Dear all,

I have the following question: Imagine I have colData (Metadata) with Gender, Age and City.

Now the city could be New York, Paris, Florida.

For my Differential Expression Analysis on city I only want to use the city Paris and Florida.

If I create my code I wonder how I can filter colData before applying this:

dds <- DESeqDataSetFromMatrix(countData = cts, colData = colData, design = ~ Age + City)

Thank you very much, Bine

Additional question: Is there a way i could also put Paris + Florida together and do a Differential Expression against New York? Thank you!

In general, put all your samples in the dds object, so that all the samples are used for library normalization and dispersion estimates. When you call results, you can use contrasts to specify what to compare to what.

You can combine cities together either by making a list in contrasts, or by making a new column of ColData which groups the samples the way you want. See this answer:

DESeq2 comparison two cell types and two genotypes