June 05, 2013

Amerindian-like admixture in northern Europe is real

genetiker, a new genome blogger questions the existence of Amerindian-like admixture in Europe. I am generally well-disposed to anyone who tries their hand at analysis of genetic data. On the other hand, if one  accuses me of writing a series of posts "chock-full of stupidity", then there's a good chance I might respond. This should also be useful for anyone wishing to understand the evidence for this admixture.

genetiker proposes that the "Amerindian-like" admixture in North Europeans is misunderstood and can be in fact explained by the existence of "North European-like" admixture in Amerindians. In support of this, he presents the results of an F4 Ratio estimation analysis which suggests that there is "Nordic admixture of the Amerindian populations in the 10 to 20 percent range."

F4 Ratio estimation produces admixture estimates but does not prove the existence of such admixture. The admixture estimates are as good as the relationship proposed for a particular set of populations. If the relationship is nonsensical, so will be the admixture estimates.

According to genetiker, the following relationship holds, with A=Sardinian, B=Orcadian, C=Dai, and O=Yoruba, with X=Amerindians.


But, is the above consistent with the data? The existence of Amerindian-like admixture was argued by Patterson et al. (2012) on the basis of the following F3 statistic (right):

F3(European; Sardinian, Amerindian)

which is signifantly negative for North Europeans. Now, consider the value of this statistic for genetiker's phylogeny.

F3(B = North European; A = Sardinian, X = Amerindian)


In the above figure I color-coded the path from B=North European to A=Sardinian (red) and from B=North European to X=Karitiana (green, if it goes via the supposed "North European" admixture, or blue, if it goes via the "Amerindian" admixture). The value of the F3 statistic is then the weighted sum of the overlap of the red/green and red/blue paths:

F3(B; A, X) = αBZ+(1-α)(BZ+ZW)

where BZ and ZW are drifts along the paths indicated in the figure. This statistic is then always positive, since the common segments in the graph are traversed in the same direction.

genetiker's model is thus falsified by the data: it predicts a positive f3(North European; Sardinian, Amerindian) statistic, but we in fact observe negative ones.

125 comments:

Onur Dincer said...

genetiker is not open to even the most moderate forms of criticism. He refuses to publish my post on the "Amerindians have both Mongoloid and Caucasoid physical features" thread despite the fact that I do not openly criticize him in that post. Here is my censored post:

"Northern Europeans are not only more Amerindian-shifted than southern Europeans, but also more East Asian-shifted than them:

http://dienekes.blogspot.com/2012/08/4-population-test-and-east-eurasian.html

Note that the population used as the East Asian representative in the above 4-population test is the Han Chinese, which is one of the purest Mongoloid populations in the world. So it seems that northern Europeans – even excluding northeastern Europeans – are more Mongoloid-admixed than southern Europeans, unless southern Europeans in general (thus not just non-Basque Iberians and southern Italians) have a Negroid admixture that is lacking or in lesser amounts in northern Europeans."

genetiker also deleted all my previously published (when there was no comment moderation yet) posts in his blog and effectively banned me from his blog (he turned on comment moderation to block my posts).

Unknown said...

Would Mongoloid admixture in Northern Europeans make them seem more Amerind-like, being that Amerinds and Mongoloids are both of north-east Asian ancestry?

Or could it be that Amerinds share a more ancient affinity with the indigenous North-European substrate that was present before the spread of Mesopotamian/Anatolian agriculturalists?

Rokus said...

Obviously the employed phylogenetic tree is not consistent with the data, being anachronistic to both the old age of the admixture and the recent age of the modern northern European cluster.

However, this still doesn't define that "Amerindian-like" admixture in North Europeans isn't geographically equivalent to "North-European-like" admixture in Amerindians.

Simon_W said...

Funny! I think I have an idea which German HBD poster might hide behind this Genetiker character. :D

apostateimpressions said...

D, is it possible that the apparent Amerindian-like admixtue is a relic from before Hg. R, which dominates northern Europe, separated from Hg. Q, which dominates Amerindians?

Or some Q may have been with R from 20,000 or more years ago, before the groups became much genetically distinct.

In either case it would seem to predate the formation of Caucasian and Mongolian races rather than indicate a mixture between the two.

Dienekes said...

However, this still doesn't define that "Amerindian-like" admixture in North Europeans isn't geographically equivalent to "North-European-like" admixture in Amerindians.

North-European-like admixture in Amerindians is inconsistent with the observed F3 statistic.

(Feel free to write down any tree where F3(North European; Amerindian, Sardinian) is negative and there is gene flow from North Europe to the Americas and not from an Amerindian-related population into the ancestors of Northern Europeans).

D, is it possible that the apparent Amerindian-like admixture is a relic from before Hg. R, which dominates northern Europe, separated from Hg. Q, which dominates Amerindians?

Well, we don't know what haplogroup R bearers were originally like. Also, we don't know when they entered Europe (they are first detected post-5kya in ancient DNA). And, finally, we don't know whether Europeans had this "Amerindian"-like admixture before the arrival of the R bearers.

The relationship of R and Q is certainly very intriguing and it'd be interesting to one day find the P-bearing population -presumably somewhere in Central Asia, tens of thousands of years ago- and see what they looked like genetically.

apostateimpressions said...

D, thanks for your answers. Yes I forgot that R may be post-5kya in northern Europe and seems to have spread here in the metal ages. It raises the question of whether R spread the Amerindian-like admixture here. Has any AI-like admixture been detected in any pre-R ancient samples?

