Interferons can trigger robust anti-tumor responses following immune checkpoint blockade (ICB). However, chronic interferon signaling can also reshape tumor cell phenotypes, selecting for immune evasion mechanisms that drive resistance. In a cohort of 108 metastatic melanoma samples collected prior to first-line ICB, immune infiltration and bulk interferon gamma (IFNG) expression correlate with initial but not durable responses. Using a regression-based approach to deconvolve cancer cell signals from bulk tumors, we find that tumors from patients who initially respond but later acquire resistance exhibit heightened cancer cell-dependent IFN signaling at baseline, accompanied by MYC downregulation, dedifferentiation, and impaired major histocompatibility complex (MHC) class II induction. Genetically, acquired resistance also associates with reduced T cell interferon signaling, lower memory T cell activity, high tumor mutational burden, clonal diversification, and elevated subclonal neoantigen burden. These findings reveal baseline transcriptomic and genetic features distinguishing acquired resistance from durable ICB response, identifying candidate targets to prevent relapse and refine stratification.

Steatotic liver disease (SLD) affects one-third of the global population, yet current non-invasive diagnostic methods are too costly or operator-dependent for population-scale screening. Here, we present 3D-FAICE, a deep learning system that uses three-dimensional facial imaging for non-invasive SLD detection. Trained and tested on 11,456 participants, the facial model achieves robust performance across internal, external, and self-controlled longitudinal cohorts and remains effective in a smartphone-based point-of-care setting. Metabolomic analysis reveals that facial risk scores correlate with glycolipid and amino acid pathways, supporting biological plausibility. Multimodal fusion of facial and metabolomic data further improves accuracy, and a cross-modal distillation strategy significantly elevates the performance of the facial-only model. These findings establish facial image-based AI as a non-invasive, scalable, and privacy-aware tool for SLD screening, with potential applications in self-monitoring and population health management.

This study aimed to develop and validate a computed tomography (CT)-based model that combines radiomics features, deep features derived from nnU-Net, and clinical imaging variables for predicting the presence of calcaneal spurs. In this retrospective study, 803 calcaneal CT images were included and randomly divided into a training cohort (n = 642) and a test cohort (n = 161). The calcaneus was segmented with an iterative semiautomatic no new U-Net (nnU-Net) workflow, and conventional radiomics features and nnU-Net deep features were extracted from the reviewed regions of interest. Stable and discriminative features were selected using the intraclass correlation coefficient, between-group testing, and least absolute shrinkage and selection operator regression. Support vector machine models with ten-fold cross-validation and grid search were used to generate the radiomics score (Rad-score) and nnU-Net score. Logistic regression was used to develop the clinical and combined models. Model performance was assessed with receiver operating characteristic curves (ROC), calibration analysis, decision curve analysis, and interpretability analysis. Age and Gissane's angle were retained in the clinical model. The combined model included the Rad-score, nnU-Net score, and Gissane's angle. In the training cohort, the combined model achieved an area under the ROC curve of 0.916 (95% confidence interval: 0.894-0.937), exceeding the nnU-Net, radiomics, and clinical models. In the test cohort, the combined model also showed the highest area under the ROC curve (0.772; 95% confidence interval: 0.700-0.843), compared with the nnU-Net model (0.705), radiomics model (0.710), and clinical model (0.691). Calibration analysis showed acceptable agreement between predicted and observed probabilities, with Hosmer-Lemeshow P values of .218 and .463 in the training and test cohorts, respectively. Decision curve analysis suggested a higher net benefit for the combined model across clinically relevant threshold ranges. A CT-based combined model integrating the Rad-score, nnU-Net score, and Gissane's angle improved the prediction of calcaneal spur presence compared with clinical or imaging models alone. This model may provide a quantitative imaging tool for risk stratification of calcaneal spurs in patients undergoing foot or ankle CT.

Osteoporosis (OP) occurs with decreased bone mass and deterioration of the bone microarchitecture. This can lead to an increased incidence of bone fractures. In traditional Chinese medicine, the herbal pair of Eucommia ulmoides and Achyranthes bidentata is used to tonify the liver and kidney in order to strengthen bones. This study examined the network pharmacology and molecular docking of the pair to investigate the pharmacological mechanisms of the drug pair against OP. Active compounds were retrieved from the Traditional Chinese Medicine Systems Pharmacology database if they had an oral bioavailability of 30% or greater and a drug-likeness score of 0.18 or greater. OP-associated targets were identified from the GeneCards and Online Mendelian Inheritance in Man databases. The compound-target-disease network and protein-protein interaction network were built using Cytoscape and the Search Tool for the Retrieval of Interacting Genes/Proteins, respectively. Key pathways were identified using the Database for Annotation, Visualization and Integrated Discovery functional enrichment analysis. To predict the potential binding interactions between the major compounds and the hub targets, molecular docking was used. In total, 49 active compounds (28 from E ulmoides and 21 from A bidentata) and 95 overlapping therapeutic targets were found. Quercetin, kaempferol, β-sitosterol, stigmasterol and wogonin were identified as key bioactive components within the network analysis. Fourteen hub targets, including interleukin-6, vascular endothelial growth factor A, tumor necrosis factor, tumor protein p53, and mitogen-activated protein kinase 1, were identified through protein-protein interaction analysis. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment revealed involvement in tumor necrosis factor, mitogen-activated protein kinase, hypoxia-inducible factor 1, Toll-like receptor, and NOD-like receptor signaling pathways. Molecular docking demonstrated favorable binding affinities between the key compounds and core proteins, suggesting potential interactions consistent with the network pharmacology predictions. The Eucommia-Achyranthes herbal pair may exert potential therapeutic effects on OP through multicomponent, multi-target, and multi-pathway interactions. Its main bioactive compounds may potentially interact with inflammatory and osteogenic signaling pathways via key targets such as interleukin-6, vascular endothelial growth factor A, tumor necrosis factor-alpha, tumor protein p53, and mitogen-activated protein kinase 1, which may be associated with bone metabolism and structural integrity. These findings provide a theoretical foundation for future experimental and clinical validation.

