Research on Docking Approaches and Simulation of Protein
|School||Harbin Institute of Technology|
|Course||Computer Science and Technology|
|Keywords||Protein - protein docking Genetic Algorithms Filter Scoring function Binding free energy|
The study of protein - protein interactions and recognition is an important topic in the field of molecular biology. Experimentally determined protein complex structure is difficult, therefore, binding mode between the computer simulation predicted molecular docking method is of great concern. The ultimate purpose of the molecular docking is predicted through computer simulation the unknown composite structure of the receptor and the ligand binding between. However, in the protein - protein docking system atoms and the increase of the number of degrees of freedom so that the analog composite formed in the conformational change is very difficult. In addition, the search process may produce millions docking conformation preliminary screening, before scoring with a fine energy function, and how these structures in order to get as much as possible near-native conformation is challenging problem in the current method study . Firstly, on the basis of in-depth study of protein docking theory, given a fast and efficient protein - protein docking method, using an evolutionary model based on information entropy protein docking, including the traditional docking model of optimization goals, as well as information entropy The composition of the objective function. Information entropy is used to measure the optimization process of the optimal solution in which a population of uncertainty, with the optimization process approximation to the optimal solution, this uncertainty is gradually eliminated, along with the optimal solution to the entropy Less The minimum value of zero, and by studying the genetic algorithm to solve the problem used key technologies, protein docking problems thinking to solve a variety of group-based genetic algorithm to solve the model using genetic algorithms. Second, the filter protein docking structure, the use of geometric matching filter combined with the the interface residue pair preference of double filtration technology, screening near natural structure to significantly increase the number of, and improve the success rate of identification and prediction of protein molecules ; purpose of the scoring function, is the structure within a reasonable period of time to make an effective evaluation to distinguish between a nearly native conformation. The scoring function used in many docking algorithm is based on thermodynamics assumption that the complex structure of the natural protein is a combination of a free energy minimum conformation. Therefore, the development of a computational fast and accurate enough binding free energy calculation method is a very important step in structure prediction. This article by the protein docking process, the electrostatic interaction between the protein and go hydration free energy analysis, free energy-based scoring function and reasonable evaluation of right and wrong docking structure made of near-native conformation distinction. Finally, this paper design and implementation of a simple graphic simulation of protein docking system, further validates the feasibility and effectiveness of the algorithm given in the paper.