Joint Probability Distribution of mRNA and Protein Molecules in a Stochastic Gene Expression Model
Abstract
Stochastic modeling of gene expression is a classic problem in theoretical biophysics. However, models formulated via chemical master equation have long been considered analytically intractable unless burst approximation is applied. This article shows that general stochastic gene expression models with an arbitrary number of gene states admit direct analysis. Based on chemical master equation and high-dimensional binomial moment method, we derive recurrence relations for binomial moments in steady state, yielding analytical expressions to arbitrary order in a hierarchical manner. Subsequently, the joint probability mass function of mRNA and protein copy number can be reconstructed. An algorithm is developed for numerical computation. Particularly, explicit expressions for low-order cumulants are presented. Compared with models under burst approximation, the first-order cumulant remains exact, whereas the second-order cumulant typically differs. We estimate the difference between two second-order binomial moments using functional analysis, therefore evaluating the validity of burst approximation.