Drug Molecular Docking Based On Improved Genetic Algorithm

Molecular docking plays a very important role in drug design, and it’s also a hot topic all over the world. In recent years, with the development of computer technology and structural biology, as the computer recognition technology has been used, more and more three-dimensional structures of target for biological macromolecules have been surveyed and evaluated, which largely improve the medicine designing. Besides that, molecular docking costs less in time and money, compared with medicine selection and practical biological tests.The purpose of molecular docking is searching the most stable conformation for small molecules of receptor and ligand, at this time, the combination of energy is also at the low level. People have a more understanding as the development of biology. They found lock and key principle, induced fit theory, the ligand receptor mode of induced fit theory-induced theory and so on, establishing rigid model, flexible model and the mixing of the two models. However, even the simple small organic molecules have the high degree of freedom. As to the receptor macromolecules, we have to consider its freedom, which leads to the complication of calculation. Moreover, the paper focuses on genetic algorithms, so when we establish the model, we must take the flexibility of small molecules into consideration.In the beginning of this paper we establish the model of the computer-aided drug design and the solve the model by the numerical optimization algorithm.In the terms of algorithm, our paper is based on information entropy of a variety of population genetic algorithm, and we put latin hypercube sampling method into the algorithm, then improve the algorithm, being a new algorithm which accelerates genetic evolution. The algorithm adapts a variety of group strategy and optimal retention policy, raising the ability of algorithm which improve the efficiency of docking. Finally, In the paper we put the new genetic algorithm into the software which be designed for the molecular docking optimization, and make experimental test s with the real example. Experiments show that the algorithm has fast convergence and high stability.