Sun Microsystems forged ahead with a number of new academic alliances last week. The company donated several servers to the BioPathways Consortium and SRI International’s BioCyc project under its academic equipment grant program, and also sealed a $40 million hardware deal with the UK’s Cambridge-Cranfield High Performance Computing Facility.
News of the BioPathways Consortium equipment grant first broke in January [BioInform 01-14-02], but the specifics of the group’s plans for the donated system were not disclosed until last week. The servers, which will be installed at Columbia University over the next few weeks, will run the GeneWays text-mining algorithm developed by Columbia’s Andrey Rzhetsky. GeneWays uses natural language processing to automatically extract molecular pathway interactions from online scientific journals.
Rzhetsky said the software has so far analyzed approximately 100,000 articles and extracted 1.5 million interactions. The software is currently available for download for developers, but the new equipment will allow anyone to access the tools and pathway data over the web once it is put in place.
The BioPathways Consortium plans to enable other researchers in the field of automated biological text-mining to offer their tools via the Columbia server as well, and is considering creating a common public repository for all mined pathways and mechanisms.
“The dream is to have a ‘review article’ that is always up to date and can summarize everything that’s known currently,” Rzhetsky explained.
Users of the recently re-named BioCyc Knowledge Library (www.biocyc.org) will also benefit from the Sun donation. BioCyc encompasses the EcoCyc and MetaCyc curated pathway/ genome databases as well as databases for 12 other organisms developed using SRI’s Pathway Tools software.
Peter Karp, director of the bioinformatics research group at SRI International and principal investigator for the BioCyc project, said the new Sun workstation has improved the performance of the project’s website and has enabled a new Blast search capability for each genome in the BioCyc collection.
At the Cambridge-Cranfield HPCF, 10 Sun Fire 15K servers totaling two teraflops will be added to a 160-processor 240-gigaflop IBM SP system and a 64-CPU SGI Origin 2000 already in place. Researchers at Cambridge and Cranfield Universities will use the system for bioinformatics among other computationally intensive tasks, according to Sun, but a spokesperson from the facility was unavailable to provide further details before press time.
The Sun Fire 15K servers will be installed this fall and available to researchers throughout the UK as part of the country’s eScience grid computing initiative.