While biodiversity is being reshaped across the globe, extinction risk assessments are lacking for most species, and a major challenge remains in understanding whether global threat status aligns with local population trends. Here, we assess whether population temporal prevalence trends are consistent with a species' global extinction risk, using over 60,000 populations of 2362 species across 978 marine and terrestrial assemblages (sampled for at least 20 years, mostly from temperate regions). We assign each population to one of five categories of temporal prevalence dynamics, and retrieve each species' extinction risk from the International Union for Conservation of Nature (IUCN) Red List. Fewer than 10% of local populations show consistent increasing or decreasing prevalence over time, with most exhibiting random patterns of temporal change, especially marine populations. Overall, higher extinction risk is associated with a higher frequency of decreasing local prevalence, and vice-versa for increasing prevalence, against a backdrop of complex links between extinction risk and local temporal dynamics. Our results suggest that directional changes in species local prevalence could be harbingers of future changes in global threat status, and highlight how leveraging assemblage monitoring data can aid conservation efforts and extinction assessments.
Most of our knowledge about the dog-human relationship comes from studies with dogs from 'WEIRD' (Western, Educated, Industrialized, Rich, and Democratic) societies. Here, we investigate cultural differences in dog-owner interactions worldwide. To achieve this, we developed a test battery comprising six well-established social-cognitive experiments and a questionnaire that assessed the psychological and practical aspects of the dog-human bond. We tested hunting dogs alongside their owners in five rural societies across culturally diverse locations in various countries: Vanuatu, Mongolia, Madagascar, Peru, and Germany. Despite dramatic cultural and environmental differences, we found that dog-human relationships were remarkably similar. Residual differences may be attributed to variations in hunting techniques and differences between WEIRD and non-WEIRD societies.
暂无摘要（点击查看详情）
Mechanical ventilation (MV) is a critical intervention for managing respiratory compromise and acute respiratory conditions. However, MV can also cause ventilator-induced lung injury (VILI) through mechanisms such as volutrauma, atelectrauma, and biotrauma. Conventional global pressure-volume (PV) loop analysis cannot capture regional mechanical heterogeneity within the lungs, which is central to amplifying VILI, particularly in the context of pre-existing lung injury. In this study, we employed four-dimensional computed tomography (4DCT) combined with propagation-based phase-contrast X-ray imaging to evaluate regional lung mechanics (overdistention and cyclic collapse) in a mouse model subjected to four MV protocols with varying peak inspiratory pressure and positive end-expiratory pressure. Regional PV loop analysis revealed significant overdistention in low tidal volume ventilation settings, which was not apparent in global PV loops, particularly in the group that was ventilated with a low tidal volume and zero PEEP (LVZP). Although global and regional collapse metrics showed no significant differences, regional overdistention indices were markedly higher than global measurements in LVZP, underscoring the limitations of global assessments. These findings highlight the utility of 4DCT for detecting subtle regional mechanical stress and support the integration of regional metrics into MV strategies to enhance lung protection and reduce the risk of VILI.
Global climate crisis and waste disposal costs drive the need for circular industrial models. This study investigates whether industrial symbiosis through co-disposal of papermaking waste and blast furnace slag can convert these materials from waste to resources. Using a system expansion Life Cycle Assessment framework, we assessed alkali-activated mortars based on global warming potential, water footprint, and toxic impacts. Results indicate that high-volume waste substitution significantly improves the material's environmental profile, achieving a net-negative Global Warming Potential of - 7.9 kg CO2 eq/m³ and a 129% net environmental benefit for human health damage compared to the baseline. These results occur because the avoidance of landfill-related greenhouse gas emissions and primary material production outweigh the impacts of chemical activation. This study outlines a structured approach to decarbonizing construction materials. It shows how technological innovation can strengthen competitiveness within circular economic systems. This work verifies the technical feasibility of regenerative material strategies and identifies activator optimization as a critical factor for advancing next-generation sustainable materials, thereby offering practical guidance to help industrial sectors meet global sustainability requirements.