Jim said...

Onur, when someone calls a claim that the PIE speakers were "mixed" a "slur", I almost stop reading right there. Your experience there is no surprise in light of that kind of attitude.

"D, is it possible that the apparent Amerindian-like admixture is a relic from before Hg. R, which dominates northern Europe, "

The map in Wkipedia for R1B (I know, but if it's wriong, please just tell me) shows R1B in north western Europe, not "northern europe". it doesn't show much R1b at all in Scandinavia.

apostateimpressions said...

Jim, R1a in NE Europe.

Perhaps R1a and R1b mixed with some Amerindian-like populaton in central Asia before they made their way into Europe. Or perhaps the apparent AI admixture is a pre-R (P) genetic relic that looks like admixture. Perhaps it was already in northern Europe before R migrated there. I really dont know, perhaps some of the more learned among us can find some solution to the puzzle.

Kurti said...

@Onur
"genetiker is not open to even the most moderate forms of criticism. He refuses to publish my post on the "Amerindians have both Mongoloid and Caucasoid physical features" thread despite the fact that I do not openly criticize him in that post"

I had the same impression Amerindians appear like having Caucasian features too.

shenandoah said...

Some of the earliest reports from European explorers to the New World described Native Americans (at least some tribes and individuals) as having Caucasoid physical features. mtDNA X2 (or that cluster anyway, because a lot more research needs to be done on it) has been classified both "Native American" and "Caucasoid". Geneticists have been aware of it for several years, although it doesn't seem to satisfy their preferred preconceived theory about Native American origins. They liked to assume that any Caucasoid genes came to America from Europe, presumably after 1492. But those of us who carry that DNA and who know our family history well enough, know better. Also, this seems to support German Dziebel's theory (as I understand it, anyway) that wholly indigenous, aboriginal Europeans probably either originated or at least co-evolved in North America.

Michael Boblett said...

Dear Dienekes:

In reply to Apostateimpressions, you wrote: “Well, we don't know what haplogroup R bearers were originally like. Also, we don't know when they entered Europe (they are first detected post-5kya in ancient DNA). And, finally, we don't know whether Europeans had this "Amerindian"-like admixture before the arrival of the R bearers.”

Of course, you’ve referred previously to Amerindian-like admixture in hunter-gatherers from Gotland. Cf. your helpful reply to my comment on your October 18, 2012, post “ADMIXTURE tracks Amerindian-like admixture in northern Europe.” (No point quoting it.)

Do you still accept the Gotland interpretations? If so, Amerindian-like admixture before R depends on the relationship between several dates: 1) the arrival time of agriculturalists, 2) the arrival time of IE, 3) the arrival time of R. But if R does NOT precede agriculture, then Amerindian-like admixture must precede R, yes?

I admit I’m biased. I want to push Amerindian-like admixture back as far as possible. That’s because I dislike speculations about pre-Columbian explorers, of which I saw far too much after your October 18 post. I name no names. But our beautiful Welsh and Irish legends deserve a more scholarly treatment, which can produce more useful insights.

Fiend of 9 worlds said...

Kurti, some do and some do not. There's even a few pockets of rh - blood types.


If you can trust discovery channel at all then supposedly some monastic orders seemed to know of the americas. There's tobacco found in ancient egyptian burials as well. Not to mention "rafting" events becoming very common when you have people on actual boats. So I would not be too surprised to learn there's been a trickle here and there for a great deal of time.

Also the innuit are apparently fairly recent newcomers to north america so there's been gene flow of some kind.

If you discard the now more and more silly "clovis first" idea and take out of africa made up numbers with a grain of salt or discard it entirely as I have come to think is wise then it might be much more ancient admixture.

Justin said...

genetiker is a troll.

Nothing more, nothing less.

I would highly suggest that he removes his blog before he is humiliated even further.

Davidski said...

Who is that guy?

terryt said...

"D, is it possible that the apparent Amerindian-like admixtue is a relic from before Hg. R, which dominates northern Europe, separated from Hg. Q, which dominates Amerindians?"

That would be my suspicion too. To me it looks likely that Y-DNAs C3 and N played a considerable part in a northward movement some time after humans had reached America. This movement separated 'Amerindians' from 'northern Europeans'.

pconroy said...

I think the Amerindian-like ancestry in Northern Europeans could stem from a Caucasian population that went East and West.

My mother and I have have 4% of a Lezghin centered component on some tests. My father has Selkup like ancestry, my mother has Chuckhi like and Athabascan like ancestry.

23andMe say my mother has some Native American ancestry.

I've mentioned before that my mother and father - both living in Ireland - have Salishan like ancestry, or Pacific North West.

This article links Salishan (Pacific NW) to Lezghian (North Caucasus), and says the connection is only 5,000 years ago:
http://www.ajol.info/index.php/ijma/article/view/60357

Just last week my mother had a new relative - 3rd-Distant, 11.1 cM and > 1,000 SNP's - with an ethnic Kazakh from Kazakhstan, who has 7 generations of family listed in his family tree...

pconroy said...

BTW, my mother's Kazakh relative has all ancestry from Aktobe, Kazakhstan and is:
mtDNA = R9b1b
Y-DNA = C3c

Fanty said...

@Jim:
Not R1b but R as a whole. Its subclades a and b combined. If the shared autosomal DNA between American natives and Europeans bases on the fact that R (Europeans) and Q (Native Americans) both branch off from P (common anchestor of Europeans and Amerindians) then ALL of R must possess it.

http://www.eupedia.com/images/content/Haplogroup_R-borders.gif

Hmm but as can be seen on that map, R dominates the middle more than the north or the south.

