Energy and information flows in autonomous systems

Abstract

Multi-component molecular machines are ubiquitous in biology. We review recent progress on describing their thermodynamic properties using autonomous bipartite Markovian dynamics. The first and second laws can be split into separate versions applicable to each subsystem of a two-component system, illustrating that one can not only resolve energy flows between the subsystems but also information flows quantifying how each subsystem’s dynamics influence the joint system’s entropy balance. Applying the framework to molecular-scale sensors allows one to derive tighter bounds on their energy requirement. Two-component strongly coupled machines can be studied from a unifying perspective quantifying to what extent they operate conventionally by transducing power or like an information engine by generating information flow to rectify thermal fluctuations into output power.