Logical Modeling and Analysis of Regulatory Genetic Networks in a Non Monotonic Framework
Résumé
We present a constraint based declarative approach for analyzing qualitatively genetic regulatory networks (GRNs) with the discrete formalism of R. Thomas. For this purpose, we use the logic programming technology ASP (Answer Set Programming) whose related logic is non monotonic. Our aim is twofold. First, we give a formal modeling of both Thomas’ GRNs and biological data like experimental behaviors and gene interactions and we evaluate the declarative approach on three real biological applications. Secondly, for taking into account both gene interaction properties which are only generally true and automatic inconsistency repairing, we introduce an optimized modeling which leads us to exhibit new logical expressions for the conjunction of defaults and to show that they can be applied safely to Thomas’ GRNs.