Abstract
SUMMARY: Understanding how amino acid insertion mutations affect protein structure can inform pharmaceutical efforts targeting diseases that are caused by protein mutants. In silico simulation of mutations complements experiments performed on physical proteins which are time and cost prohibitive. We have computationally generated the exhaustive sets of two amino acid insertion mutations for five protein structures in the Protein Data Bank. To probe and identify how pairs of insertions affect structural stability and flexibility, we tally the count of hydrogen bonds and analyze a variety of metrics of each mutant. We identify hotspots where pairs of insertions have a pronounced effect, and study how amino acid properties such as size and type, and insertion into alpha helices, affect a protein's structure. The findings show that although there are some residues, Proline and Tryptophan specifically, which if inserted have a significant impact on the protein's structure, there is also a great deal of variance in the effects of the exhaustive insertions both for any single protein, and across the five proteins. That suggests that computational or otherwise quantitative efforts should consider large representative sample sizes especially when training models to make predictions about the effects of insertions. AVAILABILITY AND IMPLEMENTATION: The data underlying this article is available at https://multimute.cs.wwu.edu.