I'm doing a retrospective analysis of some genetic data using AMOVA - Analysis of MOlecular VAriance. (Do not mistake it with ANOVA!).
1. Does there exist a closed formula for statistical power for AMOVA test?
2. If not, could anyone advise how to compute statistical power for AMOVA test?
Some additional info.
Data: 7 populations, 3 subpopulations, 485 samples all together. Microsatelite data looks like this - haploid, 5 alleles (0 means allele missing):
Pop1 485 254 125 368 0
Pop1 262 354 125 0 342
...
Pop2 470 342 125 368 0
...
...
Pop7 460 451 213 0 487
...
Population vs. Subpopulation info:
Sub1 Sub2 Sub3
Pop1 6 71 0
Pop2 0 9 0
Pop3 0 86 60
Pop4 0 45 30
Pop5 0 12 18
Pop6 0 87 8
Pop7 0 18 35
Here's the latest R package for AMOVA (not ANOVA) https://grunwaldlab.github.io/Population_Genetics_in_R/AMOVA.html
AMOVA was introduced in 1992, here's the original paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1205020/pdf/ge1312479.pdf
