“…The inferred allele age is highly dependent on the unknown effective population size between the time of mutation and the Bronze age, as well as the dominance coefficient h. Assuming dominance (h = 1), the MAP estimate varied between 3,550 (90% CI 3,200-5,100) and 7,140 (90% CI 5,525-7,650) years BP for population sizes 2N = 10 2 and 2N = 10 5 , respectively, and was about 20% older when assuming additivity (h = 0.5). In line with the low frequency for the Tollense sample, these estimates are at the lower end of the age range [36] (3,280-23,100 BP) or considerably younger [34] (17,500 BP) than recent estimates from modern data. However, we note that all such estimates make strong assumptions about the demographic history, the strength of selection, as well as the origin of the selected allele.…”