SOS Experiment

Description

Implementation of Experiment for simulating SBML models using the SOS: (S)BML (O)DE (S)olver library.

Details

The general workflow for running a simulation:

1. Create or import an SBML Model.
2. Customize the model, for example by adding perturbation Events.
3. Wrap the model in a SOSSubject, e.g. new("SOSSubject", model).
4. Optionally construct a SOSDesign for running the experiment in batch over several sets of model parameter settings.
5. Optionally construct a SOSProtocol for specifying the time points and other parameters controlling the simulation.
6. Construct an instance of this class that groups the subject, design and protocol.
7. Run simulate on the SOSExperiment, optionally specifying the number of iterations and the random seed.
8. Analyze the returned SOSResult, perhaps starting by converting it to a time series with as.ts and making some plots.

Objects from the Class

Objects can be created by calls of the form new("SOSExperiment", ...).

Slots

protocol:

Object of class SOSProtocol, where the simulation parameters are specified.

design:

Object of class SOSDesign, specifying model parameters for each run of a batch experiment.

subject:

Object of class SOSSubject, containing the Model to be simulated.

result:

Object of class SOSResult containing the result of the simulation.

Extends

Class Experiment, directly.

Methods

simulate

signature(object = "SOSExperiment"): simulate(object, nsim = 10, seed, ...): Simulates the model in the subject slot according to the design points in design and parameters in protocol for nsim iterations, using seed as the random seed. Returns an instance of SOSExperiment, which now should include a SOSResult for analysis.

Author(s)

Michael Lawrence

References

See http://www.tbi.univie.ac.at/~raim/odeSolver/ for more information on the SBML ODE Solver library.

See Also

The simulate method on SBMLDocument is a shortcut, but most users will probably find the above approach most useful.