Local identifiability analysis of large signaling pathways models

Abstract:
We introduce a simple automatic approach to investigating the identifiability of signaling pathway models. The method splits the set of model parameters into two subsets, (1) those parameters that can be determined by parameter estimation, and (2) those parameters that cannot be determined by parameter estimation and therefore render the model unidentifiable. The method is applied to three published models of major signaling pathways, namely the JAK-STAT, the EGF, and the NF-kB pathway. The performance of the method is compared to a correlation based method proposed by Jacquez and Greif (1985). In all three examples the approach proposed here requires considerably fewer parameters to be fixed to create an identifiable model than the method proposed by Jacquez and Greif.

Keywords:
identifiability, parameter estimation, JAK-STAT, NF-kB, MAP kinase, signal transduction

Wenn Sie Interesse an einer elektronischen Kopie haben, benutzen Sie bitte unser Onlineformular