Immune Infiltrating Cells-Derived Risk Signature Based on Large-scale Analysis Defines Immune Landscape and Predicts Immunotherapy Responses in Glioma Tumor Microenvironment.

ABSTRACT: The glioma tumor microenvironment (TME), composed of several noncancerous cells and biomolecules is known for its complexity of cancer-immune system interaction. Given that, novel risk signature is required for predicting glioma patient responses to immunotherapy. In this study, we systematically evaluated the TME infiltration pattern of 2877 glioma samples. TME phenotypes were determined using the Partitioning Around Medoid method. Machine learning including SVM-RFE and Principal component analysis (PCA) were used to construct a TME scoring system. A total of 857 glioma samples from four datasets were used for external validation of the TME-score. The correlation of TME phenotypes and TME-scores with diverse clinicopathologic characteristics, genomic features, and immunotherapeutic effica

SUBMITTER: Zhang N 

PROVIDER: S-EPMC8418124 | biostudies-literature | 2021

REPOSITORIES: biostudies-literature