**0**wrote:

Hello, I am interested in doing a pairwise condition with DEXSeq, I have read your earlier answer here: C: Pairwise comparasion for multiple conditions using DEXSeq, but I am still confused as to how to implement it. How would I subset the DEXSeqDataSet? I have 2 conditions, 3 replicates each, and 3 stages. Overall, 18 different samples. My data set looks like this:

condition stage

GFP_18_1 GFP 18

GFP_18_2 GFP 18

GFP_18_3 GFP 18

GFP_28_1 GFP 28

GFP_28_2 GFP 28

GFP_28_3 GFP 28

GFP_38_1 GFP 38

GFP_38_2 GFP 38

GFP_38_3 GFP 38

ICD_18_1 ICD 18

ICD_18_2 ICD 18

ICD_18_3 ICD 18

ICD_28_1 ICD 28

ICD_28_2 ICD 28

ICD_28_3 ICD 28

ICD_38_1 ICD 38

ICD_38_2 ICD 38

ICD_38_3 ICD 38

I would like to compare GFP 18 vs ICD 18 (overall not individual replicates), GFP 28 vs ICD 28, etc. How would I subset this and what would be my design matrix, currently I am using ~sample + exon + stage:exon + condition:exon + stage:condition:exon. I have also used ~sample + exon + stage:exon + condition:exon, but I am still trying to figure out the differences between the comparisons. I also would like to know what fitExptoVar variable would be best for my experiment. Thank you!