Search Results (8,959)

Background: Preterm premature rupture of membranes (PPROM) before 32 weeks of gestation with prolonged latency is associated with substantial neonatal morbidity, including Dry Lung Syndrome (DLS), pulmonary hypoplasia (PH), bronchopulmonary dysplasia (BPD), and death. Accurate individualized risk stratification remains elusive, as the interacting contributions of amniotic fluid dynamics, inflammatory status, and microbiological burden are inadequately captured by traditional statistical approaches. Methods: We performed a retrospective, exploratory–predictive analysis of 66 pregnancies complicated by second-trimester PPROM with latency exceeding 14 days. Elastic Net and Random Forest models were trained across six clinically defined predictor domains using a multi-stage block modelling strategy. To address the clinically relevant distinction between antenatal and postnatal information, results are reported separately for Model A—comprising exclusively antenatal predictors available during expectant management (gestational age at PPROM, latency, amniotic fluid trajectory, inflammatory status, vaginal microbiome at admission)—and Model B, which additionally incorporates postnatal variables and characterizes the full mechanistic perinatal risk trajectory. Binary and ordinal outcomes included DLS, PH, BPD, intraventricular hemorrhage (IVH), and neonatal death. Pairwise interaction models were additionally computed to identify cross-domain risk constellations. Results: Distinct predictor architectures emerged per outcome. Pulmonary hypoplasia was most strongly associated with temporal features of oligohydramnios—particularly the persistence and timing of SDP < 1 cm—rather than isolated measurements. For DLS, the antenatal model (Model A) achieved AUC 0.776, driven by gestational maturity and inflammatory status; surfactant administration—a postnatal variable reflecting therapeutic response rather than an antenatal risk factor—dominated only the mechanistic Model B. Neonatal death was driven by a combined profile of respiratory support burden, amniotic fluid persistence, and co-morbidity. IVH showed consistently high ordinal predictability (accuracy 0.863), with amniotic fluid dynamics and microbiological burden as leading contributors. BPD remained the least linearly separable endpoint across all configurations. Conclusions: Multi-domain machine learning reveals outcome-specific, cross-domain risk architectures following second-trimester PPROM that are invisible to conventional statistical models. Longitudinal amniotic fluid trajectory is the dominant antenatal determinant of structural pulmonary morbidity, while microbiological burden independently shapes neurological risk. These findings support prospective validation of integrated ML-based risk stratification tools for individualized antenatal counselling in this high-risk population. Full article

Cultivated-meat production relies on robust animal cell-line engineering, scalable tissue-engineering strategies, and clearly defined regulatory standards. This review examines the developmental pipeline from primary tissue biopsy to large-scale expansion and regulatory evaluation, focusing on stable and safe immortalized cell platforms. We compare muscle satellite cells, mesenchymal stromal/adipogenic progenitors and induced pluripotent stem cells, highlighting trade-offs among proliferative capacity, lineage commitment, genomic stability, and food-safety considerations. We then analyze immortalization strategies, including spontaneous senescence bypass, telomerase reactivation and CRISPR-based checkpoint modulation, highlighting their impact on genomic stability and food-safety risks. Recent advances in serum-free media, extracellular matrix-mimetic biomaterials and staged co-culture protocols have enabled centimeter-scale tissues with improved texture and marbling; however, cost, reproducibility and scalability remain bottlenecks. Integrating multi-omics surveillance with life-cycle assessment reveals that environmental benefits (land, water and antibiotic reduction) are attainable only when energy inputs and growth-factor sourcing are optimized. Finally, we examine regulatory frameworks that distinguish food-grade immortalized cells from pharmaceutical substrates and genetically modified crops. By integrating cell biology, animal biotechnology, and bioprocess engineering, this review identifies technical priorities for advancing cultivated meat from laboratory development to industrial implementation, positioning genomic monitoring as an essential framework for assessing biological stability, functional predictability, and food-production suitability. Full article

