what post hoc method Alpha diversity significance results Kruskal-Wallis (pairwise) test used in Qiime2?

Hi All,
I have a question regarding what post hoc method Alpha diversity significance results Kruskal-Wallis (pairwise) test used in Qiime2? Those papers I read always reported that they used Dunn’s tests as post hoc for Krustal-Wallis.

I have did some search in the forum and get some information and want to confirm with you.

1. FDR method of Benjamini and Hochberg were used for multiple comparisons in the Kruskal-Wallis pairwise test in Qiime2?
2. Qiime2 also use Dunn’s tests as post hoc for Krustal-Wallis pairwise test?

Can you please correct me if I am saying something wrong.
Thank you so much for your help!!

2 Likes

Hi @Lei,

Correct! see here for confirmation

Not quite, Q2 just uses the Kruskal-Wallis test again for pairwise comparisons. I believe the KW test is basically an extension of the Mann-Whitney test for multiple groups, so that may be more of an appropriate comparison than the Dunn’s test. But I would imagine you would get nearly identical results.

3 Likes

Hi @Mehrbod_Estaki,

Thank you for much for your explanation.
I saw some paper used letter to show the significant difference between groups if they do post hoc test. Please find the example below.

My question is can I still use (or it is common) to use different letters to show the significant differences between groups?

image


Thank you so much for your kind help!

HI @Lei,

No problem! Using letters to highlight pairwise differences in plots is totally valid. It is just a matter of preference from yourself and the journal of what approach you use. :+1:

2 Likes

Hi @Mehrbod_Estaki,

Thank you for your confirmation. :+1:

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.