Consistent count region–copy number variation (CCR-CNV): an expandable and robust tool for clinical diagnosis of copy number variation at the exon level using next-generation sequencing data

Purpose: Despite the importance of exonic copy number variations (CNVs) in human genetic diseases, reliable next-generation sequencing–based methods for detecting them are unavailable. We developed an expandable and robust exonic CNV detection tool called consistent count region (CCR)–CNV. Methods: In total, about 1000 samples of the truth set were used for validating CCR-CNV. We compared CCR-CNV performance with 2 well-known CNV tools. Finally, to overcome the limitations of CCR-CNV, we devised a combined approach. Results: The mean sensitivity and specificity of CCR-CNV alone were above 95%, which was superior to that of other CNV tools, such as DECoN and Atlas-CNV. However, low covered region and positive predictive value and high false discovery rate act as obstacles to its use in clinical settings. The combined approach showed much improved performance than CCR-CNV alone. Conclusion: In this study, we present a novel diagnostic tool that allows the identification of exonic CNVs with high confidence using various reagents and clinical next-generation sequencing platforms. We validated this method using the largest multiple ligation-dependent probe amplification–confirmed data set, including sufficient copy normal control data. The approach, combined with existing CNV tools, allows the implementation of CCR-CNV in clinical settings.