Individual-breed assignments in caprine populations using microsatellite DNA analysis

Genetic variability of 18 microsatellites in five Chinese goat breeds was characterized and then individual-breed assignments were implemented on the basis of microsatellite multi-genotypes. Although the low Gst value of 5.7% occurred among these breeds, the accumulated non-parental exclusion probabilities of 99.99% and the correct assignment percentage of 82% demonstrated that microsatellites were feasible for paternity testing and individual-breed assignments in caprine populations. All of loci showed high polymorphism except that MAF35 had only two alleles, and INRA0132 had potential null allele in these breeds. The non-parental exclusion probability per locus varied from 0.081 to 0.534 under case I for paternity testing and ranged from 0.161 to 0.699 under case II, respectively. Both of Excl 1 and Excl 2 were significantly correlated to the genetic diversity measures including genetic heterozygosity, polymorphism information content (PIC) and gene differential coefficients, and that suggested genetic polymorphism of marker provide enough information for microsatellite selection. In addition, the Bayesian method was proven to have superiority of the correct assignment percentage in comparison to gene frequency method and genetic distance methods, with the assumptions of locus-population Hardy–Weinberg Equilibrium and no linkage between loci. The correct assignment percentage were always higher in YC and Boer than other three goat breeds, implying assignment efficiency was positively relative to populations with greater uniformity and low population genetic variability. The correct assignment percentage rose with the increase of number of loci used, and it could reach 71.5% when nine microsatellites with the highest genetic polymorphism were combined together.