07/07/2023 - 14H00 - Salle A08 du Loria (Nancy)

"Sélection et analyse de modèles pour les réseaux biologiques : utilisation des connaissances du domaine et application aux réseaux perturbés dans les pathologies"

Résumé :
Conversion between modelling formalisms is a critical challenge in
systems biology, as it helps to get new insights into complex biological
systems. This thesis proposes an approach to automatically synthesise
Boolean networks from existing reaction networks, enabling the
conversion between two widely used formalisms. The thesis consists of
three contributions. First, we introduce a method for extracting the
structure and dynamics of a reaction network. That is, a graph that
encodes the potential direct influences between the components, and
either a (binarised) timeseries or a (partial) transition graph. Second,
we present an answer-set programming-based approach to synthesise
Boolean networks that comply with the extracted structure and dynamics.
Finally, we evaluate the resulting pipeline on both toy examples and
reaction networks from the Biomodels repository, encoded in the system
biology markup language. Our results demonstrate the effectiveness of
our approach in generating relevant Boolean networks from reaction networks.
Jury :
- Rapporteurs : François Fages, Directeur de Recherche, Inria Saclay
Loı̈c Paulevé, Chargé de Recherche CNRS, Labri
- Autres membres : Examinateurs :
Anna Niarakis, Maı̂tre de Conférences HDR Université d'Evry Val d'Essonne
Encadrants:
Taha Boukhobza Professeur, Université de Lorraine
Malika Smaı̈l-Tabbonne Maı̂tre de Conférences HDR, Université de Lorraine
Thierry Bastogne, Professeur Université de Lorraine