In this paper, the authors demonstrate that if one relies solely on allele counting methods, then there is an increasing risk of incorrectly inferring a mixture of n contributors with a mixture of n − 1 contributors as n increases, and that the number loci in a multiplex can reduce the rate at which the risk increases but it will never remove it.
The effect of masking on the assignment of the number of contributors is assessed for the European Standard Set of loci by simulation. The risk that a two-person mixture presents as single source is assessed as 2.6 × 10−13, a three-person mixture presents as a two-person 6.7 × 10−4 and a four-person mixture presents as a three-person 0.165. (Publisher Abstract Provided)