Browse Articles

A hybrid concatenation packet aggregation approach for energy-aware disparate path routing in energy harvesting WSNs

Abstract In a sensor network, a sensor node does not aggregate the data packets efficiently, since the packet is dropped for poor connection between the sensor nodes in the network. The sensor node location is varied in every area, so infrequent to share the information with its neighbor by the sender node. The communication is lost, and only received a minimum amount of packets for every communication, remaining packets are dropped. The node performs again packet transmission on the path makes the minimum residual energy level of each node. It increases the packet loss rate and energy consumption. Proposed IMPROVED ENERGY EFFICIENT DISPARATE PATH ROUTING (IEEDP) method is designed to achieve energy efficient packet aggregation along network environment. The position of the node is monitored clearly and then assign disparate path for routing. Concatenation Packet Aggregation algorithm is designed for a sensor network, it gathers the data packets without any loss for the transmission period from the source node to a destination node in a network environment, energy preservation is a sequence to increase the lifetime of the network by analyzing the residual Energy Effects on EH-WSN. It reduces packet loss rate and energy consumption.

Boycott of major AI conference exposes a growing US–China divide

Placental insufficiency and markers of fetal growth are associated with longitudinal infant growth and body composition trajectories up to two years of age

Coalescence and translation: A language model for population genetics

Probabilistic models such as the sequentially Markovian coalescent have long provided a powerful framework for population genetic inference, enabling reconstruction of demographic history and ancestral relationships from genomic data. However, these methods are inherently specialized, relying on predefined assumptions and/or limited scalability. Recent advances in simulation and deep learning provide an alternative approach: learning directly to generalize from synthetic genetic data to infer hidden evolutionary processes. Here we reframe the inference of coalescence times as a problem of translation between two biological languages: the sparse, observable patterns of mutation along the genome and the unobservable ancestral recombination graph that gave rise to them. Inspired by large language models, we develop cxt, a decoder-only transformer that autoregressively predicts coalescent events conditioned on local mutational context. We show that cxt performs competitively with state-of-the-art Markov Chain Monte Carlo-based likelihood models across a broad range of demographic scenarios, matching their accuracy in-distribution and approaching it in out-of-distribution, with the potential for improvement via fine-tuning. Trained on simulations spanning the stdpopsim catalog [J. R. Adrion et al ., eLife 9 , e54967 (2020); M. E. Lauterbur et al ., eLife 12 , e84874 (2023); G. Gower et al ., Accessible, realistic genome simulation with selection using stdpopsim. bioRxiv [Preprint] (2025). https://doi.org/10.1101/2025.03.23.644823 .], the model generalizes robustly and enables efficient inference at scale, producing over a million coalescence predictions in minutes. cxt also produces well-calibrated approximate posteriors, enabling principled uncertainty quantification. We apply cxt to population genomic data from both humans and mosquitoes, highlighting the model’s ability to deal with the complexities of empirical data.

The value of multiparametric magnetic resonance imaging in the prognostic judgment of prostate cancer

Amoeboid–mesenchymal transition and the proteolytic control of cancer invasion plasticity

