Is population differentiation underestimated by Fst in practice?
In the early 19th century, Sewall Wright, one of the population genetics founders, proposed the statistic Fst for measuring the genetic differentiation between populations.
The statistic has since been widely applied in evolution, conservation and molecular ecology studies. Recently, however, it is recognized that the statistic could underestimate population differentiation when applied to highly polymorphic microsatellites. I investigated theoretically the sets of conditions when Fst provides unbiased- and under-estimates of differentiation, and proposed to use the regression (r) of single-locus Fst estimate against diversity estimate (Hs) to identity the two cases.
A negative and significant (positive or nonsignificant) r signifies that Fst provides under- (unbiased-) estimation of population differentiation. This project intends to exploit the regression approach to make a retrospective survey of published Fst studies, aiming to address questions such as how often and when Fst is underestimated by microsatellites in practice. The study will lead to a potentially interesting and important paper.