[bionet.software] Kinetic Simulation <-- program evaluation

water@blake.u.washington.edu (Hong Li) (01/16/91)

Hi,
      We all know that there are many programs available for computer
simulation of chemical reactions.  THe speed of the simulation is still
the main concern for microcomputer programs, especially for those can
estimate kinetic parameters with fitting.  I would like to find out which
program has the best (fastest) algorithm.
      The following is a set of enzymatic reactions I involved.  I have
a program which took 3 minutes to finished the simulation process 
described below.
      Programs with simple algorithms will take hours to finish this job.

      Computer  : IMB-AT Compatable  (with 287 math coprocessor)
      CPU Speed : 16 Mhz

species and initial concentrations:
	e    1.52e-6
	o    9.75e-4
	al   1e-3
	ad   0
	h    0
	eo   0
	eo1  0
	eh   0
	eoal 0
	ehad 0

reactions and k's
	e + o -> eo       k=4.5e7
	eo -> e + o       k=2.3e4
	eo -> eo1         k=620
	eo1 -> eo         k=64
	eo1 + al -> eoal  k=9.7e5
	eoal -> eo1 + al  k=84
	eoal -> ehad      k=380
	ehad -> eoal      k=600
	ehad -> eh + ad   k=17
	eh + ad -> ehad   k=2.8e6
	eh -> e + h       k=5.5
	e + h -> eh       k=1.1e7

reaction time:
      50 seconds

accuracy:
      1.0e-5 relative local error.

      The program I used is written in C and is a stand-alone program,
ie. no separate compilation of formula (as in some FORTRAN programs) is
required.
      If you have a program handy,  I would appreciate your trying of
similar simulations and letting me know the result, better or worse.
      If the program you have won't move at all, try to reduce the reaction
time.
      I will post the summary if there are enough responds.

Thank you very much for reading.

Please send respond to:
      water@blake.u.washington.edu

Hong Li
MS Computer Science, Ph.D. candidate
Department of Chemistry
University of Wasington
Seattle, WA 98195