Associate Professor of Computational Biology, University of Pittsburgh School of Medicine; Ph.D. in Chemical Physics, University of Colorado, 1998; Reviewer for several journals including J. Chem. Phys., Bioessays, and BMC Bioinformatics; 2007 Junior Faculty Travel Award; Executive Committe Member, Center for Nonlinear Studies at Los Alamos National Laboratory, 2005-2007; Member, American Chemical Society, American Physical Society, and American Association of Immunologists.
Rule-based Modeling of Biochemical Networks
Friday, May 15, 2:30 – 3:30 PM
1175 Benedum Hall
Signaling within cells generally involves protein-protein interactions, which can produce myriad protein complexes. Such protein-protein interactions can be represented compactly and precisely using graphical reaction rules, which can be processed automatically to obtain a chemical reaction network. However, reaction networks implied by typical sets of rules are often too large for conventional simulation procedures to handle. To address this challenge, we have developed a kinetic Monte Carlo method that can take advantage of a rule-based model specification. Rules are used directly to advance a simulation, thus avoiding the computationally expensive step of generating the underlying chemical reaction network implied by the rules. Unlike previously proposed methods that adaptively generate species and reactions in response to network activity, the method is not overwhelmed when the likelihood of encountering new species each time a reaction fires becomes high. The method is applied to characterize the interaction of a trivalent ligand with a bivalent cell-surface receptor and an analogous example of surface molecule aggregation initiated by complexes within the cell. The results of the simulations suggest that extremely large aggregates of signaling molecules may form under some physiological conditions.