Integrating multiple-level molecular data to infer the distinctions between glioblastoma and lower-grade glioma

Abstract

Glioblastomas (GBMs) and lower-grade gliomas (LGGs) are the most common malignant brain tumors. Despite extensive studies that have suggested that there are differences between the two in terms of clinical profile and treatment, their distinctions on a molecular level had not been systematically analyzed. Here, we investigated the distinctions between GBM and LGG based on multidimensional data, including somatic mutations, somatic copy number variants (SCNVs), gene expression, lncRNA expression and DNA methylation levels. We found that GBM patients had a higher mutation frequency and SCNVs than LGG patients. Differential mRNAs and lncRNAs between GBM and LGG were identified and a differential mRNA-lncRNA network was constructed and analyzed. We also discovered some differential DNA methylation sites could distinguish between GBM and LGG samples. Finally, we identified some key GBM- and LGG-specific genes featuring multiple-level molecular alterations. These specific genes participate in diverse functions; moreover, GBM-specific genes are enriched in the glioma pathway. Overall, our studies explored the distinctions between GMB and LGG using a comprehensive genomics approach that may provide novel insights into studying the mechanism and treatment of brain tumors.

Keywords: DNA methylation sites; glioblastoma; lower-grade glioma; multiple-level molecular data; somatic mutations.