To this date many approaches exist that model a genetic regulatory network in order to elucidate the dynamics of the system. These methods focus, however, mainly on small-scale regulatory models. In this thesis we use a formal approach over qualitative large-scale regulatory networks, that models the equilibrium shift of the cell molecules between two steady states. We test the coherency between the network topology and gene expression data, by using a general interaction logical causal rule. The outputs of our approach are to measure the consistency of our data, diagnose inconsistent regions of the network with respect to the experimental data, and infer the qualitative variation of new network molecules. Our method reasons over the whole network of interactions using e