It is now clear that a straightforward single "Exit from Africa" model is completely untenable, but some articles (like the one discussed in this post) still try to stick to it.
The simplest explanation of a difference of the split times of PNG from Africans relative to other Eurasians from Africans is the presence of an archaic admixture, which this article tries to refute. This difference is documented in many works and it is significant, sometimes very large. For example in the article "Early modern human dispersal from Africa: genomic evidence for multiple waves of migration" the divergence time between East Africa and various Asian populations is estimated at 66.3 kY for West Asia, 96.6 for aboriginal Australia and 107.2 – for Papua New Guinea (more details
the difference between PNG_AFR and ASN_AFR is somewhat larger than in the final version (Fig 1 in the article) but the theoretical model the authors present (Fig 4) the difference is smaller still. That is, with the "bottleneck" tricks they still can't reproduce the actual (larger) difference.
The bottleneck model is not convincing also because such effect would have shown itself for some other indigenous Eurasian populations, some of which surely experienced a strong bottleneck some time in the deep past. Yet for all other populations the split times between Africans are very consistent with statistically insignificant variations. Only PNG and Australian aboriginals show large deviations.
The best explanation of that is the simplest - that they have some small admixture from archaic middle paleolithic population that resided in this region long before the world-wide expansion of the upper paleolithic population.
This corresponds to the model AX in the paper, and the authors admit that this model at relatively low level of admixture is misclassified as the model A (single OoA exit, no archaic admixture).
It is now clear that a straightforward single "Exit from Africa" model is completely untenable, but some articles (like the one discussed in this post) still try to stick to it.
The simplest explanation of a difference of the split times of PNG from Africans relative to other Eurasians from Africans is the presence of an archaic admixture, which this article tries to refute. This difference is documented in many works and it is significant, sometimes very large. For example in the article "Early modern human dispersal from Africa: genomic evidence for multiple waves of migration" the divergence time between East Africa and various Asian populations is estimated at 66.3 kY for West Asia, 96.6 for aboriginal Australia and 107.2 – for Papua New Guinea (more details
https://kirillpankratov.substack.com/p/the-world-after-toba-or-recent-out-700 )
That's just can't be explained by some tricks like a small sample size or a bottleneck effect.
Interestingly, in the preprint of the article, available here:
https://www.biorxiv.org/content/10.1101/2024.09.19.613861v2
the difference between PNG_AFR and ASN_AFR is somewhat larger than in the final version (Fig 1 in the article) but the theoretical model the authors present (Fig 4) the difference is smaller still. That is, with the "bottleneck" tricks they still can't reproduce the actual (larger) difference.
The bottleneck model is not convincing also because such effect would have shown itself for some other indigenous Eurasian populations, some of which surely experienced a strong bottleneck some time in the deep past. Yet for all other populations the split times between Africans are very consistent with statistically insignificant variations. Only PNG and Australian aboriginals show large deviations.
The best explanation of that is the simplest - that they have some small admixture from archaic middle paleolithic population that resided in this region long before the world-wide expansion of the upper paleolithic population.
This corresponds to the model AX in the paper, and the authors admit that this model at relatively low level of admixture is misclassified as the model A (single OoA exit, no archaic admixture).