Skip to main content
Fig. 4 | BMC Immunology

Fig. 4

From: Dissecting cellular states of infiltrating microenvironment cells in melanoma by integrating single-cell and bulk transcriptome analysis

Fig. 4

Identification of tumor-immune phenotypes. A Heatmap showing Pearson coefficients of correlation for relationships between each pair of cell states (top), red for positive correlations and blue for negative correlations. The cell states are colored according to cell states shown in Fig. 3A. Scatterplots showing relationships between T_CD8_Mixed (Cytotoxic and Exhausted) with T_CD8_Cytotoxic (top) and B_Non-regulatory (middle), CAF and Endo (bottom). Pearson correlations and p values are indicated. For significant correlations, linear models are shown as blue lines. B The first two components of correspondence analysis, accounting for 47% of the co-association structure between cell subsets and patients in the cohort, are shown. Cell states are displayed as red triangles, and patients as blue circles. Red texts highlight the important cell states for the first two components. Red, light blue and blue ellipses indicate the three patient groups of TMIP0, TMIP1 and TMIP2 classified by median values of CA-1 and CA-2, respectively. Circular heatmaps showing the z-score normalized cellular fractions for cell states (row, colored according to cell states shown in Fig. 3A) across the three patient groups (TMIP0, TMIP1 and TMIP2) classified by median values of CA-1 and CA-2. Median value difference of cell state fraction among groups was evaluated using Mood’s test. **** P < 0.0001 C The different structures of main cell states in the tumor microenvironment

Back to article page