Benchmarking copy number variation detection with low-coverage whole-genome sequencing.

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Low-coverage whole-genome sequencing (lcWGS) is a cost-effective method for genome-wide copy number variation (CNV) profiling.
• Technical limitations and analytical variability of lcWGS necessitate systematic evaluation for reliable application.

Purpose of the Study:

• To benchmark five CNV detection tools for lcWGS.
• To evaluate the impact of sequencing depth, FFPE artifacts, tumor purity, multi-center reproducibility, and signature stability on CNV profiling.
• To establish guidelines for robust CNV profiling using lcWGS in precision oncology.

Main Methods:

• Benchmarking of five CNV detection tools using simulated and real-world datasets.
• Evaluation focused on sequencing depth, FFPE artifacts, tumor purity, multi-center reproducibility, and signature-level stability.
• Comparative analysis of CNV feature extraction methods (Wang et al., Steele et al., Tao et al.).

Main Results:

• IchorCNA demonstrated superior precision and runtime at high tumor purity (≥50%) for lcWGS.
• Prolonged FFPE fixation induced artifactual CNVs due to DNA fragmentation, uncorrectable by current tools.
• High reproducibility was observed for the same tool across centers, but low concordance between different tools.
• Copy number features from Wang et al. showed greater stability across conditions compared to Steele et al. and Tao et al.

Conclusions:

• IchorCNA is the recommended tool for lcWGS CNV profiling when tumor purity is ≥50%.
• Strict control over FFPE fixation time or use of fresh-frozen samples is crucial to mitigate DNA fragmentation artifacts.
• The Wang et al. method provides more stable copy number features for robust profiling in precision oncology.