Plugin error from taxa: box plot

Hi there!

I run into the problem while creating a box plot after I feature grouped my table.

This is what I do:
qiime feature-table group --i-table FLORA_combo_table.qza \
–p-axis sample \
–m-metadata-file FLORA_combo_mm.txt \
–m-metadata-column ‘Run’ \
–p-mode ‘sum’
–o-grouped-table FLORA_combo_grouped_Run.qza

This works perfectly, I get my .qza file, then follows this:

qiime taxa barplot --i-table FLORA_combo_grouped_Run.qza \
–i-taxonomy taxonomy_FLORA_combo.qza \
–m-metadata-file FLORA_combo_mm.txt \
–o-visualization FLORA_combo_bars_grouped_Run.qzv

I the outcome is:
Plugin error from taxa:

Sample IDs found in the table are missing in the metadata: {‘J347R’, ‘J33PL’, ‘J33PD’, ‘J364V’}.

Debug info has been saved to /var/folders/_b/5xsnx4z919g3_pftfw91wk_m0000gn/T/qiime2-q2cli-err-9__zrvyn.log

I use the same metadata file for grouping, get an artifact, but cannot visualise it!

I cannot understand what do I do wrong, and even worse, I do find some similar topics heres, but Im still lost…

Please help!!!
Thank you,

This is where the problem is - you cannot use the same sample metadata file again, because the feature-table group command has effectively built new sample ids:

You will need to manually construct a new sample metadata file with these new sample ids. It is very likely that most of the old sample metadata won’t make sense now (because you might have multiple column values collapsed down into a single “sample”).

Thank you for a quick response!
But Im still lost…
This is how my metadata looks for grouping!
Screen Shot 2020-11-20 at 17.27.39|690x175

I group by ‘Run’, where J33PD and others are 4 names for 4 runs.
I cannot just simply use those 4 names instead of #Sample ID…
Im sorry but Im so confused here.
Is there any way to know the ‘correct’ sample name to get the box plots?

Thank you!

Perhaps it would be more clear to you if you generated a table summary after grouping:

qiime feature-table summarize \
  --i-table FLORA_combo_grouped_Run.qza \
  --o-visualization FLORA_combo_grouped_Run.qzv

The process of grouping combines multiple samples into one. This in turn effectively produces new samples.

Ha! I see now! :slight_smile:
For how long is this the case? Just curious how much behind am I :smiley: :thinking:

Thanks @thermokarst!!

Have a nice Friday & stay safe!


I don’t understand the question - can you elaborate or rephrase?


Oh sorry. before one could use same meta file for the bar plots.
So I meant since when this has changed i.e. creating new meta file for box plots?

Hi @Kotryna, nothing has changed here - sample metadata has always been required for this visualization. As I mentioned earlier, you are performing a step (feature-table group) that creates all-new samples, which requires all new sample metadata. Hope that helps!

Ok thanks!
It is a treat to have you guys available :slight_smile: