Comprehensively benchmarking applications for detecting copy number variation.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Copy number variation (CNV) is a significant genomic alteration linked to complex traits and diseases.
• Numerous computational tools for CNV detection have emerged, necessitating comparative evaluations.
• Choosing the appropriate CNV detection method is crucial for researchers based on specific study goals.

Purpose of the Study:

• To comprehensively compare the performance of ten popular CNV detection applications.
• To evaluate methods based on sensitivity, specificity, and computational resource requirements.
• To guide researchers in selecting the most suitable CNV detection tools for their specific needs.

Main Methods:

• Benchmarking of ten CNV detection applications: CNVnator, ReadDepth, RDXplorer, LUMPY, Control-FREEC, Canvas, GROM-RD, and others.
• Evaluation using the DGV gold standard variants dataset.
• Testing across a range of sequencing depths (5X to 50X) using real sequencing data.

Main Results:

• LUMPY demonstrated superior performance, achieving high sensitivity and specificity across all tested sequencing depths.
• Canvas is recommended for applications prioritizing high specificity.
• CNVnator and RDXplorer are suitable choices when high sensitivity is the primary requirement.
• CNVnator and GROM-RD showed good performance with low-depth sequencing data.

Conclusions:

• The study provides a detailed performance evaluation of commonly used CNV detection methods.
• Findings assist researchers in selecting appropriate tools for CNV analysis based on desired sensitivity, specificity, and sequencing depth.
• Results can inform the development of improved CNV prediction algorithms.