Long-read-based human genomic structural variation detection with cuteSV

Area of Science:

• Genomics
• Bioinformatics

Background:

• Long-read sequencing technologies offer potential for comprehensive structural variation (SV) discovery.
• Challenges remain in achieving high yield and performance due to complex SV signatures in noisy long reads.

Purpose of the Study:

• To introduce cuteSV, a novel approach for sensitive, fast, and scalable SV detection from long-read sequencing data.
• To address the limitations of existing tools in accurately identifying SVs from noisy long reads.

Main Methods:

• cuteSV employs tailored methods to capture diverse SV signatures.
• A clustering-and-refinement strategy is utilized for sensitive SV detection.
• The approach is benchmarked on simulated and real long-read sequencing datasets.

Main Results:

• cuteSV demonstrates superior yields and scalability compared to state-of-the-art tools.
• The method effectively identifies various types of structural variations.
• Performance benchmarks confirm the tool's efficiency and accuracy.

Conclusions:

• cuteSV provides a robust and efficient solution for structural variation detection using long-read sequencing.
• The tool enhances the comprehensive discovery of SVs, overcoming previous technical hurdles.
• cuteSV is publicly available to facilitate genomic research.