Search Results (746)

High-entropy alloys (HEAs) present a fundamental design paradox: their exceptional properties arise from complex, high-dimensional composition–process–microstructure–property (CPMP) relationships, yet the knowledge needed to navigate this space is fragmented across a vast and unstructured literature. Large language models (LLMs) offer a transformative interface to this complexity. By extracting structured facts from text, they can convert dispersed and heterogeneous evidence (i.e., findings scattered across many studies and reported with inconsistent test protocols or characterization standards) into queryable knowledge graphs. Through code generation and tool composition, they can automate simulation pipelines, surrogate model construction, and inverse design workflows. This review analyzes how LLMs can augment key stages of HEA research—from intelligent literature mining and multimodal data integration (using LLMs to automatically extract and structure data from texts and to combine information across text, images, and other data sources) to model-driven design and closed-loop experimentation—illustrated by emerging case studies. We propose concrete evaluation protocols that measure direct scientific utility, including knowledge-graph completeness, workflow setup efficiency, and experimental validation hit rates. We also confront practical limitations: data sparsity and noise, model hallucination, domain bias (where models may exhibit superior predictive performance for specific, well-represented alloy systems over others due to imbalances in training data), and the imperative for reproducible infrastructure. We argue that domain-specialized LLMs, embedded within grounded, verifiable research systems, can not only accelerate HEA discovery but also standardize the representation, sharing, and reuse of community knowledge. Full article

Large Language Model (LLM) agents depend heavily on multiple external tools such as APIs, databases and computational services to perform complex tasks. However, these tool executions create latency and introduce costs, particularly when agents handle similar queries or workflows. Most current caching methods focus on LLM prompt–response pairs or execution plans and overlook redundancies at the tool level. To address this, we designed a multi-level caching architecture that captures redundancy at both the workflow and tool level. The proposed system integrates four key components: (1) hierarchical caching that operates at both the workflow and tool level to capture coarse and fine-grained redundancies; (2) dependency-aware invalidation using graph-based techniques to maintain consistency when write operations affect cached reads across execution contexts; (3) category-specific time-to-live (TTL) policies tailored to different data types, e.g., weather APIs, user location, database queries and filesystem and computational tasks; and (4) session isolation to ensure multi-tenant cache safety through automatic session scoping. We evaluated the system using synthetic data with 2.25 million queries across ten configurations in fifteen runs. In addition, we conducted four targeted evaluations—write intensity robustness from 4 to 30% writes, personalized memory effects under isolated vs. shared cache modes, workflow-level caching comparison and workload sensitivity across five access distributions—on an additional 2.565 million queries, bringing the total experimental scope to 4.815 million executed queries. The architecture achieved 76.5% caching efficiency, reducing query processing time by 13.3× and lowering estimated costs by 73.3% compared to a no-cache baseline. Multi-tenant testing with fifteen concurrent tenants confirmed robust session isolation and 74.1% efficiency under concurrent workloads. Our evaluation used controlled synthetic workloads following Zipfian distributions, which are commonly used in caching research. While absolute hit rates vary by deployment domain, the architectural principles of hierarchical caching, dependency tracking and session isolation remain broadly applicable. Full article

Thrombosis is a cardiovascular disease characterized by the pathological formation of a fibrin clot in blood vessels. Currently available fibrinolytic enzymes have some limitations, including severe side effects, high cost, short half-life, and low fibrin specificity. Proteins from microalgae and cyanobacteria have various biological effects and are emerging as promising sources for fibrinolytic enzymes. In this study, bioinformatics tools were used to evaluate the intrinsic disorder predisposition of microalgal fibrinolytic proteins, their capability to undergo liquid–liquid phase separation (LLPS), and the presence of disorder-based functional regions, and short linear motifs (SLiMs). Analysis revealed that these proteins are predominantly hydrophilic and exhibit acidic (pI 3.96–6.49) or basic (pI 8.05–11.0) isoelectric points. Most of them are expected to be moderately (61.4%) or highly disordered proteins (6.8%) and associated with LLPS, with nine proteins being predicted to behave as droplet drivers (i.e., being capable of spontaneous LLPS), and twenty-five proteins being expected to be droplet clients. These observations suggest that LLPS may be related to the regulation of the functionality of microalgal fibrinolytic proteins. The majority of these proteins belong to the blood coagulation inhibitor (disintegrin) 1 hit superfamily, which can inhibit fibrinogen binding to integrin receptors, preventing platelet aggregation. Furthermore, the SLiM-centered analysis indicated that the main motifs found in these proteins are MOD_GlcNHglycan and CLV_PCSK_SKI1_1, which can also play different roles in thrombolytic activity. Finally, Fisher and conservation analysis indicated that CLV_NRD_NRD_1, CLV_PCSK_FUR_1, CLV_PCSK_PC7_1, and MOD_Cter_Amidation motifs are enriched in intrinsically disordered regions (IDRs) of these proteins, showing significant conservation and suggesting compatibility with proteolytic activation and post-translational processing. These data provide important information regarding microalgal proteins with potential thrombolytic effects, which can be realized through protein–protein interactions mediated by SLiMs present in intrinsically disordered regions (IDRs). Additional analyses should be conducted to confirm these observations using experimental in vitro and in vivo approaches. Full article

