Toxicology databases in the metadatabank of online databases.

Databases in science and technology dealing with toxicology subjects are of great and increasing interest to the scientific community. International hosts pay their tribute in creating toxicology clusters as a user-aid. Toxicology clusters of DataStar, DIALOG, DIMDI and STN are listed and compared. At the German National Research Centre for Environment and Health an 'Information System for Environmental Chemicals' has been developed. Within this research project approximately 400 online databases have been evaluated and the results are stored in the Metadatabank of Online Databases called DADB. A further step in this project is to assist the users of these metadatabases to choose the most appropriate database(s) for a special query. The retrieval of information out of a multitude of differently structured databases can be optimized using different evaluation criteria. We distinguished three groups of evaluation criteria: (1) general evaluation criteria, e.g. size of database, update frequency, cost of online search, etc., (2) evaluation criteria based on the characterization of environmental chemicals, and (3) evaluation criteria based on environmental-relevant data-types. We studied the influence of a number of different criteria in order to rank a set of objects, in this particular case toxicology databases. We used a six-number scoring system. The scoring is carried out using the scientific background of lattice theory and its graphical evaluation. The application of this scoring system for toxicologically-relevant databases, which is still in an early development stage, is useful in selecting the most appropriate database(s) in accordance with special items of interest. Applying our five chosen criteria, the databank CA (Chemical Abstracts) turns out to be the best, followed by four other databanks, which cannot be compared with each other.