|
|
|
|
|
|
|
|
|
|
| Randy J. Zauhar, University of the Sciences in Philadelphia |
|
| Randy J. Zauhar received a BS in Physics and BA in Anthropology from Eastern Washington University, and an MS and PhD in Molecular & Cell Biology from Penn State University. After a postdoctoral position at the Mount Sinai School of Medicine in New York, he returned to Penn State where he eventually directed the Center for Computational Biology at the Biotechnology Institute. After several years as an industry scientist with Tripos involved in the development of software for the pharmaceutical industry, he returned to academics and is currently an Associate Professor of Biochemistry at the University of the Sciences in Philadelphia, where he also directs the Master's program in Bioinformatics. His research interests range from molecular biophysics to computer-aided drug design.
|
|
Shape Signatures: A Tool for Rapid In Silico Screening and Clustering
Randy Zauhar*, Peter Meek*, Izabela Hartman*, ZhiWei Liu*, LiFeng Tian* & William Welsh+
*Department of Chemistry & Biochemistry, University of the Sciences in Philadelphia
+Departmant of Pharmacology, University of Medicine & Dentistry of New Jersey
Shape Signatures is a new method for compactly encoding shape and property information for ligand molecules and target receptor sites, and for making rapid comparisons of shapes and properties. The method uses an approach much like ray tracing to explore the shapes of molecules and protein binding sites; our Shape Signatures are simply probability distributions derived from the random path followed by a ray as it is propagated at the molecular surface by the laws of optical reflection. By using the ray to sample property information (such as the electrostatic potential) on the surface, the Signatures can incorporate this type of information as well. In contrast to methods that use a grid-based approach to represent shape, the Shape Signatures are independent of molecular orientation and can be very rapidly compared. Shape Signatures can be used to identify molecules in a database that are similar in shape and electrostatic properties to known bioactive compounds, and to find compounds likely to be complementary in shape to a receptor site. Shape Signatures can also be used to cluster compounds on the basis of similarity between Signatures.
We will present results from several drug-design and clustering studies, with focus on the design of new COX inhibitors and identification of estrogenic compounds, using both the NCI and ZINC databases augmented with Shape Signatures.
|
|
|
|
|
|
|
|
|
