Diffusion approximations to the chemical master equation only have a consistent stochastic thermodynamics at chemical equilibrium

Jordan M. Horowitz
DOI: https://doi.org/10.48550/arXiv.1506.07797
2015-07-24
Abstract:The stochastic thermodynamics of a dilute, well-stirred mixture of chemically-reacting species is built on the stochastic trajectories of reaction events obtained from the Chemical Master Equation. However, when the molecular populations are large, the discrete Chemical Master Equation can be approximated with a continuous diffusion process, like the Chemical Langevin Equation or Low Noise Approximation. In this paper, we investigate to what extent these diffusion approximations inherit the stochastic thermodynamics of the Chemical Master Equation. We find that a stochastic-thermodynamic description is only valid at a detailed-balanced, equilibrium steady state. Away from equilibrium, where there is no consistent stochastic thermodynamics, we show that one can still use the diffusive solutions to approximate the underlying thermodynamics of the Chemical Master Equation.
Statistical Mechanics,Chemical Physics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is whether in chemical reaction networks, diffusion approximations (such as the Chemical Langevin Equation (CLE) and the Low Noise Approximation (LNA)) can inherit the stochastic thermodynamic properties of the Chemical Master Equation (CME). Specifically, the author explores whether these diffusion approximations can maintain a stochastic thermodynamic description consistent with the CME in non - equilibrium states, especially in relation to the correlation between entropy flow and heat flow. ### Main problems of the paper 1. **Thermodynamic consistency of diffusion approximations in equilibrium states** - The author finds that only in the detailed - balance equilibrium steady state can diffusion approximations (such as CLE and LNA) maintain a stochastic thermodynamic description consistent with the CME. This means that in the equilibrium state, the entropy flow and heat flow of the diffusion approximations are consistent with those of the CME. 2. **Thermodynamic approximation in non - equilibrium states** - In non - equilibrium states, diffusion approximations cannot provide a stochastic thermodynamic description consistent with the CME. Nevertheless, the author shows that even in this case, diffusion approximations can still be used to approximate the heat flow of the CME, but this approximation will not be proportional to the entropy production of the trajectory as in the equilibrium state. ### Key conclusions - **Thermodynamic consistency in equilibrium states**: In the detailed - balance equilibrium steady state, diffusion approximations (such as CLE and LNA) can inherit the stochastic thermodynamic properties of the CME, especially the consistency of entropy flow and heat flow. - **Thermodynamic approximation in non - equilibrium states**: In non - equilibrium states, although diffusion approximations cannot provide a stochastic thermodynamic description consistent with the CME, they can still be used to approximate the heat flow of the CME, although this approximation is not proportional to the entropy production of the trajectory. ### Formula summary - **Entropy flow in CME**: \[ d\sigma_{\text{CME}}^{\rho}(X_t) = \ln \frac{a_{\rho}(X_t)}{a_{-\rho}(X_t + \nu_{\rho})} \] - **Entropy flow in CLE**: \[ d\sigma_{\text{CLE}}^{\rho}(X_t) = \frac{a_{\rho}(X_t) - a_{-\rho}(X_t) - \nu_{\rho} \cdot \partial_x (a_{\rho}(X_t) + a_{-\rho}(X_t))/2}{(a_{\rho}(X_t) + a_{-\rho}(X_t))/2} \] - **Entropy flow in LNA**: \[ d\sigma_{\text{LNA}}^{\rho}(X_t, \phi_t) = \frac{a_{\rho}(\phi_t) - a_{-\rho}(\phi_t) + (X_t - \phi_t) \cdot \partial_{\phi} (a_{\rho}(\phi_t) - a_{-\rho}(\phi_t))}{(a_{\rho}(\phi_t) + a_{-\rho}(\phi_t))/2} \] Through these formulas, the author shows that in the equilibrium state, diffusion approximations can be consistent with the thermodynamic properties of the CME, while in non - equilibrium states, additional approximation methods are required to handle the heat flow.