A systematic benchmark of copy number variation detection tools for high density SNP genotyping arrays

Copy Number Variations (CNVs) are crucial in various diseases, especially cancer, but detecting them accurately from SNP genotyping arrays remains challenging. Therefore, this study benchmarked five CNV detection tools—PennCNV, QuantiSNP, iPattern, EnsembleCNV, and R-GADA—using SNP array and WGS dat...

Full description

Saved in:

Bibliographic Details
Published in:	Genomics (San Diego, Calif.) Vol. 116; no. 6; p. 110962
Main Authors:	van Baardwijk, M.N., Heijnen, L.S.E.M., Zhao, H., Baudis, M., Stubbs, A.P.
Format:	Journal Article
Language:	English
Published:	Elsevier Inc 13-11-2024
Subjects:	Benchmarking Copy number variation Reference datasets SNP arrays Variant calling WGS BE CNP FN CNV FP LRR Variant calling QC Copy number variation SNP FDR kb SBL BAF NHGRI OB-HMM Benchmarking HMM GQ Reference datasets DGV MLPA SNP arrays DRAGEN GADA CNVR NGS TP
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Copy Number Variations (CNVs) are crucial in various diseases, especially cancer, but detecting them accurately from SNP genotyping arrays remains challenging. Therefore, this study benchmarked five CNV detection tools—PennCNV, QuantiSNP, iPattern, EnsembleCNV, and R-GADA—using SNP array and WGS data from 2002 individuals of the DRAGEN re-analysis of the 1000 Genomes project. Results showed significant variability in tool performance. R-GADA had the highest recall but low precision, while PennCNV was the most reliable in terms of precision and F1 score. EnsembleCNV improved recall by combining multiple callers but increased false positives. Overall, current tools, including new methods, do not outperform PennCNV in precise CNV detection. Improved reference data and consensus on true positive CNV calls are necessary. This study provides valuable insights and scalable workflows for researchers selecting CNV detection methods in future studies. •Novel tools such as EnsembleCNV, which integrates calls from multiple methods, do not yet outperform PennCNV in balancing precision and recall.•The significant lack of consensus among CNV calling results highlights the need for careful interpretation and integration of high-confidence calls.•We present reproducible CNV calling workflows, which can uniform CNV detection across studies, and benchmarking performance analyses, which can be applied to other CNV detection tools.
Bibliography:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23
ISSN:	0888-7543 1089-8646 1089-8646
DOI:	10.1016/j.ygeno.2024.110962