Estimation of genetic distance and coefficient of gene diversity from single-probe multilocus DNA fingerprinting data

DNA fingerprinting exhibits multilocus genotypes of individuals, detectedby the use of a single multilocus probe. Consequently, population data onDNA fingerprinting do not provide a complete characterization of thegenetic variation in terms of allele-frequency distributions, since neitherthe number of loci nor the locus affiliation of alleles is directlyobservable. Yet DNA fingerprinting has been proved to be a cost-effectivemethod of detecting hypervariable polymorphisms in several organisms, wherethe traditional loci fail to detect enough variation for microevolutionarystudies. In the present paper we demonstrate that the above-mentionedfeatures of DNA fingerprinting data do not cause any serious problem whenthey are used in evolutionary studies. Bias-corrected estimators of Nei'sstandard and minimum genetic distances are derived, and, by an applicationof this theory to data on seven short tandem repeat loci in three majorhuman populations, it is shown that these modified measures of geneticdistances based on DNA fingerprint patterns are quite close to Nei'sdistances based on locus-specific allele frequencies. Empirical as well astheoretical support of the adequacy of such genetic distances from DNAfingerprinting data is also discussed, and it indicates that the technicallimitations of DNA fingerprinting should not deter the use of the methodfor short-term evolutionary studies.