Bloomberg News

IBM Gives U.S. Health Researchers Data on 2.4 Million Chemicals

December 12, 2011

Dec. 8 (Bloomberg) -- U.S. researchers gained access to a database of 2.4 million chemical formulas and diagrams that International Business Machine Corp. culled from 24 years’ worth of patent applications and medical journals.

The catalogue of compounds will be housed at the National Institutes of Health in Bethesda, Maryland. Scientists can use the data to identify new candidates for drug development or new uses for existing drugs.

“The applications are very wide reaching, from life sciences to chemicals, petroleum, to food, to health,” said Chris Moore, a partner and vice president in Armonk, New York- based IBM’s global life sciences division, in a telephone interview.

The NIH has made development of drugs for rare diseases such as sickle cell anemia a priority since Francis Collins became director of the $31 billion research agency in 2009. Collins has asked Congress to create a new institute for drug development and pressed manufacturers to open catalogs of abandoned compounds.

IBM, the world’s biggest computer-services provider, created the database by “Watson-type technology,” Moore said, referring to the supercomputer that beat human competitors on the game show ‘Jeopardy.’ The data was extracted from about 4.7 million U.S., European and United Nations patents and 11 million biomedical journal abstracts from 1976 to 2000.

Researchers can search the database for free. IBM plans to sell a product that will produce more sophisticated analysis of the database that can tell customers who is conducting research on specific therapies, Moore said. He wouldn’t disclose the price.

--Editors: Adriel Bettelheim, Andrew Pollack

To contact the reporter on this story: Alex Wayne in Washington at

To contact the editor responsible for this story: Adriel Bettelheim at

The Good Business Issue
blog comments powered by Disqus