LcDel: deletion variation detection based on clustering and long reads

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Genomic structural variations, including deletions, are crucial in disease pathogenesis.
• Existing deletion detection tools often struggle with chimeric variants, impacting accuracy.
• Accurate deletion detection aids disease diagnosis, treatment, and prevention.

Purpose of the Study:

• To develop a novel tool, LcDel, for precise deletion variant detection using long reads.
• To improve the accuracy of deletion detection by addressing limitations in current methods, particularly regarding chimeric variants.

Main Methods:

• LcDel employs a two-step clustering approach for deletion detection.
• Candidate deletion sites are identified, followed by sliding window-based and coverage-based clustering.
• Hierarchical clustering is utilized to refine deletion location and length determination.

Main Results:

• LcDel demonstrates superior performance in deletion variant detection compared to existing tools.
• The tool effectively handles chimeric variants, leading to more precise clustering results.
• Benchmarking across multiple datasets validates LcDel's enhanced detection capabilities.

Conclusions:

• LcDel offers a more accurate and reliable method for detecting genomic deletions.
• The tool's ability to manage chimeric variants represents a significant advancement in structural variant analysis.
• LcDel provides valuable insights for understanding disease mechanisms and improving clinical applications.