Early predictors of 6-month non-recovery in drug-induced liver injury (DILI) remain limited. Genetic variants are stable host characteristics that may complement baseline clinical variables. We aimed to develop and validate an interpretable, clinical-genetic machine learning model for predicting 6-month non-recovery in patients with DILI.

This retrospective, single-center study included 338 patients with DILI, who were classified as recovered (n = 171) or non-recovered (n = 167) at 6 months. Candidate single-nucleotide polymorphisms and baseline clinical variables were collected during initial hospitalization. Features were selected using complementary screening approaches. Multiple machine learning models were developed and compared. Model discrimination, calibration, clinical utility, the incremental value of genetic predictors, and interpretability using SHapley Additive exPlanations (SHAP) were assessed.

Five predictors were consistently retained for model development: rs72631567, rs28521457, alanine aminotransferase, monocyte percentage, and low-density lipoprotein. Among the candidate algorithms, the light gradient boosting machine model showed the best performance, with area under the receiver operating characteristic curve (AUC) values of 0.92 (95% confidence interval [CI] 0.89–0.95) in the training set and 0.81 (95% CI 0.70–0.91) in the validation set. The model showed acceptable calibration and favorable decision-curve performance. In ablation analysis, the clinical-only model showed limited discrimination (AUC 0.57, 95% CI 0.43–0.71). SHAP analysis identified rs72631567 as the most influential predictor.

An interpretable model that integrates host genetic variants with baseline clinical variables demonstrated good internal performance for early prediction of 6-month non-recovery in DILI. These findings support external validation of genotype-informed risk stratification in patients with DILI.

Esophageal adenocarcinoma (EAC) and gastric cancer (GC) remain major causes of cancer-related mortality worldwide, largely because of late-stage diagnosis. Barrett’s esophagus (BE) is the established precursor setting for EAC, whereas gastric intestinal metaplasia is an important precursor lesion in the intestinal-type gastric carcinogenesis pathway. Aberrant DNA methylation arises early in both pathways and may offer greater sensitivity and specificity than histology or serology alone, while also illuminating the biology of neoplastic progression. This review evaluates the evidence and translational potential of DNA methylation biomarkers in upper gastrointestinal adenocarcinoma.

PubMed, Web of Science, and EMBASE were searched for studies published before May 2026 addressing DNA methylation biomarkers in BE, EAC, gastric intestinal metaplasia, and GC. Tissue-based, minimally invasive, and circulating biomarkers were reviewed, with emphasis on diagnostic performance, progression risk, biological significance, and clinical translation.

Methylation alterations accumulate during progression from precancerous lesions to invasive cancer and distinguish disease states in tissue and minimally invasive specimens. Methylation-based assays show promising diagnostic performance in BE and EAC, including nonendoscopic cytology-based approaches and risk-stratification assays, and circulating methylated DNA shows potential for noninvasive detection of EAC and GC. However, heterogeneity in populations, specimens, assay platforms, and study designs, together with limited prospective validation, remains a barrier to implementation.

The field is moving rapidly from discovery toward practice: Cytosponge-TFF3 with biomarker panels has been evaluated in real-world UK surveillance pathways, the EsoCheck/EsoGuard methylation assay is clinically available, and Esopredict is the first epigenetic prognostic assay clinically validated to risk-stratify BE, while multicancer early detection assays undergo prospective evaluation. Multicenter studies of clinical utility and cost-effectiveness remain the decisive step before routine adoption.

Atrial fibrillation and chronic kidney disease (CKD) frequently coexist, increasing thromboembolic and bleeding risks. This is a narrative review of pathophysiology and clinical evidence for anticoagulation strategies in CKD patients. Direct oral anticoagulants are preferred in CKD stages 1–4. Recent data suggest that the efficacy of apixaban and rivaroxaban is comparable to that of warfarin in end-stage renal disease. In advanced CKD, anticoagulation should be tailored with close monitoring.

