Triage tools in routine emergency care are largely static and may miss simple dynamic bedside cues available after presentation. We developed and temporally evaluated an emergency severity index (ESI)-informed Bayesian sequential prediction model for hospital admission using time-to-urination (TTU) in a prospective single-center cohort of ambulance-transported emergency department patients in Japan (February-August 2025; [Formula: see text]). The outcome was hospital admission at emergency department disposition. ESI was used as the initial pretest risk layer, TTU as a dynamic updating cue, and age and sex as refinement variables. Population-level fit to the cumulative admission curve was strong. In nested model comparison, ESI alone yielded an AUROC of 0.661 (95% CI 0.640-0.680), adding TTU improved discrimination to 0.677 (95% CI 0.658-0.698), and further adjustment for age and sex yielded the best performance (AUROC 0.741, 95% CI 0.722-0.760). Recalibration improved probability alignment without materially changing discrimination. Calibration deteriorated later in the post-arrival period, suggesting that the model is most informative in the early post-arrival window. This framework is designed to augment, rather than replace, existing triage systems.

Population pharmacokinetic (PK) models can be combined with Bayesian estimation to optimize dosing regimens. The impact of sample collection time on the accuracy and precision of Bayesian predictions was evaluated. Data from adult and pediatric patients were used to develop a cefepime population PK model for Bayesian prior use. Holdout data were used for model evaluation. Clinical dosing regimens in the latter cohort were used to conduct optimal sample-time analysis. The accuracy and precision of the Bayesian predictions were assessed as a function of infusion duration and the differences between the observed and optimal sampling times. Analyses were conducted using Pmetrics for R. An allometrically scaled 2-compartment model was fitted (n = 71 patients, 685 observations). In the holdout group (n = 116 patients, 203 observations), the posterior Bayesian fit was acceptable (R2 = 0.923; relative bias -3%; median absolute error, 11.2%; F20, 72%; and F30, 86%). Mid-interval sampling was the optimal 1-sample design for 11/16 regimens. In the 2-sample design, a peak (8/16 regimens) and trough (9/16 regimens) approach was frequently optimal. The 2-sample design yields a lower Bayesian risk of misclassification. For 0.5-hour infusions, Bayesian predictions were similarly accurate but significantly more imprecise when samples were collected >2 hour away from the optimal time versus within ±1 hour of the optimal time (ΔRMSE: 8.98 mg/L, 95% CI: 3.61-15.7 mg/L). For 3 hours infusions, no significant differences in the accuracy or imprecision of the Bayesian predictions were noted. The nonparametric cefepime population PK model fit as a Bayesian prior in the holdout group. The optimal timing of PK sample collection varied according to regimen type and infusion duration. The precision of Bayesian estimates was lower for 0.5-hour infusions when samples were collected further from the model-predicted regimen-specific optimal collection times.

This study investigates the relationship between DNA methylation patterns and clinicopathological characteristics in prostate cancer. We performed targeted next-generation sequencing (NGS) on methylation sites linked to prostate cancer and calculated the corresponding methylation rates. Clinical samples were divided into training and prediction sets. In the training set, we utilized unsupervised clustering and Support Vector Machine (SVM) modeling to distinguish between prostate cancer and non-cancer samples. For predictions, K-Nearest Neighbors (KNN) assessed sample similarity, while SVM facilitated classification. Bayesian methods integrated probabilities to predict cancer status and cluster assignments. To address uncertainties identified by Uniform Manifold Approximation and Projection (UMAP), we validated our results using Random Forest Support Vector Machine (RFSVM), which highlighted significant methylation sites for SVM training. Depth-correction methods were applied to mitigate variations in sequencing depth.In the training dataset, the leave-one-out cross-validation (LOOCV) prediction accuracy of RFSVM was 0.85 (AUC: 0.91); for RFSVM-depth, it was 0.83 (AUC: 0.93). The LOOCV prediction accuracy for Bayesian SVM (BSVM) was 0.87 (AUC: 0.94), decreasing to 0.83 (AUC: 0.91) with depth correction. In the test dataset, our Bayesian prediction achieved an accuracy of 0.8793 (sensitivity: 0.8182, specificity: 0.9167), which improved to 0.9655 with depth correction (sensitivity: 1.0, specificity: 0.9444). RFSVM demonstrated an accuracy of 0.8621 (sensitivity: 0.8182, specificity: 0.8889), dropping to 0.7759 with depth correction. Among 58 samples, predictions showed 67% complete consistency, with 84% consistency in methylation rates and 78% in specific methylation sites. Figure6 presents a comprehensive comparison of the performance of RFSVM, BSVM, XBSVM, and RFESVM on training and test datasets, including feature selection and prediction consistency. Confusion matrices, ROC curves, and Venn diagrams were used to detail the feature importance and prediction consistency of each method.These findings highlight the importance of analyzing individual methylation sites and broader methylation patterns in understanding the role of DNA methylation in prostate cancer, providing valuable insights into the effects of data preprocessing and feature selection.

