Parameter identifiability and model selection for partial differential equation models of cell invasion

Abstract

When employing a mechanistic model to study biological systems, practical parameter identifiability is important for making predictions in a wide range of scenarios, as well as for understanding the mechanisms driving the system behaviour. We argue that parameter identifiability should be considered alongside goodness-of-fit and model complexity as criteria for model selection. To demonstrate, we use a profile likelihood approach to investigate parameter identifiability for four extensions of the Fisher--KPP model, given experimental data from a cell invasion assay. We show that more complicated models tend to be less identifiable, with parameter estimates being more sensitive to subtle differences in experimental procedures, and require more data to be practically identifiable. The results from identifiability analysis can inform model selection, as well as data collection and experimental design.