HNSC extended cohort. HNSCs collected from TCGA and CPTAC were divided in HPV + , HPV - CNA low and HPV - CNA high samples based on HPV infection and level of aneuploidy [ 41 ]. KaplanMeier survival curves between HPV + and HPV - ( B ) or HPV - CNA low and HPV - CNA high ( C ) HNSC patients. Overall survivals were compared using the log-rank test. Comparison of CIBERSORTx absolute score medians between HPV + and HPV - ( D ); HPV - CNA low and HPV - CNA high ( E ); or HPV + and HPV - CNA low ( F ) HNSCs. Only immune cell types enriched in at least one HNSC subtype are shown. Comparison of sample proportion in the five TIME features between HPV + and HPV - ( G ) or HPV - CNA low and HPV - CNA high ( H ) HNSCs (see Methods). TIME drivers more frequently damaged in HPV + HNSCs ( I ), HPV - CNA low ( J ), or HPV - CNA high HNSC samples ( K ). For HPV - CNA high HNSCs only the top 13 TIME drivers are shown (full list in Additional File 2 : Table S4). CPI = cancerpromoting inflammation; CYS, cytotoxicity score; CPTAC, Clinical Proteomic Tumour Analysis Consortium; FDR, false discovery rate; HPV, human papillomavirus; ICR, immunologic constant of rejection; IS, immune score; TIS, tumour inflammation signature. Proportions were compared using Fisher’s exact test ( D-F , I-K ) or MantelHaenszel chisquare test ( G , H ). Distributions ( M , N ) were compared using KruskalWallis test. In D-K , BenjaminiHochberg correction for multiple testing was applied
