Downstream Cuffdiff analysis

[apadr007]

Hi guys,

I am working on the downstream analysis of my rna seq data using Cufflinks, Cuffcompare, and Cuffdiff. I basically want to know three simple questions that I am having a lot of difficulty with.

1.how many transcripts represent how many genes?
2.how many transcripts match previously known reference transcripts?
3.out of these how many isoforms are there?


Which output file should I be using for my quantification? If I use "genes.fpkm.tracking" from Cuffdiff, I notice there are repeated genes in "gene_short_name" -- this is obviously wrong for quantification.

What files should I use to address my 3 questions and why on earth are genes being repeated in "genes.fpkm.tracking"??

thanks

[apadr007]

Can anyone shed some light -- 33 people have view, no one has posted

[cw11]

Responding to your last point first, I think the repeated genes in gene_short_name occur because some genes have multiple transcription start IDs. I notice that each of the repeat short names has a unique tss_id.

As for your other questions, I'm not quite sure I understand what you're asking. If you're looking for fpkm values for each of your genes, then genes.fpkm.tracking would probably be the best file to look at. But it sounds as though you're more interested in generating a kind of summary of what's within the Cuffdiff output files. Is that the case?

I also just ran across (on the cufflinks website) information about CummeRbund which apparently helps users handle their output data - you could perhaps look into this.

I would suggest looking at the Cufflinks/compare/diff manual, which has a lot of information about what each of the output files means.