The usefulness of various polymorphisms in paternity testing. Experience with three southern African populations

Genetic studies have been carried out on 1,059 cases (involving 3,177 blood samples) of disputed paternity. Seventeen polymorphic systems, representing 21 loci, have been used on all the cases and an additional 11 system (12 loci) were used when an exclusion on only one of the system was obtained. The overall rate of exclusion was 34.0% and the data were analysed according to the population from which the case came (white, 'coloured' or black). HLA is the most informative system in all three populations, followed by PGM1, with the rhesus blood group the next most useful in the whites and 'coloured's, whereas ABO is the next most useful in the blacks. The probability of excluding a man falsely accused of paternity was 99.1% in the black, 99.4% in the white and 99.7% in the 'coloured' populations. The data were also used to calculate the probability of paternity in the cases in which an exclusion could not be demonstrated; values of over 98% were found in 98.5% of whites, 98.8% of 'coloureds' and 90% of blacks; a further 9.1% of blacks gave values of 95.1 - 98.0%. The range of genetic markers, representing red cell antigen, red cell enzyme, serum protein and HLA polymorphisms used in this study, meets the requirements of an efficient paternity testing service.