Mutafy: A webserver to identify high quality mutant protein structures in the Protein Data Bank

Abstract

Changes in the amino acid sequence of proteins resulting from nonsynonymous variants in the genome, can have significant effects on protein folding, stability, dynamics, and function, which may ultimately lead to diseases. The analysis of large sets of disease associated variants is a common approach for the study of pathogenic mechanisms. In-silico mutagenesis experiments based on wildtype structures of target proteins are a common approach to this aim, however these do not account for the effect of variants on folding and might not accurately reflect conformational changes. A growing number of experimentally solved protein structures harbouring disease-associated mutations, including single amino acid variants, are deposited in the worldwide Protein Data Bank (PDB). Nevertheless, identifying high-quality structures for specific missense variants of interest remains challenging due to the growing number of deposited protein structures in the PDB, and the lack of a dedicated interface and annotation system to search and retrieve mutant protein structures. As a result, mutant protein structures in the PDB are a powerful source of information which is largely underused. To address these shortcomings, we have developed Mutafy, a publicly available webserver to identify high quality mutant protein structures. Given input human genes, the webserver finds structures of the corresponding coded wildtype proteins and their available solved mutants, selects high quality structures, annotates them with information from biomedical databases to favour their interpretation and selection, and allows for the interactive exploration of the results and 3D visualisation. Mutafy is publicly available without requiring user registration at https://mutafy.rosalind.kcl.ac.uk.