plotDEXSeq no output

usegalaxy.org support
, ,
1

I tried to run plotDEXSeq on my output rds file (after converting it to rdata, database is mm10) but the resulting pdf file is empty. I’ve based the gene identifier on the DEXSeq result file but the error code says: No read counts falling in this gene, there is nothing to plot.

2

Welcome, @VI_Rodriguez

Things to check

  • The way to interpret that result is as a warning about what happened, not a tool bug/failure. Does your query gene have content in the files? That’s the first place to check, and could be a legitimate result.

  • If you are missing expected counts: are all inputs based on the same version of mm10? Meaning, the chromosome identifiers are the same? Galaxy uses the UCSC version for any server indexes. UCSC Genome Browser Downloads with identifiers like: chr1, chr2, … chrM.

  • Then, maybe explore what is going on in a genome browser? IGV or UCSC are common choices.

You can share more data if you want: please see the banner at this forum for the how-to, or you can find the help directly here:

Let’s start there. :slight_smile:

3

There doesn’t seem to be a problem with the rest of the DEXSeq result files as seen here

image
image637×834 28.5 KB

image
image1028×854 74 KB

image
image1037×790 116 KB

Here is the job information for the plotDEXSeq run: ( 11ac94870d0bb33ac1950e04d7fe814a)

image
image1098×868 66.2 KB

image
image1103×627 37.9 KB

I tried multiple gene IDs but there is still no result for plotDEXSeq, but count files have the right contents. Could it have been an issue with the initial DEXSeq run?

4

Hi @VI_Rodriguez

Thanks for sharing those details, super helpful.

So … Bioconductor tools sometimes do not “like” gene identifiers with the version at the end. This happens everywhere, not just Galaxy.

The solution is to modify: gene.N → gene (strip off the .N).

Make the change in all files. This makes the variables R friendly (alphanumeric with optional underscores). I haven’t tested out converting the . to a _ but you could see if that is an alternative successful way. If not, just strip it all out.

This mini guide includes tips about formatting gotchas with these tools: FAQ: Extended Help for Differential Expression Analysis Tools

And, those manipulations can be done in a history or in an interactive environment (depending on where you are starting from). Tutorials → GTN Materials Search (query=olympics)

5

I removed all the version numbers and it worked. Thanks for your help!

1 Like
6

Very glad that worked! :rocket: