Extracting robust information from the confrontation of knowledge and observations on a biological system

12/02/2013 - Room C209 at 14:00 - External Seminar
Anne Siegel
The "basic" knowledge on a biological system is generally represented by a graphical model. This prior knowledge is then confronted to experimental datasets in order to propose a dynamical model for the response of the system. Nonetheless, the inference of the dynamical model may be prone to errors. Indeed, there may exist several plausible dynamical models equally capable of representing the observed response of the system. In this talk, we will introduce two different approaches to globally investigate a complete family of feasible models and extract robust information from this family. The first illustration will be related on the inference of boolean model for signaling pathways (common work with J. Saez Rodriguez and C. Guziolowski). At a different biological scale, the second illustration will introduce the prediction of time-series quantitative measures of proteins from a discrete regulatory network (common work with J. Bourdon and D. Eveillard).