Algorithm Optimization And Function Study Of Co-evolution Network Of A/B Hydrolases Subfamilies

In the simplest definition,co-evolution refers to the coordinated changes that occur in pairs of organisms or biomolecules,typically to maintain or to refine functional interactions between those pairs.As one of the computational algorithm focused on residue co-evolution,SCA extracts groups of co-evolution residues that are functionally related,called “Protein Sectors”,which can provide new ideas in rational design for novel protein functions.However,SCA is fully dependent on sequence information,which would cause low accuracy in high-throughput calculation.In this study,we optimized SCA algorithm by introducing structure information and MD simulation.Taking 8 CALB related subfamilies(belongs to α/β hydrolase family)as our studying object.We first did a structure alignment between the representative structures of these 8 subfamilies,then combined the iterative sequence alignment within each subfamily to create a high-quality alignment which consist of 2182 proteins.On the basis of the original algorithm,we optimized SCA with residual movement correlations,which was calculated from molecular dynamic simulations of CALB.The optimized method is called “SCA.MD”.By doing multiple comparisons between these 2 methods,such as the residual distribution in eigenvector space and on 3D structure,we found that SCA.MD performs significantly better than SCA.Moreover,for the first time,we combined literature digging and network analysis,predicted the function of sectors according to the correlations between sector residues and annotated functional residues.This enriches the analysis after co-evolution calculation and provides reliable reference for the directed evolution of CALB related subfamilies.