Genomes to Life Contractor-Grantee Workshop III
February 6-9, 2005, Washington, D.C.
Genomics:GTL Program Projects
Sandia National Laboratories
26
Microbial Cell Modeling via Reacting/Diffusing Particles
Steve Plimpton* (sjplimp@sandia.gov) and Alex Slepoy
Sandia National Laboratories, Albuquerque, NM
We have developed a simulator called ChemCell [1] that tracks protein interactions within cells and can be used to model signaling, metabolic, or regulatory response. Cell features for microbial cells are represented realistically by triangulated membrane surfaces. Particles represent proteins, complexes, or other biomolecules of interest. They diffuse via 3d Brownian motion within the cytoplasm, or in 2d within membrane surfaces. When particles are near each other, they interact in accord with Monte Carlo rules to perform biochemical reactions, which can represent protein complex formation and dissociation events, ligand binding, etc. ChemCell is similar in spirit to MCell [2] and Smoldyn [3].
In this poster, we focus on the underlying algorithms used for reaction rules. We have recently developed a spatial version of the stochastic simulation algorithm (SSA) due to Gillespie [4] and discuss it’s implementation in ChemCell. We compare it to alternative approaches including the original SSA and the interaction rules recently developed by Andrews and Bray [3]. We also highlight issues with various reaction/diffusion algorithms relevant to parallel implementation within ChemCell, with the eventual goal of enabling whole-cell models of realistic numbers of proteins and other biomolecules.
References
- S. J. Plimpton and A. Slepoy, SAND Report 2003-4509 (2003).
- J. R. Stiles and T. M. Bartol, in Computational Neuroscience: Realistic modeling for experimentalists, edited by E. De Schutter, published by CRC Press, 87-127 (2001).
- S. S. Andrews and D. Bray, Phys Biology, 1, 137-151 (2004).
- D. T. Gillespie, J Comp Phys, 22, 403-434 (1976).
* Presenting author
