>Someone who is often overlooked in the BG (Before gemonics) simulation era
>is David Park. He developed a simulation packaged called (I think) Metasim,
>it wasn't so much the package as the algorithm he implemented. He has a
>paper in J theor biol, 1974 where he describes in some depth how to
>dynamically partition reactions which are fast and can be treated as if they
>are in quasi-steady state and those which are not. During the simulation
>these two groups change composition as fast reactions become slow and slow
>reactions become fast. Quite impressive and as far as I know it's never been
>implemented since.The advantage of this is that for large models simulations
>time can be greatly reduced and numerical integration is easier.
The BIOSSIM package developed by the Garfinkel group in the 1970s
included a matrix method for computing corrections to the time
derivatives of the state variables when the corresponding chemicals
participated in rapid equilibria. These reactions were flagged as
fast reactions which were assumed to equilibrate on the time scale of
a single integration step. We used this feature for such processes as
chelation equilibria and other non-enzymatic reactions. This approach
minimized the stiffness of the integration and greatly reduced the
computation time. Also, we assumed that the ligand binding was in a
quasi-steady state, separated out the ligand binding equations for
enzymes, and distributed the enzyme forms algebraically. The
enzymatic rate was then calculated from the instantaneous enzyme
distribution. This too improved stiffness and sped the computation.
As far as I know, no one has implemented Park's idea of dynamically
determining if a reaction is sufficiently fast.
Michael C. Kohn
Laboratory of Computational Biology and Risk Analysis
National Institute of Environmental Health Sciences
P.O. Box 12233, Mail Drop A3-06
Research Triangle Park, NC 27709-2233
919-541-4929 (voice)
919-541-1470 (Fax)
E-mail: kohn at valiant.niehs.nih.gov
Web site: http://valiant.niehs.nih.gov
---
