Search Results (2,479)

5Cr-0.5Mo ferritic steels are widely used in high-temperature power-plant components. Although artificial neural network (ANN) models have shown good performance in predicting tensile properties, they provide limited insight into predictions and generally do not quantify the uncertainty. In this study, three tree-based machine learning models—Random Forest (RF), XGBoost (XGB), and Gradient Boosting (GB)—were developed using 36 unique alloy grade–temperature observations from a validated NIMS 5Cr-0.5Mo tensile dataset. The model performance was evaluated using leave-one-grade-out (LOGO) cross-validation, with pooled out-of-fold (OOF) predictions used to assess the overall performance. SHapley Additive exPlanations (SHAP) were used to examine feature contributions, whereas Gaussian Process Regression (GPR) was evaluated as a proof-of-concept for uncertainty quantification of yield strength (YS). GB showed the strongest performance for ultimate tensile strength (UTS) and reduction in area (RA), achieving pooled OOF R² values of 0.9698 and 0.9570, respectively. RF achieved corresponding R² values of 0.9406 and 0.9488, respectively. SHAP identified the test temperature as the most influential feature across all four properties, whereas the Cr content and austenite grain size contributed significantly to the strength predictions. For YS, the GPR achieved complete empirical coverage of the 95% predictive intervals, although the relatively large mean interval width indicated conservative uncertainty estimates. Given the limited dataset and feature correlations, the SHAP results should be regarded as exploratory, rather than mechanistic. Overall, this study demonstrates the potential of interpretable, uncertainty-aware ML for small alloy datasets, while emphasizing the need for larger, compositionally diverse datasets and independent validation. Full article

DNA methylation is implicated in epileptogenesis. Sarcosine, a glycine transporter 1 (GlyT1) inhibitor and methyl donor, attenuates behavioral progression during rapid hippocampal kindling and alters hippocampal DNA methylation, but its locus-specific epigenetic effects remain poorly understood. Here, reduced representation bisulfite sequencing (RRBS) was combined with targeted gene expression analysis in the hippocampi of sarcosine-treated kindled rats. RRBS identified 563, 533, and 390 differentially methylated regions (DMRs), corresponding to 521, 499, and 374 DMR-associated genes, in vehicle-kindled versus sham (vKD vs. vSH), sarcosine-kindled versus sham (sKD vs. vSH), and sarcosine-kindled versus vehicle-kindled (sKD vs. vKD) comparisons, respectively. Pathway enrichment analysis identified 217 significantly affected pathways, including glutamatergic signaling, extracellular matrix (ECM) organization, chromatin regulation, axon guidance, and apoptotic processes. Eleven candidate genes involved in epigenetic regulation, excitatory neurotransmission, and ECM remodeling were selected for transcriptional validation. All 11 genes were significantly upregulated in kindled hippocampi, whereas sarcosine was associated with reduced expression relative to vehicle-kindled rats for eight genes (Hdac9, Fos, Smad7, Unc5a, Grik2, Gpr37l1, Cacna2d2, and Yy1). Collectively, these findings indicate that sarcosine remodels DNA methylation-associated transcriptional networks during rapid hippocampal kindling and support GlyT1 inhibition as a potential disease-modifying approach in experimental epileptogenesis. Full article

Rapid surrogate prediction of ducted propeller performance is challenging when only a limited number of independent geometries are available and operating points belonging to the same geometry are strongly correlated. This study proposes a sensitivity-informed physics-regularized backpropagation neural network (SIPR-BP) for simultaneous prediction of the thrust coefficient K_T and the scaled torque coefficient 10K_Q. A CFD database comprising 20 Ka4-70-derived parameterized geometries, each evaluated at five advance ratios, provides 100 observations and 20 complete performance curves. The framework combines three main strategies. First, two-component multi-output partial least-squares (PLS) curve surrogates are reconstructed exclusively from the training geometries of each outer fold to generate leakage-controlled conditional Sobol gate priors. Second, the operating coordinate J is separated from geometric gating and represented by five ordered curve nodes, which guarantee non-increasing K_T and 10K_Q responses over the investigated interval. Third, training-only physics-consistency reliability weighting and a three-member ensemble improve robustness to locally irregular CFD responses and initialization variability. Under a ten-round geometry-grouped holdout protocol, SIPR-BP achieves a geometry-balanced MAPE of 2.65%, RMSE of 0.0112, MAE of 0.00911, and pooled R² of 0.951. When evaluated under the same outer partitions, a two-component PLS baseline yields a MAPE of 4.38%. Across the evaluated PLS, Extra Trees, GPR, and SVR baselines, SIPR-BP reduces geometry-balanced MAPE by approximately 39.6–79.2%. The results indicate that the proposed framework improves unseen-geometry prediction while preserving the prescribed response-curve structure. Full article

