Rudy.Jonker at wur.nl
Mon Feb 7 14:33:57 CET 2011
When using the pairwise.fst method I obtain Fst values between all populations, such as:
P1 P2 P3 P4 P5
P1 0.000000000 0.006336151 0.007354116 0.006710538 0.014397947
P2 0.006336151 0.000000000 0.009318789 0.015509769 0.010039293
P3 0.007354116 0.009318789 0.000000000 0.012273043 0.005988692
P4 0.006710538 0.015509769 0.012273043 0.000000000 0.014574640
P5 0.014397947 0.010039293 0.005988692 0.014574640 0.000000000
To test if these Fst's are significant, I'd like to get a similar table with p-avlues for each combination of populations. If I do gstat.randtest I get one p-value for the entire set, also when I select all three methods (global, within & between).
If I make subsets of my data two do each Fst analysis between two populations separately (which allows testing with gstat.randtest) to obtain a p-value for the fst between these two populations, results in different Fst values than via the method above and is thus not comparable?
Is there any way to get such a p-value matrix?
Thanks in advance,
More information about the adegenet-forum