Volume 9, Number 1, p.p. 45–52
Towards self-modelling of metabolic pathways
C. Bernon,1 D. Capera,2 J.-P. Mano,1 S. Videau,1 and C. Régis1
1
Institut de Recherche en Informatique de Toulouse, University of Toulouse III, 118 route de Narbonne, 31062 Toulouse cedex 9, France
2
Upetec, 10 avenue de l'Europe, 31520 Ramonville St Agne, France
Complexity qualifies biological systems whether it be computational, systems or integrative biology in which huge amounts of data, large scale interaction networks or many coupled levels have to be considered. On the one hand, because satisfactory theories are still missing, modelling and simulations are considered as means for avoiding very costly in vivo experimentation; on the other hand, because of this inherent complexity, the models provided to biologists are usually static ones. Therefore influencing or modifying them in a dynamic way for trying to understand or discover new virtual experiments is usually impossible. A prospective solution in order to overcome this drawback would be to let a model self-build to constantly reflect experimental data and the user’s desiderata. The approach proposed here is based on the adaptive multi-agent paradigm combined with emergent behaviour at the system level coming from interactions between parts of the system (or agents). Agents of the artificial system presented here are based on a four-layer agent model, which enables endowing a multi-agent system with the ability to self-tune, self-organize and “self-evolve” in order to let a biological model build itself. Some preliminary results are also discussed.
Keywords:
cooperation, multi-agent systems, self-organization