How to name samples in CummeRbund?

[veber]

Hello,

Problem: I cannot figure out which sample is which in Cummerbund or how to name them.

Details:
I used Galaxy to run Cufflinks, Cuffmerge, and Cuffdiff. I then downloaded files, renamed them to Cufflinks convention and am trying to visualize my data with CummeRbund in RStudio. I cannot figure out which sample is which and how to name them. For example, my density plot just shows samples as q1, q2.

Can someone please help me out?

[veber]

It seems possible that the only way to do this is to rename sample names from Sample1/q1, Sample2/q2, etc. in cuffdiff output. I understand from Cufflinks manual that the first sample input into cuffdiff is Sample1/q1 and the second is Sample2/q2.

For example,

285: control.bam
287: test.bam
414: cuffmerge.gtf

425: Cuffdiff on data 285, data 287, and data 414: transcript FPKM tracking

Output
Sample 1= q1 = 285 = control.bam
Sample 2 = q2 = 287 = test. bam

Can someone confirm this and touch on whether this is normal workflow? Is there a way to enable Galaxy to keep original sample names?

[veber]

*Bump

Anyone?

[dpryan]

N.B. I'm not familiar with Galaxy. Cuffdiff will set q1/q2 in the order of the files given to it. That is, if you execute "cuffdiff OPTIONS transcripts.gtf control.bam test.bam" or similar then control is usually q1. In the future, use the "-L Label1,Label2,..." switch to make your life easier. I would assume there's a way to do that in Galaxy.

[lechu502]

sed replacment

If you run the experiment with several samples

ie. cuffdiff -o cuffdiff/all REF.gff SAMPLE1.bam,SAMPLE1a.bam,SAMPLE1b.bam SAMPLE2.bam,SAMPLE2a.bam,SAMPLE2b.bam SAMPLE3.bam,SAMPLE3a.bam,SAMPLE3b.bam

you can rename them by entering cuffdiff output directory and running:

sed -i.bak -e 's/q1/SAMPLE1/g;s/q2/SAMPLE2/g;s/q3/SAMPLE3/g' *.diff *.info *_tracking