Hello,

I'm trying to use a model like what described in Chapter 3.5 "Comparisons both between and within subjects" of EdgeR users' manual:  

Disease Patient Treatment

1 Healthy 1 None

2 Healthy 1 Hormone

3 Healthy 2 None

4 Healthy 2 Hormone

5 Healthy 3 None

6 Healthy 3 Hormone

7 Disease1 1 None

8 Disease1 1 Hormone

9 Disease1 2 None

10 Disease1 2 Hormone

11 Disease1 3 None

12 Disease1 3 Hormone

13 Disease2 1 None

14 Disease2 1 Hormone

15 Disease2 2 None

16 Disease2 2 Hormone

17 Disease2 3 None

18 Disease2 3 Hormone

design <- model.matrix(~Disease+Disease:Patient+Disease:Treatment)

Where Healthy and None are the baseline for Disease and Treatment, respectively.

My question is, in this model how can I compare the difference between Healthy None and Disease1 None.

I cannot find a suitable contrast to do the job from its model matrix. How can I change the model to include this effect, if I still wanna nest individuals and treatments within Disease?

Thank you very much.

David

Making the baseline comparison with this model is not possible using edgeR, because edgeR assumes that samples are independent whereas repeated measurements on the same individual are correlated.

You can do the analysis using limma with duplicateCorrelation(), see the section of the limma User's Guide on "Multi-level designs".

In the edgeR, you would have to analyse all the "None" samples separately to the "Hormone" samples to make the baseline comparison.