Seeking Byte-Savvy Biologists

The U.S. government should fund a new network of research centers devoted to churning out biologists skilled in the arts of computer science, a National Institutes of Health (NIH) advisory panel concluded today. The recommendation signals that NIH may be ready to throw its substantial funding weight behind biocomputing, an emerging interdisciplinary field that the panel says has been neglected by universities.

The use of computers in biomedical science has grown explosively over the last decade, with researchers relying on the machines to do everything from browse the technical literature to model the complex folding of proteins. But few biologists have the expertise necessary to tap the flood of data generated by gene studies and clinical trials. A common genetics lab these days, for instance, can produce 100 terabytes of information a year--equivalent to 1 million encyclopedias. At the same time, few computer scientists know enough biology to figure out the best ways of mining important nuggets from the data deluge. "You can count on the fingers of one hand" the number of researchers with topflight training in both fields, says geneticist David Botstein of Stanford University in California, a co-chair of the advisory panel.

To bring biologists up to speed in the brave new computing world, the 16-member panel appointed by NIH director Harold Varmus recommended that the agency create up to 20 biocomputing centers at universities and research institutes, with each getting up to $8 million a year to focus on research and training rather than hardware. Such "watering holes" could bring together biologists and computer scientists and break down the disciplinary walls that have hobbled cooperation within universities, says panel co-chair Larry Smarr, director of the National Center for Supercomputing Applications at the University of Illinois, Champaign. The panel also recommended that NIH establish a new program aimed at organizing and curating massive biological databases, such as one on the mouse genome; set aside funds for researchers to hire biocomputing help; and support scientists assembling midsize computer networks that are more powerful than desktop machines but easier to manage than supercomputers. By taking such steps, "NIH will send a powerful message" about biocomputing's importance, the panel concluded.

The recommendations are "good news" and "should cause universities to pay attention" to biocomputing, says computational biologist Larry Hunter, a researcher at the National Cancer Institute in Bethesda, Maryland, and president of the International Association for Computational Biologists. It's unclear how much cash NIH might plow into such an initiative. But NIH officials say the subject is sure to come up later this month, when Varmus and his top aides meet to begin mapping out their 2001 budget request.

Follow News from Science