Search Results (589)

This study presents two independently optimised Extreme Gradient Boosting (XGBoost) regression models, one for predicting corrosion current density ($i_{corr}$) and one for predicting corrosion potential ($E_{corr}$) parameters of carbon steel rebar embedded in mortar and immersed in simulated pore solution. An experimental dataset consisting of 216 measurements was curated from a systematic potentiodynamic scan study covering six chloride contamination levels, two carbonation states (non-carbonated and carbonated), four moisture conditions for mortar (65%, 85%, 95% relative humidity, and submerged), and three conditioning durations for simulated pore solution (36 h, 72 h and 20 days). Hyperparameters of the XGBoost models were optimised using a Bayesian optimisation framework with the Tree-structured Parzen Estimator (TPE) sampler over 300 trials. Model performance was assessed using 5-fold cross-validation and a random 80:20 train–test split. The optimised models achieved cross-validation $R^2$ scores of 0.936 and 0.953 for $i_{corr}$ and $E_{corr}$, respectively. On the hold-out test set, $R^2$ values of 0.933 and 0.945 were obtained with test RMSE values of 0.2 log₁₀(µA/cm²) and 41.9 mV, respectively. The contribution of each input feature to model predictions was quantified and visualised using the SHapley Additive exPlanations (SHAP) methodology. SHAP analysis reveals that chloride content has the highest impact on $i_{corr}$, followed by carbonation state and the low-humidity condition, while for $E_{corr}$, chloride content and the Submerged condition have the greatest impact. An interactive web application was developed using Streamlit, enabling researchers and practitioners to obtain corrosion parameter predictions. The findings provide data-driven insights into the relative importance of environmental factors governing rebar corrosion, with direct implications for the development of accurate corrosion prediction models for reinforced concrete service life assessment. Full article

Background/Objectives: Multimodal data fusion is increasingly applied in neuroinformatics to integrate heterogeneous sources of information. However, the optimal strategies for combining modalities with markedly different dimensionality, scale, and noise characteristics remain unclear. To our knowledge, this is among the first systematic and controlled benchmarks explicitly disentangling the effects of fusion strategy and feature scaling within a unified deep learning framework. Methods: Using data from 747 healthy participants from the Human Connectome Project, we evaluated multiple fusion paradigms—including early fusion, attention-based fusion, subspace-based fusion, and graph-based fusion—within a unified and reproducible framework. Importantly, we assessed how different feature scaling techniques (Standard, Min–Max, and Robust scaling) interact with fusion strategies and influence model performance. Biological sex was used as a controlled benchmark task to focus on methodological insights rather than task-specific optimization. Results: Early feature-level fusion consistently achieved the highest classification performance across all evaluated configurations. In particular, direct concatenation of cognitive and neuroimaging features combined with Standard Scaling yielded the best results (AUC–ROC = 0.96 (0.95–0.96)), outperforming unimodal baselines as well as intermediate and late fusion strategies. Conclusions: This systematic benchmark demonstrates that multimodal deep learning performance in neuroscience is driven primarily by the interaction between fusion strategy and feature scaling rather than by architectural complexity alone. By explicitly disentangling the effects of fusion level and preprocessing within a unified framework, this study provides practical methodological guidance for the design, evaluation, and reproducible deployment of multimodal deep learning models in neuroscience. Full article

Accurate prediction of the air quality index (AQI) is crucial for understanding urban pollution dynamics and protecting public health. This study proposes a dual-branch fusion framework (CL-XGB-Season) to address seasonal heterogeneity in AQI prediction by integrating temporal dynamic features and static patterns. The CNN-LSTM branch captures short-term temporal fluctuations, while a seasonally split XGBoost branch fits long-term static patterns via independent submodels for spring, summer, autumn, and winter. SHAP-based interpretability analysis revealed the dominant drivers across different seasons: the “temperature × O₃” interaction feature plays a key role in summer, characterizing the ozone formation mechanism dominated by photochemical reactions under conditions of high temperature and strong solar radiation; whereas the PM_2.5/PM₁₀ ratio is crucial in winter (where pollution is primarily driven by pollutant accumulation). The dual-branch fusion framework was validated using hourly resolution data from Chongqing for the 2020–2025 period. Results indicate that the framework achieved a prediction accuracy of 0.197 root mean square error (nRMSE) and 0.9611 coefficient of determination (R²) on the test set, outperforming eight ablation variants and five baseline models (ARIMA, Transformer, etc.) in comparative experiments. Ablation studies confirm the necessity of dual branches and seasonal modeling, with the full model reducing nRMSE by 19–63% versus single-model variants. This framework maintains stable seasonal performance and provides actionable insights for targeted air quality management. Full article

