svviz

svviz visualizes sequencing reads that support candidate structural variants (SVs) in high-throughput sequencing data to enable validation and assessment of SV evidence.

Key Features:

• Targeted Read Filtering: Filters input BAM files to identify reads potentially relevant to a candidate structural variant.
• Read Realignment: Realigns identified reads against the inferred sequence of the putative variant allele and against the reference genome for comparative assessment.
• Allele-specific Alignments: Generates separate allele-specific alignments to distinguish reads supporting the variant allele from reads supporting the reference.
• Evidence Evaluation and Zygosity Estimation: Uses allele-specific alignments to evaluate evidence for or against the putative SV and to estimate zygosity.
• Breakpoint Refinement: Supports manual refinement of breakpoints based on read alignment data.
• Sequencing Platform Compatibility: Operates with data from most modern sequencing platforms.
• Implementation: Implemented in Python.

Scientific Applications:

• Variant Validation: Confirms candidate structural variants by comparing read support for variant and reference alleles.
• Zygosity Estimation: Estimates zygosity of structural variants using counts and alignments of allele-supporting reads.
• Breakpoint Refinement: Refines SV breakpoint coordinates based on realigned reads.
• Genomic Annotation Assessment: Assesses genomic annotations affected by the variant through read-level evidence.

Methodology:

Filters BAM reads for candidate SV regions, realigns those reads to the inferred variant allele and to the reference genome, and generates allele-specific alignments to evaluate support and enable manual breakpoint refinement; implemented in Python.