I have a question about using DEXSeq on a reduced annotation set. I have a set of genes that I used to filter the flattened gtf / gff file, and used this file to count reads only in those exons.

My question is how to interpret p-value, q-value, and log2 fold change in this context.

Should I interpret a q-value < 0.05 as being a bit inflated since there are many fewer exons being tested? Is there anything else about the resulting output that will be affected in a way that changes biological interpretation?

thanks, Sumeet