Hi,

I am trying to decide on the best design for my EdgeR differential-expression analysis. I have RNA-seq data from four patients - matched blood and liver for three of the patients and only liver from one of the patients (as below).

Donor  Tissue
1      Blood
1      Liver
2      Blood
2      Liver
3      Blood
3      Liver
4      Liver

The two options I am considering for the design formula are:

1. ~ Tissue
2. ~ Donor + Tissue

Is it possible to use the second approach when one of the donors only has a liver sample? What is the best approach to get around this issue?

Many thanks for the advice.

Best,

Lucy

You need to use formula 2. if you wish to correct for donor baseline expression levels. It will work fine but the fourth donor will not contribute to the results.