2025Activity reportProject-Team‌​‌MUSCA

Network behavior​

In the case of​‌ autonomous systems (with no​​ explicit dependency on time),​​​‌ the main theoretical challenge​ is the prediction of​‌ the long time dynamics,​​ given the algebraic complexity​​​‌ associated with putative stationary​ states in high dimension.​‌ In physiological systems, the​​ intracellular reaction networks are​​​‌ not under a static​ or constant input stimulation​‌ but rather subject to​​ complex and highly dynamic​​​‌ signals such as (neuro-)hormones​ 27 or metabolites. These​‌ systems are thus non-autonomous​​ in nature. Understanding to​​​‌ what extent reaction network​ motifs are able to​‌ encode or decode the​​ dynamic properties of a​​​‌ time-dependent signal is a​ particularly challenging theoretical question,​‌ which has yet been​​ scarcely addressed, either in​​​‌ simplified case-studies 94,​14 or in the​‌ framework of “pulse-modulated systems”​​ 75.

Network reduction​​

The high dimension of​​​‌ realistic networks calls for‌ methods enabling to perform‌​‌ model reduction. Our strategy​​ for model reduction combines​​​‌ several tools, that can‌ be applied separately or‌​‌ sequentially to the initial​​ model. Both in stochastic​​​‌ biochemical systems and population‌ dynamics, large species abundance‌​‌ calls in general for​​ the functional law of​​​‌ large number and central‌ limit theorems, for which‌​‌ powerful results are now​​ established in standard settings​​​‌ of finite dimension models‌ 83. However, in‌​‌ more and more biological​​ applications, the very large​​​‌ spectrum of orders of‌ magnitude in reaction rates‌​‌ (or birth and death​​ rates) leads naturally to​​​‌ consider simultaneously large species‌ abundance with timescale separation,‌​‌ which generally results in​​ either algebraic-differential reduced models,​​​‌ or to hybrid reduced‌ models with both deterministic‌​‌ and stochastic dynamics. We​​ will apply the generic​​​‌ methodology provided by the‌ singular perturbation theory of‌​‌ Fenichel-Tikhonov in deterministic systems,​​ and Kurtz's averaging results​​​‌ in stochastic systems, which,‌ in the context of‌​‌ high dimensional reaction networks​​ or population dynamics, are​​​‌ still the matter of‌ active research both in‌​‌ the deterministic 84,​​ 76 and stochastic context​​​‌ 65, 82,‌ 93.

Other reduction‌​‌ approaches of deterministic systems​​ will consist in combining​​​‌ regular perturbation expansion with‌ standard linear model order‌​‌ reduction (MOR) techniques. We​​ will continue our previous​​​‌ work 18, 17‌ on the derivation of‌​‌ convergence and truncation error​​ bounds for the regular​​​‌ perturbation series expansion (also‌ known as Volterra series‌​‌ expansion) of trajectories of​​ a wide class of​​​‌ weakly nonlinear systems, in‌ the neighborhood of stable‌​‌ hyperbolic equilibria. The challenge​​ will be to obtain​​​‌ biologically interpretable reduced models‌ with appropriate features such‌​‌ as for instance positivity​​ and stability. Finding a​​​‌ general approach for the‌ reduction of strongly nonlinear‌​‌ systems is still an​​ open question, yet it​​​‌ is sometimes possible to‌ propose ad-hoc reduced models‌​‌ in specific cases, using​​ graph-based decomposition of the​​​‌ model 97, combined‌ with the reduction of‌​‌ weakly nonlinear subsystems.

Bridging​​ the gap between discrete​​​‌ and continuous networks through‌ most permissive semantics

Since‌​‌ their introduction in the​​ late 1960s, state- and​​​‌ time-discrete frameworks such as‌ Thomas Networks and Boolean‌​‌ Networks (BNs), which belong​​ to the subclass of​​​‌ logical models with only‌ two values possible in‌​‌ any variable, have been​​ widely adopted for reasoning​​​‌ about signaling and gene‌ networks, as they require‌​‌ few parameters and can​​ easily integrate information from​​​‌ omics datasets and genetic‌ screens. Two-way translations from‌​‌ BNs to discrete Petri​​ nets (PNs) allow one​​​‌ to transfer both theoretical‌ results and efficient algorithms‌​‌ from one model class​​ to the other (see​​​‌ 6) ; we‌ therefore regard BNs and‌​‌ discrete PNs as one​​ class, which we call​​​‌ discrete models. These‌ models represent processes with‌​‌ a high degree of​​ generalization and can offer​​​‌ coarse-grained but robust predictions.‌ That makes them particularly‌​‌ suitable for large biological​​ networks, for which ample​​​‌ global knowledge exists about‌ potential interactions with little‌​‌ precise data on actual​​​‌ molecules abundances and reaction​ kinetics. In this way,​‌ discrete models provide one​​ with an approach orthogonal​​​‌ to that enabled by​ continuous deterministic dynamics (via​‌ differential equations). In fact,​​ whereas deterministic models reflect​​​‌ in a very fine​ way the spatio-temporal dynamics​‌ in chemical reactions and​​ biological networks, the reliability​​​‌ of their predictions strongly​ depends on the precision​‌ of the knowledge concerning​​ existence and strength of​​​‌ interactions, and of precise​ measurements of all parameters,​‌ initial conditions and external​​ influences; perturbations in either​​​‌ domain are hard to​ apprehend, and their impact​‌ very difficult to predict.​​ Typically, such models are​​​‌ useful for conducting campaigns​ of large numbers of​‌ simulations with varying parameters,​​ giving a panorama of​​​‌ some possible types of​ system evolution. Obviously, there​‌ remains an irreducible coverage​​ problem: do the​​​‌ simulations we performed capture​ all the evolutions that​‌ the wild system could​​ have undergone?

By contrast,​​​‌ discrete models are fundamentally​ non-deterministic in their behavior,​‌ allowing one to extract​​ at moderate computational cost​​​‌ a complete, possibilistic overview​ of any evolution the​‌ system may take (see​​ 15, 5,​​​‌ 24). Some of​ these evolutions may be​‌ actually ruled out by​​ intrinsic quantitative features; discrete​​​‌ models thus may predict​ as possible some behaviors​‌ that will not actually​​ occur. It has long​​​‌ been believed that discrete​ models systematically over-approximate the​‌ system evolution. Our previous​​ work on the most​​​‌ permissive semantics7,​ 22 has shown that​‌ this is the case​​ if, but also only​​​‌ if, the semantics of​ the discrete system is​‌ sufficiently enriched (by one​​ additional, intermediate state PLUS​​​‌ the liberty, for every​ function, to read this​‌ ambiguous value as either​​ 0 or 1). Integrating​​​‌ into a new execution​ paradigm, called Most Permissive​‌ Boolean Networks (MPBNs), one​​ can therefore ensure that​​​‌ a carefully crafted discrete​ model predicts at least​‌ the actual possible behaviors​​ of the system. Moreover,​​​‌ MPBNs significantly reduce the​ complexity of dynamical analysis​‌ (e.g. reachability verification can​​ be done in linear​​​‌ time), enabling one to​ model genome-scale networks. A​‌ recent line of research​​ in MUSCA studies the​​​‌ dynamics of continuous Petri​ nets (CPNs). These​‌ models differ from their​​ discrete counterparts in the​​​‌ facts that transitions may​ be enabled with real​‌ rather than integer enabling​​ degrees, and may fire​​​‌ with any fraction of​ their enabling degree, leading​‌ to a continuous state​​ space. A crucial tool​​​‌ is to use abstract​ and symbolic semantics, in​‌ which state classes lump​​ together all continuous states​​​‌ that have the same​ activity, that is,​‌ the same set of​​ transitions that are eventually​​​‌ firable in some sequence​ initiated in those states.​‌ While being of great​​ interest in its own​​​‌ right, this allows one​ to capture the long-run​‌ behavior efficiently. The particularly​​ helpful mathematical properties of​​​‌ CPN allow for very​ efficient verification of reachability​‌ and limit-reachability; all attractors​​ are leaves, and vice​​​‌ versa. Emerging research directions​ are followed in MUSCA​‌ to (i) efficiently explore​​ the dynamics of biological​​ networks ; (ii) further​​​‌ study the relationships between‌ CPN, MPBN and related‌​‌ models; (iii) investigate the​​ passageways between discrete and​​​‌ continuous stochastic system models‌ ; and develop attractor‌​‌ search and reprogramming algorithms.​​

Statistical Inference, Data-fitting

Once​​​‌ again, a key challenge‌ in parameter estimation is‌​‌ due to the high​​ dimension of the state​​​‌ space and/or parameter space.‌ We will develop several‌​‌ strategies to face this​​ challenge. Efficient Maximum likelihood​​​‌ or pseudo-likelihood methods will‌ be developed and put‌​‌ in practice 1611​​, using either existing​​​‌ state-of-the art deterministic derivative-based‌ optimization 98 or global‌​‌ stochastic optimization 73.​​ In any case, we​​​‌ pay particular attention to‌ model predictivity (quantification of‌​‌ the model ability to​​ reproduce experimental data that​​​‌ were not used for‌ the model calibration) and‌​‌ parameter identifiability (statistical assessment​​ of the uncertainty on​​​‌ parameter values). A particularly‌ challenging and stimulating research‌​‌ direction of interest concerning​​ both model reduction and​​​‌ statistical inference is given‌ by identifiability and inference-based‌​‌ model reduction 86.​​ Another strategy for parameter​​​‌ inference in complex, nonlinear‌ models with fully observed‌​‌ state, but scarce and​​ noisy observations, is to​​​‌ couple curve clustering, which‌ allows reducing the system‌​‌ state dimension, with robust​​ network structure and parameter​​​‌ estimation. We are currently‌ investigating this option, by‌​‌ combining curve clustering 80​​ based on similarity criteria​​​‌ adapted to the problem‌ under consideration, and an‌​‌ original inference method inspired​​ by the Generalized Smoothing​​​‌ (GS) method proposed in‌ 96, which we‌​‌ call Modified Generalized Smoothing​​ (MGS). MGS is performed​​​‌ using a penalized criterion,‌ where the log-likelihood of‌​‌ the measurement error (noisy​​ data) is penalized by​​​‌ a model error for‌ which no statistical model‌​‌ is given. Moreover, the​​ system state is projected​​​‌ onto a functional basis‌ (we mainly use spline‌​‌ basis), and the inference​​ simultaneously estimates the model​​​‌ parameters and the spline‌ coefficients.

Model​ behavior

The theoretical challenges​‌ associated with the analysis​​ of structured population models​​​‌ are numerous, due to​ the lack of a​‌ unified methodological framework. The​​ analysis of the well-posedness​​​‌ 25 and long-time behavior​ 10, and the​‌ design of appropriate numerical​​ schemes 1, 3​​​‌ often rely on more​ or less generic techniques​‌ 92, 88 that​​ we need to adapt​​​‌ in a case-by-case, model-dependent​ way: general relative entropy​‌ 89, 70,​​ measure solution framework 81​​​‌, 66, 74​, martingale techniques 67​‌, finite-volume numerical schemes​​ 85, just to​​​‌ name a few.

Due​ to their strong biological​‌ anchorage, the formulation of​​ our models often leads​​​‌ to new mathematical objects,​ which raises open mathematical​‌ questions. Specific difficulties generally​​ arise, for instance from​​​‌ the introduction of nonlocal​ terms at an “unusual​‌ place” (namely in the​​ velocities rather than boundary​​​‌ conditions 25), or​ the formulation of particularly​‌ tricky boundary conditions 12​​. When needed, we​​​‌ call to external collaborators​ to try to overcome​‌ these difficulties.

