Baltimore, Md.--In a move that surprised and disappointed many genetics researchers, the US Department of Energy (DOE) has terminated funding for the human genome database housed at the Johns Hopkins School of Medicine here. The eight-year-old database, which is the official repository for genomic mapping data produced by the Human Genome Project and one of the nation's largest bioinformatics projects, will cease operation by July 31, officials announced on January 20.
"The database played an important role in advancing genetic research, but it has been overtaken by newer technologies and events," explained Dan Drell of DOE's Office of Energy Research, which provided most of the database's $5 million annual budget. In particular, Drell and other sources told BioInform that DOE's growing focus on large-scale gene sequencing--as opposed to mapping--led to the decision to redirect the database funding to other, as yet unidentified, bioinformatics projects. "As the heyday of traditional mapping has faded, so has the perceived need for a large community database project focused on maps," explained Stanley Letovsky, the database project's director, in a notice posted on the project's web site, http://www.gdb.org.
"This decision does not mean DOE is reducing its commitment to informatics," Drell emphasized, adding that the agency is not "pulling the plug on the human genome database and throwing everything in the trash can." Letovsky said a mothballed version of the database will remain accessible to research ers through Ed Uberbacher's Computational Bioscience Section of the Oak Ridge National Labora tory, at http://compbio.ornl.gov. In addition, the Human Gene No men clature section of the database project will continue to be curated by Sue Povey of the Uni versity College of London (http://www.gene.ucl.ac.uk/nomenclature/).
The termination decision, which abruptly followed a year of uncertainty following DOE's 1996 decision to renew project funding for only one year, shocked many researchers. "I was horrified to discover the announcement," said Janice Nicklas, a molecular geneticist at the University of Vermont, Burlington, who uses the database several times a month. She calls it "a remarkable resource. Without it, you'd have to spend weeks looking for literature. I think they are premature in saying people don't need this any more." Nicklas has protested the termination to federal officials and, as BioInform went to press last week, at least 60 other researchers had voiced concerns through a special comment form on the database web site.
Drell conceded the termination "will cause real pain for some researchers," but said the protests are unlikely to reverse the decision. "It's a done deal," he insisted, noting that officials at the National Institutes of Health (NIH), the other major funder of the database project, supported the decision.
Among those hardest hit by the termination, observers told Bio Inform, will be medical researchers hunting for disease-causing genetic mutations and scientists in smaller labs that lack their own informatics resources. Overall, database officials estimated that up to 10,000 researchers used the database in 1997. The database's web site received an average of 75,000 hits per week last year, ranging from a low of about 40,000 to a high of 145,000.
At least one informatics specialist, however, said the termination decision is a good one. "It should have been done a while ago," claimed informatics researcher Nathan Goodman of the Jackson Laboratory in Bar Harbor, Maine. He said the growth of genetics information on the internet and the advent of high-throughput labs have made such multipurpose databases--originally designed to capture and organize scattered information from scores of small labs--"obsolete. When they were invented, the human genome database and similar databases were critical in helping genetics move into a prominent position, but there is no reason to assume that approaches that were useful in the past will remain useful. We need to revisit the scientific rationale for these databases and ask whether they are providing a value commensurate with their cost." Projects that are similar to the human genome database but much smaller include databases on mouse, fruit fly, and yeast genomes.
The Howard Hughes Medical Institute originally created the human genome database in 1989 to capture data from Yale University's Human Gene Mapping Library project. In 1991, DOE, NIH, and the Japan Science and Technology Agency began funding the database. Under the stewardship of a worldwide group of scientific volunteers, the project electronically organized much of the available information on human genetics and gene mapping, helping jumpstart the Human Genome Project. It was an early pioneer in the use of the web to share bioinformatics data and also one of the first to deploy a Java client application, Mapview, to graphically display database query results. In recent years, database staff developed novel algorithms to integrate multiple maps into a single comprehensive chromosome model that can be searched and displayed.
Observers told BioInform the termination may have an unexpected silver lining for some of the two-dozen-plus database staffers who will lose their jobs. "I'm sure the headhunters are drooling over this pool of trained folks," Goodman said. Another informatics watcher quipped, "It may get ugly. If they jump to the private sector, some of these people will have to choose between doubling their salary or simply increasing it by half."
But Letovsky said the transition won't be easy for a number of his staff, who have strong ties to the Baltimore area or lack the informatics skills that would make them attractive recruitment targets. In addition, he continued, researchers who want to stay in the public sector "will find it very difficult to replicate the kind of environment the human genome database provided."