Recovering the geographic origin of early modern humans by realistic and spatially explicit simulations

Abstract

Most genetic and archeological evidence argue in favor of a recent and unique origin of modern humans in sub-Saharan Africa, but no attempt has ever been made at quantifying the likelihood of this model, relative to alternative hypotheses of human evolution. In this paper, we investigate the possibility of using multilocus genetic data to correctly infer the geographic origin of humans, and to distinguish between a unique origin (UO) and a multiregional evolution (ME) model. We introduce here an approach based on realistic simulations of the genetic diversity expected after an expansion process of modern humans into the Old World from different possible areas and their comparison to observed data. We find that the geographic origin of the expansion can be correctly recovered provided that a large number of independent markers are used, and that precise information on past demography and potential places of origins is available. In that case, it is also possible to unambiguously distinguish between a unique origin and a multiregional model of human evolution. Application to a real human data set of 377 STR markers tested in 22 populations points toward a unique but surprising North African origin of modern humans. We show that this result could be due to ascertainment bias in favor of markers selected to be polymorphic in Europeans. A new estimation modeling this bias explicitly reveals that East Africa is the most likely place of origin for modern humans.

Haplotype Trees and Modern Human Origins

“…Moreover, a recent analysis claimed that the genetic data \clearly" support the out-of-Africa replacement hypothesis (Ray et al, 2005). The basis of this discrepancy stems from fundamental issues in statistical philosophy, as debated by Knowles and Maddison (2002) and Templeton (2004b).…”

“…Hence, the model shown in Figure 9 lies completely outside the inference universe of Ray et al (2005), so none of their results are relevant to the relative merits of the out-of-Africa replacement model vs. the model shown in Figure 9. In contrast, inferences are data driven in the multilocus nested-clade analysis and not limited by prior beliefs.…”

“…Ray et al (2005) claimed that they used \an approach based on realistic simulations." For example, their simulations of the out-of-Africa replacement hypothesis assume that the total human population size between 900,000-120,000 years ago was 20,000.…”

“…In this regard, Ray et al (2005) did not consider a model with an Acheulean expansion or a model with out-of-Africa expansion coupled with some interbreeding, even though both Fig. 9.…”

“…Moreover, the goodness of fit of the alternative models can be measured in several ways, and it is not always obvious which is the most appropriate. For example, Ray et al (2005) measured the fit of their prior models to the data through the pairwise genetic distance statistic R ST . Interestingly, two species of African ungulates, sampled at the same locations, yield statistically indistinguishable values of a closely related statistic, F ST , but had completely different phylogeographic histories under nested-clade analysis (Templeton, 1998).…”