Abstract Details


Poster 7: Chemical Patterns: A Visual Approach to Interpretation and Design

Karen Schomburg1, Lars Wetzer, Matthias Rarey (presenting)1
1University of Hamburg, Center for Bioinformatics, Bundesstrasse 43, 20146 Hamburg, Germany
Chemical patterns are the generalization of chemical structure. They appear everywhere in medicinal chemistry, from synthesis to ADME and bioactivity predictions. With the wealth of data available today, they are playing an increasing role for accessing compound databases beyond a simple compound or substructure query. Despite this importance, the inspection and design of chemical patterns is by far non-trivial, leading to the situation that this powerful technique for data description and access is mostly hidden from the bench chemists.
A major reason is that while there is the quasi-standard SMARTS [1] for writing patterns a graphical representation is lacking. Since chemists mostly think about chemical structures via structure diagrams, a similar graphical representation would simplify the interpretation of patterns substantially. In 2010, we started developing a graphical language for chemical patterns with two major design goals: The language should be close enough to structure diagrams such that chemists can easily understand them and it should be generic enough to capture the full expressive power of SMARTS. Meanwhile, the graphical language is published [2] and resulted in a software tool and a freely available web service named SMARTSviewer[3]. The SMARTSviewer simplifies the interpretation of patterns substantially, especially in cases where patterns have larger substructural components like the PAINS set of patterns characterizing problematic compounds for bioassays [4].
Recently, we also developed a graphical editor for the design of SMARTS patterns [5]. Based on the SMARTSviewer language, the editor allows to first draw a scaffold which can be decorated with chemical atom and bond specifications. The editor supports logical operators for combining properties, as well as the creation of atomic environment descriptions (recursive SMARTS). In an intuitive way, the editor guides through the pattern design process and can be used even without knowledge of SMARTS. For more experienced users, the design process becomes much more efficient, for example by the alternate writing of textual SMARTS and graphical editing as well as by the use of customizable template libraries. Within this presentation, several examples of pattern interpretation and design are given.

[1] James, C.A., Weininger, D., Daylight Theory Manual. Daylight Chemical Information Systems, Inc. of Aliso Viejo, CA, available at www.daylight.com
[2] Schomburg, K., Ehrlich, H.-C., Stierand, K., Rarey, M., From Structure Diagrams to Visual Chemical Patterns, J. Chem. Inf. Model., 2010, 50 (9), 1529-1535
[3] www.smartsview.de
[4] Baell, J.B., Holloway, G.A., New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays. J. Med. Chem., 2010, 53 (7). 2719-2740
[5] Schomburg, K.T., Wetzer, L., Rarey, M., Interactive design of generic chemical patterns. Drug Discov. Today, accepted Dec 2012

Return to Programme