Model reduction​​

Even if the use​​​‌ of a structured population​ formalism leads to models​‌ that can be considered​​ as compact, compared to​​​‌ the high-dimensional ODE systems​ introduced in section 3.2​‌, it can be​​ useful to derive reduced​​​‌ versions of the models,​ for sake of computational​‌ costs, and also and​​ above all, for parameter​​​‌ calibration purposes.

To proceed​ to such a reduction,​‌ we intend to combine​​ several techniques, including moment​​ equations 91, dimensional​​​‌ reduction 9, timescale‌ reduction 4, spatial‌​‌ homogenization 6313,​​ discrete to continuous reduction​​​‌ 12 and stochastic to‌ deterministic limit theorems 19‌​‌.

Once again, all​​ these techniques need to​​​‌ be applied on a‌ case-by-case basis, and they‌​‌ should be handled carefully​​ to obtain rigorous results​​​‌ (appropriate choice of metric‌ topology, a priori estimates).‌​‌

Statistical inference, Data-fitting

The​​ calibration of structured population​​​‌ models is challenging, due‌ to both the infinite-dimensional‌​‌ setting and the difficulty​​ to obtain rich enough​​​‌ data in our application‌ domains. Our strategy is‌​‌ rather empirical. We proceed​​ to a sequence of​​​‌ preliminary studies before using‌ the experimental available data.‌​‌ Sensitivity analyses 79,​​ 69, and theoretical​​​‌ studies of the inverse‌ problems associated with the‌​‌ models 8 intend to​​ preclude unidentifiable situations and​​​‌ ill-posed optimization problems. The‌ generation and use of‌​‌ synthetic data (possibly noised​​ simulation outputs) allow us​​​‌ to test the efficiency‌ of optimization algorithms and‌​‌ to delimit an initial​​ guess for the parameters.​​​‌ When reduced or simplified‌ versions of the models‌​‌ are available (or derived​​ specifically for calibration purposes)​​​‌ 2, these steps‌ are implemented on the‌​‌ increasingly complex versions of​​ the model. In situations​​​‌ where PDEs are or‌ can be interpreted as‌​‌ limits of stochastic processes,​​ it is sometimes possible​​​‌ to estimate parameters on‌ the stochastic process trajectories,‌​‌ or to switch from​​ one formalism to the​​​‌ other.

6.1.1​‌ pyDynPeak

• Keywords:
  Data processing,​​ Endocrinology
• Scientific Description:
  Analysis​​​‌ of time series taking​ into account the inherent​‌ properties of secretion events​​ (form and pulse half-life,​​​‌ regularity of changes in​ rhythm)
• Functional Description:
  Detection​‌ of LH pulses (luteinizing​​ hormone) and analysis of​​​‌ their rhythm. Visualisation, diagnostic​ and interactive correction of​‌ the detections.
• URL:
  https://­gitlab.­inria.­fr/­musca/­pydynpeak​​
• Contact:
  Frédérique Clément

7.1.1 Bifurcation analysis​​ for gene regulatory networks​​​‌ embedding a toggle-switch model:​ Application to X chromosome​‌ inactivation

Participants: Frédérique Clément​​, Alice Fohr,​​​‌ Hélène Leman.

We​ have analyzed models of​‌ minimal gene regulatory networks​​ (GRN) introduced in the​​​‌ context of X chromosome​ inactivation (XCI), an epigenetic​‌ process occurring in placental​​ female mammals 68.​​​‌ A core module in​ these GRNs is a​‌ 2D toggle-switch (TS) model​​ resulting from a mutual​​​‌ inhibition. We have first​ performed a bifurcation analysis​‌ of the TS model,​​ where the main difficulty​​​‌ arises from the unknown​ coordinates of the fixed​‌ points. In the symmetric​​ case, we have proven​​​‌ the occurrence of a​ pitchfork bifurcation, and computed​‌ explicitly the bifurcation lines.​​ In the asymmetric case,​​​‌ we have designed a​ constructive numerical approach providing​‌ one simultaneously with the​​ bifurcation lines and fixed-point​​​‌ coordinates, and illustrated the​ occurrence of a saddle-node​‌ bifurcation using numerical continuation​​ tools. Bistability in the​​​‌ TS accounts for the​ possible choice between an​‌ activated ($X_{\mathrm{a}}$) and inactivated (​​​‌$X_i$) state​ in the dynamics of​‌ a single X chromosome.​​ We then proceeded to​​​‌ a thorough analysis of​ 4D and 6D GRN​‌ models representing the joint​​ dynamics of the X​​​‌ chromosome pair, to (i)​ ensure the stability of​‌ the mono-inactivated (${\mathrm{X}}_i{X}_a$)​​​‌ state, and, then, (ii)​ exclude the stability of​‌ the bi-inactivated (${\mathrm{X}}_i{X}_i$)​​​‌ and bi-activated (${\mathrm{X}}_a{X}_a$)​‌ states. Studying the 6D​​ GRN involves the analysis​​​‌ of TS models coupled​ through a state-dependent parameter.​‌ Combining proper changes of​​ variables and reparameterization, we​​ managed to study both​​​‌ the transient and asymptotic‌ behavior from a 2D‌​‌ phase plane analysis. We​​ further restricted the parameter​​​‌ space to meet quantitative‌ specifications on the relative‌​‌ gene expression level between​​ the $X_i$ and​​​‌ $X_a$ states.

7.1.2‌ Bifurcation analysis of a‌​‌ size-structured population model: Application​​ to oocyte dynamics and​​​‌ ovarian cycle

Participants: Frédérique‌ Clément, Louis Fostier‌​‌, Romain Yvinec.​​

In the framework of​​​‌ Louis Fostier's PhD thesis‌ 55, we have‌​‌ introduced and analyzed a​​ quasilinear size-structured population model​​​‌ with nonlinearities accounting for‌ nonlocal interactions between individuals‌​‌ 29. The recruitment​​ (immigration), growth and death​​​‌ rates are inhomogeneous in‌ time and/or space and‌​‌ depend on weighted averages​​ of the density. We​​​‌ have first proven the‌ existence and uniqueness of‌​‌ globally bounded weak solutions​​ using the characteristic curves​​​‌ and Banach fixed point‌ Theorem, after transforming the‌​‌ partial differential equation into​​ an equivalent system of​​​‌ integral equations. We then‌ investigated the long-time behavior‌​‌ of the PDE in​​ the case when the​​​‌ growth rate is separable.‌ Applying a classical time-scaling‌​‌ transformation, the problem boils​​ down to a PDE​​​‌ with linear growth rate‌ and nonlinear inflow boundary‌​‌ condition, entering the theoretical​​ framework of abstract semilinear​​​‌ Cauchy problems. We could‌ then perform a bifurcation‌​‌ analysis which revealed the​​ richness of the model​​​‌ behavior. Depending on the‌ ratio of the recruitment‌​‌ to the growth rate,​​ the model can exhibit​​​‌ multistability and stable oscillatory‌ solutions, emanating respectively through‌​‌ saddle-node and Hopf bifurcations.​​ We have illustrated these​​​‌ theoretical results on the‌ biological application motivating this‌​‌ work, oogenesis, the process​​ of production and maturation​​​‌ of female gametes (oocytes)‌ that is critical to‌​‌ reproductive fitness.

7.1.3 Rapid​​ cell turnover to model​​​‌ adipocyte size distribution

Participants:‌ Chloé Audebert, Louis‌​‌ Fostier, Romain Yvinec​​, and collaborators.​​​‌

White adipose tissue, composed‌ of adipocyte cells, primarily‌​‌ stores energy as lipid​​ droplets. The size of​​​‌ adipocytes varies significantly within‌ the tissue according to‌​‌ the amount of stored​​ lipids. A striking observation​​​‌ is that the adipocyte‌ size distribution is bimodal,‌​‌ and thus, this tissue​​ is lacking a characteristic​​​‌ size. We have proposed‌ a novel dynamical model,‌​‌ based on a partial​​ differential equation, to represent​​​‌ the adipocyte size distribution‌ 30. The model‌​‌ assumes continuous adipocyte growth,​​ with a velocity dependent​​​‌ on cell radius and‌ extracellular lipid availability, together‌​‌ with constant rates of​​ cell recruitment and death.​​​‌ We have proven the‌ existence and local stability‌​‌ of a unique stationary​​ solution for a broad​​​‌ range of growth velocity‌ functions. Choosing a parsimonious‌​‌ formulation, we have shown​​ that three parameters are​​​‌ enough to describe adipocyte‌ size distributions measurements in‌​‌ rats. These parameters are​​ robustly estimated through approximate​​​‌ Bayesian computation, and the‌ model demonstrates excellent agreement‌​‌ with experimental data. This​​ mechanistic, three-parameter framework offers​​​‌ a new and interpretable‌ approach to characterizing adipocyte‌​‌ size distributions.

7.2.1 A mechanistic​​​‌ modelling approach of the‌ host–microbiota interactions to investigate‌​‌ beneficial symbiotic resilience in​​​‌ the human gut

Participants:​ Béatrice Laroche, Eleonora​‌ Pastremoli, Lorenzo Sala​​, and collaborators.​​​‌

The health and well-being​ of a host are​‌ deeply influenced by the​​ interactions with its gut​​​‌ microbiota. Diet, especially the​ amount of fiber intake,​‌ plays a pivotal role​​ in modulating these interactions​​​‌ impacting microbiota composition and​ functionality. We have introduced​‌ a novel mathematical model​​ 77, designed to​​​‌ delve into these interactions,​ by integrating dynamics of​‌ the colonic epithelial crypt,​​ bacterial metabolic functions and​​​‌ sensitivity to inflammation as​ well as colon flows​‌ in a transverse colon​​ section. Unique features of​​​‌ our model include accounting​ for metabolic shifts in​‌ epithelial cells based on​​ butyrate and hydrogen sulfide​​​‌ concentrations, representing the effect​ of innate immune pattern​‌ recognition receptors activation in​​ epithelial cells, capturing bacterial​​​‌ oxygen tolerance based on​ data analysis, and considering​‌ the effect of antimicrobial​​ peptides on the microbiota.​​​‌ Using our model, we​ show a proof-of-concept that​‌ a high-protein, low-fiber diet​​ intensifies dysbiosis and compromises​​​‌ symbiotic resilience. Our simulation​ results highlight the critical​‌ role of adequate butyrate​​ concentrations in maintaining mature​​​‌ epithelial crypts. Through differential​ simulations focused on varying​‌ fiber and protein inputs,​​ our study offers insights​​​‌ into the system resilience​ following the onset of​‌ dysbiosis. The present model,​​ while having room for​​​‌ enhancement, offers essential understanding​ of elements such as​‌ oxygen levels, the breakdown​​ of fiber and protein,​​​‌ and the basic mechanisms​ of innate immunity within​‌ the colon environment.

In​​ the framework of Eleonora​​​‌ Pastremoli 's PhD, this​ model has been further​‌ enriched to incorporate a​​ newly identified hard mucus​​​‌ layer, with additional specific​ boundary conditions and reaction​‌ terms such as the​​ degradation rate. The completed​​​‌ model intends to capture​ accurately the micro-scale processes​‌ within the crypt, such​​ as cell-microbe interactions and​​​‌ nutrient absorption, and their​ impact on the macro-scale​‌ dynamics of the colon.​​ With this comprehensive framework,​​​‌ we can establish a​ virtual laboratory to test​‌ various hypotheses about gut​​ microbiota interactions, such as​​​‌ simulating different dietary regimes​ or studying the inflammatory​‌ response associated with microbial​​ groups. Current efforts are​​​‌ focused on enhancing the​ computational efficiency of the​‌ model simulations by developing​​ model order reduction techniques.​​​‌

7.2.2 Multiscale modeling of​ single cell-based dynamics of​‌ ovarian development

Participants: Chloé​​ Audebert, Frédérique Clément​​​‌, Fabien Crauste,​ Pawan Kumar.

