Reaction prediction, classification, and retro-synthesis using a rule-based reaction expert system

CINF 78

Jonathan H. Chen, chenjh@uci.edu, Qian-Nan Hu, qhu@uci.edu, and Pierre Baldi, pfbaldi@uci.edu. Institute for Genomics and Bioinformatics, School of Information and Computer Sciences, University of California, Irvine, Irvine, CA 92697
A rule-based reaction expert system, initially developed to support computer-based learning in organic chemistry, has been expanded in scope to address a broader set of chemoinformatics problems ranging from reaction prediction and discovery, to retro-synthetic analysis and combinatorial library design, to automated annotation and classification of large reaction databases. The current system comprises over 1,200 manually-curated reaction patterns written using the SMIRKS language with transformation rule extensions to enable robust predictions. Applications of this expert system will be demonstrated in: (1) reaction prediction, including complete mechanism diagrams, for interactive learning in organic chemistry and validation of synthesis plans; (2) retro-synthetic analysis, with automated combinatorial library design, available over the Web; and (3) automated annotation and classification of a large database containing about 4M chemical reactions. Select applications available at http://cdb.ics.uci.edu.
 

Cheminformatics: From Teaching to Research
1:40 PM-5:20 PM, Tuesday, April 8, 2008 Marriott Convention Center -- Blaine Kern C, Oral

Sci-Mix
8:00 PM-10:00 PM, Monday, April 7, 2008 Morial Convention Center -- Hall A, Sci-Mix

Division of Chemical Information

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