Does anyone have suggestions for how to deal with multiple covariates in a PERMANOVA (e.g. using the adonis function from the 'vegan' package)?
For example, if I'm trying to see the effects of diet, ethnicity, and disease status on a microbial community, how should I deal with the fact that the order of the terms matters? My current approach is to do additive feature selection using R2 as the criterion, but I'm not sure how valid this is.
For example, if I'm trying to see the effects of diet, ethnicity, and disease status on a microbial community, how should I deal with the fact that the order of the terms matters? My current approach is to do additive feature selection using R2 as the criterion, but I'm not sure how valid this is.