Scalable Label Distribution Learning for Multi-Label Classification.

Multi-label classification (MLC) refers to the problem of tagging a given instance with a set of relevant labels. Most existing MLC methods are based on the assumption that the correlation of two labels in each label pair is symmetric, which is violated in many real-world scenarios. Moreover, most existing methods design learning processes associated with the number of labels, which makes their computational complexity a bottleneck when scaling up to large-scale output space. To tackle these issues, we propose a novel method named scalable label distribution learning (SLDL) for MLC, which can describe different labels as distributions in a latent space, where the label correlation is asymmetric and the dimension is independent of the number of labels. Specifically, SLDL first converts labels into continuous distributions within a low-dimensional latent space and leverages the asymmetric metric to establish the correlation between different labels. Then, it learns the mapping from the feature space to the latent space, resulting in the computational complexity is no longer related to the number of labels. Finally, SLDL leverages a nearest neighbor-based strategy to decode the latent representations and obtain the final predictions. Extensive experiments illustrate that SLDL achieves very competitive classification performances with little computational consumption.

Grain boundary engineering for efficient and durable electrocatalysis.

Grain boundaries in noble metal catalysts have been identified as critical sites for enhancing catalytic activity in electrochemical reactions such as the oxygen reduction reaction. However, conventional methods to modify grain boundary density often alter particle size, shape, and morphology, obscuring the specific role of grain boundaries in catalytic performance. This study addresses these challenges by employing gold nanoparticle assemblies to control grain boundary density through the manipulation of nanoparticle collision frequency during synthesis. We demonstrate a direct correlation between increased grain boundary density and enhanced two-electron oxygen reduction reaction activity, achieving a significant improvement in both specific and mass activity. Additionally, the gold nanoparticle assemblies with high grain boundary density exhibit remarkable electrochemical stability, attributed to boron segregation at the grain boundaries, which prevents structural degradation. This work provides a promising strategy for optimizing the activity, selectivity, and stability of noble metal catalysts through precise grain boundary engineering.

Advances and challenges in neuroimaging-based pain biomarkers.

Identifying neural biomarkers of pain has long been a central theme in pain neuroscience. Here, we review the state-of-the-art candidates for neural biomarkers of acute and chronic pain. We classify these potential neural biomarkers into five categories based on the nature of their target variables, including neural biomarkers of (1) within-individual perception, (2) between-individual sensitivity, and (3) discriminability for acute pain, as well as (4) assessment and (5) prospective neural biomarkers for chronic pain. For each category, we provide a synthesized review of candidate biomarkers developed using neuroimaging techniques including functional magnetic resonance imaging (fMRI), structural magnetic resonance imaging (sMRI), and electroencephalography (EEG). We also discuss the conceptual and practical challenges in developing neural biomarkers of pain. Addressing these challenges, optimal biomarkers of pain can be developed to deepen our understanding of how the brain represents pain and ultimately help alleviate patients' suffering and improve their well-being.

The microenvironmental factors induced invasive tumor cells in glioblastoma.

Glioblastoma (GBM) cells have the potential to switch from being "proliferative cells" to peritumoral "invasive cells". Peritumoral GBM cells have highly invasive properties that allow them to survive surgery, leading to recurrence. The mechanisms underlying the manner in which the tumor microenvironment (TME) regulates the invasiveness of GBM remain unclear. Single-cell RNA sequencing analysis revealed heterogeneity in GBM cells, microglia and macrophages. In this study, the Oncostatin M receptor (OSMR) and leukemia inhibitory factor receptor (LIFR) expression indicated higher invasiveness in core GBM cells. Under environmental stress, the expression of OSMR and LIFR were up-regulated with the effect of hypoxic, acidic, and low-glucose conditions in vitro. Functional experiments revealed that TME stress significantly influences the proliferation, migration and invasion of GBM cells. The differences in core/peripheral TMEs in GBM affected the invasive properties, indicating the significant role of OSMR expression within the TME in tumor progression and postoperative therapy.

