JBIOT-2026
Les 2èmes Journées de la Biologie Théorique
9 et 10 Novembre 2026, Amphi IMAG, UGA, Grenoble
« Processus dynamiques et évolutifs : modèles et concepts »
Les prochaines Journées de la Biologie Théorique seront consacrées au thème : « Processus dynamiques et évolutifs : modèles et concepts » avec des conférences plénières invitées, présentations contributives de chercheurs et postdoctorants ainsi que présentations courtes de doctorants, afin de favoriser les échanges entre communautés et niveaux d’expérience variés. L’édition 2026 mettra plus particulièrement l’accent sur les approches issues de la théorie des systèmes dynamiques et des systèmes complexes pour comprendre les mécanismes d’adaptation et d’évolution dans les systèmes vivants et artificiels. Les notions d’attracteurs, de bifurcations, de chaos, de transitions d’états ou encore de paysages dynamiques, notamment inspirés du paysage de Waddington, constitueront des fils conducteurs des discussions. L’objectif sera d’explorer comment les outils de modélisation et de simulation permettent aujourd’hui de relier dynamiques adaptatives, plasticité et processus évolutifs dans des systèmes biologiques complexes.
Ces journées, gratuites pour les membres de la Société Francophone de Biologie Théorique, sont ouvertes à tous sur inscription. Elles visent à rassembler chercheurs confirmés, jeunes scientifiques et étudiants de divers horizons disciplinaires afin de renforcer les interactions entre communautés scientifiques et de contribuer à l’émergence d’un espace de réflexion interdisciplinaire autour des fondements théoriques du vivant.
Programme
| Lundi 9 novembre | Mardi 10 novembre |
|---|---|
|
|
Session posters sur les pauses café et buffet
Orateurs invités confirmés
In this talk, we present a class of non-local mathematical models developed to investigate normal and abnormal wound healing processes such as keloids (these abnormal processes lead to tissue overgrowth, remodelling and invasion similar to those observed in benign tumours). The models account for non-local cell-cell and cell-matrix interactions via different signalling molecules as well as long-distance cell protrusions. We discuss various analytical and numerical aspects associated with these non-local models from the perspective of biological applications. Due to the complexity of this model, and lack of appropriate data, it is difficult to parametrise it. Nevertheless, we discuss model parametrisation for a simplified local model.
In this talk, we present a population dynamics approach to study the homeostasis of the ovarian follicle population throughout the reproductive lifespan. In mammals, the number of female germ cells continuously decreases over time, whereas in fish, oogenesis is self-renewing. Nevertheless, in both cases, the active germ cells within growing ovarian follicles exhibit remarkable robustness properties.
We revisit a compartmental population dynamics model for ovarian follicle development [1] by incorporating follicular interactions that arise (directly or indirectly) from the regulation exerted by reproductive hormones on their activation, growth, and death rates. The characteristics of the follicle population distribution are obtained thanks to a time scale separation between follicle activation and growth [2,3], which enables us to study the relationship between the growing follicle pool and the quiescent one. Depending on the ratio of activation to growth rates, we perform a full bifurcation analysis and show that the model of growing follicles can exhibit monostability, multistability, or stable oscillatory solutions, arising respectively through saddle-node and Hopf bifurcations [4].
Finally, we discuss how to calibrate such models, either through a parametric approach using population count data from mice [5], or through a nonparametric approach using follicle size distribution data from medaka [6].
1. M. Faddy, E. Jones, and R. Edwards, An analytical model for ovarian follicle dynamics, J. Exp. Zool., 197:173–185, 1976.
2. C. Bonnet, K. Chahour, F. Clément, M. Postel, and R. Yvinec, Multiscale population dynamics in reproductive biology : Singular perturbation reduction in deterministic and stochastic models, ESAIM Proc. Surveys, 67:72–99, 2020.
3. G. Ballif, F. Clément and R. Yvinec, Averaging of a Stochastic Slow-Fast Model for Population Dynamics: Application to the Development of Ovarian Follicles, SIAM J. Appl. Maths., 82(1):359–380, 2022.
4. F. Clément, L. Fostier and R. Yvinec, Well-posedness and bifurcation analysis of a size-structured population model: Application to female gametes dynamics, SIAM J. Appl. Dyn. Syst., 24(3):2427–2472, 2025.
5. G. Ballif, F. Clément and R. Yvinec, Nonlinear compartmental modeling to monitor ovarian follicle population dynamics on the whole lifespan, J. Math. Biol., 89:9, 2024.
6. L. Fostier, M. Lesage, V. Thermes, F. Clément and R. Yvinec, Data-driven PINN inference of oocyte dynamics in fish ovaries, hal-05536280, 2025.
