Somatic Structural Variant Discovery from Tumor Specific Strings Computed Against a Draft Composite Germline Reference

VESPA-SS detects somatic structural variants from paired tumor-normal PacBio HiFi data, reaching F1 scores above 0.92 for HG008 and 0.90 for COLO829.

• Why it matters: Accurate somatic SV calling is hindered by repetitive regions, alignment ambiguity near breakpoints, germline versus artifact discrimination, and weak support for low-frequency events.
• What they did: The method computes specific strings from tumor reads against a Draft Composite Germline Reference, then clusters and refines signals using consensus sequence generation and alignment-based steps.
• The result: On HG008 and COLO829, VESPA-SS exceeds the best evaluated callers’ F1 scores (0.86 and 0.85), enabling more accurate discovery of complex cancer genome alterations.

AI summary of the abstract · may contain errors · not medical advice · read the paper.