A new paper, Genomic insights into population history and biological adaptation in Oceania, is worth reading. Read it along with Multiple migrations to the Philippines during the last 50,000 years and Multiple Deeply Divergent Denisovan Ancestries in Papuans.
I’m going to sidestep the new inference that Austronesian expansion may predate the movement out of Taiwan. I’ll revisit. Rather, let’s reflect on the Denisovans. There is strong evidence of more than one admixture from this lineage into modern humans. And, multiple papers now support a model where various Southeast Asian groups have several different pulses. Finally, the “Denisovans” have really deep divergence. Way deeper than anything in modern humans. Some of them split right after the west-east Eurasian hominin split.
All this is curious in light of small hominins in the Philippines and Flores, as well as late ‘erectus.’ I think it is likely that some of the Denisovan lineages have ‘super-archaic’ admixture, while some of the gene flow is mediated by highly admixed moderns with high Denisovan load.
Hey Razib, I just wanted to thank you for often posting the publicly available links for losers like me who wouldn’t have access to journals.
Thoughts only having read press coverage and not paper:
1) finding of Denisovan introgression contributing only to immune pheno, whereas Neander introgression affects more diverse traits. OK, but before we go spinning a story about how Denisovans lacked “useful” adaptations outside immune system (or what this might tell is about Denisovans), isn’t this already contradicted? By introgression of high altitude adaptations in Tibetans, of high altitude adaptations in Inuit, and facial morphology in Native Americans. Is this more a limitation of GWAS in Oceanians?
2) I’m still puzzled about why so many of these Maximum Likelihood / SFS models still prefer a later split of Europeans and East Asians after Oceanians split, and then two-layer admixture in East Asians, even considering Denisovan introgression, and some form of “Basal OoA” flow into Europeans. E.g. the OP fig above and this Extended Fig (https://www.nature.com/articles/s41586-021-03236-5/figures/6) with Han and Sardinians. Simple f4 stats and the qpGraph F-statistics trees don’t need this. I don’t think this is resolvable by anything like the “first wave” that the IUP BK were pointing to. So what’s up?
The high-coverage shotgun ancient DNA set from Reich should be useful for parallel examination of both sets of methods.
“I’m still puzzled about why so many of these Maximum Likelihood / SFS models still prefer a later split of Europeans and East Asians after Oceanians split,”
Perhaps Papuan admixture into Lapita people (possibly once in the Philippines and again much more strongly in islands immediately adjacent to Papua New Guinea) could produce this as a methodological artifact?
The Papuan split predates the European-East Asian split by ca. 20,000 years, and Europeans and East Asians lack Papuan admixture, so this may make Oceanians look older and more diverged than they really are, and Papuan admixture in Oceanians is very significant (with many populations having 24%-35% Papuan ancestry). This is a big distance in time and a lot of admixture.
If so, it also suggests that Denisovans, who in the Altai have about 17% Neanderthal admixture and 4% super-archaic admixture may be pulled to look like TMRCA is a little older than it is, although TMRCA of Neanderthals and Denisovans should be about the same, so you really only have 4% introgression pulling them older (about twice the amount of archaic hominin ancestry found in non-Africans without Papuan/Australian aboriginal/Filipino negrito ancestry). So the tug might not be that great.
Hey @matt can you link to the facial morphology paper? I hadn’t heard that about indigenous Americans.
Sure – https://advances.sciencemag.org/content/7/6/eabc6160 . Also Iain Mathieson – https://mobile.twitter.com/mathiesoniain/status/1336091332788424704 .
I sometimes see that Papuans/Indigenous Australians diverge from before both West Eurasians and East Asians diverge from each other, and then other times see that Papuans/Indigenous Australians diverge from East Asians after both had diverged from West Eurasians. On the other hand, the other Onge/Negrito groups seem to be consistently grouped with East Asians after their divergence from West Eurasians.
Does this mean that Papuans/Indigenous Australians are a separate lineage from the other Onge/Negrito groups, or that perhaps they are a mix of a deeply diverged lineage (distinct from both West Eurasians and East Asians) with a later lineage that’s closer to East Asians (i.e. Onge-like)?
“On the other hand, the other Onge/Negrito groups seem to be consistently grouped with East Asians after their divergence from West Eurasians.”
I’m pretty confident that this is largely a function of greater admixture with mainland East Asians.
@ohwilleke
Negritos? Sure. But Onge having east Asian admixture is news to me. How did you arrive at this conclusion?
@DaThang
According to Kale’s (anthrogenica member) graph, Onge are 64% Crown Eurasian + 36% East Eurasian.
East Eurasians are 85% Crown Eurasian + 15% Bacho-Kiro like population.
Tianyuan, Jomon and Boisman_MN are pure East Eurasians in this model.
Which bacho kiro population are you talking about? The initial upper paleolithic? That is a type of crown Eurasian as well and more divergent than ust ishim at that.
Yes the IUP Bacho-Kiro. They are basal to Crown Eurasian.
Related to this topic, the absolutely amazing Neanderthal dna from sediment paper this week (https://science.sciencemag.org/content/early/2021/04/14/science.abf1667) could really help on this line of enquiry. I might be wrong but seems while we can’t expect it to provide too much information about selected variants, provided the right caves (cool? non-acidic?) are about enough, it might be able to confirm which Denisovan populations were where and at what time, and whether they were admixed with AMH or not. (Also hopefully applicable to the questions of African complex mixture/structure and introgressions during earliest expansions of our species within the continent?)
…
@J Khan, does that model then make Onge 95% “Crown Eurasian” (from 64 direct + (36*0.85) East Eurasian)? And is K14 100% “Crown Eurasian” in that model?
