I tried to perform a hierarchical island model, which is not possible in Lositan (Fdist2) for the detection loci under selection in Arlequin 3.5. However I was given the feedback from Arlequin that it was impossible because I have only 9 microsatellite loci and at least ten is required! Do you know about this? And how can I perform my outlier detection if I still want this hierarchical island model? Thanks a lot.
If you have used Fdist (or Fdist2), then you have noticed that you need to input the average heterozygosity of your data (in order to construct the simulated samples). Doing it with just 9 markers (or even 10) can lead you to huge errors. Please note that "outlier methods" are "genome scans"... No use to do it with very few markers!!
Hope it helps
Pablo
- Badges
- Science topic