Time-resolved molecular dynamics calculations produce an ensemble of conformations for a macromolecule. Some of these structures are more probable than others. Eliminating structures with low probability and grouping more probable structures into families produces a set of conformations that are relevant. These conformations are good tagents for drug design. I am developing a set of programs that will statistically reduce an ensemble of conformations into a group of relevant families. This technique will allow the most likely conformation in an ensemble to be determined. The most likely structure of a molecule is valuable for drug design. The average structure of a molecular ensemble contains the structural information of the entire ensemble and may not represent a conformation that has any physical relevance. The most probable structure represnts a conformation that the molecule will adopt some fraction of the time. The structural ensemble will be represented by families of macromolecules. Each family will have a different level of opacity. More probable structures will be more opaque, while less probable structures will be more translucent. These programs will be written in C and C++ using the developement tools available in the Computer Graphics Lab.
Showing the most recent 10 out of 508 publications
