Hi again Kelly Okay - as I suspected all along, this turns out to be was an embarrassingly trivial error on my part (an initial lack of understanding of the different fastq file types, combined with a typo in declaration of the file type name i.e. "fastqcsanger" is not the same as "fastqcssanger"). So I'm feeling a bit red faced - however it has been a good excuse to look into some of the Galaxy source code. Thanks for your help pointing me in the right direction. On a side note: the Galaxy central wiki page on setting up NGS tools for local installations doesn't currently have sections for bfast: https://bitbucket.org/galaxy/galaxy-central/wiki/NGSLocalSetup I'd be happy to add some details there if that would help (I'm pjbriggs on bitbucket, and it seems to allow me edit privileges on the wiki?). Thanks & best wishes Peter On 24/06/11 09:21, Peter Briggs wrote:
Hi Kelly
Thanks for the quick response & clarification, I see now how the wrapper XML should function.
I can confirm that I have the tool_data_table_conf.xml file with the entry for "bfast_indexes". However the indexes only appear in the drop-down when I remove the comma-parsing lines from bfast_wrapper.xml - when I restore them and restart, my drop-down is empty. Hmm. As you say, there must be some other issue ... I'll investigate some more and get back to you.
Thanks again for your help! Best wishes Peter
On 23/06/11 19:46, Kelly Vincent wrote:
Peter,
This loc file is tab-separated, but it's just that one of its columns (the third, which holds the Galaxy formats that the line applies to) can hold comma-separated values. For instance, you could have the line: hg18_standardhg18fastqsanger,fastqcssangerhg18 standard indexes/data/hg18/bfast_index/hg18.fa where these indexes would work for both base- and color-space data. What those two lines of code do is look at the input dataset's extension to see if it matches any of the comma-separated values, and if it does then that line is included in the dropdown box.
So if your file is tab-separated and has the right column values but still isn't showing up, you've got another issue. Do you have a tool_data_table_conf.xml file that contains an entry for "bfast_indexes"? Let us know if you are still having issues.
Thanks, Kelly
On Thu Jun 23, at 12:11 PM, Peter Briggs wrote:
Hello
I'm configuring a personal Galaxy instance with NGS tools (precursor to setting up a local production version) and had some difficulty getting the Bfast mapper (NGS Mapping -> Map with BFAST) to pick up the reference genomes from the bfast_indexes.loc file.
I'm using a hg clone of galaxy-dist, and declared the reference indexes in tab-delimited format, as indicated by the comments in bfast_indexes.loc.sample. However the bfast_wrapper.xml seems to expect some kind of comma-separated value? Anyway patching bfast_wrapper.xml thusly fixed the problem for me:
--- bfast_wrapper.xml.orig2011-06-23 09:56:54.008898765 +0100 +++ bfast_wrapper.xml2011-06-23 10:01:30.048840835 +0100 @@ -62,8 +62,6 @@ <when value="indexed"> <param name="indices" type="select" label="Select a reference genome index set"> <options from_data_table="bfast_indexes"> - <filter type="multiple_splitter" column="2" separator=","/> - <filter type="param_value" column="2" ref="input1" ref_attribute="extension"/> <filter type="sort_by" column="3"/> <validator type="no_options" message="No indexes are available for the selected input dataset"/> </options>
I can imagine that using a different format in bfast_indexes.loc should also work, but tab-delimited seems more consistent with the other loc files.
Apologies if this is a known issue (I couldn't find anything in the GMOD archive, and neither bfast_wrapper.xml or bfast_indexes.loc.sample seem to have changed wrt this issue in Galaxy Central), or if I've missed something really obvious.
Thanks for your help, best wishes
Peter
-- Peter Briggs peter.briggs@manchester.ac.uk <mailto:peter.briggs@manchester.ac.uk> ___________________________________________________________ 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:
-- Peter Briggs peter.briggs@manchester.ac.uk