LAM Foundation

Abstract SCN2A variants are one of the most common genetic causes of intractable epilepsy in children, particularly in developmental and epileptic encephalopathies (DEEs) which can present with uncontrolled seizures at birth, accounting for 1-2% of all epileptic encephalopathies. There is significant genotype-phenotype heterogeneity in SCN2A-related disorders (SRD) which include neurologic symptoms of seizures, developmental delay, choreoathetosis, and autism spectrum disorder (ASD). A substantial fraction of causal variants is gain-of-functon(GOF) or mixed function, functionally associated with increased open-probability or greater sodium current flux. Individualized allele-selective antisense oligonucleotides (ASOs) were designed to target heterozygous intronic SNPs for decreased expression of mutant SCN2A transcript while preserving the wild-type copy. Efficacy measures were also individualized to phenotype. Improvements in seizure control, development, and quality-of-life were seen with no ASO-related adverse events. Haplotype phasing in a separate cohort of infants with SRD diagnosed by rapid whole genome sequencing identified 16% of patients with compatible SNPs. Allele-selective ASOs demonstrate potential to decrease seizures and impact neurodevelopment, providing a pathway for potential benefit in n-of-1 to n-of-more SRD patients.

Abstract This study combines experiments and numerical simulations (3D-Reynolds-averaged Navier–Stokes (RANS)) to achieve a deep understanding of the effects induced by varying key parameters affecting tip leakage flow in a highly loaded low-pressure turbine (LPT) rotor blade. Specifically, results for flat tip configurations are compared with a squealer tip geometry for different clearance heights and mass flow ratios simulating coolant flow ejected from the tip. Experimental results map the effects of these parameters on loss generation, while detailed insights into the interaction between the tip vortex and other vortical structures within the passage are discussed through computational fluid dynamic (CFD) simulations. Available experimental data include 2D distributions of total pressure and flow angles measured with a five-hole pressure probe downstream of the cascade for different operating conditions, enabling comparison with numerical simulations. The RANS solver provides visualizations of streamlines developing close to the tip region, offering a clear interpretation of the mechanism by which cross flow motion in the tip region interacts with the pressure gradient to generate the tip leakage vortex. The study explores how these processes vary with tip gap height and different mass flow ratios, providing a comprehensive view on the development of the secondary flow system. Finally, additional simulations with moving endwall are conducted to evaluate the impact of the relative motion between the blade and casing in the current application.

Limitar el ingreso de agua de lluvia impulsada por el viento a través de aberturas representa un desafío para la aplicación de estrategias bioclimáticas de ventilación natural en climas tropicales. Esta investigación propone y valida un método experimental para evaluar la penetración en fachadas, considerando el diámetro de las gotas y la velocidad del aire que impulsan. El método compara un modelo físico a escala con cálculos teóricos, evaluando las configuraciones: apertura libre y apertura con membrana permeable. Los resultados identifican los ángulos de incidencia más frecuentes y demuestran que la membrana puede reducir la entrada de agua entre un 21% y un 62% a altas velocidades, y entre un 88,5% y un 100% a bajas velocidades. Esta propuesta puede servir como base para futuras simulaciones computacionales que integren esta variable en la evaluación del confort y la eficiencia en fachadas.

This study presents the results of a long-term wave hindcast covering the period 1975–2024 overthe Mediterranean Sea. The hindcast was produced through a dynamical downscaling approach,based on a chain of numerical models. ERA5 global reanalysis data were downscaled using theBOLAM and MOLOCH atmospheric models, providing the wind forcing for the WW3 wave model,which simulated the sea state. WW3 adopts an unstructured computational grid with variableresolution, reaching up to 500 meters along the Ligurian and Tyrrhenian coasts (Italy), allowing fora detailed representation of the coastal wave climate.Although hindcasts do not assimilate observational data, validation against in-situ observationsshows that the generated wind and wave fields are robust and reliable, providing added valuecompared to global reanalyses. The resulting dataset represents a valuable resource for waveclimate studies, coastal risk assessment, and the analysis of long-term variability in theMediterranean region.The availability of such a long-term, high-resolution hindcast enables several potentialapplications. It provides a solid baseline for trend analysis and climate variability studies and cansupport the identification of suitable areas for offshore renewable energy development. Wepresent user cases in which the dataset was used to assess both atmospheric and marineconditions over sea areas involved in particularly sensitive operations, where atmosphericdynamics play a critical role. We show statistical analyses performed to produce monthly waitingtime tables (expressed in hours), estimating how long it typically takes for sea state conditions tofall within the operational thresholds required to safely carry out planned maritime activities.Finally, we conclude with some considerations on the computational efficiency of the modellingframework adopted, which makes the dataset particularly suitable for operational updates andfacilitates its regular extension to future years.