Invasion plasticity allows malignant cells to toggle between collective, mesenchymal, and amoeboid phenotypes while traversing extracellular matrix (ECM) barriers. Current dogma holds that collective and mesenchymal invasion programs trigger the mobilization of proteinases that digest structural barriers dominated by type I collagen, while amoeboid activity allows cancer cells to marshal mechanical forces to traverse tissues independently of ECM proteolysis. Here, we use cancer spheroid-3-dimensional matrix models, single-cell RNA sequencing, and human tissue explants to identify the mechanisms controlling mesenchymal versus amoeboid invasion. Unexpectedly, collective/mesenchymal- and amoeboid-type invasion programs—though distinct—are each characterized by active tunneling through ECM barriers, with expression of matrix-degradative metalloproteinases. CRISPR/Cas9-mediated targeting of a single membrane-anchored collagenase, MMP14/MT1-MMP, ablates tissue-invasive activity while coregulating cancer cell transcriptional programs. Though changes in matrix architecture, nuclear rigidity, and metabolic stress as well as the presence of cancer-associated fibroblasts are proposed to support amoeboid activity, none of these changes restore invasive activity of MMP14-targeted cancer cells. While a requirement for MMP14 is bypassed in low-density collagen hydrogels, invasion by the proteinase-deleted cells is associated with nuclear envelope and DNA damage, highlighting a proteolytic requirement for maintaining nuclear integrity. Nevertheless, when cancer cells confront explants of live human breast tissue, MMP14 is again required to support invasive activity. Corroborating these results, spatial transcriptomic and immunohistological analyses of human breast cancers identified MMP14 expression in tissue-infiltrating carcinoma cells that were further juxtaposed with proteolyzed type I collagen fragments, underlining the pathophysiologic importance of this proteinase in directing invasive activity in vivo.

Tea-inspired curing modulates chemical composition, volatile aroma, and sensory quality of flue-cured tobacco leaves

Abstract Curing largely determines the usability and style of tobacco. Drawing on tea processing, we developed three methods using fresh tobacco leaves: a black tea–like process with wilting, rolling, fermentation, and drying (TR), and two green tea key steps that replace rolling with steaming (TZ) or pan-firing (TC) to shape distinct aroma styles, and these were compared with conventional flue-curing (CK). Sensory quality was evaluated by a trained panel and an electronic nose (E-nose). Enzyme activities, macromolecular constituents, major chemical components, and volatile organic compounds (VOCs) were systematically analyzed. The results showed that TR achieved the highest overall sensory performance, with CK and TC comparable, and TZ the lowest. E-nose patterns clearly discriminated the treatments and covaried with aroma scores, reflecting differences in aroma style. Enzymatic reactions induced by TR processing led to a marked decrease in polyphenols, while facilitating macromolecular degradation and promoting the accumulation of key VOCs, including phenylacetaldehyde, indole, and 4,7,9-megastigmatrien-3-one. In contrast, high-temperature pretreatments in TZ and TC rapidly inactivated enzymes before fermentation, leading to higher polyphenol retention but limited macromolecular degradation and key aroma development. Overall, macromolecules are primary negative drivers of sensory quality, whereas chemical balance and key aroma constituents support higher scores. These results indicate that targeted manipulation of key processing steps can effectively direct the formation of tobacco quality and flavor differentiation, while potentially regulating the formation of certain undesirable components. This study provides valuable support for advancing diversified tobacco-curing methods and generating tobacco leaf materials with distinct flavor profiles.

Infants use helping to infer the existence and strength of caring relationships

Prosocial and antisocial behaviors can reflect the actor’s positive or negative feelings toward a specific person and their broader disposition toward others’ welfare. Observers’ inferences about these motives shape what they learn about the actor’s social relationships and moral character, thereby informing expectations about the actor’s future behaviors. In four preregistered studies, we used violation-of-expectation looking methods to investigate 14- and 15-mo-old infants’ inferences about social motives by first showing them helping and hindering events and then assessing their expectations for those actors’ future actions. If infants use helping and hindering to infer disposition-based motives, they should extend expectations about future helping to others. Alternatively, if they infer relationship-based motives, their expectations should be restricted to the initial interactants. In Experiments 1 and 2, infants first observed two actors, one who helped and another who hindered a common target. At test, infants expected the helper would help the same target in a new, perceptually distinct situation. However, this expectation did not generalize to a new target, suggesting infants inferred specific relationships, not general dispositions. In Experiments 3 and 4, we assessed whether infants used information about the existence or strength of a helper’s relationships to predict who they would prioritize helping in the future. We found that infants used patterns of selective helping among multiple targets to make transitive inferences about whose needs the helper would prioritize. Taken together, this suggests that infants use such observations of helping to infer whether, and how much, the helper cares for their social partner.

