Feature - Protein origami: function follows form |
||
|
|
||
Scientists in the Kuhlman Laboratory at the University of North Carolina at Chapel Hill are using the Open Science Grid to perform protein origami. They're running Rosetta: a powerful tool for predicting and designing the incredibly complex three-dimensional structures that proteins can adopt. Why does protein folding matter? If you change the way a protein is folded, you can change the way it functions, and can even stop it from functioning at all. Nature's ultimate nano-machines Proteins are nature's ultimate nano-machines, performing sophisticated functions on a scale much smaller than any machine we can construct. They can speed up chemical reactions, turn chemical energy into mechanical force, regulate the way cells grow and divide, and even form signaling networks for cellular communications. Brian Kuhlman, principal investigator at the Kuhlman lab, uses Rosetta on OSG to predict and design protein structures. More ambitiously, he and his colleagues are attempting to design protein "keys" that can lock up a certain protein involved in disease development. These synthetic proteins must be designed to bind specifically to their target disease protein, blocking its ability to bind to and cause disease in human cells. |
||
Designing the future Designing novel protein structures is computationally demanding. For each protein, Rosetta must perform thousands of independent simulations, sampling different combinations of amino acid sequences and conformations. "For each protein we design, we consume about 3,000 CPU hours across 10,000 jobs," says Kuhlman. "Adding in the structure and atom design process, we've consumed about 100,000 CPU hours in total so far." Why does it take so long? Even a small protein can consist of hundreds of amino acids, which can fold into thousands of structures. To decide which of these structures works best, Rosetta needs to calculate the energy of each structure. The problem grows exponentially with the size of a protein. Fortunately, these simulations do not need to be run on a large parallel machine, but can rather be farmed out to large numbers of independent processors. The distributed framework of the Open Science Grid has been ideal for this work. "What impressed me most was how quickly we were able to access the grid and start using it. We learned about it [at RENCI], and we were running jobs about two weeks later," says Kuhlman. "The successful design of a novel protein interaction still remains an unattained goal," he adds. "But we've managed to enhance protein stability and binding affinity. We're also encouraged in our efforts at designing sequences that change structure." - Anne Heavey, Open Science Grid More detail on this story appears in OSG Research Highlights. |
