Abstract
PURPOSE: This study aimed to identify genes that may be effective in diagnosing or treating diabetic retinopathy (DR), the most common complication of diabetes mellitus (DM). METHODS: Differentially expressed genes (DEGs) were identified between DR and DM in GSE146615 dataset. DEGs that were consistently up- or down-regulated under both standard glucose and high glucose conditions were identified as common genes and used to generate a protein-protein interaction network and modules. The module genes were assessed for the area under the receiver operating characteristic curve (AUC), leading to the identification of hub genes. Differentially methylated probes in GSE76169 were also compared with common DEGs to identify specific methylation markers of DR. Enrichment analysis was used to explore the biological characteristics. The Short Time-series Expression Miner algorithm was used to identify genes that were progressively dysregulated in the sequence: healthy controls < DM < DR. RESULTS: A total of 1917 common genes were identified for seven modules. The eight genes with AUC > 0.8 under high glucose and standard glucose conditions were considered as hub genes. The module genes were significantly enriched during vascular smooth muscle cell development and regulation of oxygen metabolism, while 92 methylation markers were involved in the similar terms. Among the progressively dysregulated genes, three intersection genes under both standard glucose and high glucose conditions were found to be module genes and were considered as key genes. CONCLUSION: We identified eight potential DR-specific diagnostic and therapeutic genes, whose abnormal expression can cause oxidative stress, thus favoring the course of the disease.