Pulmonary infections pose a significant global health challenge to human life and health. In patients with chronic pulmonary diseases such as cystic fibrosis and bronchiectasis, structural abnormalities of the airways and impaired mucociliary clearance contribute to recurrent and challenging pulmonary infections. These infections are frequently complicated by antimicrobial resistance, making them difficult to treat with conventional antibiotics. As a result, phage therapy has emerged as a promising alternative for treating resistant pulmonary infections. Recently, the integration of artificial intelligence (AI) has improved the efficiency of phage selection. Nevertheless, the accuracy of predicting phage-bacterial host interactions remains limited, posing a significant obstacle to the clinical translation of phage-based therapies. To address this issue, we propose a deep Siamese network framework for precision phage selection in pulmonary infections. Specifically, we employ an identical model architecture to process both phage and host genomes. Initially, the genomic sequences of both phages and hosts are encoded into feature representations using k-mer segmentation followed by the skip-gram model. Subsequently, convolutional neural networks (CNNs) and Transformers are introduced to extract local and global features, respectively. Finally, the extracted features are fused to predict phage-host interactions. Experimental results on dataset created from the NCBI genome database demonstrate that our proposed method achieves superior performance in the precise identification of phages targeting specific bacterial hosts, thereby supporting its potential application in phage therapy for pulmonary infections.

The evolution of dissolved organic matter (DOM) in urban stormwater runoff drives the migration and transformation of polycyclic aromatic hydrocarbons (PAHs). The sources of DOM and PAHs in stormwater runoff are complex, and research on the driving mechanisms behind their migration and transformation remains insufficient. In this study, the optical properties of DOM and the spatial source distribution of PAHs in urban stormwater runoff were systematically analyzed. The abundance of DOM components varied significantly across different urban functional zones, with DOM exhibiting a mixed characteristic of both autochthonous and allochthonous sources (FI = 1.4-1.9), while the contribution of local sources was limited (BIX < 1). PAHs concentrations were highest in traffic and vegetation-covered areas, primarily originating from traffic emissions. Total phosphorus (TP), chemical oxygen demand (COD), and pH were identified as the main driving factors for DOM components, with rainfall increasing the proportion of humic-like components in rivers. Furthermore, PAHs showed significant correlations with pH and COD. Finally, TP, pH, and fluorescence properties had a notable influence on the characteristics of the C3 and C4 components, while humic-like components (C1 + C2) directly and positively affected PAHs. This study provides theoretical support for the prevention and control of stormwater runoff pollution.

The 2024 Nobel Prizes in Physics and Chemistry have cemented artificial intelligence (AI) as a transformative force in scientific research, marking the mainstream emergence of AI for Science (AI4S). This editorial interrogates the core epistemological question facing the research community: whether AI-driven inquiry constitutes a fundamental paradigm shift in the scientific method or merely a transitional phase in computational data processing. We delineate the evolutionary tiers of AI capability in scientific discovery, namely, Keplerian, Edisonian, and Einsteinian AI, and analyze the pivotal shift toward agentic scientific systems that are redefining human-AI research collaboration. We further address the critical infrastructure gaps, ethical risks, and reproducibility challenges inherent to the mainstream adoption of AI4S, particularly within the environmental science disciplines. To uphold editorial rigor and scholarly integrity, we propose a framework for evaluating AI-enabled research contributions, emphasizing human oversight, transparent methodology, and responsible innovation. This editorial calls for rigorous, cross-disciplinary scholarship that harnesses AI's potential while safeguarding the core principles of scientific inquiry, positioning AI4S as a sustainable driver of discovery rather than a fleeting technological trend.

