How to simulate a rule-based model

Once rule-based features are specified in the BioModel Physiology, several numerical methods can be used to run simulations. As is the case with all Virtual Cell BioModels, a Rule-Based BioModel may have multiple Applications, including deterministic and stochastic ( rule-based) applications; in addition, for rule-based models network-free applications can be created. Each Application, in turn, can have multiple Simulations, in which different numerical methods or conditions are used.

Rule-based Deterministic and Stochastic Applications

When a deterministic or stochastic application of a rule-based model is created, the BioNetGen network-generation algorithm is used to generate the complete network of species and reactions that is obtained by applying the set of reaction rules to the set of seed species. Initial amounts of seed species are defined in the Network specification tab. Network specification parameters (number of iterations to apply rules and maximum number of molecules per species in the generated network) are used to control the size of the network, which otherwise can become infinitely large. The generated species and reactions can be viewed in the Network or the Reaction Diagram view.

After the network is generated, determiministic or stochastic simulations can be run as for any Application.

Network-Free Applications

In network-free applications of a rule-based BioModel, a network is not generated. Instead, the amounts of model Observables and initially defined species are simulated using the NFSim engine. Simulation parameters are defined in the NFSim Simulation Editor.

Rule-Based MathModel

The MathModel for deterministic or stochastic applications generated for rule-based models have a standard VCell structure that includes the full list of species and reactions, and includes observables defined by functions. The MathModel for rule-based applications has a unique syntax.