Laboratoire de Physique Théorique
Toulouse - UMR 5152

Time is a major constraint for life. Biological systems need to adapt and respond accurately to changes in their environment. This is difficult to achieve : usually, as we drive a system out of equilibrium, the system lags behind the desired state due to information loss, which, for thermodynamic systems, is in the form of dissipation. One solution is to drive the system very slowly, but for biological processes speed is essential. So, how can we compensate for this lag to move the system along a controlled trajectory in a finite time ? Can we find protocols to assist this driving via an auxiliary field that would always keep the system on the desired trajectory ? In this talk, I will present a method to control the set of trajectories in stochastic networks that has close analogies with quantum adiabatic protocols (transitionless quantum driving). We will explore the implementation of these ideas with two example applications in biology : i) the evolution of populations under selective pressure and ii) in the context of protein folding biophysics.

In the first example, I will show how to manipulate the evolutionary dynamics of cell populations by controlling selection [1] (e.g., dosage control of drugs). This could have wide potential applications, from therapeutic strategies against complex diseases to crop design in response to climate change. In the second part, I will first briefly mention a general method of such control for Markovian processes using a network theory of master equation systems [2]. As an example, we will discuss a potential naturally adapted driving protocol as a response to heat shock, by molecular chaperones acting as recovery agents for misfolded proteins in cells.

[1] Iram et al., Controlling the speed and trajectory of evolution with counterdiabatic driving, Nature Physics 17, 135 (2021).

[2] Ilker et al., Counterdiabatic control of biophysical processes, arXiv:2106.07130 (2021).