Defining an operational selectivity window for rare-earth flotation using a Box–Behnken design

Optimization of crossing point temperature for predicting coal spontaneous combustion

Predicting and controlling laser-induced breakup and multidirectional propulsion of liquid droplets

Laser-driven control of droplets is important in microfluidics, targeted delivery, and droplet-based laser–matter interactions, yet propulsion direction and breakup remain difficult to predict. Here, we demonstrate that an acoustically levitated droplet’s propulsion polarity and breakup morphology can be selected by controlling axial placement relative to the external (no-droplet) optical focus together with incident pulse energy. We combine time-resolved shadowgraphy with aberration-aware optical simulations that locate the prebreakdown irradiance maxima, and we connect these linear field calculations to laser-induced breakdown using experimentally calibrated thresholds in the liquid and in near-field air. Expressing irradiance in threshold-referenced form yields a first-crossing rule that identifies whether breakdown initiates at the illumination surface, in the interior, at the shadow surface, or in wake-side near-field air. The resulting placement–energy maps anticipate forward, backward, and near-radial droplet responses. A polarity index and a reliability metric quantify directionality and highlight narrow transition corridors in parameter space where competing initiation sites are nearly tied under experimental jitter. The predicted landscapes agree with experiments, and multiphase simulations reproduce the early-time shock-driven deformation. Because the regime structure is set by caustic geometry together with breakdown thresholds, the framework transfers across transparent liquids after threshold calibration and reduces to a surface-trigger condition in the strongly absorbing limit.

Plant spatial compartmentalization buffers bacteriome structure and function under antibiotic stress

Investigating performance and key factors for real-world deployment of grain image classification using convolutional neural networks

Abstract Accurate and efficient grain quality assessment is critical for making informed decisions throughout the grain value chain. Early detection of disease enables actions to mitigate spread and further damage, and optimal batch mixing to fulfill specified quality requirements allows for maximizing value and minimizing scrapping. Vision based machine learning and deep learning approaches are gaining attention in the agricultural sector and are useful for the development of automated grain quality assessment. These techniques can reduce the current manual inspection load and are key for objective and precise analysis. Yet, the majority of prior studies are constrained to small or controlled and curated datasets. Practical challenges associated with real-world deployment and reliability are rarely addressed. That is the focus of this work. We present and demonstrate a structured approach for investigating convolutional neural networks (CNNs) and key factors influencing performance for wheat kernel classification. The objective is to determine a CNN model that ensures high and robust classification accuracy, while elucidating and explaining how different image dataset characteristics and training parameters affect performance and reliability. We use a commercial mirror-based imaging system that captures over 90% of each kernel’s surface and contrast and compare model architectures, robustness, the effect on pre-processing and image resolution. Our results show similar and high overall performance for ResNet50V2 and EfficientNetV2B0 ( $$>96$$ % accuracy), but per-class analysis indicate that the smaller classes suffer from lack of representative examples, and that most classes benefit from pre-processing including downsampling whereas others benefit from higher resolution. Interactive visualizations reveal that another contributing factor is dubious annotation and multi-class belongingness. Thus, our step-by-step analysis of CNN performance underscores the need for representative data, proper pre-processing, and class-aware evaluation to ensure trustworthy deployment in wheat grain quality assessment.

Brain–computer interface–based neurofeedback training enables transferable control of cortical state switching in humans

Behavioral flexibility relies on transient neural dynamics that govern cortical state transitions. However, whether humans can deliberately learn to control such state transitions and generalize trained neural dynamics beyond contexts remains unclear. Here, we demonstrate that operation of a brain–computer interface (BCI) which links time evolution of sensorimotor activity with real-time feedback enables volitional control over the targeted neural population. Compared with a double-blind sham control group, trained participants modulated sensorimotor oscillations in the absence of BCI. Data-driven latent-state analysis further revealed stronger interregional phase coupling and steeper broadband spectral slope in the medial frontal cortex during transitions. The training-induced reorganization of sensorimotor dynamics was found during movement execution and associated with performance improvement, indexed by reduced reaction times for both muscle contraction and relaxation. These findings provide evidence that learned control over cortical state transitions enhances behavioral flexibility beyond the training context.

