Predicting and interpreting large scale mutagenesis data using analyses of protein stability and conservation

Abstract

Understanding and predicting the functional consequences of single amino acid is central in many areas of protein science. Here we collected and analysed experimental measurements of effects of >150,000 variants in 29 proteins. We used biophysical calculations to predict changes in stability for each variant, and assessed them in light of sequence conservation. We find that the sequence analyses give more accurate prediction of variant effects than predictions of stability, and that about half of the variants that show loss of function do so due to stability effects. We construct a machine learning model to predict variant effects from protein structure and sequence alignments, and show how the two sources of information are able to support one another. Together our results show how one can leverage large-scale experimental assessments of variant effects to gain deeper and general insights into the mechanisms that cause loss of function.

Rapid protein stability prediction using deep learning representations

“…4A). Thus, a substantial fraction of the pathogenic variants are predicted to be destabilized to an extent that is likely to cause lowered abundance ( Cagiada et al, 2021 ; Høie et al, 2022 ). Similarly, variants that are common in the human population (i.e.…”

Discovering functionally important sites in proteins

“…Previously we have used evolutionary analyses combined with stability calculations to predict variant effects on protein function and stability and showed how these measures correlate with changes in cellular abundance or function ( Cagiada et al, 2020 ; Høie et al, 2021 ). Here, we build on these ideas to construct a model to identify functional sites in proteins via the identification of SBI variants (Fig.…”

“…Analyses of large-scale mutagenesis studies have been used to probe the role of individual residues in the stability, abundance and function of a protein ( Gray et al, 2017 ; Dunham and Beltrao, 2021 ; Høie et al, 2021 ). Since most proteins need to be folded to function, it may, however, be difficult to deconvolute the effects of amino acid substitutions on intrinsic function from their effects on stability and cellular abundance ( Li and Lehner, 2020 ).…”

The “dark matter” of protein variants carries a distinct DNA signature and predicts damaging variant effects

“…We note that DMS datasets have previously been utilised to construct variant impact predictors with reasonable success (e.g. [25, 36, 37, 38]); here our work serves the additional purpose of validating the relevance of DNA signatures in annotating variant impact. It is also remarkable that the damaging mutational signature is visible only in the analysis of DMS data, but under-represented in mutational profiles obtained from tumour samples.…”

“…One way to address this issue is to perform “Deep Mutational Scanning” (DMS) [17, 18] where every position in the protein is mutated to any other 19 amino acids either in vitro or in silico , followed by assays to measure the stability and/or activity of the mutants. Experimental DMS have been applied on several cancer-related proteins and domains [19, 20, 21, 22]; computationally, variant impact predictors which incorporate sequence conservation (large protein/domain multiple sequence alignments), structural/physicochemical features (protein secondary/tertiary structures) and consequential representations of protein motions (imposing network models onto three-dimensional structures to account for molecular vibrations) have been developed, applied and evaluated in a DMS context [23, 24, 25, 26]. These computational and experimental data allows us to probe and access features of variants which reside in this mutational “dark matter”.…”