Tropical rainforests, particularly the Amazon, function as the Earth's lungs yet absorb mercury (Hg) emitted worldwide. By introducing climate-driven variations in foliar functional traits into a global model of forest Hg uptake, we uncovered an inter-continental spatial decoupling between Hg sources and sinks. Unexpectedly, the minimally industrialized rainforests of South America and Africa exhibit the world's highest atmospheric Hg accumulation rates and greatest biomass, thus disproportionately sequestering Hg released from industrialized regions. This imbalance arises from climate-specific leaf traits that enhance Hg fixation towards lower latitudes. The model constrains global forest Hg uptake to 1155 ± 422 Mg yr-1, sharply reducing prior uncertainties (320-3138 Mg yr-1) and nearly equilibrating with global litterfall deposition (1180 ± 710 Mg yr-1). These findings urge a re-assessment of the Minamata Convention's effectiveness and highlight the vulnerability of tropical forests to anthropogenic Hg inputs and to climate-induced shifts in vegetation and terrestrial Hg reservoirs.
Pathological image classification is critical for early cancer diagnosis and precise subtyping. However, pathological images exhibit significant heterogeneity and complex textures. Existing methods often fail to fully exploit local details, while frequency-domain approaches lack effective inter-subband interaction, hindering the fusion of global context and local fine-grained information. To address this, this study propose a Multi-scale Frequency-domain Hybrid Attention mechanism (MFHA). MFHA uses wavelet transform to decompose images into low-frequency subbands (global structure) and high-frequency subbands (microscopic details). By integrating multi-scale convolutions with subband fusion, it enhances high-frequency feature representation, enabling effective joint modeling of global structure and local texture. This study further introduce a spatial attention module with cosine similarity and multi-dimensional statistics to improve feature robustness and discriminability. Experiments show that our method outperforms baselines on multiple pathological datasets, with accuracy gains of 1.97% and 2.71%. This work provides a novel perspective for pathological image feature co-modeling, boosting classification accuracy and robustness. The code is available at https://github.com/AnaStartz/WSI-processing-framework.
Model trees provide an appealing way to perform interpretable machine learning for both classification and regression problems. In contrast to "classic" decision trees with constant values in their leaves, model trees can use linear combinations of predictor variables in their leaf nodes to form predictions, which can help achieve higher accuracy and smaller trees. Typical algorithms for learning model trees from training data work in a greedy fashion, growing the tree in a top-down manner by recursively splitting the data into smaller and smaller subsets. This yields a fast algorithm, but the selected splits are only locally optimal, potentially rendering the tree overly complex and less accurate than a tree whose structure is globally optimal for the training data. In this paper, we empirically investigate the effect of constructing globally optimal model trees for classification and regression. The trees we consider feature linear support vector machines at the leaf nodes and are learned using mixed-integer linear programming (MILP) formulations. We use benchmark datasets to compare them to model trees obtained using greedy and dynamic programming-based algorithms, evaluating both tree size and predictive accuracy. We also compare to classic optimal and greedily grown decision trees, random forests, and support vector machines. Our results show that MILP-based optimal model trees can achieve competitive accuracy with very small trees. We also investigate the effect on the accuracy of replacing axis-parallel splits with multivariate ones, foregoing interpretability while potentially obtaining greater accuracy.
