Ordination data as boxplot

I performed Procrustes analysis for weighted and unweighted PCoA. I got nice emperor visualization, but now I want to plot it as boxplots. I went to transformed PCoA files and found ordination.txt file inside for

weighted and unweighted unifrac. Is it 5 axes? Can I just choose the ax I want or I need to transform the data? Can I pool the same ax for weighted and unweighted unifrac and plot together (or next to each other) to show the differences?
Thank you in advance

1 Like


The unifrac distance matrix qza file as input to the beta-group-significance plugin command might work.

Here is the documentation for the 2020.2 version. You might need to use the dropdown box on the top left of your browser to select your QIIME2 version if it’s not 2020.2.

I don’t think using the PCoA ordination input will create your desired boxplots.

As for the axes, those represent the variation in your PCoA. From my understanding, the axes with the largest values are the axes in space where your samples show the most variation. Usually those are the axes with the best clusters.

1 Like

Hi, thank you for your reply!

Yes, but I have only ordination data for transformed PCoA files which I obtained from Procrustes analysis and I don’t have distance matrix for this data. So I waned to ask if there is any way to do it from ordination data.