Systematic assessment of structural variant annotation tools for genomic interpretation.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Structural variants (SVs) significantly impact human traits and disease but are challenging to analyze.
• Numerous computational tools exist for SV prioritization, but their real-world biomedical utility is not well-established.

Purpose of the Study:

• To benchmark the performance of eight widely used SV prioritization tools.
• To evaluate tool accuracy, robustness, usability, and efficiency across diverse genomic data.

Main Methods:

• Categorized tools into knowledge-driven (AnnotSV, ClassifyCNV) and data-driven (CADD-SV, dbCNV, StrVCTVRE, SVScore, TADA, XCNV) groups.
• Assessed performance using seven independent datasets, considering genomic context and biological mechanisms.
• Evaluated computational efficiency.

Main Results:

• Both knowledge-driven and data-driven tools demonstrated comparable effectiveness in predicting SV pathogenicity.
• Performance varied among individual tools, highlighting the need for careful selection based on research objectives.
• Identified areas for future tool improvement.

Conclusions:

• The study provides a critical evaluation framework for SV analysis tools.
• Offers practical guidance for biomedical researchers utilizing SV data.
• Aims to facilitate the development of improved genomic research tools.