Association of glycemic variability and prognosis in patients with traumatic brain injury: A retrospective study from the MIMIC-IV database.

BACKGROUND: Elevated glycemic variability (GV) often occurs in intensive care unit (ICU) patients and is associated with patient prognosis. However, the association between GV and prognosis in ICU patients with traumatic brain injury (TBI) remains unclear. METHOD: Clinical data of ICU patients with TBI were obtained from the Medical Information Mart for Intensive Care (MIMIC) -IV database. The coefficient of variation (CV) was utilized to quantify GV, while the Glasgow Coma Scale (GCS) was employed to evaluate the consciousness status of TBI patients. Pearson linear correlation analysis, linear regression, COX regression and restricted cubic spline (RCS) were used to investigate the relationship between CV and consciousness impairment, as well as the risk of in-hospital mortality. RESULT: A total of 1641 ICU patients with TBI were included in the study from the MIMIC-IV database. Pearson linear correlation and restricted cubic spline (RCS) analysis results showed a negative linear relationship between CV and the last GCS (P = 0.002) with no evidence of nonlinearity (P for nonlinear = 0.733). Multivariable linear regression suggested a higher CV was associated with a lower discharge GCS [ß (95 %CI) = -1.86 (-3.08 âˆ¼ -0.65), P = 0.003]. Furthermore, multivariable COX regression indicated that CV ≥ 0.3 was a risk factor for in-hospital death in TBI patients [HR (95 %CI) = 1.74 (1.15-2.62), P = 0.003], and this result was also consistent across sensitivity and subgroup analyses. CONCLUSION: Higher GV is related to poorer consciousness outcomes and increased risk of in-hospital death in ICU patients with TBI. Additional research is needed to understand the logical relationship between GV and TBI progression.

Revealing the degradation pathways of layered Li-rich oxide cathodes.

Layered lithium-rich transition metal oxides are promising cathode candidates for high-energy-density lithium batteries due to the redox contributions from transition metal cations and oxygen anions. However, their practical application is hindered by gradual capacity fading and voltage decay. Although oxygen loss and phase transformation are recognized as primary factors, the structural deterioration, chemical rearrangement, kinetic and thermodynamic effects remain unclear. Here we integrate analysis of morphological, structural and oxidation state evolution from individual atoms to secondary particles. By performing nanoscale to microscale characterizations, distinct structural change pathways associated with intraparticle heterogeneous reactions are identified. The high level of oxygen defects formed throughout the particle by slow electrochemical activation triggers progressive phase transformation and the formation of nanovoids. Ultrafast lithium (de)intercalation leads to oxygen-distortion-dominated lattice displacement, transition metal ion dissolution and lithium site variation. These inhomogeneous and irreversible structural changes are responsible for the low initial Coulombic efficiency, and ongoing particle cracking and expansion in the subsequent cycles.

Characterization and function of promoters of silicon transporter genes PeLsi1-1 and PeLsi1-2 from moso bamboo (Phyllostachys edulis).

