I have estimated null allele frequencies using FreeNA (6% on average but up to 30% in some loci). Is there an accepted way of correcting null alleles in GenAlEx? Or should I change one of the alleles in the estimated false homozygotes to '0' (as for missing data)?
I tried coding nulls as "999" and as "0" - they give me slightly different estimates of parameters.
Cheers,
Stan