Like pieces in an elaborate jigsaw puzzle, molecules interact based on the way they fit together. A tiny crevice or bump might cause a reaction to proceed, or keep it from happening at all. Working at the crossroads of mathematics, chemistry, and biology, one UW-Madison scientist uses computers to model molecular shapes and identify features that may influence interaction.

“Changing the shape of a molecule could disrupt an interaction or enhance the molecule’s ability to bind,” explains Julie Mitchell, who has a joint appointment with biochemistry in the College of Agricultural and Life Sciences and mathematics in the College of Letters and Science. “We know that binding between two molecules likely occurs at certain areas with complementary shapes.”

A mathematician and computer scientist by training, Mitchell learned about structure modeling when she sought out cross-disciplinary applications for her expertise. Her research interests led her to a postdoctoral position at the University of California”s San Diego Supercomputer Center, and ultimately to the UW-Madison.

While at San Diego, Mitchell wrote computer programs that analyze the known crystal structure of molecules and produce three-dimensional images based on the density of atoms within the molecule. In a matter of a few seconds, her programs can perform an analysis that otherwise would have to be done by trial and error over a much longer time. Researchers can then study and compare different structures to predict which features might be most important in enhancing or inhibiting binding. Mitchell makes her software freely available to nonprofit users around the world; one program, FADE, has been downloaded at least 500 times.

“My goal is to create general tools,” explains Mitchell. “Other researchers bring specific problems to the table.” One UW biochemist, Ron Raines, says that collaborating with Mitchell is helping his lab make significant advances toward a possible treatment for lung cancer. He says that Mitchell’s analysis highlighted complimentary structural areas on ribonuclease, an enzyme that destroys RNA and is toxic to cancer cells, and ribonuclease inhibitor. By altering certain key structural areas, Raines can create an enzyme that can evade the inhibitor and kills cancer cells more effectively.

Mitchell”s recent appointments came as part of the university”s cluster hire initiative, which brings new faculty members to work on cross-disciplinary issues. Many scientists, including Raines, say that collaborative work is essential to advancing their research. “It”s hard to imagine a breakthrough like the one in my lab happening without the bridges built by the cluster hire program,” says Raines.