In​‌ mammals, females are endowed​​ at birth with a​​​‌ limited number of germ​ cells (oocytes) hosted in​‌ somatic structures called ovarian​​ follicles. The pool of​​​‌ primordial (not yet activated)​ follicles corresponds to the​‌ ovarian reserve, which will​​ get progressively exhausted with​​​‌ ovarian aging. In the​ framework of the AI4scMED​‌ axis of PEPR Santé​​ numérique and Pawan Kumar's​​​‌ postdoc, we have developed​ a multiscale model representing​‌ the selection of the​​ future oocytes among the​​​‌ germ cell population, which​ occurs within germ cysts​‌ during ovarian development, and​​ involves complex interactions between​​​‌ germ cells and somatic​ cells. The model has​‌ been implemented in the​​ SiMuScale environment that enables​​ one to couple biochemical​​​‌ dynamics, namely gene regulatory‌ dynamics (GRN), with spatially-distributed‌​‌ cell population dynamics. The​​ 3D individual-based model accounts​​​‌ for the different types‌ of germ cells and‌​‌ their neighboring somatic cells​​ within a spatially explicit​​​‌ framework describing the structure‌ of a germ cyst,‌​‌ and embeds a minimal​​ GRN for oocyte differentiation​​​‌ derived from scRNA seq-based‌ studies of ovarian development.‌​‌ We have investigated how​​ the germ cyst geometry​​​‌ affects the oocyte selection‌ outcome and associated changes‌​‌ in germ cell volume​​ associated with cytoplasmic mass​​​‌ transfer from non-selected germ‌ cells. The model parameters‌​‌ have been calibrated across​​ scales, ranging from GRN​​​‌ dynamics to cytoplasmic exchanges,‌ to ensure biological realism,‌​‌ with robustness evaluated through​​ local sensitivity analyses. We​​​‌ have assessed the statistical‌ reliability of simulation outcomes‌​‌ by estimating the minimum​​ number of stochastic runs​​​‌ required for stable and‌ reproducible results. Altogether, the‌​‌ model generates 3D spatiotemporal​​ distributions of germ-cell fate,​​​‌ tracks the stepwise differentiation‌ of germ cells into‌​‌ oocytes, captures cytoplasmic transfer​​ events and associated volume​​​‌ changes, and monitors the‌ timing of unselected germ‌​‌ cell death across simulations.​​ The model results give​​​‌ insight into how only‌ a small subset of‌​‌ germ cells survive, increase​​ in size—and ultimately contributes​​​‌ to establishing the ovarian‌ reserve.

7.3.1 Computational prediction​​​‌ of intracellular signaling behavior‌ via machine learning

Participants:‌​‌ Frédéric Jean-Alphonse, Pamela​​ Romero, Romain Yvinec​​​‌, and collaborators.‌

Assessing the dynamics of‌​‌ intracellular signaling processes under​​ various conditions such as​​​‌ protein-protein interactions, protein conformational‌ changes, dose-dependent drug effects,‌​‌ and ligand binding is​​ crucial for biologists and​​​‌ pharmacologists. However, generating such‌ data can be time-consuming‌​‌ and costly. In 49​​, we used data​​​‌ obtained by Bioluminescence Resonance‌ Energy Transfer (BRET) in‌​‌ live cells to develop​​ models to predict the​​​‌ temporal dynamics of intracellular‌ second messenger cAMP (cyclic‌​‌ adenosine monophosphate). We formulated​​ the task of predicting​​​‌ time series in a‌ supervised manner and employed‌​‌ decision tree, random forest,​​ and XGBoost models within​​​‌ a multiple-input, multiple-output framework,‌ enabling simultaneous forecasting of‌​‌ multiple future time steps.​​ Our findings demonstrate that​​​‌ our model achieves, in‌ the main experiment, a‌​‌ mean absolute error of​​ less than 0.018 on​​​‌ the testing set. This‌ model is computationally efficient,‌​‌ requiring only $\approx$ 16​​ seconds for training, validation,​​​‌ and testing. Finally, we‌ identified that BRET signal‌​‌ intensities are the most​​ important feature for predictions​​​‌ among the set of‌ features in the models.‌​‌ These results highlight the​​ potential of machine learning​​​‌ models for advancing research‌ in dynamical biological processes,‌​‌ particularly in cellular signaling​​ and pharmacological systems.

7.3.2​​​‌ Biology-Informed inverse problems for‌ insect pests detection using‌​‌ pheromone sensors

Participants: Béatrice​​ Laroche, and collaborators​​​‌.

Most insects have‌ the ability to modify‌​‌ the odor landscape in​​ order to communicate with​​​‌ their conspecies during key‌ phases of their life‌​‌ cycle such as reproduction.​​ They release pheromones in​​​‌ their nearby environment, volatile‌ compounds that are detected‌​‌ by insects of the​​​‌ same species with exceptional​ specificity and sensitivity. Efficient​‌ pheromone detection is then​​ an interesting lever for​​​‌ insect pest management in​ a precision agroecological culture​‌ context. A precise and​​ early detection of pests​​​‌ using pheromone sensors offers​ a strategy for pest​‌ management before infestation. In​​ 40, we have​​​‌ developed a biology-informed inverse​ problem framework that leverages​‌ temporal signals from a​​ pheromone sensor network to​​​‌ build insect presence maps.​ Prior biological knowledge is​‌ introduced in the inverse​​ problem by the mean​​​‌ of a specific penalty,​ using population dynamics PDE​‌ residuals. We have benchmarked​​ the biological-informed penalty with​​​‌ other regularization terms such​ as Tikhonov, LASSO or​‌ composite penalties in a​​ simplified toy model. We​​​‌ used classical comparison criteria,​ such as target reconstruction​‌ error, or Jaccard distance​​ on pest presence-absence, and​​​‌ also more task-specific criteria​ such as the number​‌ of informative sensors during​​ inference. Finally, the inverse​​​‌ problem has been solved​ in a realistic context​‌ of pest infestation in​​ an agricultural landscape by​​​‌ the fall armyworm (Spodoptera​ frugiperda).

7.3.3 Surrogate modeling​‌ of interactions in microbial​​ communities through Physics-Informed Neural​​​‌ Networks

Participants: Béatrice Laroche​, Lorenzo Sala,​‌ and collaborators.

Microorganisms​​ form complex communities known​​​‌ as microbiota, influencing various​ aspects of host well-being.​‌ The Generalized Lotka-Volterra (GLV)​​ model is commonly used​​​‌ to understand microorganism population​ dynamics, but its application​‌ to the microbiota faces​​ challenges due to limited​​​‌ bacterial data and complex​ interactions. This preliminary work​‌ exposed in 33 focuses​​ on using a Physics-Informed​​​‌ Neural Network (PINN) and​ synthetic data to build​‌ a surrogate model of​​ bacterial species evolution driven​​​‌ by a GLV model.​ The approach is calibrated​‌ and tested on several​​ models differing in size​​​‌ and dynamic behavior.

7.3.4​ Automatic detection and classification​‌ of ovarian follicles from​​ 2D histological images

Participants:​​​‌ Frédérique Clément, Rosario​ Medina-Rodriguez.

In the​‌ framework of Rosario Medina's​​ postdoctoral research within the​​​‌ OVOPAUSE project, we have​ developed a deep-learning model​‌ for the automatic detection​​ and classification of ovarian​​​‌ follicles from 2D histological​ images of mouse ovaries​‌ provided by domain expert​​ collaborators. The learning core​​​‌ is based on an​ object detection paradigm using​‌ the Detectron2 framework, where​​ a Mask R-CNN model​​​‌ with a Feature Pyramid​ Network (FPN) backbone, initialized​‌ with ImageNet-pretrained weights, is​​ used as the baseline​​​‌ architecture. The analysis of​ the images and annotations​‌ raises two major challenges:​​ (i) a significant variation​​​‌ in follicle sizes depending​ on the maturity stages,​‌ and (ii) a limited​​ number of annotated instances​​​‌ for each category, particularly​ for the smallest follicles.​‌ To tackle the latter​​ problem, we have performed​​​‌ dataset curation and annotation​ augmentation in order to​‌ mitigate class imbalance and​​ stabilize the training configurations​​​‌ tailored to the available​ datasets. The first step​‌ consisted in refining the​​ annotation protocol to capture​​​‌ biologically meaningful features on​ the whole follicle level​‌ (e.g. including the outward​​ theca corona) and help​​​‌ discriminating the follicles from​ other possibly morphologically similar​‌ structures (e.g. small vessels).​​ We have applied a​​ standardized preprocessing pipeline to​​​‌ the original images, including‌ resolution scaling, cropping to‌​‌ the minimal bounding box,​​ contrast enhancement, noise reduction,​​​‌ and background normalization, before‌ subsequently tiling them into‌​‌ overlapping square pixel patches.​​ Given the limited availability​​​‌ of labeled data and‌ the significant class imbalance,‌​‌ we have deployed several​​ strategies to improve the​​​‌ model robustness, such as‌ oversampling, geometric and photometric‌​‌ augmentations, stain normalization, and​​ controlled perturbations of stain​​​‌ color and intensity to‌ simulate realistic histological variability.‌​‌ These strategies were combined​​ with a pseudo-labeling approach​​​‌ to increase the number‌ of annotated datasets. Altogether,‌​‌ these steps have improved​​ the detection score of​​​‌ small follicles, which is‌ critical in the global‌​‌ performance. In parallel, a​​ Docker-based package integrating OMERO​​​‌ server with automated follicle‌ detection scripts was developed,‌​‌ enabling scalable inference and​​ integration into existing histopathology​​​‌ workflows.

7.3.5 3D imaging-based‌ analysis of the germline‌​‌ in teleost

Participants: Frédérique​​ Clément, Marlène Davilma​​​‌, and collaborators.‌

In teleost fish, female‌​‌ fecundity depends essentially on​​ the oocyte reserve, which​​​‌ determines the number of‌ eggs spawned in each‌​‌ reproductive cycle. Unlike mammals,​​ which have a limited​​​‌ and predefined stock of‌ oocytes at birth, this‌​‌ reserve can be renewed​​ throughout a female's life.​​​‌ In adult teleosts, this‌ reserve is, on the‌​‌ one hand, used to​​ generate mature oocytes ready​​​‌ to be spawned and,‌ on the other hand,‌​‌ replenished from germline stem​​ cells present in specialized​​​‌ structures called germline cradles.‌ A main issue is‌​‌ to understand the contribution​​ of these germline stem​​​‌ cells in the renewal‌ of the oocyte reserve‌​‌ in both juveniles and​​ adults, as well as​​​‌ the involved regulatory mechanisms.‌ In the framework of‌​‌ Marlène Davilma's PhD, co-supervised​​ by Violette Thermes and​​​‌ Frédérique Clément, we have‌ implemented a 3D whole‌​‌ ovary imaging strategy in​​ Medaka to provide quantitative​​​‌ data and study the‌ cell dynamics in the‌​‌ germinal cradle. We have​​ refined ovary clearing protocols​​​‌ combined with immunolabelings (e.g.,‌ anti-Vasa, anti-PH3, anti-GFP) and‌​‌ nuclear staining (MG), and​​ imaged the ovaries using​​​‌ light sheet and confocal‌ microscopy. In addition, we‌​‌ have set up 3D​​ image analysis pipelines that​​​‌ integrate pre-trained open-source neural‌ networks suitable for precise‌​‌ segmentation. These deep-learning–based pipelines​​ have greatly improved our​​​‌ ability to handle complex‌ 3D datasets on the‌​‌ whole-ovary scale and to​​ extract robust quantitative data​​​‌ for well-defined structures such‌ as ovarian follicles, while‌​‌ providing a framework for​​ future quantitative analyses of​​​‌ more complex structures such‌ as germinal cradles. We‌​‌ have characterized the spatial​​ organization, cell composition and​​​‌ developmental dynamics of germinal‌ cradles in wild-type females‌​‌ from hatching to adulthood,​​ revealing a highly dynamic​​​‌ and heterogeneous morphological organization,‌ with poorly individualized structures‌​‌ during growth stages that​​ progressively reorganize during ovarian​​​‌ maturation.

