NAR Molecular Biology Database Collection entry number 111

http://www.biocheminfo.org/klotho/

Stuppy J.¹, Kling S.¹, Holcomb J.², Dunford-Shore B.², Sadekar S.¹, Cannon B.¹, Abbott J.¹, Tullock A.¹, Campbell T.¹, Wilson S.¹, Fabrizio F.², Wise W.B.¹ and Kazic T.¹

¹Department of Computer Science, University of Missouri, Columbia, MO, USA
²Institute for Biomedical Computing, Washington University, St. Louis, MO, USA

Contact toni@athe.wustl.edu

Klotho is a system to define and recognize the configurations of molecules. A compilation of the structures of small organic molecules involved in enzymatic reactions is available at Klotho: Biochemical Compounds Declarative Database. All structures assume a solvent pH of 7.0.

The purpose of Klotho is to provide methods to accurately specify structural information about reactants and to support a wide variety of structural and functional queries. Klotho uses a multi-layered graph grammar to specify each compound in its database. The grammar is context-sensitive to account for changes in valence, charge, and bond direction as an atom's neighboring groups and the architecture of its including molecule change. The grammar permits the generation of specific compounds and classes of compounds and the recognition and classification of architectures, structures, and functional groups. The grammar operates on constructs that specify the stereochemical configuration of a molecule or group of molecules. The most abstract, highest layer of the grammar's molecular specification is the configuration (config) rule. Each config rule specifies the architecture of the compound as being a chain, a ring, or a ring system and indicates the bond and stereo connectivity of any substituent groups. Terminals in this layer become nonterminals in the layer below it. This next layer of the grammar's output is a full expansion of the config rule showing all atoms, bonds, charges, atom numbers, and relative three-dimensional bond directions as a list of edges. The bottom layer rearranges these data into a more convenient, more easily read form (a key-list data structure). Additional output formats are provided. A query on the web site for a given compound returns that compound's synonyms, structural diagrams, other output formats (e.g., SMILES strings and PDB files) and representations from the top and bottom levels of the grammar. All output data are freely available for download and the grammar is available upon request.

This work is supported by the NIH (R01-GM56529) and NSF (DBI-9117005).

Oxford University Press is not responsible for the content of external internet sites