Many patients with chronic myeloid leukemia (CML) treated with adenosine triphosphate (ATP)-competitive tyrosine kinase inhibitors (TKIs) experience persistent adverse events (AEs) that negatively impact daily living and the ability to remain on treatment. Asciminib, an allosteric inhibitor of BCR::ABL1, was designed to enhance efficacy and reduce off-target effects vs ATPcompetitive TKIs. The phase 3 randomized ASC4FIRST trial established the overall favorable safety profile of asciminib in patients with newly diagnosed CML in chronic phase (CP). This exploratory post hoc analysis of ASC4FIRST focused specifically on the tolerability of asciminib vs imatinib and asciminib vs second-generation [2G] TKIs. Analyses were conducted within each stratum to account for differences between strata; patients prerandomized to the imatinib stratum were older and had higher cardiovascular risk than those in the 2G stratum. Within both strata, patients receiving asciminib experienced fewer difficult-to-tolerate AEs (such as gastrointestinal toxicity, rash, and pleural effusion) and fewer AEs leading to dose modifications and discontinuations due to nonhematologic and hematologic AEs vs the investigator-selected (IS) TKI comparator, with a shorter median duration of dose modification. Additionally, median onset of AEs leading to dose modification occurred later in patients receiving asciminib vs ISTKIs. The safety and tolerability of asciminib observed in the ASC4FIRST trial demonstrate asciminib's excellent benefit-risk profile as a frontline therapy for a broad range of patients with newly diagnosed CML-CP.
The conventional 4-aminoantipyrine (4-AAP) method for volatile phenols (VPs) fails to detect certain substituted phenols, leading to an underestimation of VPs in water. To overcome this "selectivity blind spot", we developed an integrated flow-injection analysis chemiluminescence (FIA-CL) platform that quantifies VPs as phenol-equivalent concentrations, providing a more inclusive readout than the 4-AAP method and enabling rapid and reliable monitoring. This sensing mechanism utilizes a competitive reaction where VPs compete with the CL probe (naproxen, NAP) for reactive iron species (Fe(IV)) generated by Fe(II)-activated periodate, resulting in concentration-dependent signal attenuation. Moreover, to address signal loss due to separation between mixing and detection units in conventional FIA-CL systems, we designed a Y-shaped flow-through cell that synchronizes mixing and detection within the same microzone for real-time signal capture. The optimized platform yielded a detection limit of 3.47 μg L-1 with a linear range of 0.011-20 mg L-1, outperforming the 4-AAP method with a 3-fold lower LOD and an 8-fold wider range. Analysis of real water samples showed that VPs concentrations measured by the platform exceeded those obtained by the 4-AAP method, with mass spectrometry confirming that this discrepancy arises from 4-AAP's failure to respond to certain substituted phenols. Thus, our platform provides a sensitive, broad-spectrum tool for VPs monitoring.
Maize-soybean intercropping is a critical ecological planting technique that embodies the essence of China's intercropping agricultural practices, enhancing farmland productivity while preserving soil health and ensuring sustainable production. However, the interplay between nitrogen management and planting density, along with interspecific interaction/competition dynamics in maize-soybean strip intercropping under rain-fed conditions, remains unclear; elucidating these factors is critical for optimizing resource utilization efficiency in arid farming systems. This study employed a split-plot field design under maize-soybean intercropping conditions, with maize planting density as the main plot factor (52500, 60000 and 67500 plants ha⁻¹) and nitrogen application rate as the subplot factor (300, 375 and 450 kg ha⁻¹). Maize and soybean monocultures were included as controls. Results demonstrated that increasing planting density and nitrogen application significantly affected crop growth, photosynthetic characteristics, yield, and economic return in the maize-soybean intercropping system. With increasing planting density from D1 to D3, maize yield increased first and then declined, whereas soybean yield decreased overall. Increasing planting density altered maize growth traits, suppressed soybean growth, decreased leaf vapor pressure deficit(VPD), and generally increased net photosynthetic rate (Pn). With increasing nitrogen application from N1 to N3, crop growth and yield formation were improved under D2, but under D3, excessive nitrogen input reduced transpiration and weakened productivity. Compared with D1, maize yield under D2 increased by 28.76%, whereas soybean yield under D3 decreased by 36.15%. Therefore, N2D2 (60,000 plants ha⁻¹ combined with 375 kg N ha⁻¹) was identified as the optimal planting strategy for maize-soybean intercropping under rain-fed conditions in southern Ningxia.