Machine learning-based prediction of soiling losses in photovoltaic modules under different cleaning frequencies: an experimental investigation

Abstract Accumulation of dust on solar panels lowers performance and limits energy production, particularly in dry locations. Dust accumulation on photovoltaic panels diminishes performance and reduces energy output, especially in arid regions. This study uses four identical modules in Roorkee, India, from October to December to examine the impact of cleaning frequency on photovoltaic (PV) performance. The reference panel is cleaned daily, while the remaining panels are cleaned weekly, biweekly, and monthly. Alongside short-circuit current measurements, environmental parameters including global horizontal irradiance, ambient temperature, wind speed, and relative humidity are continuously recorded. In this study, soiling loss (%) is examined as the primary performance indicator under various cleaning intervals to observe dust accumulation progression and its impact on the performance of the solar photovoltaic module. Experimental data are utilized to develop an empirical regression model that describes the trend of dust accumulation. The daily average soiling loss ranges between 0.17 and 0.21%. Furthermore, machine learning models, including Decision Tree, K-Nearest Neighbour, support vector regression, artificial neural network, and a stacking ensemble method, are developed for accurate prediction of soiling loss from environmental variables and cleaning frequency. The stacking model consistently achieves the best performance across all months, with root mean square error as low as 0.03–0.045, mean absolute error below 0.03, and R² = 0.999 compared to other models. Moreover, statistical analyses such as Bland–Altman plots and the Wilcoxon signed-rank test are employed to validate the significance and agreement of the predicted outcomes. The study highlights the benefits of data-driven solutions for predictive operation and maintenance of solar photovoltaic systems and provides valuable insights into the impact of cleaning frequency on reducing soiling losses.

Extended Rice–Thomson analysis and atomistic simulations revealing grain boundary effects on fracture in refractory high-entropy alloys

Understanding how grain boundaries mediate fracture remains a critical challenge in designing ductile, high-performance refractory alloys. Here, we extend the Rice–Thomson criterion to account for the angle between cracks and the impinging grain boundaries (GBs), capturing the competition between intergranular fracture and dislocation-mediated plasticity. Using machine learning interatomic potentials, we performed molecular statics simulations to probe fracture mechanisms in nanocrystalline NbMoTaW and Nb 45 Ta 25 Ti 15 Hf 15 , each with two different grain sizes, revealing trends consistent with experimental observations and the extended Rice model. Comparison with averaged R-curves for bulk samples demonstrates that GBs enhance ductility in Nb 45 Ta 25 Ti 15 Hf 15 in both grain sizes investigated. In contrast, GBs only locally improve fracture resistance in NbMoTaW when cracks are temporarily pinned at GBs inclined at high angles from the crack, but generally promote brittle intergranular fracture. These contrasting behaviors are attributed to differences in GB cohesion, reflecting clear alloying trends that align with ab-initio calculations and trends observed experimentally. Our results bridge classical fracture theory, atomistic simulations, and experimental observations, providing a comprehensive understanding of the fracture mechanisms in nanocrystalline refractory complex concentrated alloys.

Carotenoid nanoforms from clementine peel stabilized by xanthan and arabic gums with antioxidant anti inflammatory and antimicrobial activities