Postoperative acute kidney injury (PO-AKI) following noncardiac surgery remains a major clinical challenge, for which effective early warning models are still lacking. This study aimed to identify risk factors for noncardiac PO-AKI and to develop and validate a practical early warning model based on readily available clinical parameters across multiple surgical subspecialties. In this study, large-scale data from noncardiac surgical patients were extracted from the electronic medical record system of Xuanwu Hospital to analyze risk factors. Subsequently, an early warning model for PO-AKI was developed by integrating logistic regression (LR) and deep learning (DL) approaches. Internal and external validations were performed to assess model performance. A total of 19,152 patients were included in the analysis. Eleven independent risk variables were identified using LR, including age, surgical department, prolonged operation duration, contrast agent exposure, elevated serum creatinine, proteinuria, hemoglobin reduction, elevated fasting blood glucose, serum potassium abnormalities, neutrophil-to-lymphocyte ratio, and D-dimer levels. A corresponding risk score was subsequently constructed. The receiver operating characteristic analysis demonstrated areas under the curve (AUCs) of 0.7679 for the LR model and 0.7578 for the derived risk score. In the validation cohort, the AUCs were 0.7856 and 0.7122, respectively. The DL model further improved predictive performance, achieving AUCs of 0.97 and 0.93 in the validation and test sets, respectively. Precision-recall analysis showed area under the precision-recall curve values of 0.96 and 0.74, with corresponding F1 scores of 0.945 and 0.897. Big data analysis revealed an overall incidence of PO-AKI after noncardiac surgery of 3.69%. A robust early warning model incorporating 11 clinically practical parameters was developed. The integration of LR and DL approaches improved predictive performance while maintaining interpretability.

The precise association between the atherogenic index of plasma (AIP) and its derived anthropometric indicators and the future risk of liver diseases and alcoholic liver disease among Chinese middle-aged and elderly adults remains unclear. The study population was derived from the China Health and Retirement Longitudinal Study. Incident liver disease cases were identified based on self-reported questionnaire information. The AIP and its related indicators were calculated using laboratory results from participants blood samples. We employed Cox regression to assess the associations between cumulative and categorical AIP, as well as related indicators, and the occurrence of liver disease. Additionally, restricted cubic splines were used to examine potential nonlinear relationships between cumulative AIP and liver disease risk. The mean age of the 9027 participants was 61.20 ± 9.54 years, and 438 (4.85%) subjects developed liver disease during a median follow-up duration of 7 years. After full adjustment for confounders, each 1-standard deviation increase in AIP was associated with a 15% higher risk of liver disease (HR = 1.15; 95% confidence intervals: 1.02-1.30; P = .020). Similarly, higher levels of AIP-BMI, AIP-WC, and AIP-BRI were associated with increased liver disease risk. restricted cubic splines analyses visually confirmed a positive dose-response relationship between AIP and its related indicators and liver disease risk, with no significant nonlinear trends. Additionally, subgroup analyses and sensitivity analyses confirmed that higher levels of AIP are associated with a higher risk of liver disease. In this national cohort, higher AIP was associated with an increased risk of self-reported liver disease. Long-term monitoring of AIP may help identify individuals for further hepatic assessment. However, confounding from unmeasured factors including viral hepatitis cannot be ruled out.

Observational studies link psychiatric disorders to Alzheimer disease (AD), but whether these associations are causal remains unclear due to confounding and reverse causality. We aimed to dissect these relationships using genetic evidence. We performed a two-sample Mendelian randomization (MR) study to assess the causal effects of 7 psychiatric and neurodevelopmental disorders on AD and its subtypes. The primary analysis utilized the inverse-variance weighted (IVW) method, supported by comprehensive sensitivity analyses (MR Steiger test, MR-Egger, and MR-PRESSO) and a supplementary analysis excluding single nucleotide polymorphisms (SNPs) associated with potential confounders. Depression showed a significant causal association with late-onset AD (odds ratio [OR] = 1.0736, 95% confidence interval [CI]: 1.0084-1.1431, P = .0264). No significant causal associations were found for other psychiatric disorders (all P > .05). A directionality test conducted by MR Steiger confirmed our estimation of potential causal direction (P < .001). Sensitivity analyses excluding pleiotropic SNPs yielded consistent results. The causal association for depression and late-onset AD remained significant after excluding pleiotropic SNPs linked to confounders (OR = 1.0726, 95% CI: 1.0069-1.1425, P = .03). Our study provides genetic evidence supporting a causal role for depression in the etiology of late-onset AD, a link not observed for other major psychiatric disorders tested. These findings highlight the specific importance of managing depression as a potential strategy for mitigating AD risk and suggest distinct etiological pathways between different mental health conditions and neurodegeneration.

