DAISY (Differential Algebra for Identifiability of SYstems) is a software tool to perform structural identifiability analysis for linear and nonlinear dynamic models described by polynomial or rational ODE equations* (for example Michaelis-Menten) with either known or unknown initial conditions. ODE models in systems biology, medical research (physiology, cancer research for tumor growth and individualized treatments, HIV studies,...), epidemiology, ecology and many other areas, often contain many unknown parameters to be estimated from experimental data. Identifiability is a prerequisite for parameter estimation to provide reliable and accurate estimates.
DAISY is able to distinguish if the model has only one (global identifiability), or a finite number (local identifiability), or an infinite number of parameter solutions (nonidentifiability). These solutions equivalently describe the experimental I/O data, but may predict different behaviors of the internal (unmeasured) variables. In case of locally identifiable models, DAISY provides also the exact number of parameter solutions.
DAISY does not require experimental data. Thus it is useful especially in physiology and clinical medicine to reject nonidentifiable models before performing the experiment, usually expensive and difficult due to the presence of many constraints.
DAISY does not require in-depth understanding of the mathematical tools.
Practical examples of use of DAISY are provided.
* Models described by equations involving non polynomial, for example exponential or synusoidal, functions can in general be rendered polynomial so that Daisy can correctly test their identifiability.
G. Bellu, Department of Mathematics, University of Cagliari, Cagliari, Italy
M. P. Saccomani, Department of Information Engineering, University of Padova, Padova, Italy
S. Audoly, Department of Structural Engineering, University of Cagliari, Cagliari, Italy
L. D'Angió, Department of Mathematics, University of Cagliari, Cagliari, Italy
Maria Pia Saccomani
Address: Department of Information Engineering, Via Gradenigo 6/B, 35131 Padova, Italy
Telephone: +39-49-8277628
E-mail address: pia@dei.unipd.it
We acknowledge the significant contribution of Dr. Karl Thomaseth to the compilation of DAISY in both Windows and MacOS operating systems and for useful discussions.
We also thank Nicole Dai Prà and Eleonora Miani for the development of the DAISY user interface.