Problem about emperor plots

Dear Qiime2 masters,
I faced a problem about emperor plots. I could not see the sample dots of the emperor plot files generated by qiime2-2018.4 through view.qiime2.org in chrome in windows 10 system. Please find attached the related qzv and qza files.
The former emperor plots qzv files generated using the old version of qiime2 and different dataset can be visualized. And all the other qzv files generated using this dataset by qiime2-2018.4 work well.
bray_curtis_distance_matrix.qza (162.1 KB)
jaccard_distance_matrix.qza (178.7 KB)
weighted_unifrac_pcoa_results.qza (188.3 KB)
bray_curtis_pcoa_results.qza (211.6 KB)
unweighted_unifrac_pcoa_results.qza (245.6 KB)
unweighted_unifrac_distance_matrix.qza (281.2 KB)
weighted_unifrac_distance_matrix.qza (278.2 KB)
jaccard_pcoa_results.qza (281.4 KB)
jaccard_emperor.qzv (809.0 KB)
unweighted_unifrac_emperor.qzv (809.4 KB)
weighted_unifrac_emperor.qzv (809.4 KB)
bray_curtis_emperor.qzv (808.9 KB)

Could you help me check what the problem is for these emperor plot qzv files? Thanks a lot!

Best regards,
Ganyu

Hey there @agan! I just took a look at some of the provenance from your bray curtis emperor plot and noticed that you ran the command with the custom_axes command set to DaysSinceExperimentStart, which, while not unique to the Moving Pictures tutorial, does look like you maybe copied and pasted it from there? Anyway, then I opened up the metadata.tsv file embedded in it and noticed you have a Metadata column called DaysSinceExperimentStart, but it is made up entirely of zeros. This is why the plots look weird - one of your axes has no variation in it.

A quick note about Metadata in QIIME 2 - there are no “required” columns from the QIIME 2 perspective - this should be all about your study - what are the things relevant to these samples that provide additional context during interpretation? Anyway, I recommend you spend a little time reading the Metadata in QIIME 2 tutorial to familiarize yourself!

Anyway, rerunning the emperor plot step without specifying the zero-variation-column looks like this:

Hope that helps! :t_rex: :qiime2:

Thank you so much for your help, @THERMOKARST!

Best regards,
Ganyu

1 Like

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