Portail HAL UNIV-RENNES
Not a member yet
180610 research outputs found
Sort by
In vitro hepatic metabolism and associated binding of enniatin B
No full textInternational audienceEnniatin B (ENNB), a secondary metabolite produced by various Fusarium species, frequently contaminates cereals, which constitutes the main dietary source of human exposure. Critical toxicodynamic and toxicokinetic data gaps currently impede a robust and accurate risk assessment. The primary objective of this study was to characterize the in vitro toxicokinetic profile of ENNB. We investigated its clearance using human and mouse liver microsomes (HLM and MLM), as well as 2D and 3D HepaRG cell models. To facilitate reliable in vitro-in vivo extrapolation (IVIVE), we also determined key parameters: plasma protein binding, binding to microsomes and HepaRG cells, CYP450 inhibition, and the identification of HepaRG metabolites. Binding studies revealed a very high binding of ENNB to human plasma proteins and a high binding to inactivated human liver microsomes and HepaRG cells. The predicted in vivo hepatic clearance (clH,blood) of ENNB, calculated using the in vitro results from MLM, HLM, and HepaRG 2D indicated a low hepatic first pass effect. Interestingly, no observable disappearance of ENNB was found in the HepaRG 3D model. Our findings on ENNB-mediated CYP inhibition in HLM, in combination with literature results, suggest a potential for CYP3A4/5-related auto-inhibition. Finally, we successfully performed a putative identification of 13 Phase I metabolites using the human HepaRG cell line. In conclusion, the low hepatic first pass effect could imply a high oral bioavailability in vivo if intestinal barrier passage is significant, as predicted elsewhere. However, this could be counteracted by transport limited hepatic clearance that should be further investigated and a pre-systemic first-pass effect already demonstrated in vitro
Post-exercise Ankle brachial Index Increases the Detection of Peripheral Artery Disease in Patients with Exertional Limb Symptoms
No full textInternational audienc
Mesozoic magmatism in the Andes of southern Ecuador and northern Peru: Tectonic insights from whole-rock chemistry and zircon petrochronology
No full textInternational audienceThe southern Ecuador–northern Peru region marks the transition between the northern and central Andes. This study reconstructs the Mesozoic magmatic history of this key region by integrating petrography, Usingle bondPb geochronology, whole-rock and zircon geochemistry, and εHf(t) and δ18O zircon isotopic data from plutonic rocks. Our results indicate that much of the Mesozoic magmatism occurred in an extensional arc setting, with magmatic reservoirs progressively incorporating more depleted, mantle material, while crustal contributions diminished through time. Magmatic reservoirs evolved both spatially and temporally, beginning with an extensive Triassic arc dominated by granitoids exhibiting strong crustal signatures at least until 220 Ma. This was followed by mildly enriched signatures associated with a stationary Jurassic to Early Cretaceous arc active between ∼190 and ∼ 126 Ma. Somewhere in between 126 and 104 Ma, the arc underwent a significant westward migration, potentially driven by slab rollback, which coincided with the opening of the Celica–Lancones Basin and the subsequent emplacement of the Late Cretaceous Celica–Lancones arc onto oceanic basement. This migration is consistent with westward shifts observed in central Ecuador and Colombia but contrasts with coeval eastward migration documented in central and southern Peru. In addition, new Usingle bondPb ages challenge current interpretations of a missing Jurassic arc in northern Peru by providing clear evidence that Jurassic magmatism extended at least as far south as 6°S
L’engagement des adjoints de direction d’établissement scolaire public et privé
No full textInternational audienc
AI-driven optimization of Congo red photo degradation using the spinel CdCr2O4 photocatalyst: From sol-gel synthesis to DT_LSBOOST predictive modeling coupled with the dragonfly algorithm
No full textInternational audienceIn this study, a nanostructure CdCr2O4 spinel photocatalyst was successfully synthesized via a low-cost sol-gel combustion route and thoroughly characterized by XRD, TGA-DTA, SEM-EDS, FTIR, and UV-Vis spectroscopy. The catalyst exhibited a well-defined spinel structure, high crystallinity, and nanometric grain size (similar to 29 nm), with strong visible-light absorption (band gap approximate to 1.97 eV). Photocatalytic performance was evaluated using Congo red (CR) as a model pollutant under visible LED light. Optimal degradation conditions (pH 6, [CR] (0) = 10 mg/L, 1 g/L catalyst, 150 min) led to an outstanding removal efficiency of 98.45 %, with a kinetic constant of 2.11 x 10(-2) min(-1). Mechanistic studies revealed that hydroxyl (center dot OH) and superoxide (center