PERMANOVA p-value question

Smoking-unweighted-unifrac-significance.qzv (289.9 KB)

Hi everyone!

When performing PERMANOVA in Qiime2, specifically in a situation with only two groups, which p-value should I report? The first one showing that a statistically significant difference was found or the q-value in the bottom (from the forum, I saw that this would be the corrected p-value). But, with only two groups, no adjustment would be required, correct?

I attached an example. Sometimes, the differences (with two groups) between the first (up) and the second (bottom) p-values are even greater.

Thanks in advance,

1 Like

Correct — and indeed p == q in your example since you have only 1 test.

In this particular case the difference is very small — it looks like it could even be due to something like rounding error differences between the all-group (top) and pairwise (bottom) permanova tests.

So just pick one — there is a negligible difference between the two.

If you have these examples please share.

I hope that helps!

Hi Dr. Bokulich,

Thanks very much for the explanation,

1 Like

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