Background: Periodontitis is a chronic inflammatory disease characterized by dysregulated host immune responses that drive connective tissue destruction and alveolar bone loss. Human gingival fibroblasts (HGFs) are central regulators of periodontal inflammation through their production of cytokines, chemokines, matrix-remodeling enzymes, and other inflammatory mediators. Although cannabinoid receptor 2 (CB2) activation has demonstrated anti-inflammatory properties, its coordinated effects on multiple inflammatory pathways in gingival fibroblasts remain poorly understood. This study investigated the transcriptomic effects of the selective CB2 agonist HU-308 on IL-1β-induced inflammatory responses in HGFs. Methods: Primary HGFs were divided into untreated controls, IL-1β-stimulated cells (10 ng/mL), and IL-1β-stimulated cells treated with HU-308 (10 μM). Twenty-four hours after stimulation, transcriptome-wide expression profiling was performed using Affymetrix Human Clariom S microarrays, followed by targeted analysis of selected inflammation-related transcriptional domains. Selected transcripts were evaluated within predefined biological domains including cytokines, chemokines, extracellular matrix-associated genes, NO/cGMP-related genes, transporter-associated genes, and GPCR-related transcripts. Gene expression was analyzed using one-way ANOVA with Tukey’s post hoc test. Results: IL-1β induced a coordinated inflammatory transcriptional program characterized by increased expression of pro-inflammatory cytokines, chemokines, matrix metalloproteinases, glucose transporter genes, and multiple GPCR-related transcripts, while suppressing collagen-associated genes, NOS3, GPR4, and GPR78. HU-308 broadly attenuated these inflammatory responses by reducing the expression of cytokines, chemokines, matrix metalloproteinases, and several GPCR-related genes while restoring collagen-associated transcripts, nitric oxide signaling components, anti-inflammatory mediators, and selected glucose transporters toward basal levels. Schematic multidimensional visualizations were used to illustrate relative expression patterns among selected transcripts within each functional domain; these visualizations do not represent statistically derived gene networks or molecular interactions. Conclusions: Pharmacological modulation of CB2 by HU-308 exerts broad immunomodulatory effects in IL-1β-stimulated human gingival fibroblasts by coordinately regulating multiple transcriptional networks involved in periodontal inflammation. These findings demonstrate that HU-308 treatment is associated with coordinated modulation of inflammatory, extracellular matrix, nitric oxide, metabolic, and GPCR-associated transcriptional pathways in IL-1β-stimulated HGFs. The results support the hypothesis that CB2 signaling may participate in the broader regulation of these interconnected responses; however, receptor-specific studies using CB2 antagonism or CNR2 knockdown are required to establish causality. Full article

Quantitative determination of gangue content is important for efficient coal use and intelligent coal–gangue separation. We combine transmission terahertz time-domain spectroscopy (THz-TDS), multidomain feature fusion, and machine learning to predict gangue mass fraction in coal–gangue mixtures. Time- and frequency-domain signals, refractive index, absorption and extinction coefficients, and complex permittivity were extracted from samples with different gangue contents. Five-fold cross-validation was used to compare random forest, support vector regression, Gaussian process regression, an artificial neural network, and an Effective Medium Theory-constrained Physics-Informed Neural Network (EMT-PINN). EMT-PINN achieved the best performance, with a coefficient of determination (R²) of 0.81 ± 0.15, a mean absolute error (MAE) 3.17 ± 0.59%, and a root mean square error (RMSE) of 5.79 ± 0.21%, compared with R² values of 0.72 ± 0.08, 0.61 ± 0.21, 0.74 ± 0.11, and 0.64 ± 0.18 for RF, SVR, GPR, and ANN, respectively. These results demonstrate the potential of physics-informed THz spectroscopy for rapid and physically interpretable quantitative characterization of coal–gangue mixtures. Full article

