Hi! I'm new to this forum. I'm having trouble understanding whether it is logical to run a permutational dispersion analysis (permdisp) on a Unifrac distance matrix. I'm concerned because for two R programs that run permdisp, they require flags as to what the specific distance matrix is (i.e. bray-curtis, sorenson etc), but unifrac is not an option for these programs. I feel uncomfortable assuming I can call the distance matrix "bray-curtis" for the R code to run, and assume that I'm not violating any statistical assumptions.