Automated reasoning in metabolic networks with inhibition

Robert Demolombe (1) , Luis Fariñas del Cerro (2, 3) , Naji Obeid (2)

1 IRIT - Institut de recherche en informatique de Toulouse
2 IRIT-LILaC - Logique, Interaction, Langue et Calcul
3 CNRS - Centre National de la Recherche Scientifique

The use of artificial intelligence to represent and reason about metabolic networks has been widely investigated due to the complexity of their imbrication. Its main goal is to determine the catalytic role of genomes and their interference in the process. This paper presents a logical model for metabolic pathways capable of describing both positive and negative reactions (activations and inhibitions) based on a fragment of first order logic. We also present a translation procedure that aims to transform first order formulas into quantifier free formulas, creating an efficient automated deduction method allowing us to predict results by deduction and infer reactions and proteins states by abductive reasoning.

Mots clés: Logical model; Inhibition; Metabolic pathways; Automated reasoning

Domaines: Intelligence artificielle [cs.AI]; Apprentissage [cs.LG]; Logique en informatique [cs.LO]; Informatique et langage [cs.CL]

Robert Demolombe, Luis Fariñas del Cerro, Naji Obeid. Automated reasoning in metabolic networks with inhibition. 13th International Conference of the Italian Association for Artificial Intelligence (AI*IA), Dec 2013, Turin, Italy. pp. 37-47, ⟨10.1007/978-3-319-03524-6_4⟩. ⟨hal-01159682⟩