At least one particle number in the initial state is too big.

[billg]

Date: Tue, 17 Nov 2009 17:53:06 -0500
Subject: Re: At least one particle number in the initial state is too big.

Hello,

I am seeing the same error message as Mridul while trying to run an imported SBML test model using the stochastic integrator under the "concentration" framework. My test model has 6 state variables, with initial conditions set to values ranging from 0.2 - 1.0 mol/l. With the deterministic integrator, the model runs fine, however with the COPASI stochastic integrator I get the following error:

>EXCEPTION 2009-11-17 14:42:01<
At least one particle number in the initial state is too big.

I also see this error message while attempting to run the stochastic integrator on the Brusselator example model.

I am using COPASI 4.5 (Build 30) on Mac OSX 10.5.8.

Thanks.

Bill

[sven]

Date: Wed, 18 Nov 2009 05:30:14 -0500
Subject: Re: Re: At least one particle number in the initial state is too big.

Hello Bill,

> I am seeing the same error message as Mridul while trying to run an imported SBML test model using the stochastic integrator under the "concentration" framework. My test model has 6 state variables, with initial conditions set to values ranging from 0.2 - 1.0 mol/l. With the deterministic integrator, the model runs fine, however with the COPASI stochastic integrator I get the following error:
>
> >EXCEPTION 2009-11-17 14:42:01<
> At least one particle number in the initial state is too big.

the problem is most likely the same: When doing stochastic simulations, Copasi internally always works with particle numbers (this is how stochastic simulation is intended to work). If, for example, your initial concentration is 1 mol/l, and the volume is set to 1 liter (as it is the case for many models) this means the particle number will be about 6*10^23. This is much to large for any meaningful stochastic simulation.

As Stefan wrote, the easiest way to change this is to set the volume to a much smaller value.

Sven