Road traffic accidents at intersections pose a persistent challenge in Riyadh, Saudi Arabia, contributing significantly to public health burdens and economic losses. Traditional statistical approaches often fail to capture the complex, non-linear interactions among geometric design, traffic parameters, and accident severity. This study develops a multi-methodological machine learning framework to predict intersection accident severity using the Equivalent Property Damage Only (EPDO) metric. Historical data (2017–2023) from Riyadh Municipality for 150 high-risk intersections were analyzed, incorporating predictors such as service road distance (SRD), U-turn distance (UTD), median width (MW), peak hour volume (PHV), heavy vehicle percentage (HV%), and injury/frequency counts. Six algorithms, i.e., Decision Tree, Random Forest, Gradient Boosting, Support Vector Machine, Linear Regression, and Artificial Neural Network, were compared using a 70/30 train–test split and k-fold cross-validation in this study. The Gradient Boosting model achieved superior performance (R² = 0.89 with MSE = 63.43 and RMSE = 7.96) and was selected for final deployment. SHAP feature importance analysis revealed minor injuries (MIs), serious injuries (SRIs), and fatalities (FAs) as the most important dominant predictors, with geometric factors (UTD, MW) and traffic composition (HV%) providing actionable infrastructure insights. The model ranked intersections and identified the “Jeddah Road with Taif Road” (predicted EPDO = 137.22) as the highest-risk location. Evidence-based recommendations include enforcing the minimum 300 m U-turn buffers with staggering service road exits ≥150 m and restricting heavy vehicles during peak hours. The scalable framework developed in this study supports the data-driven prioritization of safety interventions and aligns with sustainable urban mobility goals and offers transferability to other metropolitan contexts worldwide. Full article

G protein-coupled receptor kinase 5 (GRK5) is an important therapeutic target involving cardiovascular diseases, cancer, and inflammatory disorders. However, the features of its activation as an essential function regulation process have been poorly understood, limiting related drug development. The work utilizes a molecular dynamics simulation coupled with an interpretable machine learning model to identify key structure and dynamics determinants distinguishing the active and inactive states of GRK5. Benefiting from the unbiased and data-driven framework, the work reveals that the active site tether (AST) is a dominant activation-associated feature, acting as a conformational switch that regulates kinase domain movements. Beyond this canonical element, we also uncover two previously underappreciated structure modules contributing to GRK5 activation, such as the coupling interaction between the α10/α11 helix interface with the N-terminal lipid-binding domain (NLBD) in the active state, and the α5 helix region that facilitates large-scale RH domain reorientation. Conformation dynamics analyses further indicate that GRK5 activation involves disruption of the interdomain interactions and interaction coupling between AST, αN-helix, kinase domain N-lobe, NLBD, and α10/α11 hinge. These observations provide valuable insights into understanding the GPK5 activation mechanism and also highlight the power of machine learning in capturing functionally conformational changes, and in turn offering a methodological guideline for the studying of the protein function mechanism. Full article

