MoBiDiC Prioritization Algorithm, a Free, Accessible, and Efficient Pipeline for Single-Nucleotide Variant Annotation and Prioritization for Next-Generation Sequencing Routine Molecular Diagnosis

Publication date: July 2018Source: The Journal of Molecular Diagnostics, Volume 20, Issue 4Author(s): Kevin Yauy, David Baux, Henri Pegeot, Charles Van Goethem, Charly Mathieu, Thomas Guignard, Raul Juntas Morales, Delphine Lacourt, Martin Krahn, Vilma-Lotta Lehtokari, Gisele Bonne, Sylvie Tuffery-Giraud, Michel Koenig, Mireille CosséeInterpretation of next-generation sequencing constitutes the main limitation of molecular diagnostics. In diagnosing myopathies and muscular dystrophies, another issue is efficiency in predicting the pathogenicity of variants identified in large genes, especially TTN; current in silico prediction tools show limitations in predicting and ranking the numerous variants of such genes. We propose a variant-prioritization tool, the MoBiDiCprioritization algorithm (MPA). MPA is based on curated interpretation of data on previously reported variants, biological assumptions, and splice and missense predictors, and is used to prioritize all types of single-nucleotide variants. MPA was validated by comparing its sensitivity and specificity to those of dbNSFP database prediction tools, using a data set composed of DYSF, DMD, LMNA, NEB, and TTN variants extracted from expert-reviewed and ExAC databases. MPA obtained the best annotation rates for missense and splice variants. As MPA aggregates the results from several predictors, individual predictor errors are counterweighted, improving the sensitivity and specificity of missense and splice variant predicti...
Source: The Journal of Molecular Diagnostics - Category: Pathology Source Type: research