Browse Articles
Discover research articles across all indexed journals
An efficient transformer architecture with depthwise separable convolutions for high-accuracy underwater acoustic target recognition
Optical Properties of CdSe/CdZnS Core/Shell Nanoplatelets at High Pressure
Power management of hybrid fuel cell fixed wing UAVs using a fuzzy reinforcement learning system optimized with meta-heuristic methods
Total Synthesis and Anticancer Study of (+)-Verticillin A
Heterogeneous graph transformer and diffusion model for disease diagnosis
Abstract With the continuous development of Electronic Health Record (EHRs), medical heterogeneous data have become increasingly abundant, containing diverse types of entities and complex semantic relationships that provide essential support for disease diagnosis. However, conventional heterogeneous graph neural networks struggle to distinguish semantic differences among multi-type nodes and k-hop neighbors, often leading to semantic confusion and vulnerability to noise, which limits their classification performance and generalization capability. To tackle these challenges, we present a novel framework named Heterogeneous Graph Transformer and Diffusion mechanism for Disease Diagnosis (TD4DD), which captures multi-scale semantic dependencies across k-hop neighborhoods, while the diffusion module performs latent-space denoising to alleviate noise interference. Specifically, a k-hop hierarchical Transformer is introduced to capture multi-scale dependencies across different hop layers, enabling the differentiation of fine-grained semantic variations among neighbors at various distances. Additionally, a diffusion module is designed to handle noise in the data by performing denoising in the latent space through auxiliary subgraphs constructed using different meta-paths, thereby generating more discriminative node representations. Finally, the model fuses structural information with denoised embeddings to accomplish disease classification. Experiments conducted on two real-world clinical datasets, MIMIC-III(7,000 patients, 5 disease categories) and MIMIC-IV(8,331 patients, 6 disease categories), demonstrate that TD4DD consistently outperforms existing baseline methods in terms of both Micro-F1 and Macro-F1 scores, showing strong generalization ability. On MIMIC-III, TD4DD achieves a Micro-F1 of 88.29 and a Macro-F1 of 86.11, while on MIMIC-IV, it reaches 83.60 and 83.94, respectively. Furthermore, ablation studies and t-SNE visualizations further validate the effectiveness of each module and the distinguishing capability of the learned embeddings.
Spatial Confinement Synthesis of Platinum Group Metal Single-Atom Alloy Catalysts for Acidic Hydrogen Evolution
An improved porcine model of infrarenal abdominal aortic aneurysm
Abstract Abdominal aortic aneurysm (AAA) remains a significant public health challenge, primarily due to its high mortality rate and the lack of effective preventive and causal strategies. This study aims to establish an improved translational triple-hit porcine model of AAA and to compare it with human AAA disease. AAA was induced in four juvenile domestic pigs by balloon catheter-based aortic dilation, enzyme-mediated extracellular matrix (ECM) degradation, and lysyl oxidase inhibition. The porcine AAA model was characterized by proteomics, histological investigation, and cytokine profiling, and compared with natural occurred human AAA disease. Infrarenal AAA was successfully established with sustained aortic dilation (> 150% of baseline diameter) occurring within 7–14 days post induction and maintained through day 28. Proteomic analysis of porcine AAA tissue identified significant shifts in protein abundance, including downregulation of proteins associated with vascular smooth muscle cell function and ECM integrity, and upregulation of immune-related proteins. Comparative analysis of porcine and human aortic tissues revealed reduced medial elastic fiber content and length, along with increased calcification in porcine AAA, reflecting structural and pathological changes observed in human AAA. Systemic cytokine profiling revealed significant increases in both pro-inflammatory and anti-inflammatory cytokines following AAA induction in pigs, with a cytokine profile largely comparable to that observed in human AAA patients. These findings suggest that the refined porcine AAA model offers a reliable and reproducible platform for investigating AAA progression and evaluating potential therapeutic interventions prior to clinical implementation.
Acid Catalyst Is Required for Hydrogenolysis of a Late Metal Hydroxo Complex
An intuitionistic fuzzy automated negotiation model for personalized and efficient shared decision-making
Mechanically Strong and Tough Ionic Liquid Gel Electrolyte for Four-Electron Zinc–Iodine Batteries
Two heuristic algorithms for location-inventory-routing models involving two warehouses within multi-echelon supply chain networks
Abstract In supply chain management, the location of facilities, inventory control, and vehicle routing are three key components. This paper incorporates a two-warehouse inventory system into the location- inventory-routing problems (LIRPs) and develops LIRP models with two warehouses in one-level, two-level, and three-level supply chain networks. This study aims to minimize the average total costs of the models by reducing their average costs. To handle these models, two innovative hybrid algorithms, viz. Clarke and Wright—genetic algorithm (CW-GA) and Clarke and Wright—firefly algorithm (CW-FA) are put forward. Computational experiments and sensitivity analyses are conducted to compare the proposed two algorithms with Baron and test the algorithms’ effectiveness and the models’ feasibility. The management implications of this study are presented from two dimensions: model and method. Finally, future research directions and the gap between models and reality are discussed.
