LIGAND - CHEMICAL DATABASE FOR ENZYME-REACTIONS

Motivation: The existing molecular biology databases focus on the sequ ence and structural aspects of biological macromolecules,i.e. DNAs, RN As and proteins. However; in order to understand the functional aspect s, it is essential to computerize the interaction of these molecules. Furthermore, living cells contain additional molecules, such as metabo lic compounds and metal ions, that may also be consider-ed as parts of the basic building blocks of life, but are not well organized in publ ic databases. LIGAND chemical database is our attempt to solve these p roblems, at least for enzymatic reactions. Results: LIGAND consists of two sections. ENZYME and COMPOUND. The ENZYME section is an extension of previous studies (Suyama et al., Comput. Applic. Biosci., 9, 9-15, 1993), and it is a flat-file representation of 3303 enzymes and 2976 enzymatic reactions in the chemical equation format that can be parsed by machine. The COMPOUND section has been newly constructed for infor mation on the nomenclature and chemical structures of compounds. It co ntains 5383 chemical compounds. Both ENZYME and COMPOUND entries conta in I ich cross-reference information, most of which is automatically g enerated by the DBGET/LinkDB system, thus providing the linkage betwee n chemical and biological databases. LIGAND is updated daily, tightly coupled with the KEGG metabolic pathway database, and forms the basis for reconstruction and computation of pathways, Availability: LIGAND c an be accessed through the DBGET/LinkDB and KEGG systems in the Japane se GenomeNet database service via http://www.genome.ad.jp/. The flat-f ile format of the LIGAND database can be downloaded by anonymous FTP v ia frp ://kegg.genome.adjp/molecules/ligand/. Contact: goto@kuicr:kyot o-u.ac.jp; nishioka@scl.kyoto-u.ac.jp; kanehisa@kuicr:kyoto-u.ac.jp.