Has anyone else noticed odd p-values with UniFrac? Even if I compare to samples that are identical (Libshuff and Parsimony give p-values of 1.000) both weighted and unweighted UniFrac give p-values of 0.
Otherwise, thanks for the great software.
Can you email the tree, name, and group file to mothur.bugs@gmail.com?
Problem solved. In Mothur v 1.14 random must be set to true as follows:
unifrac.weighted(random=T)
my mistake for not noticing the change.