We revisit a compartmental population dynamics model for ovarian follicle development [1] by incorporating follicular interactions that arise (directly or indirectly) from the regulation exerted by reproductive hormones on their activation, growth, and death rates. The characteristics of the follicle population distribution are obtained thanks to a time scale separation between follicle activation and growth [2,3], which enables us to study the relationship between the growing follicle pool and the quiescent one. Depending on the ratio of activation to growth rates, we perform a full bifurcation analysis and show that the model of growing follicles can exhibit monostability, multistability, or stable oscillatory solutions, arising respectively through saddle-node and Hopf bifurcations [4].
Finally, we discuss how to calibrate such models, either through a parametric approach using population count data from mice [5], or through a nonparametric approach using follicle size distribution data from medaka [6].
1. M. Faddy, E. Jones, and R. Edwards, An analytical model for ovarian follicle dynamics, J. Exp. Zool., 197:173–185, 1976.
2. C. Bonnet, K. Chahour, F. Clément, M. Postel, and R. Yvinec, Multiscale population dynamics in reproductive biology : Singular perturbation reduction in deterministic and stochastic models, ESAIM Proc. Surveys, 67:72–99, 2020.
3. G. Ballif, F. Clément and R. Yvinec, Averaging of a Stochastic Slow-Fast Model for Population Dynamics: Application to the Development of Ovarian Follicles, SIAM J. Appl. Maths., 82(1):359–380, 2022.
4. F. Clément, L. Fostier and R. Yvinec, Well-posedness and bifurcation analysis of a size-structured population model: Application to female gametes dynamics, SIAM J. Appl. Dyn. Syst., 24(3):2427–2472, 2025.
5. G. Ballif, F. Clément and R. Yvinec, Nonlinear compartmental modeling to monitor ovarian follicle population dynamics on the whole lifespan, J. Math. Biol., 89:9, 2024.
6. L. Fostier, M. Lesage, V. Thermes, F. Clément and R. Yvinec, Data-driven PINN inference of oocyte dynamics in fish ovaries, hal-05536280, 2025.
Microtubules are key structural elements of living cells that are crucial for cell division, intracellular transport and motility. They are dynamic polymers, which grow and shrink by addition and removal of tubulin dimers at their extremities. Within the microtubule shaft, dimers adopt a densely packed and highly ordered crystal-like lattice structure, which is generally not considered to be dynamic.
Recent experiments have shown that microtubules exhibit a lattice dynamics far away from the extremities. This dynamics manifests itself as localized incorporation of free tubulin into the microtubule shaft, either spontaneously or facilitated by microtubule associated proteins (MAPs), e.g. ATP dependent MAPs such as molecular motors and severing enzymes, but also ATP independent MAPs such as tau. A major biological corollary of such fresh tubulin incorporation events is the increased stability of dynamic microtubules.
The mechanisms underlying tubulin incorporation into the microtubule lattice remain poorly understood. Given the inherent complexity of microtubule dynamics and the limitations of experimental observations, multiple incorporation scenarios are plausible. To address this, we develop minimal coarse-grained models that capture the essential time and length scales of the system. These models allow us to test mechanistic hypotheses in silico and generate predictions to guide experimental efforts.
In this presentation, I theoretically explore mechanisms of lattice dynamics, integrating experimental findings with our existing knowledge from tip dynamics. I focus on two scenarios for lattice turnover: (1) spontaneous turnover at topological defects, where lattice irregularities facilitate dimer exchange, and (2) motor-stimulated dynamics, where molecular motors drive turnover, potentially via long-lived dimer vacancies.
Recent experiments have shown that microtubules exhibit a lattice dynamics far away from the extremities. This dynamics manifests itself as localized incorporation of free tubulin into the microtubule shaft, either spontaneously or facilitated by microtubule associated proteins (MAPs), e.g. ATP dependent MAPs such as molecular motors and severing enzymes, but also ATP independent MAPs such as tau. A major biological corollary of such fresh tubulin incorporation events is the increased stability of dynamic microtubules.
The mechanisms underlying tubulin incorporation into the microtubule lattice remain poorly understood. Given the inherent complexity of microtubule dynamics and the limitations of experimental observations, multiple incorporation scenarios are plausible. To address this, we develop minimal coarse-grained models that capture the essential time and length scales of the system. These models allow us to test mechanistic hypotheses in silico and generate predictions to guide experimental efforts.
In this presentation, I theoretically explore mechanisms of lattice dynamics, integrating experimental findings with our existing knowledge from tip dynamics. I focus on two scenarios for lattice turnover: (1) spontaneous turnover at topological defects, where lattice irregularities facilitate dimer exchange, and (2) motor-stimulated dynamics, where molecular motors drive turnover, potentially via long-lived dimer vacancies.
