galaxy / galaxy-central
Main development repository for Galaxy. Active development happens here, and this repository is thus intended for those working on Galaxy development. See http://bitbucket.org/galaxy/galaxy-dist/ for a more stable repository intended for end-users.
$ hg clone http://bitbucket.org/galaxy/galaxy-central/
Frequently Asked Questions
Galaxy license
What are the licensing terms for Galaxy?
Galaxy is distributed as open source under the conditions of the Academic Free License.
Limits of data file size
What are the current limits on data file size when using the central Galaxy server?.
There is no restriction on the size of a data file or the amount of storage space available for each user on the central Galaxy server. However, there are practical limits. The central Galaxy server stores many alignments locally, so take advantage of them rather than attempting to retrieve them from external sources. For example, moving the entire set of 28-way alignments from UCSC to Galaxy is both time consuming and impractical. There is a much more efficient way to do this with Galaxy, as it provides access to the large alignments from UCSC through a special tool that allows exactly the intersection you want to be done efficiently: the "Fetch Alignments -> Extract MAF Blocks" tool. Simply select your interval file, along with "Locally cached alignments" and "28-way multiz". See our MAFAnalysis page for more information on how to analyze multiple alignments through Galaxy.
Storage on Galaxy servers
How long will my histories and associated dataset be stored on the central Galaxy server?
When you use Galaxy, any analyses that you perform will result in a "Galaxy history", which consists of history items ( datasets ). This history is displayed in the right panel of your browser. If you are logged into the Galaxy main server, your history / datasets will never be removed, except under these scenarios:
1. If you delete a specific history using the Options link at the top of the history panel, that history and all of its associated datasets will be removed from disk 60 days after you deleted the history.
2. Those specific history items ( datasets ) that you delete from one of your histories by clicking the "X" icon in the history item will be removed from disk after 60 days, but unless you manually delete your history, you will still be able to view the history itself ( only the dataset that you deleted from your history will be removed from disk ).
If you are not logged into our Galaxy main server, the history which results from your analyses will not be associated with a Galaxy login ID ( the history will not be associated with a user ). Any history that is not associated with a user and which has not been altered for over 60 days will be deleted, and all history items ( datasets ) associated with that history will be removed from disk at the same time.
Purging unwanted histories and datasets
I have downloaded and installed a local instance of Galaxy. How can I get rid of unwanted histories and their associated datasets?
There are scripts available in the Galaxy distribution that make this process very simple. See our Purge Histories and Datasets page for details.
Galaxy Python version
What versions of the Python language do you support with your downloadable Galaxy source distribution?
We currently support Python 2.4 to 2.6. When you start up your Galaxy instance, the Python eggs for your platform will be atomatically downloaded for you. See our Eggs page for more information.
Multiple histories
I suspect there must be a way of creating multiple history query sets (e.g., for different projects). Is this possible?
Galaxy allows you to create as many histories as you want, but doing this requires you to have a Galaxy account and be logged in. When logged in, click Options in the upper right corner of the history panel. This will display a list of options in the center work panel. Save a history by providing the history with a name. If a history has more than one item and is saved, there will be an additional option "create a new empty history" which will let you create and save another history.
Connect an external datasource to Galaxy
How do I connect an external datasource to Galaxy?
Adding an external datasource is quite easy, and only requires minimal work on the datasource side. For complete details on how to do this, see our DataSources page.
Adding support for a new data type
How do I implement support for a new data type in Galaxy?
For complete details on how to support a new data type, see our AddingDatatypes page.
Configuring Galaxy so a tool can properly generate an Excel file
One of my tools generates a binary excel file and the mime type is always binary, and the file extension is always "data.ext". This causes problems when the user tries to view ( click the eye icon ) or download ( click the 'save' link ). Changing the file type to "excel" doesn't help, and the user is forced to save the file on his desktop, rename the file to an XLS extension and then open it. Is there a way to force the mime type of the downloaded files ?
In your universe_wsgi.ini file add this line to [galaxy:datatypes]...
xls = galaxy.datatypes.data:Data,application/vnd.ms-excel
...and set the output format of your tool to xls.
Running a local Galaxy mirror on port 80, or through Apache
How can I run a Galaxy mirror alongside my regular web server (i.e. on port 80)?
Via a proxy connection in Apache. Instructions can be found on the Config/ApacheProxy page.
Browser cache
If I encounter some problem when using Galaxy, should I clear my browser cache to attempt to correct the problem? In other words, is Galaxy behavior affected by my browser cache?
Yes, your browser cache definitely affects the behavior of Galaxy. The Galaxy development team often updates the central Galaxy server with new code enhancemetns. Because of this, your browser cache may be using old versions of files and clearing them out is important.
Central Galaxy server or Galaxy source distribution
Should I use the central Galaxy server environment hosted at Penn State or should I download the Galaxy source code and host it locally within my environment?
It depends upon your specifics needs. In general, we advise our customers to use the central Galaxy server, hosted by The Institute for Genomics, Proteomics and Bioinformatics at Penn State University. This environment is freely available to anyone with Internet access. Unlimited disk space is available for storing personal Galaxy histories and data. If you want to manage a local Galaxy environment, the Galaxy source distribution aims to be a "zero configuration", entirely self-contained system that provides a lightweight webserver, an embedded database and a multi-threaded job manager. All tools (and their parameters) can be specified via simple XML based configuration files.
Including HTTP links in a tool's help text
Is there a way to include an http link (eg <a href='foo'>foo</a> in the help text in a tool's xml file?
Response
This is done using a restructured text syntax similar to the following line.
.. _Screencasts: http://www.bx.psu.edu/cgi-bin/trac.cgi/wiki/GopsDesc
There is additional information on restructured text here.
Alignment Tools
Does Galaxy provide tools to concatenate only blocks that are adjacent to each other ( i.e. without any gap in the "projected" species )?
The following alignment tools available in Galaxy enable this:
- Join MAF blocks by species - this tool will take a MAF from your history and fuse MAF blocks which are genomically adjacent on all the specified species producing another MAF file
- Stitch MAF blocks given a set of genomic intervals - this tool uses a set of guide intervals that you provide and creates one fasta alignment block per interval for each desired species.
- Stitch Gene blocks given a set of coding exon intervals - this tool does the same as above, except uses coding region information (encoded in the BED format) to create one fasta alignment block per CCDs.
Interval and BED format
Extract Genomic DNA tool under Fetch sequences section returns one less nucleotide for each of my intervals. Is this a bug?
Genomic intervals (Interval and BED formats) on Galaxy are 0-based, start-inclusive, and end-exclusive, in accordance with the definition of BED format by UCSC. As a result, a BED interval defined as "chr1 100 200" will cover nucleotides from 100 to 199.
For a detailed explanation, please read this article.
This revision is from 2009-11-18 18:15