Global increases in jellyfish populations due to climate change, pollution, and overfishing have led to more encounters with humans, resulting in an uptick in jellyfish envenomation. Appropriate first-aid treatments have sparked ongoing debates among specialists regarding the use of topical chemical agents. Despite the establishment of various testing assays, issues with stability, scalability, suitability for on-site testing, and reproducibility remain. This research aims to develop a polymer-based artificial skin layer by evaluating four different polymer films-polypropylene, linear low-density polyethylene (LLDPE), a blend of polylactic acid and polybutylene adipate terephthalate (PLA/PBAT), and silicone rubber. These materials are being tested as potential alternatives to the traditional porcine intestine film used in the Tentacle Skin Blood Agarose Assay (TSBAA) for jellyfish envenomation testing. The films underwent tensile, puncture resistance, chemical stability, and permeability tests to assess their suitability. Mechanical changes were also studied after exposure to deionized water, salt water, 70% ethyl alcohol, vinegar (5% acetic acid), and baking soda slurry-chemicals that porcine tissue might encounter in TSBAA studies. Results showed that PP, LLDPE and PLA/PBAT silicone were able to form an ultra-thin film (39, 42, 47 μm MD and 38, 47, 45 μm TD, respectively) while silicone exhibits at least double the thickness of the other films. PLA/PBAT blend demonstrated a favorable balance of low puncture resistance (1.80 MPa)-approaching that of silicone rubber (0.84 MPa)-while maintaining superior chemical stability across tested solutions. Although silicone rubber exhibited the lowest puncture resistance, its relatively poor chemical durability and the highest environmental impacts up to 4-fold higher than other polymers, limit its suitability as a sustainable alternative. In contrast, PLA/PBAT films combined adequate puncture performance with improved environmental profiles compared to silicone rubber, supporting their potential as a more balanced and eco-friendly material for TSBAA applications. Life cycle assessment (LCA) further confirmed that PLA/PBAT and polypropylene films exhibited lower global warming potential (2.86 ⋅ 10- 4 kg CO2eq) and non-renewable energy use per functional unit (3.72 ⋅ 10- 3 MJ), underscoring their viability as sustainable alternatives for biomedical assay systems. This work serves as a proof-of-concept, laying a foundational step for developing standardized synthetic skin models. Future studies will focus on optimizing film properties and conducting rigorous validation using live tentacles or isolated cnidocytes, ultimately leading to a more efficient, sustainable, and widely applicable framework for mitigating jellyfish envenomation.
Land-use and land-cover change (LULCC) is a major source of anthropogenic CO₂ emissions, yet projections remain scarce. Here, we use the reduced-complexity Earth system model OSCAR to generate national LULCC carbon emission trajectories through 2100, across 150 socioeconomic and policy-relevant scenarios. Deforestation and forest regrowth dominate variability in LULCC carbon emission, with policy timing and ambition exerting strong control. Ending gross deforestation by 2030 yields large, persistent removals (about -30 Pg C by 2100), whereas net forest area balance still emits 4-9 Pg C. The strongest sinks are projected to emerge in China and Indonesia, while Brazil and the Democratic Republic of the Congo dominate global sources. The accompanying open dataset enables country-level scenario assembly and policy evaluation. Our findings underscore that early and ambitious land governance, particularly in tropical regions, is essential for transforming the land sector into a durable carbon sink aligned with global temperature goals.
Protein-protein interactions (PPIs) are dynamic and critical to adaptive homeostasis. While there have been massive efforts to catalogue proteome-wide PPIs, global quantification of changes remains a challenge. Here, we integrate dynamic protein correlation profiling - mass spectrometry (PCP-MS) and quantitative cross linking-mass spectrometry (qXL-MS) using multiplexed stable isotope labelling to characterise global PPI remodelling following the development of chronic skeletal muscle insulin resistance (IR) with or without acute insulin stimulation. We quantify >7,000 unique PPIs amongst 5,346 proteins and show changes in the interactome network dominate the proteome response. Our data show the dysregulation of protein processing in the endoplasmic/sarcoplasmic reticulum involving changes in PPIs with protein chaperones and disulfide isomerases is a major hallmark of skeletal muscle IR. Mechanistically, we show the dysregulation of PPIs with Protein-Disulfide Isomerase 6 (PDIA6) regulates cysteine oxidation and insulin sensitivity. Taken together, we show in vivo quantitative interactome mapping is a powerful approach to understand disease mechanisms and provide new insights into protein network re-organisations with IR.
Cross-modal retrieval, especially for image-text pairs, plays a vital role in areas such as medical image analysis, multimedia search, and personalized recommendations. However, challenges like modality heterogeneity, missing data, and noise still limit its effectiveness. Traditional methods often suffer from high computational complexity, while deep learning-based approaches struggle to capture long-range semantic relationships and handle missing modalities. To address these issues, we propose a graph-based Incomplete Modality Awareness Cross-Modal Retrieval method (GCMR-IMA) with four key innovations: (1) A dual-layer semantic architecture with image-text dual embedding for global semantic graph learning and long-range semantic capture; (2) A sparsified modality adjacency graph to model latent inter-modality relationships, improving modality association awareness; (3) A graph-guided missing modality awareness method, providing semantically consistent guidance for missing modalities using the global semantic graph's sparsified adjacency matrix; (4) An adaptive loss function that optimizes modality projections for context-aware missing modality detection. Extensive experiments demonstrate that GCMR-IMA continues to perform well in the presence of missing modalities. This framework enhances the robustness and effectiveness of cross-modal retrieval systems, supporting better decision-making in real-world applications.