Slumbery said...

apostateimpressions

Even if you take R1b and R1a togheter, there is still no visible correlation with this ancient Amerindian component suggested by the Dienekes results. It reaches the highest level among N1c populations, but even that is not a solid connection, because it has a much bigger area of distribution (than N1c).

Also Dienekes run the globe4 calculator on the Neolithic HG DNA results from Sweden and got Amerindian levels comparable to higest modern levels.(http://dienekes.blogspot.hu/2012/10/ancient-european-dna-assessment-with.html) This suggests the admixture is old (a way older than Bronze Age).

eurologist said...

In either case it would seem to predate the formation of Caucasian and Mongolian races rather than indicate a mixture between the two.

apostateimpressions,

I agree. Given the wide-spread and fairly even admixture in Europeans, my guess is it latest happened close to the start of Gravettian (which from what we know so far started in the northern Pontic region). I envision a second (post-Aurignacian) wave of people from the general North Pakistan and surroundings area merging on the steppes with westward-moving proto-Mongoloids arriving via the Altai. This also matches the timeline of finds in southern Siberia.

In another thread we had some fun with facial merging software and merged extreme Caucasoids (approximating Aurignacians) with Mongoloid faces, and the results looked like modern Caucasoids (think Maria Höfl-Riesch + Mongolian women --> attractive French lady ;)).

I had the same impression Amerindians appear like having Caucasian features too.

Kurti,

Your impression is probably because they are not as fully (late) Mongoloid as NE Asians, and at least part of their heritage is more proto-Mongoloid (i.e., some of them spent 10,000 - 15,000 years in Beringia and missed out on becoming extreme Mongoloids). It's something like Inuit --> Na Dene speakers --> older Native American groups in terms of getting less Mongoloid.

cris said...

@Kurti
Those caucasoid physical features are in reality proto-mongoloid ones and could be seen in similar proto-mongoloid (wich is different and appeared earlier than mongoloid proper) Ainus and Malays

AWood said...

"The map in Wkipedia for R1B (I know, but if it's wriong, please just tell me) shows R1B in north western Europe, not "northern europe". it doesn't show much R1b at all in Scandinavia."

Jim, you don't look at much data do you? R1b is very heavy in western, and southern Norway - consistently in the mid 30% range. Likewise, it shows high 20% - low 30% in Sweden.(2nd most frequent hg after I1) In Denmark we're at 40%+ range. Since when is 30-40% considered low frequency? FYI - the Wiki maps aren't the greatest to go by.

Dienekes said...

Of course, you’ve referred previously to Amerindian-like admixture in hunter-gatherers from Gotland.

Yes, those guys do have it, but they're from the NE periphery of Europe that is "special" with regards to this Amerindian-like admixture.

The distinction I was trying to make is whether this "Amerindian"-like admixture was present in places like France and Iberia where there is no hint of any recent Siberian connections. Basically whether this stuff came pre-Neolithic or post-5kya, since it cannot have come with the Neolithic (obviously, as Gok4 and Otzi don't have it).

Cuah123 said...

Slaps hand to face...guess we are forgetting that Natives were brought back to Europe. 2) We see similar percentages in Mexico. Native mothers and European Fathers.

If these admixture results are from "Americans" of NEU ancestors then I would be suspicious. As they've had plenty of time to admix.

Are these purely NEU subjects? Subjects which have never

shenandoah said...

Agriculture was already present in North America prior to 1492. The Cherokee, who are believed to have been remnants of the Hopewell / Mound cultures practiced it extensively.

Jim said...

"Jim, you don't look at much data do you? R1b is very heavy in western, and southern Norway - consistently in the mid 30% range."

None. That's why I asked, that's what I do when i don't know something, and fortunately someone with manners answered me earlier.

"Someone who defines homosexuality in humans as a disease of the post-neolithic, is not the effort worth to argue with.

Mr. Knows Wehn, it's not uncommon to find ravid homophobia in racials supremacists. It was a feature of the Christian Identity movement and you see the same thing directed at lesbins in the Aztlan movement.

Fanty said...

@Cuah

It couldnt be natives that British, Frensh, Spanish, brought home from their colonies.

Because ancient hunter-gatherer show the Amerindian like admixture aswell. Ancient farmers do not.

As far as I recall it its decreasing counterclockwise in Europe.

NE Europeans (Fin, Balts, Russians, Poles...) have the highest amounts of "amerindian like" admixture, followed by Scandinavians and Brits. Brits having only very SLIGHTLY lower amounts of it than Russians, wich suggests that this is not the same as Sibirian admixture.
Followed by Germans, followed by Frensh.

Next come SW Europeans wich have clear lower levels and the lowest levels of it have Italians and Greeks.

My personal impression of it is, that its not higher in the regions of special haplogroups but the other way around. Its lower in parts of Europe with G, E, T or J Haplogroups.

Those combined look like this:
http://www.eupedia.com/images/content/Haplogroup_EGJT.gif

Wich suggests that it was part of the native european genpool and got thinned out by farmers in some regions.

Eurasian Sensation said...

Why waste time trying to debate with a guy who is obviously some kind of white supremacist? Genetiker's blog is a joke.

Mark D said...

Michael Boblett - "But our beautiful Welsh and Irish legends deserve a more scholarly treatment, which can produce more useful insights."

It appears St Brendan may have forgotten his vows. ;)

Va_Highlander said...

