Protein-protein Docking Algorithm And Practice

Protein-protein interaction and recognition is an important subject in the field of molec-ular biology, Protein-Protein docking is one of the significant research content of this subject.Since it is difficult with experimental determination of protein docking complex structures,people pay close attention to the docking algorithms through computer simulating the bind-ing patterns of moleculars. So far, there have several successful protein docking algorithmssuch as ZDOCK, AutoDOCK, Rosetta computing platform, etc. In many protein-proteindocking algorithms, the most important scoring function is based on geometric matching, soit is very helpful to optimize docking algorithms with improving current geometric match-ing model. These protein docking algorithms have their own advantages and disadvantages,they generally have higher sensitivity with protein object computed.Protein docking algorithms have been widely used. Many important physiological pro-cesses within cells such as cell signaling, immune response, gene duplication, gene tran-scription, gene translation and cell cycle regulation are all regulated with protein-proteininteractions. In order to fully understanding and awareness of these complex physiologicalprocesses, we must parse the structures of protein complexes and understand the confor-mational changes in the protein interaction. Protein docking algorithms use computers tosimulate molecular docking, the output is the predicted complex structures. Further more,Some docking algorithms could calculate the molecular dynamics indicators in docking pro-cesses for further analysis.The main contribution of this paper is two parts of practices of protein docking algo-rithms. In concerned with the score function is based on geometric matching, we introducingthe concept of evolutionary couplings to optimize the geometric matching pattern in RosettaSymDock and present our EcSymDock algorithm. SymDock is a protocol of Rosetta used topredict protein oligomers’ structure. This protocol mainly considers oligomer’s symmetryfeature. Evolution Coupling(EC) is an optimized mutual information calculated by specific protein family MSA, it reveals certain contacts between residues. We combined EC informa-tion with SymDock protocol and present the EcSymDock method to add EC as constraintsinto SymDock protocol, let EC play the role of geometric distance constraints. After theexperiments of dimer structure prediction, the result shows our method has a better perfor-mance. Following is docking algorithm practice, by means of protein docking algorithms,we characterize the anti-cancer drug resistance reasons from the point of view of energy.EGFR(Epidermal growth factor receptor) mutation-induced drug resistance has significant-ly impaired the potency of small molecule tyrosine kinase inhibitors(TKIs) in lung cancertreatment. Computational approaches can provide powerful and efficient techniques in theinvestigation of drug resistance. In our work, the EGFR mutation feature is characterized bythe energy components of binding free energy(concerning the mutant-inhibitor complex).The dynamics of the bound mutant-inhibitor complex is simulated via AMBER and thebinding free energy of the complex is calculated based on the dynamics. Through analysis,we concluded that binding energy release strength determines the level of drug resistance.