Computational analysis and prediction of mechanisms of intragenic compensation of human pathogenic and bacterial resistance-associated mutations
Final Report Abstract
Single mutations that cause a change in the amino acid sequence of a protein are fundamental events in protein evolution. These events can be regarded as traversing a protein’s fitness landscape in its sequence space. Most such mutations entail adverse consequences for protein’s structure, function, stability, or affinity to inter-action partners. Compensation for such a deleterious event is possible either by a reverse mutation – another substitution at the same position back to the wildtype, – or by a mutation at a different protein site. Such compensatory (trailing) or permissive (preceding) mutations enable protein evolution: otherwise all sequences would be extremely conserved, since most mutations are deleterious. In this project, we aimed to explore the intragenic compensatory mechanisms be analyzing large genomic datasets from various sources: bacteria with associated resistance phenotypes and human exomes with a variety of disease-associated phenotypes. We proposed to develop machine-learning tools to predict compensatory effects of mutations by analysis of their spatial distribution in the protein three-dimensional structures, as well as their evolutionary history.
Publications
-
d-StructMAn: Containerized structural annotation on the scale from genetic variants to whole proteomes. GigaScience, 11.
Gress, Alexander; Srikakulam, Sanjay K.; Keller, Sebastian; Ramensky, Vasily & Kalinina, Olga V.
