Spatial clustering is a powerful technique for exploratory spatial data analysis and has been widely used in geoscience, economics, and social sciences. However, spatial clustering still faces challenges, such as complex data distribution, parameter sensitivity, and noise interference. In this paper, we propose a robust spatial clustering algorithm by Network Edge Pruning and Internal Connection (NEPIC). Specifically, Delaunay Triangulation Network (DTN) is leveraged to establish the spatial proximity relationships between geographical entities. Four geometric features, edge length, expansion ratio, gravitation consistency, and connection strength, are adopted to distinguish Intra- and Cross-Cluster Edges (ICEs, CCEs). An initial clustering can be obtained through CCE pruning. Internal and boundary points are further identified, and connections involving boundary points are temporarily removed to separate weakly connected clusters. The final clusters are formed by connecting internal points and reassigning each boundary point to the cluster of its nearest internal neighbor. To demonstrate the effectiveness, we compared NEPIC with four typical baselines, including K-means, DBSCAN, ASCDT, and CDC, across six synthetic datasets. The results exhibited its advantage in clustering accuracy and parameter robustness. Moreover, we applied NEPIC in a real-world case to unravel the spatial patterns of global terrorist incidents and interpret the development trends of Syrian terrorism.

Deconstructing order and emergence in spatial cluster and evolution of the Guangdong-Hong Kong-Macao Greater Bay Area using percolation theory

Tuberculosis (TB) remains a major public health challenge in Thailand, a high-burden country undergoing both epidemiological transition and pandemic-related disruption. This study examined temporal and spatial patterns of age-standardized TB incidence from 2012 to 2023 across Thailand's 13 health regions and 77 provinces. This nationwide ecological study used annual TB case notifications (2012-2023) for all 77 Thai provinces from Thailand's National Disease Surveillance System (Report 506), Bureau of Epidemiology, with mid-year and age-stratified provincial populations. Age-standardized incidence rates (ASR) were calculated using the WHO World Standard Population (2000-2025). Temporal trends were assessed by Joinpoint regression with Bayesian-Information-Criterion model selection (maximum 2 joinpoints) and validated by generalized additive models. Spatial clustering was evaluated by Global Moran's I and Local Indicators of Spatial Association (LISA) under queen contiguity, with sensitivity analyses across alternative weights and Benjamini-Hochberg false-discovery-rate correction. Provinces were classified into long-term trend categories combining effect size (AAPC) and significance, and COVID-19 impact was quantified by 2019-2023 ASR percent change. National TB incidence declined substantially from 2012 to 2023, although marked regional heterogeneity persisted. Five health regions showed the strongest long-term reductions: region 3 [AAPC: - 20.51, 95% confidence interval (CI): - 33.00 to - 13.77], region 9 (- 19.42, 95% CI: - 28.98 to - 8.57), region 4 (- 17.53, 95% CI: - 23.64 to - 12.89), region 11 (- 11.79, 95% CI: - 22.16 to - 0.04), and region 13 (- 11.19, 95% CI: - 22.44 to - 3.74). Region 6 showed an early decline followed by stabilisation, and region 12 showed a mid-period increase before a post-2019 reduction. Thailand has achieved substantial overall reductions in TB incidence over the past decade, but pronounced regional and provincial disparities persist. Localized hot-spots, biphasic regional trajectories that may reflect surveillance and programmatic transitions, and compound vulnerability during the COVID-19 period underscore the need for geographically targeted monitoring, equitable resource allocation, and pro-poor interventions to sustain progress toward TB elimination.

Antimicrobial resistance (AMR) in livestock is a growing global health concern with important implications for food security, human and animal health. However, the spatial distribution and determinants of AMR in livestock systems remain insufficiently characterised. This study examined spatial patterns and contributions of selected AMR drivers in livestock AMR, using data from the ResistanceBank database, which encompasses livestock species, bacterial pathogens, and antimicrobial categories. This study included all data compiled in the Resistancebank database, including prevalence studies published between 2000 and 2021 and 33,186 resistance data points compiled from 93 countries. Spatial dependence was evaluated using Global and Local Moran's I statistics, while the influence of livestock species, pathogens, and antimicrobial classes was analysed using beta regression and Extreme Gradient Boosting (XGBoost) machine learning models. The spatial analysis includes only countries listed in the Resistancebank database and displays them. Global Moran's I revealed significant positive spatial autocorrelation in AMR proportions (I = 0.1911, p = 0.0246; Z-score = 1.9676, expected I = - 0.0127), indicating geographic clustering of resistance. High-high clusters were identified across South and East Asia, the Middle East, parts of Sub-Saharan Africa, and South America, whereas low-low clusters occurred in Southern Africa, Southeast Asia, and several European regions. Beta regression showed that cattle (β = -0.5953, p = 0.0263) and sheep (β = -0.7873, p = 0.0034) contributed less to AMR variation than buffalo, whereas highly important antimicrobials were associated with increased proportions of resistance (β = 0.3365, p < 0.0001). The XGBoost model demonstrated slightly better predictive performance (Root Mean Squared Error (RMSE) = 0.3428) than beta regression (RMSE = 0.3471). These findings reveal pronounced spatial clustering of AMR in livestock and underscore the need for strengthened global surveillance, improved antimicrobial stewardship and integrated One Health strategies to mitigate the spread of AMR.