Financial networks are typically summarised by a single average-regime connectedness estimate that treats transmission as symmetric across calm and turbulent markets. Using a quantile vector autoregression on nine European equity indices from 2000 to 2026, we show that crash-regime connectedness is not an episodic crisis response but a persistent premium over the normal regime, holding steady across nearly six thousand rolling windows. We introduce Geopolitical Risk Realised Volatility, a within-month measure of geopolitical risk dispersion distinct from its level, and show that it predicts a delayed, statistically robust decoupling of tail connectedness, modest in magnitude and specific to the crash regime, that adds information beyond GPR Act’s level alone. A quantile-specific structural break test shows that the Brexit referendum permanently shifted the United Kingdom’s net shock-transmission position within the European equity network. These shocks affect connectedness only in the crash regime, a pattern an average-regime estimate does not capture. Full article

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection. Its development and progression involve multiple interconnected mechanisms, including uncontrolled inflammation, immunosuppression, metabolic reprogramming, intestinal barrier disruption, and multi-organ injury. Short-chain fatty acids (SCFAs), primarily acetate, propionate, and butyrate, are important metabolites produced by the anaerobic fermentation of dietary fiber and indigestible carbohydrates by gut microbiota. During sepsis, antibiotic exposure, intestinal hypoperfusion, insufficient nutritional substrates, and microbial dysbiosis may deplete SCFA-producing bacteria and lower SCFA levels, thereby aggravating intestinal barrier dysfunction, endotoxin translocation, and systemic inflammatory responses. SCFAs can influence sepsis-associated intestinal, pulmonary, cardiac, hepatic, renal, and cerebral injury by activating receptors such as free fatty acid receptor 2 (FFAR2)/G protein-coupled receptor 43 (GPR43), free fatty acid receptor 3 (FFAR3)/G protein-coupled receptor 41 (GPR41), and G protein-coupled receptor 109A (GPR109A); inhibiting histone deacetylases; and regulating immune-cell metabolism, inflammasome activation, oxidative stress, mitochondrial function, and modes of cell death. In recent years, strategies such as direct SCFA supplementation, promotion of endogenous SCFA production, restoration of SCFA-producing microbial communities, and targeting of SCFA receptors and downstream signaling pathways have shown therapeutic potential. However, their clinical translation remains limited by uncertainties regarding dose, timing, route of administration, patient stratification, and safety. This review systematically summarizes the mechanisms of action and therapeutic advances of SCFAs in sepsis, aiming to provide a reference for microbiome-based interventions and metabolism-targeted therapies in sepsis. Full article

Antenna–surface coupling is a key factor controlling the quality of Ground Penetrating Radar (GPR) data, governing the efficiency of electromagnetic energy transmission into the investigated medium. In cultural heritage applications, however, direct antenna contact is often not feasible due to the fragility of decorated surfaces, requiring non-contact configurations whose impact on high-resolution imaging remains insufficiently quantified. This study investigates the effect of antenna coupling on high-resolution 3D GPR imaging of concealed masonry at Pompeii through electromagnetic simulations and a controlled in situ comparison of contact and non-contact acquisitions on a plastered wall. The experimental campaign was conducted at the House of the Red Walls (VIII, 5, 37) on an opus mixtum masonry, selected as the most geometrically and electromagnetically challenging test case among regular Pompeian construction techniques. A 3 GHz antenna was employed, and three acquisition configurations were analysed: direct contact, and non-contact setups with antenna elevations of 3 cm and 5 cm. The results show that even moderate antenna elevation significantly reduces coupling efficiency at the air–plaster interface, leading to a progressive degradation of imaging performance. While the overall masonry arrangement remains recoverable in all configurations, increasing stand-off distance reduces the detectability of individual units and degrades geometric accuracy, with vertical mortar joints being the most affected elements. These findings demonstrate that, given the combined electromagnetic and geometric characteristics of the construction materials used at Pompeii, near-contact GPR acquisition is required for reliable imaging of masonry arrangement under the investigated conditions, highlighting the critical role of antenna coupling in high-frequency GPR surveys of fragile architectural surfaces. Full article

