To narrow down the list of chemicals, among thousands in the environment, that pose risks to people or wildlife, the US Environmental Protection Agency (EPA) is turning to computational methods. The agency has $2.4 million in anticipated funding for innovative approaches to virtually mimicking the effects of compounds on living organisms. The specific problem the EPA is concerned with at the program's outset is endocrine disruption, specifically of the brain–pituitary–gonadal and the brain–pituitary–thyroid systems.

"There are 87,000 chemicals in the environment. There is a need for improved approaches for prioritizing which chemicals should be screened and tested," explained Elaine Francis, director of EPA's endocrine disruptors research program. "We would not have to test every chemical for effects, and make better informed decisions as a regulatory agency."

Computational toxicology also could help cut animal testing, the original reason EPA began pursuing the approach 15 months ago, Francis told The Scientist.

"The time was right to begin using emerging technologies to improve hazard identification and risk assessment," she said. "It's making better use of molecular tools that did not exist 5 years ago—high-throughput approaches, genomics, proteomics, and metabolomics—coupling those with computerized approaches to help us model that information."

"I'm very happy to see the EPA coming up with funding to stimulate research in this area. I think we'll be seeing a lot more of these initiatives, for instance, from in NIH," said Thomas Knudson of Thomas Jefferson University, editor of the journal Reproductive Toxicology.

While endocrine toxicologist Peter Thomas of the University of Texas in Austin feels this program is valuable to start now, he is uncertain about how much it can accomplish. "It will be hard to model some of these responses. The reproductive endocrine system is very complicated. It's not like modeling a simpler pathway like the insulin–blood sugar system, where one hormone does one thing," he said.

According to EPA mandate, only institutions of higher education and not-for-profit institutions are eligible to apply. The agency anticipates making three to five awards of potentially $150,000 to $250,000 per year for up to 3 years.

So far, Francis said, she has received about a half-dozen inquiries. The deadline for submitting proposals for these research grants is January 21, 2004. "We purposefully left a large window of time, because of the need to develop partnerships across disciplines for this solicitation," Francis said.

The new grant program is part of EPA's larger Computational Toxicology Framework, which will be the topic of a workshop in Research Triangle Park, NC, September 29-30.