I have an abundance matrix with samples and OTU's present in those samples. I want to know if the samples are different between each other, so I ran a PERMANOVA in R with default options, using the function adonis within the vegan package.
Every time I run it this warning pops up:
'nperm' >= set of all permutations: complete enumeration.
Set of permutations < 'minperm'. Generating entire set.
Someone know what does this mean?
Hi everyone,
I came across with the same warning using adonis.
"I think you're requesting more permutations with adonis (default is 199) than is possible with your number of observations, and that the number of possible permutations is less than the default minimum (5040). When this happens functions adonis calls turn on complete enumeration so that you don't end up with adonis calculating statistics on a set of duplicate permutations."
So Malcolm Baptie
, is that a problem? I mean, does that warning mean that your analysis is not correctly done? I have been trying to reduce the number of permutations, and when I tried with permutations = 100, it only appeared one of them:Set of permutations < 'minperm'. Generating entire set.
Raquel Neri, did you manage to solve that warning?
Should I try running the permute::how and permute::check before running the adonis function to find the best number of permutations that should be used?
Thank you both,
Juan Manuel.
Hi Juan Manuel Ortiz Sanjuán ! I solved that problem by increasing the number of samples I was working with, i.e., I was working with three samples for each group I was testing (water vs soil) and then I added 8 more samples for water and ~10 more samples for soil and the warning stopped showing. I don't know if you can do this? the thing is that I'm working with metagenomic samples, so I searched for more samples available on the internet (NCBI, personal communication, etc.)
- Badges
- Science method