Postoperative pneumonia (POP) is a common and severe complication following surgery for geriatric osteoporotic fracture (OPF), which warrants close attention as a major global public health issue.  With the growth of global life expectancy, the number of patients who suffer OPF is expected to increase dramatically. However, national data on the incidence of POP and its associated risk factors remain scarce. This study aims to investigate the incidence and independent risk factors of POP among elderly patients with OPF in China via a large-scale cross-sectional survey. From September to November 2023, elderly patients who underwent OPF surgery were recruited for this cross-sectional study via stratified random sampling and a self-designed questionnaire from 594 hospitals in China. Patients' data, including demographic characteristic, medical history and postoperative information, were analyzed with descriptive statistics, and intergroup comparisons were conducted between patients with and without POP. One-to-one propensity score matching (PSM) was applied to baseline clinical datawith a caliper value of  0.01. Independent associated factors related to POP for geriatric OPF were identified by using univariate analysis and multifactorial logistic regression. P < 0.05 was considered a statistically significant difference.A total of 12,496 participants were ultimately included. Among all the patients we investigated, the overall incidence of pneumonia following geriatric OPF surgery was 2.42%. Multifactorial logistic regression analysis showed that history of fracture (OR [odds ratio] 4.483, 95% CI [confidence interval] 2.824-7.117); history of surgery (OR 2.260, 95% CI 1.457-3.507); history of trauma (OR 2.866, 95% CI 1.461-5.625); pre-operative self-care ability (OR 3.254, 95% CI 2.177-4.864); aid of crutches/walkers/wheelchairs (OR 1.692, 95% CI 1.149-2.490); combined chronic obstructive pulmonary disease (COPD) (OR 4.634, 95% CI 1.889-11.369); length of stay (LOS) (OR 1.044, 95% CI 1.016-1.073) were independent risks factors predicting pneumonia for geriatric OPF surgery during hospitalization. Our study identified a non-negligible incidence of POP in elderly patients with OPF. For older adults with a history of fracture, surgery or trauma, comorbid COPD, poor preoperative self-care ability, dependence on walking aids, and longer hospital stays, clinicians should implement evidence-based pneumonia prevention protocols to optimize postoperative recovery.
Automated weed detection is essential for site-specific herbicide application, that can result into the reduced environmental footprint of conventional agriculture. However, for field deployment of automated weeding devices, occlusion remains a critical challenge that can weaken the precision of weed identification. Here, we compare the performance of Vision Transformers (ViT-B16 & PvTv2) and Convolutional Neural Networks (EfficientNet-B0 & ResNet-50) in accurate weed detection, using controlled synthetic occlusion levels (0%, 25%, and 50%). We found that ViT-B16 has superior occlusion resilience, with image testing accuracy increasing from 80% to 86% under 50% occlusion. In contrast, the testing accuracy of PvTv2, EfficientNet-B0 and ResNet-50 dropped from 45 to 76% under similar conditions. Multivariable regression confirmed architecture type as the dominant testing accuracy driver (p ≤ 0.001), with ViTs outperforming CNNs by an average of 14.56% points. These results suggest that occlusion resilience is not uniform across architectural variants but depends critically on attention-based design. Consequently, for real time deployable automatic weed detection systems, hybrid architectures that balance ViT global context with CNN computational efficiency represent a critical future direction. Such approaches can support precise herbicide application, reduce chemical inputs, and enable more sustainable crop protection through reliable AI-driven automation.