A synthetically accessible library of N-acyl sulfonamides was constructed using a combination of copper(I)-catalyzed azide–alkyne cycloaddition (CuAAC) and N-acylation of primary sulfonamides. The proposed two-step reaction sequence had a high experimentally confirmed synthetic success rate (up to 85%) and gave reasonable product yields (up to 61%). As a result of the validation process, a 262-member compound library (out of >70K accessible combinations) was prepared. Biological profiling of the synthesized library by differential scanning fluorimetry and enzymatic assays identified several low micromolar inhibitors of human carbonic anhydrase. The interaction of the discovered hits with the biological target was studied by docking and molecular dynamics. Full article

Although the clownfish, Amphiprion ocellaris (A. ocellaris), is a popular ornamental marine fish worldwide, the mechanisms underlying color pattern variation remain unclear. Given that the Platinum-type clownfish, nearly entirely white, has high economic value, understanding the biological mechanism that accounts for the difference between orange and white colors in A. ocellaris is crucial. To investigate these coloration differences, we performed RNA sequencing analysis and identified differentially expressed genes (DEGs) by comparing white and orange skin samples from three A. ocellaris individuals. A total of 76 DEGs were detected, including 56 downregulated and 20 upregulated genes. DEG sequences were annotated using Danio rerio and Stegastus partitus as reference species, selecting the best hit based on the lowest E-value. A protein–protein interaction (PPI) network and Gene Ontology biological process terms were additionally analyzed. Several DEGs previously reported to be associated with pigmentation, including hpdb, cldn11b, sfrp5, slc2a9, slc2a11b, si:ch211-256m1.8, fhl2, rab38, and ttc39b were identified. Based on the functions of these DEGs, it is inferred that leucophores and xanthophores contribute to both white and orange coloration by modulating related genes, including slc2a11b and slc2a9. Additionally, sfrp5, sost, and sp7 genes were identified to interact with each other in the PPI analysis, with sfrp5 and sost being associated with the Wnt signaling pathway, which contributes to melanocyte specification and osteoblast differentiation. Based on these findings, we propose sost and sp7 as candidate genes that might provide insights relevant to extreme white pigmentation phenotypes, such as those observed in Platinum-type clownfish. For a clearer understanding, further studies integrating quantitative genetics and functional analyses are required. Full article

Currently, with the widespread popularization of e-commerce systems, enterprises have increasingly high requirements for the timeliness of order fulfillment. It has become particularly critical to enhance the operational efficiency of heterogeneous robotic “goods-to-person” (G2P) systems in book e-commerce fulfillment, reduce enterprise operational costs, and achieve highly efficient, low-carbon, and sustainable warehouse management. Therefore, this study focuses on determining the optimal storage location assignment strategy and order batching method. By comprehensively considering the characteristics of book e-commerce, such as small-batch, high-frequency orders and diverse SKU requirements, as well as existing system issues including uncoordinated storage assignment and order processing, and differences in the operational efficiency of heterogeneous robots, this study proposes a joint optimization framework for storage location assignment and order batching centered on a multi-objective model. The framework integrates the time costs of robot picking operations, SKU turnover rates, and inter-commodity correlations, introduces the STCSPBC storage strategy to optimize storage location assignment, and designs the SA-ANS algorithm to solve the storage assignment problem. Meanwhile, order batching optimization is based on dynamic inventory data, and the S-O Greedy algorithm is adopted to find solutions with lower picking costs. This achieves the joint optimization of storage location assignment and order batching, improves the system’s picking efficiency, reduces operational costs, and realizes green and sustainable management. Finally, validation via a spatiotemporal network model shows that the proposed joint optimization framework outperforms existing benchmark methods, achieving a 45.73% improvement in average order hit rate, a 48.79% reduction in total movement distance, a 46.59% decrease in operation time, and a 24.04% reduction in conflict frequency. Full article

