The thermodynamics of computational copying in biochemical systems

Living cells use readout molecules to record the state of cell-surface receptors that detect ligands in their environment. On the surface, these readouts appear similar to measurements made by computational devices, as extensively studied in the literature following Maxwell’s demon. But at what level do measurements made by biochemical devices map onto computational measurements made, for example, by magnetic devices? Do living cells operate at the thermodynamic limits for taking a measurement? Here, we consider a canonical biochemical network for receptor readout and show how it maps onto a measurement protocol of the type studied in the computational literature. We find that the biochemical network does not achieve thermodynamic limits of efficiency. A key consequence is that biochemical networks face a tradeoff between the energy dissipated for a measurement and the accuracy of the measurement that is more severe than and qualitatively distinct from the tradeoff required thermodynamically. We study this tradeoff. We conclude by demonstrating how biomolecules can be used to achieve optimal protocols – not under the control of a steady-state biochemical network, but rather under the control of exogenously manipulated baths of ATP and ADP. This leads to an experimental system that might be used to test theories on the thermodynamics of computation.