Topological entanglements are central to understanding and predicting the properties of polymer melts. Yet, they make equilibrium sampling computationally challenging, as decorrelation times grow rapidly with chain length. Here, we introduce a Monte Carlo scheme that bypasses typical computational bottlenecks by working in a self-assembly ensemble rather than at fixed composition. Strictly local moves efficiently propagate backbone reconnections across scales while conserving the number of linear chains, achieving near-linear scaling of decorrelation time with system size, τeq ~ V 1.0. With this method, formulated for a fully-packed lattice, we equilibrate periodic systems totalling up to  ≃ 1.1 × 109 monomers, accessing a universal melt regime insensitive to lattice details. We analyze intra- and inter-chain entanglements for chains of up to N ≃ 5 × 105 monomers, revealing that they manifest as localized knots and links rather than as global tangles. Finally, we show that the magnitude of the Gauss linking integral between neighbouring chains grows only as N1/4.
Adipocyte death is a key event in the development of white adipose tissue (WAT) inflammation, a major driver of obesity-associated metabolic dysfunction. Receptor-interacting protein kinase 3 (RIPK3) mediates necroptosis, a recently discovered mode of regulated necrosis. Necroptosis has been implicated in several inflammatory pathologies; however, the role of adipocyte necroptosis in obesity remains unclear. In the present study, we sought to investigate the role of adipocyte RIPK3 in obesity and glucose homeostasis. We demonstrated that necroptotic signalling was upregulated in WAT of mice with diet-induced obesity and was associated with body-mass index in human WAT. We also demonstrated that caspase-8, a central regulator of apoptosis, suppresses adipocyte necroptosis both in vitro and in vivo. Adipocyte-specific deletion of caspase-8 in mice reduced adiposity compared to control mice. This difference was not observed with concomitant global deletion of RIPK3. Furthermore, adipocyte-specific deletion of the RIPK3 receptor-interacting protein homotypic interaction motif (RHIM), which is required for necroptotic induction, did not influence weight gain, adiposity, or glucose homeostasis in mice with diet-induced obesity. Caspase-8 knockdown by siRNA or pharmacological inhibition in 3T3-L1 adipocytes suppressed adipogenesis, which may be independent of adipocyte Ripk3. Collectively, our findings suggest that adipocyte RIPK3 RHIM does not play a critical role in obesity and glucose homeostasis. Alternatively, we provide further evidence that caspase-8 plays an essential role in adipocyte differentiation, offering insight into the molecular mechanisms underlying obesity and metabolic dysfunction.
Dengue virus (DENV) continues to pose a significant global health threat, with increasing infection rates and limited treatment options. The viral NS2B-NS3 protease (NS2B-NS3pro), a highly conserved two-component enzyme essential for polyprotein processing and replication, is a key target for antiviral drug development. Here, we report the 2.25 Å-resolution crystal structure of DENV serotype 2 (DENV-2) NS2B-NS3pro, in which a PEG fragment is bound within a solvent-exposed pocket between β10, β11, β14 and β15 of NS3. This structure reveals a previously unrecognized solvent-exposed PEG-binding pocket, which may represent a putative ligand-binding surface for future validation and inhibitor-design studies. We also show that the glutathione-coated gold nanocluster (GSH-AuNC) directly inhibits DENV-2 NS2B-NS3pro. Bio-layer interferometry and fluorescence-based protease assays indicate that the nanocluster binds NS2B-NS3pro with a dissociation constant of 15.64 µM and inhibits its catalytic activity with an IC50 of 16.04 µM, consistent with a direct inhibitory effect at the in vitro enzymatic level. Molecular dynamics simulations further suggest that GSH-AuNC is predicted to interact with the catalytic triad of NS2B-NS3pro, forming stable electrostatic and van der Waals interactions that may interfere with substrate access. Collectively, these findings provide an in vitro structural and biochemical basis for targeting DENV-2 NS2B-NS3pro and may inform future development of small-molecule or nanomedicine-based inhibitors, although cell-based antiviral efficacy, cytotoxicity, and cellular uptake remain to be evaluated.
