Systematic and comprehensive benchmarking of an exome sequencing based germline copy-number analysis pipeline to detect clinically relevant CNVs

Purpose Detecting germline copy-number variants (CNVs) from exome sequencing (ES) is not a standard practice in clinical settings owing to several reasons concerning performance. We added a CNV pipeline to our clinical ES workflow and comprehensively characterized its performance. Methods We used a cohort of 387 individuals (351 probands and 36 family members) with both clinical chromosomal microarray (CMA) and ES data available to compare the CNV calls and comprehensively characterize the CNVs from ES. We excluded the exons with low mappability scores prior to variant calling to reduce the number of false positives. Reproducibility was assessed by running 1,000 iterations per sample using permuted random subsets of 200 controls. Results The ES-based CNV pipeline was 93% sensitive for all deletions and duplications as compared to the gold-standard baseline true-positive CNV calls defined by the CMA. The modified workflow resulted in a drastic reduction of the total number of CNVs identified per sample and the number of false positives while retaining a high sensitivity of 90%. The exome-based CNV pipeline was 100% sensitive for clinically-relevant, rare variants (including single exon deletions), and was highly reproducible. Conclusion Exome-based CNV detection can be reliably used in a clinical setting and can increase the overall diagnostic yield.