Web Exclusives: Structural Biology
Advancing Protein Studies with Worldwide Software Experiment
Structural biologists and computer programmers are teaming up in a worldwide effort to advance a field that could help us understand what certain molecules do inside our bodies. The project will test how well software programs can turn data into an accurate picture of a protein structure.
An increasingly popular way to figure out the shapes of proteins is called nuclear magnetic resonance (NMR) spectroscopy. Unlike the more established approach of X-ray crystallography, NMR places proteins in solution rather than crystallizing them in a solid form. This lets researchers see the proteins in a more natural environment and capture their interactions with other molecules—details that can help explain the proteins' functions in biology and health.
But NMR studies are very difficult, expensive and slow, especially when it comes to analyzing data. After spending weeks carrying out experiments on a single protein, researchers can spend several more weeks just making sense of the resulting information. Much of this analysis is done manually, which can introduce errors or bias.
The researchers behind the new software project, formally called the "Critical Assessment of Automated Structure Determination by NMR" (or CASD-NMR), hope to change all this.
During CASD-NMR, researchers will share their NMR data—but not the protein structures it generated—with computer scientists and others, who will then use different software programs to analyze the data automatically and produce the best rendering of the related protein. CASD-NMR organizers will compare those renderings with the structure originally determined manually by the scientists.
The results will help the community identify the most reliable software programs. They also will lead to improvements that make data analysis faster and more accurate.
Gaetano Montelione , a CASD-NMR organizer and a molecular biology professor at Rutgers University, says the effort ultimately could completely automate the analysis process, freeing up time for new experiments and making high quality NMR data analysis easier and more accessible to a larger number of scientists.
"This effort has been talked about for years, but it's only now possible," says Montelione, because of the recent capabilities of large-scale structural biology centers to provide the data needed for CASD-NMR.
Montelione directs one of these centers—the Northeast Structural Genomics Consortium , which is part of the U.S. Protein Structure Initiative. The center has used NMR to solve more than 300 protein structures, and it plans to contribute at least one new NMR data set each month to the software experiment.