In 1979, a young associate professor at Harvard Business School published his first article for HBR, "How Competitive Forces Shape Strategy." In the years that followed, Michael Porter's explication of the five forces that determine the long-run profitability of any industry has shaped a generation of academic research and business practice. In this article, Porter undertakes a thorough reaffirmation and extension of his classic work of strategy formulation, which includes substantial new sections showing how to put the five forces analysis into practice. The five forces govern the profit structure of an industry by determining how the economic value it creates is apportioned. That value may be drained away through the rivalry among existing competitors, of course, but it can also be bargained away through the power of suppliers or the power of customers or be constrained by the threat of new entrants or the threat of substitutes. Strategy can be viewed as building defenses against the competitive forces or as finding a position in an industry where the forces are weaker. Changes in the strength of the forces signal changes in the competitive landscape critical to ongoing strategy formulation. In exploring the implications of the five forces framework, Porter explains why a fast-growing industry is not always a profitable one, how eliminating today's competitors through mergers and acquisitions can reduce an industry's profit potential, how government policies play a role by changing the relative strength of the forces, and how to use the forces to understand complements. He then shows how a company can influence the key forces in its industry to create a more favorable structure for itself or to expand the pie altogether. The five forces reveal why industry profitability is what it is. Only by understanding them can a company incorporate industry conditions into strategy.
Conventional identification of stable HFpEF still depends on resource-intensive clinical assessment, whereas non-invasive acoustic analysis of sustained vowels may provide a complementary and accessible classification signal. The objective of this study was to develop and validate a classification model for stable heart failure with preserved ejection fraction (HFpEF) versus healthy controls using acoustic features extracted from the sustained vowel /ɑː/. In this retrospective case-control study, voice recordings were obtained from a primary cohort of 341 participants and an independent external validation cohort of 172 participants. The present study extracted 384 features using the Interspeech 2009 feature set and compared five deep learning models. The top-performing deep learning architecture, the multilayer perceptron (MLP), was further assessed through tenfold cross-validation, independent external validation, and feature analysis using SHapley Additive exPlanations and Local Interpretable Model-Agnostic Explanations. Additional analyses benchmarked the MLP against classical machine-learning models and examined uncertainty, calibration, prevalence sensitivity, and demographic confounding. For the binary classification of stable HFpEF versus healthy controls, the MLP classifier demonstrated the strongest performance in the five-model deep learning comparison, achieving a mean tenfold cross-validation accuracy of 0.8593 and an area under the curve (AUC) of 0.9130. On the independent external validation cohort, the primary MLP achieved an AUC of 0.836. In a broader benchmark, acoustic-feature models clearly outperformed age/sex-only models, and MLP remained competitive against strong classical tabular baselines. A regularized MLP sensitivity model achieved an external AUC of 0.8638 (95% bootstrap CI 0.8074-0.9188), whereas the best untuned SVM achieved an external AUC of 0.8628. Interpretability analyses highlighted MFCC-related spectral-envelope descriptors as influential feature families, although fold-wise stability supported family-level rather than single-feature conclusions. While the performance gap between cross-validation and external validation indicates room for further refinement, this study demonstrates that sustained-vowel IS09 acoustic features contain discriminative signal for stable HFpEF versus healthy controls under controlled conditions. The MLP was the strongest model among the tested deep learning architectures, but the broader benchmark showed that it was competitive rather than clearly superior to classical tabular models. These findings support acoustic-feature analysis as a preliminary research-stage adjunctive classification signal, while broader prospective validation in clinically heterogeneous populations is required before any clinical screening, triage, or monitoring role can be established.