Grognard:

"If you discard the now more and more silly "clovis first" idea..."

Good point.

"Clovis First" is a zombie model, increasingly untenable yet still commonly found providing a framework in which to interpret new data. While I find the above comments about Mongoloids vs Proto-Mongoloids entirely plausible, we might also bear in mind that we know almost nothing about the pre-Clovis population, other than that associated sites -- Topper, Cactus Hill, &c -- are found on the east coast of North America and that other sites are likely submerged on the continental shelf.

Onur Dincer said...

The recent evidence of the EDAR370A mutation suggests otherwise. The mutation introgressed (or originated) and reached 100% fixation in just a small region on the northern Chinese/Inner Mongolia border region. From there populations carrying it spread widely into northern Eurasia, America, southern China and SE Asia. Northern Chinese are therefore far more 'Mongoloid' than are Amerindians. Amerindians are mixed with a Mongoloid element, not a 'proto-Mongoloid' population.

.
.
.

That appears not to be the case. Malays, Ainu and Amerindians are all outside the region where the EDAR370A gene reached fixation by the original selective sweep. As a result the 'Amerindian admixture' element in Europe has no connection whatsoever to any Mongoloid genes.


You are defining Mongoloidness just based on a single marker, Terry? We have genomewide analyses for a reason. The fact that Vologda Russians, whose Mongoloid genetic admixture is unquestionable, lack the EDAR370A mutation demonstrates how unreliable the frequency of that mutation is in detecting the amount of Mongoloid admixture in a population.

pconroy said...

@MarkD,

At no point did Irish monks or priests of the Celtic Church take a vow of celibacy. In fact they often left large numbers of children.

Unknown said...

I have difficulty following a lot of the technical genetics here, as I am just an armchair anthropology geek.

So, in that context, I am trying to figure out what the theory here is. I assume we aren't suggesting a large back-migration directly from America before 5kya. If that's not the case, what does the "Amerindian admixture" data indicate?

Best guess after read all of this is that there was gene flow from the ancestors of "Amerindians" into Europe during (or before) the neolithic...? Or am I way off base here?

terryt said...

"genetiker also deleted all my previously published (when there was no comment moderation yet) posts in his blog and effectively banned me from his blog (he turned on comment moderation to block my posts)".

Congratulaions. Maju has done me the same honour. He didn't like me trying to explain that East Asian have probably moved as much since the Paleolithic as have Europeans. For some reason he is convinced no-one in East Asia has moved anywhere since they first arrived there.

"Best guess after read all of this is that there was gene flow from the ancestors of 'Amerindians' into Europe during (or before) the neolithic...?"

Exactly correct. I doubt anyone seriously claims gene exchange directly between America and Europe until relatively recently.

"You are defining Mongoloidness just based on a single marker, Terry?"

It seems we are fortunate in the case of East Asians in that a single marker does seem to be reponsible for much of what could be considered 'Mongoloidness'. Anyway, whether or not it is 'the' defining gene we can readfily see that its expansion was from North China/Ineer Mongolia. That spread represents something, presumably significant as the level of its presence coincides to a great deal with the level of Mongoloid features in the relevant populations.

"The fact that Vologda Russians, whose Mongoloid genetic admixture is unquestionable, lack the EDAR370A mutation demonstrates how unreliable the frequency of that mutation is in detecting the amount of Mongoloid admixture in a population".

Populations on the margin of the gene's expansion are not completely consistent with the gene. For example the Munda-speaking people have the gene but are not particularly Mongoloid. So at the geographic margins other factors come into play.

terryt said...

"The fact that Vologda Russians, whose Mongoloid genetic admixture is unquestionable, lack the EDAR370A mutation"

I haven't been able to find a single image of a Vologda Russian who looks remotely 'Mongoloid'. Perhaps that explains the lack of the EDAR370A mutation in the region. Perhaps the 'Mongoloid genetic admixture' overall was so small as to leave no visible manifestation of the gene. Or perhaps was so diluted by the time it reached the region that we finish up with the same result. Can you provide some images?

eurologist said...

Wich suggests that it was part of the native european genpool and got thinned out by farmers in some regions.

Fanty,

I agree. Also by metal-ages traders and invaders, and Mediterranean seafarers.

As have stated numerous times before, all points to the beginning of the Gravettian, which also coincided with notable anthropomorphic changes.


Amerindians are mixed with a Mongoloid element, not a 'proto-Mongoloid' population. ... As a result the 'Amerindian admixture' element in Europe has no connection whatsoever to any Mongoloid genes.

Terry,

I agree with Onur. You cannot use one marker to make that statement. Amerindian can be (and are) both Proto-Mongoloid derived and have (somewhat more recent) Mongoloid admixture. The Mongoloid feature set evolved over ~40,000 years (and may have also involved non-AMHs introgression).

It seems we are fortunate in the case of East Asians in that a single marker does seem to be reponsible for much of what could be considered 'Mongoloidness'

I disagree, There are tens of additional anatomical characteristics (flat face, cheek bones, eyes, eye lids, subcutaneous fat, rib cage, pelvic, legs, etc. - some of them showing potential cold adaptation) so far not and likely simply not related to EDAR.

Onur Dincer said...

Congratulaions. Maju has done me the same honour. He didn't like me trying to explain that East Asian have probably moved as much since the Paleolithic as have Europeans. For some reason he is convinced no-one in East Asia has moved anywhere since they first arrived there.

