- galaxy-dev - lists.galaxyproject.org

integrating genome view as a display application
by Marijn Vandevoorde 07 Sep '10

07 Sep '10

Hi all, I'm trying to integrate genomeview as a display application. The configuration of the display applications seems to be pretty straightforward, but genomeview needs multiple input files. I can of course write a cgi file that will take a bunch of get parameters and use these as input files for the java applet. so on the side of genomeview, there's not really a big problem. however, to integrate it with galaxy, i don't see how to send multiple datasets to genomeview. you can only add the "view with genomeview" link to one dataset (or to all, but all will be viewed seperately). so i came up with the idea of a new datatype, .gvl, the genomeview list. in it, basically, a list of links to the datasets, so galaxy can work with them. i could make a tool, gvl creator, to take several datasets as input and then create a really simple textfile containing one url per line, for each dataset. i think this is the way to go (if anyone has better suggestions, please tell!) but now come the real questions: most important: how do i get the public url of my datasets? all other tools use the filesystem path, can't really find an example. and secondly: any idea how i could maybe take a dynamic number of inputs? i suppose that's not possible. this is not that big a problem. i could combine multiple gvl creators and then a gvl merger or something, no biggie! all help really appreciated. and of course i'll be happy to share everything with whoever is interested afterwards :-) with kind regards, marijn -- Marijn Vandevoorde Tel:+32 (0)9 331 36 95 fax:+32 (0)9 3313809 VIB Department of Plant Systems Biology, Ghent University Technologiepark 927, 9052 Gent, BELGIUM marijn.vandevoorde(a)psb.vib-ugent.be http://www.psb.vib-ugent.be ================================================================== "If you type google into google, you can break the internet" --Jen Barber

1 1

super-FAQ? How to get the file from CLI from a library
by Davide Cittaro 07 Sep '10

07 Sep '10

Hi there, Suppose I have a data library with a file called "Foo"... How can I get the real file associated to "Foo"? I had to copy it somewhere else (without using the web interface). I guess a simple database query is enough... isn't it? Thanks d /* Davide Cittaro Cogentech - Consortium for Genomic Technologies via adamello, 16 20139 Milano Italy tel.: +39(02)574303007 e-mail: davide.cittaro(a)ifom-ieo-campus.it */

2 2

Apache setup
by Andreu Alibés 07 Sep '10

07 Sep '10

Hi, Anyone has a full example of the configuration of Apache and Galaxy? (I mean, more than what is already at http://bitbucket.org/galaxy/galaxy-central/wiki/Config/ApacheProxy) Thanks, Andreu -- -------------------------------------------- *Andreu Alibés, PhD **Bioinformatics Core & EMBL-CRG Systems Biology Unit Center for Genomic Regulation* C/ Dr. Aiguader 88, 08003 Barcelona, Spain Phone: +34 93 316 0202 http://sites.google.com/site/aalibes -- -------------------------------------------- *Andreu Alibés, PhD **Bioinformatics Core & EMBL-CRG Systems Biology Unit Center for Genomic Regulation* C/ Dr. Aiguader 88, 08003 Barcelona, Spain Phone: +34 93 316 0202 http://sites.google.com/site/aalibes

2 1

How to see stdout even when tool is successful
by Branden Timm 03 Sep '10

03 Sep '10

Hello, I'm writing some wrappers for Galaxy and am finding it difficult to debug proper redirection of stderr and stdout in my wrapper subprocesses. When Galaxy detects an error running a tool, it highlights the output files in red and gives you the "bug" icon so that you can view stderr and stdout. However, when Galaxy interprets the tool as having run successfully, I'm unable to find a way to view what was in stdout. Long story short, Galaxy is interpreting my tool to have run successfully even when it hasn't. I'm redirecting stderr to sdtout on my <command> line in the tool conf file so that Galaxy only relies on return codes to detect an error in the tool, not stderr. Is there a way to view stdout when a tool has run "successfully"? Cheers -- Branden Timm Great Lakes Bioenergy Research Center btimm(a)glbrc.wisc.edu

2 2

Re: [galaxy-dev] Galaxy unix user account
by Peter 03 Sep '10

03 Sep '10

On Fri, Sep 3, 2010 at 5:12 PM, Nate Coraor <nate(a)bx.psu.edu> wrote: > > Peter wrote: > >> I take it then that the "normal" way to install Galaxy is to create a >> galaxy user, and put the files under /home/galaxy - if that is what >> most installations do, I can follow suit. > > You don't have to put them in the home directory, we don't do this > for any of our servers, but we do have a home directory configured > so we can log in as the galaxy user when it's convenient to do so. OK - thanks for the clarification. >> I'm unclear how it is relevant, but we do have a cluster available, >> although I'm not sure at this point if we would try and call it from >> Galaxy... there are internal networking constraints which may >> prevent that. > > Most of the cluster schedulers I'm familiar with would require any > users on the cluster to have a home directory defined. I see what you mean now. In this case the web server which will be running Galaxy (where currently the "galaxy" user has no home directory) would not be the same machine as the cluster head node, and thus any cluster jobs would be run under a separate user account anyway (even if it was also called "galaxy"). Regards, Peter

