Stochastic thermodynamics of multicomponent molecular machines

Molecular machines transduce free energy between different forms throughout all living organisms. While truly machines in their own right, unlike their macroscopic counterparts molecular machines are characterized by stochastic fluctuations, overdamped dynamics, and soft components, and operate far from thermodynamic equilibrium. In addition, information is a relevant free-energy resource for molecular machines, leading to new modes of operation for nanoscale engines. Nonequilibrium free energy transduction in molecular machines is typically studied through the lens of stochastic thermodynamics, which permits analysis of thermodynamic quantities like work, energy, entropy, and information in nanoscale stochastic systems far from equilibrium. Many biological and synthetic molecular machines are made up of interacting components coupled together. While individual machine components have been well studied through single-molecule experiments and computational modelling, multicomponent molecular machines are relatively underexplored. Multicomponent machines permit qualitatively new features that will be explored in this thesis, including internal flows of energy and information, and the possibility of simultaneous exposure to different sources of fluctuations. In this thesis I apply existing and novel tools from stochastic thermodynamics to study molecular machines, with a special focus on understanding the behaviour of multicomponent molecular machines. The work in this thesis derives fundamental limits, explores model systems, and develops tools for inference from experimental data, all of which which allow for novel analysis of molecular machines. Ultimately, these efforts lead to the identification of design principles which I hope will help to guide future engineering of synthetic nanomachines.