Non linear physics for early immune recognition

Conference Dates

July 31-August 4, 2017


Subcellular dynamics emerge from the interactions of molecules of many different types, and it seems a priori hopeless to build predictive theories, similar to what is done in physics. In this talk, I will use the example of early immune detection to illustrate how approaches inspired by physics allow us to untangle the biological interaction “hairball”. We use in silico evolution algorithms developed in our group to reverse engineer mechanisms allowing for a sensitive and specific immune recognition similar to what is observed in adaptive immunity. This led us to the discovery of the so-called “adaptive sorting” principle which relies on the combination of proofreading mechanisms with internal negative feedforward/feedback interactions. Adaptive sorting mechanisms present counter-intuitive properties, such as ligand antagonism and non monotonic dose response curves. We will present experimental validations of our model, in collaboration with quantitative immunologist Grégoire Altan-Bonnet (NIH, Bethesda).

24.pdf (174 kB)

This document is currently not available here.