Deciphering the exact breakpoints of structural variations using long sequencing reads with DeBreak

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Long-read sequencing offers potential for structural variation (SV) characterization.
• Existing SV discovery algorithms lack sufficient sensitivity and precision.
• Accurate SV detection is vital for genomic research, especially in cancer.

Purpose of the Study:

• To introduce DeBreak, a novel computational method for comprehensive and accurate SV discovery.
• To overcome the limitations of existing SV detection tools.
• To enable precise characterization of structural variations using long-read sequencing data.

Main Methods:

• DeBreak utilizes a density-based clustering approach for SV candidate grouping.
• It incorporates local de novo assembly for reconstructing long insertions.
• Partial order alignment ensures single base-pair resolution for SV breakpoints.

Main Results:

• DeBreak demonstrates superior performance compared to existing tools on simulated and real long-read data (PacBio and Nanopore).
• The method accurately identifies multi-allele SV events using k-means clustering.
• DeBreak successfully characterizes complex structural variations.

Conclusions:

• DeBreak provides a sensitive and precise computational solution for SV discovery.
• It is a valuable tool for analyzing cancer genomes to identify tumor-driving SVs.
• DeBreak can complement existing whole-genome assembly-based SV detection methods.