Alejandro F. Villaverde

Alejandro F. Villaverde, Julio R. Banga

The concept of dynamical compensation has been recently introduced to describe the ability of a biological system to keep its output dynamics unchanged in the face of varying parameters. Here we show that, according to its original definition, dynamical compensation is equivalent to lack of structural identifiability. This is relevant if model parameters need to be estimated, which is often the case in biological modelling. This realization prompts us to warn that care should we taken when using an unidentifiable model to extract biological insight: the estimated values of structurally unidentifiable parameters are meaningless, and model predictions about unmeasured state variables can be wrong. Taking this into account, we explore alternative definitions of dynamical compensation that do not necessarily imply structural unidentifiability. Accordingly, we show different ways in which a model can be made identifiable while exhibiting dynamical compensation. Our analyses enable the use of the new concept of dynamical compensation in the context of parameter identification, and reconcile it with the desirable property of structural identifiability.

Alejandro F. Villaverde, Julio R. Banga

Dynamical compensation (DC) has been recently defined as the ability of a biological system to keep its output dynamics unchanged in the face of varying parameters. This concept is purported to describe a design principle that provides robustness to physiological circuits. Here we note the similitude between DC and Structural Identifiability (SI), and we argue that the former can be explained in terms of (lack of) the latter. We propose to exploit this fact by using currently existing tools for SI analysis to perform DC analysis. We demonstrate the feasibility of this approach with four physiological circuits, for which we confirm the correspondence between DC and lack of SI. We also warn that care should we taken when using an unidentifiable model to extract biological insight, since lack of SI can be the result of an inappropriate choice of model structure and therefore not necessarily a sign of biological robustness.