, a two-stage approach integrating split reads mapping and discordant insert size analysis.
• Stage 1: Utilizes Burrows-Wheeler transform (BWT)-based split reads mapping, allowing mismatches and small indels for broader detection.
• Stage 2: Employs discordant insert size analysis to filter false positives identified in the first stage.

Main Results:

• SVseq demonstrates higher accuracy compared to alternative methods on both simulated and empirical datasets.
• The SVseq approach is significantly faster than existing deletion detection tools.
• The split reads mapping method effectively identifies deletions near other small variations and utilizes reads with sequencing errors.

Conclusions:

• SVseq offers a more accurate and efficient solution for deletion detection in challenging low-coverage sequencing data.
• The combined approach overcomes limitations of previous methods, improving structural variation analysis.
• SVseq is a valuable tool for genomic research, particularly in identifying disease-associated genetic variations.