In molecular biology, databases form an essential complement tothe data contained in the literature. Nowadays there exists a large number of databases of heterogeneous data. On the one hand, this diversity can be explained by the variety of biological data, going well beyond sequences. On the other hand, the databases have been designed with different objectives in mind. The major problem for the management of biological data is therefore no so much their volume as their heterogeneity (nature of the data, representational formats). Consequently, the fundamental question is to integrate the biological data in order to make them accessible and to exploit them as easily as if they were contained in the same database. The review discusses the different technical solutions that have been proposed thus far. Il underlines the necessity in every case to conceptualise and to represent formally the biological entities being concerned and their relations. A simple, but complete example illustrated this approach.
|
