Abstract: Optimizing resource utilization and task execution within scalable cloud computing infrastructures remains a paramount challenge for service providers. This paper proposes and empirically evaluates a novel framework for intelligent resource orchestration, leveraging advanced learning algorithms to dynamically enhance performance. Our methodology integrates reinforcement learning principles to adaptively manage heterogeneous cloud resources, aiming to minimize task completion times and maximize system throughput. Through rigorous simulation experiments, this study demonstrates a significant improvement in resource allocation efficiency compared to conventional scheduling paradigms. The findings offer a strategic blueprint for developing autonomous and cost-effective cloud management systems, paving the way for next-generation adaptive cloud services.

Abstract: The study presents a novel process to design lightweight, high-performance cooling manifolds for power electronics using generative design (GD). The process begins with a baseline design that defines the constraints of the manifold with regard to the target cooling geometry and flow path. A GD flow optimization is then performed to minimize pressure drop and improve flow uniformity. Once a final fluid volume is obtained, a GD structural optimization is conducted to minimize weight and material usage. The final design demonstrated a 74.4% increase in heat transfer coefficient, an 87.3% improvement in uniformity, and a 63.3% reduction in weight.

Abstract: The objective of this work is to propose a simulation strategy for production planning that is compatible with the dynamism of natural gas processing, especially under open-market arrangement, in which several scheduling simulations must be performed within short time horizons. In such contexts, traditional first-principles-based ap-proaches, although accurate, require prohibitive computational times, motivating the need for an alternative simulation strategy. This work thus proposes a data-driven model built with the aid of machine learning and applied in a case study with historical data from the largest gas processing site in Brazil: Cabiúnas Petrobras asset. Main plant flowrates were selected: 18 targets and 44 input candidates – 1282 observations from three and a half years of operation. Principal Component Analysis was used for order reduction, keeping the 22 main principal components. A forward neural network (2 hidden layers and 225 neurons per layer) was built from training/test sets randomly selected and optimized hyperparameters – learning rate (0.001533) and batch size (8). Training converged in roughly 200 epochs (Adam optimizer), with early stop triggered by validation set. A mean absolute error of 0.0017 (test set) and R2=0.72 were found, a promising result considering plant complexity and data simplicity. Results showed particularly good fit for lighter products (sales gas, natural gas liquid), also indicating an opportunity for further work by including inputs related to liquid fractionation.

Abstract: The spontaneous emergence of macroscopic dissipative structures in systems driven by generalized chemical potentials is well-established in non-equilibrium thermodynamics. Some examples are, hurricanes, Bénard cells, reaction-diffusion patterns, and atmospheric/oceanic currents. Less recognized, however, are microscopic dissipative structures that form when the driving potential excites internal molecular degrees of freedom (electronic states and nuclear coordinates), typically via high-energy photons. The thermodynamic dissipation theory for the origin of life posits that the core biomolecules of all three domains of life originated as self-organized molecular dissipative structures—chromophores or pigments—that proliferated across the Archean ocean surface to absorb and dissipate the intense “soft” UV-C (205–280 nm) and UV-B (280–315 nm) solar flux into heat. Thermodynamic coupling to ancillary antenna and surface-anchoring molecules subsequently increased photon dissipation and enabled more complex dissipative processes, including modern photosynthesis, to dissipate lower-energy but higher-flux UV-A and visible light. Further thermodynamic coupling to abiotic geophysical cycles (e.g., diurnal, water cycles, winds, and ocean currents) ultimately produced today’s biosphere, efficiently dissipating the full incident solar spectrum well into the infrared. This paper details three examples of molecular dissipative structuring (nucleotides, fatty acids, pigments) and argues that dissipative structuring, rather than natural selection, is the fundamental creative force in biology at all levels of hierarchy.

