COMPUTER SIMULATIONS OF PROTEIN STRUCTURE AND DYNAMICS

点击翻译按钮获取中文摘要

The research described in this proposal focuses on the development and application of computational methods for studying the relationship between structure, dynamics and function of proteins. Computer simulations provide the most detailed theoretical approach available to study proteins at a molecular level. The goals of this proposal during the next grant period are in two areas: I. the integration of structure with dynamics information from solution NMR and computer simulations, and II. the study of the solvation of proteins and solvent effects on electrostatics by computer simulation. The first goal addresses the need to develop computational tools necessary to extract the most structural and dynamical information contained in rapidly advancing 2D-NMR experiments on proteins. In a continuing collaboration with experimental NMR groups, structure and dynamics studies will be carried out on several systems including defensin polypeptides and conotoxins, and the protein alpha-lactalbumin. The second goal addresses the need to develop a more basic understanding of how molecular forces determine protein structure and the structural changes which give rise to specific functions. Among the most important forces to be considered for such protein functions as binding, catalysis, and transport, are electrostatic interactions. Methods are being developed for improving the way these interactions are treated in microscopic simulations based on the use of the reaction field model. These methods will be applied to such problems as simulating pKa shifts in doubly charged acids and zwitterions, analyzing solvation effects on the alpha-helix dipole, and to mapping the surface properties of Azurin, an electron transfer protein. The health relatedness of this research is in the more accurate modeling of the properties of proteins in solution and the use of improved computational tools for studying protein structure and function with rational drug design strategies based on modeling.