Hi,

I am looking to do some differential expression analysis using DESeq2 to identify a range of treatment specific effects within my timecourse.

To identify treatment specific effects across time I am using the below model:

dds <- DESeqDataSetFromMatrix(countData = count, colData = sampleinfo, 
                              design = ~[covariates]+ Time + Treatment + Time:Treatment)

# Run the DESeq2 analysis
dds <- DESeq(dds, test='LRT', reduced =~ [covariates]+ Time + Treatment )

However I am also interested in identifying whether there are tissue specific treatment effects within my timecourse. I am unsure on what model to proceed with to identify this. I have seen in some forums a triple interaction is conducted like below, if this is correct how would I go about identifying which tissue each of the effects are within?

dds <- DESeqDataSetFromMatrix(countData = counts, colData = sampleinfo, 
                              design = ~[covariates]+ Tissue + Time + Treatment + Tissue:Treatment + Time:Treatment + Tissue:Time:Treatment)

# Run the DESeq2 analysis
dds <- DESeq(dds, test='LRT', reduced =~ [covariates]+ Time + Treatment )

If this is the wrong strategy I would appreciate any advice!

Yes, if you think the entire time course can be modified by tissue effects then you would include another set of interaction terms.

I recommend working with a local statistician as the number of terms can grow quickly here and you want to make sure you are interpreting the terms correctly. I have to reserve my time on the support site for software related questions.