The science of protein design depends on software that can do two things: predict the three-dimensional structure of proteins from their amino acid sequence and experiment with (or design) new shapes. In other words, says David Baker, Ph.D., director of the new Institute for Protein Design (IPD), “protein design starts off as a computer science problem.”
The Rosetta software program developed by Baker and his team is the world standard for predicting protein structures and designing new proteins. Rosetta’s ability to sort through large data sets speeds the process of designing proteins with therapeutic functions.
Enlisting the public helps, too. Rosetta@home — software based on BOINC, an open-source program that can be installed on private computers — can run on many computers at once, a resource-sharing method called distributed computing. This method gives Rosetta the computing power to process data and produce protein structures. Baker’s group also collaborates with the public through “Foldit,” a protein-folding computer game that has resulted in the mapping of proteins involved in HIV infection, among other discoveries.
Baker’s trainees, many of whom have gone on to form their own labs around the world, return to the UW every year for a Rosetta conference at which they continue to advance the software’s capabilities. “It’s a highly collaborative environment,” says Michelle Scalley-Kim, Ph.D., director of research and strategy for the IPD, “which has been critical to Rosetta’s success.”