Research Paper Volume 15, Issue 21 pp 12330—12368

Figure 7. Correlations of molecular prognostic subtypes with immunological features. (A) Infiltration comparisons of 22 immune cell types. The scores of some immune cell types had significant differences between two clusters, such as T cells CD4 memory resting, mast cells resting and dendritic cells resting with higher scores in the cluster 2. *p < 0.05, **p < 0.01 and ***p < 0.001. (B) Infiltration comparison of stromal and immune cells. The cluster 2 had higher stromal score, immune score and ESTIMATE score. *p < 0.05, **p < 0.01 and ***p < 0.001. (C) Comparisons of activities of 10 oncogenic signaling pathways. The scores of cell cycle, MYC, NRF1, TGF-beta, RAS and WNT pathways had significant differences between two clusters, and these oncogenic signaling pathways had lower activities in the cluster 2. *p < 0.05, **p < 0.01, ***p < 0.001. (D) Comparisons of tumor immune dysfunction and exclusion (TIDE) score, interferon-gamma (IFNG), T cell dysfunction (Dysfunction), T cell exclusion (Exclusion), tumor-associated macrophages M2 (TAM.M2) and myeloid-derived suppressor cells (MDSC). These terms had significant differences between two clusters, and TIDE, IFNG, Exclusion and MDSC in the cluster 1 were significantly higher than those in the cluster 2. Dysfunction and TAM. M2 in the cluster 1 were significantly lower than those in the cluster 2. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. (E) Comparisons of estimated IC50 values for 6 conventional chemotherapy agents. The estimated IC50 values for erlotinib, sunitinib, paclitaxel, VX-680, TAE684 and crizotinib in the cluster 1 were significantly lower than those in the cluster 2. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.