Grid technology for collaborative ontology development

Back to index page

Previous article: Repositories and the grid come together

Next article: OGF help IT to reduce carbon emissions

The far-reaching impact of the Gene Ontology (GO) shows the important role a shared conceptual view of the biological world can have on data interpretation in the life sciences. Subjects such as anatomy, phenotype and clinical studies are active areas for ontology development where consensus has yet to be reached, or where ontologies have been developed, but their organisation needs to be kept under review. At the very least, any ontology that is in use will be curated and subject to extensions and alterations. In all cases, effective collaborative ontology development is essential to progress, and appropriate tools are needed.

While the GO is curated centrally, many biological ontologies are, in contrast, developed by loosely organised groups of scientists who remotely participate in the ontology development process. Until recent times, technological support for bio-ontology development relied on stand-alone ontology editors running on users’ desktops. These supported the creation of new ontology versions (for example, OBOedit, COBrA and Protégé) private email, email lists and perhaps Wikis for the distribution of ontology files and discussions about their contents. Clearly, much better use could be made of storage, versioning and visualisation techniques being developed by the database and e-Science communities.

The BioSphere ontology portal provides a set of tools for editing ontologies and organising the community of developers. Through the use of GridSphere, tools and services can be provided securely. This system provides each user with their own view of the ontology which is created on-the-fly. Each user’s view is derived from the group’s view – ontology development is a process where users create new versions from an initial common source, and then, periodically, divergent views are reconciled by a curator and a new consensus emerges. For each user, the system automatically manages the edits they make and supports versioning. That is, while editing their view of the ontology, the modifications are saved in the central source document as updates or deletions. Users belong to groups and have editing or curation rights, according to the views of other members of the group. The use of an XML database allows searching within one or more ontology documents, all of which are stored in the XML syntax recently developed for the Web Ontology Language (OWL) version 1.1. Viewing ontologies as XML documents allows XML methods for versioning to be applied. Additional annotations are introduced to represent timestamps for each assertion in the ontology. The eXist XML database is accessed through an OGSA-DAI API, and while currently the OWL XML API resides in the GridSphere portlet, our ultimate aim is to relocate these data-access functions to the OGSA-DAI layer so that they can be made widely available.

While developing systems in the Java Enterprise style has its challenges – the multiplicity of layers being one – we believe that the potential for code reuse within the e-Science development community is a key benefit beyond the immediate goal of providing a useful tool for a specific user group.

http://www.geneontology.org

http://www.biosphere-portal.org