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

Academic Article


  • Abstract Long-read sequencing has demonstrated great potential for characterizing all types of structural variations (SVs). However, existing algorithms have insufficient sensitivity and precision. To address these limitations, we present DeBreak, a novel method for comprehensive and accurate SV discovery. Based on alignment results, DeBreak employs a density-based approach for clustering SV candidates together with a local de novo assembly approach for reconstructing long insertions. A partial order alignment algorithm ensures precise SV breakpoints with single base-pair resolution, and a k-means clustering method can report multi-allelic SV events. DeBreak outperforms existing tools on both simulated and real long-read sequencing data from both PacBio and Nanopore platforms. An important application of DeBreak is analyzing cancer genomes for potentially tumor-driving SVs. DeBreak also demonstrates excellent performance in supplementing whole-genome assembly methods.
  • Digital Object Identifier (doi)

    Author List

  • Chen Y; Wang A; Barkley C; Zhao X; Gao M; Edmonds M; Chong Z