I have a theory that y-dna D (along with other lineages) was brought to South East Asia by Paleolithic East Eurasian HGs. It’s expansion was limited to South East Asia somehow, hence it’s absence in South Asia and Sahul/Oceania. Or it was lost while migrating across Wallace line due to strong bottleneck.
I think it’s comparable to the Austroasiatic/Munda migration to South Asia, where the migration was Male mediated with roughly 30% autosomal contribution.
@Matt
In Kale’s model the West Eurasian lineage is 100% Crown Eurasian without any contribution from Bacho-Kiro like population. K14 and Sunghir are pure West Eurasian. Goyet is 84% West Eurasian +16% from Bacho-Kiro like population that contributed to East Eurasians.
@J Khan
In my nomenclature I just refer to IUP Bacho Kiro as outer crowns. So what as this other ‘crown’ supposed so be? The middle crown ust ishim or an inner crown group which the paper has shown to be 61% ancestral to tianyuan. I refer to that inner crown as inner crown clade B.
Addendum: I have read your most recent comment about ‘west eurasians’ being crown. In that case I will assume that by crown so far you have been referring to what I call inner crown. I refer to ‘west eurasians’ as inner crown clade A.
In that case kale has been modeling tianyuan as inner crown clade B + IUP bacho kiro just like the paper. But he ended up with very different numbers in comparison to the paper (15 vs 39) even though his model of 16% in goyet did not differ from the paper’s 19% in goyet.
To know what I am talking about look at my imgur link in the first open thread of April 2021 here.
Anyway, even though kale’s numbers disagree with the paper (especially when it comes to ‘east eurasians’) he/she has given us onge’s position on the inner clade B vs IUP bachelor kiro paradigm. It seems that onge has less iup bacho kiro then. So if tianyuan and onge are on an inner crown clade B to IUP outer crown cline, then tianyuan is closer to the IUP end of the cline and onge is closer to the inner clade B end. Maybe all the talk of a han to one cline was inadvertently referring to this cline
@DaThang
Kale’s model: https://imgur.com/a/cbBEUyH
My interpretation using your nomenclature:
Outer crown splits from inner crown leading to outer crown A (IUP BK) and outer crown B (the lineage that contributes to Eurasians). Outer crown A and B share some drift before splitting.
Inner crown splits into Inner Crown SE, Inner Crown EE, Inner Crown WE and Ust’Ishim. None of them share any significant drift with each other. Ust’Ishim goes extinct like IUP BK.
East Eurasian = 100% Inner Crown EE + 15% Outer Crown B, followed by bottleneck and shared drift before splitting into East Eurasian branches
West Eurasian = 100% Inner Crown WE, followed by bottleneck and shared drift before splitting into West Eurasian branches
Onge (South Eurasian) = 64% Inner Crown SE + 36% East Eurasian branch that is closest to Jomon, followed by bottleneck and drift
I think the cline you are talking about is better referred to as Jomon-like to ASE (ancient south eurasian) cline, with the pure south eurasian being 100% Inner Crown SE derived with no Jomon-like ancestry. All Onge and Negrito should be on this cline, closer to the ASE end.
The way I defined this is based on two thing:
-Ust Ishim being middle crown, everything is defined by being inside or outside of the Ust Ishim split.
and
-Neanderthal mixture event, outlining the outer limit of outer crowns
As such, Ust Ishim will be a middle crown, not an inner crown (though more inner than outers obviously from a relative point if view, but still outer relative to the inner crown which is defined as being inside/after the Ust Ishim split).
>I think the cline you are talking about is better referred to as Jomon-like to ASE (ancient south eurasian) cline, with the pure south eurasian being 100% Inner Crown SE derived with no Jomon-like ancestry. All Onge and Negrito should be on this cline, closer to the ASE end.
When I said han to onge cline, I was using the words from some old papers, like the minor ‘east Eurasian’ component in ANE being described as on a han-onge cline a few years ago. What I mean is that this observation might have been accidentally based on an actual cline, which exists between not han and onge but more truly between inner clade B (onge is closer to this) and outer crown (tianyuan is closer to this).
Kale results look different from the paper’s. In kale’s diagram, 85% of the ancestry comes from something more divergent than ust ishim. Basically ust ishim is places between the ‘west Eurasian’ inner crown group and the ‘east eurasian’ inner crown group. So it is more different than what I thought it would be. The following is a picture of the paper’s result with some additional labelling: https://imgur.com/a/cOpSfR0
The paper basically gives this: there are crown Eurasians more divergent than Ust Ishim, I call them outer crowns. Then there is Ust Ishim which is an important reference for the whole thing and 83% of its ancestry comes from the middle crown reference, and 17% from outer crowns (which also contributes to goyet and tianyuan). Then after the middle crown split, there is a common inner crown zone, this could have a bunch of branches but there is only 1 relevant branch which leads to the common ancestor of ‘west Eurasians’ and ‘east Eurasians’. This inner crown branch splits into clade A (‘west Eurasian’) and clade B (‘east Eurasian’). Clade A is ancestral to Kostenki, Sunghir, and 81% of Goyet in the paper’s schema and Clade B is ancestral to 61% of Tianyuan. So Tianyuan is a near 3:2 mix of inner crown clade B and some outer crown respectively. It would be interesting to see where Onge falls in this particular schema, which looks different from kale’s result.
Way off topic, but not expecting another Open Thread any time soon …
I read Scheidel’s Escape from Rome after its recommendation here (and here and here: I may be slow to pick up a hint, but as Beatrice said,
I have a good eye …; I can see a church by daylight.
For those who don’t want to read the whole, tedious (because he is building a case, block by block by bloody effin’ block) thing, Scheidel has provided a TL;DR summary
PS: In looking for those links, I came across this from March 2020, which is looking pretty prescient, if anything a tad conservative.