Eric Neumann is VP of bioinformatics at Beyond Genomics. He chaired OMG’s Life Science Research Task Force and the Bio-Ontologies Consortium. He left I3C after one year when the organization refused to have open elections for its board of directors. He is co-founder and steering member of the Biopathways Consortium.
There has been much ado in the bioinformatics community about the Interoperable Informatics Infrastructure Consortium, but after nearly two years in existence, the organization has offered no solutions and raised many questions. Is it a data standards initiative? Is it a forum for vendors? Or is it a collective of folks who hope to help the life sciences industry out of a quandary by developing an informatics architecture that will solve the integration problems that have long haunted it? And how can I3C succeed where other such initiatives have yet to deliver, anyway?
I3C’s founders, who conceived it during a Sun Microsystems meeting in February 2001, envisioned an open consortium that would develop informatics solutions and promote working prototypes based on the XML web-services model, avoiding the slow specification process that has hampered other standards groups such as LSR-OMG. I3C was to provide practical bridges between standards emerging for the XML web-services model from groups such as W3C, UDDI, and OASIS. I3C’s participants would propose and construct “interoperable spanners,” and would do it quickly.
I was not the only one who believed that I3C would find the quickest path for offering models and specs to help life sciences companies use their existing applications to integrate and manage all the new data types and protocols they face. And I3C would do this in a unique way, using ontologies to unify diverse, even ad hoc, data types. Rather than oppose the proliferation of XML models by administering standards, different XML formats would refer to a common set of ontologies, and thereby become interconnected without having to adhere to an agreed-upon set of data-type standards.
The early I3C days were exciting. But by the summer of 2001, heavyweight participants — IBM, Sun, Millennium, and BIO — became embroiled for several months over issues of code-related IP and reasonable and non-discriminatory royalties. I3C chose to go royalty-free, but the struggle resulted in the appointment, rather than the planned open election, of an I3C board composed of those who had been part of the legal debate.
What an unmitigated disaster! While the original drivers of the organization have either had to take a back seat or, unsure of the agenda, have stopped participating, the I3C, which still is seeking a director, is on a campaign to recruit as many new members as possible — a sure path to chaos. The necessary roadmap for I3C has never been completed (Millennium took over the process to fit its own needs), and so no clear next steps have been agreed upon. Now it’s becoming a “me too” party for individual interests, but the infrastructure is nowhere to be found.
If you’re interested in the I3C interoperability story, keep an eye on how things progress over the next six months. If I3C can become a fair, responsive, open organization with open elections, and if it can develop a common, flexible, XML infrastructure model for all of the life sciences, then it will be an important and useful organization for the life sciences IT community. On a positive note, I believe a lot of useful interoperability work can be accomplished without large numbers of discrete standards, as Beyond Genomics and other companies are beginning to see in our own efforts. If I3C recognizes such results, it could achieve its goals efficiently.
But if I3C remains a rallying point for big vendors unsure themselves of how to proceed, then it will have no legitimacy. There are pressing informatics needs for biotechs and pharmas, and little time to debate whose standard is better.
Opposite Strand is a forum for readers to express opinions and ideas. Submissions may be e-mailed to [email protected]