Mental health is now recognised as a major global concern affecting people from diverse backgrounds. There is growing evidence that the gut microbiota plays a crucial role by producing metabolites that significantly influence a person's mood and behaviour. Despite its importance, there is a significant gap in the profiling and understanding of the gut microbiota's influence on mental health among Malaysians. Therefore, this study aims to determine gut microbiome profiles among patients with major depressive disorders (MDD) of different treatment groups attending psychiatric clinics in the state hospitals and compare them to healthy individuals in the community of Klang Valley, Malaysia. This cross-sectional study will be carried out in Klang Valley, Malaysia. Eligibility of the patients will be assessed by the psychiatrists prior to recruitment of patients. Patients with MDD will be categorised into monotherapy and polypharmacy while healthy individuals will be used as a comparison group. Demographic data will be recorded. Stool samples will be subjected to DNA extraction and 16S rRNA gene-sequencing analysis to determine the microbial compositions of the gut microbiome. This study will be conducted following the procedure set by the National Medical Research Register Malaysia. The Medical Research Ethics Committee (MREC), Ministry of Health Malaysia, has ratified this study and granted ethical approval to conduct this study (NMRR ID-22-00893-JVW). All the participants will be given an information sheet and will sign a consent form to participate. Participants have the right to withdraw from the study at any time without an explanation. All information gathered will be used for research purposes only and treated as confidential. The results will be submitted to peer-reviewed journals for publication as well as presented at national and international conferences. Informing policy makers at all levels is a crucial aspect of the dissemination and will be done from local to international levels. This study was registered under National Medical Research Register Malaysia (NMRR ID-22-00893-JVW), a mandatory procedure of the Ministry of Health Malaysia before the start of a research.
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and arboviruses represent major global public health threats, each with distinct epidemiological and pathogenic profiles that often overlap in endemic regions. Reports of co-infection and cross-reactivity between these viruses highlight the need for deeper understanding of their interactions. Co-infections pose diagnostic challenges due to overlapping clinical symptoms and immune modulation, which may exacerbate disease severity. Moreover, antibodies generated from prior virus infections may influence the outcome of subsequent infections through mechanisms such as cross-reactivity and antibody-dependent enhancement (ADE). This review summarises current knowledge on the clinical manifestations and immunological consequences of co- and subsequent infections involving SARS-CoV-2 and arboviruses. Emphasis is placed on underlying pathophysiological mechanisms, including cross-reactivity and ADE, that shape host immune responses. Understanding these interactions is essential for improving diagnostic accuracy and guiding public health strategies to mitigate the risks associated with co-infections and sequential viral infections.
Climate change significantly undermines livestock productivity worldwide, with profound implications for milk production. Unprecedented climatic changes indiscriminately affect the dairy production across the global south with profound effects on the Trans-Gangetic plains, especially the high milk production tract of Haryana. Hence, this study employs panel data analysis to assess the impact of climatic variables, including annual minimum, maximum, and mean temperatures, heavy rainfall incidence, temperature-humidity index (THI), and potential evapotranspiration (PET) on milk yield and production in buffalo (Bubalus bubalis), indigenous cattle (Bos indicus), and crossbred cattle (Bos taurus × Bos indicus) in Haryana, India, from 2004 to 2019. The study was based on data spanning 16 years of Integrated Sample Survey in Haryana, adopting a stratified three-stage sampling design (districts → villages/households → animals) with complete livestock population enumeration. Annually, 1,148 villages were surveyed for recording milk yield and milk production. Crossbred cattle, indigenous cattle and buffalo had daily milk yields of 8.08, 5.23 and 7.64 kg/day with annual production of 788.24, 338.81 and 6045.02 MT, respectively. The yearly population estimates based on complete enumeration, viz., indigenous cattle (n₁ = 0.179 million), crossbred cattle (n₂ = 0.291 million), and buffalo (n₃ = 2.22 million). The findings reveal that high temperatures (> 38 °C) combined with elevated humidity (> 70%) during July and August significantly reduce milk production, whereas winter temperatures exhibit negligible effects. Notably, PET (p < 0.01 for May and June across all species) emerges as a critical climatic indicator alongside THI and heatwaves, necessitating its integration with solar radiation, ambient temperature, and vapour pressure in climate impact assessments. These results underscore PET's role in shaping adaptive strategies for sustainable livestock production amid global warming.