When Sun Microsystems held the first meeting of its Informatics Advisory Council late last year, the roster read like a bioinformatics who’s who directory.
A squad of public- and private-sector bioinformatics leaders who attended the two-day meeting in Palo Alto included John Barton of Rosetta; David Benton of Glaxo-SmithKline; Chris Dagdigian of Blackstone Technology; Ed Kiruluta of DoubleTwist; Jim Lindelien of TimeLogic; Maciek Sasinowski and Brian Kamery of Incogen; Christoph Sensen of Canada’s National Research Council; Mark Showers, Monsanto’s director of IT; and Shankar Subramaniam of UC San Diego.
Sun’s discovery informatics market segment manager Loralyn Mears gathered the group to chat about how Sun’s life sciences team and its partner vendors could “facilitate industry collaboration.”
One idea tossed around at the meeting was Ernie Retzel’s two-year-old plan for a Global Genomics Grid. Retzel, another Sun advisory council member, directs the University of Minnesota’s Center for Computational Genomics and Bioinformatics, which is developing public-domain Java visualization tools for modeling the effects of perturbed genes in entire organisms. Retzel’s even grander ambitions is to establish a distributed computing and database network among public plant genomics laboratories.
He envisions a computing “collaboratory” for plant genome research such as is used in astrophysics and astronomy research. “If you have enough computer resources you can assemble a team to solve a problem and disassemble it without anyone moving anywhere,” he says. “I’m looking to tie together the developers and folks still in the academic and public sector to begin to leverage their computer and intellectual resources so that, frankly, we can compete with the private sector.”
He aims to harness enough CPUs across more than a dozen laboratories to enable high-throughput genome sequencing throughout the public sector. University of Wisconsin computer science professor Miron Livny has already pledged a pool of 720 Condor CPUs that will form the foundation of the grid. With his own 90 CPUs and another 48 from USDA’s Ithaca, NY site, Retzel says the grid is on its way: “100 processors here, 100 there, and pretty soon you have a real computer.”
Retzel says Purdue statistician Rebecca Doerge has been a strong supporter of the effort. And he is in discussions with computer scientists and biologists at campuses including Clemson, Cornell, University of Georgia, University of Illinois, Iowa State, University of Iowa, University of Minnesota-Duluth, North Carolina State, University of North Carolina-Chapel Hill, UC Davis, UC San Diego, and Cold Spring Harbor Laboratory.
Although Sun is not funding the project outright, two of the computer maker’s Discovery Informatics partners — Time Logic and Incogen — are collaborating with Retzel. And Retzel says Sun has contributed resources that are enabling him to pick up his pace. “If I can make this work, it will probably be the most important thing I do in my career,” he adds.
And even if that’s all that comes of Sun’s new push into life sciences, it will go a long way toward fulfilling the company’s promise to revolutionize the field.