Biom format for MetaCoMET


I am trying to export my feature table into a biom format usable by MetaCoMET. Although I’ve been able to create my BIOM format, the file hangs when I try to load it on MetaCoMET (it finishes processing, but hangs when I view it). I was hoping you fine folks may be able to help me, because I believe it may be a formatting error in the biom file that I’m passing in. In particular, the biom file generated by Qiime 2 looks different from the example file on the MetaCoMET site, so I figured I might have exported it incorrectly. The example biom MetaCoMET provides is biom v 1, so the code that I used exports the feature table and then converts it into json:

#Export to biome v2.1
qiime tools export \
  table.qza \
  --output-dir absolute_bacteria
#Convert HDF5 to json
biom convert \
  -i absolute_bacteria/feature-table.biom \
  -o absolute_table_json.biom \
  --table-type="OTU table" \

Does anyone have any experience or suggestions on properly exporting from Qiime 2 to MetaCoMET?

This export workflow looks good to me — this sounds like an issue on Metacomet’s end and you may want to get in touch with their user support to figure out what’s going wrong. Honestly, it sounds like maybe nothing is actually going wrong — you might just have a really big feature table that’s causing something to crash when you are attempting to view results?

Sorry we can’t be much use on this end! None of the qiime2 developers have experience using metacomet, but maybe other forum users who do can chime in :slight_smile:

if you get in touch with the metacomet developers and they are interested in making their tool more compatible with qiime2 (i.e., by accepting feature table QZA files), tell them to get in touch on the forum and we can help :slight_smile:

Good luck!

Hello! Just to follow-up on this, it was indeed an issue with Metacomet. Basically, their parser did not accept the biom formats, even though the files generated by Qiime seem to be valid against the spec. After a lot of experimentation, I was able to mangle the biom format into the proper format. Basically, they’re very inflexible about null fields in the metadata.

I tried to reach out to the Metacomet people around when I first posted this, and they never got back to me. I guess that’s just how it goes sometimes!


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