Biology research goes in silico

Long before a new jumbo jet takes off the runway, all of its systems and subsystems down to the tiniest of individual parts have been virtually designed, built and tested through the use of computer models and simulations. Similarly, Pacific Northwest National Laboratory is developing virtual cells that can be explored, tested and manipulated within the world of computers to make important discoveries in systems biology.

"Computational biology involves creating sophisticated mathematical and equation-based computer models and simulations to understand and predict how cells behave, interact and respond to their environments," said Dave Dixon, associate director of theory, modeling and simulation at the U.S. Department of Energy's William R. Wiley Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility operated by PNNL.

Dixon leads the Virtual Cell project, which is focused on integrating and analyzing biological data and developing and improving advanced computational models from the data. With the help of conceptual models, new drugs could be designed, for example, and capabilities of microorganisms could be harnessed to convert contaminants into less harmful substances or to efficiently generate energy by creating hydrogen.

The broad scope of the research includes atomic-level simulations of biomolecular complexes, network analysis and kinetic modeling of cell signaling and metabolic pathways, and development of bioinformatics tools for data mining and analysis with a focus on data from high throughput experiments.

In one study, researchers are using computational methods to simulate cellular signaling pathways and compare reactions that originate at the cell surface with those that begin within the cell. Because a mutant form of a specific protein complex called Ras is found in 30 percent of all human tumors, researchers are investigating how the Ras protein serves as a switch to control when cells develop and grow and how they differentiate from one another.

Another group of researchers is using computational data analysis tools to statistically analyze biochemical data from some of the world's most powerful mass spectrometers at EMSL. "We're analyzing how protein networks within cells communicate with each other, using the same statistical approach that is used to analyze communication networks," said senior research scientist Bill Cannon.

"The Internet is a system of computers talking to one another in different ways," he said. "Similarly, we're trying to understand how cells communicate as part of a complete system." Cannon and his colleagues are using statistical tools related to network communications and artificial intelligence to study Shewanella oneidensis cells, which potentially could be used in bioremediation.

In yet another study, researchers use slices of images captured by a microscope to explore water density in cells and simulate radiation damage. In addition to understanding how radiation affects exposed cells, researchers are investigating how effects spread to adjacent cells, which is known as the bystander effect.

"There are a lot of opportunities for computational science, mathematics and statistics to bring together data sets and extract valuable information from the mountains of data collected," Dixon said. "The models we build help bring the pieces together and drive where the research is headed to get a description of a biological system."

For more on computational biology, see www.biomolecular.org/research/virtualcell/index.html. For more about the virtual lung, see www.pnl.gov/news/2001/01-33.htm