P-values for pairwise FST

Question

[Emy]

Hello everyone,

I've computed FST values for my six populations using plink2, employing the --fst CATPHENO method=wc function. I'm now looking to calculate the p-value using 10,000 bootstrap replicates. Should I use the mean_fst and sd_fst derived from my data, or is there a standard value for mean_fst and sd_fst that should be considered instead?

Could someone please advise me on the correct procedure for calculating p-values for pairwise FST results?

Thanks

[dthorbur]

How did you conduct the analysis? Did you use a windowed approach, is it a single genome-wide value, or calculated only for a specific loci. If using a windowed approach you can use a circular genomic permutation analysis to give a p-value to each outlier window. Method published here. This permutation approach is more robust than a random permutation since it takes the distribution of your outliers into account during each permutation. You can then calculate using a one-tailer p-value from the permutations.