TopHat2 to align RNA with or without annotation file

Question

[plicht]

I want to align my RNA reads against the human reference genome with TopHat2 in order to count features with HTseq-count and perform differential expression analysis. In TopHat2, you have the option to specify an annotation file (GTF/GFF).

I am wondering if you should parse an annotation file here, since downstream HTseq-count will also make use of an annotation file? Sould I include the annotation file in both programs? Or only in HTseq-count?

[lieven.sterck]

you can provide it in the Tophat2 step but you will for sure need to provide it for the htseq-count step

[plicht]

What is the advantage of providing the annotation file to TopHat2? Is it just a matter of computing time?