Map with Tophat, DEGs with EdgeR. Possible?

[Gonza]

Dear all,

Please some general and technical advice with the following. I would like to run edgeR as an alternative to cufflinks. I have successfully mapped all my reads to the TAIR10 genome using tophat2. If I now want to use edger I am guessing I do not have to map again as I already did with tophat2.

The general question is, can I use the tophat2 files to run edger? and if so, how do you even begin to think about it? If you follow the tuxedo pipeline you will use the "accepted_hits.bam" in the cufflinks scripts, but i have no idea how to run edger from here.

Any thoughts/ideas are welcome.

Many thanks.
-G

[mikep]

Yes you can. You just need a script to determine counts per gene (like htseq-count), then you are good to go.

[Gonza]

Thanks for your answer.
Can you please provide me an example as to how that would work?

-G

[mikep]

Well here's the page on htseq-count

http://www-huber.embl.de/users/anders/HTSeq/doc/count.html

With pretty simple instructions. The output is raw counts per gene, which you can then feed directly into edgeR

[Gonza]

ok, thanks again. Will look into it.
-G