Environmental exposures, such as air pollution and extreme temperatures, have complex effects on human health. These effects are often characterized by non-linear exposure-lag-response relationships and delayed impacts over time. Accurately capturing these dynamics is crucial for informing public health interventions. The Distributed Lag Non-Linear Model (DLNM) is a flexible statistical framework for estimating such effects in epidemiological research. However, standard DLNM implementations typically assume a homogeneous exposure-lag-response association across the study region, overlooking potential spatial heterogeneity, which can lead to biased risk estimates. To address this limitation, we introduce DLNM-Clust: a novel mixture of DLNMs that extends the traditional DLNM. Within a Bayesian framework, DLNM-Clust probabilistically assigns each geographic unit to one of latent spatial clusters, each of which is defined by a distinct DLNM specification. This approach allows capturing both common patterns and singular deviations in the exposure-lag-response surface. We demonstrate the method using municipality-level time-series data on the relationship between air pollution and the incidence of COVID-19 in Belgium. Our results emphasize the importance of spatially aware modeling strategies in environmental epidemiology, facilitating region-specific risk assessment and supporting the development of targeted public health initiatives.

Assigning realistic spatial locations to secondary activities — such as shopping or leisure — in synthetic populations remains a major challenge in activity-based transport modeling. Existing methods, including the frequently used Relaxation-Discretization Algorithm (RDA), succeed in reproducing realistic travel distances but fall short in capturing the spatial clustering of real-world activity patterns. To address this limitation, we introduce a guidance force that draws secondary activities toward dense clusters of relevant points of interest during the relaxation phase of the RDA. This modification enhances the spatial realism of synthetic activity distributions, producing clustered patterns that better reflect observed urban structures. It achieves this without major distortion to trip distance fidelity or requiring additional data inputs. Applied to the region of Lyon, France, our enhanced method shows stronger agreement with household travel surveys and public transport usage data, demonstrating its value for more accurate transport simulations and its potential impact on downstream task. A reference implementation of this work is available in EQUASIM. 1 1 Available at https://github.com/eqasim-org/eqasim-france/pull/385 . • New Relaxation-Discretization algorithm for secondary activity localization. • Incorporation of a guidance force drawing activities toward attractive areas. • Principled definition of this force using an inverse-square attraction model. • New convergence criterion ensuring stability and fidelity of distance distributions. • Validation using household travel survey data and public transport ticketing data.

Worldwide hotspots and ecological drivers of canine Echinococcus granulosus sensu lato: Space-time scan statistics and Maxent modelling from a systematic evidence base.

Interpretable behavioral clusters of gamblers through unsupervised learning.

Computational delineation and cellular profiling of murine cortical cell layers using multiplex immunofluorescence imaging.

Magnetic fields play a crucial role in compact object physics, particularly in white dwarfs (WDs), where high densities can sustain strong magnetic fields. Observations have revealed magnetized WDs (MWDs) with surface fields reaching approximately 10 9 G, although high-field MWDs are fewer in number in current catalogs owing to their intrinsic faintness and limitations in conventional electromagnetic surveys. In this study, we apply unsupervised machine learning (ML) techniques to systematically analyze a sample of hydrogen-atmosphere (DA) WDs. Using Uniform Manifold Approximation and Projection (UMAP) for dimensionality reduction and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) for cluster identification, we classify distinct subpopulations within the DA WD sample. Each cluster exhibits unique intrinsic properties such as mass, surface gravity, temperature, and age. Our analysis further reveals that these subgroups effectively differentiate MWDs from non-magnetic or weakly magnetic counterparts. Moreover, utilizing a set of previously confirmed MWDs, we estimate the field strengths of all other MWDs lacking magnetic field measurements. This study underscores the effectiveness of ML-based approaches in astrophysical discovery, particularly detecting magnetized compact objects when direct measurements are unavailable.