Background: A major technical challenge in single-cell transcriptomics is the absence of an integrative analytic pipeline that can simultaneously leverage gene regulatory network (GRN) architecture, AI-assisted gene panel discovery, and functional relevance analyses to generate coherent biological insights. Existing approaches often treat these components independently, focusing on clusters, marker genes, or predictive features without integrating them into a mechanistically grounded framework. Consequently, comprehensive screening that links regulatory association, gene signature screening, and functional interpretation within single-cell datasets remains limited, underscoring the need for an integrated strategy. Methods: We developed an integrative bioinformatics pipeline based on Gene regulatory network–AI–Functional Analysis (GAFA), combining latent-space integration, unsupervised clustering, diffusion pseudotime analysis, lineage-resolved generalized additive modeling, GRN inference, and machine learning-based gene panel discovery. This framework enables systematic mapping of cell-state structure, reconstruction of differentiation and effector trajectories, and identification of transcriptional and regulatory features strongly associated with clinical outcomes. As a case study, we applied the pipeline to NK cell transcriptomes from six CML patients (two early relapse, two late relapse, two durable treatment-free remission—TFR; 15 samples) collected at TKI discontinuation and 6–12 months after therapy cessation. Results: We reanalyzed publicly available scRNA-seq data from a previously published CML cohort to evaluate NK-cell transcriptional programs associated with treatment-free remission and relapse. We resolved six transcriptionally distinct NK cell states spanning CD56^bright-like cytokine-responsive, early activated, terminally mature, cytotoxic, lymphoid trafficking, and HLA-DR⁺ immunoregulatory populations, each exhibiting outcome-specific compositional differences. Pseudotime analysis revealed two major NK cell lineages—a maturation trajectory and a cytotoxic effector trajectory. TFR samples displayed balanced occupancy of both lineages, whereas early relapse samples showed marked depletion of the maturation branch and preferential accumulation in cytotoxic end states. AI-guided feature selection and random forest modeling identified an 18-gene panel that distinguished NK cells from TFR and relapse samples in an exploratory manner. Among them, CST7, FCER1G, GNLY, GZMA, and HLA-C were conventional NK-associated genes, whereas ACTB, CYBA, IFITM2, IFITM3, LYZ, MALAT1, MT2A, MYOM2, NFKBIA, PIM1, S100A8, S100B, and TSC22D3 were novel. The GRN inference further uncovered outcome-specific regulatory modules, with RUNX3, EOMES, ELK4, and REL regulons enriched in TFR, whereas FOSL2 and MAF regulons were enriched in relapse, and their downstream targets linked to IFN-γ signaling, metabolic reprogramming, and immunoregulatory feedback circuits. Conclusions: This AI-enabled single-cell analysis demonstrates how NK cell state composition, differentiation trajectories, and regulatory network rewiring collectively shape TFR versus relapse following TKI discontinuation in CML. The integrative pipeline provides a modular framework that could be extended to additional datasets for data-driven biomarker discovery and mechanistic stratification, and highlights candidate transcriptional regulators and NK cell programs that may be leveraged to improve remission durability, pending validation in larger patient cohorts. Full article

Measles remains a significant public health threat despite widespread vaccination, with recent resurgences driven by vaccine hesitancy and coverage gaps. Existing mathematical models often fail to capture the substantial temporal heterogeneity in incubation periods, vaccine-induced protection, and recovery processes that characterize measles transmission. We develop and analyze an SVEIR epidemic model incorporating four independent distributed time delays with exponential survival factors, capturing the realistic variability in these epidemiological processes. The model features compartment-specific mortality rates, disease-induced mortality, and imperfect vaccination with failure probability $\theta$. Using next-generation matrix methods adapted for delay kernels, we derive the delay-dependent reproduction number ${\mathcal{R}}_0^d$ and prove, via systematic construction of Volterra-type Lyapunov functionals, that it constitutes a sharp threshold: the disease-free equilibrium is globally asymptotically stable when ${\mathcal{R}}_0^d\le 1$, while a unique endemic equilibrium emerges and is globally stable when ${\mathcal{R}}_0^d>1$. Normalized forward sensitivity analysis reveals that the transmission rate $\beta$ and recruitment rate $\Lambda$ exhibit maximal positive elasticity, while the vaccination rate p, vaccine failure probability $\theta$, and incubation delay ${\tau }_3$ possess the largest negative elasticities. Critically, ${\tau }_3$ exerts exponential influence via $e^{-n_3{\tau }_3}$, making interventions that delay infectiousness—such as post-exposure prophylaxis—unusually potent. We derive an explicit expression for the critical delay ${\tau }_3^{\mathrm{cr}}$ at which ${\mathcal{R}}_0^d=1$, demonstrating that prolonging the effective incubation period sufficiently can shift the system from endemic persistence to extinction. Numerical simulations using Dirac delta kernels confirm all theoretical predictions. These findings provide three actionable insights for public health: (1) maintaining high vaccination coverage among new birth cohorts remains paramount; (2) improving vaccine quality (reducing $\theta$) yields substantial returns; and (3) the incubation delay represents a quantifiable, measurable target for evaluating the population-level impact of time-sensitive interventions. The framework is broadly applicable to infectious diseases characterized by significant temporal heterogeneity. Full article

