Browse Articles
Discover research articles across all indexed journals
Observation of resonance of kagome flat band doublet
Deep learning enhanced prediction framework for bio oil yield from organic solid waste with chemically informed features
Abstract Accurate prediction of bio-oil yield from pyrolysis of organic solid waste is still the main problem in the field of biomass valorization, mostly because of high variability in the composition and the heterogeneity of experimental conditions reported in the literature. Machine-learning techniques described in existing literature frequently depend on datasets that are too narrow, have limited feature representations, or shallow models that cannot sufficiently capture nonlinear thermochemical interactions that control devolatilization and liquid formation. In this study, the present study develops a deep learning-based predictive framework that addresses these gaps by using a harmonized dataset of 245 samples derived from diverse biomass sources and pyrolysis conditions. The model employs a chemically guided feature-engineering strategy that uses elemental ratios, ash-corrected volatility, and an energy-density index, followed by Variance Inflation Factor (VIF)-driven feature selection to lessen multicollinearity while keeping the mechanistic relevance. The resulting Hybrid DNPO achieves an R 2 of 0.980 and an RMSE of 1.14 on new data, making it better than all benchmark regression models, including Light Gradient Boosting (LGB). The results show that a thermochemically grounded, chemically informed deep neural prediction framework is proposed, integrating feedstock compositional descriptors and operating conditions to predict bio-oil yield from organic solid waste pyrolysis, significantly improving the performance of prior models from the literature, and can serve as a reliable tool for guiding experimental design, biomass screening, and process optimization in eco-friendly bio-oil production. By leveraging a larger, diverse dataset and advanced feature engineering, the methodology also offers a solid foundation for future studies aiming to enhance the accuracy and robustness of biomass-to-biofuel conversion predictions.
A model for drug transport across two membranes of Gram-negative bacteria by an MFS tripartite assembly
Abstract Transport of proteins and small molecules across cellular membrane is crucial for bacterial interaction with the environment and survival against antibiotics. In Gram-negative bacteria that possess two layers of membranes, specialized macromolecular machines are required to transport substrates across the cell envelope, often via an indirect stepwise process. The major facilitator superfamily (MFS)-type tripartite efflux pumps use proton electrochemical gradient to extrude drugs in diverse bacterial species, but the architecture of the assembly and structural mechanisms remain elusive. A representative MFS-type tripartite efflux pump, EmrAB-TolC, mediates resistance to multiple antimicrobial drugs through proton-coupled EmrB, a member of the DHA2 transporter family. Here, we report the high-resolution (3.13 Å) structure of the EmrAB-TolC pump, revealing a distinct, asymmetric architecture emerging from the assembly of TolC:EmrA:EmrB with a ratio of 3:6:1 and contacts that are essential for the pump assembly. Key residues involved in drug transport are identified and corroborated by mutagenesis and antibiotic sensitivity assays. The structural and functional data support a model for one-step drug transport by the MFS pump across the entire envelope of Gram-negative bacteria.
Ultra-compact canvas-type acoustic metasurfaces for uniform sound field in indoor environments
Bidirectional catalysts with dual-atom dynamic d-band centre modulation and support self-reconstruction for de/hydrogenation in MgH2/Mg
Hidden Markov models reveal ontogenetic plasticity in green and loggerhead sea turtles
FineST: contrastive learning integrates histology and spatial transcriptomics for nuclei-resolved ligand-receptor analysis
Abstract Spatial transcriptomics (ST) has emerged as a powerful tool for analyzing cell-cell communication (CCC) across various biological processes, ranging from embryonic development to cancer progression. However, its limited resolution and high data sparsity hinder the detailed characterization of CCC patterns within complex tissues. Here, we introduce FineST , a deep contrastive learning model that leverages a histology foundation model to fuse ST and histology images, enabling Fine -grained S patial T ranscriptomics analysis. This approach facilitates precise nuclei segmentation, high-resolution RNA expression imputation, and the identification of intricate ligand-receptor interactions. Using both colorectal cancer VisiumHD and breast cancer Xenium datasets, we demonstrate that FineST significantly outperforms existing methods in high-resolution RNA imputation, cell type prediction, and CCC pattern discovery. With focused application to the Visium platform, FineST reveals novel biological insights into tumor-immune interactions across multiple cancer types, including invasive fronts in breast cancer, tertiary lymphoid structures in nasopharyngeal carcinoma, and PD-1 therapy resistance barriers in hepatocellular carcinoma. These findings highlight a new paradigm in ST analysis through the integration of readily available histology images.
