IMSindel

IMSindel detects intermediate-size insertions and deletions (indels) from short-read next-generation sequencing data to enable accurate identification of structural variants relevant to human disease studies.

Key Features:

• Detection range: Detects intermediate-size insertions and deletions (indels) of 50 base pairs (bp) and above.
• Input signals: Utilizes BWA soft-clipped fragments and unmapped reads from short-read next-generation sequencing.
• Algorithms: Integrates de novo assembly with gapped global-local alignment and split-read analysis to reconstruct and characterize indels.
• Short-read mitigation: Addresses limitations of short read lengths typical of next-generation sequencing for detecting larger indels.
• Validation and benchmarking: Benchmarked on whole exome sequencing across multiple samples with Sanger sequencing of all predicted indels to determine false positive and false negative rates.
• Performance metrics: Reported an F-measure of 0.84 in one sample versus GATK 0.56, PINDEL 0.52, and ScanIndel 0.46, with similar results across additional samples.
• Biological relevance: Detects indels that can cause frameshift mutations or larger structural changes affecting gene function.

Scientific Applications:

• Disease association studies: Improves detection of intermediate-size indels for studies linking indels to human diseases.
• Structural variant discovery: Enables discovery and characterization of larger indels and frameshift-causing variants.
• Whole exome sequencing analysis: Applied to whole exome sequencing datasets for accurate indel calling and benchmarking.
• Benchmarking and validation: Facilitates method comparison and validation using Sanger sequencing to assess precision and sensitivity.

Methodology:

Performs de novo assembly combined with gapped global-local alignment and split-read analysis using BWA soft-clipped fragments and unmapped reads from short-read NGS data.