DeCode to Market Clinical Genomics Software Through IBM
DeCode and IBM announced an agreement last week that will bring the Icelandic population genetics company’s Clinical Genome Miner (CGM) software platform to the commercial market.
Under the terms of the three-year agreement, DeCode will integrate CGM with IBM’s hardware and software. While CGM will run on other platforms, IBM will be the “preferred provider” for the DeCode system, said Anne-Marie Derouault, director of business alliances and distribution channel management at IBM. The companies will jointly market and sell the combined system, and will cooperatively provide consulting, implementation, and integration services.
The architecture for the joint solution includes IBM’s eServer pSeries running AIX and its xSeries running Linux, WebSphere internet infrastructure software, the DB2 database with DB2 Intelligent Miner, and DiscoveryLink data integration middleware.
In addition, IBM will now provide DeCode with an IT infrastructure for its in-house use that includes eServer systems based on the new Power4 microprocessor and the “Shark” Enterprise Storage Server disk storage system.
CGM allows users to mine their own phenotypic and genotypic datasets, along with DeCode’s annotated human and animal genome data.
Derouault noted that the agreement is IBM’s first partnership with a software provider in its newest focus area — information-based medicine. IBM and DeCode estimate that the market for tools and technology in this field could grow to as much as $100 million in three years, Derouault said.
Tomorrow’s Bioinformatics Trends Debut at PSB 2003
The Pacific Symposium on Biocomputing is known for two things: the hands-down sweetest venue in bioinformatics (Hawaii every January), and its focus on emerging bioinformatics approaches.
After wrapping up PSB 2003 on January 7th, co-chair Russ Altman mused a bit about this year’s highlights with BioInform last week. Yes, the weather was beautiful, but this year’s 365 attendees also took in some high-quality science, Altman said.
“There is clearly a movement away from sequence analysis towards understanding function,” wrote Altman in an e-mail to BioInform. Sessions in ontologies, natural language processing, gene regulation, and comparative genomics also generated some buzz, Altman said.
One trend that stood out in Altman’s view “is the expectation that bioinformatics predictions will be backed up with experimental or other biological data.” While this is not yet pervasive, “there seems to be increasing realization that experimental-informatics partnerships are critical,” said Altman.
Lion in SRS Licensing Deals with J&J, Cellzome
Lion Bioscience said last week that it had signed a two-year global licensing agreement with Johnson & Johnson Pharmaceutical Research & Development and a three-year licensing agreement with Cellzome for its SRS integration technology.
Financial terms were not disclosed for either deal.
J&J will use SRS 7.0 as well as SRS Prisma, a version of the product that automatically updates internal databases as new information becomes available. Particulars on the Cellzome license were not provided.
Incyte Evaluating SGI Altix Clusters
Incyte, which currently operates a 3,000-processer 32-bit Beowulf cluster, is “investigating uses” of SGI’s recently launched Linux-based Altix 3000 supercluster server family, SGI said last week at LinuxWorld in New York.
Incyte, which migrated from a proprietary Unix-based platform to the Linux cluster amid much publicity several years ago, is now finding that the operating costs of the commodity cluster are a bit higher than initially expected, and that the system utilization is not as efficient as hoped for, according to Dan Stevens, SGI’s marketing manager for life and chemical sciences. As part of a total-cost-of-ownership study that Incyte was performing on the system, it opted to evaluate the Altix system.
In preliminary benchmark tests running Blast that were performed on a 1-GHz Altix prototype, Incyte reported a seven-fold speedup over its 2.2 GHz Beowulf: The Beowulf processed 39 sequences per processor per hour, while the Altix processed 277 sequences per processor per hour.
Incyte has not committed to purchasing the Altix system, but according to Ken Jacobsen, Incyte’s senior vice president of information services, “We were very impressed by the fact that the SGI Altix 3000 family server showed a 7x productivity improvement against our existing Intel Pentium-based Linux cluster.”