Identification and validation of key biomarkers for the early diagnosis of diabetic kidney disease

Background: Diabetic kidney disease (DKD) is the leading cause of end-stage renal disease. This study explored the core genes and pathways associated with DKD to identify potential diagnostic and therapeutic targets. Methods: We downloaded microarray datasets GSE96804 and GSE104948 from the Gene Expression Omnibus (GEO) database. The dataset includes a total of 53 DKD samples and 41 normal samples. Differentially expressed genes (DEGs) were identified using the R package "limma". The Metascape database was subjected to Gene Ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses to identify the pathway and functional annotations of DEGs. A WGCAN network was constructed, the hub genes in the turquoise module were screened, and the core genes were selected using LASSO regression to construct a diagnostic model that was then validated in an independent dataset. The core genes were verified by in vitro and in vivo experiments. Results: A total of 430 DEGs were identified in the GSE96804 dataset, including 285 upregulated and 145 downregulated DEGs. WGCNA screened out 128 modeled candidate gene sets. A total of eight genes characteristic of DKD were identified by LASSO regression to build a prediction model. The results showed accuracies of 99.15% in the training set (GSE96804) and 94.44% and 100%, respectively, in the test (GSE104948-GPL22945 and GSE104948-GPL24120). Three core genes (OAS1, SECTM1, and SNW1) with high connectivity were selected among the modeled genes. In vitro and in vivo experiments confirmed the upregulation of these genes. Conclusion: Bioinformatics analysis combined with experimental validation identified three novel DKD-specific genes. These findings may advance our understanding of the molecular basis of DKD and provide potential therapeutic targets for its clinical management.