Social information aids learning: By making assumptions about other people's knowledge and intentions, people can draw strong and accurate inferences from limited data. In this study, we systematically tested people's ability to reason from information providers with different intentions. The task was an adaptation of Shafto et al.'s (2014) rectangle game, where learners guessed a rectangle's size and location based on provided clues. We examined reasoning based on information from four types of providers: a helpful provider, a provider who sampled randomly, and two misleading providers (who could mislead but not lie). We also varied whether people were given a cover story describing the provider in advance or whether they could infer how helpful a provider was based on what the provider shared. Participants learned efficiently from helpful providers, aligning closely with the predictions of a normative Bayesian model, even without a cover story. However, while people usually recognized unhelpful providers, they struggled to identify and respond appropriately to misleading strategies. Overall, our results suggest a helpful bias: In our task, participants assumed helpful intent unless given explicit feedback, and even then, they did not fully adjust in line with Bayesian predictions. People also struggled to overcome this bias when learning from randomly sampled information, especially when they had experience being an information provider themselves (Experiment 3). (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Heartwood traits in trees are critical for timber quality but are notoriously difficult to phenotype due to their late expression and the need for destructive sampling. In this proof-of-concept study, we demonstrate that combining genome-wide association studies (GWAS) with Bayesian genomic prediction models provides an effective strategy to overcome these challenges. By using GWAS to preselect trait associated single nucleotide polymorphisms (SNPs) and integrating them into predictive models, we substantially improve the accuracy of genomic predictions for heartwood related traits in oaks. Our approach facilitates reliable selection of superior genotypes within a breeding population, long before heartwood traits can be directly measured, thus enabling early and cost-effective breeding decisions. We also identify the number of rings in sapwood as a genetically controlled, easily measured proxy trait that enhances selection strategies for heartwood content. Together, these findings provide a scalable framework for integrating genomics into operational tree breeding programs and demonstrate how combining GWAS and genomic prediction can accelerate the improvement of complex wood traits in long lived forest tree species. • Heartwood traits are difficult to phenotype as they require destructive sampling. • Genomic prediction enables early selection of superior heartwood genotypes. • Sapwood ring number provides an easy proxy for heartwood content. • Genomics-based breeding framework in oaks transferable to other tree species.

ABSTRACTGenome-wide association studies have identified risk loci for aging brain disorders, but mechanistic interpretation depends on linking these loci to genes and to the tissues, cell types, and molecular modalities in which those genes act. Here we introduce FunGen-xQTL Multi-Brain (FGMB), a multi-context regulome-wide association atlas for transcriptome-wide association studies (TWAS) built from molecular datasets assembled by the ADSP Functional Genomics Consortium. FGMB providescis-genetic prediction models for 17,375 protein-coding genes across 36 molecular datasets, 18 contexts, and 3 regulatory modalities, yielding more than 293,000 imputable gene-level or splice-event models. FGMB evaluates eight established and newer Bayesian or multivariate prediction methods, including cross-context models that borrow information across tissues and cell types. Applied to Alzheimer’s disease, FGMB identified 327 TWAS associations and used joint fine-mapping of variants and predicted molecular traits to prioritize 146 gene–molecular-trait pairs, distinguishing regulatory associations from linkage disequilibrium (LD) hitchhiking.

Cognition under uncertainty can be formalized through Bayesian inference, but biologically plausible neural implementations remain a challenge. This study develops a Bayesian neural model for lifespan prediction that integrates fractional-order dynamics into classical Leaky Integrate-and-Fire and Izhikevich neuron models. The inclusion of fractional derivative introduces long-term memory, and thus enhancing both biological plausibility and representational capacity of the Bayesian neural model. Experimental results demonstrate that fractional-order neuron models consistently provide closer alignment with both human predictions and optimal Bayesian predictions. The large-scale fractional-order Izhikevich model shows the most robust convergence and cortical plausibility. These findings highlight the role of fractal neural dynamics in probabilistic cognition and bridging theoretical Bayesian models with realistic spiking behavior. The study demonstrates how biologically inspired spiking neuron models can approximate Bayesian inference, suggesting pathways for computational neuroscience to design models that learn, predict, and adapt with the efficiency of cortical computations. Further, in this study, neural populations represent priors from demographic lifetables. However, a uniform likelihood and posterior that yield median lifespan predictions as probability distributions within the Neural Engineering Framework have been retained from the previous study. To investigate the influence of neural population size on biological plausibility and Bayesian optimality, experimental conditions systematically increase the neural population size and compare the predictive outcomes.