There has been significant interest in solving robot path planning problems using fuzzy logic-based methods. Recently, the Genetic Algorithm-based Hierarchical Interval Type-2 Fuzzy (GA-HIT2F) system has been introduced as a novel planner in this domain. However, this method lacks adaptability to changes in target points, and insufficient flexibility can lead to planning failures in local minimum traps, making it difficult to apply to complex scenarios. In this paper, we identify the limitations of the original GA-HIT2F approach and propose an enhanced Q-Learning-aided Adaptive Hierarchical Interval Type-2 Fuzzy (QL-HIT2F) algorithm for path planning. The proposed planner incorporates reinforcement learning to improve a robot’s capability to avoid collisions with special obstacles. Additionally, the average obstacle orientation (AOO) is introduced to optimize the robot’s angular adjustments. Two supplementary robot parameters are integrated into the reinforcement learning action space, along with fuzzy membership parameters. The training process also introduces the concepts of meta-map and sub-training. Simulation results from a series of path planning experiments validate the feasibility and effectiveness of the proposed QL-HIT2F approach. Full article

Temporal knowledge graphs (TKGs) incorporate temporal information into traditional triplets, enhancing the dynamic representation of real-world events. Temporal knowledge graph reasoning aims to infer unknown quadruples at future timestamps through dynamic modeling and learning of nodes and edges in the knowledge graph. Existing TKG reasoning approaches often suffer from two main limitations: neglecting the influence of temporal information during entity embedding and insufficient or unreasonable processing of relational structures. To address these issues, we propose DERP, a relation-aware reasoning model with dynamic evolution mechanisms. The model enhances entity embeddings by jointly encoding time-varying and static features. It processes graph-structured data through relational graph convolutional layers, which effectively capture complex relational patterns between entities. Notably, it introduces an innovative relational-aware attention mechanism (RAGAT) that dynamically adapts the importance weights of relations between entities. This facilitates enhanced information aggregation from neighboring nodes and strengthens the model’s ability to capture local structural features. Subsequently, prediction scores are generated utilizing a convolutional decoder. The proposed model significantly enhances the accuracy of temporal knowledge graph reasoning and effectively handles dynamically evolving entity relationships. Experimental results on four public datasets demonstrate the model’s superior performance, as evidenced by strong results on standard evaluation metrics, including Mean Reciprocal Rank (MRR), Hits@1, Hits@3, and Hits@10. Full article

Background/Objectives: Increasing evidence supports the hypothesis that dietary intervention can improve pain among individuals with headaches, including migraine, a highly prevalent condition that can be disabling. Non-pharmacologic treatments for migraine are particularly attractive. In this secondary analysis of 182 participants enrolled in a randomized controlled trial of a dietary intervention designed to increase omega-3 (n-3) compared with a control diet, we examined the effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), both thought to decrease inflammatory processes. Methods: Path models with two time points (baseline and 16 weeks after randomization), were used to test the relationships between exposures of n-3 blood levels and self-reported dietary intake on outcomes of pain interference using the PROMIS pain interference scale and the Headache Impact Test (HIT-6). Model building was based on our published conceptual model. Results: Good fit was demonstrated for both models (EPA model: CFI = 0.984, RMSEA = 0.039, and SRMR = 0.045; DHA model: CFI = 0.981, RMSEA = 0.040, and SRMR = 0.040). Both EPA and DHA in the blood at 16 weeks were associated with lower levels of pain interference, but the effect for EPA was stronger (B = −0.56, p < 0.001 for EPA, and B = −0.43, p = 0.057 for DHA). Conclusions: Our findings are consistent with an indirect pathway linking diet to pain interference through blood levels of EPA and DHA in migraine. Full article