High-intensity interval training (HIIT) is widely recognized as an effective strategy for improving cardiorespiratory fitness and metabolic health. Beyond these physiological benefits, growing evidence indicates that HIIT may also induce beneficial immunometabolic adaptations. A key exercise-responsive metabolite in this context is lactate, which is increasingly being recognized not as a metabolic waste product but as a bioactive signaling metabolite capable of coordinating metabolic, inflammatory, and immune processes. This narrative review examines current evidence suggesting a potential role for exercise-induced lactate in immune responses associated with HIIT. We summarize the molecular pathways through which lactate may interact with immune cells, including uptake via monocarboxylate transporters (MCT1/MCT4) and SLC5A12, receptor-dependent signaling through GPR81/HCAR1, and epigenetic regulation via histone lactylation. We further discuss the cell-specific effects of lactate on macrophages, dendritic cells, neutrophils, and T lymphocytes, highlighting how these mechanisms may influence immune-cell metabolism, inflammatory regulation, and functional remodeling. A central concept emerging from the current literature is that the biological actions of lactate are highly dependent on the kinetics, duration, and physiological context of exposure. Unlike pathological lactate elevations observed in conditions such as cancer, sepsis, or mitochondrial myopathies—the latter potentially involving an exaggerated lactate response during exercise due to impaired oxidative metabolism—HIIT generates transient systemic lactate elevations as part of a coordinated neuroendocrine and metabolic response. When combined with adequate recovery, these repeated metabolic perturbations may promote hormetic adaptations characterized by improved inflammatory regulation, enhanced immune resilience, and more efficient immunometabolic homeostasis. Conversely, excessive training loads or inadequate recovery may shift these responses toward maladaptive immune stress. Overall, current evidence suggests a paradigm shift in exercise immunology in which lactate should be regarded as one component of an integrated immunometabolic signaling network rather than simply as a marker of anaerobic metabolism. Future mechanistic studies integrating lactate kinetics, immune-cell phenotyping, transporter expression, and lactate-dependent post-translational modifications are needed to clarify the extent to which lactate may contribute to exercise-induced immune remodeling and to guide the development of immunologically informed HIIT protocols. Full article

Meso-scale ultrasonic vibration-assisted end milling (UVAEM) is an advanced manufacturing technique. UVAEM of Inconel 718 superalloy presents significant modeling challenges due to complex thermo-mechanical interactions and meso-scale size effects. This study introduces a machine learning framework that systematically compares Gaussian Process Regression (GPR), Support Vector Regression (SVR), Random Forest (RF), and Ridge Regression for predicting cutting force, tool wear, and surface roughness, thereby contributing to sustainable manufacturing. Experiments followed a Taguchi L16 orthogonal array with two replicates (n = 32), varying cutting speed (10–40 m/min), feed rate (0.01–0.025 mm/tooth), depth of cut (0.10–0.25 mm), vibration amplitude (0–9 μm), and tool coating (TiAlN, TiSiN, nACo, Uncoated). Tool coating was one-hot encoded, and strict leave-one-out cross-validation (LOOCV) with within-fold standardization ensured unbiased generalization metrics. Following nested hyperparameter tuning, SVR achieved the highest accuracy (R² = 0.9552, 0.9473, 0.9294), marginally outperforming GPR (R² = 0.9543, 0.9420, 0.9290). Ridge regression was competitive for tool wear (R² = 0.9235), while random forest ranked last due to limited ensemble diversity at n = 32. Pearson correlation identified depth of cut as the dominant driver of cutting force and surface roughness (r = 0.767, 0.759), cutting speed as the primary driver of tool wear (r = 0.663), and vibration amplitude as consistently beneficial across all three responses. SVR and GPR are recommended as reliable surrogate models for process optimization in UVAEM of Inconel 718. Full article