Abstract: This paper presents a distributed AI training system that pools GPU high-bandwidth memory, host DRAM, and SSD into a coordinated parameter-serving hierarchy to support multiterabyte, sparsity-dominated deep models without sharing raw features across machines. The design shards and caches only the working parameters in GPU memory via multi-GPU hash tables, communicates intra-node over NVLink, and performs inter-node synchronization using RDMA-backed collective updates to preserve convergence under data parallelism. A four-stage pipeline overlaps network transfers, SSD I/O, CPU partitioning, and GPU compute while file-level compaction mitigates I/O amplification, yielding high throughput without inflating latency at scale. On industrial click-through-rate workloads with multi-terabyte embeddings, the system outperforms a large in-memory CPU cluster while maintaining production-grade accuracy, improving both training speed and price-performance for distributed AI. Overall, the architecture offers a pragmatic blueprint for scaling distributed learning through memory-hierarchy co-design and communication-aware parameter serving rather than brute-force cluster expansion.

Abstract: Amaranth (R2) is used as a color additive in cosmetics. In Korea, R2 is permitted only as a cosmetic colorant and is prohibited in products intended for infants and children under 13 years of age; in Europe, it is regulated solely as a cosmetic colorant rather than a hair dye ingredient. Despite its regulatory relevance, dermal absorption data for R2 are lacking. In this study, percutaneous absorption of R2 was evaluated using the Franz diffusion method with excised rat dorsal skin. Quantitative analysis of R2 was developed and validated using high-performance liquid chromatography (HPLC) in accordance with Korean Ministry of Food and Drug Safety guidelines, demonstrating acceptable linearity (r² = 0.9996–0.9999), accuracy (95.5–104.4%), and precision (0.3–5.8%). Two formulations (skin lotion and cream), each containing 1% R2, were applied at 113 mg/cm² for 24 h. Dermal absorption was assessed by analyzing receptor fluid, skin wash, stratum corneum, epidermis, and dermis. Total dermal absorption of R2 was 3.4 ± 2.7% for the lotion and 0% for the cream, corresponding to in vitro skin permeabilities of 34.5 ± 27.0 μg/cm² and 0 μg/cm², respectively. Total recovery ranged from 80.3 ± 8.2% to 91.4 ± 19.4%. These results provide essential data for cosmetic risk assessment of R2.

Abstract:

This study examines how different programming paradigms are associated with learning experiences and cognitive challenges as encountered by non-computer science novice learners. Using a case-study approach situated within specific instructional contexts, we integrate survey data from undergraduate students with large-scale public question-and-answer data from Stack Overflow to explore paradigm-related difficulty patterns. Four instructional contexts—C, Java, Python, and Prolog—were examined as pedagogical instantiations of imperative, object-oriented, functional-style, and logic-based paradigms using text clustering, word embedding models, and interaction-informed complexity metrics. The analysis identifies distinct patterns of learning challenges across paradigmatic contexts, including difficulties related to low-level memory management in C-based instruction, abstraction and design reasoning in object-oriented contexts, inference-driven reasoning in Prolog-based instruction, and recursion-related challenges in functional-style programming tasks. Survey responses exhibit tendencies that are broadly consistent with patterns observed in public Q&A data, supporting the use of large-scale community-generated content as a complementary source for learner-centered educational analysis. Based on these findings, the study discusses paradigm-aware instructional implications for programming education tailored to non-major learners within comparable educational settings. The results provide empirical support for differentiated instructional approaches and offer evidence-informed insights relevant to curriculum-oriented teaching and future research on adaptive learning systems.

Abstract: As the reliability and validity of forensic evidence, particularly in feature comparison disciplines, confront on-going scrutiny, forensic practitioners must ensure their processes, whether for investigative, intelligence or evidential purposes are robust, scientifically grounded, and validated. In forensic facial identification, morphological analysis is internationally recognized as the preferred method for facial image comparison, and is applied during the analysis and comparison steps of the Analysis, Comparison, Evaluation, Verification (ACE-V) process, commonly applied in feature comparison. While several international proficiency tests have assessed forensic facial examiners’ accuracy in comparing mated and non-mated pairs (black box tests), fewer opportunities have focused on evaluating inter-laboratory procedures and methods. To address this gap, members of a small border and immigration focused expert working group participated in an inter-laboratory collaborative exercise designed to analyse and harmonize best practices across member laboratories. There are limited published validation studies of facial image comparison methods. This paper presents the results of a collaborative exercise that compares the methodologies of three different agencies, highlighting key similarities and differences in examiner process and decision making, and provides a foundation for the development of similar future initiatives.

