Hello Irene, Yes, if using CuffDiff with a reference GTF file, all inputs must in present. One way to create this GTF is using CuffCompare. Our example protocol is to run CuffMerge (which runs CuffCompare as part of the process) with all inputs prior to running CuffDiff. The updated RNA-seq tutorial covers the details: http://main.g2.bx.psu.edu/u/jeremy/p/galaxy-rna-seq-analysis-exercise The Cufflinks website has some sample protocols where CuffCompare is used alone (without proceeding to CuffDiff). http://cufflinks.cbcb.umd.edu/tutorial.html#discovery I know that we have pointed there before, but it really is the best resource (what the content on Galaxy is based on), along with the primary paper and tophat.cufflinks@gmail.com support. Best, Jen Galaxy team On 8/10/12 12:44 AM, i b wrote:

Dear all, given 3 samples, 1 control and 2 treated replicates when I do cuffcompare to produce the gtf input for cuffdiff, do I have to run it with or without the cufflink control? E.g. Cuffcompare with only my 2 treated replicates?--------------use this gtf to run cuffdiff Cuffcompare 1 contro and 2 treated replicates?--------------use this gtf to run cuffdff

Thanks a lot ib ___________________________________________________________ The Galaxy User list should be used for the discussion of Galaxy analysis and other features on the public server at usegalaxy.org. Please keep all replies on the list by using "reply all" in your mail client. For discussion of local Galaxy instances and the Galaxy source code, please use the Galaxy Development list:

http://lists.bx.psu.edu/listinfo/galaxy-dev

To manage your subscriptions to this and other Galaxy lists, please use the interface at:

http://lists.bx.psu.edu/

-- Jennifer Jackson http://galaxyproject.org