PERMANOVA Plots( Y-axis) Interpretation

[AHK]

Hi all,
I am running the Pairwise permanova test.
The group significance plots show values on the y-axis greater than 1.
Given that the value of 1 is that 100% of all the taxa found in one sample (or groups of samples) was found in another; What does it mean when you have values greater than one?
Thank you in advance!

[thermokarst]

Hi @AHK - what beta diversity metric are you using when computing the beta group significance plots?

[AHK]

I am using the follwing metrics bellow (note: XXX is the variable I am investigating)

qiime diversity beta-group-significance \
  --i-distance-matrix core-metrics-results/weighted_unifrac_distance_matrix.qza \
  --m-metadata-file metadata.tsv \
  --m-metadata-column XXX \
  --o-visualization core-metrics-results/weighted-unifrac-XXXX-significance.qzv \
  --p-pairwise

Thank you

[lizgehret]

Hello @AHK,

Would you be willing to share the core-metrics-results/weighted_unifrac_distance_matrix.qza file with us? I know that metadata can be sensitive, but if you are willing to share the distances themselves, we can take a look to see if there's an issue with the calculation of the distances, or with the output visualization.

If you are OK with sharing the distances, you can post them in this thread or send me a Direct Message.

Thank you! :qiime2:

[AHK]

Hi,
Thank you for getting back to me.
This file generated when I ran the whole dataset:
weighted_unifrac_distance_matrix.qza (5.2 MB)
This file generated when I ran a portion of the dataset:
[weighted_unifrac_pcoa_results.qza (218.8 KB)
Thank you!

[lizgehret]

Hi @AHK,

Thanks for providing those files! @colinbrislawn was able to confirm for me that you are in fact seeing weighted UniFrac distances greater than 1 (from the datasets you provided), so I wouldn't worry about any issues with those distance calculations or the output viz you were seeing on your end.

It is actually not unreasonable to see distances greater than 1 - @jwdebelius outlines the reasoning and mathematics behind why this can occur here, for your reference.

Hopefully this helps provide some clarification on your results!

Happy :qiime2:ing!

[AHK]

That helps. Thank you!