Gas kick events in drilling operations are characterized by strong coupling dynamics, subtle early-stage evolution, and severe class imbalance, which limit the effectiveness of conventional feature-independent monitoring methods. To address these challenges, this paper proposes a structure-aware intelligent monitoring framework for early gas kick detection. First, multivariate drilling parameters are modeled as an interacting graph, and a graph neural network (GNN) is introduced to capture relational dependencies and anomaly propagation behaviors at the structural level. Second, to mitigate abnormal sample scarcity and enhance temporal discriminability, a representation enhancement strategy integrating conditional tabular generative adversarial networks (CTGAN) and shapelet-based temporal patterns is developed. Finally, a multi-level interpretability mechanism combining graph attention analysis and SHAP attribution is constructed to provide transparent insights into both structural interactions and feature contributions. Experiments conducted on real drilling datasets demonstrate that the proposed GNN baseline achieves the highest accuracy (0.7302) among various machine learning and deep learning models. With representation enhancement, the GNN+CTGAN+Shapelet model further improves accuracy to 0.7507 and F1-score to 0.7347, validating the effectiveness of the enhancement strategy. Interpretability results reveal that the model decisions are primarily driven by flow-rate and standpipe-pressure-related temporal evolution patterns, which are consistent with drilling engineering knowledge. Overall, the proposed framework provides a structurally consistent, robust, and interpretable solution for intelligent gas kick monitoring in modern drilling operations. Full article

Computed tomography (CT) enables non-destructive insight into internal log structure, yet fully automated interpretation of CT images remains limited by inconsistent annotations, boundary ambiguity, and insufficient spatial context in 2D slice-based analysis. These challenges restrict the industrial deployment of deep learning for wood quality assessment. This study applies artificial intelligence (AI) and deep learning to the automated analysis of computed tomography (CT) scans of wood logs for detecting internal qualitative features and segmenting bark. Using convolutional neural networks (CNNs), trained models accurately distinguish healthy and damaged regions and segment bark, including discontinuous parts. We introduce a novel pseudo-spatial representation by merging consecutive slices into red–green–blue (RGB) format, which improves prediction accuracy and model robustness across logs. To enhance interpretability, Gradient-weighted Class Activation Mapping (Grad-CAM) highlights regions contributing most to defect detection, particularly knots. Comprehensive evaluation using Sørensen–Dice similarity coefficients and confusion matrices confirms the effectiveness of the proposed approach under industrial conditions. These findings demonstrate that AI-driven CT image analysis can address key limitations of current log-grading workflows and enable more reliable, objective, and scalable quality assessment for timber-dependent economies. Full article

In this study, we propose a structured valuation framework for non-fungible tokens (NFTs), a distinct class of digital assets whose pricing mechanisms remain insufficiently understood. Based on previous empirical studies and illustrative case analyses of three major NFT collections, we synthesize insights from non-cash-flow asset theory, market microstructure, and behavioral finance to construct a four-layer valuation framework consisting of the Asset, Market, Technology, and Ecosystem layers. We identify three NFT-specific mechanisms—verified digital scarcity, pseudonymous signaling, and on-chain herding—that modify or extend traditional valuation paradigms. Empirical evidence from the literature suggests that rarity-driven asset features and social-influence dynamics are dominant price determinants, while wash trading, fragmented liquidity, and platform incentive structures generate persistent distortions in price discovery. Case analyses of CryptoPunks, Bored Ape Yacht Club, and Pudgy Penguins demonstrate how differing risk exposures across the four layers translate into distinct valuation trajectories. With this framework, we obtain a basis for improved risk assessment, regulatory oversight, and business model design in NFT markets. Full article

