Stanford University's National Center for Physics-Based Simulation of Biological Structures, or Simbios, said this week that it plans to use $10.6 million in funding from the National Institutes of Health's National Centers for Biomedical Computing program to advance simulation technology to better understand drug activity and to investigate how brain activity controls movement.
The NCBC grant — awarded to Russ Altman, professor and chair of bioengineering; Scott Delp, professor of bioengineering; and Vijay Pande, associate professor of chemistry — will fund Simbios for the next four years.
Created in 2004, Simbios is one of six NIH-funded National Centers for Biomedical Computing. Existing centers at Brigham and Women's Hospital and Vanderbilt University were also awarded grants under the latest NCBC funding round, along with a new center, called Integrating Data for Analysis, Anonymization, and Sharing, or iDASH, at the University of California, San Diego (BI 10/29/2010).
Altman said in a statement that Simbios has developed core software that it will now use "to investigate problems we couldn’t previously study."
In particular, Simbios will focus on using physical simulations to understand how drugs work and to study how neural activity ultimately generates movement.
While neuroscience is beginning to understand the brain-movement relationship, "what’s completely missing is the biomechanical model — how is the neural activity driving the muscles, which drive the bones that generate the movement," said Krishna Shenoy, associate professor of electrical engineering at Stanford and co-leader of the neuroprosthetics effort with Delp. "If you can successfully map neural activities to a realistic muscle model, then you’ve effectively also solved the neuroprosthetics problem."
Brian Shoichet, professor of pharmaceutical chemistry at UC San Francisco and a developer of drug-docking software, will lead Simbios’ drug activity initiative, which will investigate how simulations can improve predictions of drug interactions with the target protein B2AR.
In addition, Simbios is partnering with Stanford’s Pervasive Parallelism Lab to define domain-specific languages that programmers can use to rapidly write simulation code that will run efficiently on many different hardware platforms.
Domain-specific languages are expected to "hide the complexity" of parallel computing, "while still allowing people access to the power of the underlying hardware," said Pat Hanrahan, a professor of computer science at Stanford and a member of the PPL.
So far, the center has developed the simulation toolkit SimTK for dynamic modeling of biological structures; the OpenSim application for simulating movement; and the OpenMM software library for accelerating the Amber and Gromacs molecular dynamics packages on graphics processing units.