Important Agricultural Heritage Systems (IAHS) represents an important model for sustainable development, which faces increasing pressure under rapid modernization. This study investigates 205 Zhejiang Important Agricultural Heritage Systems (Zhejiang-IAHS) using spatial analytical methods and the GeoDetector model to examine their spatiotemporal evolution, spatial patterns, and driving factors. The results indicate that: (1) The temporal frequency of Zhejiang-IAHS formation follows an inverted U-shaped trajectory, and the spatial gravity center oscillated between Shaoxing and Jinhua across different historical periods; (2) Zhejiang-IAHS exhibits a polycentric and uneven clustering structure, high-density clusters are predominantly located in plains and basin areas, while different Zhejiang-IAHS types display distinct geographical affinities; (3) Regarding the driving mechanisms of the present-day distribution, the GeoDetector analysis reveals that the explanatory power of social factors and economic factors is significantly stronger than that of natural conditions. Specifically, contemporary elements such as cultural infrastructures are identified as dominant forces associated with the spatial clustering and retention of these heritage systems. Interaction analysis further reveals a synergistic mechanism wherein static resource endowments are activated by dynamic socioeconomic conditions, with the interaction between transportation networks and tourism development being particularly pronounced. These findings suggest that while natural endowments provided the initial foundation for heritage formation, modern socioeconomic vitality plays a decisive role in their current identification and preservation. These findings highlight that activating static natural endowments through dynamic social factors and economic factors is key to the contemporary persistence of IAHS, offering a theoretical framework for understanding their spatiotemporal evolution.

Typhoid fever (Salmonella Typhi) has declined dramatically in Korea over three decades, but residual seasonality, demographic patterns and provincial spatial concentration have not been systematically reanalysed. Using all 3535 typhoid cases reported to the Korea Disease Control and Prevention Agency between 2001 and 2024 (3051 domestic and 484 imported [13.7%]), we quantified the long-term annual trend with the Hamed-Rao modified Mann-Kendall test (τ = -0.667, p = 0.0011) and a negative-binomial generalized linear model (GLM) yielding a Sen slope of -7.3%/year (95% CI -9.9 to -4.4); a parametric-bootstrap analysis of the runner-up 1-breakpoint GLM identified 2018 as the most likely structural change point (95% CI [2009, 2022]). A continuous Morlet wavelet transform of the weekly series revealed annual periodicity that exceeded a 1000-simulation AR(1) red-noise null by a factor of 2.60. Direct age standardization and age-band-specific Mann-Kendall tests showed statistically significant declines in every age stratum. Provincial spatial analysis (16 sido) gave a global Moran's I of 0.403 (p = 0.0083, 9999 permutations) under a Queen + k-nearest-neighbor-2 baseline scheme, with the southeastern coastal cluster (Gyeongnam-Busan-Ulsan-Gyeongbuk axis) preserved across 12 alternative weighting schemes after empirical Bayes shrinkage of sigungu-level rates and across two equal-length time strata; one province (Ulsan) survived Benjamini-Hochberg FDR adjustment of the 16 LISA p-values. Korean typhoid has continued its long-term decline; residual incidence concentrates in the southeastern coastal corridor, and imported cases-including ciprofloxacin-resistant H58 strains from South Asia-now form a substantial fraction of notifications, motivating pre-travel counseling and post-travel vigilance.

We present a global dataset of spatiotemporally clustered drought events for 1980-2024, derived from daily precipitation, potential evapotranspiration, soil moisture, and surface runoff data. Drought conditions were consistently defined using a 10th percentile threshold and clustered in space and time using a three-dimensional implementation of the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. The dataset represents droughts as coherent spatiotemporal events rather than isolated grid-cell anomalies. For each drought event, it provides detailed metadata on spatial extent, temporal duration, severity, and centroid position. By applying a consistent event-detection framework across atmospheric forcing, root-zone soil moisture, and runoff response, the dataset supports systematic analysis of global drought dynamics and compound extremes. The dataset is openly available at https://doi.org/10.5281/zenodo.18292641, providing a reusable resource for climate, hydrology, and hazard research.

The transition from seed to seedling involves major changes in nuclear organization and gene expression, yet the extent to which this developmental transition requires chromatin reprogramming remains largely unexplored. Here, we report that Arabidopsis dry seed embryos accumulate the histone variant H2B.8, which contributes to higher-order chromatin organization by forming spatial clusters that structure the 3D nuclear space. H2B.8 predominantly assembles into heterotypic nucleosomes, is enriched at euchromatic transposons and lowly expressed genes and, during imbibition, modulates the transcriptional activation of a subset of these genes. Water uptake triggers a decrease in H2B.8 transcripts and the eviction of the H2B.8 variant from chromatin, in a process that operates independently of DNA replication but requires de novo transcription. Histone eviction is not restricted to H2B.8, as imbibition also induces turnover of the H3.3 histone variant and therefore initiates a replication-independent chromatin reprogramming process. These findings highlight a fundamental mechanism of epigenetic regulation during early plant development.