BACKGROUND:Arboviral infections impose significant public health challenges globally, yet routine surveillance typically captures only symptomatic infections, underestimating the true extent of exposure. Insights into how regional and demographic factors influence population immunity are essential for targeted surveillance and prevention, but such multidimensional insights remain limited. This study aimed to quantify population-level arboviral sero exposure and delineate the effects of regional and demographic factors on immunity to inform targeted surveillance and prevention. METHODS:We utilized a programmable phage display platform, ArboScan, which evaluates antibody binding to overlapping peptides that represent the proteomes of 691 human and zoonotic arboviruses. We profiled baseline antibody reactivity in serum samples from 400 healthy individuals, collected before the dengue outbreaks reported in Hainan in 2019. Antibody reactivity was quantified as normalized fold-change (FC) values relative to negative controls, and analyzed by region, sex, and age. Normality was assessed using the Shapiro-Wilk test. Two-group comparisons were conducted using independent two-sample t tests for normally distributed data or Mann-Whitney U tests otherwise; comparisons among > 2 groups were performed using One-way Analysis of Variance for normally distributed data. RESULTS:Regional ranking by mean product fold change (MPFC) showed northern enrichment for bluetongue virus (MPFC = 3.56), whereas southern cohorts were enriched for mosquito-borne arboviruses-dengue virus (MPFC = 3.54), Alagoas vesiculovirus (MPFC = 3.50), and Venezuelan equine encephalitis virus (MPFC = 3.39). Females exhibited higher FC than males for selected arboviral families (P < 0.001). By family-level analysis, Flaviviridae, Togaviridae, and Phenuiviridae showed no age-stratified differences (P > 0.05). High fold-change values were detected for non-arboviral viruses such as human cytomegaloviruses and human adenoviruses across all regions. CONCLUSIONS:Our findings reveal distinct regional and demographic patterns of arboviral antibody reactivity in China, reflecting differing histories of exposure and potentially informing region-specific surveillance strategies. The stable antibody levels across age groups, together with higher fold-change values in females, underscore the influence of biological and social factors on arboviral immunity. The ArboScan platform, and programmable peptide display platforms in general, offer a scalable approach to characterize population-level immunity and could enhance early detection and public health preparedness in arbovirus-endemic areas.

Fe-N-C single-atom nanozymes (SANzymes), which exhibit the properties of well-defined atomic structures and carefully controlled coordination environments, have become a hot research topic in biomedical fields. Unfortunately, the lower accessibility and intrinsic activity of the FeN4 sites severely limit their enzyme-like activity. Here, a densely exposed surface FeN4 structure was constructed on layered nitrogen-doped hierarchical porous carbon support through two steps of pyrolysis strategy. Using a honeycomb porous carbon support, the Fe-N-C catalyst boasted a high specific surface area with numerous Fe anchoring sites and was equipped with efficiently accessible active FeN4 structures. The Fe edge effect could modulate the electronic structure of individual Fe atoms, thereby boosting the intrinsic oxidase-like activity of the FeN4 molecules. As a result, Fe-N-C SANzymes were efficiently able to catalyze O2 with 3,3',5,5'-tetramethylbenzidine (TMB) as a substrate, achieving higher catalytic kinetic values than previously reported SANzymes. The colorimetric sensor using Fe-N-C SANzymes further detected uric acid (UA) with a wide detection range and a low detection limit. Then the visual sensing of the colorimetric system allowed the smartphone to identify colors by HSV patterns and obtain quantitative analysis. Moreover, the developed Fe-N-C colorimetric method showed satisfactory results in clinical samples, and proved to be a simple-operated and reliable method for detection of UA. This work not only highlights the advantages of the rationally designed edge effect of iron single atoms, but also presents the promising applicability of single-atom nanozymes in clinical diagnosis and related fields.

Predicting runoff water quality is crucial to mitigate non-point source pollution within the urban watershed. However, the complex physical runoff transportation process makes it difficult to predict effectively. This study proposed a flexible framework integrating hydrology-hydraulic datasets from the physical-driven model and machine learning networks to enhance prediction accuracy and efficiency. High-resolution measurement data was provided by online monitoring equipment installed in a highly urbanized watershed in the Pearl River Delta. Interpretable analysis derived by Shapley Additive Explanations (SHAP) approach was further utilized to determine the driving forces for predicting runoff water quality. Results show the average concentrations of chemical oxygen demand (COD), ammonia nitrogen (NH3-N), and suspended solids (SS) in the given water were 15.28 ± 2.84 mg/L, 2.63 ± 1.48 mg/L, and 12.02 ± 0.55 mg/L, respectively. Modeling comparison shows that the random forest networks performed the best among the given machine learning models. The R2 values for COD, NH3-N, and SS predictions were 0.78, 0.77, and 0.81, respectively. RMSE values were 0.58, 0.31, and 0.17, respectively. SHAP analysis revealed that precipitation, slope, and impervious areas ratio strongly affected the runoff water quality. The data presented herein shows the proposed modeling framework could capture the dynamic characteristics of pollutants in surface water.

