Paternity probabilities of biologic fathers and unexcluded, falsely accused men using blood group markers

A frequent legal argument raised in defense of men accused of paternity, but not excluded by genetic tests, is that the probabilities of paternity of falsely accused men are similar to those of biologic fathers. This assertion was tested in a computer simulation experiment that used a database of 15,000 actual paternity cases to provide red cell and HLA phenotypes of mothers, children, and putative fathers. Tests had a combined probability of exclusion of 97.3 percent. Equal numbers of true and false fathers were generated from the data by computer to achieve a prior probability of paternity of 0.5. True fathers' phenotypes were those of unexcluded men from actual cases (Group A) or of mothers from actual cases (Group B) in which paternity was not excluded. The false father group was created by assigning the phenotypes of racially identical men who were selected at random from among cases other than their own. Probabilities of paternity were calculated for the men in each group and were classified into descriptive intervals. The frequency of men in each group was compared in each interval. The frequency distributions of probabilities of paternity for true fathers and unexcluded, falsely accused men (false fathers) were markedly dissimilar.