Exploring ENPP5 as a diagnostic biomarker for sepsis: a comprehensive bioinformatics analysis

作者：Jiamin Gao, Yan Jun Li, Jingping Huang, Cailing Wei, Jieling Chen, Aichun Huang, Na Liu, Yibo Lu, Shixiong Yang · 发表于：BMC Infectious Diseases · 年份：2025 · DOI：10.1186/s12879-025-11152-6 · 研究领域：Sepsis Diagnosis and Treatment、Ferroptosis and cancer prognosis、Cancer-related molecular mechanisms research

BACKGROUND: The rising mortality rates in sepsis highlight the current lack of reliable therapeutic biomarkers. This study aims to identify markers associated with biological functions to offer new strategies for sepsis diagnosis. METHODS: We conducted differential expression analysis to identify differentially expressed messenger RNAs (DEmRs), long non-coding RNA (DElncRs), and microRNAs (DEmiRs) in sepsis compared to healthy controls. Enrichment analysis was performed using DEmRs, and a lncRNA-miRNA-mRNA competing endogenous RNA network was constructed. Least absolute shrinkage and selection operator (LASSO) and random forest models were applied to identify diagnostic mRNAs. The optimal diagnostic model was determined through decision curve analysis, resulting in the identification of seven hub genes. The key gene, determined by its highest importance and the largest area under the receiver operating characteristics curve (AUC) value, was further validated. Additionally, we analyzed the correlation of the key gene with microenvironment cell infiltration and immune genes. RESULTS: A total of 4,450 intersected DEmRs (GSE66099, GSE13904, GSE154918, GSE8121) that were significantly involved in the cell cycle. We obtained 13 mRNAs, and further screened seven hub genes, including PPARD, ZSCAN2, ABI2, ENPP5, FMNL3, CD3E, and CAMK4. Subsequently, ENPP5 was as the key gene based on importance and AUC value. Moreover, Neutrophil cells and macrophages had a high abundance in sepsis pa...