Neuromyelitis optica spectrum disorder (NMOSD) is a rare multifocal inflammatory disease that primarily affects the optic nerve and spinal cord, leading to visual loss and paralysis. Although current research has identified numerous biomarker changes in the cerebrospinal fluid (CSF), the causal relationships between these changes and NMOSD are not yet fully understood, and they are susceptible to confounding factors. In this study, Mendelian randomization (MR) analyses were conducted to estimate the role of CSF metabolites in aquaporin-4 (AQP4)-IgG positive and AQP4-IgG negative NMOSD. We strictly adhered to the 3 principles of MR when selecting instrumental variables, effectively reducing the impact of confounding variables. Additionally, we conducted a comprehensive and meticulous sensitivity analysis to ensure the robustness of positive outcomes. We have elucidated the roles of several CSF metabolites in NMOSD, with specific glycerophosphocholine species showing risk associations in AQP4-IgG positive and AQP4-IgG negative subtypes, while ascorbate and N-acetylhexosamines demonstrate protective effects across both subtypes, thereby providing valuable insights. No evidence of heterogeneity or horizontal pleiotropy was observed in MR analyses. These metabolites hold the potential to inform pathogenesis and guide the development of therapeutic strategies for these conditions.

This study investigates the potential mechanisms through which the Rhubarb-Peach Kernel herb pair may exert effects in gastric cancer (GC) using network pharmacology and molecular docking approaches. Active compounds and their corresponding targets were retrieved from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP), with additional compound-target information obtained from DrugBank. Gastric cancer-related targets were collected from GeneCards, Online Mendelian Inheritance in Man (OMIM), the Therapeutic Target Database, and DrugBank. The overlapping targets were used to construct a compound-target-disease network and a protein-protein interaction network. Gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses were subsequently performed. Nine candidate compounds were docked against 8 core targets to assess their predicted structural interactions. A total of 27 active compounds were identified, among which β-sitosterol, aloe-emodin, and hederagenin were identified as major candidate compounds. Tumor protein p53, caspase 3, and JUN were among the principal network targets. Enrichment analysis indicated that the overlapping targets were mainly associated with the p53 signaling pathway and other pathways involved in cancer-related processes, including apoptosis, cell survival, and inflammatory regulation. Molecular docking revealed differences in predicted binding scores among the 9 candidate compounds and 8 core targets, with several compound-target pairs showing comparatively lower docking energies. This in silico study suggests that the Rhubarb-Peach Kernel herb pair may act through multiple compounds, targets, and biological pathways relevant to GC. However, the identified targets are not necessarily specific to GC, and the molecular docking findings represent computational predictions rather than evidence of actual biological binding. These results should therefore be considered preliminary and hypothesis-generating, and further in vitro and in vivo studies are required to evaluate their biological and therapeutic relevance.

Observational studies have shown that air pollution may affect functional outcome following ischemic stroke. Our study aimed to explore the relationship between genetically predicted specific air pollutants and poststroke functional outcomes using the Mendelian randomization (MR) methodology. The instrumental variables for air pollution (nitrogen oxide [NOx], nitrogen dioxide [NO2], PM2.5, PM2.5-10, and PM10) were derived from whole-genome association study data of individuals of European descent. Summary data for functional outcomes after ischemic stroke were sourced from the Ischemic Stroke Genetics Functional Outcome Network. Primary analysis employed the "IVW" method for 2-sample MR. In sensitivity analysis, Cochran Q test, inverse variance-weighted (IVW) method, and MR-Egger method were used for heterogeneity assessment. In the horizontal pleiotropy analysis, we utilized the MR-Egger regression and MR pleiotropy residual sum and outlier (MR-PRESSO) test. Finally, the robustness of the results is ensured. Gene prediction suggested no association between several air pollutants and functional outcome after ischemic stroke (PM2.5 [odds ratio (OR) 0.87; 95% confidence interval (CI) 0.26-2.95; P = .825], NOx [OR, 0.77; 95% CI: 0.27-2.21; P = .629], NO2 [OR 1.73; 95% CI: 0.61-4.90; P = .304], PM2.5-10 [OR 2.48; 95% CI: 0.28-22.12; P = .416], and PM10 [OR 2.54; 95% CI: 0.41-15.87; P = .319]). The supplementary MR methods confirmed similar results. Further analysis, not adjusted for stroke severity, supports these findings. Air pollutants (NOx, NO2, PM2.5, PM2.5-10 and PM10) may not have an impact on functional outcome after ischemic stroke.