Equipment fault prediction in maglev train systems poses substantial challenges: fault events are inherently rare, class distributions are severely imbalanced, and individual equipment units are subject to complex spatial and functional couplings that single-device statistical approaches fundamentally cannot capture. To address these challenges, this study proposes a Knowledge Graph-driven Graph Neural Network (KG-GNN) framework. A fault knowledge graph encompassing equipment, fault, temporal, and environmental entities is constructed to unify multi-source maintenance data. Graph connectivity is established via three spatial relation types (co-location, co-zone, and co-level), with edge weights derived from Laplacian-smoothed Lift scores quantifying fault co-occurrence strength. A two-layer GATv2Conv-based graph attention network is designed: the first layer employs four-head attention with explicit edge-weight integration to capture heterogeneous neighborhood influences, while the second layer produces compact node embeddings via single-head attention. A Top-20 sparsification strategy suppresses weak-association noise, and training under severe class imbalance is stabilized through Focal Loss and F2-Score-guided early stopping. On the test set, the proposed method achieves an F2-Score of 0.5703, Recall of 0.6825, and AUC-ROC of 0.9329 (single-run evaluation); multi-seed evaluation (5 seeds) yields F2 = 0.5645 ± 0.0035, Recall = 0.6789 ± 0.0095, and AUC-ROC = 0.9298 ± 0.0026, outperforming the MLP baseline by 18.3% in F2-Score and substantially exceeding GCN (F2 = 0.1476 ± 0.0176) and GATConv (F2 = 0.4284 ± 0.0097). Ablation studies confirm the individual contributions of authentic graph topology, precise edge weighting, and graph sparsification to overall performance. Full article

This review examines the emerging role of machine learning (ML) in pharmaceutical chemistry, with emphasis on molecular design, synthetic feasibility, and structure–property–performance (SPP) relationships. By enabling pre-synthesis prediction of physicochemical properties, reaction pathways, and pharmaceutical performance, ML can reduce empirical trial-and-error experimentation and support more efficient exploration of chemical space. A structured narrative review design with PRISMA-aligned systematic search elements was used to evaluate 101 studies, enabling transparent literature identification, eligibility screening, and thematic synthesis across heterogeneous ML applications in pharmaceutical chemistry. This review examines structure–property relationships (SPRs) and property–performance relationships (PPRs), with emphasis on key pharmaceutical endpoints such as solubility, permeability, stability, dissolution, and bioavailability. An integrated SPP framework is proposed to connect molecular structure, intermediate properties, and final performance outcomes while incorporating retrosynthetic analysis and experimental feedback and closed-loop optimization. Recent frontier developments are also discussed, including molecular foundation models, multimodal language–graph models, diffusion-based molecular generation, E(3)-equivariant models, and MolMIM-like latent-space optimization. This review also covers co-folding and joint ligand–protein modeling, Boltz-2-like affinity prediction, AlphaFold 3-related biomolecular interaction modeling, and absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction. Key limitations include dataset leakage, benchmark inconsistency, assay variability, conformational and protonation-state effects, reproducibility challenges, regulatory constraints, and the gap between computational prediction and prospective experimental validation. Future progress is expected to depend on hybrid physics–ML models, uncertainty-aware prospective validation, autonomous experimentation, explainable artificial intelligence, and sustainability-aware molecular design. Overall, ML is evolving from a predictive tool into a chemically informed decision-support framework for rational, synthesis-aware, and experimentally validated pharmaceutical development. Full article