KEY MESSAGE: Promoters of moso bamboo silicon transporter genes PeLsi1-1 and PeLsi1-2 contain elements in response to hormone, silicon, and abiotic stresses, and can drive the expression of PeLsi1-1 and PeLsi1-2 in transgene Arabidopsis. Low silicon 1 (Lsi1) transporters from different species have been shown to play an important role in influxing silicon from soil. In previous study, we cloned PeLsi1-1 and PeLsi1-2 from Phyllostachys edulis and verified that PeLsi1-1 and PeLsi1-2 have silicon uptake ability. Furthermore, in this study, the promoters of PeLsi1-1(1910 bp) and PeLsi1-2(1922 bp) were cloned. Deletion analysis identified the key regions of the PeLsi1-1 and PeLsi1-2 promoters in response to hormone, silicon, and abiotic stresses. RT-qPCR analysis indicated that PeLsi1-1 and PeLsi1-2 were regulated by hormones, salt stress and osmotic stress. In addition, we found that the driving activity of the PeLsi1-1 and PeLsi1-2 promoters was regulated by 2 mM K2SiO3 and PeLsi1-1-P3 ~ P4 and PeLsi1-2-P4 ~ 5 were the regions regulated by silicon. Overexpression of PeLsi1-1 or PeLsi1-2 driven by 35S promoter in Arabidopsis resulted in a threefold increase of Si accumulation, whereas transgenic plants showed deleterious symptoms and dwarf seedlings and shorter roots under 2 mM Si treatment. When the 35S promoter was replaced by PeLsi1-1 or PeLsi1-2 promoter, a similar Si absorption was achieved and the transgene plants grew normally. This study, therefore, demonstrates that the promoters of PeLsi1-1 and PeLsi1-2 are indeed effective in driving the expression of moso bamboo Lsi1 genes and leading to silicon uptake.

Variational Label Enhancement for Instance-Dependent Partial Label Learning.

Partial label learning (PLL) is a form of weakly supervised learning, where each training example is linked to a set of candidate labels, among which only one label is correct. Most existing PLL approaches assume that the incorrect labels in each training example are randomly picked as the candidate labels. However, in practice, this assumption may not hold true, as the candidate labels are often instance-dependent. In this paper, we address the instance-dependent PLL problem and assume that each example is associated with a latent label distribution where the incorrect label with a high degree is more likely to be annotated as a candidate label. Motivated by this consideration, we propose two methods VALEN and MILEN, which train the predictive model via utilizing the latent label distributions recovered by the label enhancement process. Specifically, VALEN recovers the latent label distributions via inferring the variational posterior density parameterized by an inference model with the deduced evidence lower bound. MILEN recovers the latent label distribution by adopting the variational approximation to bound the mutual information among the latent label distribution, observed labels and augmented instances. Experiments on benchmark and real-world datasets validate the effectiveness of the proposed methods.

Concave Grain Boundaries Stabilized by Boron Segregation for Efficient and Durable Oxygen Reduction.

The oxygen reduction reaction (ORR) is a critical process that limits the efficiency of fuel cells and metal-air batteries due to its slow kinetics, even when catalyzed by platinum (Pt). To reduce Pt usage, enhancing both the specific activity and electrochemically active surface area (ECSA) of Pt catalysts is essential. Here, ultrafine, grain boundary (GB)-rich Pt nanoparticle assemblies are proposed as efficient ORR catalysts. These nanowires offer a large ECSA and a high density of concave GB sites, which improve specific activity. Atoms at these GB sites exhibit increased coordination and lattice distortion, leading to a favorable reduction in oxygen binding energy and enhanced ORR performance. Furthermore, boron segregation stabilizes these GBs, preserving active sites during catalysis. The resulting boron-stabilized Pt nanoassemblies demonstrate ORR specific and mass activities of 9.18 mA cm-2 and 6.40 A mg-1 Pt (at 0.9 V vs. RHE), surpassing commercial Pt/C catalysts by over 35-fold, with minimal degradation after 60 000 potential cycles. This approach offers a versatile platform for optimizing the catalytic performance of a wide range of nanoparticle systems.

Electroacupuncture regulates glucose metabolism by inhibiting SGLT1 levels, inhibiting microglial polarization, and alleviating Parkinson's disease.