Commodity futures markets exhibit pronounced non-stationarity, nonlinearity, and multifractal characteristics that challenge traditional linear models. We employ a multiscale framework integrating four methodologies—MF-DCCA, PG irreversibility index, MSWPE, and JS-divergence segmentation—to analyze these features using daily closing prices for WTI crude oil, agricultural commodities (US soybeans, meal, oil, and wheat), the US dollar index, and Chinese No. 2 soybeans, spanning from 2 January 2018 to 1 October 2025, sourced from Investing.com and Matteo Iacoviello’s GPR database. The analysis yields three key results. First, scale dependence varies across commodities and is shaped by supply adjustment elasticity: energy markets show scale-dependent amplification and directional sign reversals, while agricultural markets maintain near-monofractal structures. Second, multiple methods converge on a characteristic time scale of approximately 20 days, linking physical logistics rhythms with financial pricing. Third, the persistence of structural reconstruction after shocks depends on systemic penetration depth, with exogenous macroeconomic uncertainty exerting stronger and more lasting effects than market-internal events. Together, supply elasticity, physical logistics rhythms, and systemic penetration depth constitute the three fundamental determinants of nonlinear commodity futures dynamics, with implications for cross-commodity allocation, multi-horizon risk management, and geopolitical scenario analysis. Full article

Machining-induced residual stress, cutting force, and temperature govern the fatigue life, dimensional stability, and surface integrity of milled components, yet predictive models for these responses are still routinely validated only in-sample, concealing overfitting on small, single-laboratory datasets. This study re-examines a published AISI 1045 end-milling dataset (N = 24, combining one-factor-at-a-time and Taguchi L9 trials) using six regression paradigms: Multiple Linear Regression (MLR), random forest, gradient boosting, Support Vector Regression (SVR), Gaussian process regression (GPR), and a shallow neural network (ANN)—under leave-one-out cross-validation (LOO-CV). The previously reported in-sample R² of 0.84 (from a smaller n = 9 subset) was substantially higher than the LOO-CV R² of 0.167 obtained here on the full dataset; although this gap cannot be attributed to cross-validation alone, it shows a substantial in-sample/out-of-sample performance gap. SVR gave the strongest, bootstrap- and nested-CV-confirmed cross-validated residual-stress prediction (R² = 0.575); its apparent force advantage (R² = 0.558) was statistically indistinguishable from GPR and did not survive nested tuning, so it is reported cautiously. GPR was narrowly best for temperature (R² = 0.492); the ANN and SVR underperformed the linear baseline there, though nested tuning traced this largely to a fixed hyperparameter rather than the kernel method itself. Random forest permutation importance identified feed rate as the dominant residual-stress predictor, consistent with the original ANOVA. The contribution is a cross-validated, multi-paradigm reassessment with explicit uncertainty and sensitivity analysis, together with a candidate low-cost screening surrogate for AISI 1045 process planning—not a replacement for XRD or FE—and a broader caution to match model complexity to sample size. Full article

Carcass traits, which reflect growth performance and muscle development, are economically important in beef cattle, yet their genetic determinants remain poorly characterized. Both single-population Genome-wide association studies (GWAS) methods, such as BLINK, and cross-population meta-analysis approaches are widely used to identify genetic variants, yet their comparative performance in genomic prediction for complex traits in structured populations remains underexplored. Few studies have directly compared these methods in genomic prediction. To address this gap, this study aims to (i) identify positional candidate genes associated with carcass traits and (ii) evaluate the context-dependent advantages of Covariate Adjustment (CA) and meta in genomic prediction. In this study, we analyzed carcass weight (CW), live weight (LW), and dressing percentage (DP) in 279 crossbred Angus cattle genotyped with the PHR0105_Bt140K_v1.0 SNP chip. GWAS was performed on the full population using BLINK, and results from three subpopulations were combined via meta-analysis, with significance thresholds for both approaches determined by a shuffle-based method. Candidate genes located within ±10 kb of significant SNPs were associated with different carcass traits, including STRIT1, SEL1L3, NOC4L and ANK1 for DP; SNCA and DNAH5 for CW; and GYPC, GPR158, and GUCY1A1 for LW. Prediction accuracy under MAS and MABLUP showed meta slightly outperformed BLINK in MAS, while BLINK was better with covariate adjustment; after incorporating kinship in MABLUP, meta achieved higher accuracy and population partitioning was negligible. Overall, MABLUP yielded the highest accuracy (0.52–0.79) versus MAS (0.37–0.54) in all traits. These findings provide a methodological basis for selecting appropriate GWAS strategies in structured populations and highlight candidate genes. Full article

