Meeting Summary: The last two decades have witnessed impressive breakthroughs in non-equilibrium statistical physics. These have recently allowed us to calculate that some computations in cells are performed with extraordinary thermodynamic efficiency, orders of magnitude greater than computations in modern artificial computers. This raises several deep issues that we will investigate in this working group, including:
1) Is there any meaningful sense in which biochemical systems “compute”, or is that just a fancy way of saying that biochemical networks evolve according to a discrete time Markov process that is highly modular?
2) To what degree do the high thermodynamic efficiencies in cellular computations simply reflect the way that any chemical reaction network that implements those computations has to behave, and to what degree do those efficiencies instead reflect selective pressure during evolution?
3) Do the biochemical processes implementing the “input-output functions” in cellular circuits exploit statistical correlations between their inputs and other variables in the cells in order to improve their thermodynamic efficiency?