Software interoperability, componentry, and reuse are acknowledged as important goals for the bioinformatics community. Nevertheless, numerous heterogeneous molecular biology database systems are developed without considering facilities provided by and/or experience gained during the development of similar systems. The main reason for this undesirable situation is the lack of an adequate framework for describing existing systems, including their architecture, policies and operational constraints.
Characterizations of existing heterogeneous molecular biology database systems are essential for sharing methods, mechanisms, and practical experience. Inadequate characterizations of existing systems conceal, for example, the fact that these systems share similar facilities (e.g., schema conversion, query processing) with object database interfaces such as those supported by Genera (Letovsky 1995) and OPM (Chen and Markowitz 1995a), and with each other.
In this paper, we proposed criteria for characterizing heterogeneous molecular biology database systems in terms of their assumptions, schema and data converters, extent of detecting and resolving semantic conflicts, type of database correlations, query interfaces and processing, and synchronization with component databases. Initially, we hoped to use our criteria for a comprehensive characterization of several heterogeneous molecular biology database systems, including GenomeTopographer, IGD, Entrez, Docking-D/RELIWE, and LinkDB. Unfortunately we were not able to find enough information for such a characterization. We hope that the criteria presented in this paper will be applied in the future to characterize and document heterogeneous molecular biology database systems, and thus facilitate comprehensive evaluations and comparisons of these systems.
Acknowledgements. The work of Victor M. Markowitz was supported by the Office of Health and Environmental Research Program of the Office of Energy Research, U.S. Department of Energy under Contract DE-AC03-76SF00098. The work of Otto Ritter was supported by the European Commission within the BIOMED 1 program under Contract GENE-CT93-0003.