Retinoblastoma gene expression profiling based on bioinformatics analysis

Abstract

Abstract Background Retinoblastoma (RB) is frequently occurring malignant tumors that originate in the retina, and their exact cause and development mechanisms are yet to be fully comprehended. In this study, we identified possible biomarkers for RB and delved into the molecular mechanics linked with such markers. Methods In this study GSE110811 and GSE24673 were analyzed. Weighted gene co-expression network analysis (WGCNA) was applied to screen modules and genes associated with RB. By overlapping RB-related module genes with differentially expressed genes (DEGs) between RB and control samples, differentially expressed retinoblastoma genes (DERBGs) were acquired. A gene ontology (GO) enrichment analysis and a kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis were conducted to explore the functions of these DERBGs. To study the protein interactions of DERBGs, a protein–protein interaction (PPI) network was constructed. Hub DERBGs were screened using the least absolute shrinkage and selection operator (LASSO) regression analysis, as well as the random forest (RF) algorithm. Additionally, the diagnostic performance of RF and LASSO methods was evaluated using receiver operating characteristic (ROC) curves and single-gene gene set enrichment analysis (GSEA) was conducted to explore the potential molecular mechanisms involved with these Hub DERBGs. In addition, the competing endogenous RNA (ceRNA) regulatory network of Hub DERBGs was constructed. Result About 133 DERBGs were found to be associated with RB. GO and KEGG enrichment analyses revealed that the important pathways of these DERBGs. Furthermore, the PPI network revealed 82 DERBGs interacting with each other. By RF and LASSO methods, PDE8B, ESRRB, and SPRY2 were identified as Hub DERBGs in patients with RB. From the expression assessment of Hub DERBGs, it was found that the levels of expression of PDE8B, ESRRB, and SPRY2 were significantly decreased in the tissues of RB tumors. Secondly, single-gene GSEA revealed a connection between these 3 Hub DERBGs and oocyte meiosis, cell cycle, and spliceosome. Finally, the ceRNA regulatory network revealed that hsa-miR-342-3p, hsa-miR-146b-5p, hsa-miR-665, and hsa-miR-188-5p may play a central role in the disease. Conclusion Hub DERBGs may provide new insight into RB diagnosis and treatment based on the understanding of disease pathogenesis.

Keywords

• Retinoblastoma
• Bioinformatics analysis
• Biomarkers
• ceRNA
• Regulatory network