Abstract:

Knowledge about the germination potential of Mandacaru seeds is fundamental for maintaining breeding programs and germplasm banks. Thus, we aimed to study the germination of stored and freshly harvested mandacaru seeds in order to investigate seed viability as a function of storage imposition, in addition to characterizing seed anatomy and conducting biochemical evaluation. Germination tests were conducted in a completely randomized design in a 2×6 factorial scheme, with two storage conditions and six temperatures (15, 20, 25, 30, 35, and 40°C), with 4 replications of 25 seeds each. Anatomical evaluation tests and biochemical tests had 5 and 10 replications for each storage condition, respectively. It is concluded that the range of 25-35°C is ideal for germination of C. jamacaru seeds, and temperatures below 20°C and above 35°C are detrimental to germination. X-ray computed microtomography was efficient for characterizing seed anatomy and differentiating their tissues, allowing accurate and clear evaluation of their internal structures, and proper storage was efficient in minimizing the deleterious effects of H₂O₂ and MDA accumulation.

Abstract: Trypanosoma vivax is a hemoparasite of major veterinary importance, causing trypanoso-miasis in domestic and wild ruminants. While cattle are widely recognized as susceptible hosts, water buffaloes are increasingly reported to develop acute, subacute, and chronic infections with severe health and production impacts. This review critically evaluates the current knowledge on T. vivax in buffaloes, focusing on pathogenesis, epidemiology, clin-ical manifestations, diagnostic approaches, therapeutic challenges, and control strategies. Data from Africa and South America are synthesized, with particular emphasis on out-breaks in the Amazon Biome, especially Marajó Island (Brazil), where buffalo farming represents a key economic activity. Advances in molecular diagnosis, such as PCR-based methods, are compared with traditional parasitological and serological tools, and their applicability in field conditions is discussed. Current chemotherapeutic options, emerging reports of drug resistance, and perspectives for vaccine development are examined. In ad-dition, integrated control strategies considering mechanical vectors, iatrogenic transmis-sion, and biosecurity practices are highlighted. This review identifies critical gaps in re-search and provides practical recommendations for surveillance and disease manage-ment in buffalo herds. The information presented here aims to support veterinarians, re-searchers, and policymakers in designing sustainable strategies to mitigate the impact of T. vivax in tropical livestock systems.

Abstract: Background: Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disorder globally. Mazdutide has shown clinical benefits in weight management and metabolic regulation, indicating its potential as a therapeutic agent for NAFLD. This study aimed to investigate the efficacy and mechanism of action of Mazdutide against early-stage NAFLD. Methods: A NAFLD mouse model was induced by a 12-week high-fat diet, followed by a 4-week treatment with subcutaneous Mazdutide (100, 200, or 400 μg/kg). In vitro, a cellular NAFLD model was established by treating hepatocytes with 1 mM free fatty acids for 24 h, followed by co-treatment with Mazdutide (10, 20, or 50 nM) or the endoplasmic reticulum (ER) stress inhibitor 4-phenylbutyric acid (4-PBA). Serum and hepatic lipid profiles, liver injury markers, and pro-inflammatory cytokines were quantified. Liver histopathology was assessed by hematoxylin and eosin and Oil Red O staining. Protein expression related to ER stress, inflammation, and lipid metabolism was analyzed by immunohistochemistry and Western blot. Results: Mazdutide treatment significantly ameliorated systemic and hepatic lipid metabolism disorders, reduced liver injury markers and hepatic steatosis, and mitigated inflammation and oxidative stress in NAFLD mice and hepatocytes. Mechanistically, Mazdutide alleviated ER stress by modulating the PERK-eIF2α-ATF4-CHOP pathway, suppressed the NF-κB-mediated inflammatory response, and downregulated key lipogenic regulators, including SREBP-1, C/EBPβ, and PPARγ. Conclusion: Our findings demonstrate that Mazdutide alleviates hepatic ER stress in NAFLD, leading to suppressed inflammatory responses and improved lipid metabolism, which ultimately attenuates disease progression.

