Frontiers in Immunology (Nov 2021)

Identification of Methylation Immune Subtypes and Establishment of a Prognostic Signature for Gliomas Using Immune-Related Genes

  • Zhengang Hu,
  • Hao Zhang,
  • Fan Fan,
  • Zeyu Wang,
  • Jiahao Xu,
  • Yunying Huang,
  • Ziyu Dai,
  • Hui Cao,
  • Xun Zhang,
  • Zhixiong Liu,
  • Zhixiong Liu,
  • Quan Cheng,
  • Quan Cheng,
  • Quan Cheng

DOI
https://doi.org/10.3389/fimmu.2021.737650
Journal volume & issue
Vol. 12

Abstract

Read online

DNA methylation patterns are essential in understanding carcinogenesis. However, the relationship between DNA methylation and the immune process has not been clearly established—this study aimed at elucidating the interaction between glioma and DNA methylation, consolidating glioma classification and prognosis. A total of 2,483 immune-related genes and 24,556 corresponding immune-related methylation probes were identified. From the Cancer Genome Atlas (TCGA) glioma cohort, a total of 683 methylation samples were stratified into two different clusters using unsupervised clustering, and eight types of other cancer samples from the TCGA database were shown to exhibit excellent distributions. A total of 3,562 differentially methylated probes (DMPs) were selected and used for machine learning. A five-probe signature was established to evaluate the prognosis of glioma as well as the potential benefits of radiotherapy and Procarbazine, CCNU, Vincristine (PCV) treatment. Other prognostic clinical models, such as nomogram and decision tree, were also evaluated. Our findings confirmed the interactions between immune-related methylation patterns and glioma. This novel approach for cancer molecular characterization and prognosis should be validated in further studies.

Keywords