Bioinformatics Vol. 19 no. 18 2003
pages 2428-2435
© 2003 Oxford University Press
Susceptibility of non-linear systems as an approach to metabolic responses
Departamento de Bioquímica y Biología Molecular I, Facultad de Ciencias Químicas, Universidad Complutense, 28040, Madrid, Spain
Received on January 19, 2003
; revised on April 15, 2003
; accepted on June 20, 2003
Motivation: Recent developments of several analytical techniques and fast sampling procedures are making it possible to study non-linear dynamics on an automated basis. Approaches that suggest experimental setups and fully exploit the data are needed. To prevent unreliable fits of parameters to experimental data, model-independent methods would be advantageous. One such method consists in representing the response of a system by means of a transformation or functional, of an excitation, i.e. a flux of a metabolite.
Results: A functional of unknown form can be expanded in series if its functional derivatives are known. An algorithm for calculating such generalized derivatives from impulse-perturbation experiments was developed. The only assumption was that the implicit kinetics is time-independent, i.e. the system is time-invariant. The method is illustrated on a part of fructose catabolism. A reaction network that involves 14 metabolites and 11 enzymes and includes several branches and feedback loops is considered. It is shown that the method provides good approximations to the responses of this complex system.
Availability: FUNDER is available from http://bbm1.ucm.es/torralba/funder/down/
Supplementary information: http://bbm1.ucm.es/torralba/funder/
Contact: torralba{at}bbm1.ucm.es