ABSTRACT
AIM: To prevent neovascularization in diabetic retinopathy (DR) patients and partially control disease progression. METHODS: Hypoxia-related differentially expressed genes (DEGs) were identified from the GSE60436 and GSE102485 datasets, followed by gene ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Potential candidate drugs were screened using the CMap database. Subsequently, a protein-protein interaction (PPI) network was constructed to identify hypoxia-related hub genes. A nomogram was generated using the rms R package, and the correlation of hub genes was analyzed using the Hmisc R package. The clinical significance of hub genes was validated by comparing their expression levels between disease and normal groups and constructing receiver operating characteristic curve (ROC) curves. Finally, a hypoxia-related miRNA-transcription factor (TF)-Hub gene network was constructed using the NetworkAnalyst online tool. RESULTS: Totally 48 hypoxia-related DEGs and screened 10 potential candidate drugs with interaction relationships to upregulated hypoxia-related genes were identified, such as ruxolitinib, meprylcaine, and deferiprone. In addition, 8 hub genes were also identified: glycogen phosphorylase muscle associated (PYGM), glyceraldehyde-3-phosphate dehydrogenase spermatogenic (GAPDHS), enolase 3 (ENO3), aldolase fructose-bisphosphate C (ALDOC), phosphoglucomutase 2 (PGM2), enolase 2 (ENO2), phosphoglycerate mutase 2 (PGAM2), and 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3). Based on hub gene predictions, the miRNA-TF-Hub gene network revealed complex interactions between 163 miRNAs, 77 TFs, and hub genes. The results of ROC showed that the except for GAPDHS, the area under curve (AUC) values of the other 7 hub genes were greater than 0.758, indicating their favorable diagnostic performance. CONCLUSION: PYGM, GAPDHS, ENO3, ALDOC, PGM2, ENO2, PGAM2, and PFKFB3 are hub genes in DR, and hypoxia-related hub genes exhibited favorable diagnostic performance.