Abstract
To discover novel biological targets in glioblastoma, genomic and immunological analysis were performed using The Cancer Genome Atlas (TCGA) data set. The RNA-seq data of 156 primary glioblastoma cases were subjected to CIBERSORT to detect tumor infiltrating cell fractions. Principal component analysis was performed on this data to detect factors that strongly contribute to the first principal component, and hierarchical clustering was performed. Survival curves were compared for each of the derived clusters. Finally, Gene Set Enrichment Analysis (GSEA) using HALLMARK Gene Set was performed. In the principal component analysis, we detected seven factors (NK cells resting, T cell regulatory, NK cells activated, Macrophage type 0, T cell gamma delta, Macrophage type 2, Macrophage type 1) which strongly contribute to the first principal component. Based on these seven factors, hierarchical cluster analysis resulted in T cell regulatory (Treg), Macrophage type 0 (M0), Macrophage type 2 (M2) and Macrophage type 1 (M1) clusters. There was no significant difference between these groups in CD8 T cell. M2 and M1 clusters displayed better OS with a significant difference. TNFA signaling via NFκB in Treg group, IFNα response, IFNγ response and ALLOGRAFT response in M2 group, G2M CHECKPOINT, GLYCOLYSIS, WNTβ catenin signaling, MITOTIC SPINDLE and TGFβ signaling in M1 group were upregulated. In conclusion, tumor microenvironment of glioblastoma can be divided into 4 immunological subtypes, Treg, M0, M1, and M2. Because of the contribution of innate immunity for shaping the tumor microenvironment of glioblastoma, immunotherapies targeting these innate immune cells are anticipated.
Parsimonious Description of Glucose Variability in Type 2 Diabetes by Sparse Principal Component Analysis
