Data mining the drug database: Is there such a thing as drug-like space?

COMP 334

Chihae Yang, cyang@leadscope.com1, Guangyu Sun1, David Bower1, and James F. Rathman, rathman.1@osu.edu2. (1) Leadscope, Inc, 1393 Dublin Road, Columbus, OH 43215, (2) Department of Chemical and Biomolecular Engineering, The Ohio State University, Koffolt Laboratories, 140 West 19th Avenue, Columbus, OH 43210-1110
The goal of the drug database is to assist chemists in leveraging past successes as a starting point for finding new drugs during the discovery process. A database of marketed drugs has been created to provide a detailed source of information for clean pharmaceutical compounds. Part of the historical information was provided from FDA collaborators through a Cooperative Research and Development Agreement. Based on this drugs database, we analyze the chemical space covered by the drugs in different therapeutic indications over time from 1970 to the present. Classifying these structures into indication categories defined by the National Library Medicine results in scaffolds representing the therapeutic indications. Chemical space analysis is performed using structural feature descriptors provided by Leadscope. One goal of this work is to explore how compound clustering in this new “drug space” compares with analysis of the same compounds in a feature space defined by a conventional chemical database. Various analysis methods including principal component analysis and partial ordinal modeling methods are employed to model therapeutic categories in terms of chemical features.
 

Poster Session
6:00 PM-8:00 PM, Tuesday, August 21, 2007 BCEC -- Ballroom Foyer, Poster

Sci-Mix
8:00 PM-10:00 PM, Monday, August 20, 2007 BCEC -- Exhibit Hall - B2, Sci-Mix

Division of Computers in Chemistry

The 234th ACS National Meeting, Boston, MA, August 19-23, 2007