Maju banned me too. Because according to him use of the term hybrid to designate racial mixes is racist.:D

It seems we are fortunate in the case of East Asians in that a single marker does seem to be reponsible for much of what could be considered 'Mongoloidness'. Anyway, whether or not it is 'the' defining gene we can readfily see that its expansion was from North China/Ineer Mongolia. That spread represents something, presumably significant as the level of its presence coincides to a great deal with the level of Mongoloid features in the relevant populations.

Populations on the margin of the gene's expansion are not completely consistent with the gene. For example the Munda-speaking people have the gene but are not particularly Mongoloid. So at the geographic margins other factors come into play.


Amerindians as a whole are high in the EDAR370A mutation frequency. But the EDAR370A mutation is not in fixation in most of full or almost full Mongoloid populations even in Asia. So the EDAR370A mutation frequency cannot be used to detect the proportion (even roughly) of Mongoloid ancestry of a population.

I haven't been able to find a single image of a Vologda Russian who looks remotely 'Mongoloid'. Perhaps that explains the lack of the EDAR370A mutation in the region. Perhaps the 'Mongoloid genetic admixture' overall was so small as to leave no visible manifestation of the gene. Or perhaps was so diluted by the time it reached the region that we finish up with the same result. Can you provide some images?

I wrote "whose Mongoloid genetic admixture is unquestionable", did not write anything about phenotypes.

terryt said...

A bit more in response to Onur's comment, if I may:

"You are defining Mongoloidness just based on a single marker, Terry? We have genomewide analyses for a reason".

The population containing the EDAR370A variation that developed in, and expanded from, the Mongolia/China border region obviously contained many genes other than that single gene. And at least some of those other genes may have originated in that same region, and expanded with the population. But as the population expanded, and mixed with resident populations, the various genes would have been differentially eliminated. Most would not have dissappeared exactly in proportion to the distance travelled. And some would have travelled further than others.

The advantage of the EDAR370A gene for our purpose is that it is responsible for some of the most recognisable characters of what we see as a 'Mongoloid phenotype'. And we can trace its individual expansion pattern. Other genes that originated in the same region could display a fairly different individual expansion pattern. As a result a particular proportion of Mongoloid a-DNA, say 10%, could very easily contain a different 10% of genes present in the original Mongolia/China population. That explains why even though Vologda Russians contain a reasonable Mongoloid genetic admixture they do not have the EDAR370A variant.

The EDAR370A mutation, although diploid, can be treated in exactly the same manner as can any haploid gene. The resulting geographic pattern is just as revealing as is any haplogroup pattern. And let's not forget that each gene, in effect, actually has its own evolutionary history.

Michael Boblett said...

Mark D said...

"Michael Boblett - "But our beautiful Welsh and Irish legends deserve a more scholarly treatment, which can produce more useful insights."

It appears St Brendan may have forgotten his vows. ;)"

Look, I know you were kidding. And I’m not sure if this fits into a blog on genetics. But here’s a little cultural anthropology stuff about St. Brendan: I think it's relevant to this blog as an example of just how little all this pre-Columbian Celtic babble has to do with actual Insular scholarship.

So please bear with me:

Yes, there was a historical man named Brendan. Yes, he went sailing. But his actual voyages were to relatively nearby places, all or most in the British Isles.

The man had guts. He also had a pre-Christian name containing the ancient, arguably pre-IE root “Bran.” This means “crow” or “raven.” And there is a whole literature of travel narratives involving Bran, most notably the voyages of Bran the Blessed, a Welsh figure whose stories precede and help shape the Brendan narratives.

So we can go two ways with this. One is to speculate about lots of real-time voyages involving various guys who all happened to have the same name. The other way is to look at the meanings of these narratives as metaphors for visionary journeys. More probable, given that Crow/Raven was widely regarded among Celts as a psychopompos or guide of souls.

In Brendan’s case, actual voyages and metaphorical voyages got mixed up, that’s all. But the former certainly didn’t take him to New Hampshire. More like Scotland, which was exotic enough. Kind of like the Upper Amazon a century ago.

Me, I prefer the visionary stuff. It’s part of my actual heritage, not something tacked on by sci-fi writers who know nothing about us..

Michael Boblett said...

In a very helpful response to my post, Dienekes said...

''Of course, you’ve referred previously to Amerindian-like admixture in hunter-gatherers from Gotland.' (Quote from me.)

"Yes, those guys do have it, but they're from the NE periphery of Europe that is "special" with regards to this Amerindian-like admixture.

"The distinction I was trying to make is whether this "Amerindian"-like admixture was present in places like France and Iberia where there is no hint of any recent Siberian connections. Basically whether this stuff came pre-Neolithic or post-5kya, since it cannot have come with the Neolithic (obviously, as Gok4 and Otzi don't have it)."


Thanks for the clarification, which certainly addresses my concerns. Duh! (Michael slaps his forehead,) So NE Europe’s Amerindian-like admixture (rather a mouthful, maybe just call it ALA or something) is an artifact of Siberian influence that probably didn’t exist in Franco-Cantabria during the actual Paleolithic – when NE Europe was uninhabited?

Good, good. The Mesolithic’s not just a lot of Cro-Magnons basking in relatively warm weather. Other folks are already wandering in. I like that.

Onur Dincer said...

The advantage of the EDAR370A gene for our purpose is that it is responsible for some of the most recognisable characters of what we see as a 'Mongoloid phenotype'.

