Modelling, observability and experiment: a case study - Positive feedback loop in a genetic regulatory network -
Abstract
We propose an new methodology for modelling of biological regulatory networks inspired by the design and validation of large computing systems. We take into account the capability to validate a model by a set of biological experiments. So defining a model goes with experimental methods and conditions to validate or invalidate it. As in the design of large sized softwares, we will distinguish two activities: first to build an accurate model specifying the assumed behaviour, second to design plans of experiments to verify a posteriori the model predictions. We wish to experiment, through the case of the modelling of the mucus production by the bacterium Pseudomonas aeruginosa, the application of this working methodology.