Browse Articles
Discover research articles across all indexed journals
Genome-wide iterative fine-mapping for non-Gaussian phenotypes
Cortical modulation of resting state BOLD signals in white matter
Dendrobium officinale Kimura & Migo polysaccharide ameliorates skin photoaging by promoting angiogenesis
Scientific rationale and guidelines to use cane sugar as a performance enhancer in lime mortars
Declawing in Cat is associated with neuroplastic sensitization and long-term painful afflictions
A unified FLC-blockchain framework for optimized carbon credit trading in multi-microgrid systems
Improvement in genetic evaluation of quantitative traits in sheep by enriching genetic model with dominance effects
Abstract Although dominance effects play a major role in quantitative genetics, most studies on quantitative traits have often neglected dominance effects, assuming alleles act additively. Therefore, the aim followed here was to quantify the proportion of variation in the early growth of Baluchi sheep that was attributed to dominance effects. Data collected over a 28-year period at the Baluchi sheep breeding station was used in this study. Traits evaluated were birth weight (BW), weaning weight (WW) and average daily gain (ADG). Each trait was analyzed with a series of twelve animal models which included different combinations of additive genetic, dominance genetic, maternal genetic and maternal permanent environmental effects. The Akaike’s information criterion (AIC) was used to rank models. The predictive ability of models was measured using the mean squared error of prediction (MSE) and Pearson’s correlation coefficient between the real and predicted values of records (r( $$\:y$$ , $$\:\widehat{y}$$ )). Correlations between traits due to additive and dominance effects were estimated using bivariate analyses. For all traits studied, including dominance effects improved the likelihood of the fitting model. In addition, models that included dominance effects had the better predictive ability as provided higher r( $$\:y$$ , $$\:\widehat{y}$$ ) and lower MSE. However, accounting for dominance effects significantly increased the computing burden evidenced by considerably longer computing time and a huge amount of memory required. By including dominance effects in the model, additive genetic variance did not change, but residual variance decreased significantly up to 41%, which indicated that the dominance component distangelled from residual variance. For BW, WW and ADG, dominance genetic variance was 6.61, 1.91, and 2.73 times greater than additive genetic variance and contributed 87%, 65% and 73% to the total genetic variance, respectively. Estimates of dominance heritability ( $$\:{\varvec{h}}_{\varvec{d}}^{2}$$ ), were 0.29 ± 0.06, 0.15 ± 0.07 and 0.20 ± 0.07 for BW, WW and ADG, respectively. Additive heritability ( $$\:{\varvec{h}}_{\varvec{a}}^{2}$$ ), was 0.05 ± 0.01 for BW, 0.08 ± 0.02 for WW and 0.07 ± 0.02 for ADG, respectively. By including dominance effects in the model, the accuracy of additive breeding values increased by 8%, 8% and 11% for BW, WW and ADG, respectively. Correlation between additive breeding values obtained from the best model and the best model without dominance effects were close to unity for all traits studied, indicating negligible changes in the additive breeding values and little chance for re-ranking of top animals across models. While additive genetic correlations were all positive and high, the dominance genetic correlation between WW and ADG was positively high (0.99), and between other pairs of traits was negative. Although the inclusion of dominance effects in the model did not change the ranking of top animals and had high computational requirements, it improved the predictive performance of the model and led to a significantly better data fit and an increase in the accuracy of additive breeding values. Therefore, including dominance effects in the model for genetic evaluation of the early growth of Baluchi lambs can be a reasonable recommendation.
