Prediction and interpretation of deleterious coding variants in terms of protein structural stability

Abstract

The classification of human genetic variants into deleterious and neutral is a challenging issue, whose complexity is rooted in the large variety of biophysical mechanisms that can be responsible for disease conditions. For non-synonymous mutations in structured proteins, one of these is the protein stability change, which can lead to loss of protein structure or function. We developed a stability-driven knowledge-based classifier that uses protein structure, artificial neural networks and solvent accessibility-dependent combinations of statistical potentials to predict whether destabilizing or stabilizing mutations are disease-causing. Our predictor yields a balanced accuracy of 71% in cross validation. As expected, it has a very high positive predictive value of 89%: it predicts with high accuracy the subset of mutations that are deleterious because of stability issues, but is by construction unable of classifying variants that are deleterious for other reasons. Its combination with an evolutionary-based predictor increases the balanced accuracy up to 75%, and allowed predicting more than 1/4 of the variants with 95% positive predictive value. Our method, called SNPMuSiC, can be used with both experimental and modeled structures and compares favorably with other prediction tools on several independent test sets. It constitutes a step towards interpreting variant effects at the molecular scale. SNPMuSiC is freely available at https://soft.dezyme.com/.

Complementary computational and experimental evaluation of missense variants in the ROMK potassium channel

“…In recent years, several studies-some accompanied by publicly accessible tools-have supported the idea that structural data, in addition to classical conservation analyses based on multiple sequence alignments (MSAs), provide a means to predict the effect of single amino acid substitutions on biological function [17][18][19][20]. We recently showed that a consideration of not only structure but also structural dynamics can further improve the accuracy of these predictions [21].…”

Benchmarking analysis of deleterious SNP prediction tools on CYP2D6 enzyme

“…SNPMusic, the stability-oriented knowledge-based classifier, uses protein structure, artificial neural networks, and combinations of solventdependent statistical potentials to predict whether destabilizing or stabilizing mutations cause disease (Ancien, Pucci, Godfroid, & Rooman, 2018).…”

Curating the gnomAD database: Report of novel variants in the globin‐coding genes and bioinformatics analysis

“…The main purpose of this study was to analyze the protein structural and functional impact of the sequence variants in the HBA1 , HBA2 , and HBB genes reported in the gnomAD database, and to improve the description of the observed clinical outcomes. Although there are many computational and bioinformatic methods available for predicting a possible pathogenic effect of an amino acid substitution (Ancien, Pucci, Godfroid, & Rooman, ; Mahmood et al, ), none of these consider the different phenotypes that can result as a consequence of these variations. Moreover, the differential impact of these changes in the different conformations of a protein should be considered (Juritz et al, ) and many current methods usually analyze only one structure per protein.…”