TMEM104 is identified as a hub gene and potential therapeutic target for metabolic dysfunction-associated steatotic liver disease

Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as nonalcoholic fatty liver disease (NAFLD), is a common cause of chronic liver disease. This study identified key MASLD-associated hub genes and explored their biological roles, cellular context, and potential for pharmacological modulation. We obtained the patients’ transcriptome and scRNA-seq data from the Gene Expression Omnibus (GEO) database. Subsequently, weighted gene co-expression network analysis (WGCNA) was used to identify gene modules related to disease progression. Differentially expressed genes (DEGs) were screened by the DESeq2 package. LASSO regression and SVM-RFE were applied for feature selection and validation in an independent cohort. Enriched biological pathways were identified through GSEA and candidate small molecules were predicted via Comparative Toxicogenomics Database and evaluated by molecular docking. Single-cell transcriptomic analysis of MASLD was performed with the Seurat package to construct an atlas and examine the cell-type-specific distribution of the key genes. Finally, the expression of the selected key genes in vitro cell lines was quantitatively analyzed by real-time quantitative polymerase chain reaction (qRT-PCR). Through WGCNA, we identified 323 hub genes with the strongest association with MASLD. Intersecting these hub genes with the 1037 DEGs yielded 46 candidate genes. LASSO and SVM-RFE converged on two shared feature genes, TIMM44 and TMEM104, and only TMEM104 remained significantly upregulated in the validation cohort. GSEA revealed that high-expressed TMEM104 was related to reactive oxygen species pathways, including PI3K–Akt–mTOR, TNF–NFκB activation and ADRB3–UCP1 metabolic pathways. Molecular docking analysis indicated favorable binding of TMEM104 to indomethacin and methyltestosterone. Single-cell analysis showed enriched TMEM104 expression in hepatocytes and neuroendocrine cells. Consistent with this, RT-qPCR in an in vitro NAFLD-like steatosis model further confirmed its upregulation under lipid overload conditions. This study identified a previously unrecognized hub gene, TMEM104, implicated in oxidative stress, inflammatory, and metabolic pathways in driving MASLD. The results highlighted TMEM104 as a promising biomarker and a candidate for therapeutic development, which warrants further functional validation.