exported unweighted UNIFRAC distance matrix text file is filled with 0s at end

[Sam_Degregori]

Hi all,

I am trying to export an unweighted unifrac distance matrix qza file from qiime2 into R to run MRM analyses. I exported it as a text file using:

qiime tools export
--input-path core-metrics-results_taxmerged_400/unweighted_unifrac_pcoa_results.qza
--output-path exported

Assuming this is the correct way I then get a text file showing only the rownames, but when I copy and paste them as column names to complete the matrix it appears that the last three columns are all 0's as such:

But they have nonzero values across the rows, shown below, which means the 0s shouldn't be there.

I know you can redo this in R but for some reason I get different values and would like to stick with the qiime2 distance matrix.

Hopefully, this makes sense. Any help would much be appreciated. Thank you,
Sam

[jwdebelius (Justine Debelius)]

Hi @Sam_Degregori,

It looks like you're exporting the PCoA instead of the distance matrix. Is that what you intended?

Best,
Justine

PS: the difference with R is a known issue

very different weighted unifrac values for qiime2 versus phyloseq

[Sam_Degregori]

Sigh...is there a way to forget 4 hours of pointless googling...
Thank you for the catch, I'll report back if I somehow don't get what is expected.

And yes my googling also led me to that forum which is quite interesting as a lot of my previous work was based on the phyloseq unifrac outputs instead of the qiime2 ones. Wonder if it will get fixed

[jwdebelius (Justine Debelius)]

Hi @Sam_Degregori,

We have all been there! Will it help if i confess i spent half of yesterday trying to figure out why an analysis wasn't working before i figure out I hadn't aligned my data the way I though i had?

And i'm not sure about the phyloseq timing? You could ping the github, but I have no clue what's happening there.

Best,
Justine