Bugfix for bwa_meth data manager

Question: Bugfix for bwa_meth data manager

14 months ago by

Dear all, I found a bug in the data manager for the making bwa-meth indices. I'm not sure if this is of interest but here goes:

Old line 27: data_table_entry = dict(value=args.dbKey, dbkey=args.dbkey, name=args.name, path=args.dbKey)

New and corrected:

data_table_entry = dict(value=args.dbkey, dbkey=args.dbkey, name=args.name, path=args.dbkey)

It needed to be "dbkey" and not "dbKey".

Best regards, Christopher Previti

datamanager bugfix bwa-meth • 457 views

ADD COMMENT • link •

modified 14 months ago • written 14 months ago by christopher.previti • 0

14 months ago by

Jennifer Hillman Jackson ♦ 25k

United States

Jennifer Hillman Jackson ♦ 25k wrote:

Hello,

Thanks for reporting the problem. The code is correct in the GitHub source code repository (here) but that update has not been published out to the Tool Shed repository (here). Data Managers are still in the process of being updated to use Conda (a bigger project), so that may explain the delay in pushing out a single-line bug fix.

I opened a ticket asking the IUC to publish the most current wrapper version (at a minimum, correct this line). Please follow progress here.

Again, thank you! Jen, Galaxy team

ADD COMMENT • link written 14 months ago by Jennifer Hillman Jackson ♦ 25k

14 months ago by

christopher.previti • 0

christopher.previti • 0 wrote:

Thanks for the infos, I'll try it with the updated version. Cheers, Christopher

ADD COMMENT • link written 14 months ago by christopher.previti • 0

14 months ago by

christopher.previti • 0

christopher.previti • 0 wrote:

Dear Jen and the Galaxy team,

This data manager is still throwing an error...

Here's the message:

Dataset generation errors Dataset 2591: bwa-meth index

The Galaxy framework encountered the following error while attempting to run the tool:

Traceback (most recent call last): File "/opt/galaxy/galaxy/lib/galaxy/jobs/runners/__init__.py", line 630, in finish_job job_state.job_wrapper.finish( stdout, stderr, exit_code ) File "/opt/galaxy/galaxy/lib/galaxy/jobs/__init__.py", line 1383, in finish self.tool.exec_after_process( self.queue.app, inp_data, out_data, param_dict, job=job ) File "/opt/galaxy/galaxy/lib/galaxy/tools/__init__.py", line 2234, in exec_after_process data_manager.process_result( out_data ) File "/opt/galaxy/galaxy/lib/galaxy/tools/data_manager/manager.py", line 288, in process_result data_manager_dict[ key ].update( value ) ValueError: dictionary update sequence element #0 has length 6; 2 is required Tool execution generated the following error message: Unable to finish job

I'm running the latest Galaxy version and this was a "standard" hg19 genome that I tried to index. I hope it helps somebody...

Cheers, Christopher

ADD COMMENT • link written 14 months ago by christopher.previti • 0

Hi Chris - This is a python error that comes up when data is passed that doesn't meet the expected format (a key-value pair problem). The version of the DM you are using isn't published yet, and some DMs are known to need more tuning, but we can try to troubleshoot if you want to see if there is a usage problem (versus a tool/wrapper problem).

The alternative is to create an Issue at the tool's Github repository. This will allow direct feedback and fixes for the version you are using.

Was the hg19 genome initially loaded using the data manager that fetches a genome and creates a dbkey? If not, how was the genome initially added to Galaxy? Do other DMs work using it?
Click on rerun icon for the jobs and screenshot the settings entered into the data manager form (both the fetch DM and bwa-meth DM).
Click on the "i" view details for the jobs and screenshot both. It is ok to redact private content.
Screenshot or capture the text of the first few lines of the hg19 fasta. The path to the data can be found on disc if you know where to look already or find the path in the GUI under Admin > Server > Data tables registry > all_fasta (named all_fasta.loc on disc).
Put the screenshots/text someplace public (dropbox, imgur, etc) and share the link(s) back.

ADD REPLY • link modified 14 months ago • written 14 months ago by Jennifer Hillman Jackson ♦ 25k

14 months ago by

christopher.previti • 0

christopher.previti • 0 wrote:

Ok, I think I got it.

I think it was two issues :

Bwa mem had to be in the PATH
The indices that were generated were not put in their proper place. Below you can see a (badly formatted) part of the loc file. I made an index for the chicken genome:

galGal5_PhiX_Lambda galGal5_PhiX_Lambda Chicken Dec 2015 (Gallus_gallus-5.0/galGal5) (galGal5) with PhiX and Lambda phage genomes /opt/galaxy/galaxy/tool-data/galGal5_PhiX_Lambda/bwameth_index/galGal5_PhiX_Lambda/opt/galaxy/galaxy/database/jobs_directory/002/2929/dataset_5878_files/galGal5_PhiX_Lambda.fa

The path to the reference/indices should be:

/opt/galaxy/galaxy/tool-data/galGal5_PhiX_Lambda/bwameth_index/galGal5_PhiX_Lambda/galGal5_PhiX_Lambda.fa

I'm working on bwameth, which also has...issues. Will post as soon as I find a fix.

Cheers, Christopher

ADD COMMENT • link written 14 months ago by christopher.previti • 0

14 months ago by

christopher.previti • 0

christopher.previti • 0 wrote:

Oops small correction, it's only the path in the loc file that is wrong, the files themselves are in the right spot.

ADD COMMENT • link written 14 months ago by christopher.previti • 0

Please log in to add an answer.

Similar posts • Search »