About the Technical Committee
The current focus of this TC is to promote model sharing through publications. Publications provide an opportunity to share models through a media that virtually every researcher encounters. A model shared directly through a publication will not only facilitate review of the publication, but also sharing of the model to the wider journal audience. Some questions are: What is to be shared? Where will the shared entity be stored? Who pays for it? How can we make this happen?
What is a “Model”?
A model is a formal description of a biological system. It is the description of the biology that will eventually be converted into an executable form. The model is not the computer software or the binary executable because that can be in many different forms and implemented in many different ways.
When you want to make an observation about how your model behaves, you perform what is known as a ‘simulation experiment’. This includes running a time course; measuring parameter sensitivities; performing a flux balance analysis; etc.
In non-exchangeable modeling, the executable/source code/algorithms are typically a combination of the two above things, sometimes inextricably linked. Part of the goal here is to be able to separate the model from the experiment, and to divorce both from the realm of source codes and executables entirely.
The minimum information you need is the system of equations, but many modeling languages (like SBML) store much more information than just that. SBML is an example of a model description language that can be converted into computer source code, such as Matlab, C etc and then executed. The advantage of this is that we retain all the biological information in the SBML and we’re not constrained to one particular type of mathematical formalism. SBML models for example can be translated into ODE, stochastic, Boolean or FBA models. Plus at a future date we can retarget the SBML to a different computer language.