Contemporary burden of subarachnoid hemorrhage: a comparative assessment of global and Asian trend.

• Multiscale wildfire exposure mapping for safeguarding rural cultural heritage. • Automated harmonisation and spatial downscaling of environmental data with Empirical Bayesian Kriging Regression Prediction. • Integration of GeoAI and spatial analysis for fire occurrence pattern modelling and cultural heritage exposure. • Proposal of a Total Exposure Index (TEI) to support heritage-oriented prevention and planning. Rural cultural heritage faces escalating threats, notably the rising incidence of wildfires, driven by climate change and evolving land-use dynamics. This study presents a GeoAI framework that combines spatial analysis and environmental data downscaling to model wildfire distribution and quantify heritage site exposure. The results demonstrate the effectiveness of the method in capturing real spatial patterns of exposure and providing tools and indicators to support territorial management and planning for the protection of rural cultural heritage. Overall, the findings reveal the close interplay between cultural landscapes and ecological processes, highlighting the need for integrated environmental and heritage governance to enhance resilience.

Combining dynamic Bayesian prediction of the crop coefficient with automated lysimetry for highly accurate water-use control

BackgroundBayesian dose prediction software supports model-informed precision dosing to improve therapeutic outcomes. This study evaluated clinicians’ awareness, usage, attitudes, and perceived barriers related to Bayesian dosing software.MethodsA cross-sectional study was conducted using a validated electronic questionnaire distributed through eight international professional organizations. Descriptive statistics, non-parametric tests, and logistic regression analyses were performed using SPSS version 29.ResultsOf 234 respondents, 190 (81.2%) were eligible and completed the survey. Most were from North America (67.4%), pharmacists (76.3%), and worked in academic hospitals (78.4%). Awareness of Bayesian dosing software was high (82.6%) among respondents, yet only 65% reported its use in clinical practice. The most used software included InsightRx (40.2%), DoseMe (22.5%), and MWPharm (18.6%), mainly for vancomycin (94.1%) and aminoglycosides (50%). The overall median attitude score (IQR) was 4.0 (2.0), indicating positive attitudes, whereas the barrier score was 3.0 (2.0), reflecting high implementation challenges. The most frequently reported barriers were prohibitive licensing costs (65%), lack of institutional support for use and maintenance (50.3%), and limited awareness of its use (49.7%). In multivariable analysis, “region” was the sole predictor of awareness with participants from the Middle East significantly less aware (AOR = 0.14; 95% CI: 0.05–0.42; p = 0.002). Software type was significantly associated with professional role, with pharmacists more likely to select commonly used tools (AOR = 7.2; 95% CI: 1.4–17.9; p = 0.003). A non-significant negative correlation was observed between overall attitude and barrier scores (r = −0.49; p = 0.072).ConclusionDespite high awareness and positive attitudes toward Bayesian dosing software, its clinical use remains limited, primarily focused on vancomycin and aminoglycosides, mostly restricted to three commercial tools. Targeted interventions are required to address key implementation barriers related to licensing costs, institutional support, and awareness.

Bayesian model prediction of colorectal cancer incidence in Jiangsu Province, China, a cancer registry-based study

Artificial intelligence-based Bayesian regression predictions and entropy generation analysis of Oldroyd-B fluid flow through porous structures

Plant hardiness zone mapping for the conterminous USA using the Empirical Bayesian Kriging Regression Prediction method

Introduction: Infectious disease outbreaks remain a persistent global health burden, particularly as populations experience the concurrent circulation of multiple infectious agents that interact across space and time. Forecasting such complex epidemic systems requires models that can capture shared transmission mechanisms, pathogen-specific dynamics, and uncertainty arising from incomplete surveillance.Methods: In this study, we develop a unified Bayesian hierarchical spatiotemporal framework for predicting multi-pathogen outbreak trajectories while integrating human mobility patterns, environmental exposures, and structured reporting uncertainties. We conducted a comprehensive simulation experiment to evaluate the model’s ability to recover known parameters, distinguish pathogen-specific transmission effects, and generate calibrated forecasts under varying levels of reporting noise and spatial heterogeneity. We further applied the method to CDC FluView surveillance weekly data from the United States, spanning January 2017 to December 2025.Results: In the simulation study, the model showed good parameter recovery under different levels of reporting noise and spatial heterogeneity, with stable estimates and satisfactory convergence. The model effectively distinguished pathogen dynamics, with posterior means for baseline incidence (alpha_{1} and alpha_{2} at -0.74 and -0.87). Human mobility (eta = 0.35, 95% CI: 0.06-0.72) was a significant driver of spatial transmission, while overdispersion parameters (phi_{1} = 11.61, phi_{2} = 5.74) accounted for variability beyond the mean structure. Diagnostic Rhat ​values near 1 confirmed model convergence and robust chain mixing. In the real data application, influenza showed a higher baseline incidence than measles (alpha_{1} = -0.74, alpha_{2} = -0.87), temperature had a positive effect on transmission (beta = 0.19), while humidity effects were weaker and more uncertain, and the mobility parameter (eta = 0.35) indicated that human movement contributed to spatial spread; influenza also exhibited greater variability (phi_{1} = 11.61 vs phi_{2} = 5.74), and the model captured seasonal patterns while closely tracking the observed incidence over time.Conclusions: Across all scenarios, the model demonstrated robust parameter recovery, reduced bias in reproduction number estimates, and improved predictive accuracy relative to conventional compartmental and independent-pathogen Bayesian models. This performance was consistent in both simulation and real data settings, where the model distinguished pathogen-specific dynamics, captured the contribution of human mobility to spatial transmission, and accounted for variability in case counts through overdispersion. The results support the use of this approach for stable inference and reliable forecasting in complex multi-pathogen systems.