But there are many other typical Mongoloid traits, including more recognizable ones among them, that have nothing to do with the EDAR370A mutation. As Eurologist wrote above:

"There are tens of additional anatomical characteristics (flat face, cheek bones, eyes, eye lids, subcutaneous fat, rib cage, pelvic, legs, etc. - some of them showing potential cold adaptation) so far not and likely simply not related to EDAR."

Other genes that originated in the same region could display a fairly different individual expansion pattern. As a result a particular proportion of Mongoloid a-DNA, say 10%, could very easily contain a different 10% of genes present in the original Mongolia/China population. That explains why even though Vologda Russians contain a reasonable Mongoloid genetic admixture they do not have the EDAR370A variant.

So you are beginning to accept that frequencies of the EDAR370A mutation cannot be used to support your statements "Northern Chinese are therefore far more 'Mongoloid' than are Amerindians" and "the 'Amerindian admixture' element in Europe has no connection whatsoever to any Mongoloid genes". Good to see that.

BTW, somewhere in reply to me you wrote:

"Amerindians are mixed with a Mongoloid element, not a 'proto-Mongoloid' population."

I never said Amerindians are Proto-Mongoloid, I said Amerindians are probably less diverged from Proto-Mongoloids than East Eurasians are.

shenandoah said...

"genetiker also deleted all my previously published (when there was no comment moderation yet) posts in his blog and effectively banned me from his blog (he turned on comment moderation to block my posts)".

I'm amused to hear you of all people complaining about censorship, lol.

terryt said...

"Maju banned me too. Because according to him use of the term hybrid to designate racial mixes is racist"

Yes. I remember that. I agreed with you. The term was exactly correct.

"But the EDAR370A mutation is not in fixation in most of full or almost full Mongoloid populations even in Asia. So the EDAR370A mutation frequency cannot be used to detect the proportion (even roughly) of Mongoloid ancestry of a population".

I disagree. Populations in which the EDAR370A mutation is not in fixation are less Mongoloid-looking htna those in which it is at full fixation. Look, for example, at SE Asian populations.

"Amerindian can be (and are) both Proto-Mongoloid derived and have (somewhat more recent) Mongoloid admixture".

I don't think the Mongoloid admixture is at all 'recent'. I'd presume it was carried in with the first immigrants although later ones probably had a highr proportion of the Mongoloid phenotype, including the EDAR mutation.

"There are tens of additional anatomical characteristics (flat face, cheek bones, eyes, eye lids, subcutaneous fat, rib cage, pelvic, legs, etc. - some of them showing potential cold adaptation) so far not and likely simply not related to EDAR".

And at this stage we do not know where those mutations first arose. My money goes on them having first appeared somewhere within the region where the EDAR mutation entered the modern human genome.

"So you are beginning to accept that frequencies of the EDAR370A mutation cannot be used to support your statements 'Northern Chinese are therefore far more "Mongoloid" than are Amerindians' and 'the "Amerindian admixture" element in Europe has no connection whatsoever to any Mongoloid genes'. Good to see that".

That's not what I'm saying at all. I'm saying that not all Mongoloid admixture involves genes for Mongoloid phenotype but the EDAR370A gene is responsible for a good proportion of the Mongoloid phenotype. Away from its centre of origin it is a good indicator of the level of Mongoloid phenotype and so was probably originally connected with other genes responsible for other elements of that phenotype.

Onur Dincer said...

I disagree. Populations in which the EDAR370A mutation is not in fixation are less Mongoloid-looking htna those in which it is at full fixation. Look, for example, at SE Asian populations.

Whatever correlation exists between the frequency of the EDAR370A mutation and the degree of Mongoloidness is very loose and partial so much so that the frequency of the EDAR370A mutation cannot be used as an indicator of the degree of Mongoloidness.

I don't think the Mongoloid admixture is at all 'recent'. I'd presume it was carried in with the first immigrants although later ones probably had a highr proportion of the Mongoloid phenotype, including the EDAR mutation.

I would not call it higher proportion of the Mongoloid phenotype but higher proportion of the neo-Mongoloid phenotype.

That's not what I'm saying at all. I'm saying that not all Mongoloid admixture involves genes for Mongoloid phenotype but the EDAR370A gene is responsible for a good proportion of the Mongoloid phenotype. Away from its centre of origin it is a good indicator of the level of Mongoloid phenotype and so was probably originally connected with other genes responsible for other elements of that phenotype.

I completely disagree with these. You make too much of the EDAR370A mutation and its phenotypic effects. It is not a sine qua non of Mongoloidness (even loosely speaking), be it genetic or phenotypic.

terryt said...

Sorry for yet another comment here but I might need to explain further:

The collection of genes that together produce the 'Mongoloid' phenotype form such a distinctive and characteristic set that it is impossible to believe the collection developed randomly across some huge region. Instead they must have accumulated, and become fixed, within a relatively confined region and later expanded from that region. Where was that 'relatively confined region'?

"There are tens of additional anatomical characteristics (flat face, cheek bones, eyes, eye lids, subcutaneous fat, rib cage, pelvic, legs, etc. - some of them showing potential cold adaptation)"

Exactly. Cold. That must have been north of the Qin Ling Mountains. The region south of those mountains is almost sub-tropical.

So, we have a suite of genes that gave rise to the Mongoloid phenotype having developed in some region north of the Qin Ling Mountains, exactly where we can be sure at least one gene within the set originated (EDAR370A).

"Amerindian can be (and are) both Proto-Mongoloid derived and have (somewhat more recent) Mongoloid admixture".

The Amerindians cannot be said to have been part of any 'Proto-Mongoloid' population that was partway along the path to becoming 'Fully Mongoloid'. They are almost certainly a hybrid (how Maju loves that term!) formed from an eastward-moving 'Caucasian' population from central/West Asia (Y-DNA Q and mt-DNA X) and a northward-moving 'Mongoloid' population from North China/Mongolia (Y-DNA C3 and mt-DNAs A, B, C and D). As time went by the latter set dominated the groups moving into America, representing the '(somewhat more recent) Mongoloid admixture'. And in the later stages it was also joined by Y-DNA N, the combination replacing to some extent what had been an 'original' thread of Y-DNA Q and mt-DNA X across Siberia.

To me that is the only explanation that makes sense of ALL the evidence.

eurologist said...

I don't think the Mongoloid admixture is at all 'recent'. I'd presume it was carried in with the first immigrants although later ones probably had a highr proportion of the Mongoloid phenotype, including the EDAR mutation.

Terry,

Well, if you look at the actual time scales involved, your comment is not too far from calling the first wave of settlers proto-Mongoloid derived.

By some calculations, the first newcomers (which may have been two different populations: a coastal one and an inland one) spent 15,000 to 20,000 years in Beringia before crossing into the Americas. That brings us to before 30,000 - 35,000 years ago, which is (i) in alignment with Siberian and Altai finds, and (ii) corresponds to the Gravettian - IMO the last possible time of admixture of proto-Mongoloids with proto-Caucasians that can explain today's European distribution (and some of the ancient DNA finds).

Na-Dene speakers came later, and when you travel around in the US SW and NW Canada and Alaska, you can immediately tell. In comparison, most other American Natives in the US, Mexico, or south thereof don't have that many outright Mongoloid features.

Lastly, many (but not all) Inuit of course are closest to having a full Mongoloid feature set.

apostateimpressions said...

Some P (R) are Caucasian and some (Q) are Mongolian. Do we really know that Q acquired the Mongolian/ proto-Mongolian phenotype through mixture with east Asians? Could it not be that east Asians originally got those genes from P people? Is it possible that P had genes that could lead to either phenotype and that one appeared more here, the other more there due to relative isolation, environment selection, drift or whatever? I just wonder whether we really can with certainty rule out that possibility.

terryt said...

"your comment is not too far from calling the first wave of settlers proto-Mongoloid derived".

Except that I wouldn't use the term 'proto-Mongoloid. Hybrid? Yes.

"brings us to before 30,000 - 35,000 years ago, which is (i) in alignment with Siberian and Altai finds, and (ii) corresponds to the Gravettian - IMO the last possible time of admixture of proto-Mongoloids with proto-Caucasians that can explain today's European distribution (and some of the ancient DNA finds)".

Yes. But again I would suggest that some population in East Asia was already 'fully Mongoloid'. The Beringia population would be a mix of some sort of 'Gravettian' from the west with members of this eastern Mongoloid population.

"Na-Dene speakers came later, and when you travel around in the US SW and NW Canada and Alaska, you can immediately tell".

I agree. I've seen them myself.

"In comparison, most other American Natives in the US, Mexico, or south thereof don't have that many outright Mongoloid features"

Yes. The 'Caucasoid' element is more apparent than is the 'Mongoloid' element in most North American Natives at least.

terryt said...

"Whatever correlation exists between the frequency of the EDAR370A mutation and the degree of Mongoloidness is very loose and partial"

It's not as 'loose' as you like to imagine. Have you seen the distribution map of the EDAR gene that Razib put up some time back?

"I completely disagree with these. You make too much of the EDAR370A mutation and its phenotypic effects. It is not a sine qua non of Mongoloidness (even loosely speaking), be it genetic or phenotypic".

I think I've answered the problem you try to see here with my comment of a couple of hours later than your own.

Onur Dincer said...

It's not as 'loose' as you like to imagine. Have you seen the distribution map of the EDAR gene that Razib put up some time back?

Yes. I am speaking based on that map.

I think I've answered the problem you try to see here with my comment of a couple of hours later than your own.

I have already provided my explanations for those issues. No need to reiterate.

terryt said...

"I have already provided my explanations for those issues. No need to reiterate".

Well what do you mean by the term 'proto-Mongoloid'? Most usually it seems to refer to a population who lived in a region that later contained people we would today consider to show Mongoloid features. But that covers a huge region and by that definition even the people of Timor could be considered 'proto-Mongoloid'. With the influx of Indonesians to that island (and to West Papua) the people will eventually have a substantial Mongoloid element in their genetic makeup.

Or perhaps the term refers to the group of people who originally lived in the region where the Mongoloid phenotype first appeared, and from where it spread. But we have no idea of where that region was except to suggest it was fairly well to the north of the Yangtze. We don't know who those people were so that makes the term meaningless in that context.

One thing we can be sure of. The fact that the phenotype can be so closely defined by such characteristics as having 'flat face, cheek bones, eyes, eye lids, subcutaneous fat, rib cage, pelvic, legs, etc.' as well as having straight, black hair indicates strongly that the phenotype developed in a resonably confined region before expanding widely. It cannot possibly have originated slowly over the whole of Eastern Eurasia.

"Do we really know that Q acquired the Mongolian/ proto-Mongolian phenotype through mixture with east Asians?"

I can't think of any other possibility.

"Could it not be that east Asians originally got those genes from P people?"

That's unlikely because the set of genes is a regional characteristic so presumably developed somewhere within the region where the set is found today. Possibly in the region where it has been less diluted through admixture with populations not carrying the set of genes.

