Decoupling of fast and slow variables in chemical Langevin equations with fast and slow reactions

Under the assumption of existence of a macro-scopically infinitesimal time increment, biochemical reaction network are accurately represented by chemical Langevin equation systems. In this work, we consider such networks in the presence of fast and slow reactions. With the ultimate goal of proposing a systematic framework to derive non stiff models of the slow dominant dynamics, we focus on the preliminary step of obtaining a new stochastic differential equation system with decoupled fast and slow variables in order to apply, in future work, model reduction techniques available for such systems