Hybrid DRASTIC-EBKRP-Hydrochemistry approach for groundwater vulnerability assessment in a semi-arid context under agricultural pressure

Predicting how green policies reshape power business environments remains notoriously difficult. The underlying dynamics are nonlinear, the uncertainties substantial, and conventional models often fall short. This study develops a Bayesian neural network framework designed specifically for forecasting and optimizing green policy outcomes within the Fujian power system, placing particular weight on quantifying prediction uncertainty to support sound decision-making. Our methodology weaves together stochastic variational inference and multi-objective optimization, thereby capturing the channels through which policies transmit their effects to environmental outcomes. Drawing on empirical data spanning 2018–2024, we find that this approach outperforms standard machine learning techniques by roughly 4–5% points in prediction accuracy while delivering markedly better uncertainty calibration. Scenario analyses reveal that moderate-to-high policy intensity tends to achieve favorable cost-effectiveness, with renewable energy incentives, carbon pricing, and regulatory enforcement standing out as especially potent drivers of transformation. Perhaps more importantly for practitioners, the framework demonstrates that well-designed moderate-intensity strategies can surpass maximum-intensity approaches once diminishing returns enter the picture. By enabling joint assessment of environmental gains, economic efficiency, and operational stability under uncertainty, this work offers a practical foundation for evidence-based policy design—though readers should bear in mind that our validation remains grounded in the Fujian regional context.

How does the brain optimize sensory information for decision-making in new tasks? One hypothesis suggests that learning reduces redundancy in neural representations to improve efficiency, whereas another, based on Bayesian inference, predicts that learning increases redundancy by distributing information across neurons. We tested these hypotheses by tracking population responses in macaque cortical area V4 as monkeys learned visual discrimination tasks. We found strong support for the Bayesian predictions: Task learning increased redundancy in neural responses over weeks of training and within single trials. This redundancy did not reduce information but instead increased the information carried by individual neurons. These insights suggest that sensory processing in the brain reflects a generative rather than discriminative inference process.

Corrigendum to “Plant hardiness zone mapping for the conterminous USA using the Empirical Bayesian Kriging Regression Prediction method” [Applied Geography 189 (2026) 103927

Accurate prediction of desalination jet behavior in coastal environments is essential for optimizing discharge design and minimizing environmental impacts. This study presents a Multifidelity Gaussian Process (MFGP) framework for predicting the behavior of desalination discharges under different configurations. The framework integrates computationally efficient low-fidelity (LF) Reynolds averaged Navier–Stokes (RANS) simulations with high-fidelity (HF) Large Eddy Simulation (LES) data, using experimental true values (TV) measurements to correct residual bias and ensure consistency across fidelity levels. Two representative scenarios were investigated, an inclined dense desalination jet in shallow ambient conditions (RANS-LES validated with PIV) and a vertical thermal desalination jet (RANS-LES combined with LIF data). The formulation systematically links LF, HF, and TV datasets through hierarchical inference, enabling bias correction and uncertainty quantification. Results show that the MFGP accurately predicts desalination jet behavior including dilution and geometrical characteristics while reducing prediction error compared with single fidelity models. The framework achieves high accuracy using only a fraction of the computational and experimental effort. This study demonstrates that multifidelity modeling provides an efficient and reliable approach for the design, operational assessment, and optimization of desalination discharges. • Bayesian multifidelity GP framework for desalination jet prediction. • Three-level integration of RANS, LES, and LIF experimental data. • Achieves high accuracy with quantified predictive uncertainty. • Demonstrates robust scalability across complex jet configurations.