[galaxy-dev] Cufflinks with reference annotation and without reference annotation

[Jennifer Jackson]

Repost:

Crystal,

If you provide a gene annotation to Cufflinks, the transcripts produced 
will match those in the annotation exactly. If you assemble without a 
gene annotation, the transcripts produced will match the reference in 
some cases, but, in others, will not match the reference due to small 
and/or large errors. Because '=' denotes an exact match between an 
assembled transcript and a reference transcript, more '=' are to be 
expected when Cufflinks has a gene annotation.

Finally, a couple procedural issues:

*please send questions about analyses and tool usage to the galaxy-user 
mailing list, not galaxy-dev or individual developers;
*please do not send duplicate emails as it can confuse our tracking 
system and slow down our response rather than speed it up.

Good luck,
J.

On 8/17/11 9:42 AM, Crystal Goh wrote:
>
>
> Hi, I am Crystal. I have some problem with Cuffdiff output. Hope can get
> some advice. Thanks.
>
>
> After aligning RNA-seq reads with Tophat, I used the Tophat output for
> Cufflinks.
>
> For Cufflinks, I tried two approaches and compared the results:
> 1st approach: Put zebrafish Ensembl GTF as reference annotation
> 2nd approach: without reference annotation.
>
>
>  From the output of above 2 approaches, I continued with Cuffcompare
> (with reference annotation) and Cuffdiff,
> Attached word document is the workflow and parameters I set for these 2
> approaches.
>
>
> When I compared the output of Cuffdiff between these 2 approaches, a
> total of 48584 tracking id with class code "=" was observed in trancript
> FPKM tracking file from Approach 1, whereas there is only 1248 tracking
> id with class code '=' from Approach 2 (I attached transcript FPKM
> tracking files from approach 1 and 2)
>
> In my opinion, I should observe 48584 tracking id with class code '='
> and additional tracking id with other class codes in transcript FPKM
> tracking file from Approach 2.
>
> Can I get advice on this?
>
>
> Thank you.
>
> Best regards,
> Crystal
>
>
> ___________________________________________________________
> Please keep all replies on the list by using "reply all"
> in your mail client.  To manage your subscriptions to this
> and other Galaxy lists, please use the interface at:
>
>    http://lists.bx.psu.edu/

-- 
Jennifer Jackson
http://usegalaxy.org
http://galaxyproject.org/Support