Objectives: The increasing sensitivity of forensic analysis methods allows to investigate less and less amount of biological samples. For samples of low quality or quantity, there are stochastic events that require intensive statistical analysis.
Methods: There are several models how to calculate the probability of a given set of alleles. We have described three of them and compared them to verify their accuracy.
Results: The two models proposed in [1] extend so far the most widely used model by the possibility of dropout and peak areas of individual alleles. The first one is incorrect, while the second model highly improves the possibility of DNA mixture analysis.
Conclusions: We have shown the inaccuracy of one of the recently proposed models. We have added the possibility of determining the dropout probability into the second model, otherwise this model overestimates the probabilities calculated.