Anthropogenic water pollution severely threatens human society worldwide, yet the water pollution induced by combined sewer overflow (CSO) remains unclear within climate change and urbanization. Hence, this study integrated the general circulation model (GCM) and shared socioeconomic pathway (SSP) projections with water quality modeling, to analyze spatiotemporal patterns and future trends of CSO-induced water pollution under changing environments. Results demonstrated that the given area (Dresden, Germany) encountered significant CSO-induced pollution, with 14,860 kg (95 % confidence interval, CI: 9,040-15,630 kg) of particulate matter (SS), organic compounds (COD, TN, TP), and pharmaceuticals (Carbamazepine, Gabapentin, Ciprofloxacin, Sulfamethoxazole) being discharged annually. Climate change and urbanization exacerbated the severity of CSO-induced pollution, causing the discharged pollutants to reach a maximum annual load of 34,900 kg (CI: 21,400-44,100 kg), with up to 82.19 % of organic compounds and 75.28 % of pharmaceuticals being discharged by the top 25 % of extreme CSOs. GIS-based spatial analysis indicated the regional heterogeneities of CSO-induced pollution, the high-frequency CSOs were predominantly located in highly-impervious areas, while the high-load discharges mainly occurred in densely-populated areas. Scenario analysis revealed stronger temporal variabilities of CSO-induced pollution in the future, with the seasonal anomalies of discharged loads ranging from -86.18 % to 76.89 %. In addition, pharmaceutical pollution exhibited significant uncertainties under changing environments, and the CI of discharged load expanded by up to 131.71 %. The methods and findings herein yielded further insights into water quality management in response to changing environments.

Creatinine serves as a crucial biomarker for various diseases, and the development of effective and stable nonezymatic sensor is essential due to the limitations of enzyme-based biosensors, such as poor conductivity and environmental instability. In this study, a composite material comprising a copper organic framework and MXene (Cu-MOF@MXene) is synthesized through a one-step reaction, enabling enzyme-free electrochemical sensing of creatinine. The Cu-MOF@MXene sensor demonstrates good linearity within the range of 20.0 mu M to 150.0 mu M and a low detection limit (LOD) of 20.069 mu M. A screen-printed electrode (SPE) coated with Cu-MOF@MXene is fabricated for monitoring urine and human serum samples. The results indicate that the SPE possesses high sensitivity, selectivity, and stability for nonenzymatic creatinine detection, highlighting potential advancement in clinical testing and medical diagnostics.

AbstractGlobalization has led to an increasing geographical separation of primary input, consumption and production, and consequently to a substantial transboundary transfer of air pollution and associated health burdens through international trade. Here, we develop an integrated framework to determine the consumption‐ and income‐based global atmospheric emissions, and quantify the drivers of associated health impacts from 2000 to 2015, and evaluate the impacts of international trade on PM2.5‐related deaths by hypothetical scenarios. Results show that consumption transferred more primary PM2.5 emissions (2.2 Mt, 23.5%) and caused more additional mortality (241,000 deaths) through international trade than primary input (emission: 1.1 Mt, 12.3%, mortality: 167,000 deaths) in 2015. Top three key sectors contributed to more than half of emission flow driven by consumption (commercial, construction, electrical and machinery) and primary inputs (commercial, petroleum, and mining). Health benefits of reduced emissions intensity, which avoided 1.4 million deaths, were largely offset by not only increases in consumption and primary input levels but also population vulnerability, resulting in the increase in mortality (0.8 million) from 2000 to 2015. Changes in primary input (1.2 million deaths) contributed more to the rise in health burdens than changes in consumption (1.0 million deaths). Hypothetical scenarios show that the participation of Western Europe in international trade contributed to the reduction in global health burden, while the USA gained health benefits from international trade. Accordingly, our findings provide profound suggestions for future policy decisions from different perspectives and demonstrate that optimizing global supply chain through cooperation would mitigate the PM2.5‐related health impacts.

