A partial-propensity variant of the composition-rejection stochastic simulation algorithm for chemical reaction networks

摘要

We present the partial-propensity stochastic simulation algorithm with composition-rejectionsampling (PSSA-CR). It is an exact formulation of the stochastic simulation algorithm (SSA) forwell-stirred systems of coupled chemical reactions. The new formulation is a partial-propensityvariant [R. Ramaswamy, N. Gonzalez-Segredo, and I. F. Sbalzarini, J. Chem. Phys. 130, 244104(2009)] of the composition- rejection SSA [A. Slepoy, A. P. Thompson, and S. J. Plimpton, J. Chem.Phys. 128, 205101 (2008)]. The computational cost of this new formulation is bounded by aconstant for weakly coupled reaction networks, and it increases at most linearly with the number ofchemical species for strongly coupled reaction networks. PSSA-CR thus combines the advantagesof partial-propensity methods and the composition-rejection SSA, providing favorable scaling of thecomputational cost for all classes of reaction networks.