12 November 2008
Jasmin Fisher
Microsoft Research
Cambridge, UK
Computational modelling of biological systems is becoming increasingly important in efforts to better understand complex biological behaviours. 'Executable Biology' is a pioneering approach focused on the design of executable computer programs that mimic biological phenomena. While traditional mechanistic models in biology are usually described by diagrams (giving a fairly static picture of cellular processes), executable biology seeks to translate such static diagrams into dynamic models using formal computational methods that were originally designed for the construction and analysis of complex man-made systems (i.e. computers and computer programs). In this talk, I will illustrate the usefulness of this framework to model signalling pathways using the following examples: (1) our modelling work of the EGFR/Notch signalling crosstalk during the process of cells fate determination in C. elegans vulval development. This model brings forward intricate timing considerations in the operation of these signals, which were also validated experimentally; (2) a model of the Notch/Wnt dialogue in keratinocytes, predicting an additional mode of interaction between these two signalling pathways; (3) a more recent model describing metabolic disturbance in fat tissue with relation to diabetes and obesity; and (4) a detailed analysis of a molecular model describing the EGFR pathway, leading to a more abstract view of the different modules of this network.