Bioinformatics identification of key genes associated with risk and prognosis of neuroblastoma

Abstract Objective Neuroblastoma is a childhood malignancy with high morbidity and mortality. Through bioinformatics analysis, we identified key biomarkers associated with neuroblastoma risk and prognosis. Methods The gene modules most associated with neuroblastoma risk were derived by weighted gene co-expression network analysis (WGCNA). Modular genes were intersected with differentially expressed genes (DEGs) between high-risk neuroblastoma (HR-NB) patients and non-high-risk neuroblastoma (NHR-NB) patients to obtain risk genes, and enrichment analysis was performed. After incorporating risk genes into Cox regression analysis, Least Absolute Shrinkage and Selector Operation (LASSO) algorithm, and Kaplan-Meier survival analysis, key genes were identified and introduced into three external datasets for validation. In addition, we performed short time-series expression miner (STEM) analysis and single sample genome enrichment analysis (ssGSEA). Finally, we evaluated the difference in DNA methylation levels to identify meaningful methylation marks. Results The 26 risk genes we screened were mainly associated with ion channels and ion transport across membranes. We identified 5 key genes (ANO6, CPNE2, DST, PLXNC1, SCN3A) for neuroblastoma risk and prognosis, which correlated closely with known neuroblastoma biomarkers. All key genes showed a progressive down-regulation trend with increasing risk levels of neuroblastoma. The level of immune infiltration of 14 immune cells was significantly different between HR-NB and NHR-NB, and most immune cells were negatively correlated with key genes. Furthermore, the expression of ANO6, CPNE2, DST and PLXNC1 was modified by DNA methylation. Conclusion This study identified 5 key genes for neuroblastoma risk and prognosis that were potential biomarkers and potential therapeutic targets for neuroblastoma.