Molecular Docking Based On Optimal Search Theory Research

With people’s attention, pharmaceutical industry has an tremendous development. Due to the outstanding performance and highly efficient of finding lead compounds, molecular docking has been widely used in drug researching. Optimization of conformational search algorithm and the scoring function in the molecular docking is the main content of this thesis.First of all, a summary of status of molecular docking in drug design is introduced, then thesis analyzes the difference between traditional drug design process and rational drug design process, also emphasizes the significance of molecular docking in rational drug design process. After intensive research of the theory of molecular docking, constructing a new optimal model for the molecular docking with optimization methods.In the inspiration of the Lamarckian genetic algorithm, a conformational search strategy named IGA model is proposed in molecular docking based on the immune genetic algorithm. Because of the genes restructuring, immune mutation, immune selection and immune memory, the IGA model can use the prior knowledge in the genetic process effectively. This algorithm improves the efficiency and the adaptability of molecular docking process, To verify the optimizing ability of IGA model, tests are conducted on seven protein-ligands complex from the Brookhaven protein database. Compared with the simulated annealing algorithm and the Lamarckian genetic algorithm, results show that the IGA model performs better than the rest.Compared with the classic search model, the molecular docking model can adapt the optimal methods in optimal search. Constructing the initial probability distribution function and the detection function for optimal search model is the key to establish the optimal search model. To perform the local search, the model employs the commonly used algorithm of Broyden-Fletcher-Goldfarb-Shanno(BFGS) for nonlinear unconstrained optimization, the basic idea of the proposed heuristic is to test an intermediate point ready for local search against a database which stored the information of other points which have been previously reached by local search.