Tacrolimus therapeutic drug monitoring after liver transplantation is characterized by significant interindividual variability and sparse, irregularly sampled concentration data, which limits the applicability of conventional pharmacokinetic and data-intensive modelling approaches. We developed a hierarchical prediction framework integrating density-based spatial clustering of applications with noise-derived patient stratification with self-memory algorithm-based nonlinear grey Bernoulli model (SA-NGBM) using retrospective data from 129 liver transplant recipients. Patients were stratified into homogeneous subgroups using routinely available clinical indicators, and cluster-specific SA-NGBM models calibrated on representative patients were used to predict subsequent tacrolimus trough concentrations. Performance was further evaluated in an independent same-center validation cohort of 60 patients. In the development cohort, the overall mean absolute relative prediction error for the next tacrolimus concentration decreased from 41.4% with the nonclustered baseline to 21.2% with the clustered SA-NGBM framework. In the independent validation cohort, consistent performance gains were observed, with the mean absolute relative prediction error decreasing from 56.8% to 27.3% in the largest patient subgroup. Full longitudinal concentration profiles were required for only 4 representative patients. Overall, this clustered SA-NGBM framework reduces prediction error under sparse and irregular therapeutic drug monitoring conditions and provides a data-efficient stratified modelling strategy. However, further refinement and prospective validation are required before clinical implementation.

Coffee production in Peru plays a crucial socio-economic role, supporting over 200,000 families and contributing significantly to export income. However, the spatial variation in coffee farming across ecological and socio-economic regions remains poorly understood. This study examines spatial patterns of coffee farm clustering in three Peruvian mountainous regions (Moyobamba, Tingo María, and Tocache) using descriptive statistics, geospatial visualization, and unsupervised clustering techniques. Farm-level reports and government geospatial records covering 2019–2023 were analyzed to evaluate cultivation area, altitude, and spatial distribution. Kernel density mapping, Moran’s I spatial autocorrelation, and Local Indicators of Spatial Association (LISA) were applied to identify statistically significant clustering patterns, while regression analysis and DBSCAN clustering were used to evaluate spatial trends and production hotspots. Moran’s I indicated moderate spatial clustering (0.34, p &lt; 0.001), while regression analysis showed a weak negative association between altitude and cultivation area (β = −1.144 × 10−4, adjusted R2 = 0.023). Results suggest that measured environmental variables explain only a limited proportion of spatial variation in coffee production, indicating that additional unmeasured factors, potentially including socio-economic influences, may contribute to observed clustering patterns. These findings highlight the value of spatial analysis for understanding production heterogeneity and for supporting regionally adapted agricultural planning strategies.

Epidemiological studies of Talaromyces marneffei (T. marneffei) infection are largely confined to HIV-positive populations and specific geographic regions, whereas its population-scale characteristics and environmental factors associated with infection in China are lacking. This study investigates spatiotemporal patterns, host susceptibility, and environmental correlates of T. marneffei detection in Chinese mainland. Targeted next-generation sequencing (tNGS) data from patients hospitalized for acute respiratory tract infections (ARTIs) from 2022 to 2024 were used for epidemiological, spatiotemporal, and co-detection analyses. Geographical detector models with the q-statistic were employed to quantify the associations of meteorological, host distribution, and social factors on detection risk, where the q-statistic measures the proportion of spatial variance explained by each factor. Among 2,316 reported cases, we identified significant spatial clustering, with bimodal seasonal peaks in detection rate. Males and individuals aged 41-50 showed the highest susceptibility. Pneumocystis jirovecii was the predominant co-detected pathogen; 5 pathogens were positively and 16 were negatively correlated with T. marneffei. Univariate analysis using the q-statistic revealed that dew point temperature had the strongest explanatory power for T. marneffei detection at 48.55%. Bivariate interaction analysis demonstrated that paired factor combinations exhibited enhanced explanatory power for disease prevalence compared with single factors. Average air pressure, which alone explained only 1.4% of the spatial variance, showed markedly higher explanatory power when paired with other variables. This study revealed spatiotemporal heterogeneity, population susceptibility, and environmental factors associated with T. marneffei detection in Chinese mainland, which can guide disease monitoring and control through enhanced surveillance and tailored interventions in epidemic hotspots and among high-risk groups.

