A comprehensive benchmarking of WGS-based deletion structural variant callers

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Molecular Genetics

Background:

• Whole-genome sequencing (WGS) advancements facilitate structural variant (SV) discovery.
• Numerous SV detection tools exist, but comparative performance data is limited.
• Selecting appropriate SV detection tools for WGS data analysis remains challenging.

Purpose of the Study:

• To evaluate and benchmark the performance of various SV detection tools.
• To provide evidence-based guidance for selecting optimal SV detection methods.
• To assess tool performance on both mouse and human WGS datasets.

Main Methods:

• Performance evaluation of SV detection tools using mouse and human WGS data.
• Utilized a comprehensive polymerase chain reaction-confirmed gold standard SV set (mouse) and genome-in-a-bottle variant set (human).
• Focused on deletion detection to provide an optimistic estimate of overall SV detection capability.

Main Results:

• Significant variability in performance was observed among different SV detection tools.
• Several tools demonstrated a favorable balance between sensitivity and precision for SV detection.
• Identified SV callers optimal for low- and ultralow-pass sequencing data and various deletion lengths.

Conclusions:

• The choice of SV detection tool critically impacts discovery accuracy and comprehensiveness.
• This study provides essential performance metrics to guide tool selection in WGS analysis.
• Recommendations are offered for optimizing SV detection based on data type and variant characteristics.