Urban traffic flow prediction is a fundamental task in intelligent transportation systems, yet it remains exceptionally challenging due to the entangled nature of spatial heterogeneity, non-stationary temporal dynamics, and multi-scale periodicity in real-world road networks. Existing graph neural network (GNN)- and transformer-based methods often treat spatial and temporal modeling as largely independent components, thereby overlooking the synergistic interactions between structural topology and sequential patterns. To address these limitations, we propose ST-GNNFormer, a novel Hybrid Spatio-Temporal Graph Transformer that tightly couples adaptive graph learning with multi-scale temporal attention for fine-grained traffic prediction. Specifically, ST-GNNFormer consists of four collaborating modules: (i) an Adaptive Dynamic Graph Learning (ADGL) module that infers time-varying adjacency from learnable node embeddings conditioned on temporal context, capturing both structural and semantic proximity; (ii) a Spatial Graph Transformer (SGT) that integrates graph-structure biases into multi-head self-attention to propagate spatially correlated features over the learned topology; (iii) a Temporal Multi-Scale Transformer (TMST) that simultaneously models short-term fluctuations and long-range periodicity across multiple temporal granularities; and (iv) a Cross-Scale Fusion (CSF) gate that adaptively merges spatial and temporal representations at each encoder layer. Extensive experiments on four public benchmarks-METR-LA, PEMS-BAY, PEMS04, and PEMS08-demonstrate that ST-GNNFormer consistently outperforms 8 competitive baselines by up to 7.5% in MAE at the 60-minute prediction horizon while maintaining competitive computational efficiency. Ablation studies and visualization analyses confirm the individual contributions of each module and reveal physically meaningful attention patterns that align with real-world traffic dynamics.
This research examines the intersection of education and political systems, exploring how instructional practices align with neoliberal policies. Education is often viewed as a means to address societal issues such as unemployment and inequality; however, its emphasis on creativity, critical thinking, and ethical considerations may sometimes be overlooked. A shift toward fostering critical engagement and societal development could be benefi cial. Using a descriptive approach, this study retrieved articles from the Web of Science and Scopus databases. In Iran, the education system has undergone notable changes since the Islamic Revolution in 1979, primarily aligning with ideological objectives. While enrollment rates have increased, the system appears to face challenges such as outdated curricula, structural ineffi ciencies, centralized governance, and limitations in teacher training. Although reforms have aimed at reinforcing ideological priorities, their impact on fostering innovation and global competitiveness remains uncertain. Additionally, regional disparities and gender inequalities continue to be areas of concern. Key challenges may include an overreliance on rote learning, limited adoption of modern pedagogical methods, and insuffi cient coordination between research and policy implementation. Teacher motivation could also be aff ected by inadequate wages and institutional support, potentially infl uencing the overall quality of education. Furthermore, both teachers and students encounter barriers to accessing equitable and high-quality education, which may hinder educational progress. Psychological concerns among Iranian students appear to be rising, possibly due to academic stress, suboptimal educational environments, and family dynamics. Research suggests that supportive family and school settings may play a signifi cant role in improving mental wellbeing, motivation, and self-perception. Adolescents, in particular, seem to benefi t from strong familial bonds, which could positively impact their mental health and academic performance. Based on these fi ndings, it may be advisable to consider reducing political infl uence in education, modernizing curricula, investing in teacher re-training, and integrating psychological support within schools. Decentralizing governance and fostering innovation could contribute to a more dynamic and responsive education system. While Iran's education system has made strides in enrollment, addressing these structural and pedagogical challenges could enhance its ability to prepare students for the demands of modern society while supporting their psychological and social well-being. (Neuropsychopharmacol Hung 2026; 28(2): 115-130) Keywords: Educational, Iran, Psychology, current trends.
Spatial transcriptomics (ST) enables the precise mapping of gene expression within tissue architecture, however its application is often limited by low spatial resolution and sparse sampling. While existing deep learning methods leverage histology images, spatial coordinates, or low-resolution expression data to predict high-density profiles, these methods are limited in either capturing the intrinsic constraints between histological context and spatial topology or ignoring the complex local neighborhood relationships between spots. To address these limitations, we propose SpaBiT, a multimodal framework designed to enhance ST resolution via a bidirectional attention mechanism. At its core, SpaBiT employs a bidirectional cross-attention module to facilitate precise information exchange between image features and neighborhood-aware representations learned via a graph attention network. This design explicitly models the synergistic constraints between local morphology and spatial graph topology, yielding high-fidelity, high-density gene expression maps. SpaBiT exhibits competitive performance in reconstructing complex spatial gene expression, outperforming the benchmark models utilized in this study across various quantitative metrics, providing a robust tool for deciphering complex tissue microenvironments. The source code and datasets are available at https://github.com/wenwenmin/SpaBiT. Supplementary data are available at Bioinformatics online.
