Biomarker prediction for membranous nephropathy prognosis by microarray analysis

ABSTRACT Aim This study aimed to explore biomarkers for membranous nephropathy (MN) diagnosis and to provide novel insights into the pathogenesis of this disease. Methods A microarray data set, GSE73953, was used, which contained 15 immunoglobulin A nephropathy (IgAN) samples, 8 MN samples and 2 healthy controls. Pretreatments were performed for the raw data, and then, the differentially expressed genes (DEG) were screened out using the limma package. The function and pathways of these genes were demonstrated through enrichment analysis. Protein–protein interaction (PPI) network analysis was performed to uncover interactions of DEG from the protein level. MN‐related genes were further selected, integrating the Comparative Toxicogenomics Database (CTD). Results In total, 446 and 231 DEG were identified in the comparisons of MN versus control and MN versus IgAN, respectively. JUN , NFKB1 , TGFB1 and PPBP were the predominant DEG, and the latter two were especially differentially expressed between the MN and IgAN groups. UBL4A and EIF4G1 were the two most important DEG for MN because they were downregulated compared with both control and IgAN groups. The above‐mentioned genes were highlighted in the PPI networks and mainly enriched the ribosome‐ and platelet‐related function/pathways. Conclusion Several potential biomarkers were identified in MN, and some of them could well distinguish the MN from IgAN. Disruption of ribosome‐ and platelet‐related function or pathways might contribute to MN progression. EIF4F and UBL4A might be two novel biomarkers for MN prognosis. Nevertheless, more experimental validations are needed.