Identification of pathogenic missense mutations using protein stability predictors

Abstract

Attempts at using protein structures to identify disease-causing mutations have been dominated by the idea that most pathogenic mutations are disruptive at a structural level. Therefore, computational stability predictors, which assess whether a mutation is likely to be stabilising or destabilising to protein structure, have been commonly used when evaluating new candidate disease variants, despite not having been developed specifically for this purpose. We therefore tested 13 different stability predictors for their ability to discriminate between pathogenic and putatively benign missense variants. We find that one method, FoldX, significantly outperforms all other predictors in the identification of disease variants. Moreover, we demonstrate that employing predicted absolute energy change scores improves performance of nearly all predictors in distinguishing pathogenic from benign variants. Importantly, however, we observe that the utility of computational stability predictors is highly heterogeneous across different proteins, and that they are all inferior to the best performing variant effect predictors for identifying pathogenic mutations. We suggest that this is largely due to alternate molecular mechanisms other than protein destabilisation underlying many pathogenic mutations. Thus, better ways of incorporating protein structural information and molecular mechanisms into computational variant effect predictors will be required for improved disease variant prioritisation.

Insight into molecular characteristics of SARS-CoV-2 spike protein following D614G point mutation, a molecular dynamics study

“…Also, the score function module from Rosetta (version 3.12) (Combs et al, 2018 ), a well-known package in peptide design in parallel was used to calculate and compare the energy of the structure in the wild and mutant spike. Both FoldX and Rosetta tools calculate the energy of the structure by applying linear combination of physics and statistics-based energy terms algorithms (Gerasimavicius et al, 2020 ). Tools listed in our study provide useful information about the characteristics of protein conformations by calculating the different terms of energy.…”

Understanding the origins of loss of protein function by analyzing the effects of thousands of variants on activity and abundance

“…Indeed, cellular studies of disease-causing variants in a number of genes have shown that many variants are degraded in the cell (Meacham et al, 2001;Yaguchi et al, 2004;Olzmann et al, 2004;Ron and Horowitz, 2005;Yang et al, 2011Yang et al, , 2013Arlow et al, 2013;Nielsen et al, 2017;Chen et al, 2017;Matreyek et al, 2018;Scheller et al, 2019;Abildgaard et al, 2019;Suiter et al, 2020). For this reason, several methods for predicting and understanding disease-causing variants include predictions of changes in protein stability (Yue et al, 2005;De Baets et al, 2012;Casadio et al, 2011;Ancien et al, 2018;Wagih et al, 2018;Gerasimavicius et al, 2020). While stability-based predictions can be relatively successful and may provide mechanistic insight into the origins of disease, it is also clear that variants can cause disease via other mechanisms such as removing key residues in an active site or perturbing interactions or regulatory mechanisms.…”

Understanding protein structural changes for oncogenic missense variants

“…This is consistent with the observations from Figures 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , and 17 , which do not show any consistent changes in the overall protein structures upon mutation, precluding any further statistical analysis. While we did not find overall principles that can be used to classify WT and oncogenic variables in this dataset, the judicious use of 3D structure prediction methods remains a valuable tool to understand oncogenic mutations further, as demonstrated previously [ 11 , 12 , 32 , 33 , 34 , 35 ]. We recognize that comparing these results with those that could be obtained from neutral non-pathogenic variants could provide more details on the problem and perhaps could help in differentiating the structural effects in different types of variants.…”