Burkholderia pseudomallei (B. pseudomallei) is the pathogen of fatal tropical infectious disease melioidosis. This study screened sRNA BprsO from extracellular vesicles (EVs) of B. pseudomallei as a new biomarker and constructed a lateral flow strip assay (LFSA) with dual signal amplification for its visual biosensing by coupling catalytic hairpin assembly (CHA) with MXene@PtCu nanozymes. The platinum nanoparticles (Pt NPs) in the nanozymes provided a large number of active sites for the assembly of nucleic acid and antibodies, and enhanced the catalytic activity of Cu NPs through their synergistic effect. The large specific surface area of MXene avoided the aggregation of Pt and Cu NPs for improving the signal output. Moreover, CHA technology not only recycled the target, but also successfully introduced biotin-labeled probes to the nanozymes surface, ultimately converting the nucleic acid concentration into easily accessible visual detection signals. Benefiting from the twofold amplification of MXene@PtCu and CHA, the proposed colorimetric assay exhibited superior performance for BprsO detection with a dynamic range from 50 aM to 5nM, and a detection limit of 5 aM. Notably, the method was successfully implemented in the context of LFSA, thereby illustrating its capacity for the visual detection of BprsO at a concentration as low as 15 fM and achieving a detection sensitivity of 1.26 fM. This work validates that the screened BprsO can serve as a specific biomarker for B. pseudomallei, and proposes a strategy with high sensitivity and specificity that offers a robust visual platform for early diagnosis and monitoring of B. pseudomallei.

Oxazolidinones, novel synthetic antibacterials, inhibit protein biosynthesis and show potent activity against Gram-positive bacteria, including Mycobacterium tuberculosis (MTB). In this study, we aimed to compare the in vitro activity of linezolid (LZD) and four oxazolidinones, including tedizolid (TZD), contezolid (CZD), sutezolid (SZD), and delpazolid (DZD), against multidrug-resistant tuberculosis (MDR-TB) and pre-extensively drug-resistant tuberculosis (pre-XDR-TB) isolates from Hainan. We established their epidemiological cut-off values (ECOFFs) using ECOFFinder software and analyzed mutations in rrl (23S rRNA), rplC, rplD, mce3R, tsnR, Rv0545c, Rv0930, Rv3331, and Rv0890c genes to uncover potential mechanisms of oxazolidinone resistance. This study included 177 MTB isolates, comprising 67 MDR and 110 pre-XDR-TB isolates. Overall, SZD exhibited the strongest antibacterial activity against clinical MTB isolates, followed by TZD and LZD, with CZD and DZD showing equivalent but weaker activity (SZDMIC50 = TZDMIC50 < LZDMIC50 < CZDMIC50 = DZDMIC50; SZDMIC90 < TZDMIC90 = LZDMIC90 < CZDMIC90 = DZDMIC90). Significant differences in MIC distribution were observed for TZD (p < 0.0001), CZD (p < 0.01), SZD (p < 0.0001), and DZD (p < 0.0001) compared to LZD but not between MDR-TB and pre-XDR-TB isolates. We propose the following ECOFFs: SZD, 0.5 µg/mL; LZD, TZD, and CZD, 1.0 µg/mL; DZD, 2.0 µg/mL. No statistically significant differences in resistance rates were observed among these five drugs (p > 0.05). We found that eight MTB isolates (4.52% [8/177]) resisted these five oxazolidinones. Among these, only one isolate, M26, showed an amino acid substitution (Arg79His) in the protein encoded by the rplD gene, which conferred cross-resistance to TZD and CZD. Three distinct mutations were identified in the mce3R gene; notably, isolate P604 displayed two insertions that contributed to resistance against all five oxazolidinones. However, no significant correlation was observed between mutations in the rrl, rplC, rplD, mce3R, tsnR, Rv0545c, Rv0930, Rv3331, and Rv0890c genes with oxazolidinone resistance in the clinical MTB isolates tested. In summary, this study provides the first report on the resistance of MTB in Hainan to the five oxazolidinones (LZD, TZD, CZD, SZD, and DZD). In vitro susceptibility testing indicated that SZD exhibited the strongest antibacterial activity, followed by TZD and LZD, while CZD and DZD demonstrated comparable but weaker effectiveness. Mutations in rplD and mce3R were discovered, but further research is needed to clarify their role in conferring oxazolidinone resistance in MTB.

