Thesis : Semantic approaches for the meta-optimization of complex biomolecular networks

PhD defense: Ali AYADI

Team: SDC

Title: Semantic approaches for the meta-optimization of complex biomolecular networks

Abstract: Systems biology models aim to understand the behaviour of a cell trough a complex biomolecular network. In the literature, most research focuses on modelling isolated parts of this network, such as metabolic networks.
However, to fully understand the cell’s behaviour we should analyze the biomolecular network as a whole. Available approaches do not address these requirements sufficiently. In this context, we aim at developing a platform that enables biologists to simulate the state changes of biomolecular networks with the goal of steering their behaviours.
The platform employs rules, knowledge and experience, much like those that an expert biologist might derive. This platform consists of four modules: a logic-based modelling module, a semantic modelling module, a qualitative discrete-time simulation module and an optimization module. For this purpose, we first present a logic-based approach for modelling complex biomolecular networks including the structural, functional and behavioural aspects. Next, we propose a semantic approach based on four ontologies to provide a rich description of biomolecular networks and their state changes. Then, we present a method of qualitative discrete-time simulation to simulate the biomolecular network behaviour over time. Finally, we propose a multi-objective optimization method for optimizing the transittability of complex biomolecular networks in which we take into account various criteria such as minimizing the number of external stimuli, minimizing the cost of these stimuli, minimizing the number of target nodes and minimizing patient discomfort. Based on these four contributions, a prototype called the CBNSimulator was developed. We describe our approaches and show their applicability through real cases studies, the bacteriophage T4 gene 32, the phage lambda, and the p53 signaling network. Results demonstrate that these approaches provide the necessary elements to model, reason and analyse the dynamic behaviour and the transition states of complex biomolecular networks.

The jury is composed of Cecilia ZANNI-MERK Professor, INSA Rouen/University of Normandie, LITIS (Thesis director), Saoussen KRICHEN Professor, ISG de Tunis/University of Tunis, LARODEC (Thesis director), Claudia FRYDMAN Professor, University of Aix-Marseille, LSIS (Reviewer), Lina SOUALMIA Associate professor with habilitation, University of Rouen, LITIS (Reviewer), Edward SZCZERBICKI Professor, University of Newcastle Australia, François de BERTRAND de BEUVRON Associate professor, INSA Strasbourg, ICube and Olivier POCH, Senior researcher, Université de Strasbourg, ICube.

The defense will be held in English and will take place on Friday September 28th, 2018, at 3:00 pm in the amphitheater A302 of the Pole API building in Illkirch (300 boulevard Sebastien Brant).