If the affinities and docking modes of drugs to proteins are predicted before the drugs are synthesized, the development of drugs would be much more efficient. Also, by looking at the binding mode, the organic chemist can find ways to optimize drug structure more logically. Thus, it is very important to predict docking modes correctly. In our laboratory, docking programs have been developed and some of them are very useful, but it is still very difficult to predict docking modes when the molecule is flexible. For searching enormous conformation and configuration space efficiently, the searching and minimzation algorithms can be treated separately. There are several reports that more than 90% of the computation time is spent for minimization in the docking program; as the optimization of minimization would affect the computation time directly and significantly. So the minimization technique is of the most interest. Currently, I am writing an interface program for studying molecules using C++ with the STL and GNU libraries. To write a simple, well-structured and encapsulated interface program is important for beginning to develop relatively complicated programs without spending too much time removing bugs. It will take one month to make the basic interface for flexible molecules with several energy estimation methods. I then want to write a more rapid minimization program based on the precalculation of information to avoid redundant calculations. It is necessary to look at the binding modes or configurations of molecules by computer graphics, and for this purpose I need to use the Computer Graphics Laboratory computers.
Showing the most recent 10 out of 508 publications