1 0

Galaxy unix user account
by Peter 03 Sep '10

03 Sep '10

Hi all, I'm trying to install Galaxy on CentOS, and doing things under my own user account the basics look fine (sh setup.sh work, sh run.sh works, and I can access the galaxy webpages on localhost). I'm now trying to do this via a new Unix/Linux user account just for Galaxy, as recommended here: http://bitbucket.org/galaxy/galaxy-central/wiki/Config/ProductionServer For security I have deliberately setup a "galaxy" user with no home directory and with login disabled (based on the apache user): sudo /usr/sbin/adduser -s /sbin/nologin -M -l -r -c "Galaxy" galaxy I have downloaded Galaxy using hg, moved the directory to /opt/galaxy-dist and made the new galaxy user the owner of this folder and its contents. Then: sudo -u galaxy sh setup.sh That seems to work, next: sudo -u galaxy sh run.sh This fails as it tries to write to ~/.python-eggs (i.e. my home directory, the galaxy user doesn't have a home directory), so based on the error message's suggestion to set PYTHON_EGG_CACHE what I have tied is editing run.sh to include the following line near the start: export PYTHON_EGG_CACHE="/opt/galaxy-dist/.python-eggs" That seems to work, but is is a good idea? Do you instead recommend giving the galaxy account a home directory, and installing galaxy there? It would make setting environment variables simpler... Thanks, Peter

2 3

Galaxy on Python 2.7
by Peter 03 Sep '10

03 Sep '10

Hello all, I note that Galaxy currently only supports Python 2.4, 2.5 and 2.6. http://bitbucket.org/galaxy/galaxy-central/src/tip/scripts/check_python.py http://bitbucket.org/galaxy/galaxy-dist/src/tip/scripts/check_python.py The last of the Python 2.x series, Python 2.7, was released back in July 2010. Are there any known issues with Galaxy on Python 2.7? Peter P.S. I filed an issue on this: http://bitbucket.org/galaxy/galaxy-central/issue/386/python-27-support

3 3

The job is finished..but not in the history
by amerite 02 Sep '10

02 Sep '10

hi, I'm writing some wrappers to connect galaxy with some scripts/programs. I can't understand why, for some, the job in the history never end. Here an example of the log output <output> 127.0.0.1 - - [20/Aug/2010:14:18:38 +0200] "GET /history HTTP/1.1" 200 - "http://localhost:8080/tool_runner/index" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" galaxy.jobs DEBUG 2010-08-20 14:18:41,524 job 81 dispatched galaxy.jobs.runners.local DEBUG 2010-08-20 14:18:41,694 executing: python /home/extuser/galaxy-dist/tools/myTools/wrapper_pileup.py --ir /home/extuser/raffo/chromosomes/hg19 --im /home/extuser/galaxy-dist/database/files/000/dataset_40.dat --out /home/extuser/galaxy-dist/database/files/000/dataset_87.dat --mm 20 127.0.0.1 - - [20/Aug/2010:14:18:42 +0200] "POST /root/history_item_updates HTTP/1.1" 200 - "http://localhost:8080/history" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" 127.0.0.1 - - [20/Aug/2010:14:18:46 +0200] "POST /root/history_item_updates HTTP/1.1" 200 - "http://localhost:8080/history" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" ... ... ... 127.0.0.1 - - [20/Aug/2010:14:19:18 +0200] "POST /root/history_item_updates HTTP/1.1" 200 - "http://localhost:8080/history" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" galaxy.jobs.runners.local DEBUG 2010-08-20 14:19:20,831 execution finished: python /home/extuser/galaxy-dist/tools/myTools/wrapper_pileup.py --ir /home/extuser/raffo/chromosomes/hg19 --im /home/extuser/galaxy-dist/database/files/000/dataset_40.dat --out /home/extuser/galaxy-dist/database/files/000/dataset_87.dat --mm 20 127.0.0.1 - - [20/Aug/2010:14:19:22 +0200] "POST /root/history_item_updates HTTP/1.1" 200 - "http://localhost:8080/history" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" 127.0.0.1 - - [20/Aug/2010:14:19:27 +0200] "POST /root/history_item_updates HTTP/1.1" 200 - "http://localhost:8080/history" "Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.19) Gecko/2010072022 Iceweasel/3.0.6 (Debian-3.0.6-3)" ... ... ... and so on </output> If i check the "running" dataset clicking the eye i see the final result that I expect.... Then, checking in the table 'dataset' (in universe.sqlite) i obtain 87|running why the flag isn't changed to ok? any ideas? Tnx :) raf

2 2

Adding links to other Galaxy tools in help section of a tool
by Johannes Eichner 02 Sep '10

02 Sep '10

Dear Galaxy development team, I have set up a customized instance of Galaxy at the University of Tuebingen, Germany. I wondered, if it is possible to add links to other Galaxy tools in the help section of a certain tool. This would make sense for tools that are applied successively in a workflow. Is it possible to add links to tools naturally preceeding or succeeding a certain tool in a workflow? This would make the workflow structure more obvious for the user. Kind regards, Johannes -- Dipl.-Inf. (Bioinf.) Johannes Eichner Univ. Tuebingen, WSI-RA, Sand 1, D-72076 Tuebingen, Germany Phone: (+49/0) 7071 / 29 70436 Fax: (+49/0) 7071 / 29 5091

2 1

Tool not showing up
by Branden Timm 01 Sep '10

01 Sep '10

Hi All, I've created a new tool but I'm unable to get it to show up under "My Tools". I've added it to tool_conf.xml, as shown here: http://codepad.org/sJHNzabY And my tool configuration can be viewed here: http://codepad.org/9r4tnNYV The only tool that shows up under My Tools is soap2. I've confirmed that both the tools/macs folder and macs.xml itself are owned by the Galaxy user. I've tried in several different browsers in case it was a caching issue, but that doesn't seem to be it either. Any ideas? Cheers -- Branden Timm Great Lakes Bioenergy Research Center btimm(a)glbrc.wisc.edu

2 3