The genetic divergence of three populations (Q1098798)

This paper extends the study of genetic divergence in \textit{G. A. Watterson}'s [ibid. 27, 298-317 (1985; Zbl 0572.92016)] two samples drawn from two populations after the original population giving rise to them has split. The author now considers three samples drawn from three populations after the split. The assumptions holding are that (a) mating is random, (b) there are infinitely many alleles at each locus, with all mutants being new alleles, (c) no selective differences exist. The probability distribution of the gene frequencies in the samples is found. Marginal distributions lead to the known single-sample and two-sample results. Watterson had obtained some three-sample results by simulation, and the author compares his formulae with these in special cases. The agreement proves to be excellent. The author also obtains the probability that the three samples are monomorphic for the same allele, and concludes the paper with a numerical example.