7.4.1 Petri nets faithfully​​ fluidify most permissive boolean​​​‌ networks

Participants: Stefan Haar‌, and collaborators.‌​‌

The analysis of biological​​ networks has benefited from​​​‌ the richness of Boolean‌ networks (BNs) and the‌​‌ associated theory. These results​​​‌ have been further fortified​ in the recent years​‌ by the emergence of​​ Most Permissive (MP) semantics,​​​‌ combining efficient analysis methods​ with a greater capacity​‌ of explaining pathways to​​ states hitherto thought unreachable,​​​‌ owing to limitations of​ the classical update modes.​‌ While MPBNs are understood​​ to capture any behavior​​​‌ that can be observed​ at a lower level​‌ of abstraction, all the​​ way down to continuous​​​‌ refinements, the specifics and​ potential of the models​‌ and analysis, especially attractors,​​ across the abstraction scale​​​‌ remain unexplored. In 48​, we have fluidified​‌ MPBNs by means of​​ Continuous Petri nets (CPNs),​​​‌ a model of (uncountably​ infinite) dynamic systems that​‌ has been successfully explored​​ for modeling and theoretical​​​‌ purposes. CPNs create a​ formal link between MPBNs​‌ and their continuous dynamical​​ refinements such as ODE​​​‌ models. The benefits of​ CPNs extend beyond the​‌ model refinement, and constitute​​ well established theory and​​​‌ analysis methods, recently augmented​ by abstract and symbolic​‌ reachability graphs. These structures​​ are shown to compact​​​‌ the possible behaviors of​ the system with focus​‌ on events which drive​​ the choice of long-term​​​‌ behavior in which the​ system eventually stabilizes. These​‌ results bring an important​​ keystone to this novel​​​‌ methodology for biological networks,​ namely the proof that​‌ extant PN encoding of​​ BNs instantiated as a​​​‌ CPN simulates the MP​ semantics. In spite of​‌ the underlying dynamics being​​ continuous, the analysis remains​​​‌ in the realm of​ discrete methods, constituting an​‌ extension of all previous​​ work.

7.4.2 Searching for​​​‌ attractors: To infinity and​ beyond

Participants: Stefan Haar​‌.

Discrete and nondeterministic​​ modeling of regulatory and​​​‌ signaling networks allows to​ characterize many of their​‌ crucial dynamical properties, with​​ a reasonable computational effort.​​​‌ A dynamical feature of​ particular interest in its​‌ own right, as well​​ as in terms of​​​‌ biological relevance, is the​ landscape of attractors and​‌ their attraction basins. In​​ the recent years, we​​​‌ have developed new approaches​ to discovery and global​‌ cartography of this basin​​ landscape. We have further​​​‌ enlarged the set of​ models used, by moving​‌ to Petri nets on​​ the one hand and​​​‌ an opening to continuous​ dynamics on the other​‌ 47. Surprisingly, these​​ openings are rewarded not​​​‌ only by a sharpening​ of the analysis, but​‌ also the emergence of​​ compact and readable data​​​‌ structures, and fast search​ algorithms. Altogether, this work​‌ highlights the cross-fertilization between​​ discrete and continuous approaches.​​​‌

7.4.3 Modeling compartmentalization in​ cell signaling networks

Participants:​‌ Léo Darrigade, Stefan​​ Haar, Frédéric Jean-Alphonse​​​‌, Romain Yvinec,​ Chloé Weckel.

In​‌ the course of the​​ COMPARTIMENTAGE exploratory action, we​​​‌ have initiated a new​ thematics on the compartmentalization​‌ of cell signaling, with​​ a special focus on​​​‌ the compartimentalization of G​ Protein-Coupled Receptors

In the​‌ framework of Leo Darrigade's​​ post-doc, we have designed​​​‌ a piecewise deterministic Markov​ process of intracellular GPCR​‌ trafficking and cAMP production.​​ The stochastic part of​​​‌ the model accounts for​ the formation, coagulation, fragmentation​‌ and recycling of intracellular​​ vesicles carrying the receptors,​​ while the deterministic part​​​‌ of the model represents‌ the chemical reactions mediating‌​‌ the response to the​​ activated receptor. The stochastic​​​‌ part of the model‌ accounts for the formation,‌​‌ coagulation, fragmentation and recycling​​ of intracellular vesicles carrying​​​‌ the receptors, while the‌ deterministic part of the‌​‌ model represents the chemical​​ reactions mediating the response​​​‌ to the activated receptor.‌ We have studied the‌​‌ long time behavior analysis​​ of the model and​​​‌ shown the existence of‌ a stationary measure under‌​‌ mild conditions. Studying the​​ existence of an accessible​​​‌ Doeblin point for our‌ process, we have obtained‌​‌ a general condition to​​ prove convergence in Total​​​‌ Variation towards a unique‌ stationary measure. This condition‌​‌ relies on the properties​​ of the flow associated​​​‌ with the membrane signaling‌ and the intracellular-vesicle signaling,‌​‌ and can be verified​​ on concrete examples, case​​​‌ by case. Under stronger‌ conditions (e.g. the flow‌​‌ is exponentially contractive), we​​ have further shown the​​​‌ exponential ergodicity in a‌ weak topology.

In the‌​‌ framework of Chloé Weckel's​​ PhD thesis, we have​​​‌ designed a complementary mathematical‌ framework applicable to generic‌​‌ GPCRs. We have designed​​ a compartmental model based​​​‌ on systems of ordinary‌ differential equations (Chemical Reaction‌​‌ Networks), and we have​​ studied the effects of​​​‌ internalization and recycling on‌ the receptor-induced cell responses.‌​‌ This approach helped us​​ to address two reciprocal​​​‌ questions: (i) How does‌ trafficking influence the cell‌​‌ responses? and (ii) Does​​ the cell response affect​​​‌ trafficking? We have demonstrated‌ that receptor trafficking can‌​‌ either enhance or reduce​​ te pharmacological effect of​​​‌ ligands, depending on plasma‌ membrane versus endosomal signaling‌​‌ properties, and revealed that,​​ in turn, the signal​​​‌ feedback on receptor trafficking‌ can induce complex phenomena‌​‌ like multi-stability.

7.5.1​​​‌ Trafficking of luteinizing hormone‌ receptor directs the differential‌​‌ signal activation between luteinizing​​ hormone and chorionic gonadotropin​​​‌

Participants: Frédéric Jean-Alphonse,‌ Eric Reiter, and‌​‌ collaborators.

Luteinizing hormone​​ (LH) and human choriogonadotropin​​​‌ (hCG) support distinct reproductive‌ events via differential activation‌​‌ of the luteinizing hormone​​ receptor (LHCGR). LH-mediated LHCGR​​​‌ trafficking is known to‌ be key in activating‌​‌ and regulating its signal​​ responses, yet whether LH​​​‌ and hCG differentially direct‌ LHCGR trafficking is unknown.‌​‌ In 38, using​​ bioluminescence resonance energy transfer​​​‌ (BRET) trafficking biosensors and‌ highresolution TIRF imaging, we‌​‌ have demonstrated that LH​​ induces rapid internalization and​​​‌ recycling via an APPL1linked‌ very early endosomal pathway,‌​‌ while hCG-mediated receptor trafficking​​ and recycling is slower,​​​‌ and preferentially involves beta-arrestins‌ and accumulation in endocytic‌​‌ compartments positive for early​​ and late endosomal markers.​​​‌ Receptor internalization was differentially‌ required for Gq, Gi‌​‌ and Gs protein-mediated signals,​​ revealing distinct LH- vs​​​‌ hCG-trafficking signatures that may‌ be fundamental to preserving‌​‌ unique hormone signaling patterns​​ and their impact at​​​‌ the genomic level. These‌ results support different LH‌​‌ vs hCG modes of​​ action on the LHCGR​​​‌ through differential post-endocytic sorting‌ of the receptor, providing‌​‌ a potential “location bias”​​ mechanism underlying the distinct​​​‌ physiological roles of these‌ two gonadotropins.

7.5.2 LHR‌​‌ and G$\alpha$s​​​‌ trafficking drive sustained cAMP​ signalling from endosomes to​‌ control steroidogenesis

Participants: Frédéric​​ Jean-Alphonse, Eric Reiter​​​‌, and collaborators.​

A growing number of​‌ G protein-coupled receptors signal​​ through G proteins from​​​‌ intracellular compartments following endocytosis.​ While for few receptors​‌ the physiological relevance of​​ such mechanism has been​​​‌ established, the relationships between​ the spatio-temporal organization of​‌ cellular signaling and the​​ physiological responses are still​​​‌ to be elucidated for​ most receptors. Signaling by​‌ the luteinizing hormone receptor​​ (LHR) is essential to​​​‌ regulate sex steroids production​ in gonads, but how​‌ G protein-dependent signals from​​ endosomes are functionally important​​​‌ for steroidogenesis remains unexplored.​ In 58, we​‌ have demonstrated that transient​​ LHR activation promotes a​​​‌ prolonged G$\alpha$s/cAMP​ signaling from endosomes, which​‌ requires ligand-induced independent receptor​​ and G$\alpha$s​​​‌ trafficking. We have shown​ that endosomal trafficking is​‌ specifically required for ligand-induced​​ cAMP accumulation in the​​​‌ nucleus and gene expression,​ as well as for​‌ steroid production in Leydig​​ cells. These results contribute​​​‌ to further understand the​ molecular mechanisms by which​‌ LHR, through signaling compartmentalization,​​ controls gonadal steroidogenesis.

7.5.3​​​‌ A single domain intrabody​ as a novel tool​‌ to bias the subcellular​​ trafficking of the follicle-stimulating​​​‌ hormone receptor

Participants: Pascale​ Crépieux, Frédéric Jean-Alphonse​‌, Eric Reiter,​​ Romain Yvinec, Chloé​​​‌ Weckel, and collaborators​.

