Structural Sensitivity of Neural and Genetic Networks

Hedi Ben-Amor (1) , Jacques Demongeot (1) , Sylvain Sené (2, 1)

1 TIMB
2 IXXI - Institut Rhône-Alpin des systèmes complexes

This paper aims at giving new results on the structural sensitivity of biological networks represented by threshold Boolean networks and ruled by Hopfield-like evolution laws classically used in the context of neural and genetic networks. Indeed, the objective is to present how certain changes and/or perturbations in such networks can modify signicantly their asymptotic behaviour. More precisely, this work has been focused on three diferent kinds of what we think to be relevant in the biological area of robustness (in both theoretical and applied frameworks): the boundary sensitivity (external fields, hormone flows, ...), the state sensitivity (axonal or somatic modulations, microRNAs actions, ...) and the updating sensitivity.

Mots clés: robustness; neural networks; genetic regulatory networks; threshold boolean network; structural sensitivity; boundary sensitivity; state sensitivity; updating sensitivity

Domaines: Bio-informatique [q-bio.QM]; Bio-Informatique, Biologie Systémique [q-bio.QM]; Systèmes dynamiques [math.DS]; Réseau de neurones [cs.NE]; Neurosciences

Hedi Ben-Amor, Jacques Demongeot, Sylvain Sené. Structural Sensitivity of Neural and Genetic Networks. MICAI 2008: Advances in Artificial Intelligence, 7th Mexican International Conference on Artificial Intelligence, Oct 2008, Mexico, Mexico. pp.937-986, ⟨10.1007/978-3-540-88636-5_92⟩. ⟨hal-00346566⟩