Experiences applying meta-data to bioinformatics

Bioinformatics is facing the daunting challenge of providing geneticists an d biologists effective, efficient access to data currently distributed amon g dynamic, heterogeneous data sources. Complicating the problem is the spee d at which the underlying science and technology evolve, leaving the termin ology, databases and interfaces to catch up. As the genomics community move s from sequences to functional genomics, the pressure to find a solution is increasing. Realistically addressing this problem, whether through a data warehouse, multi-database, federated database, or other approach, requires development of an scalable, flexible infrastructure that can quickly adapt to meet user needs in this extremely dynamic environment. This is best acco mplished by extensively using meta-data to reduce the application's mainten ance costs. Using the DataFoundry project as an example, this paper discuss es the first steps and practical problems of developing a meta-data-based i nfrastructure capable of meeting the demands of an active scientific commun ity. It also demonstrates how much bioinformatics must still progress befor e it can truly satisfy its users. Published by Elsevier Science Inc.