Question: Using DESeq2 inside Galaxy
gravatar for marcelops
11 months ago by
United States
marcelops20 wrote:


I have been trying to use DESeq2 on Galaxy, and am having issues with this package.

To illustrate the problem, I have 6 datasets (T1, T2, T3, C1, C2, and C3), being T the treatment samples, and C the control samples.

Here are examples of the content of each sample (I am showing the first lines of T1 and C1 only, but the other datasets are all similar):

gene1   331
gene2   74
gene3   50
gene4   1676.27
gene5   496.99
gene6   0

gene1   361
gene2   59
gene3   30
gene4   1906
gene5   639
gene6   12

In the package DESeq2 1.8.2 on Galaxy, I am using the following arguments:

  • Factor: Treatment
  • 1: Factor Level: Treated
    • Count Files: T1, T2, T3
  • 2: Factor Level: Control
    • Count Files: C1, C2, C3

Then I got the following error:

DESeq2 run information

sample table:
dataset_1.dat     Treated
dataset_2.dat     Treated
dataset_3.dat     Treated
dataset_4.dat     Control
dataset_5.dat     Control
dataset_6.dat     Control

design formula:

primary factor: Treatment


I couldn't find the documentation on how to use the Galaxy package DESeq2, and I am not sure about the format of the input files.

Has anyone successfully used DESeq2 inside Galaxy? Could you please let me know how your inputs look like, or if you have any info on how to properly use this package?



rna-seq galaxy deseq2 • 1.3k views
ADD COMMENTlink modified 11 months ago by Bjoern Gruening4.5k • written 11 months ago by marcelops20

Hello, Some tests are running to determine if htseq-count is producing the correct input. This tool form is new to me as well, so am testing a few things out to see where the corner cases are that could trigger errors. Feedback from me early next week.

Thanks for the details here, very helpful. You usage case should work. But that is part of the test. If you can reproduce this at and want to submit a bug report (with a link to this post), that could be helpful if there are other minor input issues. If you do not actually have errror datasets, but rather "green" failure datasets, a shared history link sent to is another way to allow us to review. All datasets for review must be left undeleted. And it is best to not submit a very large history, as these are difficult to import. Just the inputs for this test (sam inputs, reference gff used, htseq-counts, then this tool's datasets).

Thanks, Jen, Galaxy team

ADD REPLYlink written 11 months ago by Jennifer Hillman Jackson21k

sorry can't I use raw counts file coming from featurecounts???

ADD REPLYlink written 5 months ago by sa63_tanha0

Btw. why do you think this is an error? Is your dataset red? I can not see any error message in it, for me it looks like normal stdout messages about the design matrix.

ADD REPLYlink written 11 months ago by Bjoern Gruening4.5k
gravatar for Bjoern Gruening
11 months ago by
Bjoern Gruening4.5k
Bjoern Gruening4.5k wrote:


we are using DESeq2 very successfully, so it should work. However your inputs are looking strange. This should be count data from htseq-count.

We have a small tutorial here: Maybe this gets you started.

Cheers, Bjoern

ADD COMMENTlink written 11 months ago by Bjoern Gruening4.5k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 58 users visited in the last hour