A growing demand for tools to build and decrypt genetic networks that control cellular processes is felt in biology. We argue that the use of the declarative approach is relevant and applicable to answer questions from biologists about these networks, which are in general partially known. The main idea is to model knowledge about both the structure and the dynamic of a network by a set of constraints representing all the solutions, to check its consistency, to repair a possible inconsistency by an automatic constraint removal, and to infer properties on the structure and dynamic of the network. To demonstrate the feasibility of the approach, we formalize the discrete networks of R. Thomas and relevant biological properties, offer a tool based on constraint logic programming in cooperation with a SAT solver, and validate it on significant biological applications.
|Institution||Université Joseph Fourier - Grenoble 1|