Protein-protein interaction prediction using co-evolution based method

Zurich Seminars in Bioinformatics - Tao Fang

• 12:15 om ZOOM Call

It has been found that protein families consisting of orthologous proteins usually show divergent sequence similarities due to amino acid mutations in evolutionary history while structures of protein families are usually much more conserved. This phenomenon can be explained by two types of mutant positions in the protein sequence. One is conserved regions across the protein family and even a slight change in this region could result in critical loss of protein structure or function. Another is the so-called compensated mutation or coevolution region in protein. This means if one residue mutates, other residues that in close contact or functionally related to the mutant residue often need to show compensated mutation to locally restabilize protein structure and function.

This type of residue-residue coevolution can be and have been intensively used to study protein three-dimensional (3D) structure and functions within the same protein (family), while considering intra-and inter-protein co‑evolution involve similar physical interaction forces and evolutionary constraints, I’m trying to apply and adapt these methods to predict protein-protein interactions networks.