dot O-2(-)) radicals played dominant roles in the degradation process. To model and optimize the system, a hybrid machine learning approach combining Decision Tree with Least Squares Boosting (DT_LSBOOST), optimized using the Dragonfly algorithm, was implemented. The model demonstrated excellent prediction accuracy (R = 0.9998, RMSE = 0.66) and successfully identified optimal operating conditions with <1 % deviation from experimental results. Stability and reusability tests confirmed the photocatalyst retained >90 % efficiency after five successive cycles, with no significant structural degradation. Compared to state-of-the-art materials, CdCr2O4 proved highly competitive in visible-light-driven photocatalysis, establishing its suitability for advanced wastewater treatment applications
Localization effects in mixed-ligand gold bis(dithiolene) complexes as single-component conductors
No full textInternational audienceMixed-ligand gold bis(dithiolene) complexes involving two non-innocent dithiolene ligands with different electronic characteristics have been developed, involving one TTF dithiolate (BMT-TTFS22- = 4 ',5 '-bis(methylthio)tetrathiafuvalene-4,5-dithiolate) as a highly electron-rich ligand and either a benzene-1,2-dithiolate (bdt) or a pyrazine-2,3-dithiolate (pzdt) as an electron-poor ligand. The monoanionic closed-shell complexes are oxidized (by electrocrystallization) to neutral radical species, which behave as single-component conductors. The notably different electronic properties of the two dithiolene ligands lead to an exacerbated spin density localization on TTF dithiolate, with the resulting SOMO localized on the less electron-rich ligand. The dissymmetry imposed by the presence of two different ligands leads to a head-to-tail arrangement in the solid state and stack dimerization. The solid-state properties of the two radical complexes [Au(BMT-TTFS2)(bdt)](center dot) and [Au(BMT-TTFS2)(pzdt)](center dot) are deduced from transport measurements under pressure (up to 21 GPa) and spin-polarized band structure calculations. The 1D electronic structure with strongly dimerized chains and a direct, large band gap explains the observed semiconducting behavior. At variance with weakly dimerized systems adopting a Mott insulator behavior sensitive to pressure effects (toward a metallic state), [Au(BMT-TTFS2)(bdt)](center dot) and [Au(BMT-TTFS2)(pzdt)](center dot) show a robust gap under pressure, a direct consequence of the reinforced dimerization of BMT-TTFS2 moieties in the solid state
D2SFNet: Dual-domain spatial-frequency network for few-shot medical image segmentation
No full textInternational audienceFew-shot learning has attracted growing attention in medical image segmentation due to its ability to achieve accurate results with limited labeled data by leveraging prior knowledge. However, existing few-shot segmentation methods are typically restricted to a single dataset and rely solely on spatial features, which limits their ability to model fine-grained anatomical structures and overlooks the utility of related datasets commonly available in clinical practice. To address these challenges, we propose a dual-domain spatial-frequency network (D2SFNet) that integrates frequency-domain information and data from heterogeneous domains. Specifically, we design a dual-domain joint training strategy that incorporates both the target and auxiliary datasets into the learning process, where the target dataset provides task-specific information while the auxiliary dataset contributes generalizable representation cues. To mitigate domain shifts in dual-domain training and enhance intra-class consistency, we introduce a novel joint alignment mechanism combining intra-and inter-domain alignment. Moreover, we employ the discrete cosine transform to extract complementary frequency-domain representations, which are dynamically fused with spatial features through a novel dynamic spatial-frequency representation (DSFR) module. Extensive experiments on three widely used medical image segmentation benchmarks demonstrate that D2SFNet consistently outperforms existing state-of-the-art methods. The source code is available at https://github.com/qchi-code/D2SFNet
Feedback-MPPI: Fast Sampling-Based MPC via Rollout Differentiation – Adios low-level controllers
No full textInternational audienceModel Predictive Path Integral control is a powerful sampling-based approach suitable for complex robotic tasks due to its flexibility in handling nonlinear dynamics and non-convex costs. However, its applicability in real-time, highfrequency robotic control scenarios is limited by computational demands. This paper introduces Feedback-MPPI (F-MPPI), a novel framework that augments standard MPPI by computing local linear feedback gains derived from sensitivity analysis inspired by Riccati-based feedback used in gradient-based MPC. These gains allow for rapid closed-loop corrections around the current state without requiring full re-optimization at each timestep. We demonstrate the effectiveness of F-MPPI through simulations and real-world experiments on two robotic platforms: a quadrupedal robot performing dynamic locomotion on uneven terrain and a quadrotor executing aggressive maneuvers with onboard computation. Results illustrate that incorporating local feedback significantly improves control performance and stability, enabling robust, high-frequency operation suitable for complex robotic systems.</div
