Next, for each locus in the genome, the genetic diversity was estimated in three ways: (1) based on IBD probabilities with flanking markers; (2) based on expected heterozygosity with flanking markers; (3) the true expected heterozygosity of the marker itself.
In place of actual data on expected heterozygosity for most of these countries, the authors used as the observed diversity the predicted expected heterozygosity from the linear regression of expected heterozygosity in 53 populations with migratory distance from East Africa.
For most of the countries, however, information on expected heterozygosity was unavailable.
Post hoc Scheffé tests on expected heterozygosity also showed significant differences between dsub14 and all other loci (data not shown).
Using the software SPAGeDi 1.1 [57], we computed for each taxon and population two estimators of diversity across loci, one based on allele identity, the expected heterozygosity HE [58], and the other based on allele size, the variance of allele size V [59].
The observed and expected heterozygosities, based on SSR markers, were 0.74 and 0.81, respectively, and these were more than twice higher than the values calculated for SNPs (0.30 and 0.34, respectively).
They are based on three measures of heterozygosity: HT is the expected heterozygosity based on the whole population, Hs is the mean of the expected heterozygosities in each subpopulation (in our case farm) and HI is the observed frequencies of heterozygotes (I = for individuals).
Under the Cornuet and Luikart method, the estimates of expected heterozygosity based on allele frequencies (H E ), and on the number of alleles and sample size (H EQ ) were compared, based on the assumption that the allele diversity is reduced faster than the heterozygosity.
This test is based on the comparison of the logarithm of the ratio between expected heterozygosities obtained for each locus in two populations: Ln RH = Ln [((1/ 1- Hpop1))-1)/((1/ 1- Hpop2))2 -1)].
Powell et al. (1996 b ) compared the expected heterozygosities and the estimated genetic similarities based on different molecular marker types for the evaluation of a set of soybean accessions.
