Bug in `fst.hudson` function (?)
Hi,
Thanks for writing this package! I was hoping to use the fst.hudson
function, but in looking at the code noticed the following two lines:
#fst = mean(set.fst[1,],na.rm=T) / mean(set.fst[2,],na.rm=T)
fst = mean(set.fst[1,]/set.fst[2,],na.rm = T)
It appears that this means you are taking the mean of ratios, rather than the ratio of means. Since the Bhatia 2013 paper made the "recommendation that FST be estimated as a ratio of averages." I was hoping to know if I was reading your choice correctly, and if that was a bug or a deliberate decision if so.
Thanks again! -Nigel