Fast and accurate prediction of small-molecule 3-D structures

CINF 89

Ryan W. Benz, rbenz@uci.edu, School of Information and Computer Sciences, University of California, Irvine, ORU Genomics and Bioinformatics, Irvine, CA 92697-3445 and Pierre Baldi, pfbaldi@uci.edu, Institute for Genomics and Bioinformatics, School of Information and Computer Sciences, University of California, Irvine, Irvine, CA 92697.
Three-dimensional molecular representations are vital for a wide variety of computational chemistry and chemoinformatics methods. In this work, we present our latest efforts to improve upon existing computational 3D structure generation methods and produce an open, fast, and accurate 3D structure generator for small molecules. Our hierarchical method incorporates the use of rigid segment and torsion angle databases composed of nearly 2 million fragments from 400,000 experimental structures, along with fast numerical methods to cover gaps in the database when needed. This approach is robust and always returns a structure given a valid 2D input. The accuracy was assessed by comparing the RMSD between the generated and experimental structures, and was typically below 1 Ang. In addition to a description of the method and its performance, we will also present new efforts to incorporate machine learning into our generation method.
 

Challenges in Structure Searching
9:00 AM-11:30 AM, Wednesday, April 9, 2008 Marriott Convention Center -- Blaine Kern C, Oral

Division of Chemical Information

The 235th ACS National Meeting, New Orleans, LA, April 6-10, 2008