A novel method addressing NGS-based mappability bias for sensitive detection of DNA alterations

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Next-generation sequencing (NGS) has revolutionized cancer research by enabling large-scale genome sequencing.
• NGS facilitates the detection and analysis of structural variations, including copy number variations (CNVs), crucial in cancer development.
• Existing CNV detection tools struggle with biases introduced by repetitive DNA sequences, particularly in CNV regions, leading to inaccurate identification.

Purpose of the Study:

• To develop a novel computational method for accurate CNV identification in cancer genomes.
• To specifically address and overcome the challenges posed by DNA repeats and mappability bias in CNV detection.
• To enhance the sensitivity and accuracy of CNV detection, especially in complex genomic regions.

Main Methods:

• A three-phase approach was developed, starting with alignment of uniquely mapped DNA reads.
• The method incorporates base quality and mismatch parameters at nucleotide-level precision.
• A probabilistic membership model was employed to optimally allocate non-uniquely mapped reads in DNA repeat regions.

Main Results:

• The proposed method demonstrates high sensitivity, exceeding [Formula: see text] in simulations.
• Validation on real data showed over 95% overlap with existing genomic variant databases.
• The methodology achieved superior breakpoint prediction compared to other bias correction CNV detection methods.

Conclusions:

• The novel method effectively identifies CNVs in both coding and non-coding DNA regions, including challenging repeat regions.
• This approach significantly improves the accuracy and reliability of CNV detection in cancer genomics.
• The findings offer a more robust tool for analyzing structural alterations in cancer genomes, advancing the field.