Viability, invariance and reachability for controlled piecewise deterministic Markov processes associated to gene networks

UniversitéParis-Est, Laboratoire d’Analyse et Mathématiques Appliquées, UMR 8050, Boulevard Descartes, Cité Descartes, 77450 Champs-sur-Marne, France
Dan.Goreac@univ-mlv.fr

Received: 11 February 2010
Revised: 22 November 2010

Abstract

We aim at characterizing viability, invariance and some reachability properties of controlled piecewise deterministic Markov processes (PDMPs). Using analytical methods from the theory of viscosity solutions, we establish criteria for viability and invariance in terms of the first order normal cone. We also investigate reachability of arbitrary open sets. The method is based on viscosity techniques and duality for some associated linearized problem. The theoretical results are applied to general On/Off systems, Cook’s model for haploinsufficiency, and a stochastic model for bacteriophage λ.