Evaluation of CNV detection tools for NGS panel data in genetic diagnostics

Abstract

Although germline copy-number variants (CNVs) are the genetic cause of multiple hereditary diseases, detecting them from targeted next-generation sequencing data (NGS) remains a challenge. Existing tools perform well for large CNVs but struggle with single and multi-exon alterations. The aim of this work is to evaluate CNV calling tools working on gene panel NGS data and their suitability as a screening step before orthogonal confirmation in genetic diagnostics strategies. Five tools (DECoN, CoNVaDING, panelcn.MOPS, ExomeDepth, and CODEX2) were tested against four genetic diagnostics datasets (two in-house and two external) for a total of 495 samples with 231 single and multi-exon validated CNVs. The evaluation was performed using the default and sensitivity-optimized parameters. Results showed that most tools were highly sensitive and specific, but the performance was dataset dependant. When evaluating them in our diagnostics scenario, DECoN and panelcn.MOPS detected all CNVs with the exception of one mosaic CNV missed by DECoN. However, DECoN outperformed panelcn.MOPS specificity achieving values greater than 0.90 when using the optimized parameters. In our in-house datasets, DECoN and panelcn.MOPS showed the highest performance for CNV screening before orthogonal confirmation. Benchmarking and optimization code is freely available at https://github.com/TranslationalBioinformaticsIGTP/CNVbenchmarkeR .

Fig. 1. Benchmark design and augmented datasets.

a The panel shows the benchmark design and the objective of applying the results in the diagnostics routine. b To evaluate the diagnostics scenario, a new dataset was built for each run belonging to the original dataset. The augmented datasets contained all the samples originally sequenced in the run and, in the case of the MiSeq datasets (upper), a set of 51 samples with no known CNV from different runs (MLPA multiplex ligation-dependent probe amplification; aCGH array comparative genomic hybridization; NGS next-generation sequencing; CNV copy-number variant).

Fig. 2. Benchmark results with default parameters: per ROI metrics.

Shows results when executing tools with the default parameters and computing the per ROI metrics. ExomeDepth and DECoN tools obtained same sensitivity and specificity in panelcnDataset (ROI region of interest; PPV positive predictive value; F1 F1 score).

Fig. 3. Optimization results at ROI level.

Shows sensitivity and specificity on validation sets when executing tools with the optimized parameters in comparison to the default parameters (ROI region of interest).

Fig. 4. Benchmark results for the diagnostics scenario: whole diagnostics strategy metrics.

Shows sensitivity and specificity on the augmented in-house datasets when executing tools with the optimized parameters in comparison to the default parameters.