Variable fragments from​‌ heavy-chain only antibodies of​​ camelids (VHH) stabilize active​​​‌ G protein-coupled receptor conformations,​ enabling their 3D structural​‌ determination and facilitating detailed​​ structure–activity studies when expressed​​​‌ intracellularly, by linking specific​ receptor states to downstream​‌ signaling pathways. Recently, intra-VHHs​​ have been instrumental in​​​‌ tracking active GPCRs in​ various subcellular compartments. In​‌ 61, we have​​ reported the isolation and​​​‌ characterization of iPRC2, an​ intra-VHH selectively recognizing the​‌ intracellular loops of the​​ human follicle-stimulating hormone receptor​​​‌ (hFSHR). iPRC2 expression inside​ hFSHR-expressing cells decreases cAMP​‌ accumulation in response to​​ FSH binding, but requires​​​‌ G$\alpha$s for​ optimal interaction with the​‌ receptor. Importantly, iPRC2 reroutes​​ internalized hFSHR from very​​​‌ early endosomes, leading to​ an increased accumulation of​‌ active receptor in early​​ endosomes, and consequently, diminishes​​​‌ its recycling towards the​ cell surface. A compartmentalized​‌ modeling approach further supports​​ that the iPRC2-induced rerouting​​​‌ of hFSHR is sufficient​ to explain the decreased​‌ cAMP accumulation in response​​ to FSH binding. Hence,​​​‌ in contrast to previously​ described intra-VHHs that recognize​‌ active G protein-coupled receptor​​ intracellular location, iPRC2 provokes​​​‌ per se a location​ bias, through its ability​‌ to reroute hFSHR and​​ appears to be an​​​‌ innovative tool to examine​ the functional consequences of​‌ G protein-coupled receptor accumulation​​ in selective subcellular compartments.​​​‌

7.5.4 Endogenous ligands of​ bovine FFAR2/GPR43 display distinct​‌ pharmacological properties

Participants: Frédéric​​ Jean-Alphonse, Eric Reiter​​​‌, and collaborators.​

Free fatty acids (FFAs)​‌ have been identified as​​ ligands for members of​​​‌ the G protein-coupled receptor​ (GPCR) family, called free​‌ fatty acid receptors (FFARs).​​ Among these receptors, there​​​‌ is a particular interest​ in the physiological roles​‌ of FFAR2 and its​​ potential use as a​​ therapeutic target for various​​​‌ health disorders. Despite great‌ progress in other species,‌​‌ pharmacological properties of the​​ bovine FFAR2 (bFFAR2) are​​​‌ not fully understood. In‌ 41, we have‌​‌ evaluated how a selection​​ of FFAs (C2:0 to​​​‌ C8:0, and branched FFAs)‌ activate and regulate bFFAR2‌​‌ signaling. We used HEK293A​​ cells and BRET assays​​​‌ to measure G$\mathrm{\alpha ‌}$i/G$\alpha$q coupling‌​‌ and signaling, $\beta$-arrestin​​ 2 recruitment, and receptor​​​‌ internalization/trafficking. SRE and NFAT-RE‌ dependent transcription was assessed‌​‌ by luciferase reporter assay.​​ bFFAR2 presents a dual​​​‌ coupling to G$\mathrm{\alpha ‌}$i and G$\mathrm{\alpha ‌‌}$q, and recruits $\mathrm{\beta }$-arrestin 2 when stimulated​​​‌ with short and medium-chain‌ FFAs up to eight‌​‌ carbons. Straight-chain FFAs with​​ 4 to 7 carbons​​​‌ plus 3-methyl-butanoic acid showed‌ the greatest potency to‌​‌ activate bFFAR2 upstream and​​ downstream signaling, while C2:0,​​​‌ C3:0 and 2-methylpropanoic acid‌ (2MP) were the least‌​‌ potent. 2MP exhibited minimal​​ pharmacological activity towards $\mathrm{\beta ‌}$-arrestin 2, and although‌ it induced receptor internalization,‌​‌ bFFAR2 trafficking to the​​ early endosome was not​​​‌ observed. Overall, the number‌ of carbons of straight-chain‌​‌ FFAs and methyl position​​ of branched FFAs differentially​​​‌ regulates the activation of‌ bFFAR2.

8.1.1 Participation in International​​​‌ Programs

• Bill & Melinda‌ Gates Foundation, ContraBody (2021-February‌​‌ 2025, PI Eric Reiter​​ , 1.8 million U.S.​​​‌ dollars) “Non-hormonal contraception by‌ nanobody produced from within‌​‌ the body”. In partnership​​ with University of Modena​​​‌ E Regio Emilia, Italy,‌ MabSilico, France and InCellArt,‌​‌ France. Involved MUSCA members​​ : Eric Reiter ,​​​‌ Pascale Crépieux , Frédéric‌ Jean-Alphonse , Romain Yvinec‌​‌
• Medical Research Council, MICA​​ (2022-February 2025, PI Waljit​​​‌ Dhillo, 642,000 euros) “Investigating‌ kisspeptin receptor signalling to‌​‌ improve the treatment of​​ reproductive disease”. Involved MUSCA​​​‌ member: Eric Reiter

8.1.2‌ Visits to international teams‌​‌

8.2.1 Horizon Europe​​​‌

• EU Pathfinder, ABCardionostics (2024-2028,‌ PI Gisèle Clofent-Sanchez, 3.6‌​‌ million euros). Involved MUSCA​​ member: Anne Poupon ,​​​‌ WP4 coordination

9.1.1 Scientific‌​‌ events: organisation

9.1.2 Journal‌

9.1.3 Invited and‌​‌ contributed presentations

Chloé Audebert​​

• Mathematical modeling of adipocyte​​​‌ size distribution, invited seminar,‌ MAP5 working group on‌​‌ Modeling, Analysis and Simulation,​​​‌ April 11
• Parameter estimation​ and medical measurements, online​‌ invited seminar, Mathematical -​​ Biological Seminar, Faculty of​​​‌ Science, Charles University in​ Prague, March 25

Frédérique​‌ Clément

• Modélisation de la​​ fonction ovarienne, invited talk,​​​‌ 49th colloque des Sciences​ de l'Animal de Laboratoire​‌ (AFSTAL), November 19-21, Nantes​​

Pascale Crépieux

• Targeting FSHR​​​‌ and LHR with intracellular​ antibodies for structure-activity studies,​‌ invited seminar, Olson Center,​​ University of Nebraska Medical​​​‌ Center, June 2, Omaha​ (USA),
• Modulation of gonadotropin​‌ receptor activity with intracellular​​ VHH, invited talk, Aging​​​‌ Pituitary/Gonadal Axis Spring Retreat,​ June 3, Omaha (USA)​‌
• Des fragments d’anticorps intra-cellulaires​​ ciblant les récepteurs hormonaux,​​​‌ pour l’innovation thérapeutique, June​ 20, 10ème journée de​‌ Biotechnocentre, online,

Marlène Davilma​​

• Unveiling the role of​​​‌ mir-187 in adulte ovarian​ follicle growth and female​‌ fecundity in medaka (Orizyas​​ latipes), invited talk, Colloque​​​‌ Différenciation et fonctions des​ gonades, March 21, Paris​‌

Alice Fohr

• Dynamical systems​​ and stochastic processes applied​​​‌ to developmental biology :​ Modeling X-chromosome inactivation, poster​‌ communication, Journées Math-Bio-Santé, November​​ 5-7, Montpellier

Louis Fostier​​​‌

• A PINN-based approach to​ the calibration of structured​‌ population models, poster communication,​​ Journées Math-Bio-Santé, November 5-7,​​​‌ Montpellier
• A population dynamics​ model for fish oogenesis,​‌ invited talk to the​​ mini-symposium “Models and methods​​​‌ for applications in Biology​ and Medicine”, Biennale de​‌ la SMAI, June 2-5,​​ Carcans-Maubuisson

• Size-structured population models​​​‌ with applications to cell​ populations

  contributed talk, Differential​‌ Equations and Applications to​​ Biology, June 16-20, Le​​​‌ Havre

  invited seminar, Math​ Bio seminar of Institut​‌ Denis Poisson, September 24,​​ Tours

  invited seminar, Laboratoire​​​‌ de Mathématiques Appliquées du​ Havre, November 27, Le​‌ Havre

Stefan Haar

• Searching​​ for Attractors: To infinity​​​‌ and beyond, invited talk,​ AUTOMATA 2025, June 30​‌ - July 2, Lille​​
• Continuous Petri nets faithfully​​​‌ fluidify most permissive boolean​ networks, contributed talk, 23rd​‌ International Conference on Computational​​ Methods in Systems Biology​​​‌ (CMSB 2025), September 10-12,​ Lyon

Frédéric Jean-Alphonse

• LHR​‌ and Gs trafficking drive​​ sustained cAMP signaling from​​​‌ endosomes to control steroidogenesis,​ invited talk, neuroGPCR Symposium,​‌ September 17, Bordeaux

Pawan​​ Kumar

• Decoding Biology Hackathon​​​‌ (Selected participant and special​ jury award with Cellf-Driving​‌ Agents team), OWKIN, September​​ 29 - October 1,​​​‌ Paris
• Glioblastoma Moonshot Hackathon​ (finalist with GlioMatrix team),​‌ OWKIN and Servier, February​​ 3–4, Paris
• Extending tumor​​​‌ hypoxia modeling from 2D​ to 3D: A mechanistic,​‌ data-driven approach informed by​​ 2D histopathology, poster presentation,​​​‌ Advanced Lecture Course on​ Computational Systems Biology (CompsysBio2025),​‌ October 6-10, Aussois
• Multiscale​​ modeling of ovarian development:​​​‌ Mechanisms underlying the selection​ of future oocytes, contributed​‌ talk, Journées Math-Bio-Santé, November​​ 5-7, Montpellier

Rosario Medina​​​‌ Rodriguez

• Detection and classification​ of ovarian follicles in​‌ histopathology images, poster presentation,​​ Spring School of Deep​​​‌ Learning for Medical Imaging​ 2025, April 21-25, Lyon​‌

Eleonora Pastremoli

• Towards a​​ digital twin of the​​​‌ gut microbiota: Multiscale modeling​ and host interaction, invited​‌ talk to the minisymposium​​ “Digital twins for biological​​​‌ systems: Advancing multiscale modeling​ and hybrid solutions”, Biennale​‌ de la SMAI, June​​ 2-5, Carcans-Maubuisson

Lorenzo Sala​​​‌

• Physics-informed neural networks for​ data-integrated modeling of microbial​‌ interactions, contributed talk, 16th​​ Conference on Dynamical Systems​​ applied on Biology and​​​‌ Natural Sciences (DSABNS25), January‌ 20-24, Napoli (Italy)
• Physics-informed‌​‌ neural networks for generalized​​ Lotka-Volterra models: towards parameter​​​‌ estimation in microbial communities,‌ invited talk to the‌​‌ mini-symposium “Models and methods​​ for applications in Biology​​​‌ and Medicine”, Biennale de‌ la SMAI, June 2-5,‌​‌ Carcans-Maubuisson
• Hybrid inference for​​ microbial community models: Physics-informed​​​‌ neural networks for parameter‌ estimation in generalized lotka-volterra‌​‌ system, invited talk, Optimization,​​ and Control in Biomedicine​​​‌ Worshop, September 28-29, Frankfurt‌ am Main (Germany)
• Hybrid‌​‌ data integration with PINNs:​​ Mechanistic modeling of biological​​​‌ systems using omics data,‌ invited talk, Multi-Omics and‌​‌ Data Integration Conference, October​​ 16-17, Nice
• Physics-informed neural​​​‌ networks for modeling plant‌ gene expression dynamics under‌​‌ stress, invited talk, Interactive​​ Mathematics Day, Institut des​​​‌ Hautes Études Scientifiques, November‌ 5, Bures-sur-Yvette
• Physics-nformed neural‌​‌ networks for parameter inference​​ in biological systems, invited​​​‌ talk 4ièmes Journées annuelles‌ du GDR MECABIO Santé,‌​‌ November 26-28, Avignon

