Abstract
BACKGROUND: The aim of this study was to find novel diagnostic markers for renal fibrosis based on bioinformatics analysis. METHODS: In this study, we first identified differentially expressed genes and modular genes associated with renal fibrosis. We utilized the GSE76882 dataset (81 renal fibrosis patients and 91 control samples) to identify these genes and then intersected them with the DEGs in the GSE120495 dataset (5 renal fibrosis patients and 5 control samples). To gain insights into the biological processes involved, we performed GO and KEGG analyses on the intersection genes and constructed protein interaction networks. Next, we employed the LASSO regression algorithm and the random forest algorithm to further screen the candidate biomarkers. The sensitivity and specificity of the candidate biomarkers were assessed using ROC curves. Finally, we induced renal fibrosis using the Unilateral Ureteral Obstruction (UUO) model (characterized by renal tubular damage and fibrosis in response to obstructed urinary flow) to observe the impact of the candidate biomarkers on the progression of renal fibrosis. RESULTS: A total of 383 differential genes and 689 renal fibrosis-related module genes were identified from the GSE76882 dataset, and 213 genes were obtained by intersecting the two. A total of 2431 differentially expressed genes were identified from the GSE120495 dataset, and 34 genes were obtained by intersecting them with the results of the previous step. Five candidate biomarkers (SFN, ARHGAP9, VSIG4, ISG20, CD3G) were identified by LASSO regression analysis and random forest algorithm. The comprehensive AUC values of the five genes in both datasets were greater than 0.7. Finally, in vivo experiments were performed to verify that VSIG4 may promote renal function injury induced by UUO. CONCLUSIONS: VSIG4 is highly expressed in UUO and may be an important mediator involved in renal fibrosis.