This study characterizes AP-20-A, a lytic podovirus infecting Sinorhizobium meliloti, isolated from agricultural chernozem. Its 49.4 kbp genome shows negligible intergenomic similarity with known rhizobiophages (<2%). Core structural proteins—the major capsid protein (MCP) and terminase large subunit (TerL)—show closest homology to podoviruses infecting Paenibacillus, rather than to alphaproteobacterial viruses, suggesting cross-phylum horizontal gene transfer. This exchange is ecologically plausible, as Paenibacillus and Sinorhizobium co-exist in the rhizosphere. Over 63% of predicted proteins are functionally uncharacterized, with structural homologs detected in bacteria, archaea, and eukaryotes. We report the first identification in a rhizobiophage of a Tad2-like domain, predicted to block the bacterial Thoeris type II anti-phage defense. AP-20-A infected 56% of native S. meliloti strains; agrocenose isolates showed higher resistance than phytocenose isolates, evidence of local co-evolution. Among susceptible strains, 60% entered putative pseudolysogeny (with one strain exhibiting growth stimulation), whereas a symbiotically elite inoculant strain was completely lysed within hours. Some host strains carry additional AbiE systems; whether these independent defense–counterdefense layers interact during infection remains unknown. We conclude that resident phages represent a selective force that can disrupt inoculant establishment, underscoring the need to integrate soil virome assessment into agricultural microbiome management. Full article

A regeneration process for powdered spent activated carbon was developed through binder-assisted forming and thermal regeneration, and the process parameters were optimized by using response surface methodology (RSM). The effects of calcination time, calcination temperature, and binder ratio on the iodine adsorption value of spent activated carbon were investigated by using a Central Composite Design. The quadratic regression model exhibited high accuracy and statistical significance (R² = 0.9934, p < 0.0001), indicating good agreement between the predicted and experimental results. The optimal regeneration conditions were determined as a calcination time of 39.2 min, a calcination temperature of 848 °C, and a binder ratio of 10.15%. Under the optimized conditions, the VOCs (dichloromethane) adsorption capacity increased sharply from 9.1 mg/g to 108.2 mg/g. Characterization results showed that the regeneration process effectively restored the pore structure and improved the surface properties of the activated carbon. Thermogravimetric analysis demonstrated the effective removal of adsorbed pollutants. XPS analysis revealed a decrease in oxygen-containing functional groups, particularly –COOH groups, accompanied by an increase in C=O content, while the C–O groups changed only slightly. These structural and surface modifications contributed to the recovery of adsorption performance and provide guidance for the regeneration of powdered spent activated carbon. Full article

For the first time in this study, the ability for methylotrophic growth on methanol was demonstrated in representatives of the genus Sphaerotilus. The analysis of 20 genomes and the physiological verification of genomic predictions regarding C1 compound metabolism were carried out using Sphaerotilus montanus HS^T, Sphaerotilus hippei DSM 566^T, and Sphaerotilus sulfidivorans D-501^T as model strains. Genes involved in the direct oxidation of methanol to carbon dioxide were identified, including the lanthanide-dependent methanol dehydrogenase XoxF, the NAD-dependent methanol dehydrogenase Mdh2, genes of the tetrahydromethanopterin (H₄MPT) and tetrahydrofolate (H₄F) pathways, and the NAD-dependent formate dehydrogenase. In addition, a number of genes associated with C1 assimilation were identified, including genes of the Calvin–Benson–Bassham cycle and the incomplete serine cycle. Experimental data suggest that the bacteria are capable of using two strategies of methylotrophic growth: methanol oxidation via the lanthanide-dependent methanol dehydrogenase XoxF and the H₄MPT pathway, as well as oxidation via the NAD-dependent methanol dehydrogenase Mdh2 and the H₄F pathway. Both strategies provide CO₂ assimilation via the Calvin–Benson–Bassham, but additionally the second strategy demonstrates additional involvement of the incomplete serine cycle in the process of the C1 compounds. A hypothetical model of C1 compound assimilation in representatives of the genus Sphaerotilus was constructed. Full article

