Applying Semantic Web Techniques to Bioinformatics

While advances in database and parallel processing technology over the past decade have improved the celerity with which bioinformatics computations can be performed, many unnatural barriers persist between bioinformaticians and their data, which is often scattered over dozens of machines in incompatible data stores in a myriad of formats. Users of biological data should be allowed to concentrate on their research and not the job of interfacing disparate systems and data sets. In this paper we present an approach to unifying diverse bioinformatics data sources, such as GenBank, PDB, and literature databases, using Semantic Web techniques. We describe a unified naming convention called LSID, which permits uniform, Internet-based retrieval of biological data, and schemas based on RDF that enable us to integrate protein and genetic sequence data, literature references, Web Service definitions, and annotations into a common framework. Finally, we demonstrate a user interface that takes advantage of the increased uniformity and expressive power of this data model and allows users to collect pieces of data together according to purpose or research objective rather than data format.