Background: In an earlier murine model of myocardial infarction (MI), we showed that CD8 cells and myeloid dendritic cells (mDCs) infiltrate the infarcted myocardium within the first week. However, in humans, the spatial interplay between CD8⁺ T cells and dendritic cells in the spatial context of human myocardial infarction remains underexplored. Objective: In the present study, we applied spatial transcriptomics and functional assays to characterize immune–stromal dynamics in infarcted myocardium and peripheral blood. Methods & Results: Spatial transcriptomics analysis of infarcted human myocardium at days 2 and 6 post-MI, combined with peripheral blood flow cytometry and EPC colony-forming assays, was performed. Cell composition, pathway enrichment, and cell-to-cell communication analyses were conducted to map immune–stromal cells’ dynamics across time points. Spatial mapping identified dynamic shifts in immune, fibroblast, and endothelial populations, with fibroblasts and endothelial cells remaining abundant throughout. CD8⁺ T cells accumulated in ischemic regions while their circulating levels declined. Gene Ontology and pathway analyses of CD8A⁺ transcripts revealed enrichment of proinflammatory and NF-κB survival programs. ITGAX/CD33/THBD⁺ APCs progressively increased within infarct zones, activating antigen-presentation and leukocyte chemotaxis pathways. Early (day 2) APC–endothelial crosstalk showed the strongest predicted recruitment signals for CD8⁺ T cells, which diminished by day 6. Finally, EPC colony-forming capacity showed a tendency for reduction in MI patients and inversely correlated with coronary lesion burden, indicating impaired vascular repair potential. Conclusions: This integrative spatial and functional study demonstrates that APC-driven CD8⁺ recruitment and EPC dysfunction are key features of human MI. Immune–endothelial niches facilitate early cytotoxic T-cell infiltration, while progenitor depletion limits vascular regeneration. These findings provide mechanistic insight into immune–vascular imbalance during infarct healing and highlight potential therapeutic targets to modulate inflammation and restore vascular repair. Full article

Metastatic colorectal cancer (mCRC) accounts for 90% of CRC-related mortality. This review synthesizes insights from comparative genomics tracing evolutionary trajectories from primary tumor to disseminated disease. Multi-region sequencing reveals metastatic seeding often occurs early—before clinical detection—challenging linear progression models. The metastatic bottleneck reduces clonal diversity while enriching for dissemination-competent traits including SMAD4 loss, PTEN inactivation and metabolic reprogramming. Organ-specific adaptation yields distinct molecular signatures: liver metastases exhibit Wnt hyperactivation and TGF-β-driven immune suppression; peritoneal tumors display mucinous features; brain metastases show HER2 enrichment. The immune microenvironment evolves toward immunosuppressive configurations, with Microsatellite instability high (MSI-H) tumors acquiring B2M or JAK1/2 mutations. Circulating tumor DNA (ctDNA) enables real-time tracking of clonal dynamics, detecting molecular residual disease months before radiographic progression. Therapeutic resistance follows predictable evolutionary trajectories—from RAS/BRAF mutations to EGFR ectodomain alterations, HER2/MET amplifications and lineage plasticity—with metastasis-specific mechanisms including microenvironmental protection and cellular dormancy. The clinical future lies in interception: leveraging liquid biopsies for early detection, targeting both tumor-intrinsic vulnerabilities and permissive metastatic niches and adapting therapy dynamically to anticipate resistance. Understanding this genomic odyssey is essential for transforming mCRC into a controllable chronic condition. Full article