Abstract: Gene expression is inherently stochastic, and promoter switching–induced transcriptional bursting generates substantial cell-to-cell variability in mRNA abundance. Such variability is commonly characterized by the mean and variance; however, these low-order statistics fail to capture the geometric features of mRNA copy number distributions and may obscure mechanistic differences in promoter dynamics. In this work, we analyze a two-state stochastic gene transcription model and derive explicit analytical expressions for higher-order moments of mRNA abundance. We show that skewness and kurtosis provide mechanistically informative signatures of transcriptional bursting, explicitly depending on promoter switching kinetics and burst size. In particular, positive skewness increases with slower promoter switching and larger burst sizes, even when the mean expression level is fixed, while elevated kurtosis distinguishes burst-dominated, low-expression regimes from Gaussian-like high-expression regimes. Our results demonstrate that distinct promoter dynamics can produce identical mean expression levels and variances while exhibiting markedly different skewness and kurtosis. Incorporating higher-order statistics, therefore, extends conventional mean–variance analyses and enables improved discrimination between competing stochastic gene expression mechanisms in single-cell data.

Abstract: The global health crisis driven by Antimicrobial Resistance (AMR) necessitates an urgent pivot toward novel therapeutic agents, with traditional medicinal plants serving as a critical resource. The Asteraceae genus Verbesina, particularly utilized in Mexican ethnobotany, has garnered scientific attention due to its potent bioactive profile against infection and inflammation. This review provides a comprehensive and critical synthesis of the pharmacological landscape of Verbesina species, focusing specifically on the dual role of its major secondary metabolites, the Sesquiterpene Lactones (SLs), as both cytotoxic and antimicrobial agents. We systematically compile and analyze reported in vitro data, including IC50 values from cancer and non-cancerous cell lines, and MIC values against clinically relevant drug-resistant strains like S. aureus and E. coli. A core focus is placed on establishing the therapeutic index (SI = IC50/MIC) for lead compounds, providing a crucial indicator of drug feasibility. Furthermore, we review the proposed molecular mechanisms of SL action, such as the crucial role of the α-methylene-γ-lactone moiety in alkylating cellular targets, which underpins both their antiproliferative and bactericidal effects. By critically bridging ethnopharmacology with modern mechanistic data, this review validates the translational potential of Verbesina metabolites and highlights clear directions for bioassay-guided isolation and optimization as next-generation anti-resistance scaffolds.

Abstract: Background. In actinic keratosis (AK), clinical clearance after field-directed therapies does not necessarily correspond to histological resolution, resulting in subclinical persistence and risk of recurrence. Objective. To provide a practical, up-to-date framework for non-invasive monitoring of treatment response in AK, integrating clinical assessment and dermoscopy with high-resolution imaging techniques, reflectance confocal microscopy (RCM), line-field confocal optical coherence tomography (LC-OCT), and high-frequency ultrasound (HFUS), and to discuss emerging optical biomarkers based on Raman spectroscopy. Results. For each modality, we summarize pre- and post-treatment imaging patterns, proposed response criteria, recommended follow-up timing, and correlations with clinical outcomes (including clearance and AKASI) and, when available, histological findings. The available evidence is derived from a limited number of observational studies, predominantly involving RCM and LC-OCT, whereas data on HFUS and Raman spectroscopy remain comparatively scarce. RCM and LC-OCT allow in vivo assessment of epidermal architectural normalization and reduction of intraepidermal keratinocyte atypia. HFUS captures quantitative trajectories of superficial dermal remodeling, including changes in the subepidermal low-echogenic band (SLEB) and dermal echogenicity after photodynamic therapy and other field treatments. Dermoscopy remains the first-line tool for routine follow-up but may fail to detect minimal subclinical persistence. Finally, we discuss the potential role of in vivo Raman spectroscopy for dynamic molecular endpoints and its possible integration with artificial intelligence–based analytical approaches. Conclusions. A standardized multimodal follow-up strategy improves the accuracy of treatment-response assessment compared with clinical evaluation alone. We propose a technique-specific checklist of minimal response criteria and a pragmatic temporal assessment scheme, and outline a research roadmap to support validation and clinical implementation of non-invasive imaging–guided monitoring in actinic keratosis.

