This work proposes a fuzzy modeling-based approach for describing signal transduction networks. Many key steps in signal transduction mechanisms have been investigated and described qualitatively in the literature, however, only little quantitative information is available. Fuzzy models can make use of this situation as fuzzy rules can be based upon the qualitative information that is found in the literature whereas training of the model can be performed with data that is available. This combination of a fuzzy rule set based upon qualitative information with parameters to be determined from data can result in models where fewer parameters need to be estimated than if fundamental or black-box models were used. The presented fuzzy modeling procedure is used to describe two signal transduction pathways, one for IL-6 and one for TNF-a signaling. It is shown that the resulting models are capable of capturing the dynamics of key components of both signal transduction pathways.
Chemical Engineering Science 64, No. 9, pp. 2044-2056 (2009)