Hey Jun,
Like Jen mentioned, you should use multiple output elements if at all
possible. The format for specifying outputs you are describing should only
be used if the tool does not know the number of outputs to be used prior to
execution of the tool. I believe it is a known limitation of this mechanism
multiple output mechanism that such tools cannot be used in workflows -
many scientific workflow platforms more sophisticated than Galaxy cannot
handle such tools - it is hard to reason about workflows without knowledge
of their structure ahead of time.
That said, there is an open and high priority Trello card outlining
dataset collections that will hopefully provide some more robust
alternatives to address use cases currently addressed with this multiple
output mechanism -
.
-John
On Fri, Nov 8, 2013 at 8:47 PM, Jennifer Jackson <jen(a)bx.psu.edu> wrote:
Hi Fan,
The <outputs> </outputs> block should contain one line for each of the
output files. I believe that you need to name these differently. "output1",
"output2", etc. Or you can add in text and use variables, if you add in the
" label" option.
Others can correct or add to my comments.
Good luck!
Jen
Galaxy team
On 11/8/13 5:41 PM, Jun Fan wrote:
Hi Jen
Thanks for your reply. Yes, it is my own tool.
The outputs element is
<outputs>
<data format="gff3" name="output"/>
</outputs>
In the command element, the three output files are defined as below
$output /$__new_file_path__/primary_${output.id}_samWithPeptides_visible_sam
/$__new_file_path__/primary_${output.id}_longestORFs_visible_fasta
Is there anything wrong here?
Best regards!
Jun
*From:* Jennifer Jackson [mailto:jen@bx.psu.edu <jen(a)bx.psu.edu>]
*Sent:* 09 November 2013 01:05
*To:* Galaxy Dev
*Cc:* Jun Fan
*Subject:* multiple output tool in workflow
Hi Jun,
There is probably a problem with the tool design itself, but that may be
what you are asking how to solve. I wouldn't think this is a problem with
workflows at first pass.
Is this your own tool? Or a tool from the tool shed (the repo developer is
usually the one to make changes, unless you want to try)?
This is the primary tool development wiki, the " <output> tag set" is
where I would double check the tool first.
http://wiki.galaxyproject.org/Admin/Tools/ToolConfigSyntax
I am moving this over to the galaxy-dev(a)bx.psu.edu mailing list since it
is a tool development question.
Jen
Galaxy team
On 11/8/13 10:19 AM, Jun Fan wrote:
Hi all,
I am trying to creating a workflow from history. One of the tool
used generates multiple outputs in the format of gff3, fasta and sam. Gff3
will be visualized in IGV and the fasta file is doing further BLAST
analysis. Now the problem is that the automatically generated workflow does
not connect the having-multiple-output tool and the BLAST tool. I failed
even I tried to connect these two tools in the workflow by hand. I am
guessing this is due to only the main output type (gff3) is recognized in
the workflow. How could I solve this problem?
Best regards and have a nice weekend!
Jun
___________________________________________________________
The Galaxy User list should be used for the discussion of
Galaxy analysis and other features on the public server
at
usegalaxy.org. Please keep all replies on the list by
using "reply all" in your mail client. 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/
--
Jennifer Hillman-Jackson
http://galaxyproject.org
--
Jennifer
Hillman-Jacksonhttp://galaxyproject.org
___________________________________________________________
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/
To search Galaxy mailing lists use the unified search at:
http://galaxyproject.org/search/mailinglists/