|
The study of protein structures is directly related not only to drug design and de novo design of artificial proteins with specific functions but also the elucidation of the pathogenic mechanism for the disease that is caused by misfolding of proteins (such as mad cow disease and Alzheimer's disease). The goal of the present project is to develop an efficient simulation algorithm for the prediction of the three-dimensional structures of proteins from the first principles. In order to attain the goal, we have not only to have an efficient method for evaluation of solvation effects but also a powerful simulation algorithm that can avoid getting trapped in states of energy local minima. For the former, the RISM theory has been extensively improved. For the latter, we have developed more than 10 generalized-ensemble algorithms, among which it was found that the replica-exchange multicanonical algorithm, multicanonical replica-exchange method, replica-exchange simulated tempering, and multidimensional replica-exchange method are particularly powerful and suitable for protein structure predictions. By the development of these new methods, we conclude that we succeeded in attaining our goal. Now that the Human Genome Project has just completed the determination of the DNA base sequences that corresponds to the amino acid sequences of the proteins, the prediction of the three-dimensional structures of proteins will be one of the most important topics of research. The accomplishment of the present project will thus be very useful in the post-genome era not only in basic science but also in bioindustry.
|