June 14, 2006 | As an embryonic capability in drug development, most people outside of informatics and IT have barely even heard of the Semantic Web, much less know what to do with it. At its best, the Semantic Web aspires to enable enhanced intelligent connectivity between all forms of data and information available in a user’s work environment (see “Masters of the Semantic Web,” October 2005 Bio-IT World, page 28). Its vision of simple and direct access to knowledge, both within and outside the organization, has the potential to accelerate and add knowledge to decision making with little or no need to know or understand where data or information is physically located and what form it is in.
Isolated pockets of data and information are pervasive across the current drug development environment. Whether it is the division between regulated clinical data and discovery research data, internal research data and external collaborator data, or different research data storage conventions among different company divisions, the inability to easily access relevant information, wherever it resides, limits an organization’s ability to make properly informed decisions. As a result, time and effort are often wasted, and precious resources are spent on repetitive or fruitless pursuits. The promise of the Semantic Web is not only to provide access to diverse data resources but also to enable more complex, intelligent questions to be asked, with the potential for roadblocks and other problems to be solved more effectively and quickly.
The Path to Reality
Advances in Semantic Web technology in the life sciences are occurring in three areas, each of which will be critical to broad adoption of the approach:
- Establishment of uniform standards and application best practices
- Development of Semantic Web infrastructure to ease adoption of the approach and enable access to current data repositories
- Validation of key Semantic Web-based applications by pharmaceutical industry early adopters
With the World Wide Web Consortium (W3C) taking the lead, the establishment of standards and uniform best practices for the Semantic Web are well under way. With the establishment of uniform standards from the beginning, it should be possible to design and implement Semantic Web solutions that are consistent and compatible across the life science landscape, ranging from GenBank and the scientific literature to pharmaceutical industry discovery databases and biospecimen repositories. The W3C Semantic Web for life sciences group is actively developing the publicly accessible BioDash demonstration platform (www.w3.org/2005/04/swls/BioDash/Demo/).
Life science and IT vendors are key to enabling broad adoption of the Semantic Web commercially. While very early in the adoption process, advances, including development of RDF-compatible middleware at Oracle to enable Semantic Web access to current Oracle 10g database and Teranode’s transformation of its product to an RDF-compatible platform, will help accelerate broader adoption of the approach by the drug development industry.
Even with successful development of standards and access to the Semantic Web infrastructure, Semantic Web technology overall will not be successful without broad acceptance and adoption by the drug development industry. With ongoing pressure to streamline the drug development process, there have been significant efforts toward increasing knowledge-based efforts, with a focus on both increased efficacy and improved drug safety. For the Semantic Web to be successful, it must rapidly demonstrate value and utility, with the potential for long-term benefits of improved efficiencies and cost savings. Initial pilot studies are ongoing and include using the Semantic Web to link separate internal databases and enable new knowledge capture and evaluating the Semantic Web as a pathway modeling and analysis tool.
The results of these and other pilot studies will be critical to timely adoption of Semantic Web technologies. If successful, broad adoption of Semantic Web technology could occur in as little as five years.
Alan S. Louie is a research director for Health Industry Insights. E-mail: firstname.lastname@example.org.