Abstract Antimicrobial resistance is a major global health threat, causing 1.27 million deaths and contributing to nearly 5 million more annually. Citrus fruits are widely cultivated for their nutritional and health benefits, and their pigments may offer additional bioactivity. This study aimed to extract carotenoids from clementine peels and quantify carotene, lycopene, astaxanthin, and anthocyanins using colorimetric methods. The carotenoids were then encapsulated using xanthan-Arabic gums matrices to stabilize the pigments. The native carotenoids and their nanoform were evaluated for antioxidants, anti-inflammatory, and antimicrobial activities. Clementine peels contained total carotenoids (30.8 mg/kg), anthocyanins (13.69 mg/kg), lycopene (3.12 mg/kg), and astaxanthin (1.60 mg/kg). The physical characterization of the carotenoid nanoparticles was performed using FTIR, confirming interactions between carotenoids and the polymer matrices. Zetasizer analysis revealed a particle size of 17.05 nm and a zeta potential of − 26.7 mV, indicating good stability, while TGA demonstrated thermal stability up to 300 °C. Antioxidant activity ranged from 18.67% to 69.10% (phosphor-molybdenum and DPPH assays). Anti-inflammatory activity showed maximum protein denaturation inhibition at 83.87% (carotenoids), 74.45% (nanoform), and 69.10% (polymers) ( p  < 0.05). Antimicrobial activity was highest against Listeria monocytogenes (21.3 mm) and Penicillium verrucosum (20 mm), moderate against Escherichia coli (13.0 mm) and Aspergillus flavus (17.0 mm), and lower against Bacillus cereus (10.3 mm) and Pseudomonas aeruginosa (9.3 mm). Overall, xanthan and Arabic gums effectively stabilized carotenoids and enhanced their biological activities.

Large infinities and definable sets

Large cardinal axioms extend the standard set of axioms for mathematics by asserting that very large infinite sets exist. A prominent line of current research in mathematical logic is identifying ever stronger principles of this kind; this serves as an avenue toward mitigating the phenomenon of Gödel incompleteness. However, there is a tension between large cardinal axioms and principles asserting global simplicity of the mathematical universe, as well as forms of the Axiom of Choice. New kinds of infinity recently identified shed light into this tension and raise important mathematical and philosophical questions.

Towards efficient context-aware classification with compact VLM architectures: indoor fire case study

Abstract Accurate and reliable fire detection in indoor environments is critical for ensuring timely emergency responses and enhancing safety in both residential and industrial settings. While recent advances in deep learning have significantly improved fire and smoke detection, most existing systems remain limited to binary classification and often fail to distinguish between hazardous and benign fire events, leading to frequent and disruptive false alarms. To address this limitation, we propose a lightweight and efficient framework for context-aware fire classification. Specifically, input images are first encoded by a visual encoder and then interpreted by a Vision Language Model (VLM) to produce descriptive natural-language captions or semantic embeddings. A language model then processes these representations to perform high-level semantic classification into three categories: no fire, controlled fire (or fire under control), and dangerous fire. This design enables the system to reason about visual context and scene semantics, allowing for nuanced differentiation between visually similar but contextually distinct fire events. Our empirical results, evaluated on both re-labeled public datasets and our custom ConFire dataset, demonstrate that our approach achieves high accuracy while significantly reducing computational overhead. These findings highlight the effectiveness of integrating vision-language reasoning into fire classification tasks, paving the way for next-generation safety monitoring systems that are both intelligent and resource-efficient.

The rhythm of aging: Stability and drift in the individual rate of senescence

Human aging is marked by a steady rise in the risk of dying with age–a process demographers call senescence. Over the past century, life expectancy has risen dramatically, but is this because we are aging slower, or simply starting it later? Vaupel hypothesizes that the pace at which individuals age may be constant, with gains in longevity coming from the delayed onset of senescence rather than its slowing down. We test this idea using a framework that decomposes the pace of senescence into three components: a biological baseline, a long-term trend, and the cumulative impact of period shocks. Applying this to cohort mortality data above age 80 from 12 countries, we find that once period shocks are accounted for, there is no statistical evidence of a long-term trend, consistent with Vaupel’s hypothesis. Analyses using lower starting ages yield the same qualitative conclusion. Rather than indicating a change in the process that drives senescence, these variations are consistent with echoes of shared historical events. These results suggest that while longevity has shifted, the rhythm of human aging may be conserved.