Phase-dependent modulation of the MJO during cross-equatorial northerly surges (CENS)
Structured coherent thermal emission from non-Hermitian metasurfaces
Delayed goal-directed processing underlies inhibitory control challenges in adult ADHD
Coupled polarization dynamics and charge tunneling enable reconfigurable heterojunctions
Establishment of a rapid Brucella detection method based on MCDA-CRISPR dual signal amplification system for reducing transfusion-transmitted diseases
Synchronous activation of striatal cholinergic interneurons induces local serotonin release
Comparative effects of surface and underwater lighting methods on coastal fishery resources in Terengganu
Geo-spatial prospective life cycle sustainability of InGaN and InGaP compound semiconductors
Abstract This is the first study which presents integrated geo-spatial prospective life cycle and supply chain sustainability modelling of two compound semiconductors: Indium Gallium Nitride (InGaN) and Indium Gallium Phosphide (InGaP), in 80 international supply chain scenarios, incorporating 11 countries across 4-time horizons—024, 2030, 2040 and 2050. The results show environmental sustainability is geographic- and time- dependent and varied by properties of the supply chain characteristics. The manufacturing of InGaN and InGaP excel in UK based scenarios (~ 70% and 66% impact reduction from 2040 to 2050). Scenarios involving shared fabrication in the UK and US show strong sustainability performance in 2024, those for fabrication in Taiwan (for both materials) and US (InGaN) demonstrate increasing sustainability potential by 2050. Scenarios involving fabrication in China consistently led to a higher environmental impact. However, all 80 configurations demonstrate marked reductions in environmental impacts, primarily due to global electricity grid decarbonisation, and improved emissions controls. Despite this improvement in the clean room energy impact, epitaxial growth and substrate preparation remain the hotspots, calling for process innovation, cleaner precursors (e.g., replacing arsine or phosphine), and advanced material recycling. InGaN generally performs better than InGaP in most categories, attributed to its simpler material inputs and lower toxicity potential. InGaP scenarios exhibit higher marine ecotoxicity, carcinogenic toxicity, and mineral resource scarcity, driven by complex chemistries and Gallium Arsenide (GaAs) substrates. Interestingly, InGaP scenarios significantly outperform InGaN in stratospheric ozone depletion due to limited use of halogenated chemicals. This study provides compelling evidence to support reshoring or nearshoring of compound semiconductor fabrication to regions with cleaner energy profiles and stronger environmental regulations. Scenarios involving the UK, USA and Taiwan (specially in 2050), consistently achieve higher sustainability scores across global warming, toxicity, and resource depletion categories.
Growth, productivity and profitability of potato (Solanum tuberosum L.) as influenced by nitrogen fertilizer and intra-row spacing in Ethiopia highlands
Abstract Nitrogen fertilizer and intra-row spacing are critical agronomic practices influencing potato (Solanum tuberosum L.) production. In Ethiopia, smallholder farmers often apply nitrogen and manage plant spacing without evidence-based guidelines, resulting in low productivity. A field experiment was conducted during the 2023 main rain fall season on a farmer’s field in the Ethiopian highlands to assess the growth and seed tuber yield of the potato variety ‘Belete’ under different nitrogen rates and intra-row spacings. The study evaluated four nitrogen levels (0, 55, 110, and 165 kg N ha⁻¹) and three intra-row spacings (20, 30, and 40 cm) in a factorial randomized complete block design with three replications. Interaction effects of nitrogen and spacing significantly affected days to 50% flowering, stem number, yields of very small, small, and large-sized tubers, average tuber weight, marketable yield, and total tuber yield. The highest marketable tuber yield (41.38 t ha⁻¹) was recorded with 110 kg N ha⁻¹ and 20 cm spacing. However, partial budget analysis revealed that the combination of 110 kg N ha⁻¹ and 30 cm spacing provided the highest net benefit (236,614 ETB ha⁻¹) and marginal rate of return (12,692.11%). These results underscore the need to optimize nitrogen application and plant spacing for enhanced seed tuber productivity and economic returns. Therefore, applying 110 kg N ha⁻¹ with 30 cm intra-row spacing is recommended for profitable potato production, improving income and food security for smallholder farmers in the study area and similar agro-ecologies.
Aerial image segmentation using multilevel thresholding based on multi strategy Osprey optimization algorithm
A do-it-yourself water quality sensor network to elucidate contaminant signatures and improve land management advice
Abstract Quantifying contaminant loads to estuaries is essential for setting effective limits on resource use and safeguarding ecological values. Traditional monitoring programmes often rely on infrequent sampling, which can substantially underestimate loads and obscure key transport processes. While commercial high-frequency sampling stations are prohibitively expensive, open-source, do-it-yourself technologies now offer affordable alternatives for continuous monitoring. Here, we deployed ten low-cost, in situ monitoring stations equipped with research-grade sensors across an intensively farmed catchment draining to a sensitive estuarine environment in the Bay of Plenty, New Zealand. Using artificial neural network models trained on concurrent grab samples, we converted sensor measurements into reliable 15-min estimates of contaminant concentrations. High-frequency load calculations revealed nitrogen, phosphorus, and sediment exports to be 6–87% greater than estimates derived from traditional monthly sampling. Moreover, time-series outputs uncovered distinct sub-catchment contaminant mobilisation and transport dynamics that would otherwise remain undetected. These findings demonstrate that open-source, high-frequency monitoring can substantially improve contaminant load quantification, provide new insights into catchment processes, and inform the development of land management and policy strategies that reflect the unique spatio-temporal patterns of contaminant export.