Extracting Tumor Tissue Immune Status from Expression Profiles: Correlating Prognosis with Tumor-Associated Immunome

Investigating the expression of genes in cancer-associated immune cells (immunome) is imperative for prognosis prediction. However, evaluating the expression of immune-associated genes within cancer biopsy is subject to significant inconsistencies related to the sampling methodology. We present immFocus, a method for extracting immune signals from total RNA sequencing of tumor biopsies, intended for immunity depiction and prognosis evaluation. It is based on reducing the variation which biopsy preparation adds to the apparent expression levels of immune genes. We employed immFocus to normalize gene expression with an immune index using data obtained from kidney renal clear cell carcinoma biopsies. Genes that became less variable due to normalization were found to be preferentially immune-related. Moreover, immune-related genes tended to become more prognostic due to the normalization. These results demonstrate, for the first time, that whole transcriptome sequencing can be used for interrogation of a cancer immunome and for advancing immune-based prognosis.