Haplotype based F statistics
Hazel Perry
Hi all,
I've used stacks to analyse a range of populations, looking for SNPs which
differ between them. For most of this I've used Fst values and then run
outlier analysis which has worked well for the most part. I am also
considering using the haplotype based F statistics in the Phistats file but I
am not completely sure what they all show. I think that the Phist values
are simply a haplotype Fst value calculated using the sequences rather than
allele frequencies. Having looked at the Jost paper describing the Dest
value, my understanding is that this is measure of differentiation between the
populations which is not affected by high levels of He, but as this is being
calculated for tags not SNPs I am assuming that stacks is just treating the
tags as multi-allelic markers? Then the other measure I looked at was the
Fst', but this I am struggling to find anything about. I'm sure Fst' is
widely used and there is literature on it but because search engines don’t like
the ' they only spit out results about Fst which is not helpful!
So can anyone confirm that what I think these measures are (Dest and Phist) is
correct and point me in the direction of finding some information about Fst'?
Hazel
Julian Catchen
This publication has a good description of the different haplotype-based
measures implemented in Stacks:
Bird, C. E., Karl, S. A., & Smouse, P. E. (2011). Detecting and
measuring genetic differentiation. In C. D. Schubart, Phylogeography and
Population Genetics in Crustacea (pp. 31–55). CRC Press.
http://books.google.com/books?id=8abMBQAAQBAJ&lpg=PA31&ots=JAu-5ABf-_&dq=info%3AeN8wV0zJHT8J%3Ascholar.google.com&lr&pg=PA35#v=onepage&q&f=false
Best,
julian