Hello,
I used qiime 2 to perform Human gut microbiome investigation.
Did you integrate automatically in the script options to adjust for potential confounders ?
Thanks
Hello Meriem!
Can you post the .qza or .qzv file so we can take a look?
Thanks for the replay.
the qza and qzv files are heavy. Can you indicate with details How do you aim to check in those files the adjustment for potential confounders ?
Ah, I made a mistake.
I thought you had processed data with qiime2 and had done a statistical test, like ANCOM, and were asking if this plugin automatically adjusted for confounded factors.
It sounds like you have raw data and you are thinking ahead about the statistical tests you might perform later.
Perhaps I am still misunderstanding your question...
What is your experimental design?
What confounders do you expect?
I used Permanova test to generate PCO plots for beta diversity analysis and Kruskal wallis test for alpha diversity analysis.
Do you think that those tests adjust for confounded factors like age, sex....?
As i previously mentioned the study is about Human gut microbiome investigation.
2 Likes
No, none of these tests adjust for confounding factors, unless you specify it in the model.
The beta-diversity section of the PD-Mice tutorial may be helpful!
In this example, there are many metadata categories, including cage_id, genotype, and donor.
They use the ADONIS test, testing for the effect of donor while controlling for genotype using this formula: --p-formula genotype+donor
Note how the confounding factor of genotype is included in the model.
Thanks for all clarifications.
Do you know if there is an equivalent of ADONIS TEST in R tool and if so what are the input files ?
Meriem, I have great news; ADONIS was first implemented in R as part of the vegan package.
The Qiime2 plugin uses
vegan::adonis() internally. It's the same R code.
Let us know if you need a hand running the ADONIS test within Qiime2!
Thank you so much for your help @colinbrislawn
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.