Background: Warfarin pharmacogenomics is critical due to its narrow therapeutic index and significant interpatient variability. While machine learning (ML) can predict anticoagulation control status (ACS), its “black-box” nature limits clinical translatability. Explainable Artificial Intelligence (XAI) addresses this by providing interpretable insights. This study applied ML and XAI to a warfarin pharmacogenomic dataset to predict poor ACS and explain model decisions. Methods: A post hoc analysis was conducted on a cross-sectional dataset of 232 patients receiving warfarin for ≥6 months. Data included age, gender, interacting drugs, SAMe-TT2R2 score, and genotypes for CYP2C9, VKORC1, and CYP4F2. Poor ACS was defined as time in therapeutic range (TTR) < 70%. The dataset was split into training (70%) and testing (30%) cohorts. Three models, Random Forest, XGBoost, and Logistic Regression, were developed and evaluated using AUC-ROC, sensitivity, and specificity. XAI techniques, including permutation importance and SHapley Additive exPlanations (SHAP), were employed for global and local interpretability. Results: Of 232 patients, 141 (60.8%) had poor ACS. XGBoost and Random Forest demonstrated comparable predictive accuracy (AUC-ROC: 0.67), outperforming Logistic Regression. Sensitivity was 0.83 and 0.79 for XGBoost and Random Forest, respectively. However, specificity was modest for both ensemble methods (Random Forest: 0.48; XGBoost: 0.41) and extremely low for Logistic Regression (0.04), indicating poor discrimination, particularly for identifying patients with adequate anticoagulation control. Globally, important predictors included the age, SAMe-TT2R2 score, CYP2C9 (*2/*2), female gender, and VKORC1 (C/T). XAI revealed predictions were primarily driven by VKORC1, CYP4F2, SAMe-TT2R2 scores, and drug interactions. Concordance between XAI predictions and actual ACS was 78% for adequate and 88.6% for poor ACS. SHAP analysis showed VKORC1 provided a stable risk signal (mean absolute SHAP: 1.44 ± 0.49 in concordant cases), while CYP2C9 was a high-variance, high-impact driver of discordance (mean SHAP: 3.44 ± 3.79 in discordant cases). Conclusions: ML models, particularly ensemble methods, show modest ability to predict poor warfarin control with limited ability to correctly identify patients with adequate control from our dataset. XAI transforms these models into interpretable tools, with SHAP analysis attributing predictions to specific genetic and clinical features. While predictive accuracy remains modest, this approach enhances transparency and provides a foundation for generating hypotheses that may ultimately support clinical decision-making in pharmacogenomic-guided warfarin therapy. Full article

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by genetic heterogeneity. Post-transcriptional regulation—particularly alternative polyadenylation (APA)—plays a critical role in the pathogenesis of ASD. APA controls mRNA stability, translational efficiency, and subcellular localization through modulating the length of the 3′ untranslated region of mRNA. APA profiling can uncover functionally relevant post-transcriptional alterations often missed by conventional gene expression analyses. However, current ASD analyses still largely rely on differential gene expression or individual APA event detection, which ignores the collective explanatory power of ASD risk genes or co-dysregulated functional gene modules within specific cell types. In this study, we present an integrative computational framework that combines matrix factorization and machine learning to identify ASD-associated gene modules driven by APA and to predict cell-type-specific ASD-related cells. Applied to human brain single-nucleus RNA sequencing (snRNA-seq) data, our approach systematically uncovers APA regulatory patterns that are specific to cell type, brain region, and sex in ASD. The identified APA modules are significantly enriched in pathways related to synaptic function, neurodevelopment, and immune response, with the strongest signals observed in excitatory neurons of the prefrontal cortex. Using APA genes from these modules as features, we built a classification model that effectively distinguishes ASD cells from normal cells. Moreover, we found that integrating APA with gene expression—two complementary modalities—substantially improves prediction accuracy, underscoring APA as an independent and biologically informative regulatory layer. Our work delineates a high-resolution APA regulatory landscape in ASD, offering novel insights and potential therapeutic avenues beyond transcriptional abundance. Full article

Background/Objectives: Feature selection approaches should satisfy all evaluation criteria required by state-of-the-art chemoinformatic models. Our aim is to develop a methodology that is robust, interpretable and computationally efficient. Methods: This study presents a robust methodology for developing highly interpretable and computationally efficient predictive models, with a specific application in the discovery of SARS-CoV-2 main protease inhibitors. We evaluated various descriptor selection procedures to identify a transparent and reproducible approach that provides actionable insights for data-driven decisions. The models were trained and tested using molecules from the CHEMBL database and further validated on an external set of compounds. Results: Our findings demonstrate that a recently proposed procedure, combining the FeatureWiz algorithm with stepwise feature selection, is the only approach that satisfies all evaluation criteria required by state-of-the-art chemoinformatic models. In particular, we found that models based on two-dimensional descriptors and Ordinary Least Squares regression achieved the best results. Conclusions: Our framework and the choices made offer significant advantages in a decision-making context due to their inherent interpretability and computational efficiency. Our derived models, benchmarked against those in the literature, serve as effective, transparent tools for the rapid and reliable prediction of biological activity, providing a validated framework for data-driven decisions in drug discovery and beyond. Full article