Urban agglomerations serve as a critical vehicle for advancing high-quality regional urbanization, with the Lanzhou-Xining (Lanxi) urban agglomeration representing an advanced stage of urbanization in western China. As a nationally significant urban agglomeration in the western region, the Lanxi urban agglomeration also functions as a vital ecological barrier. Studying the supply and demand dynamics of ecotourism in this area is therefore important for the development of ecotourism and for promoting sustainable regional tourism practices. Based on this premise, this study focuses on the supply and demand conditions of new urbanization and ecotourism across 19 cities (districts and counties) within the Lanxi urban agglomeration. An evaluation index system is developed to assess the development of regional new-type urbanization and ecotourism supply and demand. The study employs methods including the coupling coordination model, the obstacle degree model, and the geographical detector to examine the comprehensive development levels of the regional systems, their spatial and temporal distribution characteristics, and the associated obstacle factors. The results indicate the following: (1) During the study period, the average comprehensive development level of the three major systems within the Lanxi urban agglomeration exhibited a steady upward trend overall. (2) Over the study period, the degree of coupling coordination among the three systems shifted from moderate imbalance (with coordination values between 0.2 and 0.3) to mild imbalance (with coordination values between 0.3 and 0.4). Significant regional differences and spatial clustering were observed in the coupling coordination levels. The overall spatial distribution of new-type urbanization and ecotourism supply and demand in the Lanxi urban agglomeration generally presented a pattern characterized by higher values in the central area (Lanzhou and Xining) and lower values in the peripheral areas. (3) The four dimensions represented by eight influencing factors-namely tourism scale, tourism benefits, tourism resources, and infrastructure-are identified as key determinants influencing the coupled and coordinated development of regional new urbanization and ecotourism supply and demand.

Tick-borne relapsing fever is an infectious disease caused by spirochetes of the genusBorrelia,leading to fever and spirochetemia in humans and animals.Borrelia theileri is the agent of bovine relapsing fever, which is transmitted by ixodid ticks of the genus Rhipicephalus and has been reported to cause mild disease in cattle. It has been infrequently reported in several countries over a broad geographic area; however, it has not been reported in Israel previously. Blood samples were collected from beef and dairy cattle in 23 locations in Israel. Real-time and conventional PCR were used to detectBorrelia spp.and co-infection with bovine piroplasmids. PCR products were sequenced,and phylogenetic analysis of the B. theileri sequences was performed. Positive cattle were identified in a limited number of locations with no clear spatial clustering. Borrelia theileri DNA was detected in 6 out of 439 (1.4%) blood samples. Of these, four were from beef and two from dairy cattle. No significant differences were found according to age (P = 0.642) and production type (P = 0.802) among the B. theileri-positive cattle. Phylogenetic analysis showed clustering of the cattle-derived isolate together with B. theileri amplified from ixodid ticks, supporting a potential transmission cycle involving cattle and Rhipicephalus spp. ticks. Theileria annulata and Babesia bigemina DNA was detected in 11 (4.1%) and one (0.4%) of 268 unvaccinated cattle, respectively. Co-infection with Theileria annulata was found in one beef and one dairy cattle. Tick-borne relapsing fever is present in the cattle population of Israel. Co-infection of B. theileri and bovine piroplasmids may drive clinical disease and financial losses in beef and dairy cattle. This is the first report of B. theileriinfection in Israel. Future control programs should address the exposure of both beef and dairy cattle to ticks and tick-borne diseases.

Chromatin organization underlies essential genome functions, but its nanoscale organization remains challenging to capture and quantify with precision. Atomic force microscopy (AFM) offers direct structural readouts of DNA and chromatin, yet translating these rich images into reproducible biological metrics has been limited by the lack of standardized, scalable analysis tools. Here we present DNAsight, an automated analysis framework that integrates machine learning-based segmentation with modular, base-pair-calibrated quantification of DNA spatial organization, looping, nucleosome spacing, and protein clustering. Applied across diverse chromatin-associated proteins, DNAsight reveals protein-specific organizational signatures, including topology-dependent compaction by integration host factor, condition-dependent changes in loop-like DNA structures in cohesin-CTCF-precocious dissociation of sisters 5A reactions, and promoter-driven multimerization of GAGA factor clusters. The framework further enables direct extraction of nucleosome spacing distributions from raw AFM images, providing a label-free route to investigate chromatin fiber architecture. Together, these advances establish DNAsight as a generalizable and scalable approach for converting AFM measurements into quantitative insights into the physical principles of chromatin organization.