Portail HAL Nantes Université
Not a member yet
120169 research outputs found
Sort by
L'affaire de l'annulation de l'autorisation environnementale de l'A69
No full textInternational audienc
Ultrafiltration and activated carbon to secure shellfish hatcheries
No full textInternational audienceHuman pressures accumulating on the marine environment, and among the different impacts, pollution, such as that caused by pesticides, is known to jeopardize the shellfish life cycle. The SOAP (Securing shellfish hatcheries by coupling processes) project aimed to improve safety in shellfish hatcheries through the development of an innovative seawater treatment combining ultrafiltration (UF) and activated carbon (AC) processes for the disinfection and chemical decontamination of seawater. The main objectives were to (i) identify the organic micropollutants (OMP) present in coastal seawater, (ii) select optimal treatment conditions to remove these molecules, (iii) study the impact of water quality, after treatment by UF and UF + AC, on shellfish (oyster Magallana gigas and clams Ruditapes philippinarum) and microalgae feed production, and (iv) transfer the best processes to industrial partner sites to evaluate their performance under real conditions (industrial scale with real seawater). Oyster fertilization tests showed high performance of the processes (UF + granular AC or powdered AC) to treat seawater spiked with micropollutants and hatching rates obtained were similar or even higher than the control (sand filtration 30 mu m and UV-disinfected seawater). At the industrial scale, UF enabled to retain bacteria and the combination with granular AC confirmed the efficiency to remove the chemical pollution present in seawater, thus protecting the shellfish and microalgae produced in treated water. However, possible contamination after UF in pipes or AC must be considered. Regarding the microalgae production for feeding clams, UF without granular AC yielded the best results in terms of ease of use and of production. In real conditions, for 1 year, the UF pilot plant proved its robustness and viability, regardless of seawater quality or the operating conditions imposed by the industrial partners. UF-AC seawater treatment process appears to be efficient for the removal of micropollutants and mollusk pathogens
A generalized hybrid machine learning framework for predicting biohydrogen production via dark fermentation of organic wastes
No full textInternational audienceThe rising global demand for sustainable energy has directed significant attention towards biohydrogen production via dark fermentation of organic wastes. Accurate yield prediction is crucial for optimizing process conditions and enhancing overall process. This study aims to develop a robust and interpretable predictive framework that integrates kinetic modeling with a hybrid Bayesian Optimization-Artificial Neural Network (BO-ANN) approach for precise biohydrogen yield prediction. The core novelty lies in representing each substrate not as a simple category, but by its quantitative kinetic parameters from the Modified Gompertz equation, providing a biologically meaningful input. A comprehensive database compiled from the literature incorporates key process variables, including temperature, pH, residence time, and substrate concentration, along with kinetic parameters from the Modified Gompertz equation characterizing each substrate. The BO algorithm was employed to optimize the ANN architecture, and 5-fold cross-validation was used to evaluate model generalization ability. The proposed hybrid model achieved outstanding predictive performance (R² = 0.9980, RMSE = 0.0117, MAE = 0.0062), confirming its accuracy and robustness. Furthermore, SHAP analysis and correlation metrics provided interpretable insights into feature contributions, particularly the relevance of kinetic descriptors. Overall, the proposed BO-ANN framework offers a scalable, interpretable, and biologically grounded tool to improve predictive accuracy and support the design of more efficient and sustainable biohydrogen production systems
Interleukins 15 and 18 synergistically prime the antitumor function of natural killer cells through noncanonical activation of mTORC1
No full textInternational audienceThe multiprotein complex mTORC1 is essential for the increase in protein synthesis and bioenergetic metabolism that supports the proliferation of many cell types, including natural killer (NK) cells, which are important innate effectors of the antitumoral response. Here, we investigated the mechanisms of mTORC1 activation in NK cells by interleukin-15 (IL-15) and IL-18, which promote NK cell function and are components of a cytokine cocktail used to preactivate NK cells for cancer immunotherapy. Through genetic and pharmacological approaches, we showed that IL-15 activated mTORC1 through the PI3K/Akt/ERK pathway, whereas IL-18 signaled through the p38 effectors MK2 and MK3 in both murine and human primary NK cells. Both pathways synergized to promote NK cell proliferation and effector functions in an mTORC1-dependent manner. Moreover, both pathways operated independently of the inhibitor TSC and the activator Rheb, revealing a noncanonical mode of mTORC1 activation by cytokines. Treating mice with IL-15 and IL-18 in combination led to increased NK cell numbers and improved antitumoral activity, suggesting that this cytokine combination could be exploited to enhance NK cell potential in therapeutic settings
Agriculture, aquaculture et pêche : impacts des modes de production labellisés sur la biodiversité. Résumé du rapport scientifique de l'étude
No full textLa loi Climat et résilience de 2021 a instauré la mise en place d’un affichage environnemental sur les produits alimentaires afin d’informer les consommateurs du coût environnemental de leurs achats. La construction de cet affichage a suscité un important travail méthodologique ouvert aux acteurs. Un bilan intermédiaire a souligné la difficulté à appréhender toutes les dimensions de la biodiversité. C’est dans ce cadre que les ministères en charge de la transition écologique, et de l’agriculture et de l’alimentation, ainsi que l’ADEME, ont sollicité INRAE et l’Ifremer courant 2022 pour mieux documenter ce volet biodiversité, en se focalisant sur les pratiques de production. Afin d’éclairer plus largement les politiques publiques, les pouvoirs publics ont choisi de s’appuyer sur les labels dont les cahiers des charges certifient des pratiques et parce que leur développement les place au cœur de nombreux débats sur les relations entre production et consommation durables. L’étude, intitulée « BiodivLabel », a été menée par un comité pluridisciplinaire d’experts scientifiques issus d’organismes publics de recherche et d’enseignement supérieur
