Fig. 6

Tumor-stroma interplay reshaped the extracellular matrix in tumor thrombus associated with poor survival. a Dot plot showing inferred interactions between epithelial cells (proximal tubule cells from ARTs and malignant cells from PTs and TTs) and macrophages, endothelial cells, and myofibroblasts. Circle size indicates the significance of the interaction, and circle color indicates the mean expression of ligand and receptor genes. The red letters represent ligands, and the black letters represent receptors. b Scatterplots demonstrating the extracellular matrix assembly (left) and EMT (right) signature score versus total myofibroblast cell fraction in bulk RNA-seq of PT-TT paired samples in the validation cohort. Cell-type fractions were inferred using CIBERSORTx. The Pearson coefficient (R) and associated p value are reported for each correlation. c Scatterplot depicting inferred cell-to-cell interactions between tumor cells and myofibroblasts by combining scRNA-seq data and bulk RNA-seq data. d Box plot showing the calculated correlation between extracellular matrix remodeling-related genes and myofibroblast relative abundance involved in PTs and TTs. p values were determined by a two-sided Wilcoxon test. e Scatterplot representing inferred cell-to-cell interactions between myofibroblasts and endothelial cells by combining scRNA-seq data and bulk RNA-seq data. f Box plot showing the calculated correlation between extracellular matrix remodeling-related genes and endothelial cell relative abundance involved in PTs and TTs. p values were determined by a two-sided Wilcoxon test. g Signature scores for extracellular matrix assembly (left) and collagen-containing extracellular matrix (right) in paired PT-TT bulk RNA-seq samples in the validation cohort. p values were determined by a two-sided Wilcoxon test. h Progression free survival for the TCGA KIRC cohort based on high tumor-stroma and stroma-tumor interaction signatures versus low signature expression