Advanced modeling and identification techniques for metabolic processes.

This paper will be devoted to demonstrating that a better understanding of complex metabolic processes requires a deep and reliable interpretation of much experimental data. Indeed this aim cannot be satisfied without the use of advanced modeling and identification techniques and their deep critical analysis. In fact, complete procedures should consider all the following steps: (1) definition of scopes, (2) prior information and hypotheses, (3) choice of experimental conditions, (4) derivation of possible classes of models, (5) parameter estimation, (6) errors evaluation, (7) validation and ordering of the identified models. This paper will mainly consider steps 4, 6, and 7 which are certainly the less assessed ones. However some real advances have been obtained in recent years which deserve to be more widely known and applied in practical problems. This paper will review such important contributions and will show, by means of applied examples, how they can be useful in avoiding ambiguities and incorrect interpretation of data and in evaluating the credibility of the inferred results.