Fig. 7

Multi-omics characterization of NRP1 expression patterns and its context-dependent prognostic value in ccRCC. A H&E-stained tissue section from ccRCC patient for spatial transcriptome. B Cell type annotation of spatial transcriptomics (ST) data using deconvolution analysis. C Spatial expression map of NRP1 overlaid on the ST data. D, E UMAP plots of the ST dataset displaying cell clusters (D) and NRP1 expression levels across cell types (E). F,G UMAP plots of an independent scRNA-seq dataset (GSE171306) showing annotated cell types in ccRCC (F) and NRP1 expression (G). H Violin plots showing the distribution of NRP1 expression across different cell types. I Correlation analysis between NRP1 expression and the inferred proportion of major cell types across TCGA samples, based on BayesPrism deconvolution. J Scatter plot with marginal density curves showing a strong inverse correlation between NRP1 expression and tumor cell proportion (Spearman R = −0.640, p < 0.001). K Box plot comparing NRP1 expression levels between samples with high versus low tumor cell proportion. L, M Kaplan–Meier survival analysis stratified by NRP1 expression in Low-Tumor (L) and High-Tumor (M) groups