RESUMO
AIMS: Growing evidence suggests that microRNAs (miRNAs) are crucial in controlling how diabetic retinopathy (DR) develops. We intend to mine miRNAs with diagnostic and predictive value for DR and to investigate new drug therapeutic targets. METHODS: After performing a differential analysis on the miRNA and mRNA datasets for DR and neovascularization (NEO), miRNA-mRNA networks were created. Combine the results of enrichment analysis, Protein-Protein Interaction Networks (PPI), and Cytoscape to identify key miRNAs. DrugBank was used to find drugs that interacted with transcription factors (TF) predicted by TransmiR. Finally, whole blood and clinical data were collected from 58 patients with type 2 diabetes mellitus (T2DM), and RT-qPCR, logistic analysis, and ROC were used to verify the value of key miRNAs. RESULTS: Differential analysis indicated the presence of genes and miRNAs that co-regulate DR and NEO. Enrichment analysis showed that key genes are inextricably linked to neovascularization. Combining the results of PPI and Cytoscape identified four key miRNAs, namely hsa-mir-(4328, 4422, 548z and -628-5p). RT-qPCR, logistic, and ROC results showed that decreased expression levels of hsa-mir-(4328, 4422, 548z and -628-5p) signal the risk of evolution to DR in T2DM patients. Finally, we constructed a TF-miRNA network to find the 15 TFs and the 35 drugs that interact with these TFs. CONCLUSION: hsa-mir-(4328, 4422, 548z and -628-5p) in whole blood are protective factors for DR as novel biomarkers for diagnosis and prediction. In addition, our research provides new drug directions for the treatment of DR, such as Diosmin, Atorvastatin, and so on.