Structural identifiability approach can be used for any type of biosystems represented through data-based mathematical models. Its aim is to compute the solution for the parameters of the models given by an experimental scheme.
Structural identifiability for nonlinear models has been tested using some methods, among which the most important are based on Taylor series, generating series, similarity transformation or differential algebra.
However, until now none of these methods could have been applied for a general nonlinear biomodel. Even with more complex and complete theoretical results and good computation techniques, the implementation algorithms depend on the structure of the model, degree of nonlinearity, number of observables, number of parameters, and so on.
- Applicability for a whole class of non-linear models.
- Better manipulation of structural identifiability problem, from computational and flexibility point of view.
- Computational information are displayed at each step, and not ”all or nothing” response.
- Lack of memory problems can be handled by the user, through the derivation order manipulation.
- Identifiability tableaus summarize the information about the parameters of the model.
- If the model is not structurally globally identifiable, at least structural local identifiability can be tested.