This bibliometric study maps olive (Olea europaea L.) yield prediction research as a coherent scientific domain for the first time. A Scopus query (27 February 2026) yielded 84 peer-reviewed articles (2002–2025), from which co-authorship network analysis, Bradford’s and Lotka’s Laws, Latent Dirichlet Allocation topic modelling (LDA), and OLS regression on citation counts were applied. Publication output increased nearly fourfold across three periods: 1.7 articles yr⁻¹ (2002–2014), 4.4 yr⁻¹ (2015–2019), and 6.7 yr⁻¹ (2020–2025). The 84 articles involve 382 authors, 61 journals, and 1551 citations (H-index = 22). Network analysis reveals a concentrated Spanish–Italian co-authorship axis. OLS regression (adj. R² = 0.267) identifies article age and abstract length as the only significant citation predictors, consistent with cumulative exposure time and study scope as structural drivers. Term-frequency screening against 18 a priori concepts finds that transfer learning, federated learning, hyperspectral imaging, digital twins, and SHAP-based explainability are absent or marginal. The field is producing more papers than ever on a narrowing methodological base geographically concentrated in the Mediterranean basin. Priority gaps—explainable AI, multi-region datasets, sensor-fusion pipelines, and federated data infrastructure—align directly with European Farm to Fork and Horizon Europe objectives. Full article

This study presents a numerical investigation of biogas–natural gas co-firing for greenhouse heating, integrating lumped-parameter energy balance, multi-objective optimization, and life cycle assessment (LCA) for a Syrian coast case study (48 dairy cows, 100 m² greenhouse). Five blends (0–100% biogas) were evaluated using a zero-dimensional model implemented in MATLAB R2024a (The MathWorks, Inc., Natick, MA, USA) and verified with Python (version 3.11, Python Software Foundation, Beaverton, OR, USA). The 70% biogas–30% natural gas blend exhibited the most favorable trade-off among conditionally feasible scenarios (requiring external biogas sourcing) with a model-predicted system thermal efficiency of 84.5% (LHV basis) and a model-estimated thermal NOx reduction of 75–85%, which represents a mathematical extrapolation beyond the experimentally validated range of 0–50% biogas and excludes prompt NOx (5–20% of total) and should be interpreted as an indicative trend requiring experimental confirmation. For self-sufficient operation using only on-site biogas production (24 m³ day⁻¹), the maximum achievable blend is 32% biogas, offering a 13.8% cost reduction and a 13.5% GWP reduction. Pure biogas achieves a 41.5% GWP reduction and 48.5% lower daily operating costs under the assumption of expanded on-site production capacity but requires 3.3 times the current production volume. Multi-objective optimization reveals stakeholder-specific optima ranging from 50% to 91% biogas, with a robust compromise region of 65–75%. All predictions for NOx emissions above 50% biogas are mathematical extrapolations requiring experimental validation. For farms without access to external biogas markets, the 32% blend (self-sufficient optimum) is the currently implementable solution, offering a 13.8% cost reduction. For farms with access to regional biogas markets, the 70% blend represents the conditional techno-economic optimum, achieving a 15.3% cost reduction but requiring 29.12 m³ day⁻¹ of external biogas procurement. Full article

This study investigates the impact of globalization, energy consumption, economic growth, urbanization, and trade openness on environmental degradation in case of Pakistan. The dynamic autoregressive distributed lag (DARDL) technique was applied to measure the short- and long-term estimates, which are robust and have higher predictive power compared to the conventional ARDL technique with a dataset that spans from 1980 to 2023 for Pakistan. The empirical findings confirm a significant long-run cointegrating relationship among the variables under investigation. Specifically, globalization and trade openness are found to exert a significant negative impact on CO₂ emissions, suggesting their potential role in fostering environmental amelioration. Conversely, energy consumption and economic growth demonstrate a significant positive impact on CO₂ emissions, indicating their contribution to increased environmental degradation. Notably, urbanization does not show a significant relationship with CO₂ emissions, a finding attributed to Pakistan’s inefficient and environmentally unfriendly urban infrastructure, despite a substantial 2.93 percent annual rate of urbanization. Furthermore, the widely posited Environmental Kuznets Curve (EKC) hypothesis is not supported by the data; instead, economic growth appears to be associated with an exponential increase in CO₂ emissions. The findings offer crucial policy insights for sustainable development initiatives for Pakistan and generally for developing nations. Full article

