Prediction of Protein Fold and Analysis of Protein Folding Procedure

点击翻译按钮获取中文摘要

Our purpose is to predict the protein fold by computing the major free energy landscape of a protein molecule, since native conformations of proteins assume the conformations with low free energy values. In addition, the analysis of protein folding procedures are investigated.First, we chose two kinds of peptides with about 10 amino acids, and draw the free energy landscapes after enhanced conformational search. The resulted landscapes were consistent with the experiments of CD and NMR spectra.Further, we noticed that the landscapes were strongly dependent on the force field parameters, and so we made our effort on this problem. We developed a new algorithm to calibrate the free energy landscapes by the quantum chemical calculations, and we succeeded to calibrate several free energy landscapes, which were given by using different force fields, into the common landscape.In addition, we developed another method for the enhanced conformational search, by applying Tsallis statistics. When the parameter q in Tsallis statistics was treated as one of the variables for the motion of equation, the wide energy region can be covered by this new method for a simple peptide system composed of five amino acids.