Question

Running cufflinks for spliceR

1

Entering edit mode

hliang ▴ 20

@hliang-6719

Last seen 10.1 years ago

United States

Dear spliceR developer, It's convenient that spliceR can take cuffdiff output from cufflinks tools. But I wonder what's the recommended way to run cufflinks tools. I've read the spliceR paper and bioConductor vignette, but could not find the related information. specifically: 1. is novel annotation necessary? 2. cufflinks can do novel annotation with or without reference genome, which way would be recommended? 3. cuffdiff needs gtf input file. Will the canonical gtf annotations (e.g. UCSC) be good enough? Or is the merged gtf (from multiple gtf files by cufflinks/cuffmerge) better? 4. Should the gtf file used by readCufflinks() be the same as the one used by cuffdiff? Thanks -Hanquan [[alternative HTML version deleted]]

Annotation spliceR Annotation spliceR • 1.6k views

ADD COMMENT • link updated 9.5 years ago by k.vitting.seerup ▴ 120 • written 10.2 years ago by hliang ▴ 20

score 0 · Answer 1 · 2015-05-28

Dear Hanquan

The problem is that how you run cufflinks/cuffdiff very much depends on you experimental setup and what you want do do with your RNA-seq data.

Generally I would run cufflinks with annotation as guideline - and I would use a lenient annotation (gencode, ensemblor UCSC). Then I would use the merged GTF (from all the individual cufflinks runs) as input to cuffdiff.

You can get more information in the protocol from the people behind cufflinks/cuffdiff here: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3334321&tool=pmcentrez&rendertype=abstract

If you provide more information about your experimental setup I can give more specific advice.

Hope this helps.

Kristoffer