ASSIGNMENT OF INDIVIDUALS TO POPULATIONS USING MICROSATELLITES

The article presents two methods for the assignment of an anonymous DN A-sample re populations with, known allele frequencies at a number oi marker loci. The first method calculates the conditional probability f or the sample Eo belong to one of the two possible populations given t he complete marker genotype at several loci. The second method is a si mple test for thr sample to belong to a known population and is based on the distribution of complete marker genotype probabilities. Simulat ion studies were performed for the first method in order to estimate t he expected probability for two populations with random allele frequen cy differences. The results indicate that sis loci with four to five a lleles each are sufficient to obtain an expected probability of more t han 0.05 in 98.1 and 99.8 % of all possible situations. The simulation studies for the second method were used to estimate the power of the test for three different scenarios. The results show char with six loc i with five alleles per locus a power between 0.90 and 0.98 can be ach ieved. However, this is only true if the distribution of allele freque ncies in the known population is not extremely skewed.