Eric​​ Reiter

• Selection and functional​​​‌ characterization of anti-CXCR4 antagonistic‌ VHHs, invited talk, Bioproduction‌​‌ congress (BIOPC2025), September 22,​​ Lyon.
• Targeting gonadotropin receptors​​​‌ with single domain antibodies‌ to control reproduction and‌​‌ treat associated dysfunctions, invited​​ talk, 13th Antibody Industrial​​​‌ Symposium (AIS2025), June 25,‌ Tours
• Recombinant antibody fragments‌​‌ as new modulators of​​ gonadotropin receptor activity, invited​​​‌ talk, Joint congress of‌ ESPE (European Society for‌​‌ Paediatric Endocrinology) and ESE​​ (European Society of Endocrinology),​​​‌ May 12, Copenhagen (Denmark).‌
• Development of VHH to‌​‌ target GPCRs and modulate​​ their activities, invited talk,​​​‌ Symposium on Nanobodies and‌ Neuroscience, April 24, Bergen‌​‌ (Norway)

Chiara Villa

• Calibration​​ and analysis of phenotype-structured​​​‌ PDEs of cancer adaptive‌ dynamics, invited talk, Young‌​‌ women in Mathematical Biology,​​ March 31 - April​​​‌ 4, Bonn (Germany)
• Coupling‌ physiologically-based kinetic models of‌​‌ endocrine axes with structured​​ cell population dynamics models:​​​‌ an integrative approach of‌ reproductive toxicity, poster, Biennale‌​‌ de la SMAI, June​​ 2-5, Carcans-Maubuisson

Romain Yvinec​​​‌

• Stochastic Becker-Döring model: large‌ population and large time‌​‌ results for phase transition​​ phenomena, invited seminar, MOME2​​​‌ : Journée de modélisation‌ mathématique pour l'écologie, November‌​‌ 17, Amiens
• Ovarian follicle​​ population dynamics along lifetime,​​​‌ invited talk, Journées Math-Bio-Santé,‌ November 5-7, Montpellier
• Kinetic‌​‌ biased and compartmentalized signaling:​​ a system biology approach,​​​‌ contributed talk, 5th annual‌ meeting of the IRN‌​‌ iGPCRnet, July 7-9, Barcelone​​ (Spain)

Chloé Weckel

• Spatiotemporal​​​‌ modeling of signaling pathways:‌ Impact of endosomal compartmentalization‌​‌ and application to gonadotropin​​ receptors

  seminar at the​​​‌ TOP days (Tours, Orléans,‌ Poitiers), May 13-14, Tours‌​‌

  contributed talks at Advanced​​ Lecture Course on Computational​​​‌ Systems Biology (CompsysBio2025), October‌ 6-10, Aussois, and GPCR‌​‌ forum virtual conference, November​​ 12-14

  poster presentations, 5th​​​‌ annual meeting of the‌ IRN iGPCRnet, July 7-9,‌​‌ Barcelone (Spain) and Journées​​ Math-Bio-Santé, November 5-7, Montpellier​​​‌

• Characterizing the signaling pathways‌ of gonadotropin receptors considering‌​‌ the impact of endosomal​​ compartmentalization, contributed talk, Journées​​​‌ de Biostatistiques, November 3-5,‌ Montpellier

9.1.4 Leadership within‌​‌ the scientific community

Frédérique​​ Clément

• member of the​​​‌ direction board of RT‌ REPRO
• member of the‌​‌ scientific board of PIXANIM​​ (Phénotypage par Imagerie in/eX​​​‌ vivo de l'ANImal à‌ la Molécule)
• scientific member‌​‌ of the FC3R COR​​​‌
• member of the steering​ committee of OI NFC​‌ (New Frontiers in Cancer)​​ Paris-Saclay

Pascale Crépieux

• member​​​‌ (and board member) of​ CNRS section 24 ,​‌ “Physiologie, physiopathologie, biologie du​​ cancer”

Frédéric Jean-Alphonse

• coordinator​​​‌ of Key Question 1​ (How can target activity​‌ be modulated through antibody​​ binding?), LabEx MAbImprove
• member​​​‌ of the Early career​ scientist comittee (ECS) of​‌ iGPCRnet

Béatrice Laroche

• member​​ of the steering committee​​​‌ of the INRAE metagrogram​ HOLOFLUX

Anne Poupon

• coordinator​‌ of “Central Development Instrument​​ 1 (Interdisciplinary Innovation)”, LabEx​​​‌ MAbImprove

Romain Yvinec

• member​ of the steering committee​‌ of RT REPRO
• member​​ of the Directory committee​​​‌ of iGPCRnet

9.1.5 Scientific​ expertise

• Frédérique Clément ,​‌ reviewer for the Czech​​ Science Foundation, member of​​​‌ the selection board for​ PhD fellowships from OI-NFC​‌
• Pascale Crépieux , member​​ of the selection board​​​‌ for the recruitment of​ CNRS CRCN in Section​‌ 24
• Béatrice Laroche ,​​ member of the selection​​​‌ board for DR2 INRAE​ “Génétiques animale et végétale,​‌ santé animale, physiologie”, member​​ of the selection board​​​‌ for the recruitment of​ Inria CRCN in Centre​‌ Inria de l'Université Grenoble​​ Alpes
• Eric Reiter ,​​​‌ member of the selection​ board for PhD and​‌ Postdoc fellowships from PEPR​​ SAFE, member of the​​​‌ selection board for the​ Endocrinology, Diabetes and Metabolism​‌ award 2025, FNRS and​​ FWO (Belgium)
• Romain Yvinec​​​‌ , member of the​ selection board for the​‌ recruitment of a MCU​​ in Université de Tours​​​‌ on job profile “Analyse​ des Équations aux Dérivées​‌ Partielles (EDP) et Applications”​​ (MCF 0782)

9.1.6 Research​​​‌ administration

• Frédérique Clément is​ invited member of the​‌ scientific council of Graduate​​ School Life Sciences and​​​‌ Health of University Paris-Saclay,​ and member of Bureau​‌ du comité des équipes-projets​​ du Centre Inria de​​​‌ Saclay
• Béatrice Laroche is​ director of MaIAGE
• Eric​‌ Reiter is deputy director​​ of UMR PRC
• Romain​​​‌ Yvinec is co-head of​ the Bios team in​‌ UMR PRC, co-head of​​ the regional federative structure​​​‌ CaSciModOT (Calcul Scientifique et​ Modélisation Orléans Tours)

9.2.1 Teaching​

• Pascale Crépieux , M2​‌ Biology of Reproduction (2h),​​ M2 Infectiology, Immunity, Vaccinology​​​‌ and Biopharmaceuticals (3h), Université​ de Tours
• Alice Fohr​‌ , L1 Mathematics/Computer Science​​ (Algebra and Geometry, 24h)​​​‌ , L1 Biology/Chemestry/Earth sciences​ (Mathematics for modeling, 28h),​‌ L2 Physics/Mathematics (Analysis and​​ Geometry, 24h), Université Paris​​​‌ Saclay
• Louis Fostier ,​ L1 Computer Science, Université​‌ de Tours, Algebra and​​ Analysis (54h)
• Stefan Haar​​​‌ , M2 Bioinformatics (Analysis​ of dynamics in biological​‌ networks, 24h), Université Paris​​ Saclay
• Frédéric Jean-Alphonse ,​​​‌ M2 Physiopathology (2h), Université​ de Tours
• Lorenzo Sala​‌ , M2 Mathématiques et​​ applications (3h30) Université Paris-Cité,​​​‌ M2 Mathématiques pour les​ Sciences du Vivant (2h)​‌ Université Paris-Saclay
• Romain Yvinec​​ , M2 Infectiology, Immunity,​​​‌ Vaccinology and Biodrugs (3h),​ M2 Physiopathology (2h), Université​‌ de Tours
• Chloé Weckel​​ , L1 Biology (Mathematical​​​‌ tools, 6h), L2 Biology​ (Statistics, 18h), Université de​‌ Tours

9.2.2 Supervision

• PhD:​​ Juliette Gourdon, “Manipulation of​​​‌ the intracellular traffic and​ endosomal signaling of gonadotropin​‌ receptors, LH/CGR and FSHR,​​ by nanobodies: deciphering the​​ molecular mechanisms and the​​​‌ consequences on reproduction”, defended‌ on June 3, supervisors:‌​‌ Frédéric Jean-Alphonse and Eric​​ Reiter
• PhD: Louis Fostier,​​​‌ “Multiscale mathematical modeling of‌ oogenesis in fish”, defended‌​‌ on October 10, supervisors:​​ Frédérique Clément and Romain​​​‌ Yvinec , associate supervisor:‌ Violette Thermes
• PhD in‌​‌ progress: Aloïs Dauger, “Dynamics​​ of adipose tissue during​​​‌ changes in food intake‌ using mathematical and numerical‌​‌ approaches”, started September 2023,​​ supervisors: Chloé Audebert and​​​‌ Hedi Soula
• PhD in‌ progress: Aleksandra Tomaszek, “Adipocyte‌​‌ size dynamics: models, mathematical​​ and numerical approaches, data​​​‌ comparison”', started October 2023,‌ supervisors: Chloé Audebert and‌​‌ Laurent Boudin
• PhD in​​ progress: Marlène Davilma, “Role​​​‌ of miRNAs in the‌ control of oocyte reserve‌​‌ in fish”, started October​​ 2023, supervisors: Frédérique Clément​​​‌ and Violette Thermes
• PhD‌ in progress: Alice Fohr,‌​‌ “Modeling of X chromosome​​ inactivation”, started September 2024,​​​‌ supevisors: Frédérique Clément and‌ Hélène Leman
• PhD in‌​‌ progress: Souhila Founas, “Development​​ of a territorial approach​​​‌ for the representation of‌ multirisk scenarios using Petri‌​‌ nets”, started October 2023,​​ supervisors: Corinne Curt and​​​‌ Stefan Haar
• PhD in‌ progress: Jérôme Kowalski “Whole-body‌​‌ vascular transport and pharmacokinetics​​ models: Application to imaging,​​​‌ in particular of liver”,‌ started December 2022, supervisors:‌​‌ Irène Vignon-Clementel and Lorenzo​​ Sala , associate supervisor​​​‌ Dirk Drasdo
• PhD in‌ progress: Eleonora Pastremoli, “Towards‌​‌ a digital twin of​​ the gut microbiota: A​​​‌ multidisciplinary approach for an‌ in-depth understanding of composition,‌​‌ function and interaction with​​ the host”, started October​​​‌ 2023, supervisors : Béatrice‌ Laroche and Lorenzo Sala‌​‌
• PhD in progress: Pamela​​ Romero Jofré, “Computational modeling​​​‌ of biased signaling in‌ G protein-coupled receptors”, started‌​‌ October 2024, supervisors: Misbah​​ Razzaq and Romain Yvinec​​​‌
• PhD in progress: Chloé‌ Weckel, “Spatiotemporal modeling of‌​‌ signaling pathways: impact of​​ endosomal compartmentalization and application​​​‌ to gonadotropin receptors”, started‌ October 2024, supervisors: Stefan‌​‌ Haar and Romain Yvinec​​ , associate supervisor: Frédéric​​​‌ Jean-Alphonse
• PhD follow-up committee‌ of Virginie Loison (ED‌​‌ STIC), member Chloé Audebert​​
• PhD follow-up committee of​​​‌ Federica Padovano (ED 386),‌ member Chloé Audebert
• PhD‌​‌ follow-up committee of Viviana​​ Gavilanes Guerrero (ED 515),​​​‌ member Chloé Audebert
• PhD‌ mobility internship of Sagbo‌​‌ Mélain Zinsou (MAP5), supervisor​​ Lorenzo Sala
• Master Internship:​​​‌ Tayssir Aittoumani, M1 Biology‌ of Reproduction, Université de‌​‌ Tours, supervisors: Pascale Crépieux​​ and Gilles Bruneau
• Master​​​‌ internship: : Malo Denoual,‌ M2 Modeling and Ecology,‌​‌ Université de Rennes, supervisors:​​ Béatrice Laroche and Lorenzo​​​‌ Sala
• Master internship: :‌ Omar Gawas, M2 Applied‌​‌ Mathematics, Université de Bordeaux,​​ supervisors: Guillaume Gautreau and​​​‌ Lorenzo Sala
• Master internship:‌ Julien Joachim, M2 MPRI‌​‌ (Master Parisien de Recherche​​ en Informatique), Université Paris-Saclay,​​​‌ supervisors: Marc de Visme‌ and Stefan Haar
• Master‌​‌ internship: : Victor Schmitt,​​ M2 Modeling and Ecology,​​​‌ Université de Rennes, supervisors:‌ Silvia Bottini and Lorenzo‌​‌ Sala
• Master internship: Lou-Anne​​ Trapeaud M2 Biology of​​​‌ Reproduction, Université de Tours,‌ supervisor: Eric Reiter
• Postdoc:‌​‌ Léo Darrigade, “PDMP-based modeling​​ of the GPCR compartmentalized​​​‌ signaling”, ended on March‌ 31, supervisor: Romain Yvinec‌​‌
• Postdoc: Pawan Kumar, “Multiscale​​ modeling of single cell-based​​​‌ dynamics of ovarian development”,‌ started October 2024, supervisors:‌​‌ Chloé Audebert , Frédérique​​​‌ Clément and Fabien Crauste​
• Postdoc: Geoffrey Lacour, “Mathematical​‌ modeling and the development​​ of numerical tools based​​​‌ on neural networks for​ the analysis and prediction​‌ of microbial consortia”, supervisors:​​ Béatrice Laroche and Lorenzo​​​‌ Sala
• Postdoc: Rosario Medina-Rodríguez,​ “Deep learning models for​‌ automatic ovarian follicle detection​​ from 2D and 3D​​​‌ imaging data”, started June​ 2024, supervisor: Frédérique Clément​‌
• Postdoc: Chiara Villa, “Coupling​​ physiologically-based kinetic models of​​​‌ endocrine axes with structured​ cell population dynamics models:​‌ An integrative approach of​​ reproductive toxicity”, February-August, supervisors:​​​‌ Frédérique Clément , Romain​ Yvinec