Abstract: Dyslipidemia is a prevalent metabolic disorder and a major cardiovascular risk factor, often influenced by sedentary lifestyles and family history. This study analyzed the association between sedentary behavior, adiposity indicators, and family history with dyslipidemia in young adults from Trujillo in 2025. A cross-sectional correlational design was applied to 137 participants (41 men and 96 women). Sedentary behavior was measured using the IPAQ questionnaire, while family history of dyslipidemia or cardiovascular disease was recorded. Anthropometric indicators included waist circumference (WC), body mass index (BMI), relative fat mass (RFM), body roundness index (BRI), and conicity index (CI). Lipid profiles were assessed with Mission monitoring equipment. HDL-c was low (38.90 ± 16.45 mg/dL in men; 47.42±15.82 mg/dL in women), while LDL-c was slightly elevated (103.39 ± 36.36 mg/dL and 102.74±33.60 mg/dL). Average cholesterol and triglyceride concentrations were normal in both genders. WC, RFM, and BRI correlated with LDL-c, with RFM showing the strongest association (OR = 4.108; 95% CI: 1.266–13.332). Triglycerides were linked to BMI, WC, BRI, and sedentary lifestyle, with WC being most significant (OR = 6.125; 95% CI: 2.007–18.690). In conclusion, RFM and WC emerged as the most robust predictors of dyslipidemia, underscoring their utility for early detection of elevated LDL-c and hypertriglyceridemia in young populations.

Abstract: Neural network interpretability methods have produced powerful approaches for gradient-based feature attribution, energy landscape analysis, and uncertainty quantification. State-of-the-art methods rely on structural weight analysis, Monte Carlo dropout uncertainty estimation, and attention mechanisms for interpretable predictions with quantified confidence. However, existing methods face fundamental challenges: unreliable explanations with poor uncertainty quantification on complex medical imaging tasks, difficulty identifying important network weights due to fixed thresholds, and computational overhead from attention mechanisms operating without uncertainty guidance. We introduce Adaptive Uncertainty-Guided Interpretable Networks (AUGIN), a framework combining adaptive structural weight analysis, uncertainty-aware prediction intervals, and uncertainty-guided attention through three interconnected modules. Our approach models adaptive threshold computation evolving with layer depth and architecture, integrates feature-level uncertainty into prediction interval generation, and introduces uncertainty-guided attention focusing on uncertain regions. Experiments on ISLES stroke prediction, BraTS brain tumor segmentation, and CT-CTA thrombectomy outcome prediction demonstrate superior performance: 0.7891 AUC-ROC on ISLES (1.59 percentage point improvement), 0.8567 Dice score on BraTS, and 0.7456 attention IoU, while maintaining excellent uncertainty calibration (0.8123) and explanation fidelity (0.8234). Our work advances neural network interpretability by providing accurate predictions with reliable, uncertainty-quantified explanations for safety-critical medical imaging applications.

Abstract:

An archaeological exposure near Hachum, featuring a Ditch profile interpreted as part of a Neolithic earthwork, was characterized using DNA analyses of bacterial 16S rRNA from soil samples. The results showed that the middle and lower parts of the Ditch fill could be clearly distinguished from each other and from the surrounding area based on the composition of soil bacterial DNA. Genera detected predominantly in the lower part of the Ditch suggest that, after the Ditch was completed, organic matter, animal dung, and possibly even human feces were accumulated at the bottom. The investigations demonstrate that analyses of soil bacterial communities can provide valuable insights into the history and function of a Neolithic earthwork and, more generally, represent an important additional source of information for interpreting archaeological contexts that are devoid of or poor in finds.

