I have been wondering whether or not to include my vital status condition (Alive or Dead) along my inflammatory groups, which are artificially created prior to DESeq2 analysis. I am aware of https://support.bioconductor.org/p/101447/ and the FAQ where it is recommended to run DESeq2 along with all samples and all conditions, except when the data among a group has high variance.
However since my inflammatory group condition is artificially created out of the data set (by doing a mean of Z-score of an inflammation signature, and ranking the samples into multiple groups), I am not sure if combining every condition is a then a good idea.
What is confusing is that DESeq2 with vital status alone as design or with both vital status and inflammatory group as design, while extracting vital status as contrast, doesn't give the same result. and I don't know which one should be more appropriate and correct.