Hi, 

I am learning how to conduct expression analysis for large population/sample size.

We would like to analyze RNA-seq expression from several hundred TCGA samples using DESeq2. One of the things we would like to correct for is tumor purity. We have the tumor purity for each of the patient ID# ranging from 0-1. 

I would like to include this in the expression analysis model and would greatly appreciate help and advice on how to proceed. 

Where should a covariate or possible confounding factor be adjusted for in the analysis described in here: http://master.bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html

Thank you for your time and help.