We consider a product X of n finite intervals of integers, a map F from X to itself, the asynchronous state transition graph Γ(F) on X that Thomas proposed as a model for the dynamics of a network of n genes, and the interaction graph G(F) that describes the topology of the system in terms of positive and negative interactions between its n components. Then, we establish an upper bound on the number of fixed points for F, and more generally on the number of attractors in Γ(F), which only depends on X and on the topology of the positive circuits of G(F). This result generalizes the following discrete version of Thomas ’ conjecture recently proved by Richard and Comet: If G(F) has no positive circuit, then Γ(F) has a unique attractor. This result also generalizes a result on the maximal number of fixed points in Boolean networks obtained by Aracena, Demongeot and Goles. The interest of this work in the context of gene network modeling is briefly discussed.

To face the frequent lack of quantitative data on the behavior of gene networks, Thomas introduced in the 70’s a boolean method allowing to capture the main qualitative features of the dynamics of such systems [48]. He proved its usefulness in the context of lambda phage genetic switch [48, 55, 50, 49,

The Boolean network is a mathematical model of biological systems, and has attracted much attention as a qualitative tool for analyzing the regulatory system. The stable states and dynamics of Boolean networks are characterized by their attractors, whose properties have been analyzed computationally, yet not much work has been done from the viewpoint of logical inference systems. In this paper, we show direct translations of Boolean networks into logic programs, and propose new methods to compute their trajectories and attractors based on inference on such logic programs. In particular, point attractors of both synchronous and asynchronous Boolean networks are characterized as supported models of logic programs so that SAT techniques can be applied to compute them. Investigation of these relationships suggests us to view Boolean networks as logic programs and vice versa. 1