The outstanding success of recent biotechnology endeavors has largely been propelled by the possibility of large scale genomic data analysis. The possibility of querying and analyzing all or a large subset of genes from one or more genomes has become real, and has opened up new opportunities to finally address hitherto unanswered biological problems. Genome databases, in particular, and bioinformatics in general, are the critical and unavoidable pieces of technology that aid biotechnology researchers interested in harvesting genetic solutions within the wealth of information offered by genomics.

Two important issues need to be addressed immediately to advance the state of genome databases and its effectiveness. First, there are numerous databases around the world for a multitude of organisms that are publicly accessible. Some even store data for the same organism. They sometimes complement, augment and replicate each other, yet, in all likelihood, they are represented differently, or associated with different semantics. For advanced and effective research, the integration of all such databases is essential so that many interesting queries can be asked and answers found. For example, are there common biological processes among certain groups of organisms of different phylogenie for which information are stored in a number of databases? For effectively answering this query involving all accessible databases, semantic and syntactic heterogeneity must be resolved in a clean way.

Second, it is not really possible to anticipate the nature of user queries ahead and develop interfaces accordingly, which IS precisely the method employed in leading contemporary bioinformatics databases. The form based access to information is undoubtedly prohibitive and limiting for cutting edge new age research. To the extent possible, researchers should be able to ask questions as desired in an ad hoc fashion, superimpose knowledge on the database and pose hypothetical questions, validate theories, etc. in a database independent manner. Such possibilities can be suitably supported by a query language similar to SQL or Datalog. For the so called naive user, a QBE like interface, or a more traditional form based interface, may be used. For more advanced users, a language based interface seems the right choice. Visual query languages may well be an alternative. In any event, the query interfaces must be able to support queries over a set of data, and possibly across databases in a user transparent way. Above all, the database must be ``intelligent'' enough to cooperate with users in their quest for information, and passive.