Modern computational biology is increasingly applied in preclinical studies, and mathematical models can provide valuable insights into biological system behavior. Numerical modeling tools can significantly and rapidly help in predicting the cellular response to different treatments, numerous newly synthesized compounds tested on different model systems. This study is devoted to the application of a biphasic numerical model to explain and predict the behavior of colorectal carcinoma cell lines in investigated co-treatments. The model was used to estimate parameters related to cell proliferation and death and to predict cellular behavior through the determination of treatment efficiency and effectiveness. This study showed that the experimental results can be mathematically confirmed, and the data for the most effective treatment can be obtained. The most efficient co-treatment concentration was identified by the model as the condition associated with the lowest proliferation-related parameter and the greatest reduction in cell viability. The model indicated that the most efficient concentration does not appear to induce a rapid adaptive cellular response and may therefore represent a suitable candidate for subsequent treatment cycles. The model suggests that the investigated treatments may have limited therapeutic potential in both cell lines due to the sustained viability of rapidly proliferating cells and evidence of continued de-differentiation. Full article

Bacterial communities play vital roles in coastal biogeochemical cycling and ecological stability. Despite their importance, a significant knowledge gap exists regarding their spatiotemporal dynamics and assembly mechanisms in the tropical coastal waters of the Leizhou Peninsula, China. To investigate the bacterial community structure, co-occurrence networks, and assembly processes, we conducted 16S rRNA gene amplicon sequencing on water samples collected seasonally from August 2022 to June 2023. The bacterial communities were dominated by Proteobacteria and Cyanobacteria, and were characterized by a distinct warm-season peak in the relative of Cyanobium. Alpha diversity indices exhibited significant seasonal fluctuations, reaching a minimum in August (autumn) and a maximum in December (winter). These variations were strongly regulated by water temperature and phosphate availability. Redundancy analysis (RDA) identified salinity as the primary deterministic factor shaping community composition. Seasonal environmental heterogeneity, rather than spatial variation, primarily governed bacterial community dynamics. We also observed a seasonal succession in community assembly mechanisms with deterministic filtering dominated in autumn, whereas stochastic processes prevailed in other seasons. Predicted functional profiles indicated a stable core metabolism, although local anthropogenic inputs stimulated specific metabolic adaptations in industrial and aquaculture zones. Our findings reveal that seasonal environmental filtering (especially temperature and salinity) and a shifting balance between stochastic and deterministic assembly processes govern bacterial dynamics in this tropical coastal ecosystem, with anthropogenic inputs modulating local metabolic functions. This study provides fundamental insights into the mechanisms maintaining microbial diversity and stability in tropical coastal waters facing seasonal and human pressures. Full article

In characterizing unconventional reservoirs, conventional Digital Rock Physics (DRP) has long been constrained by three fundamental bottlenecks: the trade-off between imaging resolution and field of view, challenges in reconstructing multiscale pore topology, and the prohibitive computational cost of direct numerical simulation (DNS) at the pore scale. The deep integration of artificial intelligence and rock physics has given rise to a new paradigm—Intelligent Digital Rock Physics (IDRP). This paper provides a systematic review of the evolutionary trajectory of IDRP, with a focus on how machine learning is reshaping the end-to-end workflow from imaging and segmentation to reconstruction and simulation. First, we survey image super-resolution and 3D pore structure generation techniques based on convolutional neural networks (CNNs), generative adversarial networks (GANs), and diffusion models, elucidating their mechanisms for surpassing optical diffraction limits and incorporating macroscopic petrophysical constraints. Second, we outline algorithmic strategies for fusing multi-source heterogeneous data (e.g., Micro-CT and SEM) and representing dual-porosity or multi-continuum systems. Third, we critically examine the application of machine learning surrogates in single- and multiphase flow prediction, highlighting how physics-informed machine learning (PIML) and reinforcement learning (RL)—by embedding governing equations such as Navier–Stokes or Muskat–Leverett into loss functions—achieve both computational acceleration and physical consistency. We further identify key limitations of current IDRP approaches, including insufficient validation of generated topological realism, narrow generalization across lithologies, inadequate representation of dynamic wettability, and limited model interpretability. Finally, we propose a forward-looking roadmap centered on multimodal foundation models for rocks, coupled with neural operators and uncertainty quantification frameworks, emphasizing the critical pathways for translating IDRP into engineering digital twins for unconventional hydrocarbon development, coalbed methane production enhancement, Enhanced Geothermal Systems, and geological CO₂ storage. This review offers a comprehensive reference for researchers at the intersection of geophysics, rock mechanics, and artificial intelligence. Full article