Ground-penetrating radar (GPR) and electrical resistivity tomography (ERT) provide complementary, but non-unique, observations of concrete and masonry conditions. GPR is primarily sensitive to dielectric contrasts, interfaces, reinforcement geometry and electromagnetic attenuation, whereas electrical measurements respond to ionic conduction, moisture state, material connectivity and electrode configuration. This paper presents a structured critical review updated to 21 July 2026. It distinguishes surface or bulk resistivity measurements from electrical resistance tomography around specimens and from multi-electrode geophysical ERT, an important terminological separation that is frequently blurred in the literature. This review’s contribution is a terminological separation of electrical measurement classes, an evidence-coding scheme that distinguishes corroboration from independent validation, and a staged, conditional research agenda built from that scheme; it is not a claim that GPR–ERT integration is routinely sufficient on its own. Verified evidence is synthesised across reinforced concrete, masonry, coastal infrastructure, laboratory calibration, field validation and forward modelling. The review shows that GPR is mature for reinforcement mapping and conditional detection of interfaces and delamination, while resistivity methods are well established for durability-related screening. True ERT can image spatial conductivity changes associated with moisture ingress and cracks, but inversion regularisation, electrode contact, reinforcement and three-dimensional effects limit resolution and quantitative recovery. Integrated GPR–ERT studies now include controlled masonry experiments, heritage structures, a field heritage pier and laboratory calibration on reinforced concrete; therefore, the principal remaining gap is not the absence of integration. It is the shortage of independent destructive field verification, scale-aware transfer rules and uncertainty-calibrated decision thresholds across structural types. A revised evidence matrix and detectability taxonomy show that neither method directly identifies active corrosion or chloride concentration. The paper concludes with a staged research agenda based on co-registration, physics-informed feature extraction, forward-modelled resolution assessment and targeted ground truth. Full article

High Friction Surface Treatment (HFST) relies heavily on wear-resistant aggregates to ensure roadway safety, yet the conventional evaluation of aggregate abrasion resistance is time-consuming and resource-intensive. In this study, a machine learning framework was developed to predict the abrasion-induced angularity evolution of recycled high-alumina aggregates from their initial morphological characteristics, thereby enabling rapid abrasion-resistance screening. Six regression models were compared under leave-one-group-out cross-validation, and an improved whale optimization algorithm (IWOA) was proposed to tune the Gaussian process regression (GPR) model, incorporating five enhancements and a regularized fitness function to restrain overfitting. The models were trained on 42 samples from six aggregates, whose angularity, Form 2D, micro-texture, sphericity, and F:E ratio were measured with the AIMS II device before and after successive abrasion cycles. The IWOA-GPR model achieved the best performance, with an R² of 0.8909, an RMSE of 150.98, an MAE of 120.55, and a MAPE of 4.60%. The SHAP analysis identified the abrasion revolutions, the initial Form 2D, and the initial angularity as the dominant contributors to the worn angularity. Moreover, the early angularity loss after the first 500 revolutions correlated strongly with the measured Los Angeles abrasion value (r = 0.935), which allows the LAA of a candidate aggregate to be estimated after a single abrasion cycle. The proposed framework therefore provides a rapid and reliable tool for screening wear-resistant aggregates for HFST applications and supports the clean utilization of recycled solid wastes in anti-skid pavements. Full article