ANCOM Analysis?

(Ida Pantoja) #1

Hi all,

I did my microbiome analysis already using QIIME 2 v 2017.9 but now I want to do some statistical analysis to confirm the changes I am seeing in my taxa box plot. I was wondering if ANCOM analysis will help on this or if you can suggest another test I can use in this case in QIIME2 or even any other tool like R. Basically, what I want to know if which changes in the organisms abundances are significantly different or not across my samples.
A would appreciate any suggestion.

Thanks a lot!


Hi, I think you already answered your question - you can use ANCOM in q2 to reveal which features are significantly changed

(Ida Pantoja) #3

Thanks so much! Wanted to be sure!


(Ida Pantoja) #4

Hey again,
In the Volcano Plot, what “W” stands for?



The bigger meaning, the more difference, if I am not mistaking

(Nicholas Bokulich) #6

Check out the original publication for more details, but here’s a great explanation on this forum: Specify W cutoff for anacom?

(Ida Pantoja) #7


I have another question about the Ancom command. This is what I am trying to run:
qiime composition ancom
–i-table MRE-comp-taxonomy_ResistantStarch_Only65_orderedbytimepoints-l7.qza
–m-metadata-file mapping_file_resistant_starch_only65_orderedbytimepoints_09_07_18_ANCOM
–m-metadata-category Diet_Time_Point
–o-visualization MRE-l7-ancom-Diet_Time_Point.qzv

It worked but in the category I want to be more specific for the analysis and selected only six samples of about 33 samples for the “Diet_Time_Point” category. How can I do that? So far, I am getting the analysis for all the samples in that category and the abundance table is big.

I would appreciate your help!



There is a tutorial “Filtering data”, you can found there how to filter your table according different parameters, including samples.

(Ida Pantoja) #9

Thanks! I looked it and is working.

What is happening if I don’t I Volcano Plot? I am doing something wrong?

Unable to generate volcano plot, please check the ANCOM statistical results (above).


(Evan Bolyen) #10

Hey @idapantoja,

You probably need to update your installation, 2017.9 is pretty old at this point. I do remember something about the volcano plot being conditional, but I don’t remember the details, @mortonjt was that the case?

(Jamie Morton) #11

Correct, you may be experiencing an outdated version of bokeh