How to visualize RPKM in a browser?

[dmotti]

Hallo everyone,

I'm new in the forum. Recently I've been dealing with some NGS data coming from the Illumina platform. I used the classical Tophat-Cufflinks pipeline to analyze the data.
But I'm dealing with the issue of visualizing RPKM values in a genome browser.

UCSC does not support GTF files, nor GFF, I can upload the wig files, but that would help only with the raw reads.

In another thread on this forum someone suggested a modification of the GTF file that would fit in Gbrowse. I tried to do that and uploaded the result on Gbrowse in MGI. No output. the track is just not visualized (even if no error output is given).

In this very moment I'm somehow stuck there, can anyone help me with that?

D.

[Siva]

Hallo everyone,

I'm new in the forum. Recently I've been dealing with some NGS data coming from the Illumina platform. I used the classical Tophat-Cufflinks pipeline to analyze the data.
But I'm dealing with the issue of visualizing RPKM values in a genome browser.

UCSC does not support GTF files, nor GFF, I can upload the wig files, but that would help only with the raw reads.

In another thread on this forum someone suggested a modification of the GTF file that would fit in Gbrowse. I tried to do that and uploaded the result on Gbrowse in MGI. No output. the track is just not visualized (even if no error output is given).

In this very moment I'm somehow stuck there, can anyone help me with that?

D.

Hi
I may be wrong but I think since GFF or GTF files are tab delimited files containing details of assembled features (transcripts or whatever), you may not be able to visualize it in a browser at least not like an aligned file. You could feed the GFF/GTF coordinates to a statistical software and output a graph/chart.

Siva

[dmotti]

Hi Siva,

Thanks for your answer. Indeed you're right, GTF and GFF files represent discrete elements, they would only show as a coverage on the browser.
however, there is the possibility to unable an abundance representation of the elements by setting a score for each of them that should be between 1:1000.
Each element at that point would be represented in a different color. I had to force my RPKM values into a scale of 1 to 1000, and considering I have a much wider range it does not represent exactly what I'm looking for.


In this thread:

http://seqanswers.com/forums/showthread.php?t=2804&highlight=.expr+file

vkrishna suggested a modified version of the GTF file that could be read by Gbrowse. I tried that and it did not work, but the kind of representation I'm looking for is exactly the one suggested by the user.



The ENCODE project presents several tracks derived from RNA Seq data that can be visualized in Genome Browser, so there must be a way to submit RPKMs to the browser.

Any suggestion?

D.

[jdanderson]

Hello dmotti,

Did you ever find a solution to the visualization of cufflinks output files? I am currently in the same boat. I was thinking about using DESeq, but it seems that it only wants the read counts, not any other type of value.

Any guidance would be most appreciated.

Regards,
Johnathon

[dawn1313]

Hi,

It seems I haven't found the exact response to what kind of genome browser for cufflinks outputs. Basically I want to visualize the reads abundance (hopefully also as a log scale) across the genome from either the Tophat output 'accepted_hits.bam' or the Cufflinks output 'transcripts.gtf'.

Is there any suggestion? Thank you very much!
Dawn