qiime version 2023.2
Hi qiime team. Thank you for making ancombc an official plugin.
I ran into an issue with the barplot visualization of differentials.
Used the composition ancombc and composition tabulate to generate differentials
However, the composition da_barplot fails to generate barplots in the qzv file
from qiime2.plugins.composition.visualizers import da_barplot
ancom_diff_abn_barplot = da_barplot(data=ancombc_differentials).visualization.save("ancombc_barplot.qzv")
Thanks for reaching out! Happy to hear that you're using ANCOM-BC in q2-composition.
Could you share what operating system you're using, and how you have QIIME 2 installed? I'm wondering where this error is coming from in the R traceback:
System has not been booted with systemd as init system (PID 1).
Can't operate. Failed to connect to bus: Host is down
Hi Liz. Thanks for the quick feedback.
I install qiime2 in Windows Subsystem for Linux running Ubuntu 22.04
On a broader comment about ancombc (as I have been using the un-official version from Morton in the past GitHub - mortonjt/q2-ancombc: qiime2 plugin for ANCOMBC), can I make a short feature request for the tabulate plugin. Can you please add the optional argument to sort and filter the table by lfc and/or q_val. Thank you very much.
I am also curious how the barplot is handled in this new ancom implementation. My own implementation of differential plotting with Morton's plugin was to place the lfc on the x-axis, the taxa on y-axis and the q_val on the secondary y-axis (sorted from 0 to 0.05), like this:
Thanks for following up and providing those details!
I found what seems to be a pretty helpful thread on AskUbuntu that discusses the error message you're receiving in detail and a few ways to solve it or move forward. Check it out and see if this helps!
This is a great suggestion, and is definitely on our radar! The current version of
tabulate is our first iteration, and I would definitely like to add filtering and sorting capabilities to it as well. We have a lot on our plates for the 2023.5 release, but hopefully this is something we can address over the summer!
This implementation uses the log-fold change on the x-axis, and the feature-id's on the y-axis. You can view an example visualization here for reference:
Thanks Liz. I'll give this a try and provide feedback here. Looking forward to where you take this plugin. It's already great. Thank you for the work you do for the community.
An off-topic reply has been split into a new topic: error using qiime composition da_barplot
Please keep replies on-topic in the future.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.