B-diversity, AMOVA, HOMOVA and PCOA

Commands in mothur
1

Hello,

I have some issues regarding how I can present the results of B-diversity analyses.

I computed Unifrac metrics, and according to the results of AMOVA, I have differences between some of my groups.
The results of HOMOVA also tell me that the groups that are different with AMOVA also have heterogenous variance (which is not really a good new…)

The biplot is not very great due to the high variance I guess. I don’t see which groups are supposed to be different. Is there another way to present the B-diversity? Or PCOA/NMDS are ok because I have small p-value with AMOVA?

Also, I am shy to ask, but how can I find out which group has the higher B-diversity, between 2 groups that are different, according to the AMOVA?

Thanks a lot!

2

I prefer PERMANOVA for comparing groups. you can use the R2 as a proxy for variance explained by your grouping variable.

3

Thanks for you reply!

What do you use for PERMANOVA?

4

R adonis in vegan