Head, thoracic, abdominal, and pelvic polytrauma is clinically complex and requires coordinated multidisciplinary care, creating an urgent need for standardized guidance. This guideline aims to provide an evidence-based foundation and clinical pathway for the diagnosis, assessment, damage control, surgical timing, and acute-phase management of patients with head, thoracic, abdominal, and pelvic polytrauma. Led by the Senior Department of Neurosurgery, Chinese PLA General Hospital, a multidisciplinary working group comprising experts in thoracic surgery, general surgery, orthopedics, critical care medicine, anesthesiology, radiology, emergency medicine, nursing, and related disciplines developed the guideline through systematic review of the relevant literature, consideration of the characteristics of trauma care in China’s military and civilian medical systems, and two rounds of expert voting and consensus decision-making. The guideline addresses 22 key clinical questions covering imaging assessment, damage-control resuscitation, prioritization of craniocerebral and torso injury management, anticoagulation and hemostatic management, intensive care, and early rehabilitation. A total of 22 recommendations were formulated by integrating evidence quality, clinical feasibility, and implementability, together with corresponding implementation points and quality-control indicators. The guideline emphasizes stabilization of vital signs, damage-control principles, multidisciplinary collaboration, standardized clinical pathways, and a tiered trauma-care system. Implementation of these recommendations may improve the consistency and timeliness of multidisciplinary care and provide a framework for reducing early mortality while preserving neurological and functional outcomes.

Celiac disease remains underdiagnosed despite an established diagnostic pathway, reflecting phenotypic heterogeneity, delayed recognition, and dependence on resource-intensive testing. This narrative review synthesizes evidence on artificial intelligence and phenomics in celiac disease (CD) across four operational layers: electronic health record phenotyping, Human Phenotype Ontology-based semantic encoding, machine-learning pre-screening from routine clinical data, and deep-learning-assisted histopathology. A targeted literature search of PubMed/MEDLINE, Embase, and Google Scholar covered publications from January 2010 through December 2025, using combinations of CD/coeliac disease with artificial intelligence, machine learning, deep learning, phenomics, Human Phenotype Ontology, computable phenotype, electronic health records, and natural language processing, supplemented by targeted searches and citation chaining. We present a curated minimum viable CD phenome and discuss clinical actionability, age-specific considerations, and current pediatric validation gaps. Selected models have demonstrated promising CD detection or pre-screening performance in specific datasets, but prospective, multicenter evidence with external validation remains insufficient to establish clinical utility. Artificial intelligence should augment expert clinical care rather than replace it.

Dysregulated lipid metabolism contributes to hepatocellular carcinoma (HCC) progression, but the prognostic value and mechanistic roles of lipid metabolism-related long noncoding RNAs (LRLs) remain insufficiently characterized. This study aimed to construct and validate an LRL-based prognostic model and to investigate the biological function and metabolic mechanism of AC026412.3 in HCC.

Transcriptomic and clinical data from the The Cancer Genome Atlas Liver Hepatocellular Carcinoma cohort were analyzed to identify LRLs based on their correlation with curated lipid metabolism genes. Differential expression, univariate Cox, least absolute shrinkage and selection operator (LASSO), and multivariate Cox analyses were performed to construct a prognostic signature, which was evaluated using Kaplan–Meier survival and time-dependent receiver operating characteristic (ROC) analyses. Functional enrichment analyses Gene Ontology [GO], Kyoto Encyclopedia of Genes and Genomes [KEGG] and gene set enrichment analysis [GSEA], mutation profiling, tumor mutational burden, immune infiltration estimation, and consensus clustering were applied to characterize associated features. A key LRL was identified through integrated bioinformatic screening and prioritization. Its biological role was assessed by quantitative reverse transcription polymerase chain reactionq (RT-PCR), western blotting, BODIPY staining, colony formation, Transwell assays, and xenograft models. RNA sequencing followed by pathway enrichment analysis was conducted to explore underlying mechanisms.

A three-LRL signature (AL031985.3, NRAV, and AC026412.3) stratified HCC patients into distinct risk groups with significantly different survival outcomes and demonstrated independent prognostic value. AC026412.3 was markedly upregulated in HCC and associated with poor prognosis. Functional assays demonstrated that AC026412.3 promoted proliferation, invasion, and tumor growth while reducing lipid accumulation. Mechanistically, AC026412.3 upregulated solute carrier family 22 member 5 (SLC22A5), enhanced fatty acid β-oxidation, and increased adenosine triphosphate (ATP) production, thereby driving metabolic reprogramming.

This study establishes a robust LRL-based prognostic model and identifies AC026412.3 as a key regulator of lipid metabolic reprogramming via the SLC22A5–fatty acid β-oxidation axis, highlighting its potential as a biomarker and therapeutic target in HCC.