Vinyl ethermaleic anhydride (VEMA) copolymers were synthesized by reversible addition-fragmentation chain transfer (RAFT) copolymerization and hydrolyzed to generate dicarboxylate-functionalized adsorbents for heavy metal removal from water. The RAFT-synthesized copolymer had superior removal efficiencies compared to a conventional free-radical synthesized analogue, while maintaining adsorption performance over multiple cycles. The metals could be removed under batch and flow conditions. Competitive experiments demonstrated adsorption in the order: Cu(II) > Fe(III) > Ni(II) > Co(II) > Zn(II). The impact of pH and metal concentration was investigated. Overall, hydrolyzed VEMA copolymers are promising reusable adsorbents in water treatment applications.
Traditional linear approaches may not adequately capture the non-linear relationships and interactions among laboratory-derived sprint test metrics. This study aimed to predict flying sprint performance of elite-level male track cyclists using multiple linear regression and random forest models based on anaerobic ergometer metrics. A total of 333 elite male track cyclists completed a 30-s all-out cycle-ergometer sprint test and indoor-velodrome flying 100-m and 200-m sprints. Eight predictors derived from the ergometer test result (body mass, peak power output, 30-s mean power, mean cadence, relative peak power, 5-s maximal mean power normalized to body mass, maximal 5-s power decline, and percentage power drop) were used to develop multiple linear regression and tuned random forest models. Model performance was evaluated on a held-out test set, with random forest hyperparameters optimized via nested cross-validation on the training data. The 100-200-m split time (flying 200-m minus flying 100-m time) was examined in unadjusted and peak power-adjusted models to assess the independent association of power decline metrics with second-half sprint performance. In mutually adjusted multiple linear regression models, 30-s average power was the only independent predictor of flying 100-m and 200-m times (p < 0.001). Test-set predictive performance was modest for linear regression (R2 = 0.326 for 100 m; R2 = 0.329 for 200 m) and marginally higher for the tuned random forest models (R2 = 0.369 for 100 m; R2 = 0.393 for 200 m). Random forest importance analyses consistently ranked relative peak power and 30-s average power as the most influential predictors across both outcomes. Percentage power drop was not independently associated with the 100-200-m split in either unadjusted or peak power-adjusted models. Sprint performance could be predicted with modest accuracy from routinely collected ergometer metrics, with sustained power and relative peak power emerging as the primary contributors. Power decline metrics showed no independent association with second-half sprint performance after accounting for peak power, suggesting that fatigue resistance, as measured by a 30-s all-out cycle-ergometer sprint test, may not directly translate to competitive second-half sprint outcomes.
Long-term time series forecasting (LTSF) underpins critical applications from energy management to weather prediction, yet achieving reliable multi-step-ahead accuracy remains challenging. Existing LTSF approaches, dominated by MLP- and Transformer-based architectures, either rely on simple linear mappings or introduce increasingly complex hand-crafted inductive biases, raising the question of whether a more expressive nonlinear modeling core could offer a useful alternative. In this work, we investigate whether Kolmogorov-Arnold Networks (KANs), which use learnable basis functions on network edges to model nonlinear relationships, can serve as effective modeling components for LTSF, and under which design choices they are most useful. Motivated by this question, we propose KANMixer, a compact KAN-centered architecture consisting of a multi-scale pooling frontend, KAN-based temporal mixing blocks, and KAN-based prediction heads. Unlike KAN-based forecasting models that combine KAN with decomposition-heavy or mixture-based pipelines, KANMixer is designed as a simple and controlled architecture for examining the role of KAN components in LTSF. Under a unified five-run reproduction protocol on seven standard benchmarks, KANMixer achieves competitive performance against representative LTSF baselines, especially on ETT-style datasets, while showing dataset-dependent limitations. Additional statistical tests, ablations, efficiency profiling, Gaussian-noise evaluation, and hyperparameter sensitivity analysis show that the practical value of KAN depends on basis-function choice, architectural placement, and computational constraints. These results suggest that KANs are promising but not plug-and-play components for LTSF, and that their benefits should be evaluated together with robustness and efficiency trade-offs.