Differentiation is the process by which a cell acquires a given phenotype through the expression of genes over time. This dynamic is widely understood to result largely from the action of a gene regulatory network (GRN). A particularly fruitful way to study it is to model the dynamics of a cell under the action of its GRN as a stochastic process: GRN inference then becomes the calibration of this model from experimental data. Temporal snapshots of scRNA-seq data make this problem both promising and challenging: as protein trajectories cannot be observed at the single-cell level, it’s not straightforward to establish direct causal relationships between regulator activity and target responses.
Two main strategies can be distinguished. Generic models, such as systems of stochastic differential equations, can be calibrated with non-parametric methods based on optimal transport and benefit from highly efficient existing tools. Mechanistic models, such as piecewise deterministic Markov processes directly parameterized by a GRN, are better suited to incorporating biological priors but require dedicated methods. They also raise the additional challenge of selecting a gene set that is both biologically relevant and tractable for the model.
After discussing these issues, I will focus on CardamomOT, an algorithm that calibrates a PDMP-based model of gene expression within a novel mechanistic optimal transport framework. It jointly reconstructs the GRN and the unobserved protein trajectories. On in silico and experimental datasets, it accurately recovers the velocity fields driving cellular trajectories and the hidden protein levels, alongside reliable GRN structures. Finally, I will show that the calibrated model can serve as a generative model to predict cellular responses to unseen perturbations.
Two main strategies can be distinguished. Generic models, such as systems of stochastic differential equations, can be calibrated with non-parametric methods based on optimal transport and benefit from highly efficient existing tools. Mechanistic models, such as piecewise deterministic Markov processes directly parameterized by a GRN, are better suited to incorporating biological priors but require dedicated methods. They also raise the additional challenge of selecting a gene set that is both biologically relevant and tractable for the model.
After discussing these issues, I will focus on CardamomOT, an algorithm that calibrates a PDMP-based model of gene expression within a novel mechanistic optimal transport framework. It jointly reconstructs the GRN and the unobserved protein trajectories. On in silico and experimental datasets, it accurately recovers the velocity fields driving cellular trajectories and the hidden protein levels, alongside reliable GRN structures. Finally, I will show that the calibrated model can serve as a generative model to predict cellular responses to unseen perturbations.
Inscription et contributions
Gratuit pour les membres de la SFBT (adhésion annuelle), sinon, paiement via les liens ci-dessous :
| Inscription | Montant | Lien Paypal |
|---|---|---|
| Étudiant | 25 € | Inscription |
| Chercheur | 50 € | Inscription |
Une facture peut être obtenue en la demandant à : tresorier@sfbt.org
Formulaire
Pour vous inscrire et si vous souhaitez contribuer en présentant vos travaux (présentation orale ou poster), vous pouvez nous soumettre un titre et un résumé jusqu’au 11 octobre via ce formulaire d’inscription.
La meilleure présentation orale d’un·e doctorant·e sera récompensée par le prix Delattre décerné par la SFBT.
Pour toutes questions, n’hésitez pas à nous contacter : contact@sftb.org.
Lieu de la conférence
Amphi IMAG, UGA, Grenoble
Organisation
Comité d’organisation Mathilde Badoual (Université Paris Cité, IJCLab), Annabelle Ballesta (Inserm, Institut Curie), Ibrahim Cheddadi (Université Grenoble Alpes, TIMC), Eric Fanchon (CNRS, TIMC), Arnaud Millet (INSERM, Biomics), Martin Rosalie (Univestité de Perpignan, LGDP), Angélique Stéphanou (CNRS, TIMC)
Comité scientifique Mathilde Badoual (Université Paris Cité, IJCLab), Annabelle Ballesta (Inserm, Institut Curie), Arnaud Chauvière (Université Grenoble Alpes, TIMC), Ibrahim Cheddadi (Université Grenoble Alpes, TIMC), Laurence Cherfils (Université de La Rochelle, LaSIE), Eric Fanchon (CNRS, TIMC), Nicolas Glade (Université Grenoble Alpes, TIMC), Martin Rosalie (Univestité de Perpignan, LGDP), Angélique Stéphanou (CNRS, TIMC)
Partenaires impliqués :
La Société Francophone de Biologie Théorique (SFBT) https://sfbt.math.cnrs.fr/
L’Université Grenoble Alpes https://www.univ-grenoble-alpes.fr/
Institut Rhônalpin des Systèmes Complexes https://www.ixxi.fr/
Le Réseau Thématique Math-Bio-Santé https://math-bio-sante.math.cnrs.fr//
Le Labex Tec21 https://www.tecxxi.fr/
La SFR Olimpics https://olimpics.univ-grenoble-alpes.fr/
© 2026 SFBT, Remerciements au réseau Mathrice pour l’hébergement du site web.