Browse Articles
Discover research articles across all indexed journals
A federated transformer-enhanced double Q-network for collaborative intrusion detection
The long non-coding RNA UGDH-AS1 encodes an NK-cell-inhibiting micropeptide in triple-negative breast cancers
Physiological responses of pak choi (Brassica Rapa subsp. Chinensis (L.) Hanelt) to cerium and yttrium in two acidic soils with contrasting textures
Electrifying long-haul freight trucks reduces societal costs in the United States
Abstract Electrifying long-haul heavy-duty vehicles (HDVs) entails high private costs but offers substantial reductions in external costs by substituting diesel combustion with electricity generation. We combine technoeconomic analysis and life-cycle assessment of lithium-ion battery electric (BE) and diesel HDVs to estimate total private costs and monetized climate and health damages in the United States. In 2025, BE-HDVs are estimated to have 46% higher private costs ($0.71 mile⁻¹) than diesel trucks, decreasing to 33% ($0.52 mile⁻¹) by 2035. However, their external costs are 64–69% lower in 2025 and 70–80% lower in 2035. Overall, BE-HDVs yield positive net societal benefits by 2035, contingent on policies that accelerate their adoption.
BlueEdge neural network approach and its application to automated data type classification in mobile edge computing
Abstract Owing to the increasing number of IoT gadgets and the growth of big data, we are now facing massive amounts of diverse data that require proper preprocessing before they can be analyzed. Conventional methods involve sending data directly to the cloud, where it is cleaned and sorted, resulting in a more crowded network, increased latency, and a potential threat to users’ privacy. This paper presents an enhanced version of the BlueEdge framework—a neural network solution designed for the automated classification of data types on edge devices. We achieve this by utilizing a feed-forward neural network and optimized features to identify the presence of 14 distinct data types. Because of this, input data can be preprocessed near its source, and not in the cloud. We utilized a comprehensive dataset comprising 1400 samples, encompassing various data formats from around the world. Compared with rule-based methods, experimental assessment achieves better performance, and results in reduced data transmission (reduced by 62%) and processing latency (78 times faster than cloud-based systems), with resource efficiency comparable to low-end mobile devices. Additionally, our strategy demonstrates strong performance under various data conditions, achieving accuracy levels of over 85% on datasets that may include variations and a noise level as high as 20%. The approach used here is capable of processing data for IoT devices used in education, which can lead to more efficient connections with the cloud and better privacy preservation.
Motorized chromosome models of mitotic chromosome folding
Shallow entangled circuits for quantum time series prediction on IBM devices
Abstract Forecasting temporal dynamics underpins many areas of science and engineering, from large-scale atmospheric prediction to nanoscale quantum control. Classical approaches, including autoregressive models and deep neural networks, have advanced sequential learning often at the expense of known model order, or large dataset and parameters, resulting in computational cost. Here, we investigate whether quantum entanglement can serve as a resource for temporal pattern learning using shallow and structured quantum circuits. We have proposed a Quantum Time Series (QTS) framework that encodes normalised sequential data into single-qubit rotations and captures temporal correlations through forward and cross-entanglement layers. Among several encoding schemes, phase encoding-based sparse entanglement provides hardware efficiency by scaling to larger qubit systems with linear circuit depth and two-qubit complexity of $$\mathcal {O}(n)$$ for qubit size n . This offers a reduction in parameters and depth compared with deep variational quantum circuits such as Heisenberg-inspired circuits, and random-parametric unitary architectures. Experiments on synthetic and geophysical datasets show that shallow QTS circuits reproduce complex temporal pattern from limited data by leveraging structured quantum entanglement. Executions on IBM’s Heron and Eagle-class processors demonstrate robustness and scalability up to 100 qubits. These results suggest that structured entanglement may offer a short-term memory effect for time-series analysis, providing a scalable route for near-term quantum applications.
Cryo-EM structures of plant Augmin reveal coiled-coil assembly, antiparallel dimerization, and NEDD1 binding
Abstract Microtubule (MT) branch nucleation requires Augmin and NEDD1 proteins, which recruit and activate the gamma-tubulin ring complex (γ-TuRC). Augmin is a fork-shaped assembly of eight coiled-coil subunits, while NEDD1 is a β-propeller protein bridging MTs, Augmin, and γ-TuRC. We reconstitute Arabidopsis thaliana Augmin assemblies and determine 3.7-7.3-Å cryo-EM structures of its V-junction and extended regions using crosslinking mass spectrometry. These structures reveal a complete plant Augmin model showing multi-coiled-coil interfaces stabilizing its 40-nm hetero-octameric fork architecture. The dual calponin homology (CH) domains at the V-junction terminus adopt open and closed conformations for MT binding. A 12-Å cryo-EM structure shows Augmin undergoes anti-parallel dimerization through conserved surfaces on its extended region. We determine the NEDD1 β-propeller structure with Augmin, revealing direct binding inside the V-junction that enhances dimerization. Direct coupling and evolutionary analyses identify co-varying residue pairs validating the eight-subunit model and NEDD1 interface. Cooperativity between dual CH domains and NEDD1 binding may regulate V-junction binding to MT lattices. This V-shaped dual binding anchors Augmin along MTs, creating platforms for γ-TuRC recruitment and branched MT nucleation.
