Fig. 4: Immune landscape, pathway, and gene set analysis of RSS.

a Boxplots of immune, stroma and tumour purity scores by ESTIMATE are in the first column. CIBERSORT results are shown in the second and third columns. Significant differences are observed in tumour purity, immune score, and stromal scores between subtypes in ESTIMATE results. The CIBERSORT algorithm demonstrates the immune cell infiltrations in different subtypes. Plasma cells, CD4 + T cells, CD8 + T cells, regulatory T cells (Tregs), M2 macrophages, and neutrophils are among the most distinguished immune cell signatures between RSSs. Star annotations regarding the p-values are as follows: “****”p < 0.0001; “***”p < 0.001; “**”p < 0.01; “*”p < 0.05; “ns”: P > 0.05. b MCPcounter immune signatures are shown in the first heatmap coloured with z-scores PROGENy pathway activity scores are in the second heatmap, coloured with z-scores (Blue- lower enrichment/activation and Red- higher enrichment/activation). High fibroblast and endothelial cell enrichments in RSS2 can be seen in the first heatmap. TGFβ, NFκB and TNFα activities in RSS2, and MAPK, WNT and EGFR deactivations in RSS3 are the most apparent results. c GSVA dot-plots of selected genesets, colours represent the fold-change, and the sizes of the dots represent the p-value (inversely). EMT and angiogenesis-related genes are highly expressed in RSS2 compared to the others. Whereas expressions of immune and inflammatory response genes are found to be lower in RSS1. All three graphs are based on all rectal samples, microarray and RNA-Seq together (n = 870, Supplementary Table 3).