There is growing optimism regarding the potential therapeutic and preventive benefits of regulating intestinal microbiota for various diseases. Diet is one of the most straightforward and safest methods for modulating the intestinal microbiota; however, considerable individual differences have been observed in the microbiota response to dietary interventions. These individual differences pose substantial challenges in application, which are primarily attributed to variations in the commensal flora and bacterial competition for nutrients. Our previous research indicated that the microscopic localization of bacteria provides valuable insights into the mechanisms by which specific intestinal bacterial species acquire nutrients within a competitive gut environment. Furthermore, our analysis revealed that the combination of bifidobacterial species and the nutrient source found in the localization analysis determined individual differences in microbiota response. These findings suggest that bacterial colonization facilitates the efficient, preferential, and presumably exclusive utilization of solid nutrient sources in the human gut. Moreover, the impact of a single nutrient source on the gut and human body may vary depending on the presence or absence of the primary species colonizing that source. In this review, we examined the micrometer-scale localization of intestinal bacteria and individual variability in microbiota responses to diet, drawing upon the results of our previous studies.
Visible-light inter-satellite communication is a promising physical-layer option for secure and interference-resilient 6G satellite networking. However, most analytical studies still assume Lambertian emission, which limits insight into emitters with asymmetric or multi-lobe radiation patterns. This paper presents a controlled analytical framework for Lambertian, Z-Power, and non-symmetric power-weighted (NSPW) beams using consistent transmitter-receiver modeling, channel-gain, receiver-noise, signal-to-noise ratio (SNR), and bit error rate (BER) formulations, including solar-background effects under Fraunhofer-line operation. The analysis considers six design dimensions: link distance, irradiance angle, transmitted optical power, receiver-bandwidth scaling, optical-filter background leakage, and beam azimuth rotation. The results show a clear operating-regime transition: Lambertian emission is competitive for near-aligned links, whereas non-Lambertian beams offer markedly higher robustness at wide irradiance angles. In a representative proximity-case stress point (0.5 km, [Formula: see text] irradiance angle), Z-Power and NSPW links achieve about 4.7 dB and [Formula: see text] dB, respectively, while the Lambertian baseline remains near [Formula: see text] dB, corresponding to gains of approximately 60 dB and 45.7 dB. The bandwidth, distance-scaling, and link-budget discussions clarify that these values are beam-profile sensitivity margins rather than a flight-qualified payload budget. Overall, the findings provide a practical roadmap for beam selection, link-margin interpretation, and attitude-aware adaptation in robust 6G visible-light inter-satellite communication systems.
Plyometric training (PT) is widely used in track and field to improve stretch-shortening cycle efficiency, sprint speed, and explosive power. However, no systematic review has synthesized randomized controlled trials (RCTs) examining the effects of PT on both injury-related and performance outcomes in track and field athletes. We aimed to evaluate the effects of PT on performance-related metrics in this population. PubMed®, Web of Science®, Scopus®, and SPORTDiscus® were systematically searched to identify RCTs involving track and field athletes of any competitive level. Eligible studies compared PT with participants continuing their usual routine training without the inclusion of plyometric exercises and reported performance-related outcomes. A meta-analysis was conducted using RevMan 5.4.1 software to evaluate the effects of PT on sports performance. The methodological quality was evaluated using the Tool for the Assessment of Study Quality and Reporting in Exercise (TESTEX) scale, while risk of bias was evaluated using the Cochrane RoB 2 tool. Thirty RCTs met inclusion criteria, with 27 contributing to meta-analyses. Out of 30 studies, one RCT assessed the impact of PT on injuries and reported a significant reduction in lower limb injury incidence following PT. Meta-analyses showed significant improvements in 30 m sprint performance (-3.53%, p = 0.02), countermovement jump height (mean difference 5.11%, p = 0.03), vertical jump height (2.95%, p = 0.02), and standing long jump distance (2.55%, p = 0.01). PT also improved VO₂ max (3.05%, p = 0.04) and running economy at 14 km·h⁻¹ (-1.96%, p = 0.05). Heterogeneity ranged from low to substantial across outcomes. PT enhances sprinting, jumping, neuromuscular, and selected endurance-related performance outcomes in track and field athletes. Although preliminary evidence suggests potential benefits including injury reduction, robust exposure-based RCTs are needed to establish definitive preventive effects.
