Fig. 5

A Dendrogram of gene modules identified from weighted gene co-expression network analysis (WGCNA) (x axis represents different genes, colors represent different gene modules). B Correlation between the eigenvalue of each WGCNA module and the myeloid diversity (y axis represents log-transformed p values, x axis represents correlation coefficients, colors represent different gene modules, shapes represent correlation significance (triangle: significant; round: non-significant)). C Functional gene set enrichment of yellow module genes (y axis represents different pathways, x axis represents log-transformed p values). D Correlation between the random forest-predicted myeloid diversity and the original myeloid diversity in the TCGA cohort. E Top 15 contributing genes in the random forest model from the TCGA cohort, arranged by their relative importance (y axis represents different genes, x axis represents the importance in the random forest model). F Correlation between the random forest-predicted myeloid diversity and the original myeloid diversity in the NCC cohort. E Top 15 contributing genes in the random forest model from the NCC cohort, arranged by their relative importance