Jost's D and G'st

[Nadia Faure]

Hello,
Is there a way to calculate Jost's D and G'st pairwise values between populations and their significance?
For Fst, the gl.fst.pop() function is very handy, thank you!! But there don't seem to be any equivalent functions for G'st and Jost's D, right?

I've managed to calculate these two metrics with the mmod package:
pairwise_Gst <- pairwise_Gst_Hedrick(genind)
pairwise_JostD <- pairwise_D(genind)
But I can't get the pvalues with a bootstrap of 10000
Do you have any ideas?

Thanks for your help,
Nadia

[Jose Luis Mijangos]

Hi Nadia,

You can try:

> install.packages("dartR.base")
> library(dartR.base)
> res <- gl.report.fstat(possums.gl)

Check the function documentation to learn about the bootstrapping options. 

Cheers,

Luis 

[Nadia Faure]

Hi Luis,
Thanks a lot for this nice function! Is there a way to get a matrix of pvalues rather than CI?
Cheers,
Nadia

[Jose Luis Mijangos]

Hi Nadia,

See code below:

library(dartR)
t1 <- platypus.gl
t1 <- gl.filter.callrate(t1,threshold = 1)
popNames(t1)
# [1] "SEVERN_ABOVE" "SEVERN_BELOW" "TENTERFIELD"
# testing whether FST between "TENTERFIELD" and "SEVERN_BELOW" is significantly
# different from FST "TENTERFIELD" and "SEVERN_ABOVE"
t2 <- gl.keep.pop(t1,pop.list = c( "TENTERFIELD","SEVERN_BELOW") )
t3 <- gl.t2 <- gl.keep.pop(t1,pop.list = c( "TENTERFIELD","SEVERN_ABOVE") )
# calculating stats
res_t2_tmp <- utils.basic.stats(t2)
res_t3_tmp <- utils.basic.stats(t3)
res_t2 <- res_t2_tmp$perloc$Fstp
res_t3 <- res_t3_tmp$perloc$Fstp
# Identify that the data conforms to a normal distribution using a
# Shapiro–Wilk test of normality
plot(density(res_t2,na.rm = TRUE))
res_shapiro_t2 <- shapiro.test(res_t2)
plot(density(res_t3,na.rm = TRUE))
res_shapiro_t3 <- shapiro.test(res_t3)
# If data is NOT normally distributed use Mann–Whitney U test
res_wilc <- wilcox.test(res_t2,res_t3)
# If data is normally distributed use Student's t-Test
res_ttest <- t.test(res_t2,res_t3)
# testing whether FST is different from 0
res_fst <- gl.fst.pop(t1,nboots = 100)
res_fst$Pvalues

Cheers,

Luis