Atlas-CNV: a validated approach to call single-exon CNVs in the eMERGESeq gene panel.

TitleAtlas-CNV: a validated approach to call single-exon CNVs in the eMERGESeq gene panel.
Publication TypeJournal Article
Year of Publication2019
AuthorsChiang, T, Liu, X, Wu, T-J, Hu, J, Sedlazeck, FJ, White, S, Schaid, D, de Andrade, M, Jarvik, GP, Crosslin, D, Stanaway, I, Carrell, DS, Connolly, JJ, Hakonarson, H, Groopman, EE, Gharavi, AG, Fedotov, A, Bi, W, Leduc, MS, Murdock, DR, Jiang, Y, Meng, L, Eng, CM, Wen, S, Yang, Y, Muzny, DM, Boerwinkle, E, Salerno, W, Venner, E, Gibbs, RA
JournalGenet Med
Date Published2019 Mar 20
ISSN1530-0366
Abstract

PURPOSE: To provide a validated method to confidently identify exon-containing copy-number variants (CNVs), with a low false discovery rate (FDR), in targeted sequencing data from a clinical laboratory with particular focus on single-exon CNVs.

METHODS: DNA sequence coverage data are normalized within each sample and subsequently exonic CNVs are identified in a batch of samples, when the target log ratio of the sample to the batch median exceeds defined thresholds. The quality of exonic CNV calls is assessed by C-scores (Z-like scores) using thresholds derived from gold standard samples and simulation studies. We integrate an ExonQC threshold to lower FDR and compare performance with alternate software (VisCap).

RESULTS: Thirteen CNVs were used as a truth set to validate Atlas-CNV and compared with VisCap. We demonstrated FDR reduction in validation, simulation, and 10,926 eMERGESeq samples without sensitivity loss. Sixty-four multiexon and 29 single-exon CNVs with high C-scores were assessed by Multiplex Ligation-dependent Probe Amplification (MLPA).

CONCLUSION: Atlas-CNV is validated as a method to identify exonic CNVs in targeted sequencing data generated in the clinical laboratory. The ExonQC and C-score assignment can reduce FDR (identification of targets with high variance) and improve calling accuracy of single-exon CNVs respectively. We propose guidelines and criteria to identify high confidence single-exon CNVs.

DOI10.1038/s41436-019-0475-4
Alternate JournalGenet. Med.
PubMed ID30890783
Grant ListU01HG8664 / / National Human Genome Research Institute /
HG008898 / / National Human Genome Research Institute /
HG006542 / / National Human Genome Research Institute /