Leveraging artificial intelligence for predictive customer churn modeling in telecommunications: a framework for enhanced customer relationship management
Abstract Customer churn remains a critical challenge in the telecommunications industry, impacting profitability and long-term customer value. This study proposes an Artificial Intelligence (AI)-driven framework integrated within Customer Relationship Management (CRM) systems to proactively identify and retain high-risk customers. Using a Random Forest classifier on a publicly available telecom dataset ( N = 2,668), the model achieved an accuracy of 95.13% and an AUC of 0.89 . Techniques such as SMOTE and class weighting were applied to address class imbalance (14.6% churn). Comparative experiments with XGBoost , SVM , and ANN confirmed the robustness of the proposed model. Feature importance analysis revealed that total day minutes, total day charge, and customer service calls were the most influential predictors. The study contributes by linking explainable AI insights to CRM operationalization, providing actionable strategies for proactive customer engagement and retention.
Chemically gated artificial nanochannels for programmable subcellular signal modulated transport regulation
Increased plasma fibronectin mirrors intimal phenotypic switching of vascular smooth muscle cells in moyamoya arteriopathy
n-doping of organic semiconductors catalysed by organometallic complexes
Data-driven analysis reveals distinct genomic and environmental contributions to bacterial growth curves
Assembly and lipid-gating of LRRC8A:D volume-regulated anion channels
Abstract Volume-regulated anion channels (VRACs) are ubiquitously expressed vertebrate ion channels that open in response to hypotonic swelling. VRACs assemble as heteromers of LRRC8A and LRRC8B-E subunits, with different subunit combinations resulting in channels with different properties. Recent studies have described the structures of LRRC8A:C VRACs, but how other VRACs assemble, and which structural features are conserved or variant across channel assemblies remains unknown. Herein, we used cryo-EM to determine structures of a LRRC8A:D VRAC with a 4:2 subunit stoichiometry, which we captured in two conformations. The presence of LRRC8D subunits widens and increases hydrophobicity of the selectivity filter, which may contribute to the unique substrate selectivity of LRRC8D-containing VRACs. The structures reveal lipids bound inside the channel pore, similar to those observed in LRRC8A:C VRACs. We observe that LRRC8D subunit incorporation disrupts packing of the cytoplasmic LRR domains, increasing channel dynamics and opening lateral intersubunit gaps, which we speculate are necessary for pore lipid evacuation and channel activation. Molecular dynamics simulations show that lipids can reside stably within the pore to close the channel. Using electrophysiological experiments, we confirmed that pore lipids block conduction in the closed state, demonstrating that lipid-gating is a general property of VRACs.
Location allocation and capacity optimization for a PV and battery integrated hybrid community electric vehicle charging station
Super El Niño events drive climate regime shifts with enhanced risks under global warming
Effect of microbubble-assisted gemcitabine delivery with repeated ultrasound exposure in a pancreatic cancer organ-on-a-chip model
Abstract The fibrotic stroma of solid tumours poses a physical barrier to drug delivery and effective treatment. Interaction between cancerous epithelial cells and their surrounding stromal partners results in the development of a rigid, collagenous matrix environment with reduced interstitial flow, crucial for drug delivery to cancer cells, particularly in pancreatic ductal adenocarcinoma (PDAC), an aggressive pancreatic cancer with poor prognosis. Therefore, evaluating novel drug delivery mechanisms using appropriate stroma-mimicking 3D culture models is essential. We previously demonstrated, using a 21-day cultured microfluidic PDAC model that mimics the rigid, collagenous stroma, reduced interstitial flow through the tumour model. In this study, we evaluated the use of microbubbles and ultrasound as an alternative method for disrupting our model’s fibrotic stroma to restore interstitial flow and improve gemcitabine delivery and efficacy. Literature shows microbubbles in 2D and 3D static cultures enhance drug delivery and effects by increasing cell membrane permeability through oscillation and bursting under ultrasound (sonoporation). Here, we observed continuous microbubble oscillation and bursting under repeated ultrasound exposure, leading to continuous matrix-microbubble and PDAC cell-microbubble interactions, which improved the gemcitabine effect. This study emphasises the need for disease-specific in vitro models to assess novel drug delivery mechanisms and improve therapeutic outcomes.