I'm using Arelquin to look at the genetic structure between a number of different populations. I want to compare the populations by producing pairwise FST values, however I don't know what model for the genetic distances I should use.
The two models provided are Slatkin's distance or Reynold's distance. They both aim to linearize the FST values by transforming them [Slatkin's = FST/(1-FST), Reynold's = -ln(1-FST)].
How do I know if the FST values need to be transformed, and if they do need to be transformed, what model should I use?