BACKGROUND: Parkinson's disease (PD) is a common central neurodegenerative disease in middle-aged and elderly people. The progressive degeneration and death of dopaminergic neurons leads to insufficient dopamine (DA) neurotransmitters. Acupuncture and moxibustion can alleviate the aging of neurons. Therefore, studying the neuroprotective effects of electroacupuncture (EA) in PD mice is particularly important. METHODS: Intraperitoneal injection of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP, 20 mg/kg) was used to establish a PD mouse model, and lipopolysaccharide (LPS) was used to induce microglia polarization. Western blotting, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL), Nissl staining and immunohistochemistry were used to detect neuronal apoptosis and injury, α-syn expression and microglial accumulation in PD mice. In addition, the levels of inflammatory factors were determined using enzyme-linked immunosorbent assay (ELISA). Flow cytometry was used to detect the Ca2+ content. The fluorescein isothiocyanate (FITC) labeling method was used to assess glucose uptake. A reagent kit was used to detect glucose and lactate levels. RESULTS: MPTP induced the selective loss of DA neurons in the SN of mice, altered Ca2+ homeostasis, and induced an inflammatory response. In addition, maintaining Ca2+ homeostasis depends on the activity of transient receptor potential channel 1 (TRPC1). EA therapy promotes TRPC1 expression, which has a negative regulatory effect on sodium-glucose cotransporter 1 (SGLT1). Under the action of EA, TRPC1 protein expression increased, Ca2+ concentrations increased, and the effect of SGLT1 was inhibited, thereby facilitating glucose metabolism, blocking the activation of the phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) pathway, restraining M1 polarization of microglia, and alleviating the PD process. CONCLUSION: EA promotes TRPC1/Ca2+ pathway activation, inhibits SGLT1-mediated regulation of glucose metabolism and PI3K/AKT pathway activation, inhibits microglial M1 polarization, and alleviates PD.

Broad-spectrum degradation of fluoroquinolone antibiotics by Hemin-His-Fe nanozymes with peroxidase-like activity.

Fluoroquinolones are a widely used class of antibiotics, with a large variety, which are frequently monitored in the aqueous environment, threatening ecological and human health. To date, effective degradation of fluoroquinolone antibiotics remains a major challenge. Focused on the broad-spectrum degradation of fluoroquinolone antibiotics, a novel biomimetic peroxidase nanozyme named Hemin-His-Fe (HHF)-peroxidase nanozyme was synthesized through a green and rapid "one-pot" method involving hemin, Fmoc-L-His and Fe2+ as precursors. After systematic optimization of the reaction conditions, fluoroquinolone antibiotics can be degraded by the HHF-peroxidase nanozyme when supplemented with H2O2 in acidic environments. Through validation and analysis, it was proved that the generated strong oxidative hydroxyl radicals are the main active species in the degradation process. In addition, it was verified that this method shows great universal applicability in real water samples.

Characterization of cigarette smokeomics by in situ solid-phase microextraction and confined-space direct analysis in real time mass spectrometry.

Characterization of chemical composition in cigarette smoke is essential for establishing smoke-related exposure estimates. Currently used methods require complex sample preparation with limited capability for obtaining accurate chemical information. We have developed an in situ solid-phase microextraction (SPME) method for online processing of smoke aerosols and directly coupling the SPME probes with confined-space direct analysis in real time (cDART) ion source for high-resolution mass spectrometry (MS) analysis. In a confined space, the substances from SPME probes can be efficiently desorbed and ionized using the DART ion source, and the diffusion and evaporation of volatile species into the open air can be largely avoided. Using SPME-cDART-MS, mainstream smoke (MSS) and side-stream smoke (SSS) can be investigated and the whole analytical protocol can be accomplished in a few min. More than five hundred substances and several classes of compounds were detected and identified. The relative contents of 13 tobacco alkaloids were compared between MSS and SSS. Multivariate data analysis unveiled differences between different types of cigarette smoke and also discovered the characteristic ions. The method is reliable with good reproducibility and repeatability, and has the potential to be quantitative. This study provides a simple and high-efficiency method for smokeomics profiling of complex aerosol samples with in situ online extraction of volatile samples, and direct integration of extracted probes with a modified ambient ionization technique.

Label Distribution Learning by Exploiting Fuzzy Label Correlation.