Disrupting the Photochemical Landscape of a β-Diketone via Electrostatic Perturbation of Ground-State Tautomers
Metabolic profiling of therapy-induced senescent cancer cells via TPEF, MALDI-MS, and RNA-sequencing
Abstract Despite advances in cancer therapies, treatment failure from resistance and recurrence remains a major clinical challenge. Therapy-induced senescence (TIS), a state of stable cell cycle arrest with sustained metabolic activity, has emerged as a driver of inflammation, tumor persistence, and relapse. However, the heterogeneity of TIS complicates its detection and targeting. Here, we applied a multi-modal strategy to characterize metabolic alterations in senescent cancer cells induced by doxorubicin or γ-irradiation across three tumor cell lines: MCF7, HeLa, and TPC-1. Mitochondrial dysfunction was assessed using MitoTracker and JC-1 staining, while two-photon excitation fluorescence (TPEF) microscopy enabled label-free visualization of intracellular NAD(P)H and FAD distribution. Lipid remodeling was evaluated by MALDI mass spectrometry imaging, and RNA sequencing was performed on control, senescent, and engulfing-senescent MCF7 cells to identify differentially expressed genes and enriched pathways. Senescent cells displayed mitochondrial dysfunction, with altered NAD(P)H/FAD distribution and decreased membrane potential. TPEF confirmed redistribution of coenzymes, reflecting redox changes. Lipidomics revealed consistent remodeling, notably involving cardiolipin precursors. Transcriptomic profiling showed engulfing-senescent MCF7 cells possess a distinct signature marked by increased lipid metabolism, endocrine signaling, and cell–cell communication. Overall, our findings reveal conserved and cell type–specific metabolic traits of TIS, highlighting metabolic vulnerabilities for senolytic intervention.
Kondo Chains of Organic Radicals on Metallic Surfaces: A Model System of Heavy Fermion Quantum Criticality
Geological and physicochemical characterization of Kobe Sorsa graphite deposit, Ethiopia
Peptide Prenylation Follows Divergent Substrate Engagement Rules
Anti-cancer activity of propolis extracts from stingless bees of Brunei Darussalam
<i>Operando</i> Infrared Spectroscopy Insights into Methane C–H Bond Activation and Conversion on Platinum
Doped C20 fullerenes as a new generation of efficient nanosorbents and nanosensors for rapid detection of dimethyltryptamine in drug detection
Abstract The quick and accurate identification of the powerful psychoactive compound Dimethyltryptamine (DMT) is still a significant drawback in forensic science and investigation concerning clinical toxicology. To overcome the issues associated with classical analytical instrumentation, a novel class of innovative highly sensitive nanosensors based on pristine and doped C 20 fullerenes is presented. Using a heavy computational workflow grounded in Density Functional Theory (DFT) at the computational level B3LYP-D3/6-311G(d, p) in the CPCM solvation model (water phase), we systematically examined the sensing properties of pristine C 20 , and the boron (BC 19 ), germanium (GeC 19 ), and silicon-doped (SiC 19 ) C 20 fullerenes to DMT. We use calculated values based on theoretical properties relating to the performance (adsorption energy (Eads), HOMO-LUMO gap (HLG), electrical conductivity (σ), and recovery time (τ)). According to the data, each nanomaterial will have its own unique and promising applications. The BC 19 and SiC 19 nanostructures presented extremely strong adsorption energies for DMT of -40.78 kcal.mol − 1 and − 18.82 kcal.mol − 1 , respectively, and recovery times that indicated effectively irreversible binding. The combination of high Eads and negligible responsibility change in electrical conductivity of BC 19 and SiC 19 suggests they would work well as candidates for adsorption and removal applications where stable analyte capture is desired. On the other hand, the GeC 19 nanosensor showed an unprecedented and selective response, with adsorption of DMT leading to remarkable increases in the electrical conductivity of the nanomaterial of over 16 orders of magnitude, from 3.4 × 10 − 15 S.m − 1 to 1.9 × 10 2 S.m − 1 while exhibiting a relatively strong adsorption energy of -25.75 kcal.mol − 1 . This unique alteration identifies GeC 19 as the top-performing disposable electrochemical sensor for fast, sensitive, and selective detection of DMT. These interactions were also confirmed by NBO, NCI, and QTAIM analyses, which indicated strong charge transfer (NBO), attractive non-covalent interactions (NCI), and medium strength hydrogen bonding (QTAIM) in the BC 19 @DMT, GeC 19 @DMT, and SiC 19 @DMT complexes, respectively. This work not only provides the first theoretical evidence for C 20 -based DMT detection, but also provides a clear pathway for experimental imagining of novel, task-specific nanosensors with important implications for future applications in forensic and point-of-care diagnostics.