dadaria
(Daria Igudesman)
October 25, 2022, 9:29pm
1
Hi QIIME community,
I am attempting to create 2D PCoA plots using QIIME2. I found instructions in a previous thread (below), but they seem to be outdated, as the below commands do not work in my version of QIIME. I am aware that it is possible to hide the axis 3 when viewing plots, but this does not change the appearance of the plot otherwise (including the proportion of the variance explained by each axis), which makes me think that a different technique is needed.
Thank you so much in advance!
Is it available to Make 2D PCoA Plots? - User Support - QIIME 2 Forum
qiime tools export --output-dir exported unweighted_unifrac_pcoa_results.qza
make_2d_plots.py -i exported/ordination.txt -m sample-metadata.tsv -o 2D-plots
Hello Daria,
In newer versions of Qiime2 (get the update here ), you can make a standard Emperor plot, then toggle on 2D mode in the options:
Do you see this option inside your Emperor plots? What version of Qiime2 do you have installed?
Hi @dadaria ,
To add a few more solutions:
Exporting to R or Python:
I need to render 2D PCoA plots with confidence ellipsoids. After going through help forum posts here I found others looking to do the same and roadmap plans from 2017 to add this functionality--it seemed that this had been added. I generated a jackknifed Emperor plot using "qiime diversity beta-rarefaction", which I had hoped would have confidence ellipsoids, but it did not, so I exported the data from this visualization.qzv file using "qiime tools export" and then tried to render the 2D plot us…
You can also directly view PCoAs using Vega to make 2D and customize:
Hi @Parix !
That depends on what visualization you're working with. Many visualizations (e.g. taxa barplot ) have colorbrewer palettes baked in for you to use.
Other visualizations may offer different approaches. (e.g. emperor plots allow editing in Vega - very powerful, but there's a bit of a learning curve)
[image]
Big-picture, it's important to remember that QIIME 2 uses a plugin-based model, where all bioinformatics knowledge is implemented in plugins, rather than in a sing…
A separate python package, dokdo:
This tutorial has gotten so big that apparently the text body exceeded the maximum number of characters one can enter in a post (99,000)! Therefore, I will leave a link to Dokdo API for those who are interested in seeing the most up-to-date tutorial. Please let me know if you have any questions.
Click here to see the original documentation
system
(system)
Closed
November 26, 2022, 9:51am
4
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.