|Back Forward Contents At Rensselaer|
|School of Science|
NSF grant funds research into virtual discovery of pharmaceuticals
Data team: Breneman, Embrechts, and Bennett
Curt Breneman is a chem-ist by training and temperament, but his metaphor of choice is automotive.
Suppose you resolve to design the perfect high-performance vehicle. You assemble a complete database of every quality that goes into a first-class set of wheels—gear ratios, rates of fuel consumption—but while analyzing the data, you discover an unexpected connection. Often, very fast cars are painted red. Should your dream car be red?
“Now, red doesn’t make a car go fast, but a lot of fast cars are red,” says Breneman, associate professor of chemistry. “When you’re dealing with pharmaceuticals, there’s a gazillion combinations like that, and there’s no way to know for sure which ones are important. To do this process right, what you need is a lot of source materials to mine.”
With his colleagues, Mark Embrechts, associate professor of decision sciences and engineering systems, and Kristin Bennett, associate professor of mathematics, Breneman is the recipient of a three-year, $1.2 million grant from the National Science Foundation for the virtual discovery and design of pharmaceuticals.
“Can we predict how good a combination of molecules will be for headaches, for instance, or something more important? We think we can predict the qualities of drugs that haven’t even been invented yet,” Embrechts says.
Breneman has developed a sophisticated modeling tool called Transferable Atom Equivalent to accurately determine the properties of new molecules—even very complex molecules with thousands of atoms.
What’s required is an immense amount of data, what Breneman calls “a library of atom types,” and that’s where Embrechts and Bennett come in. Their job is data mining—processing reams of information about atomic properties and interactions.
Ultimately, the researchers believe their method of pharmaceutical design will streamline the current time-consuming process of testing drugs. As a result, costs could be reduced and drugs to treat and cure such diseases as smallpox and the common cold might be approved faster by the Food and Drug Administration.