9.2.3 Juries

Chloé​‌ Audebert

• PhD Jury of​​ Charlotte Dugourd-Camus, Université Claude​​​‌ Bernard Lyon 1, December​ 4

Pascale Crépieux

• PhD​‌ jury of Tainara Michelotti​​ (referee), Université de Clermont-Ferrand,​​​‌ December 5

Stefan Haar​

• PhD Jury of Aurélie​‌ Kong Win Chang (referee),​​ université de Grenoble, May​​​‌ 16

Frédéric Jean-Alphonse

• PhD​ Jury of Ana Novak,​‌ Université d'Orléans, November 3​​
• PhD Jury of Lara​​​‌ Baschieri (referee), University of​ Modena and Reggio Emilia,​‌ December 25 2024

Béatrice​​ Laroche

• HDR Jury of​​​‌ Elie Desmond-Le Quéméner, Université​ de Montpellier, December 11​‌
• HDR Jury od Pascal​​ Zongo (referee), Université de​​​‌ Tours, February 27
• PhD​ Jury of Ségolène Lireux​‌ (referee), Université de Besançon,​​ April 10
• PhD Jury​​​‌ of Julien Lombard, Université​ de Lille, January 27​‌
• PhD Jury of Martin​​ Garic, Sorbonne Université, September​​​‌ 11

Lorenzo Sala

• PhD​ Jury of Louis Fostier,​‌ Université de Tours, October​​ 10

Romain Yvinec

• PhD​​​‌ Jury of Emrys Reginato​ (referee), Université Grenoble Alpes,​‌ November 24

9.3.1 Participation in Live​​​‌ events

Anne Poupon has​ participated in a round​‌ table for the launching​​ of the SERENA program​​​‌ (regional mentoring program for​ female researchers and teacher-researchers),​‌ July 3, Tours.

Romain​​ Yvinec has animated a​​​‌ workshop on ovarian follicle​ detection in Fête de​‌ la science, October 11-12,​​ Tours.

Chloé Weckel has​​​‌ been involved in the​ Scientific mediation mission “Les​‌ Sciences, c'est leur chance”​​ settled in two primary​​​‌ school classes in Tours​ (32 hours) with the​‌ project entitled “From air​​ to breathing: discovering that​​​‌ they are scientists”. This​ project was awarded a​‌ “la main à la​​ pâte" prize from the​​​‌ French Science Academy.

Chloé​ Weckel has been involved​‌ in the “Chiche” action​​ from Inria: Intervention in​​​‌ five high school classes​ to introduce the world​‌ of digital and mathematical​​ research (5h)

9.3.2 Others​​​‌ science outreach relevant activities​

Chloé Audebert has participated​‌ in the organization of​​ Paris “Maths C pour​​​‌ L” one-week internship for​ undergraduate students.

International journals

• 28 article‌L.Livio Casarini,‌​‌ N.Nafis Rahman,​​ E.Eric Reiter,​​​‌ P.Pascale Crépieux,‌ A.Adolfo Rivero-Müller,‌​‌ K.Kim Jonas,​​ T. R.T Rajendra​​​‌ Kumar, A.Alfredo‌ Ulloa-Aguirre, S.Stephen‌​‌ Franks, D.Daniel​​ Bernard, S. G.​​​‌Stine Gry Kristensen,‌ J.John Davis,‌​‌ G.George Bousfield,​​ J.James Dias,​​​‌ C.Claus Andersen,‌ D.David Handelsman,‌​‌ I.Ilpo Huhtaniemi and​​​‌ M.Manuela Simoni.​ Comment on “Atlas of​‌ Fshr Expression From Novel​​ Reporter Mice”.Endocrinology​​​‌1665May 2025​HALDOI
• 29 article​‌F.Frédérique Clément,​​ L.Louis Fostier and​​​‌ R.Romain Yvinec.​ Bifurcation analysis of a​‌ size-structured population model: Application​​ to oocyte dynamics and​​​‌ ovarian cycle.SIAM​ Journal on Applied Dynamical​‌ Systems24Issue 3​​September 2025, 2427​​​‌ - 2472HALDOI​back to text
• 30​‌ articleL.Louis Fostier​​, A.Aloïs Dauger​​​‌, R.Romain Yvinec​, M.Magali Ribot​‌, C.Chloe Audebert​​ and H.Hédi Soula​​​‌. Rapid cell turnover​ to model adipocyte size​‌ distribution.Journal of​​ Theoretical Biology6182026​​​‌HALback to text​
• 31 articleJ.John​‌ Hanna, P.Pavlos​​ Varsos, J.Jérôme​​​‌ Kowalski, L.Lorenzo​ Sala, R.Roel​‌ Meiburg and I.Irene​​ Vignon-Clementel. A comparative​​​‌ analysis of metamodels for​ 0D cardiovascular models, and​‌ pipeline for sensitivity analysis,​​ parameter estimation, and uncertainty​​​‌ quantification.Computers in​ Biology and Medicine193​‌July 2025, 110381​​HALDOI
• 32 article​​​‌A.Amaury Herbet,​ M.Marie Hautière,​‌ F.Frédéric Jean-Alphonse,​​ D.Delphine Vivier,​​​‌ C.Christophe Leboeuf,​ N.Narciso Costa,​‌ A.Aloïse Mabondzo,​​ G.Guilhem Bousquet,​​​‌ F.Franck Denat,​ E.Eric Reiter and​‌ D.Didier Boquet.​​ Targeting the activated allosteric​​​‌ conformation of the endothelin​ receptor B in melanoma​‌ with an antibody-drug conjugate:​​ mechanisms and therapeutic efficacy​​​‌.BJC Reports3​1January 2025,​‌ 3HALDOI
• 33​​ articleP. J.Paguiel​​​‌ Javan Hossie, B.​Béatrice Laroche, T.​‌Thibault Malou, L.​​Lucas Perrin, T.​​​‌Thomas Saigre and L.​Lorenzo Sala. Surrogate​‌ modeling of interactions in​​ microbial communities through Physics-Informed​​​‌ Neural Networks..ESAIM:​ Proceedings and Surveys81​‌October 2025, 104-122​​HALDOIback to​​​‌ text
• 34 articleF.​Françoise Irlinger, C.​‌Christine Keribin, A.-S.​​Anne-Sophie Sarthou, B.​​​‌Béatrice Laroche and S.​Sandra Helinck. Pink​‌ discoloration defects associated with​​ microbial structure and metabolome​​​‌ changes in commercial bloomy​ cheeses.International Journal​‌ of Food Microbiology442​​July 2025, 111363​​​‌HALDOI
• 35 article​A.Alan Jan,​‌ P.Perrine Bayle,​​ N.Nacer Mohellibi,​​​‌ C.Clara Lemoine,​ F.Frédéric Pepke,​‌ F.Fabienne Béguet-Crespel,​​ I.Isabelle Jouanin,​​​‌ M.Marie Tremblay-Franco,​ B.Béatrice Laroche,​‌ P.Pascale Serror and​​ L.Lionel Rigottier-Gois.​​​‌ A consortium of seven​ commensal bacteria promotes gut​‌ microbiota recovery and strengthens​​ ecological barrier against vancomycin-resistant​​​‌ enterococci.Microbiome13​1May 2025,​‌ 129HALDOI
• 36​​ articleF.Florent Kempf​​​‌, R.Rosanna Drumo​, A.-M.Anne-Marie Chaussé​‌, P.Pierrette Menanteau​​, T.Tereza Kubasova​​​‌, S.Sylvie Roche​, A.-C.Anne-Christine Lalmanach​‌, R.Rodrigo Guabiraba​​, T.Thierry Chaumeil​​​‌, G.Guillaume Larivière-Gauthier​, I.Ignacio Caballero-Posadas​‌, B.Béatrice Laroche​​, I.Ivan Rychlík​​, I.Isabelle Virlogeux-Payant​​​‌ and P.Philippe Velge‌. The immune response‌​‌ modulated by inoculation of​​ commensal bacteria at birth​​​‌ impacts the gut microbiota‌ and prevents Salmonella colonization‌​‌.Gut microbes17​​1December 2025,​​​‌ 2474151HALDOI
• 37‌ articleJ.Jérôme Kowalski‌​‌, L.Lorenzo Sala​​, D.Dirk Drasdo​​​‌ and I.Irene Vignon-Clementel‌. Derivation and validation‌​‌ of compartment models: implications​​ for dynamic imaging.​​​‌IEEE Transactions on Medical‌ ImagingNovember 2025.‌​‌ In press. HALDOI​​
• 38 articleC.Clara​​​‌ Lazzaretti, E.Elia‌ Paradiso, S.Samantha‌​‌ Sperduti, R.Rachel​​ Richardson, C.Carmela​​​‌ Perri, R.Riccardo‌ Bellei, S.Sara‌​‌ d'Alessandro, L.Lara​​ Baschieri, C.Claudia​​​‌ Fusco, F.Frederic‌ Jean-Alphonse, E.Eric‌​‌ Reiter, M.Monica​​ Lispi, M.Manuela​​​‌ Simoni, A.Aylin‌ Hanyaloglu and L.Livio‌​‌ Casarini. Trafficking of​​ luteinizing hormone receptor directs​​​‌ the differential signal activation‌ between luteinizing hormone and‌​‌ chorionic gonadotropin.International​​ Journal of Biological Macromolecules​​​‌3183June 2025‌, 145247HALDOI‌​‌back to text
• 39​​ articleC.Clara Lazzaretti​​​‌, S.Samantha Sperduti‌, G.Ginevra Pelagatti‌​‌, C.Carmela Perri​​, L.Lara Baschieri​​​‌, C.Claudia Fusco‌, S.Serena de‌​‌ Carlini, G.Giulia​​ Canu, M.Manuela​​​‌ Varani, A.Alessia‌ Nicoli, D.Daria‌​‌ Morini, M. T.​​Maria Teresa Villani,​​​‌ F.Francesca Fanelli,‌ E.Eric Reiter,‌​‌ M.Manuela Simoni,​​ A.Aylin Hanyaloglu and​​​‌ L.Livio Casarini.‌ Heteromerization of G Protein-coupled‌​‌ Estrogen Receptor With the​​ LH Receptor Biases G​​​‌ Protein Signaling.Endocrinology‌1671January 2026‌​‌, bqaf174HALDOI​​
• 40 articleT.Thibault​​​‌ Malou, N.Nicolas‌ Parisey, K.Katarzyna‌​‌ Adamczyk-Chauvat, E.Elisabeta​​ Vergu, B.Béatrice​​​‌ Laroche, P.-A.Paul-André‌ Calatayud, P.Philippe‌​‌ Lucas and S.Simon​​ Labarthe. Biology-Informed inverse​​​‌ problems for insect pests‌ detection using pheromone sensors‌​‌.Peer Community Journal​​52025, e19​​​‌HALDOIback to‌ text
• 41 articleT.‌​‌ C.Tainara Cristina Michelotti​​, V.Valérie Lamothe​​​‌, F.Frédéric Jean-Alphonse‌, E.Eric Reiter‌​‌, M.Muriel Bonnet​​ and G.Guillaume Durand​​​‌. Endogenous ligands of‌ bovine FFAR2/GPR43 display distinct‌​‌ pharmacological properties.Frontiers​​ in Cell and Developmental​​​‌ Biology13August 2025‌, 1645031HALDOI‌​‌back to text
• 42​​ articleA.Astrid Musnier​​​‌, Y.Yannick Corde‌, A.Adrien Verdier‌​‌, M.Mélanie Cortes​​, J.-R.Jean-René Pallandre​​​‌, C.Christophe Dumet‌, A.Adeline Bouard‌​‌, A.Abdelraouf Keskes​​, Z.Zakaria Omahdi​​​‌, V.Vincent Puard‌, A.Anne Poupon‌​‌ and T.Thomas Bourquard​​. AI-enhanced profiling of​​​‌ phage-display-identified anti-TIM3 and anti-TIGIT‌ novel antibodies.Frontiers‌​‌ in Immunology16March​​ 2025, 1499810HAL​​​‌DOI
• 43 articleT.‌T Rajendra Kumar,‌​‌ J.Juliette Gourdon and​​ E.Eric Reiter.​​​‌ Transgenic mouse models to‌ study follicle-stimulating hormone receptor‌​‌ (FSHR) function in reproduction​​​‌.LE STUDIUM Multidisciplinary​ Journal912025​‌, 1-7HALDOI​​
• 44 articleM. C.​​​‌Madeline C Rocks,​ P.Priyanka Bhatnagar,​‌ A.Alice Verticchio Vercellin​​, L.Lorenzo Sala​​​‌, B.Brent Siesky​, G.Gal Antman​‌, K.Keren Wood​​, R.Riccardo Sacco​​​‌ and A.Alon Harris​. Mathematical Modeling and​‌ Artificial Intelligence to Explore​​ Connections Between Glaucoma and​​​‌ the Gut Microbiome.​Medicina612February​‌ 2025, 343HAL​​DOI
• 45 articleL.​​​‌Lorenzo Sala, M.​Mohamed Zaid, F.​‌Faith Hughes, M.​​Marcela Szopos, V.​​​‌ H.Virginia H Huxley​, A.Alon Harris​‌, G.Giovanna Guidoboni​​ and S.Sergey Lapin​​​‌. Eye2Heart: A reduced​ mathematical model bridging cardiovascular​‌ and ocular hemodynamics.​​Mathematical Biosciences and Engineering​​​‌232January 2026​, 421 - 448​‌HALDOI
• 46 article​​L. G.Luiz Guilherme​​​‌ S. da Silva,​ R.Romain Yvinec,​‌ G.Guilherme Prata,​​ V.Varshant Dhar,​​​‌ J.John Reinitz and​ A.Alexandre Ramos.​‌ Two-State Stochastic Model of​​ In Vivo Observations of​​​‌ Transcriptional Bursts.Brazilian​ Journal of Physics55​‌4May 2025,​​ 150HALDOI

