Fig. 1

Lung squamous carcinoma is associated with inflammatory monocytes. a Kaplan–Meier plots of overall survival in lung squamous carcinoma (LUSC) patients (n = 380) from The Cancer Genome Atlas (TCGA) categorized by mRNA subtype. b Overall survival comparing each mRNA subtype with all others. c Hierarchical clustering of (n = 4291) genes expressed in 348 patients from the LUSC TCGA RNA-seq dataset. d Overall survival comparing Classical versus Secretory subtypes. e Network visualization of Ingenuity Pathway Analysis (IPA) of all differentially expressed genes in the upper portion of the heat map (shown in panel c) that are statistically significant (p < 0.05) in terms of patient survival (in the comparison high ( ≥ median) vs. low ( < median) gene expression), termed “survival genes”. (Gray: least significant, Red: most significant) f. Gene ontology analysis to investigate the biological processes most linked with genes differentially expressed, moving between the Classical and Secretory subtypes (see the Supplementary Methods for details). Individual bars represent most statistically significant GO terms in either the Classical (red bars) or Secretory (green bars) subtype. g Gene Set Enrichment Analyses (GSEA) of the LUSC TCGA dataset. The GSEA is performed going from Secretory to Classical; the GSEA ‘mountain plots’ show only the two most divergent subtypes. Gene set names were shortened to fit this figure. h Table showing 9 genes from the upper portion of the heat map that are associated with reduced overall survival and are markers of monocytes and macrophages. Genes with log-rank p≤0.001 are highlighted in red