The emergence of multidrug-resistant Staphylococcus aureus underscores the urgent need for novel therapeutic agents targeting essential bacterial pathways. The lipoteichoic acid synthase (LtaS) is crucial for the synthesis of lipoteichoic acid in the cell wall of Gram-positive bacteria and represents a promising and vulnerable target for antimicrobial drug development. This study employed a comprehensive computational pipeline to identify potent inhibitors of the LtaS enzyme. A library of natural compounds was retrieved from the COCONUT database and screened against the crystal structure of the extracellular domain of LtaS (eLtaS) (PDB ID: 2W5R, obtained from the Protein Data Bank) through a multi-stage molecular docking strategy. This process started with High-Throughput Virtual Screening (HTVS), followed by Standard Precision (SP) docking, and culminated in Extra Precision (XP) docking to refine the selection of hits. The top-ranking compounds from XP docking were subsequently subjected to MM-GBSA binding free energy calculations for further filtration. The stability and dynamic behavior of the resulting candidate complexes were then evaluated using 100 ns molecular dynamics (MD) simulations, which confirmed the structural integrity and binding stability of the ligands. Density Functional Theory calculations revealed that screened ligands exhibit improved electronic stabilization and charge-transfer characteristics compared to a reference compound, suggesting enhanced reactivity and stability relevant for hit identification. Finally, ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) profiling was conducted to assess the drug-likeness and pharmacokinetic safety of the lead compounds. These findings support them as promising orally active leads for further optimization. Our integrated approach shortlisted eight initial hits (A–H) that showed interesting scaffold diversity and finally identified two compounds, herein referred to as Compound A and Compound B, which demonstrated stable binding, favorable free energy, and an acceptable Absorption, Distribution, Metabolism, and Excretion, and Toxicity (ADMET) profile. These candidates emerge as promising starting points for developing novel anti-staphylococcal agents targeting the LtaS enzyme that cand be further proved by experimental validation. Full article

Background: There is a critical need for feasible, non-equipment based, safe, and cost-effective exercise interventions to promote muscle strength, dynamic postural balance, and independent mobility in adolescents with cerebral palsy (CP) or spina bifida (SB). Objectives: This study aimed to examine the feasibility and preliminary response of a novel exercise program: Functionally Loaded High-Intensity Circuit Training (FUNHIT) and conventional High-Intensity Circuit Training (HIT) in adolescents with CP/SB. Methods: Enrolled participants were allocated to FUNHIT or HIT or Controls in our randomized control trial. The interventions were delivered 2×/week × 4 weeks. Feasibility was assessed through process, operational, and scientific metrics. Outcome measures included maximum walking speed, Four Square Step Test (FSST), Timed Up and Go (TUG) and its dual-task variants, Lateral Step-Up Test (LSUT), Fear of Falling (FoF) and physical activity (PA) questionnaires. Results: We tested 5 participants (1 CP, 4 SB) in our study. Recruitment and retention rates were acceptable (63% enrollment, 100% retention and adherence). FUNHIT (n = 2) participants showed improvements in maximum walking speed (8–12%), FSST (15–29%), LSUT (22–33%), and TUG (4%). The HIT participant (n = 1) demonstrated improved TUG dual-task performance (40%) and FSST (30%) only. Control participants (n = 2) had varied changes (from 0–24%) in mobility, strength, balance. No adverse events were reported. Participants successfully followed (100%) the prescribed exercise dosage over the four-week period. Conclusions: FUNHIT and HIT are feasible and safe interventions for adolescents with ambulatory CP and SB who retain motor function, showing promising preliminary improvements in muscle strength, dynamic balance, and independent mobility. Our findings need to be validated in larger samples. Full article

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most prevalent chronic liver disorder in both children and adults. Pediatric MASLD, however, is not simply an early form of adult disease, as it exhibits distinct developmental, histological, and metabolic features. Emerging evidence suggests that these characteristics arise from a complex, multi-hit continuum that begins in utero. Maternal obesity, gestational diabetes, and poor diet quality during pregnancy have been associated with greater hepatic steatosis in offspring, raising the possibility that intrauterine exposure to dyslipidemia, hyperglycemia, and elevated free fatty acid flux may contribute to early hepatic lipid deposition. After birth, feeding behaviors such as a prolonged breastfeeding appear protective, whereas formula feeding, especially high added-sugar formulations, may accelerate rapid weight gain and increase susceptibility to later steatosis. Early childhood diets high in added sugars, saturated fats, and ultra-processed foods may further promote hepatic lipogenesis and inflammation and interact with underlying genetic susceptibility. Given the heterogeneity of available human cohort studies and mechanistic model systems, this narrative review summarizes converging evidence from prenatal, postnatal, and early childhood nutritional exposures and their relationship to offspring hepatic lipid accumulation, emphasizing early-life windows for intervention to reduce the burden of pediatric MASLD. Full article

