Unweighted PCoA
Weighted PCoA
My interpretations looking at the unweighted and weighted PCoA are that there is significant correlation with respect to the different bacterial species present in red, blue, and orange samples. Additionally, based on the weighted PCOA it seems there is similar abundance of the species present in all of the samples.
What do you think about the outliers? Thank you!
Hi @macrobiome,
You should use a statistical test to assess this. PERMANOVA or ANOSIM (which you can access through qiime diversity beta-group-significance) are appropriate if your samples are independent.
Thank you @gregcaporaso. The statistical analysis is something I've been struggling with because my samples are paired i.e. I am trying to study the correlation between feces, intestine, and tumor microbiomes from different mice.
I tried to use pairwise-distances but I am unable to do comparison because it requires different treatment (state) group columns which I do not have (I noticed there was some discussion on removing it just like pairwise differences). I thought maybe a Mantel test would help and hence made my other post. Is ANOSIM something I can use in this scenario?
Your insights would be incredibly helpful. Thank you!
I just posted a reply to this and your other message here.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.
