DESeq2 unknown treatment option?

[bruce01]

I had considered that as the most realistic option, just wondered if there was a way to use unknowns. Thanks for your input.

[dpryan]

The best approach is to simply exclude the unknown responders.

[bruce01]

Hi all,

I am wondering how I can include 'unknown' treatment/condition of samples in my design. I have two factors, one is method of mRNA preparation (RiboZero or PolyA), the second is 'responder' or 'non-responder' to a drug that has been administered. There is some missing data here, so about 1/4 of samples are classified 'undetermined response'. At first I thought to include a simple 'unknown' condition, but samples are either responsive to treatment or not. Therefore I thought that classifying a third condition in this way would not make sense. Is there a way to include no information for those samples not designated 'responder'/'non-responder'?

Thanks, Bruce.