A new series of angular tricyclic pyrrolo[3,2-c]quinoline derivatives (PQs) was designed and synthesized to further explore the previously promising antiproliferative activity exhibited by the 4-benzodioxole-substituted hit 7d. Accordingly, several structural modifications mainly focused on the benzodioxole moiety were introduced, allowing us to gain new insights into the activity and biological profile. NCI antiproliferative screening (SRB colorimetric assay), together with MTS-based assay against six other tumour cell lines, enabled us a deeper understanding of the selectivity and potency patterns. This led to the identification of a new promising hit, compound 7p, which exhibited cytotoxic activity in the low micromolar range against MCF-7 and HeLa cells. Further biological evaluations, including apoptosis induction, clonogenic, and scratch tests, provided additional biological insights into the anticancer potential of these compounds, supporting the subsequent lead optimization process for more potent anticancer activity. The integrated in silico docking results evidenced a clear multi-target profile, as testified by the broad anticancer activity, and suggest a good potential for rational polypharmacology. Full article

The operating temperature of lithium batteries directly affects their charge–discharge performance. This study is based on the LF50K prismatic power battery. The battery’s thermal model and the computational fluid dynamics (CFD) control equation were established. After completing the model verification, a thermal management system with a bionic leaf vein flow channel was designed. The study focused on investigating the effects of varied flow passage configurations, inlet–outlet flow channel angles, flow channel widths, flow rates, leaf vein angles, and inlet–outlet positions on the cooling effect of the lithium battery module. The results show that, as the inlet–outlet angle and width of the bionic leaf vein fluid flow channel increase, the battery cooling effect deteriorates; the increase in the angle and flow channel width has an adverse impact on battery heat dissipation. The significant reduction in the battery’s maximum temperature observed with an elevated fluid flow rate underscores the positive contribution of flow rate to the cooling process. The effect of the leaf vein angle on the cooling of lithium batteries shows a fluctuating trend: when the angle rises from 30° to 45°, the battery’s peak temperature shows a slow upward tendency; conversely, with the angle further increasing from 45° to 80°, the maximum temperature shows a gradual downward tendency. Specifically, at an angle of 45°, Battery No. 5 hits a maximum temperature of 306.58 K (around 33.43 °C), with the maximum temperature difference also reaching 6.38 K. After optimizing the structural parameters, when operating under the maximum ambient temperature conditions in 2024, the maximum temperature of the battery module decreased by 7 K, and the temperature difference decreased by 5.47 K, enabling the battery to achieve optimal operating efficiency. This study lays a foundation for a further optimization of the thermal management system for lithium-ion batteries in subsequent research. Full article

Frequent fluctuations in virtual item transactions make data access in virtual games highly dynamic. These heat changes denote temporal variations in data popularity driven by trading activity, which in turn cause traditional storage systems to struggle with timely heat adaptation, increased latency, and energy waste. This study proposes an AI-driven modeling framework for virtual game data centers. The heat feature vector composed of transaction frequency, price fluctuation, and scarcity forms the state space of a Markov decision process, while data migration between multi-layer storage structures constitutes the action space. The model captures temporal locality and spatial clustering in transaction behaviors, applies a sliding-window prediction mechanism to estimate access intensity, and enhances load perception. A scheduling mechanism combining an R2D3 (Recurrent Replay Distributed DQN from Demonstrations) policy network with temporal attention and mixed integer programming jointly optimizes latency, energy consumption, and resource constraints to achieve global data allocation tuning. Experiments on a simulated high-frequency trading dataset show that the system reduces access delay to 420 ms at a transaction intensity of 1000 per second and controls the total migration energy consumption to 85.7 Wh. The Edge layer achieves a peak hit rate of 63%, demonstrating that the proposed method enables accurate heat identification and energy-efficient multi-layer scheduling under highly dynamic environments. Full article