The Research Of Osteoarthritis Chinese Medicine Molecular Docking Based On Particle Swarm Algorithm

With the very rapid development of computer-aided drug design technique, it has significantly accelerated the drug design process and effectively enhanced the efficiency of the process from drug discovery to clinical applications. At present, a lot of researchers are focusing on the computer-aided Chinese medicine molecular design. The thesis expands the simulation study and experiments for the interaction between traditional Chinese medicine compounds and targets. The purpose is to strive to explore the positive significance for the development of Chinese medicine mechanism as well as modern Chinese medicine. As one of the computer-aided Chinese medicine molecular design methods, molecular docking is to understand the mechanism of mutual from the cellular and molecular level of traditional Chinese medicine compound, and to provide some necessary theoretical basis for the treatment of diseases, development of new drugs and clinical validation. With the help of the molecular docking software, the subject puts the treatment drug molecules of osteoarthritis as the basic data source provided by Co-operation unit Fujian University of Traditional Chinese Medicine and completes the research between the osteoarthritis protein targets and drug molecules on the micro-level from three parts as follows.Firstly, by improving the conformational search algorithm of molecular docking software, the topic studies the docking between treatment osteoarthritis drug molecules and their targets. Based on the principle of molecular docking methods, the thesis uses empirical free energy as evaluation function and particle swarm optimization algorithm as the optimization search algorithm during the process of docking, at the same time, in order to maintain the diversity of particles and avoid the algorithm falling into a local minima, the algorithm combines with adaptive local search strategy to complete the search. After comparing it with other docking search algorithms, the paper verifies the effectiveness of the particle swarm optimization in docking conformational search.Secondly, the thesis parallels particle swarm algorithm and realizes the parallelization of molecular docking conformational search algorithm to reduce the docking time-consuming. Using SPMD and master-slave model as parallel programming ideas, the paper uses MPI programming to realize particle swarm coarse granularity parallel algorithm and applies it in the conformational search and analyzes, then compares the results with the work already done, at the same time, considers the experiment in parallel speedup and parallel efficiency. It shows that docking efficiency has been improved.Finally, in order to promote the parallel processing methods of molecular docking, the thesis performs multiple molecular docking tasks parallel on multiple processors. It puts forward a hybrid parallel docking combining with parallel particle swarm optimization algorithm and changes the parallel strategy according to the docking tasks and the number of processors flexibility. Ultimately, the thesis analyzes the practicality of this approach by computing the speedup and parallel efficiency.With the improved conformation search algorithm of molecular docking software AutoDock, the subject can get the effective molecular groups, thus can do further biological experiments to provide powerful reference for new drugs of osteoarthritis. The research focuses on the realization and analysis of the particle swarm algorithm and its parallel algorithm to highlight the role of the docking process between the treatment osteoarthritis drug molecules and their targets. It can provide meaningful computer analysis ideas and methods for the research of traditional Chinese medicine active ingredients identification and molecular docking virtual screening.