Database strategies for genetic information and biological data

Summary form only given. Two key characteristics of modern biological research are: (a) the experimental data collections are very large and difficult to interpret; and (b) data interpretation generally requires information from genetic and other databases that are distinct from the ones in which the experimental data are stored. I discuss current strategies for constructing a database federation that can support composite queries that require data from different databases. I discuss current efforts to construct reusable software templates for biological data that makes querying and extracting data from different experimental databases much easier. The role of XML and various transport and interface mechanisms (TCP/IP, CORBA, Java) are also covered.