Researchers have proposed to exploit label correlation to alleviate the exponential-size output space of label distribution learning (LDL). In particular, some have designed LDL methods to consider local label correlation. These methods roughly partition the training set into clusters and then exploit local label correlation on each one. Each sample belongs to one cluster and therefore has only one local label correlation. However, in real-world scenarios, the training samples may have fuzziness and belong to multiple clusters with blended local label correlations, which challenge these works. To solve this problem, we propose in LDL fuzzy label correlation (FLC)-each sample blends, with fuzzy membership, multiple local label correlations. First, we propose two types of FLCs, i.e., fuzzy membership-induced label correlation (FC) and joint fuzzy clustering and label correlation (FCC). Then, we put forward LDL-FC and LDL-FCC to exploit these two FLCs, respectively. Finally, we conduct extensive experiments to justify that LDL-FC and LDL-FCC statistically outperform state-of-the-art LDL methods.

The Causal Effect of Serum Lipid Levels Mediated by Neuregulin 4 on the Risk of Four Atherosclerosis Subtypes: Evidence from Mendelian Randomization Analysis.

Background: Neuregulin 4 (NRG4) was known to be associated with serum lipid levels and atherosclerosis. However, it is unknown whether the role of NRG4 in lipid homeostasis is causal to atherosclerosis and whether the effect is beneficial across different atherosclerosis subtypes. Methods: We investigated the causal role of the levels of serum low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol, and triglycerides regulated by NRG4 in subtypes of atherosclerosis through two sample Mendelian randomization. Aggregated genome-wide association study (GWAS) summary data for serum lipid level of 1.32 million individuals with European ancestry were obtained from the Global Lipids Genetics Consortium. GWAS summary data for four atherosclerosis subtypes (peripheral, coronary, cerebral and the other atherosclerosis) were obtained from FinnGen Consortium. Generalized inverse-variance-weighted Mendelian randomization and several sensitivity analyses were used to obtain the causal estimates. Results: A 1-SD genetically elevated LDL-C level mediated by NRG4 was validated to be nominally associated with the risk of peripheral atherosclerosis (log (odds ratio)= 4.14, 95% confidence interval 0.11 to 8.17, P = 0.04), and the other associations were not significant or could not be validated by sensitivity analyses. Conclusion: LDL-C lowering mediated by NRG4 is likely to prevent peripheral atherosclerosis.

Curcumin for gastric cancer: Mechanism prediction via network pharmacology, docking, and in vitro experiments.

BACKGROUND: Curcumin originates from the natural herb turmeric, and its antitumor effects have been known about for a long time. However, the mechanism by which curcumin affects gastric cancer (GC) has not been elucidated. AIM: To elucidate the potential mechanisms of curcumin in the treatment of GC. METHODS: Network pharmacological approaches were used to perform network analysis of Curcumin. We first analyzed Lipinski's Rule of Five for the use of Curcumin. Curcumin latent targets were predicted using the PharmMapper, SwissTargetPrediction and DrugBank network databases. GC disease targets were mined through the GeneCard, OMIM, DrugBank and TTD network databases. Then, GO enrichment, KEGG enrichment, protein-protein interaction (PPI), and overall survival analyses were performed. The results were further verified through molecular docking, differential expression analysis and cell experiments. RESULTS: We identified a total of 48 curcumin-related genes with 31 overlapping GC-related targets. The intersection targets between curcumin and GC have been enriched in 81 GO biological processes and 22 significant pathways. Following PPI analysis, 6 hub targets were identified, namely, estrogen receptor 1 (ESR1), epidermal growth factor receptor (EGFR), cytochrome P450 family 3 subfamily A member 4 (CYP3A4), mitogen-activated protein kinase 14 (MAPK14), cytochrome P450 family 1 subfamily A member 2 (CYP1A2), and cytochrome p450 family 2 subfamily B member 6 (CYP2B6). These factors are correlated with decreased survival rates among patients diagnosed with GC. Molecular docking analysis further substantiated the strong binding interactions between Curcumin and the hub target genes. The experimental findings demonstrated that curcumin not only effectively inhibits the growth of BGC-823 cells but also suppresses their proliferation. mRNA levels of hub targets CYP3A4, MAPK14, CYP1A2, and CYP2B6 in BGC-823 cells were significantly increased in each dose group. CONCLUSION: Curcumin can play an anti-GC role through a variety of targets, pathways and biological processes.

