Many multi-allelic loci in the human genome are almost always due to variation in the number of repeats of a short DNA sequence and a number of investigators have demonstrated that the frequency distribution of such loci differs between human ethnic groups, making them a potentially valuable tool in the analysis of population relationships (Balazs et al., 1989; Bowcock et al., 1994; Deka, Chakraborty & Ferrell, 1991; Flint et al., 1989). However, a number of issues have still to be addressed before the full benefit of multi-allelic loci analysis can be felt. These can be divided into questions about the most appropriate ways to analyse allele frequencies, and technical questions about data collection. This chapter is devoted to the latter.

We discuss here one class of multi-allelic locus, the minisatellite or variable number tandem repeat (VNTR) which usually has unit sizes between 10 to 50 bp with alleles varying in size from 500 bp to 30 kb. The issues surrounding the collection of population allele frequencies at minisatellite loci can be divided into two: there are purely technical issues concerning the most effective way of extracting and recording data for analysis, and there are questions about how allele frequencies at one locus compare with those at another, and indeed with other types of DNA, protein and antigen polymorphism.