Invited​​​‌ conferences

• 47 inproceedingsS.​Stefan Haar. Searching​‌ for Attractors : To​​ infinity and beyond.​​​‌AUTOMATA 2025 - 31st​ International Workshop on Cellular​‌ Automata and Discrete Complex​​ SystemsLille, FranceJune​​​‌ 2025HALback to​ text

International peer-reviewed conferences​‌

• 48 inproceedingsS.Stefan​​ Haar and J.Juri​​​‌ Kolčák. Continuous Petri​ Nets Faithfully Fluidify Most​‌ Permissive Boolean Networks.​​LNBICMSB 2025 -​​​‌ 23rd International Conference on​ Computational Methods in Systems​‌ BiologyLyon, FranceSpringer​​September 2025HALback​​​‌ to text
• 49 inproceedings​P.Pamela Romero,​‌ J.Juliette Gourdon,​​ R.Romain Yvinec,​​​‌ F.Frederic Jean-Alphonse and​ M.Misbah Razzaq.​‌ Computational Prediction of Intracellular​​ Signaling Behavior via Machine​​​‌ Learning.Springer Communications​ in Computer and Information​‌ Science - CCIS Series​​SIMBig 2025 - 12th​​​‌ International Conference on Information​ Management and Big Data​‌Lima, PeruOctober 2025​​HALback to text​​​‌

Conferences without proceedings

• 50​ inproceedingsL.Lorenzo Sala​‌. Hybrid Inference for​​ Microbial Community Models: Physics-Informed​​​‌ Neural Networks for Parameter​ Estimation in Generalized Lotka-Volterra​‌ System.2025 Workshop​​ on Modeling, Optimization, and​​​‌ Control in BiomedicineFrankfurt,​ GermanyAugust 2025HAL​‌
• 51 inproceedingsL.Lorenzo​​ Sala. Physics-Informed Neural​​​‌ Networks for Generalized Lotka-Volterra​ models: towards parameter estimation​‌ in microbial communities.​​SMAI 2025 - 12ème​​​‌ Biennale Française des Mathématiques​ Appliquées et IndustriellesCarcans-Maubuisson​‌ (Gironde), FranceJune 2025​​HAL
• 52 inproceedingsL.​​​‌Lorenzo Sala. Physics-informed​ neural networks for data-integrated​‌ modeling of microbial interactions​​.DSABNS 2025 -​​​‌ 16th Dynamical Systems applied​ on Biology and Natural​‌ SciencesNapoli, ItalyJanuary​​ 2025HAL
• 53 inproceedings​​​‌R.Romain Yvinec.​ Ovarian follicle population dynamics​‌ along lifetime.Journées​​ Math Bio Santé 2025​​​‌Montpellier, FranceNovember 2025​HAL
• 54 inproceedingsR.​‌Romain Yvinec. Stochastic​​ Becker-Döring model: large population​​ and large time results​​​‌ for phase transition phenomena‌.MOME2 : Journée‌​‌ de modélisation mathématique pour​​ l'écologieAmiens, FranceNovember​​​‌ 2025HAL

Doctoral dissertations‌ and habilitation theses

• 55‌​‌ thesisL.Louis Fostier​​. Analysis and calibration​​​‌ of size-structured population models:‌ application to cell populations‌​‌.Université de Tours​​October 2025HALback​​​‌ to text

Reports &‌ preprints

• 56 miscL.‌​‌Luís Almeida, A.​​Alexandre Poulain, A.​​​‌Albin Pourtier and C.‌Chiara Villa. Mathematical‌​‌ modelling of the contribution​​ of senescent fibroblasts to​​​‌ basement membrane digestion during‌ carcinoma invasion.May‌​‌ 2025HAL
• 57 misc​​M.Maya Briani,​​​‌ R.Roberto Natalini and‌ M.Magali Ribot.‌​‌ A relaxation scheme for​​ the equations of isentropic​​​‌ gas dynamics on a‌ network with jump transmission‌​‌ conditions.July 2025​​HAL
• 58 miscJ.​​​‌Juliette Gourdon, M.‌Maya Haj-Hassan, V.‌​‌Vinesh Jugnarain, C.​​Christophe Gauthier, S.​​​‌Sabine Alves, F.‌Florian Guillou, É.‌​‌Éric Reiter and F.​​Frédéric Jean-Alphonse. LHR​​​‌ and Gαs trafficking drive‌ sustained cAMP signalling from‌​‌ endosomes to control steroidogenesis​​.May 2025HAL​​​‌DOIback to text‌
• 59 miscS.Sara‌​‌ Hamis, A. P.​​Alexander P Browning,​​​‌ A. L.Adrianne L‌ Jenner, C.Chiara‌​‌ Villa, P. K.​​Philip K Maini and​​​‌ T.Tyler Cassidy.‌ Growth rate-driven modelling reveals‌​‌ how phenotypic adaptation drives​​ drug resistance in BRAFV600E-mutant​​​‌ melanoma.May 2025‌HAL
• 60 miscT.‌​‌Thibault Malou, N.​​Nicolas Parisey, K.​​​‌Katarzyna Adamczyk-Chauvat, E.‌Elisabeta Vergu, B.‌​‌Béatrice Laroche, P.-A.​​Paul-André Calatayud, P.​​​‌Philippe Lucas and S.‌Simon Labarthe. Biology-Informed‌​‌ inverse problems for insect​​ pests detection using pheromone​​​‌ sensors.January 2025‌HALDOI
• 61 misc‌​‌P.Pauline Raynaud,​​ J.Juliette Gourdon,​​​‌ V.Vinesh Jugnarain,‌ L.Lucille Berthet,‌​‌ F.Frederic Jean-Alphonse,​​ O.Oceane Vaugrente,​​​‌ C.Camille Gauthier,‌ A.Amandine Vallet,‌​‌ P.Priyanka Anujan,​​ D.Delphine Naquin,​​​‌ T.Thomas Boulo,‌ C.Christophe Gauthier,‌​‌ D.Danièle Klett,​​ Y.Yves Combarnous,​​​‌ A.Aylin Hanyaloglu,‌ E.Eric Reiter,‌​‌ G.Gilles Bruneau and​​ P.Pascale Crépieux.​​​‌ A single domain intrabody‌ as a novel tool‌​‌ to bias the subcellular​​ trafficking of the follicle-stimulating​​​‌ hormone receptor.February‌ 2025HALDOIback‌​‌ to text
• 62 misc​​L.Lorenzo Sala,​​​‌ M.Mohamed Zaid,‌ F.Faith Hughes,‌​‌ M.Marcela Szopos,​​ V.Virginia H. Huxley​​​‌, A.Alon Harris‌, G.Giovanna Guidoboni‌​‌ and S.Sergey Lapin​​. Eye2Heart : a​​​‌ validated lumped-parameter model bridging‌ cardiovascular and ocular dynamics‌​‌.April 2025HAL​​