Identification of lipid metabolism-related signature in nonalcoholic fatty liver: evidences from transcriptomics and single cell RNA-sequencing analysis.

Considering the complex and close-knit relationship between non-alcoholic fatty liver disease (NAFLD) and the metabolic status, this study aimed to identify lipid metabolism-related genes (LMGs), construct an effective diagnostic model, and uncover the underlying LMG-related mechanism in NAFLD. NAFLD datasets and the LMGs were downloaded from the GEO database and a previous publication. Key LMGs were identified through differential expression analysis, Venn diagrams, WGCNA, and three machine learning algorithms; and then nomogram models were constructed. To evaluate the model's performance, we employed calibration curves, decision curve analysis (DCA), clinical impact curve (CIC), and receiver operator characteristic (ROC) analysis. Additionally, we constructed immune infiltration analysis and scRNA-seq analysis to explore the immune infiltration patterns and underlying mechanisms. Finally, qRT-PCR and Western blotting were performed to validate the expression of the key genes. The critical genes (PRKAA2 and ME1) were identified as significantly correlated with lipid metabolism in NAFLD. The nomogram model, based on the two key genes, demonstrated strong diagnostic performance for NAFLD, achieving an AUC of 0.945. The immune infiltration analysis revealed that PRKAA2 and ME1 were significantly associated with immune cells, particularly NK cells and T cells. scRNA-seq analysis classified cells into six subtypes: NK cells, monocytes and macrophages, T cells, B cells, proliferating cells, and plasmacytoid dendritic cells. Among these, T cells and NK cells were the most functionally relevant, with ME1 showing predominant expression and a strong correlation with NK cells. Experimental validation via qRT-PCR and Western blotting confirmed that only ME1 was significantly up-regulated in NAFLD. A lipid metabolism-related signature was successfully constructed, offering clinical potential for predicting NAFLD status.