The breakdown of the reaction-diffusion master equation with non-elementary rates

Abstract

The chemical master equation (CME) is the exact mathematical formulation of chemical reactions occurring in a dilute and well-mixed volume. The reaction-diffusion master equation (RDME) is a stochastic description of reaction-diffusion processes on a spatial lattice, assuming well-mixing only on the length scale of the lattice. It is clear that, for the sake of consistency, the solution of the RDME of a chemical system should converge to the solution of the CME of the same system in the limit of fast diffusion: indeed, this has been tacitly assumed in most literature concerning the RDME. We show that, in the limit of fast diffusion, the RDME indeed converges to a master equation, but not necessarily the CME. We introduce a class of propensity functions, such that if the RDME has propensities exclusively of this class then the RDME converges to the CME of the same system; while if the RDME has propensities not in this class then convergence is not guaranteed. These are revealed to be elementary and non-elementary propensities respectively. We also show that independent of the type of propensity, the RDME converges to the CME in the simultaneous limit of fast diffusion and large volumes. We illustrate our results with some simple example systems, and argue that the RDME cannot be an accurate description of systems with non-elementary rates.

8 pages, 3 figures

Cited by in corpus (5)

• Simulation and inference algorithms for stochastic biochemical reaction networks: from basic concepts to state-of-the-art
• Memory functions reveal structural properties of gene regulatory networks
• Minimal cellular automaton model with heterogeneous cell sizes predicts epithelial colony growth
• Diffusion-dynamics laws in stochastic reaction networks
• A Diffusion-Based Embedding of the Stochastic Simulation Algorithm in Continuous Space