Distinguishing mixture donors from relatives, illustrated by an example

In a previous publication [1] we have reported on the false positive risks for relatives of donors of DNA mixtures, where the results were presented as averages of many mixtures with the same characteristics (number of donors and dropout probabilities). In this report we give a worked out example for a specific mixture for illustration purposes. We note that the false positive risk for relatives is especially elevated when in the LR calculation the same probability of dropout is used for all unknown donors. Although visual inspection of the mixture profile may help eliminate these false positives, awareness of the possibility that not a donor but a relative of one is found is important, not least when LR-driven database searches are carried out with this, quite widespread, model.