Abstract:

Background: This study aimed to identify sociodemographic and occupational factors associated with facing moral dilemmas among workers at the Federal University of Pelotas-RS Teaching Hospital who worked on-site during the pandemic. Methods: A cross-sectional study was conducted in 2020 with all workers, including health professionals, support staff, and administrative personnel. Questions about moral dilemmas were grouped into two outcomes: witnessing behaviors or attitudes, and feeling pressured to act in disagreement with what they believed was right. Associations were estimated using Poisson regression with robust variance, based on a hierarchical model. Results: A total of 1,158 workers participated, most of whom were women (76.1%). The prevalence of moral dilemmas was 44% for witnessing and 15% for feeling pressured. Younger age, higher education, being a resident, working both day and night shifts, lack of PPE, and having an active or high-strain job were positively associated with both types of dilemmas, whereas the availability of social support and adequate resting areas reduced their occurrence. Conclusions: Reducing moral dilemmas requires promoting democratic leadership, ensuring adequate staffing, strengthening professional autonomy, encouraging social support, and creating rest spaces. These arrangements are essential for promoting workers’ psychological well-being.

Abstract:

Intraoperative assessment of pancreatic quality, followed by sampling for the potential isolation of Langerhans islets for subsequent autotransplantation, is currently a key component of post-total pancreatectomy diabetes mellitus treatment. The aim of this study was to quantitatively evaluate pancreatic parenchymal stiffness using optical coherence elastography (OCE) imaging, and to investigate the utility of the OCE method as a potential indicator of islet yield after pancreatectomy. A total of 41 freshly excised human pancreatic specimens, containing pancreatic ductal adenocarcinoma (PDAC) and surrounding non-tumorous tissues post-pancreatectomy, were studied. In this research, the stiffness (Young’s modulus, kPa) and its color-coded 2D distribution were calculated for various pancreatic samples using compression OCE. Stiffness values were compared between intact pancreatic parenchyma (islet-poor and islet-rich) and pancreatic lesion groups (parenchymal fibrosis and/or PDAC invasion). The data were confirmed by histological analysis. In addition, the measured stiffness values for various morphological groups of the pancreatic samples were compared with the number of isolated islets obtained from pancreatic samples after collagenase treatment. The study demonstrated that OCE can effectively distinguish areas of pancreatic lesions and identify intact pancreatic parenchyma containing Langerhans islets. A highly significant increase in mean stiffness (p<0.0001) was observed in postoperative pancreatic samples exhibiting signs of parenchymal fibrosis or PDAC invasion compared to unaffected, intact pancreatic parenchyma. For the first time, a relationship between stiffness values and the number of isolated pancreatic islets was demonstrated, in particular, the number of isolated islets significantly decreased (≤110 pcs/g) in samples exhibiting stiffness values above 150 kPa and below 75 kPa. The optimal stiffness range for the efficient isolation of islets (≥120 pcs/g) from pancreatic tissue was identified as 75–150 kPa. The study introduces a novel approach for rapid and objective intraoperative assessment of pancreatic tissue quality using real-time OCE data. This technique facilitates the identification of regions affected by pancreatic lesions and supports the selection of intact pancreatic parenchyma, potentially enhancing the accuracy of Langerhans islet yield predictions during surgical resection.

Abstract: Natural Language Processing (NLP) is rapidly transforming research methodologies across disciplines, yet African languages remain largely underrepresented in this technological shift. This paper provides the first comprehensive overview of NLP progress and challenges for the six national languages officially recognized by the Senegalese Constitution: Wolof, Pulaar, Sérère, Diola, Mandingue, and Soninké. We synthesize linguistic, sociotechnical, and infrastructural factors that shape their digital readiness and identify gaps in data, tools, and benchmarks. Building on existing initiatives and research works, we analyze ongoing efforts in text normalization, machine translation, and speech processing. We also provide a centralized GitHub repository that compiles publicly accessible resources for a range of NLP tasks across these languages, designed to facilitate collaboration and reproducibility. A special focus is devoted to the application of NLP to the social sciences, where multilingual transcription, translation, and retrieval pipelines can significantly enhance the efficiency and inclusiveness of field research. The paper concludes by outlining a roadmap toward sustainable, community-centered NLP ecosystems for Senegalese languages, emphasizing ethical data governance, open resources, and interdisciplinary collaboration.