IBM taps 'Watson-type' tech for drug research tool

By now, you've probably heard (at least once) about how IBM's ($IBM) Watson supercomputer bested some of the brightest human contestants on the game show Jeopardy! Now the tech giant is using what one company official called "Watson-type" technology for a system that can rapidly track data on chemical compounds and patents to aid drug researchers and others in speeding the development of products--including cancer drugs.

IBM is also giving away a huge chemical database to the NIH that was amassed with the company's business analytics and its strategic IP insight platform (SIIP). The technology behind the database--which holds information on 2.4 million compounds collected from patents and medical journals from 1976 to 2000--was developed through IBM's work with AstraZeneca ($AZN), Bristol-Myers Squibb ($BMY), DuPont and Pfizer ($PFE), according to IBM's release.

The system bundles data-mining, natural-language processing and analytics to scour patent databases and medical journals for chemical data, The New York Times reported. While researchers will be able to tap the donated database for free, IBM has a paid version of the system that provides analytics for specific chemicals, Bloomberg reported. So Big Blue plans to make a bundle on the system while providing some of the earlier output from it for zilch.

"It provides a landscape that shows who is working with what chemicals and drugs," Chris Moore, head of business analytics and optimization in IBM's global services unit, told the Times, adding that the software can be applied for a range of areas, including product strategy, recruitment and patent research. More called the tech "Watson-like" in an interview with Bloomberg.

Automated tools like IBM's insight platform can save life sciences companies hours that would be spent on manual compound and patent searches. Some other players with tech relevant to this area include the startup Relay Technology Management, which has software that aggregates scientific data for biopharma companies, and Cambridge Semantics, which uses semantic web tech to give pharma customers easy visibility of life sciences data from disparate sources.

IBM's chemical database will be contained in the National Center for Biotechnology Information's PubChem resource, which can be accessed here.

- here's IBM's release
- get more from the NYT's blog
- and the Bloomberg article