OBJECTIVES:Tacrolimus has been a cornerstone of immunosuppressive therapy over the past two decades. Due to its narrow therapeutic window and pharmacokinetic variability, drug monitoring is vital for enhancing the efficacy and safety during therapy. In the present study, we evaluated the analytical performances of the MAGLUMI® Tacrolimus assay based on chemiluminescent immunoassay (CLIA), and compared with LC-MS/MS and the previously validated ARCHITECT Tacrolimus assay based on chemiluminescent microparticle immunoassay (CMIA). METHODS:We assessed the precision, limit of blank (LoB), limit of quantification (LoQ), limit of detection (LoD) and linearity of the MAGLUMI® Tacrolimus assay using patient whole blood samples. Interference was assessed by introducing potential interferents into clinical samples. We also analyzed the correlation and agreement with the gold standard method (LC-MS/MS) and another previously validated high-performing ARCHITECT Tacrolimus (CMIA) assay by including 125 whole blood samples from patients and 44 spiked samples. RESULTS:MAGLUMI® Tacrolimus (CLIA) assay exhibits superior precision, as coefficients of variation (CVs) for reproducibility and between-run precision were 0.55-3.63 % and 2.18-5.14 %, respectively. The LoB and LoQ were 0.1 μg/L and 0.5 μg/L. All samples in LoD verification had tacrolimus concentrations above LoB. The assay exhibited excellent linearity (r=0.99990, 0.5-50 μg/L) with no interference. Additionally, the results of the MAGLUMI® Tacrolimus (CLIA) assay showed strong correlation and concordance with LC-MS/MS and the CMIA assay. CONCLUSIONS:The MAGLUMI® Tacrolimus (CLIA) assay has excellent performance and strong concordance with LC-MS/MS and the ARCHITECT assay, making it a good alternative for tacrolimus measurement.

Floods threaten human lives globally, yet the flood risk to the elderly (above 65) remains uncertain within warming climates and population aging. Hence, this study incorporated the General Circulation Model and Shared Socioeconomic Pathway projections into the hydraulic modeling framework, to analyze the flood risk to the elderly in Europe under climatic and socioeconomic changes. Results demonstrated that central Europe has experienced an increase in both surface runoff and streamflow (exceeding 50%), which have jointly contributed to intensified flooding in the major basins (Loire, Rhine, Elbe, and Danube) within the region. Among them, the Elbe basin exhibited a significant increase in 100-year flood peak (similar to 107%) and elderly population (similar to 15%), resulting in 51,300 (CI: 45,300-60,500) of the elderly population being exposed to high-hazard floods under a high greenhouse gas scenario (SSP5-8.5), with at most 58% (29,800, CI: 25,100-33,700) of them being densely settled or in low-& middle-income groups. Among aging cities, Prague was severely affected by floods, with 40%-54% of the exposed elderly in high-risk areas. Followed by Dresden and Hamburg, where up to 18% of the exposed elderly were threatened by high-risk floods. This study revealed regional inequalities induced by flood exposure within the context of warming climates and population aging. The methods and findings are expected to provide additional insights into sustainable flood risk management under global change.

Urban rainwater runoff is an important source of nonpoint source pollution due to its transport of diverse contaminants, including polycyclic aromatic hydrocarbons (PAHs) and chlorinated derivatives. Importantly, these chlorinated polycyclic aromatic hydrocarbons (Cl-PAHs) exhibit elevated toxicological potential compared to their non-halogenated parent compounds. In this study, we proposed an approach that combined multivariate receptor model with integration of SHapley Additive exPlanations and Random Forest model. This method identifies the possible sources and reveals the impact of source apportionment results and environmental driving factors (such as geographical and meteorological data) on pollutant concentrations. Sixteen PAHs and nine Cl-PAHs were detected in 79 runoff samples from all three sites. The ∑16PAHs average concentration (2923.93 to 6071.83 ng/L) was significantly higher than the ∑9Cl-PAHs (384.34 to 1314.73 ng/L). The source apportionment was conducted by positive matrix factorization (PMF), and six potential pollution sources for PAHs and three for Cl-PAHs were quantified. PAHs primarily originate from the combustion of fossil fuels such as traffic, industrial emissions and coal tar, while Cl-PAHs are mainly derived from atmospheric deposition and industrial emissions. Meanwhile, the self‑organizing map classified PAHs and Cl-PAHs into 2 and 3 groups, respectively. The k-means algorithm yielded 4 clusters for runoff samples. Among machine learning models, Random Forest (RF) demonstrated optimal predictive performance and integrated with SHapley Additive exPlanations (RF-SHAP) revealed the effects of driving factors on the predicted concentration of PAHs and Cl-PAHs in urban runoff samples.

