Simulating the Model with a Stochastic Solver

SimBiology software includes three stochastic solvers. The stochastic solvers more accurately calculate the change in species amounts with a small number of molecules.

1. Get the active configuration set for the model, Mobj.

   cs = getconfigset(Mobj)

    Configuration Settings - default (active)
        SolverType:                  ode15s
        StopTime:                    10
   
      SolverOptions:
        AbsoluteTolerance:           1.000000e-006
        RelativeTolerance:           1.000000e-003
        SensitivityAnalysis:         false
   
      RuntimeOptions:
        StatesToLog:                 all
   
      CompileOptions:
        UnitConversion:              false
        DimensionalAnalysis:         true
   
      SensitivityAnalysisOptions:
        Inputs:                      0
        Outputs:                     0
   

   The configset object, cs, contains all the simulation settings, including the stop time and solver.

2. Set the SolverType to the ssa stochastic solver, and list the configuration set again.

   set(cs, 'SolverType', 'ssa');
   cs 

   Notice that the SolverOptions for the stochastic solver are different from the ODE options.

   Configuration Settings - default (active)
        SolverType:                  ssa
        StopTime:                    10.000000
   
      SolverOptions:
        LogDecimation:               1
   
      RuntimeOptions:
        StatesToLog:                 all
   
      CompileOptions:
        UnitConversion:              false
        DimensionalAnalysis:         true
   
      SensitivityAnalysisOptions:
        Inputs:                      0
        Outputs:                     0

3. Change the value for the LogDecimation property.

   cs.SolverOptions.LogDecimation = 10;
   

   Increasing this setting lets you record fewer data points and decrease run time.

4. Run the simulation.

   [t_ssa, x_ssa] = sbiosimulate(Mobj, cs);

5. Plot the results.

   FH1 = figure;
   set(gcf, 'color', 'white');
   plot(t_ssa,x_ssa(:,1:4));
   title('Gene Regulation with SSA Solver');
   xlabel('Time (second)');
   ylabel('Amount (molecule)');
   axis([0, 10, 0, 50]);

   The resulting plot may resemble the following figure.

   

6. Another method to visualize stochastic simulations is through ensemble runs.

   simDataObj = sbioensemblerun(Mobj, 10, cs);
   sbiosubplot(simDataObj);

   The resulting plot may resemble the following:

   

   • Click Back or Forward to navigate through the plots.

   Note   To remove the model object from the Workspace type delete(Mobj);

See Also

• Simulating Models

• Stochastic Solvers — Brief description of when to use stochastic solvers and the types available in the software.

Back to Top

Simulating the Model

More Examples of Using the Command Line

© 1984-2012- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS