Validation of copy number variation analysis for next-generation sequencing diagnostics

Abstract

Although a common cause of disease, copy number variants (CNVs) have not routinely been identified from next-generation sequencing (NGS) data in a clinical context. This study aimed to examine the sensitivity, and specificity of a widely used software package, ExomeDepth, to identify CNVs from targeted NGS datasets. We benchmarked the accuracy of CNV detection using ExomeDepth v1.1.16 applied to targeted NGS datasets, through comparison to CNV events detected through whole genome sequencing (WGS) for 25 individuals, and determined the sensitivity and specificity of ExomeDepth applied to these targeted NGS datasets to be 100% and 99.8%, respectively. To define quality assurance metrics for CNV surveillance through Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use:http://www.nature.com/authors/editorial_policies/license.html#terms Corresponding Author: Graeme CM Black, graeme.black@manchester.ac.uk. Competing interestsThe authors of this manuscript have no competing interests to declare.Authors' contributions JME, GCMB, SCR and BP designed and coordinated the study. JME, CC, StB, SaB, SG, RLG, PIS, BH, MM, ARW, WGN, SCR, GCMB contributed genetic and/or phenotypic data. JME wrote the manuscript and all authors provided important revisions and intellectual content. ExomeDepth, we undertook simulation of single exon (n=1000) and multiple-exon heterozygous deletion events (n=1749), determining a sensitivity of 97% (n=2749). We identified that the extent of sequencing coverage, the inter-and intra-sample variability in the depth of sequencing coverage, and the composition of analysis regions are all important determinants of successful CNV surveillance through ExomeDepth. We then applied these quality assurance metrics during CNV surveillance for 140 individuals across 12 distinct clinical areas, encompassing over 500 potential rare disease diagnoses. All 140 individuals lacked molecular diagnoses after routine clinical NGS testing, and through application of ExomeDepth we identified 17 CNVs contributing to the cause of a Mendelian disorder. Our findings support the integration of CNV detection using ExomeDepth v1.1.16 with routine targeted NGS diagnostic services for Mendelian disorders. Implementation of this strategy increases diagnostic yields and enhances clinical care. Europe PMC Funders Group

Atlas-CNV: a validated approach to call Single-Exon CNVs in the eMERGESeq gene panel

“…Prior studies have demonstrated successful CNV detection in DNA sequence data generated from gene panels, but the burden of analyzing single-exon CNVs is a challenge because of high false-positive rates 5,6 . Recently developed tools such as CoNVaDING 10 and modified versions of ExomeDepth 11,12 , have been designed to identify single-exon CNVs, but lack quality metrics that enable differentiation of different levels of confidence. As a result, clinical laboratories may often ignore single-exon CNVs.…”

“…The ability to detect CNVs accurately is critical for both genetic diagnostics and to advance understanding their impact on gene function. Next-generation sequencing (NGS) based targeted gene panels are commonly used in clinical genetic testing and various methods [5][6][7][8][9][10][11][12][13] have been developed to identify exonic CNVs in gene panel sequence data. Gene panels afford a qualitatively different opportunity to assess small CNVs due to their typically deeper sequence coverage when compared with Whole Genome or Exome Sequencing (WES).…”

“…Existing CNV tools that have been validated for the clinical setting include VisCap 5 , CoNVaDING 10 , and DeCON 11 , ExomeDepth 12 , and others 6,7,13,14 . CoNVaDING and ExomeDepth are methods which are successful at detecting single-exon CNVs by evaluating individual exon suitability for variant detection and selecting highly correlated samples as reference controls.…”

Molecular genetic testing methodologies in hematopoietic diseases: current and future methods

“…Besides providing point mutational and indel data as in Sanger sequencing, these technologies can also provide information regarding copy number variants (CNV) and structural gene rearrangements. Validation of mutation, CNV, and gene rearrangement calling is complex and requires significant bioinformatics expertise and experience with clinical validation.…”

“…Additionally, currently next-generation sequencing methods are in widespread use clinically. Third-generation sequencing methods are just beginning to emerge as clinically relevant, though heavily investigated on the translational research frontier.Besides providing point mutational and indel data as in Sanger sequencing, these technologies can also provide information regarding copy number variants (CNV)52 and structural gene rearrangements. Validation of mutation, CNV, and gene rearrangement calling is complex and requires significant bioinformatics expertise and experience with clinical validation.…”