Drug-target interaction (DTI) prediction is a crucial step in modern drug discovery. Accurate and efficient predictions can substantially reduce costs and development time. Applications of deep learning methods for this purpose have been extensively studied in recent years, yielding instrumental contributions to this field. However, existing methods face issues pertaining to efficient learning of drug and target feature representations, which is detrimental to generalisability and performance in cold-start scenarios. Most approaches extract representations from SMILES strings for drugs and FASTA sequences for target proteins, which encode limited 3D structural information. Additionally, many models lack explainability, being black boxes that provide little physical insight into the underlying mechanisms behind such interactions. We propose 3DICE, a novel framework leveraging co-attention-based fusion and massively pre-trained 3D structural encoders for both drugs and proteins. Uni-Mol and ESM-IF1 are employed to generate high-fidelity, 3D structure-aware embeddings which enable richer geometric and chemical understanding. Cross-modal fusion modules further augment representations to model intermolecular binding relationships. Importantly, this mechanism also provides intrinsic interpretability, highlighting and enabling qualitative analysis of most influential atoms or residues. Experiments conducted on two canonical benchmark datasets display the competitiveness of our model in real-world scenarios. 3DICE outperformed state-of-the-art models across multiple metrics on the DrugBank and KIBA datasets. Additional experiments provide a more rigorous analysis of interpretability than is typically reported in prior DTI studies, and we find that attention consistently highlights decision-critical regions which is not intrinsically class-specific. Our model and dataset are freely available at: https://github.com/austinatose/3DICE. Supplementary data are available at Bioinformatics online.
Diabetic Retinopathy (DR) is a medical condition in which high blood sugar levels damage the retina's blood vessels. Existing Solutions for multi-class DR identification are computationally intensive and also suffer from low accuracy. There is an immense need for an automated, computationally efficient approach for monitoring DR progression in diabetic patients. The study proposed Random Forest (RF), Logistic Regression (LR), Decision Tree (DT), and Gaussian Naive Bayes (GNB) models for the classification of retinal images into five DR classes (No, Mild, Moderate, Proliferate, and Severe) using spatial features extracted through a Convolutional Neural Network (CNN), textural features extracted through a Grey Level Co-occurrence Matrix (GLCM), and hybrid features by combining these features. The CNN, EfficientNet, PyramidCNN, and Pyramid Vision Transformer (PVT) were also evaluated for the classification of DR stages. The results revealed that the RF model with hybrid features outperformed, with an accuracy of 98.00% and high performance across all evaluation metrics, with a 1.00% increase over existing approaches. The EfficientNet model also performs competitively with 97.00% accuracy. The ML models also emerged as computationally efficient in terms of training and inference time for deployment in low-resource clinical environments for automated monitoring of DR progression in diabetic patients.
Aminophylline, a methylxanthine derivative, acts as a smooth muscle relaxant through non-competitive inhibition of phosphodiesterase. It has been suggested to facilitate ureteroscopic lithotripsy by relaxing ureteral smooth muscle. The aim of this systematic review is to assess the available evidence on locally administered aminophylline in patients undergoing ureteroscopic lithotripsy and, where feasible, to perform an exploratory meta-analysis to estimate pooled effects. A systematic search was conducted through November 2025. Three randomized controlled trials (RCTs) met the inclusion criteria. The primary outcome was residual stone proportions (derived from the stone-free rate). Secondary outcomes included operative time, postoperative ureteral stenting, and auxiliary procedures. The risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB-2) tool. This review was registered with PROSPERO (CRD420251229352). Three RCTs enrolling 310 patients (155 receiving aminophylline and 155 receiving placebo) were included. Local aminophylline was associated with significantly fewer residual stones (RR 0.3; 95% CI: 0.13, 0.69; p = 0.025; I² = 0%) and fewer auxiliary procedures (RR 0.15; 95% CI: 0.06, 0.38; p <0.001; I² = 0%). Reductions in operative time and postoperative ureteral stenting were observed in the primary analysis; however, these did not remain statistically significant after applying the Hartung-Knapp-Sidik-Jonkman (HKSJ) correction. This exploratory meta-analysis provides preliminary evidence that locally administered aminophylline may reduce residual stone fragments and auxiliary procedures during ureteroscopic lithotripsy. Reductions in operative time and ureteral stenting did not retain statistical significance in sensitivity analyses and should be considered hypothesis-generating. These findings should be interpreted with caution, given the low certainty of the evidence. Adequately powered randomized trials are required to validate these results. https://www.crd.york.ac.uk/prospero/, identifier CRD420251229352.