Indoor air quality (IAQ) plays a critical role in the health and cognitive performance of students, making its assessment essential for sustainable building design in educational environments. This study evaluates whether the ventilation flow rates prescribed by the Spanish Regulation for Thermal Installations in Buildings (RTIB), together with the occupancy densities defined by the Technical Building Code (TBC), are sufficient to maintain CO₂ concentrations within regulatory limits in classrooms and library reading rooms. A validated three-dimensional CFD model was developed to simulate airflow patterns and CO₂ distribution under typical operating conditions. The model was experimentally validated using measurements from a dedicated test room in the KUBIK experimental building of Tecnalia, demonstrating high predictive accuracy with average relative errors between 14% and 20%. Results indicate that, under current RTIB and TBC design criteria, (modelled for a 36 m² classroom with 24 occupants and a fresh air supply of 1080 m³/h), CO₂ levels frequently exceed the 910 ppm regulatory thresholds established by the RTIB’s direct method, highlighting potential shortcomings in existing standards for educational spaces. Additionally, two mechanical ventilation configurations were analyzed, revealing that floor-supply ventilation promotes more homogeneous pollutant dispersion and lower concentration peaks compared with ceiling-mounted systems. These findings underline the need to reconsider ventilation design strategies in educational buildings and demonstrate the value of CFD modelling as a tool to support evidence-based decisions toward healthier and more sustainable indoor environments. Full article

Acid-generating mining waste and polymer waste are two of the most persistent environmental problems facing the mining and manufacturing sectors, respectively. We have investigated the co-recovery of these disparate waste streams for the production of unfired cementitious paving blocks. We established a statistically optimized formulation using response surface methodology (RSM) and a central composite design (CCD). We systematically evaluated three process variables: air-curing time (4–37 days), dosage of the waste mixture (5–68% by weight of dry solids: acid-generating mining waste, hydrated lime, and recycled polymer in a waste-to-polymer mass ratio of 1:1), and type of polymeric aggregate (recycled PET flakes versus granulated rubber). Compressive strength ranged from 4.5 to 42.1 MPa across the 40 experimental conditions. The resulting quadratic model was clearly significant (F = 186.31, p < 0.0001) with solid predictive parameters (R² = 0.9796; R²pred = 0.9627; adequate precision = 42.47). Desirability-based optimization, which limited air curing to industrially feasible timeframes (7–28 days) and maximized waste utilization within a 10–50% by weight, identified PET with 12.4 days of curing and a 50% by weight waste mixture as the optimal configuration, predicting a compressive strength of 37.3 MPa. This value exceeds the 32 MPa threshold for Type I heavy-traffic paving blocks; however, confirmatory tests yielded 34.09 ± 1.08 MPa, indicating that production-scale use should include control of moisture content, compaction, and batch homogeneity. Scanning electron microscopy with energy-dispersive spectroscopy (SEM/EDS) and X-ray diffraction (XRD) demonstrated that PET inclusions promoted a denser and more continuous interfacial transition zone than shredded rubber, driven by physical entanglement and the pronounced microfilling effect of the fine waste particles. Full article