Carrie,

We've had issues when  gene ids contain a '#' character, which is the R comment 
character.
We've had issues when the group names in a plot weren't exactly the same as 
those used in the cuffdiff inputs.

Are you linking in all the individual output files from cuffdiff?
I included a cuffdiff tool in the cummeRbund toolshed respository because the 
original cuffdiff galaxy tool didn't  output all 21 files that cummeRbund 
prefers.
Are you also including the "genome reference and Feature file". The gtf may be 
needed for some functions.

JJ


On 6/3/14, 11:54 AM, Jennifer Jackson wrote:
Hi Carrie,

I'll start this off by stating that I am not as familiar with this wrapper as many of the others. I 
am bringing in the repository owner (Jim Johnson) to the thread so that he can offer help. You can 
also reach him (or any other repository owner) when logged in under the "Repository 
Actions" menu under "Contact Repository Owner".

Just from a quick look, I am wondering if the R package dependencies are installed with all 
necessary components and available to the "Galaxy user". Or, if the ones you tried are 
the correct ones to be using with this tool. The "attributes" warning is one that is 
given by R.

Going forward, using the galaxy-...@bx.psu.edu mailing list will be best for 
questions like this one. This list is being retired on Friday (replaced by 
Galaxy Biostar). Please join us there for usage questions (although there is 
some cross-over in questions and we are open to both, the community is still 
actively using galaxy-dev for primary local install help and you will reach 
more developers there): https://wiki.galaxyproject.org/Support#Biostar

Hopefully Jim will see this thread, but also try contacting him direct if no 
reply is sent here.

Sorry I could not help more,

Jen
Galaxy team

On 6/2/14 2:57 PM, Ganote, Carrie L wrote:
Hi list,

Is the cummeRbund wrapper on Galaxy yet functional? I've got some replicates 
that I'm trying to test for a user on our local Galaxy, but I just can't get 
the program to behave.

Using cuffdiff v2.2.0 604fa75232a2
cummeRbund version 0.0.7, 137aab1d9ac1
R 2.15.2 and 3.0.1 (tried both)
Last galaxy-dist/stable changeset: 7a7985a007fb

I ran cuffdiff like so:
Transcripts     83: Cuffmerge on data 6, data 67, and others: merged transcripts
Name     control
Add replicate     44: Tophat2 on data 1 and data 6: accepted_hits (ran as 
single-end fastq)
Add replicate     54: Tophat2 on data 2 and data 6: accepted_hits (ran as 
single-end fastq)
Name     sample
Add replicate     59: Tophat2 on data 3 and data 6: accepted_hits (ran as 
single-end fastq)
Add replicate     64: Tophat2 on data 4 and data 6: accepted_hits (ran as 
single-end fastq)
Library normalization method     geometric
Dispersion estimation method     pooled
False Discovery Rate     0.05
Min Alignment Count     10
Use multi-read correct     No
Perform Bias Correction     No
Include Read Group Datasets     No
Set Additional Parameters? (not recommended for paired-end reads)     No

Here's what I get when I run cummeRbund:
(From Galaxy, with R 2.15.2)
Creating database //path/job_working_directory/005/5051/galaxy_dataset_16602.dat
Reading //path/dataset_files/009/dataset_9861.dat
Checking samples table...
Populating samples table...
Writing genes table
Reshaping geneData table
Recasting
Writing geneData table
Reading //path/dataset_files/009/dataset_9861.dat
Error in function (classes, fdef, mtable)  :
   unable to find an inherited method for function 'make.db.names' for signature 
'"SQLiteConnection", "NULL"'
Calls: readCufflinks ... make.db.names -> .valueClassTest -> is -> <Anonymous>
Execution halted

(From R interpreter on command line, R = 3.0.1)
Reading //path/dataset_files/009/dataset_9842.dat
Checking samples table...
Populating samples table...
Writing genes table
Reshaping geneData table
Recasting
Writing geneData table
Reading //path/dataset_files/009/dataset_9842.dat
Error in (function (classes, fdef, mtable)  :
   unable to find an inherited method for function 'make.db.names' for signature 
'"SQLiteConnection", "NULL"'
In addition: Warning message:
attributes are not identical across measure variables; they will be dropped

I tried changing the group names in cuffdiff but no luck. No spaces or funny 
characters.

Thanks for any insight,

Carrie Ganote


___________________________________________________________
The Galaxy User List is being replaced by the Galaxy Biostar
User Support Forum at https://biostar.usegalaxy.org/

Posts to this list will be disabled in May 2014.  In the
meantime, you are encouraged to post all new questions to
Galaxy Biostar.

For discussion of local Galaxy instances and the Galaxy
source code, please use the Galaxy Development list:

   http://lists.bx.psu.edu/listinfo/galaxy-dev

To manage your subscriptions to this and other Galaxy lists,
please use the interface at:

   http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:

   http://galaxyproject.org/search/mailinglists/



--
James E. Johnson, Minnesota Supercomputing Institute, University of Minnesota
___________________________________________________________
The Galaxy User List is being replaced by the Galaxy Biostar
User Support Forum at https://biostar.usegalaxy.org/

Posts to this list will be disabled in May 2014.  In the
meantime, you are encouraged to post all new questions to
Galaxy Biostar.

For discussion of local Galaxy instances and the Galaxy
source code, please use the Galaxy Development list:

 http://lists.bx.psu.edu/listinfo/galaxy-dev

To manage your subscriptions to this and other Galaxy lists,
please use the interface at:

 http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:

 http://galaxyproject.org/search/mailinglists/

Reply via email to