Efficient simulation of stochastic gene regulatory networks

Abstract

Gene expression is inherently stochastic, and the dynamics of gene regulatory networks (GRNs) is governed by the Chemical Master Equation (CME). In most cases, the solution of the CME is not available, and the stochastic simulation algorithm (SSA) requires a high computational effort. In this work we illustrate the performance of a method recently developed for the simulation of stochastic gene regulatory networks that allows computational speeds up to 6500 times higher than SSA. Exploiting intrinsic structural properties of GRNs, the method accurately approximates the Chemical Master Equation (CME) with a Partial Integral Differential Equation (PIDE), which is solved numerically by means of a semi-lagrangian method. The method is available within the toolbox SELANSI https://sites.google.com/view/selansi.