Title: ``Reaction Modeling Suite: An Integrated Modeling Environment for Reaction Networks and its Biological implications''
Speaker: Dr. Venkat Venkatasubramanian, Laboratory for Intelligent Process Systems, School of Chemical Engineering, Purdue University
Place: Stanley Coulter (SC) 239; Tuesday, 4:30pm

Abstract

One of the important recent developments in biology is the quantitative systems perspective. Rather than investigate individual genes or proteins one at a time, which has been the successful mode of biology for the past fifty years, systems biology investigates the integrated behavior and relationships of all of the elements in a particular biological system in a quantitative manner. Driving this approach is the realization that the amount and rate of accumulation of biological information is increasing rapidly. However, we are approaching a situation where the bottleneck to gaining further quantitative understanding of biological systems will no longer be the lack of data but the limitations in developing and solving complex quantitative models to extract knowledge and insight from the data. To model and compute fast enough to make computation a useful aid in the thinking process, we need new computational frameworks to describe biological systems.

While this situation may appear to be somewhat new to quantitative systems biologists, it is a problem that has been faced before by scientists and engineers in modeling complex catalytic reaction networks in the face of data explosion from combinatorial chemistry and high throughput experimentation. To address this challenge, we have developed an integrated automated knowledge extraction environment called the Reaction Modeling Suite (RMS). RMS is an integrated suite of tools based on artificial intelligence and optimization techniques that enables the expert to initiate the kinetic model development in a high-level reaction chemistry language. A software system then interprets this information into a reaction sequence, automatically sets up the appropriate equations, optimizes the model parameters while keeping them in physically and chemically allowed bounds, and does statistical analysis of the results. Thus, a modeling process that normally would have taken days or weeks to complete is done in a matter of hours and one has the results on the validity of the proposed hypotheses quickly. This will be illustrated with examples from catalysis as well as the MAP Kinase pathway.