Hepatocellular carcinoma (HCC) is a leading cause of cancer-related mortality worldwide. The in-depth study of genes and metabolites related to nucleotide metabolism will provide new ideas for predicting the prognosis of HCC patients. This study integrated the transcriptome data of different cancer types to explore the characteristics and significance of nucleotide metabolism-related genes (NMGRs) in different cancer types. Then, we constructed a new HCC classifier and prognosis model based on HCC samples from TCGA and GEO, and detected the gene expression level in the model through molecular biology experiments. Finally, nucleotide metabolism-related products in serum of HCC patients were examined using untargeted metabolomics. A total of 97 NMRGs were obtained based on bioinformatics techniques. In addition, a clinical model that could accurately predict the prognostic outcome of HCC was constructed, which contained 11 NMRGs. The results of PCR experiments showed that the expression levels of these genes were basically consistent with the predicted trends. Meanwhile, the results of untargeted metabolomics also proved that there was a significant nucleotide metabolism disorder in the development of HCC. Our results provide a promising insight into nucleotide metabolism in HCC, as well as a tailored prognostic and chemotherapy sensitivity prediction tool for patients.

The discrepancy between serum triglyceride levels and the clinical severity of hyperlipidemic acute pancreatitis (HLAP) complicates risk stratification. Traditional lipidomics, which primarily rely on linear abundance, often fail to distinguish the HLAP-specific lipidome from the metabolic background of hypertriglyceridemia (HTG). To overcome this limitation, DeepLipiDecipher was developed as a knowledge-guided graph neural network framework that integrates lipid chemical structures with metabolic topology to identify latent lipotoxic features. In a retrospective cohort of 433 subjects, DeepLipiDecipher demonstrated robust classification performance (AUC = 0.810), effectively distinguishing the HLAP phenotype and outperforming conventional machine learning models. Interpretability analysis revealed that HLAP susceptibility correlated with a distinct structural lipid profile, marked by the synergistic enrichment of polyunsaturated and ether-linked phospholipids, rather than total lipid mass. Moreover, computational causal inference implicated a pathogenic mechanism wherein SMPD3-mediated ceramide accumulation induced basal cytotoxicity, and PTGS2 hyperactivation promoted the peroxidation of these vulnerable lipids, triggering systemic inflammation. These results highlight the value of incorporating network topology into lipidomic analysis and suggest the network-inferred SMPD3-Ceramide-PTGS2 immunometabolic axis as a potential therapeutic target for preventing the progression from metabolic dysfunction to acute organ injury.

Background/Objectives: Gallstone disease is a prevalent digestive disorder worldwide, with incidence increasingly linked to obesity, insulin resistance, and dyslipidemia. Emerging evidence suggests that intestinal microbial communities and their bioactive compounds play a central role in pathogenesis. Here, we aimed to clarify whether diet-related microbial alterations and circulating metabolites contribute to gallstone formation. Methods: We integrated dietary inflammatory index (DII) evaluation, genetic analyses of large-scale cohorts, and a cholesterol gallstone mouse model induced by a lithogenic diet (LD). Serum and fecal samples were subjected to metabolomic and metagenomic profiling, followed by multi-omics integration to identify links between microbial taxa, metabolites, and gallstone risk. Results: Higher DII scores were associated with increased gallstone risk. Genetic evidence supported bile acid and amino acid metabolism as potential mediating pathways, with Akkermansia muciniphila linked to decreased N-acetylarginine levels and CAG-448 showing an inverse association with glycodeoxycholate (GDCA). In LD-fed mice, shotgun metagenomics revealed enrichment of lithogenic taxa such as Bacteroides stercorirosoris and Enterocloster, whereas protective taxa, including Akkermansia muciniphila and CAG-448, were markedly depleted. Untargeted metabolomics confirmed elevations of GDCA and N-acetylarginine together with broader bile acid imbalance, amino acid stress, and long-chain acylcarnitine accumulation. Correlation analyses further showed that protective taxa were inversely associated with risk metabolites, whereas gallstone-enriched microbes displayed the opposite pattern. Conclusions: This study provides evidence consistent with a contributory role of gut microbiota–metabolite dysregulation in gallstone pathogenesis. Specific taxa (A. muciniphila, CAG-448) and metabolites (GDCA, N-acetylarginine) may serve as potential biomarkers or targets for microbiota- and diet-based prevention strategies.

The present study investigated the distinct metabolic profiles of two primary traditional Chinese medicine (TCM) syndromes in patients with cholelithiasis—liver Qi stagnation (LQS) and liver-gallbladder damp-heat (LGDH). By integrating untargeted and targeted metabolomics, we aimed to identify objective serum and urine biomarkers for syndrome differentiation and to elucidate their underlying biological basis. In total, 91 participants, comprising 33 healthy controls (HCs), 30 cholelithiasis patients with LQS syndrome, and 28 cholelithiasis patients with LGDH syndrome, were enrolled. Untargeted analysis of serum and urine samples quantified 602 and 636 metabolites, respectively. To strengthen the robustness of the dataset, complementary targeted metabolomics was conducted as an orthogonal validation step, providing reliable quantification for 210 serum and 204 urinary. The present findings provide molecular insights into the TCM principle of syndrome differentiation (Bian Zheng Lun Zhi). The identified biomarker panel may offer a basis for developing objective tools for the stratification of cholelithiasis syndromes and may support future efforts towards personalized therapeutic strategies.

Background: Clinical lipidomics can capture disease-associated molecular alterations at high resolution, yet translating complex lipid species data into interpretable biological insight remains challenging. Existing workflows often emphasize statistical discrimination while underutilizing the structural information embedded in lipid species. To address this gap, we developed LipiDecipher, a structure-oriented analytical framework designed to summarize lipidomic alterations into interpretable structural patterns and to provide database-supported biological contextualization. Methods: LipiDecipher integrates differential lipid analysis, structure-resolved summarization, multivariate discrimination, and knowledge-based lipid-to-protein/pathway contextualization. We applied this framework to a retrospective serum lipidomics dataset comprising healthy controls and patients with acute myocardial infarction or post-PCI recurrent myocardial infarction. To improve transparency and robustness, the revised analysis includes sex-disaggregated reporting, covariate-adjusted sensitivity analyses for sex and age, and internal separation stability assessment of category-specific LDA projections through resampling-based feature stability analysis, repeated cross-validation, and permutation testing. Results: The framework identified distinct lipid alterations across study groups, including changes in phosphatidylinositols, ceramides, and triglyceride remodeling patterns. These alterations became more interpretable when summarized at the structural level, including lipid class composition, acyl-chain length, and degree of unsaturation. Internal discrimination analyses suggested separability between groups, while repeated resampling highlighted a subset of recurrently selected lipid features. Knowledge-based mapping prioritized lipid-associated biological contexts related to glycerophospholipid metabolism, sphingolipid metabolism, membrane remodeling, inflammatory signaling, and energy-related processes. Importantly, these protein- and pathway-level outputs are presented as database-supported hypotheses rather than direct evidence of target engagement or pathway activation in the studied cohort. Conclusions: LipiDecipher provides a structure-oriented and interpretation-focused framework for clinical lipidomics. In a retrospective acute myocardial infarction cohort, it enabled the prioritization of candidate lipid signatures and biologically plausible hypotheses from complex lipidomic data. These findings support its use as a hypothesis-generating analytical tool, while external validation and experimental follow-up remain necessary before mechanistic or clinical claims can be established.