Studying the aftereffect and changes in sensitivity to physical and mental time references using a time adaptation paradigm
Polyaniline nano-material backed lens antenna for X-band LEO satellite transceivers
Abstract This paper presents the design and implementation of an open-ended waveguide lens antenna engineered to generate a conical-shaped radiation pattern for the X-band fully-duplex communication subsystem of a low Earth orbit (LEO) satellite dedicated to Earth remote sensing. The antenna is designed to maintain reliable links with ground stations at a minimum satellite elevation angle of 10°, corresponding to ± 63° off-nadir, for a near-circular orbit at an altitude of 700 km. Operating within the 9.75–10.25 GHz band, the antenna provides broad coverage over a working sector of approximately 126° × 126° in azimuth and elevation, with peak radiation directed at ± 63° from nadir. It achieves a minimum gain of 5 dBic in these directions and employs right-hand circular polarization (RHCP) for downlink and left-hand circular polarization (LHCP) for uplink. The axial ratio remains below 3 dB across the working sector at the center frequency of 10 GHz, while the input reflection coefficient stays better than − 10 dB over a wide impedance matching bandwidth of 8.75–11.25 GHz. The 3-dB axial ratio bandwidth spans from 9.75 to 10.25 GHz. To further enhance performance particularly circular polarization purity, gain, and axial ratio bandwidth a metallic circular backing plate coated with a polyaniline (PANI) nano-material absorber is integrated between the dielectric lens and the reflector disc. The PANI layer, characterized by tunable dielectric properties and intrinsic microwave loss, improves impedance matching at the lens–waveguide interface and effectively suppresses surface currents and backward radiation. Material characterization via X-ray diffraction (XRD) and scanning electron microscopy (SEM) confirms the semi-crystalline structure and micro-porous morphology of the synthesized PANI, which contribute to enhanced electromagnetic absorption. As a result, the axial ratio at 10 GHz is reduced from 1.0 dB to 0.05 dB, the gain at ± 63° is increased from 4.1 dBic to 6.0 dBic, and the 3-dB axial ratio bandwidth is expanded from 400 MHz to 500 MHz. These findings demonstrate the potential of integrating functional PANI nano-materials into high-performance antenna architectures for advanced satellite communication and Earth observation applications.
High efficiency classification of thyroid cytopathological images based on knowledge distillation and vision transformer
Intraspecific higher order interactions enhance ecological community stability
Abstract Ecosystem stability is influenced by interspecific interactions, with recent research focusing on how third-party species modify these interactions. However, the impact of "intraspecific higher-order interactions," where the presence of one species affects intraspecific interactions within another species, on ecosystem stability has been less explored. This study addresses this gap by developing a mathematical model to investigate how such interactions influence stability. The analysis shows that when higher-order interactions increase intraspecific competition within another species, stability improves, especially in large, complex ecosystems. However, this effect is not observed if the interactions solely increase or decrease competition without a mixture of both. These findings highlight the importance of both positive and negative effects on intraspecific competition for enhancing stability in complex ecosystems. This emphasizes the need for further research on the role of higher-order interactions in shaping ecosystem stability, particularly in complex ecosystems.
Characterization of nanoplastics and small-sized microplastics in sewage treatment
Abstract The presence of nanoplastics (NPs) in sewage treatment plants (STPs) remains a critical yet underexplored environmental issue. Here, we present a novel investigation into the occurrence, recovery, and characterization of nanoplastics and small-size microplastics (50–2500 nm) in raw and treated sewage effluent from a full-scale STP (treating 4000 m3/day) operating with activated sludge. To our knowledge, this research includes the first confirmed assessment of nanoplastics in such a system and applies nano-flow cytometry to wastewater analysis for the first time globally. It is also the first study addressing micro- and nanoplastics in wastewater in Saudi Arabia, advancing plastic particle analysis in complex matrices. Particles in the 50– < 100 nm range accounted for 44% of total particles detected in STP effluents. Overall, plastic particles accounted for 16% (± 10%) of total particles within 50-2500 nm in raw sewage, increasing to 41% (± 13%) in treated effluent. This increase highlights the inefficiency of conventional treatment in fully removing plastic particles and suggests preferential removal based on size or density. The composition of a representative selection of particles was characterized by micro-Raman spectroscopy and Scanning Electron Microscopy with Energy-Dispersive X-ray. Identified polymers included polystyrene, polyvinyl chloride, polyethylene, polytetrafluoroethylene, polyamide, and polypropylene. These findings provide important insights into treated sewage composition, particularly for reuse in arid regions.
