Gneiss on collapsed FeatureTable

Hi everyone,

When using a collapsed table in Gneiss I am getting an error, possibly due to the FeatureID are being replaced with collapsed lineage. Has anyone run into similar issue?

Thanks,
Ali

Hello Alik,

According to the Gneiss developer

Maybe something else is causing the error. Can you post the full error message please?

Colin

Following up on this, @asr17, can you make sure to provide the following:

  • The version of QIIME 2 you are running
  • The exact command(s) you are running (copy and paste please)
  • The exact error received, copy and paste the results when run with the --verbose flag, or else attach the log file referenced at the end of the error message.

Thanks!

Hi @colinbrislawn and @thermokarst,

I am using qiime2-2018.2

Here is my command:
qiime gneiss balance-taxonomy
–i-table Gneiss-comp-L7.qza
–i-tree hierarchy.qza
–i-taxonomy taxonomy1.qza
–p-taxa-level 7
–p-balance-name ‘y0’
–m-metadata-file sample-metadata.txt
–m-metadata-column Tillage
–o-visualization y0_Tillage-L7.qzv
–verbose

Below is the error I am getting:

/home/ali/miniconda3/envs/qiime2-2018.2/lib/python3.5/site-packages/gneiss-0.4.2-py3.5.egg/gneiss/util.py:181: FutureWarning: ‘.reindex_axis’ is deprecated and will be removed in a future version. Use ‘.reindex’ instead.
Plugin error from gneiss:
“None of [frozenset({‘k__Fungi;p__Ascomycota;c__Sordariomycetes;o__Hypocreales;f__Hypocreaceae;g__Trichoderma;s__Trichoderma_asperellum’, ‘k__Fungi;p__Basidiomycota;c__Tremellomycetes;o__Cystofilobasidiales;f__Mrakiaceae;g__Mrakia;s__Mrakia_frigida’, ‘k__Fungi;p__Ascomycota;c__Sordariomycetes;o__Hypocreales;f__Hypocreales_fam_Incertae_sedis;g__Acremonium;s__Acremonium_persicinum’, 'k__Fungi;p__Ascomycota;c__Leotiomycetes;o__Helotiales;f__Helotiales_fam_Incertae_sedis;g__Chalara;s__u …] are in the [index]”

I just copied the first couple lines of the featureTable,
Thanks for your help,
Ali

Hello Ali,

Thanks for posting that! That’s just a Warning, telling the developers to update the script. Warnings are often OK, but Errors meant something went wrong.

Did you get any errors? If you only got this warning, I think the script ran successfully!

Colin

Hi Colin,

I guess the line below the warning is the error If I am not mistaken:

Plugin error from gneiss:
“None of [frozenset({‘k__Fungi;p__Ascomycota;c__Sordariomycetes;o__Hypocreales;f__Hypocreaceae;g__Trichoderma;s__Trichoderma_asperellum’, ‘k__Fungi;p__Basidiomycota;c__Tremellomycetes;o__Cystofilobasidiales;f__Mrakiaceae;g__Mrakia;s__Mrakia_frigida’, ‘k__Fungi;p__Ascomycota;c__Sordariomycetes;o__Hypocreales;f__Hypocreales_fam_Incertae_sedis;g__Acremonium;s__Acremonium_persicinum’, 'k__Fungi;p__Ascomycota;c__Leotiomycetes;o__Helotiales;f__Helotiales_fam_Incertae_sedis;g__Chalara;s__u …] are in the [index]”

In addition I don’t see any output file when I run my script,
Thanks,
Ali

1 Like

If I had to guess, the taxonomy file no longer matches the FeatureTable, since the FeatureTable is collapsed, but the taxonomy file is not.

I don’t see a way how the collapsed table can be passed into balance-taxonomy, since balance-taxonomy does collapsing internally.

So for now, you can run the regression on the collapsed table – but you are really one your own at that point.

Is there a reason why you need to collapse the taxonomy beforehand?

Hi @mortonjt,

Collapsing beforehand would merge many singletons features into the same taxon labels thus eliminating the need to exclude them before running Gneiss. However, I still filter any remaining singletons features from the collapsed table afterwards.
If the balance-taxonomy collapses internally why sometimes the same taxon can be found in both the numerator and denominator?
Best,
Ali

When I was referring to collapsing before, I was just referring to annotating the features with higher levels of taxa. Sorry for potential confusion.

Note it should be ok to collapse before running the regression. However, this isn’t expected to work with balance-taxonomy, since the inputs are drastically different. Might be able to hack the taxonomy input so that those indexes match.

But before you try to hack these commands, I’d first suggest trying to follow the original tutorials, investigate the output and attempt to collapse if you find that your signal is being driving by a low abundance organisms.

Hi @mortonjt,

Following the original tutorial Gneiss tends to overlook a good number of low abundance taxa. Is there a way to hack the taxonomy and collapsed tables so that indexes in both would match.
Thanks,
Ali

Just a follow up, I managed to do round-trip converting between biom and tsv files, I have the indexes now matched across both tables.
Thanks @mortonjt @colinbrislawn @thermokarst for all your help and suggestions,
Best,
Ali

4 Likes

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.