Error: group_humann2_uniref_abundances_to_go

[subinamehta]

@bebatut
I ran the ASAIM workflow on usegalaxy.eu. The tool group_humann2_uniref_abundances_to_go gets completed but doesn't give an output apart from the header. I looked at the stderr and it showed me the error pasted below.

Error:

Traceback

(most recent call last):
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/bin/map_to_slim.py", line 9, in
from goatools.obo_parser import GODag
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/lib/python2.7/site-packages/goatools-0.6.4-py2.7.egg/goatools/init.py", line 5, in
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/lib/python2.7/site-packages/goatools-0.6.4-py2.7.egg/goatools/go_enrichment.py", line 18, in
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/lib/python2.7/site-packages/fisher/init.py", line 1, in
from cfisher import *
File "numpy.pxd", line 148, in init cfisher (src/cfisher.c:5842)
ValueError: numpy.dtype has the wrong size, try recompiling
Loading table from: /data/dnb02/galaxy_db/files/007/507/dataset_7507949.dat
Treating /data/dnb02/galaxy_db/files/007/507/dataset_7507949.dat as stratified output, e.g. ['UniRef50_A3DEF7', 'g__Clostridium.s__Clostridium_thermocellum']
Loading custom groups file: tmp_data/uniref_go_mapping_output.txt
Loading mapping file from: tmp_data/uniref_go_mapping_output.txt
Original Feature Count: 4844; Grouped 1+ times: 2442 (%50.4); Grouped 2+ times: 1525 (%31.5)
Loading table from: tmp_data/humann2_go_abundances.txt
Treating tmp_data/humann2_go_abundances.txt as stratified output, e.g. ['GO:0000030', 'g__Clostridium.s__Clostridium_thermocellum']
Loading custom groups file: tmp_data/formatted_humman2_go_slim.txt
Loading mapping file from: tmp_data/formatted_humman2_go_slim.txt
Original Feature Count: 516; Grouped 1+ times: 1 (%0.2); Grouped 2+ times: 0 (%0.0)
Traceback (most recent call last):
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/bin/src/format_humann2_output.py", line 90, in
format_humann2_output(args, go_annotations)
File "/usr/local/tools/_conda/envs/mulled-v1-9bc79a813cb75ba2c9ce416588a3693084acbc7c8949cf5154583f312deb663a/bin/src/format_humann2_output.py", line 66, in format_humann2_output
namespace = go_annotations[go_id]["namespace"]
KeyError: 'UNGROUPED'

[subinamehta]

Hi ..do we have an update on the new version for Group Abundance tool!

[jraysajulga]

@bebatut It seems like you fixed this on your stand-alone tool as seen in this commit. The issue here is that it has not been updated on the usegalaxy.eu dependency side. Reinstalling the tool might fix it.

For example:

Prior to 8/4/16, the namespace = go_annotations[go_id]["namespace"] code was on line 66.

humann2_output.py", line 66, in format_humann2_output
namespace = go_annotations[go_id]["namespace"]

Post 8/4/16 (after the remediating commit), the code was bumped up to line 70.

70             namespace = go_annotations[go_id]["namespace"]

The error @subinamehta posted shows an error from the previous version.

On that note... If we reinstall it, should we remove the Galaxy Toolshed packages in the wrapper and use Conda recipes instead? Here are my proposed changes.

[bebatut]

Thanks @jraysajulga for investigating. I did not remember this change...

Yes we should remove the full tool_dependencies.xml and use only the conda recipe.
BTW, did you check if the conda recipe is up to date?

[jraysajulga]

@bebatut. The conda recipes are up-to-date if we update Humann2 to 0.11.2. How about the group_abundances script? The conda recipe doesn't seem to work... Should we keep that portion in the tool_dependencies.xml?

[bebatut]

tool_dependencies.xml file should go away. It is not the recommended way in Galaxy anymore

so we should make the conda recipe work. What is failing for it?

[jraysajulga]

$ conda install -c bioconda group_humann2_uniref_abundances_to_go
Collecting package metadata: done
Solving environment: failed

PackagesNotFoundError: The following packages are not available from current channels:

  - group_humann2_uniref_abundances_to_go

Current channels:

  - https://conda.anaconda.org/bioconda/osx-64
  - https://conda.anaconda.org/bioconda/noarch
  - https://repo.anaconda.com/pkgs/main/osx-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/free/osx-64
  - https://repo.anaconda.com/pkgs/free/noarch
  - https://repo.anaconda.com/pkgs/r/osx-64
  - https://repo.anaconda.com/pkgs/r/noarch
  - https://conda.anaconda.org/conda-forge/osx-64
  - https://conda.anaconda.org/conda-forge/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.

[jraysajulga]

@bebatut I looked more into it. The Conda recipe is still drawing from version 1.2.0 (has the error described above). So, on the ASaiM/group_humann2_uniref_abundances_to_GO GitHub repo, we should publish another version (ex: 1.2.1) that contains the UNGROUPED fix. Then we can update the Conda recipe (meta.yaml) to use the updated source: url: https://github.com/ASaiM/group_humann2_uniref_abundances_to_GO/archive/v1.2.1.zip