Hello, I am working with an RNAseq dataset composed of samples in two Conditions (Healthy, Sick) and then each group of samples is treated with three different treatments (T1, T2, T3). What I want to do is to compare treated samples between Conditions (e.g. T1 Sick vs T1 Healthy, T2 Sick vs T2 Healthy ...). In addition, I have a batch effect that I want to include in the model and I was thinking of using the following design.

  dds <- DESeqDataSetFromMatrix( matrix, coldata, design = ~  Batch +  Condition + Treatment + Condition:Treatment)

Then in results(), I specify the following to do the comparison T1 Sick vs T1 Healthy (setting Healthy as the reference level):

res <- results(dds, contrast = list("ConditionSick.TreatmentT1"))

Is it correct? Do you have any suggestions on this?

Another solution I had come up with is to subset the dataset and then do a different analysis for each treatment. Is it wrong?

Thanks in advance for your help!

To compare one subset of samples to another, make a new column of colData that is 'condition_treatment', and use that ( + batch) as your design.