We continue the analysis of random series associated to the multidimensional harmonic oscillator $-\Delta + |x|^2$ on $\mathbb{R}^d$ with d \geq 2$$. More precisely we obtain a necessary and sufficient condition to get the almost sure uniform convergence on the whole space $\mathbb{R}^d$ . It turns out that the same condition gives the almost sure uniform convergence on the sphere $\mathbb{S}^{d-1}$ (despite $\mathbb{S}^{d-1}$ is a zero Lebesgue measure of $\mathbb{R}^d$). From a probabilistic point of view, our proof adapts a strategy used by the first author for boundaryless Riemannian compact manifolds. However, our proof requires sharp off-diagonal estimates of the spectral function of $-\Delta + |x|^2$ . Such estimates are obtained using elementary tools.

Indeed, deeply advanced learning for data scientists has been applied to previously energy optimized smart grids through Deep learning and Deep Reinforcement learning techniques. In particular, emerging and gaining ground for smart grids is the use of deep neural network based reinforcement learning approaches. For discrete and continuous action space we use the deep reinforcement learning algorithms. We were able to construct these algorithms into a robust real physical smart grid optimizer with the use of the MATLAB environment. It was observed that the agent could correctly identify the features in the training data related to energy supply and demand and ‘learn’ to make actions that will get it the most rewards. In the scope of our multilevel, systems based method of smart grid modelling, EGO-ENIF model and traditional methods are deployed in tandem for the sake of improvement at different levels. Besides improving the predictive accuracy of the energy consumption patterns, this approach enables energy distribution decision making to be made in real time. With the help of advanced machine learning concept and the established model, we can achieve a more stable and efficient smart grid based on resource optimization and conditions adjustment.

Computer vision applications span various fields including healthcare, security, autonomous vehicles, and augmented reality, enabling machines to interpret and analyze visual data. Facial Emotion Recognition (FER) is a subclass of healthcare applications that leverages computer vision to analyze and interpret human emotions from facial expressions. Facial emotion recognition also plays a vital role in human-computer interaction, with applications in security, and affective computing. This study suggests a deep learning (DL) based hybrid model integrating MobileNetV2 for efficient feature extraction and a Vision Transformer (ViT) for capturing global facial dependencies. The dataset obtained from Kaggle is used for training, which is then preprocessed and augmented. The trained model is deployed on a smartphone as an edge device, enabling real-time emotion recognition with improved privacy, low latency, and minimal computational overhead. During testing, facial images captured by the smartphone are preprocessed using the Haar Cascade algorithm before being fed into the model for classification. Performance evaluation using accuracy, recall, precision and F1-score demonstrates a high classification accuracy of 98.51%, confirming the model’s effectiveness. The proposed approach enhances on-device FER capabilities, making it a promising solution for emotion-aware applications in mobile healthcare and intelligent human-computer interactions

Côte d’IvoireArchives nationales de Côte d’Ivoire (ANCI), Primature, AbidjanDocumentation diverse : séries 0.60, 0.161, 0.116, 3.30, 3.31, 3.32, 3.33, 3.34, 3.35, 3.39, 1 FF.Fonds ancien (années 1920) : 1EE 13, 2EE2 (28), 1DD23, 2DD171-286, 4DD5 à 9. État civil.

La création d’allocations familiales pour les travailleurs salariés constitue avec le Code civil l’autre versant d’une politique familiale. Elle est mise en place également dès la période coloniale et se prolonge au-delà : ses lois sont adoptées en décembre 1955 et mises en pratique en Côte d’Ivoire en février 1956. Ces allocations viennent concrétiser l’article 237 du Code du travail d’outre-mer de 1952, une victoire démocratique pour les syndicats et les députés de