Elucidation of user autonomous driving system preference mechanisms under the extension of internal and external factors
Abstract This study is based on the extended Technology Acceptance Model, integrating Habit Theory and Regret Theory to construct a model of users’ continuous usage intentions. It conducts a comparative analysis of users’ continuous usage intentions for both assisted driving systems and driverless systems. Data was collected through an online questionnaire survey and analyzed using Structural Equation Modeling. The results indicate that, within assisted driving systems, perceived importance and driving habits significantly influence continuous usage intentions; conversely, in driverless systems, driving habits are paramount. In both systems, users’ perceived importance and experience of regret have a significant impact on driving habits, with experience of regret indirectly affecting continuous usage intentions through driving habits. User scale exerts direct or indirect effects through various variables. Regarding control variables, significant differences exist between the two systems; users of assisted driving systems prioritize economic benefits, while users of driverless systems focus on after-sales service. This research theoretically establishes a new framework, enriching and refining relevant theories; practically, it provides references for system improvement and promotion; and it suggests avenues for future research into control variables and segmented user groups. The study reveals the driving factors behind users’ continuous usage of different driving systems, constructing a universally applicable theoretical model with significant academic and practical implications.
Limitation of super-resolution machine learning approach to precipitation downscaling
Abstract The present study explores the potential of super-resolution machine learning (ML) models for precipitation downscaling from 100 to 12.5 km at hourly timescale using the Conformal Cubic Atmospheric Model (CCAM) data over the Australian domain. Two approaches were examined: the perfect approach, which trains the ML model using coarsened high-resolution data as input (i.e., CCAM 12.5 km data coarsened to 100 km), and the imperfect approach, which uses original coarse-resolution data as input (i.e., CCAM model simulation at 100 km resolution) and in both the cases high-resolution data (i.e., CCAM 12.5 km simulation) is used as target. In the perfect case, the ML model (MLPerfect) accurately reproduces high-resolution climatology and extremes. However, the MLPerfect model with CCAM 100 km simulation data as input (i.e., in the imperfect setting) underestimates the magnitude of the output and introduces spatial inconsistencies, while the MLImperfect model captures high-resolution structures but underestimates extremes. This suggests that the super-resolution MLPerfect model approach is inappropriate for precipitation downscaling because of the spatial inconsistencies between the coarse and high-resolution simulations. Additionally, we introduced sensitivity-based diagnostics beyond standard evaluation methods to understand model behaviour and identify structural issues. These diagnostics reveal that both models increase precipitation inputs non-linearly without creating spurious spatial relationships. However, the MLImperfect model outputs precipitation in high-altitude regions regardless of input, highlighting the structural issue of the MLImperfect model. Our study highlights the challenges in using super-resolution ML models for precipitation downscaling, introduces several useful diagnostics for assessing the super-resolution ML models and their physical realism, and provides ideas to explore to improve ML-based precipitation downscaling.
Optimised knowledge distillation for efficient social media emotion recognition using DistilBERT and ALBERT
Abstract Accurate emotion recognition in social media text is critical for applications such as sentiment analysis, mental health monitoring, and human-computer interaction. However, existing approaches face challenges like computational complexity and class imbalance, limiting their deployment in resource-constrained environments. While transformer-based models achieve state-of-the-art performance, their size and latency hinder real-time applications. To address these issues, we propose a novel knowledge distillation framework that transfers knowledge from a fine-tuned BERT-base teacher model to lightweight DistilBERT and ALBERT student models, optimised for efficient emotion recognition. Our approach integrates a hybrid loss function combining focal loss and Kullback-Leibler (KL) divergence to enhance minority class recognition, attention-head alignment for effective contextual knowledge transfer, and semantic-preserving data augmentation to mitigate class imbalance. Experiments on two datasets, Twitter Emotions 416 K samples, six classes, and Social Media Emotion 75 K samples, five classes, show that our distilled models achieve near-teacher performance 97.35% and 73.86% accuracy, respectively. with only a < 1% and < 6% accuracy drop, while reducing model size by 40% and inference latency by 3.2×. Notably, our method significantly improves F1-scores for minority classes. Our work sets a new state-of-the-art in efficient emotion recognition, enabling practical deployment in edge computing and mobile applications.