| Literature DB >> 33594424 |
Jianyang Du1,2, Hang Ji1,2, Shuai Ma1, Jiaqi Jin1,2, Shan Mi1,2, Kuiyuan Hou1,2, Jiawei Dong1, Fang Wang1,2, Chaochao Zhang3, Yuan Li4, Shaoshan Hu1.
Abstract
m6A RNA methylation is an emerging epigenetic modification, and its potential role in immunity and stemness remains unknown. Based on 17 widely recognized m6A regulators, the m6A modification patterns and corresponding characteristics of immune infiltration and stemness of 1152 low-grade glioma samples were comprehensively analyzed. Machine-learning strategies for constructing m6AScores were trained to quantify the m6A modification patterns of individual samples. Here, we reveal a significant correlation between the multi-omics data of regulators and clinicopathological parameters. We identified two distinct m6A modification patterns (an immune-activated differentiation pattern and an immune-desert dedifferentiation pattern) and four regulatory patterns of m6A methylation on immunity and stemness. We show that the m6AScores can predict the molecular subtype of low-grade glioma, the abundance of immune infiltration, the enrichment of signaling pathways, gene variation and prognosis. The concentration of high immunogenicity and clinical benefits in the low-m6AScore group confirmed the sensitive response to radio-chemotherapy and immunotherapy in patients with high-m6AScore. The results of the pan-cancer analyses illustrate the significant correlation between m6AScore and clinical outcome, the burden of neoepitope, immune infiltration and stemness. The assessment of individual tumor m6A modification patterns will guide us in improving treatment strategies and developing objective diagnostic tools.Entities:
Keywords: immunotherapy; m6A; machine learning; pan-cancer; stemness; tumor immune microenvironment
Year: 2021 PMID: 33594424 DOI: 10.1093/bib/bbab013
Source DB: PubMed Journal: Brief Bioinform ISSN: 1467-5463 Impact factor: 11.622