A Local Outlier Factor-Based Detection of Copy Number Variations From NGS Data

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Copy number variation (CNV) is a significant genomic structural variation linked to human disorders.
• Next-generation sequencing (NGS) has advanced CNV detection, but low-amplitude CNVs remain difficult to identify accurately.

Purpose of the Study:

• To introduce CNV-LOF, a novel computational method for detecting CNVs across a full range of amplitudes from NGS data.
• To address the limitations of existing methods in accurately identifying low-amplitude CNVs.

Main Methods:

• CNV-LOF analyzes read depths from a local perspective, assigning an outlier factor to genome segments.
• It employs a boxplot procedure based on outlier factors to identify CNVs, avoiding distribution assumptions.
• The method was evaluated using simulation experiments and real sequencing data.

Main Results:

• CNV-LOF demonstrated superior performance compared to five existing methods in terms of F1-measure, sensitivity, and precision.
• The method achieved high consistency with peer methods on real sequencing samples.
• CNV-LOF successfully identified low and moderate amplitude CNVs missed by other approaches.

Conclusions:

• CNV-LOF offers a robust and accurate approach for CNV detection from NGS data, especially for low-amplitude variations.
• The method's ability to detect a full range of CNV amplitudes makes it a promising tool for discovering novel CNVs in whole genome sequencing.