Hello Im trying to do an RNA seq differential expression analysis using DESeq2 by doing a likelihood ratio test The DESeq2 manual states " The likelihood ratio test can be performed by ...providing a reduced design formula, e.g. one in which a number of terms from design(dds) are removed"

They then shows this as an example script you would use to this : dds <- DESeq(dds, test="LRT", reduced=~1)

1) how do you determine how many terms are removed from design(dds) ? 2) Is the number of terms removed always 1 like they show in the example script above ?

You can remove as many coefficients as you want. This is up to you as the analyst.

You can either provide a reduced design formula, as above, or you can provide model matrices directly:

dds <- DESeq(dds, test="LRT", full=mm, reduced=mm0)

Where mm and mm0 are created by model.matrix().