A new fluorinated hydrazone derivative as a multitarget therapeutic agent: Synthesis, crystal structure, spectroscopic characterization, Hirshfeld surface analysis, DFT/TD-DFT studies, NLO properties, in silico molecular docking, ADMET profiling, and biomimetic oxidation activity
No full textInternational audienceA new hydrazone molecule HL: (E)-2-fluoro-N'-(1-(4-hydroxy-6-methyl-2-oxo-2H-pyran-3-yl)ethylidene)benzo-hydrazide, was synthesized by condensation of 2-fluorobenzohydrazide with dehydroacetic acid. The structure of HL was confirmed using spectroscopic analysis, including NMR (1H, 13C), UV-visible and infrared, and single crystal X-ray diffraction. This compound adopted a zwitterionic form HL' stabilized by intramolecular hydrogen bonding interactions between [N+-H...-O] groups and crystallized in the monoclinic system with space group P21/c. Indeed, Hirshfeld surface analysis was performed to visualize and quantify the intermolecular interactions within the crystalline structure, revealing the presence of intermolecular contacts involving H and sdot;and sdot;and sdot;Omicron and H and sdot;and sdot;and sdot;F hydrogen bonds and non-conventional C-H and sdot;and sdot;and sdot;H, C-H and sdot;and sdot;and sdot;it, and it and sdot;and sdot;and sdot;lp interactions, as well as it-it stacking. DFT calculations were carried out using the omega B97X-D functional with the 6-31+G(d) basis set. Based on DFT conceptual principles, key global molecular reactivity descriptors were obtained, including chemical hardness, electronic chemical potential, electronegativity, and electrophilicity index. Furthermore, the nonlinear optical properties of the compound HL and its tautomer HL <-> HL" were explored, revealing promising potential for applications in second-and third-order NLO materials. Molecular docking studies were also conducted to evaluate the in silico biological activity of HL against cholinesterase enzymes, specifically acetylcholinesterase and butyrylcholinesterase. Furthermore, the physicochemical and pharmacokinetic properties of the molecule were assessed through ADMET analysis, confirming its favorable drug-likeness characteristics. In this study, we aim to evaluate the catalytic activity of in-situ complexes formed using HL as a catalyst with CuII salts, which are commonly used in the oxidation of catechol to o-quinone
Application of deep learning and machine learning models with enhanced feature extraction for the prediction of plant extraction yields using supercritical CO2: An optimization and comparative analysis
No full textInternational audienceThe efficient extraction of essential oils (EOs), particularly volatile compounds, from medicinal, aromatic, or oil-rich crop plants using supercritical carbon dioxide extraction (scCO(2)) is crucial for industries such as pharmaceuticals, cosmetics, and food. However, optimizing this process presents challenges due to the intricate molecular diversity of the compounds and the complex interplay of scCO(2) parameters. To address these limitations, this study introduces a hybrid predictive framework that combines deep learning and machine learning, utilizing 694 scCO(2) experimental data points sourced from the literature across 21 plant species. Four major molecular compounds per plant were selected as input features, alongside key process parameters, including temperature, pressure, extraction time, co-solvent ratio, and CO2 flow rate. Morgan fingerprints were computed for these compounds, and a convolutional neural network (CNN) was utilized to extract their high-level representations into compact vectors. These vectors were integrated with normalized process parameters and fed into a CNN-Multilayer Perceptron (CNN-MLP) hybrid architecture. Performance was compared with Support Vector Regression (SVR), Random Forest (RF), Gaussian Process Regression (GPR), and XGBoost, all optimized using OPTUNA. The CNN-MLP achieved the best performance, with an R-2 of 0.974 and a Root Mean Squared Error (RMSE) of 1.431 on the test set. A paired t-test (p = 0.810) and Bland-Altman analysis (mean difference: 9.35 %) confirmed the model's robustness. To further assess generalizability, external validations were conducted using unseen experimental conditions. The CNN-MLP was tested on three extraction profiles and demonstrated strong predictive performance, with Pearson correlations ranging from 0.95 to 0.98