Identification of Differentially Expressed Key Genes Between Glioblastoma and Low-Grade Glioma by Bioinformatics Analysis.

Yang Xu,Rongxin Geng,Fan'en Yuan,Qian Sun,Baohui Liu,Qianxue Chen

DOI: https://doi.org/10.7717/peerj.6560

2019-01-01

Abstract:Gliomas are a very diverse group of brain tumors that are most commonly primary tumor and difficult to cure in central nervous system. It's necessary to distinguish low-grade tumors from high-grade tumors by understanding the molecular basis of different grades of glioma, which is an important step in defining new biomarkers and therapeutic strategies. We have chosen the gene expression profile GSE52009 from gene expression omnibus (GEO) database to detect important differential genes. GSE52009 contains 120 samples, including 60 WHO II samples and 24 WHO IV samples that were selected in our analysis. We used the GEO2R tool to pick out differently expressed genes (DEGs) between low-grade glioma and high-grade glioma, and then we used the database for annotation, visualization and integrated discovery to perform gene ontology analysis and Kyoto encyclopedia of gene and genome pathway analysis. Furthermore, we used the Cytoscape search tool for the retrieval of interacting genes with molecular complex detection plug-in applied to achieve the visualization of protein-protein interaction (PPI). We selected 15 hub genes with higher degrees of connectivity, including tissue inhibitors metalloproteinases-1 and serum amyloid A1; additionally, we used GSE53733 containing 70 glioblastoma samples to conduct Gene Set Enrichment Analysis. In conclusion, our bioinformatics analysis showed that DEGs and hub genes may be defined as new biomarkers for diagnosis and for guiding the therapeutic strategies of glioblastoma.