Dataset Information

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 .

SUBMITTER: Moreno-Cabrera JM

PROVIDER: S-EPMC7784926 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

Moreno-Cabrera José Marcos JM Del Valle Jesús J Castellanos Elisabeth E Feliubadaló Lidia L Pineda Marta M Brunet Joan J Serra Eduard E Capellà Gabriel G Lázaro Conxi C Gel Bernat B

European journal of human genetics : EJHG 20200619 12

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, CoN ...[more]

PMID: 32561899

Similar Datasets

Project description:Familial hypercholesterolemia (FH) is an inherited, autosomal dominant metabolic disorder mostly associated with disease-causing variant in LDLR, APOB or PCSK9. Although the dominant changes are small-scale missense, frameshift and splicing variants, approximately 10% of molecularly defined FH cases are due to copy number variations (CNVs). The first-line strategy is to identify possible pathogenic SNVs (single nucleotide variants) using multiple PCR, Sanger sequencing, or with more comprehensive approaches, such as NGS (next-generation sequencing), WES (whole-exome sequencing) or WGS (whole-genome sequencing). The gold standard for CNV detection in genetic diagnostics are MLPA (multiplex ligation-dependent amplification) or aCGH (array-based comparative genome hybridization). However, faster and simpler analyses are needed. Therefore, it has been proposed that NGS data can be searched to analyze CNV variants. The aim of the study was to identify novel CNV changes in FH patients without detected pathogenic SNVs using targeted sequencing and evaluation of CNV calling tool (DECoN) working on gene panel NGS data; the study also assesses its suitability as a screening step in genetic diagnostics. A group of 136 adult and child patients were recruited for the present study. The inclusion criteria comprised at least “possible FH” according to the Simon Broome diagnostic criteria in children and the DLCN (Dutch Lipid Clinical Network) criteria in adults. NGS analysis revealed potentially pathogenic SNVs in 57 patients. Thirty selected patients without a positive finding from NGS were subjected to MLPA analysis; ten of these revealed possibly pathogenic CNVs. Nine patients were found to harbor exons 4−8 duplication, two harbored exons 6−8 deletion and one demonstrated exon 9−10 deletion in LDLR. To test the DECoN program, the whole study group was referred for bioinformatic analysis. The DECoN program detected duplication of exons 4−8 in the LDLR gene in two patients, whose genetic analysis was stopped after the NGS step. The integration of the two methods proved to be particularly valuable in a five-year-old girl presenting with extreme hypercholesterolemia, with both a pathogenic missense variant (c.1747C>T) and exons 9−10 deletion in LDLR. This is the first report of a heterozygous deletion of exons 9 and 10 co-occurring with SNV. Our results suggest that the NGS-based approach has the potential to identify large-scale variation in the LDLR gene and could be further applied to extend CNV screening in other FH-related genes. Nevertheless, the outcomes from the bioinformatic approach still need to be confirmed by MLPA; hence, the latter remains the reference method for assessing CNV in FH patients.

Project description:IntroductionTesting for homologous recombination deficiency (HRD) as a biomarker in relation to poly (ADP-ribose) polymerase inhibitor (PARPi) treatment in ovarian cancer is done by sequencing of the BRCA1/2 genes and/or by assessing a genomic instability signature. Here we present data obtained with two different methods for genomic instability testing: the Oncomine™ Comprehensive Assay Plus (OCA Plus) NGS panel and the OncoScan CNV assay.MethodsThe retrospective analytical study included 80 ovarian cancer samples of patients previously referred to clinical Myriad testing (reference cohort), and 50 ovarian cancer samples from patients collected as part of the Pelvic Mass study. OCA Plus NGS libraries were sequenced with the Ion S5™XL Sequencer and analyzed with the Ion Reporter™ Software v5.20 for calculation of the genomic instability metric (GIM). In addition, all samples were tested with the OncoScan CNV FFPE Assay and analyzed with a previously published R-algorithm for generation of an in-house genomic instability score (in-house GIS).ResultsThe OCA Plus assay had a concordance to the reference of 89% on samples with a tumor fraction ≥ 30% (auto-calculated or via molecular estimation). A total of 15 samples in the reference cohort had a calculated tumor fraction < 30% in the OCA Plus assay. In these, the concordance to reference was only 60%. For the OncoScan CNV in-house GIS a local cutoff point of ≥ 50 was calculated. This gave a concordance to the reference of 85%, with 91% of the samples in the reference cohort passing quality control (QC) on tumor fraction. Both assays had a high sensitivity for the detection of genomic instability in samples with pathogenic or likely pathogenic BRCA1/2 mutations, with 12/13 being GIM positive (OCA Plus assay) and 13/13 being in-house GIS positive (OncoScan CNV assay).ConclusionsThe OCA Plus assay and the OncoScan CNV assay show a high but not complete concordance to reference standard homologous recombination deficiency (HRD) detection. The main reason for QC failure or non-concordance in our study was a low tumor fraction estimated in the assay, despite the selection of material by a pathologist with an inclusion criterion of > 30% tumor. QC steps should include careful tumor content evaluation, and results on samples with < 30% tumor should not be reported.

Dataset Information

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

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets