Comprehensive Tumor Microenvironment Analysis

The gene set variation analysis (GSVA) was performed to identify specific pathways of each subtype (24 (link)). We downloaded Hallmark and KEGG gene sets from the Molecular Signatures Database and further transformed the gene expression matrix into gene set matrix using the GSVA package. Afterwards, we performed gene sets difference analysis using the limma package and the screening threshold were set to |log2 fold change (FC)| >0.2 and adjusted P-value <0.05. Adjusted P-value was obtained from the Benjamini–Hochberg multiple test correction.
Referring to Charoentong et al. study (25 (link)), we obtained the markers of 23 immune cells including: innate immune cells (activated dendritic cells, CD56+ natural killer cells, CD56− natural killer cells, eosinophils, immature dendritic cells, macrophages, mast cells, MDSC, monocytes, natural killer cells, neutrophils, and plasmacytoid dendritic cells) and adaptive immune cells (activated B cells, activated CD4+ T cells, activated CD8+ T cells, Gamma delta T cells, immature B cells, natural killer T cells, Treg cells, follicular helper T cells, Th1 cells, Th2 cells, and Th17 cells). Endothelial cells and fibroblasts, also the important components of TME, played a crucial role in tumor inflammation, angiogenesis, invasion, and metastasis. The markers of endothelial cell and fibroblast were retrieved from the MCP-counter (26 (link)) (Table S3). Based on these markers, we applied the single sample gene set enrichment analysis (ssGSEA) algorithm to evaluate the infiltration abundance of 25 TME cells.