IBM has taken an undisclosed minority equity stake in protein database provider Structural Bioinformatics of San Diego, marking Big Blue’s first investment in a life sciences company.
As part of the deal, IBM has also become the strategic information technology partner of Structural Bioinformatics.
The two companies will collaborate to make the content of Structural Bioinformatics’ databases more accessible to researchers worldwide through the Internet on a subscription basis. Joint marketing initiatives are being planned as well.
Anne-Marie Derouault, director of business development for IBM Life Sciences, told BioInform that the deal with Structural Bioinformatics is a good example of IBM’s strategy to increase its prominence in life sciences.
“It [the agreement] involves the equity investment and it also includes a strategic partnership on helping them standardize on the DB2 platform and the Websphere platform to build their Web infrastructure and data management infrastructure,” said Derouault.
The relationship with Structural Bioinformatics is the latest initiative by IBM’s Life Sciences business unit, which was started in August with a $100 million allotment to form partnerships and develop IT solutions for genomics, biotech, pharma, and other life sciences companies.
One of the first partnerships in that effort has been with Incyte Genomics, which is using IBM’s DiscoveryLink middleware in its forthcoming Genomics Knowledge Platform data integration product. Both Incyte and Structural Bioinformatics are relying on IBM for data storage, but they are using different products to do so, said Derouault.
Derouault noted that the investment in Structural Bioinformatics is not part of Big Blue’s $100 million fund but rather an additional amount.
Structural Bioinformatics will use IBM’s DB2 Universal Database as its development platform and IBM Websphere as its Internet software infrastructure for providing protein structures over the Web.
Managing proteomic data is a challenging endeavor largely because of the amount and complexity of the data involved. Derouault said that of the estimated two million proteins only a few thousand have been identified so far.
And even when it is known which genes code for which proteins, scientists still have to learn how proteins interact with each other in order to prevent disease and develop drugs.
With about 100,000 genes and two million proteins in human biology, the amount of combinations for those will require a powerful data storage system that also scales well, said Derouault. DB2 will fill this need for Structural Bioinformatics by managing the volumes of data and also providing the middleware to integrate various protein datasets, she added.
Besides providing software, the computer giant will supply hardware including a high-performance cluster of eServer xSeries servers running Linux, which will replace an IBM SP2 supercomputer.
One use for the 128-processor Linux system will be to help in high-resolution protein modeling and dynamics calculations that track the changing shapes of protein molecules.
“We can get almost a 300-fold improvement in our computational power for dynamics calculations,” said Edward Maggio, chairman, president, and CEO of Structural Bioinformatics.
In addition, SBI will begin marketing its Variome structural variant database modules to pharmaceutical companies in early 2001 with DB2 as the preferred database. The company will still assist customers that want to use Oracle, Maggio added.
Aside from helping Structural Bioinformatics manage and process its data, IBM will also provide advice on how to send data in a secure fashion to pharma customers, said Maggio.
The company is also interested in using the technologies for processing and annotating data that IBM’s IT research group in Almaden, Calif., is developing.