Stem cells, featuring remarkable self-renewal ability, unique multi-differentiation potential, distinguished paracrine effect and strong immunomodulatory function, have emerged as a powerful and competitive candidate for treating severe diseases and injuries. Direct injection of stem cells always suffers from limited cell retention in target tissues. Growing evidence suggests that nanofibrous microspheres with natural extracellular matrix-like topography, interconnected pores, high porosity, high specific surface area, and distinct injectability may serve as promising stem cell carriers that facilitate cell attachment, spreading, proliferation, retention and expression of specific genes. Over the past few decades, tremendous efforts have been devoted to developing nanofibrous microspheres with diverse composition, size, morphology and structure by using a variety of fabrication techniques. In this review, we provide an overview of recent progress in the development of nanofibrous microspheres with a focus on their preparation method, chemical composition, physicochemical property, and bioactivity, and highlight challenges and perspectives for future research directions.
AI-driven skin lesion diagnosis systems are revolutionizing dermatology practice but perform worse on darker skin populations, which threatens diagnostic equity in dermatology. Existing debiasing strategies rely on explicit skin tone annotations or adversarial removal of demographic information, which may be unavailable in practice and can harm diagnostic accuracy. We aimed to design a dermatology AI system that reduces skin-tone-related performance disparities across diverse skin populations without using skin tone labels during training. We propose a novel sketch-guided multimodal fusion framework that combines color (RGB) images with algorithmically generated structural sketches. Separate encoders extract representations from each modality, which are integrated by a gated fusion module that adaptively weighs color and structure features. A feature distillation loss aligns color features with their sketch counterparts to encourage structure-aware representations while retaining clinically relevant color cues. We trained and evaluated the model on the Fitzpatrick17k and Diverse Dermatology Images (DDI) datasets. The fairness performance was assessed with Equalized Opportunity, Equalized Odds, and Predictive Quality Disparity across skin tone groups. Out-of-domain robustness was examined using a DermaAmin and Atlas Dermatologico split. On Fitzpatrick17k, the proposed model yielded competitive accuracy and F1-score, while showing lower mean disparity in fairness evaluation than baseline methods. It also demonstrated reduced subgroup disparity across skin tone groups on the evaluated fairness metrics. In the out-of-domain evaluation setup, the model also exhibited improved fairness. On the DDI dataset, the framework showed consistent performance across different skin tone groups with reduced variation. Our proposed model shows promise in reducing skin-tone-related bias in dermatology while preserving utility without relying on explicit skin tone annotations. The observed improvements in skin tone fairness across two datasets suggest that our approach may help reduce measured subgroup disparities in automated skin lesion assessment, although clinical utility and real-world impact remain to be established through prospective validation.
Myrcene is a high-value monoterpene extensively applied in the fragrance, flavor, and agricultural industries, yet its efficient microbial production remains challenging due to pathway competition and limited metabolic flux. Compartmentalization offers a unique strategy to spatially organize heterologous metabolic pathways in Saccharomyces cerevisiae, enabling improved pathway efficiency through physical separation from competing cytosolic metabolism. In this study, we engineered the S. cerevisiae nucleus as a synthetic metabolic compartment for myrcene biosynthesis. Screening of myrcene synthases identified two highly active enzymes from Snapdragon Oc15 and Picea abies that function efficiently in S. cerevisiae. Myrcene production was detected only when myrcene synthase and the engineered GPP synthase mERG20p were co-localized to the nucleus, whereas cytosolic expression failed to yield detectable myrcene under their co-expression. Reconstruction of the complete mevalonate (MVA) pathway in the nucleus further increased myrcene titers. By identifying and optimizing rate-limiting steps, we substantially enhanced metabolic flux toward myrcene, achieving a final titer of 23.4 mg L-1 in flask-shaking fermentation. This work demonstrates the feasibility of repurposing the yeast nucleus for myrcene efficient biosynthesis and provide a new strategy for further improving microbial production of myrcene.