(1) Preclinical Research,
Merck KGaA, 64271 Darmstadt, Germany.
(2) GMD-IPSI, 64293 Darmstadt Germany.
New biomolecular and computational methods have led to an explosion in the available data about receptor/ligand complexes. Unfortunately, existing data is spread over different databases or is not available in electronic format at all. Therefore we have developed a comprehensive database system for receptor/ligand complexes which combines crystallographic information from the Brookhaven Protein Databank (PDB) with biomolecular information like sequence alignments and information about mutations from various other data sources. RELIBase allows complex queries (including substructure searches, similarity searches and searches for specific interactions) regarding both small molecule and protein aspects.
As a technical platform an object-oriented database management system is used, which offers a structured data model, flexible retrieval mechanisms, persistent storage and sharing of data in a multi-user environment. A rather complex database schema was designed as the foundation for linking heterogenic entities. The receptor/ligand complexes in the PDB constitute the core data of the RELIBase. The relevant information is identified and extracted using an automatic parser. In order to allow a detailed chemical analysis of receptor/ligand interactions information about bond orders and hybridisation which is not stored in the PDB was defined for each ligand. Based on this database a detailed analysis of receptor ligand interactions has been undertaken.