Feed

Diabetic nephropathy (DN) is the most common complication of diabetes, with immune-mediated inflammation playing a significant role in its pathophysiology. The complement system, a key proinflammatory factor, is implicated in DN. This study was aimed at identifying diagnostic biomarkers related to the complement system in DN using bioinformatics methods. We analyzed three datasets (GSE96804, GSE104948, and GSE1009) from public databases, employing differential expression analysis and machine learning to identify complement system-related genes (CSRGs) as potential biomarkers. A diagnostic nomogram was constructed based on these genes, and immune microenvironment differences between the DN and control groups were explored. Gene interaction networks, enrichment analysis, and drug predictions were also conducted. Mendelian randomization (MR) was used to examine the causal links between identified biomarkers and DN. Our analysis identified five biomarkers (CKB, ANXA1, HSPA1L, CYP27B1, and XYLT1) associated with DN. A diagnostic nomogram based on these biomarkers showed high accuracy (model correction slope close to 1 and AUC close to 1). Immune infiltration analysis revealed significant differences in immune cell subsets between the DN and control groups. Gene set variation analysis (GSVA) indicated that the oxidative phosphorylation (OXPHOS) pathway was activated in CKB, HSPA1L, CYP27B1, and XYLT1 while inhibited in ANXA1. MR confirmed HSPA1L as a risk factor for DN (OR = 1.625, 95% CI: 1.272-2.076, p = 9.96e - 05). In conclusion, these five CSRGs may play significant roles in DN progression, providing a foundation for further research into DN pathogenesis and potential molecular markers for clinical diagnosis.