A study into rare GPR146 gene variants in humans and mice
No full textInternational audienceBackground and aims: G-protein coupled receptor 146 (GPR146)-deficient mice exhibit a moderate 21 % reduction in plasma cholesterol. This is associated with decreased phosphorylation of ERK1/2 and reduced SREBP2 activity in the liver, which leads to lower VLDL secretion. Insight into the role of GPR146 in humans is however limited. We therefore set out to study rare genetic variants in GPR146 to improve our understanding of this new player in lipid metabolism.Methods: We used whole genome sequencing data from UK Biobank participants to search for rare coding variants in GPR146. We first carried out gene-based burden tests (using SAIGE-GENE-framework) and examined the association of individual variants with plasma cholesterol levels. One of the variants (P62L) was also studied using the Global Lipids Genetics Consortium (GLGC) data set and in a knock-in mouse model.Results: We found that the combination of rare genetic variants identified in GPR146 is significantly associated with plasma cholesterol levels. Three rare variants, i.e. P62L, I129I, and A175T were individually associated with reduced plasma cholesterol. In the GLGC cohort, the P62L variant was associated with reductions in both HDL and LDL cholesterol. Follow-up experiments show lower plasma cholesterol levels in GPR146 P61L male and female mice (-13 %, p < 0.05 and -15 %, p < 0.005, respectively) when compared to controls due to a reduction in HDL cholesterol. The GPR146 P61L mice did not exhibit a change in VLDL secretion. In line, the ERK1/2 signalling pathway and Srebp2 mRNA expression in liver homogenates, and the secretion of apoB by primary hepatocytes of GPR146 P61L and wild-type mice were unchanged.Conclusions: This study shows that rare GPR146 gene variants are associated with lower plasma cholesterol levels in humans. One of these variants, P62L is associated with reductions of HDL cholesterol and LDL cholesterol in humans while the ortholog in mice confers a loss of GPR146 function leading to only reduced HDL cholesterol. How GPR146 affects HDL metabolism in humans and mice remains to be resolved
Evolving landscape of thrombotic microangiopathy in kidney transplant recipients in the post–C5 inhibitor era
No full textInternational audienceA comprehensive analysis was performed on all consecutive biopsy-proven thrombotic microangiopathy (TMA) complicating kidney transplantation in the post-C5 inhibitor era (from 2009) to identify pathological profiles, determine causes, and establish risk factors associated with death-censored graft survival, in 2 French centers. Pathological criteria were assessed according to the TMA Banff Working Group, followed by an unbiased analysis to identify distinct subgroups. One hundred nineteen cases were identified, 8 (6.7%) primary TMA, and 23 (19.3%) antibody-mediated rejection. In 98 cases (82.4%), more than 1 potential trigger was involved. Latent class analysis identified 2 groups: acute TMA pattern (n = 79 [66.4%]), enriched for fibrin thrombi in glomerular capillaries and arterioles and mesangiolysis, and chronic active TMA pattern (n = 40 [33.6%]), enriched for collapsed capillaries. Both had similar presentations, were not indicative of specific causes, but had different outcomes. In multivariate analysis, grade 3 hypertension, low hemoglobin levels, proteinuria, baseline serum creatinine, and the value of a Banff-based chronicity index (ct + ci + 2xcg + cv) were associated with poorer death-censored graft survival, whereas fibrin thrombi in glomerular capillaries were associated with a better outcome. Kidney transplant-associated TMA is a severe condition in which the pathological pattern may reflect the disease stage, rather than the often intricate underlying mechanism
Using protein blocks to build custom fragment libraries from protein structures
No full textInternational audienc
Compact Cassegrain Antenna based-on a Low-Profile Multi-faceted Reflectarray
No full textInternational audienceThis paper presents the design of a compact Cassegrain antenna that uses a multi-faceted reflectarray as a main reflector. The multi-faceted surface comprises several panels with different angular orientations, to conform a cylindrical parabolic profile along one plane. A prototype of this antenna concept has been manufactured and tested to provide a high-gain in dual-linear polarization at Ka band. When compared to an equivalent single-facet aperture, the multi-faceted approach effectively benefits from the parabolic profile optics to reduce the differential spatial phase delay. Consequently, the required phase excursion is smaller compared to the single-facet flat topology, resulting in enhanced beamwidth stability in the sectorization plane. The proposed low-profile multi-faceted antenna demonstrates better in-band performance than other similar works in the literature, while maintaining similar compactness
Adaptive Multi-fidelity Surrogate Modelling for High-quality Shape Optimization
No full textInternational audienceSurrogate modeling and active learning methods for simulation-driven design optimization of innovative vehicles are negatively influenced by numerical noise, which is often unavoidable for numerical solvers. We propose a new approach for uncertainty estimation with noisy data, which models the uncertainty as separate contributions from the noise-affected training points, their interpolation, and the multi-fidelity corrections. The surrogate model and the associated interpolation uncertainty are reconstructed with Stochastic Radial Basis Functions (SRBF), which use a range of RBF fits with different kernels. The noise in the training points is filtered out by reconstructing filtered data in the training points, which are then interpolated with standard SRBF. For the filtering, several RBF surrogates with a number of kernels smaller than the number of training points are least-squares fitted through the data; the uncertainty contribution from the noise filtering is estimated as the variance of these reconstructions. Tests on analytical functions and airfoil shape optimization show that active learning based on this estimator is more efficient and more robust than existing approaches