Circulating tumor cells (CTCs) carry intact tumor molecular information, making them invaluable for personalized cancer monitoring. However, conventional capture methods, relying on passive diffusion, suffer from low efficiency due to insufficient collision frequency, severely limiting clinical utility. Herein, a magnetic micromotor-functionalized DNA-array hunter (MMDA hunter) is developed by integrating enzyme-propelled micromotors, magnetic nanoparticles, and nucleic acid aptamers into distinct functional partitions of a DNA tile self-assembly structure. This design ensured independent and compatible running of autonomous propulsion, targeted recognition, and magnetic enrichment, enabling efficient capture and subsequent identification of CTCs in clinical blood samples. The autonomous motion of the MMDA hunter is powered by O2 bubbles generated through the dual enzymatic cascade reactions of glucose oxidase and catalase under physiological glucose conditions. Compared to static Fe3O4 arrays (without micromotors), the MMDA hunter shows more than 2-fold improvement in capture efficiency. Meanwhile, it achieved superb precision, simple operation, rapid response, high biocompatibility, excellent stability, and superior specificity for CTC enrichment. This method provides a reliable tool for tumor diagnosis in multiple clinical application scenarios, even in primary medical care, simultaneously offering a clever solution for the bottleneck of functional-module interference in multifunctional nanomaterials.

DNA nanostructure, as polymeric material with remarkable molecular recognition property, has been widely used in bioassay. However, it still faces some challenges to overcome complexity of signal-amplified strategies and to realize efficient separation of reaction products. Herein, we present an innovative signal-amplified approach by integrating the toehold-mediated strand displacement reaction with DNA tile self-assembly technology to construct superparamagnetism-functionalized DNA polymeric materials, establishing a new signal-amplified analytical strategy for nucleic acids. This strategy enables highly sensitive, rapid, and efficient nucleic acid detection, making it a promising candidate for point-of-care testing (POCT). The analytical performance of this strategy was validated using target DNA (tDNA) and PIWI-interacting RNA-36026 (piRNA-36026), achieving limits of detection (LOD) of 2.4 × 10-10 M and 2.7 × 10-10 M. Moreover, it successfully detected single-base mutations and demonstrated stability over seven days. Comparative experiments confirmed the superior signal-amplified efficiency of DNA arrays. Recovery experiments yielded recoveries of 88.53 %-101.89 % for tDNA and 87.58 %-108.61 % for piRNA-36026. Ultimately, the feasibility of this strategy for real-world applications was validated through detecting piRNA-36026 in cell lysates. In conclusion, this work introduces an innovative and efficient signal-amplified method, while expanding the application prospects of multifunctional DNA polymeric materials in biomedical diagnostics.

Creatinine is an important indicator of renal function, and its accurate and efficient detection is important to clinical diagnosis and disease monitoring. As the structure combining MXene layers and gold nanoparticles (AuNPs) provides many active sites and hot spots for surface-enhanced Raman scattering (SERS), ultrasensitive detection of creatinine can be realized. Herein, homogeneous ultrathin Ti3C2-MXene films are produced on a large scale by interfacial self-assembly, and a uniformly distributed AuNP monolayer and three-dimensional ribbon-like AuNP assemblies (RAuNPs) are then separately assembled on the MXene surface by simply regulating the nanoparticle concentration and the drying temperature. The two SERS substrates of AuNPs/MXene and RAuNPs/MXene are capable of sensitive detection of creatinine in aqueous solutions without labels. The SERS activity is verified using Rhodamine 6G (R6G). In the SERS detection of creatinine, the AuNPs/MXene substrate shows a linear range from 1 × 10-4 to 1 × 10-8 M with a limit of detection (LOD) of 287 nM, whereas the RAuNPs/MXene substrate exhibits a good linear relationship in the range from 1 × 10-3 to 1 × 10-10 M with a LOD of 2.64 nM. The outstanding SERS properties of RAuNPs/MXene suggest promising potential pertaining to the rapid and sensitive detection of creatinine.