High accuracy of genetic discrimination among chicken lines obtained through an individual assignment test

In the current study, we carried out assignment tests applying the Bayesian and distance-based methods, using 20 microsatellite genotypes in four chicken lines. The Bayesian method showed slightly higher performance of assignment than the distance-based method. In the assignment using the Bayesian method, >or=90% accuracy of assignment was attained by using only two of the most heterozygous markers, whereas in the case of the least heterozygous markers, six were needed to reach the same level of accuracy. In the assignment of the most closely related line pair (F(ST) = 0.1736), at least 12 markers selected by random ordering and at least 15 individuals per line were needed to stably obtain high accuracy of assignment (>or=97%), whereas using only six random markers achieved 97-100% of accuracy between the two most distinct lines (F(ST) = 0.3651) without reference to the sample size per line.