Does DEICODE meet the assumption of same "multivariate spread" among groups for PERMANOVA?

[Jigyasa3]

Hey @gregcaporaso

I wanted to ask if all DEICODE transformed data meet the main assumption for PERMANOVA? [https://stackoverflow.com/questions/35410860/permanova-multivariate-spread-among-groups-is-not-similar-to-variance-homogeneit]

I tested it on my data after DEICODE transformation. Only some datasets have the same multivariate spread while others don't.

Do you have any recommendations for the datasets that are not homogeneous after DEICODE transformation?

[cameronmartino]

I am not sure I understand, PERMANOVA should be performed on the distance matrix which is an Aitchison distance, should work fine with PERMANOVA.