A gap-free genome assembly of Fusarium oxysporum f. sp. conglutinans, a vascular wilt pathogen.

Fusarium oxysporum is an asexual filamentous fungus that causes vascular wilt in hundreds of crop plants and poses a threat to public health through Fusariosis. F. oxysporum f. sp. conglutinans strain Fo5176, originally isolated from Brassica oleracea, is pathogenic to Arabidopsis, making it a model pathosystem for dissecting the molecular mechanisms underlying host-pathogen interactions. Assembling the F. oxysporum genome is notoriously challenging due to the presence of repeat-rich accessory chromosomes. Here, we report a gap-free genome assembly of Fo5176 using PacBio HiFi and Hi-C data. The 69.56 Mb assembly contained 18 complete chromosomes, including all centromeres and most telomeres (20/36), representing the first gap-free genome sequence of a pathogenic F. oxysporum strain. In total, 21,460 protein-coding genes were annotated, a 26.3% increase compared to the most recent assembly. This high-quality reference genome for F. oxysporum f. sp. conglutinans Fo5176 provides a valuable resource for further research into fungal pathobiology and evolution.

Addressing Skewed Heterogeneity via Federated Prototype Rectification With Personalization.

Federated learning (FL) is an efficient framework designed to facilitate collaborative model training across multiple distributed devices while preserving user data privacy. A significant challenge of FL is data-level heterogeneity, i.e., skewed or long-tailed distribution of private data. Although various methods have been proposed to address this challenge, most of them assume that the underlying global data are uniformly distributed across all clients. This article investigates data-level heterogeneity FL with a brief review and redefines a more practical and challenging setting called skewed heterogeneous FL (SHFL). Accordingly, we propose a novel federated prototype rectification with personalization (FedPRP) which consists of two parts: federated personalization and federated prototype rectification. The former aims to construct balanced decision boundaries between dominant and minority classes based on private data, while the latter exploits both interclass discrimination and intraclass consistency to rectify empirical prototypes. Experiments on three popular benchmarks show that the proposed approach outperforms current state-of-the-art methods and achieves balanced performance in both personalization and generalization.

Impact of volume indices in bioelectrical impedance measurement on the assessment of cardiac function indices by echocardiography in hemodialysis patients.

INTRODUCTION: Cardiovascular events resulting from volume overload are a primary cause of mortality in hemodialysis patients. Bioelectrical impedance analysis (BIA) is significantly valuable for assessing the volume status of hemodialysis (HD) patients. In this article, we explore the correlation between the volume index measured by BIA and the cardiac function index assessed by echocardiography (ECG) in HD patients. METHODS: Between April and November 2018, we conducted a cross-sectional study involving randomly selected 126 maintenance HD patients. Comprehensive data on medical history and laboratory test results were collected. Subsequently, we investigated the correlation between volume indices measured by BIA and cardiac function parameters by ECG. RESULTS: We discovered a significant correlation between the volume indices measured by BIA and various parameter of cardiac function. The Left Ventricular Hypertrophy (LVH) group exhibited higher levels of the percentage of Extracellular Water (ECW%) and the percentage of Total Body Water (TBW%) compared to the Non-LVH group. Extracellular Water (ECW) and Third Interstitial Fluid Volume (TSFV) were identified as independent risk factors for Left Ventricular Mass (LVM), and both demonstrated a high predictive value for LVM. ECW% emerged as an independent risk factor for the Left Ventricular Mass Index (LVMI), with a high predictive value for LVMI. CONCLUSION: ECW and TSFV were found to be positively associated with cardiac function parameters in HD patients.