Hybrid approximation of stochastic process algebras for systems biology

We present a technique to approximate models of biological systems written in a "distilled" version of stochastic Concurrent Constraint Programming (sCCP), a stochastic programming methodology based on logic programming. Our technique automatically associates to a stochastic model an hybrid automaton, i.e. a dynamical system where continuous and discrete dynamics coexist. The hybrid automata generated in this way are, in certain cases, capable of capturing aspects of the dynamics of stochastic processes that are lost in approximations based solely on ordinary differential equations. In particular, they work better for those systems whose sCCP model contains explicit logical mechanisms of control. In the paper we outline the general technique to perform this association and we discuss some issues related to its applicability.