Michael Stumpf

Time and Condition Dependent Biological Networks

In biology molecular interaction networks are used to summarize the interactions between proteins and other molecules inside a cell. Generally they are considered as static objects, but in reality they are highly dynamic and change in response to environmental and internal signals and processes; but also in case-control studies we would expect the regulatory networks to differ between afflicted carriers and healthy controls. Here we develop a set of approaches that capture this dynamic behaviour. Surprisingly, perhaps, the underlying mathematical framework of Bayesian non-parametric approaches is also

computationally affordable. We have been applying this methodology in a range of different contexts that range from fundamental developmental processes in model organisms all the way to complex molecular networks underlying motor-neuron diseases. Based on our work to date it has already become clear that allowing for the networks to change either over time or between different conditions offers a much more natural description of real-world biological systems.
I will conclude by discussing the functional and evolutionary implications that result from this more detailed perspective.
See bio