posted October 22, 2013

NSF Funds Integrated Metabolic and Regulatory Modeling Grant

A major, multi-year grant from the National Science Foundation (NSF) will support a significant new step in an on-going research project at Hope College that is developing and refining computer-based models of metabolic activity.

If the team’s previous work provided a road map, researchers Dr. Aaron Best and Dr. Matthew DeJongh say, the new effort should yield the equivalent of an online navigation system.

Best, who is the Harrison C. and Mary L. Visscher Associate Professor of Genetics at Hope, and DeJongh, who is an associate professor of computer science, have received the new three-year, $400,000 grant as part of an NSF collaborative award totaling $650,000 with Dr. Nathan Tintle of the mathematics faculty at Dordt College.  Best, DeJongh and Tintle have worked together on the research since Tintle’s days as a Hope colleague from 2005 to 2011.

The researchers have collaborated for nearly a decade, with support from a series of NSF grants, to build and improve a computer program that creates genome-scale metabolic models, condensing into a few hours a process that required about a year when done manually.  The research at Hope has been conducted in collaboration with Argonne National Laboratory as part of an open-source initiative that has made the tool available to scientists around the world.  The software, named “Model SEED,” has been applied to thousands of microbial genomes.  DeJongh and a team of Hope students also developed “CytoSEED,” a software package that enables viewing, manipulating and analyzing metabolic models created using the “Model SEED” and which itself is used around the globe.

Through the new award, the research team will add the ability to model how the organisms behave in response to changes in their environment.

“The first years of the project were designed to allow us to build models efficiently, but the models are at some level starting points,” Best said.  “One of the primary activities that this grant will fund is increasing the sophistication of the model:  from the capabilities the organism has, to layering on top of that information about when the organism uses those capabilities.”

“We talk about our models as being like roadmaps,” DeJongh said.  “A road map tells you how to get from point A to point B.”

“What we’re trying to do now is use real-time data to figure out how the organism is using different capabilities at different times,” DeJongh said.  “We’re hoping to develop statistical tools to elucidate parts of the system that we don’t understand yet.”

Crucially, the grant is also enabling the researchers to test and improve the accuracy of the computer models through laboratory work.  “We can make predictions using the model, and then grow a sample in the lab and use genetic sequencing to determine whether or not those conditions are actually happening,” Best said.

A key component of the project, Best and DeJongh noted, is that the research will continue to involve students at both Hope and Dordt.  Students at Hope work directly as collaborative researchers with Best and DeJongh during both the school year and summer, but the project is also integrated into the academic program in other ways.  For example, computer science students at the college revised the “CytoSEED” program to make it compatible with a new version of the platform on which it operates, and the researchers envision involving students in biology courses to conduct laboratory validation of the computer model.