"Is it possible that P had genes that could lead to either phenotype and that one appeared more here, the other more there due to relative isolation, environment selection, drift or whatever? I just wonder whether we really can with certainty rule out that possibility"

We know that haplotypes are no close indicator at all of aDNA. They are quite capable of introgressing into different populations.

eurologist said...

But again I would suggest that some population in East Asia was already 'fully Mongoloid'.

Terry,

And I think you get a timing problem if you suggest that. AFAIK, there is no evidence of AMHs north and west of, say, Beijing before ~35,000 ya, but plenty just after. In that general area, AMHs appear around 40,000 ya. Judging by other localized populations, a full-fledged Mongoloid phenotype surely required ~20,000 years or so to develop (ancient human admixture may have accelerated this, of course). However, so far, there is no evidence AMHs existed in these colder climates between ~60,000 - 40,000 ya. Thus, if we accept part of the Mongoloid phenotype to point to cold adaptation (body build and subcutaneous fat) and snow adaptation (hair and epicanthic fold), we are running out of time and place for this to have happened.

Now, I admit that admixture with local ancient humans (e.g., Xujiayao people) could have accelerated this, but that is asking for a lot, given no uniparental markers and presumably less than ~3% autosomal admixture.

To me, even with some ancient human admixture, it easier to envision cold-adaptation and formation of the Mongoloid phenotype occurred in the ~40,000 to 20,000 ya time frame, such that Caucasians and Beringians only received (had)proto-Mongoloid characteristics, while there still was plenty of time for full-fledged Mongoloids to cross the Bering Strait before 15,000 ya.

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

Terry,

I do not see any clear evidence at all for your theory of a Garden of Eden (to use your terminology) type of Mongoloid origins. All the major races are probably results of many thousand years of admixing of populations with different phenotypic traits from different regions and the gradual relative convergence of those populations as a result of continual gene flow. Inter-racial differences result from many thousand years of isolations (the longer the isolation, the greater the racial differences).

terryt said...

"Yes. I am speaking based on that map".

Then you will have noticed the proportion of the gene falls off as we move south from its proposed region of origin. This exactly mirrors the gradual change in phenotype as we move south, with a small proportion reaching Austronesian-speaking people in the Pacific. At the margins, though, the simple correlation starts to break down. Some Amerinds have a very high proportion although are not especially 'Mongoloid'. But they must have passed through very cold regions before reaching America and so the gene may have been under positive selection.

"All the major races are probably results of many thousand years of admixing of populations with different phenotypic traits from different regions and the gradual relative convergence of those populations as a result of continual gene flow".

I agree totally. However East Asians (as a whole) have such a well-defined set of features that it is impossible to believe that the set developed simultaneously over the whole of Eastern Eurasia.

"AFAIK, there is no evidence of AMHs north and west of, say, Beijing before ~35,000 ya, but plenty just after. In that general area, AMHs appear around 40,000 ya."

That fits perectly the date suggested for the origin of the EDAR variant in modern humans. Thirty thousand years, from memory. And that date is long before any human had reached anywhere near America so we have plenty of time for that mutation, along with others associated with modern humans in northern China, to have become fully Mongoloid.

"Judging by other localized populations, a full-fledged Mongoloid phenotype surely required ~20,000 years or so to develop (ancient human admixture may have accelerated this, of course)".

Yes. I tend to go for introgression of the DEAR genes rather than mutation within the modern human line. So 20,000 years was probably not required.

"However, so far, there is no evidence AMHs existed in these colder climates between ~60,000 - 40,000 ya."

But there is adequate evidence for some sort of archaic humans right through from the Upper Amur to the Altai from the Lower Paleolithic.

"To me, even with some ancient human admixture, it easier to envision cold-adaptation and formation of the Mongoloid phenotype occurred in the ~40,000 to 20,000 ya time frame, such that Caucasians and Beringians only received (had)proto-Mongoloid characteristics"

I can't at all see how we can have any sort of proto-Mongoloid population spread across any substantial area. The whole population of East (and SE) Asia cannot have all been slowly becoming 'Mongoloid' over a substantial period of time.

terryt said...

Getting back to these 'proto-Mongoloids'. I've seen it claimed that Polynesians are actually just proto-Mongoloids rather than being a product of a hybrid between a Mongoloid people and a Papuan people. On that score we know that at least some considerable north to south movement of the Mongoloid phenotype has occurred. Early Neolithic Vietnamese (such as the Hoabinhian people) were Papuan rather than Mongoloid but these days the Vietnamese are the most Mongoloid-looking people in SE Asia. There was a recent paper on the subject but I can't find it. But it was very much in agreement with these earlier one:

https://www.jstage.jst.go.jp/article/ase/116/2/116_070405/_article

Quote:

"The Man Bac skeletons support the ‘two-layer’ hypothesis in discussions pertaining to the population history of Southeast Asia".

http://s1.zetaboards.com/anthroscape/topic/4646060/1/

Quote:

"In general terms, Southeast Asia is thought to have been originally occupied by indigenous people (sometimes referred to as Australo-Melanesians) that subsequently exchanged genes with immigrants from North and/or East Asia, during the Holocene, leading to the formation of present-day Southeast Asians (Callenfels, 1936; Mijsberg, 1940; Von Koenigswald, 1952; Coon, 1962; Jacob, 1967). More recent studies based on late Pleistocene and early Holocene human remains represented by specimens from Niah Cave in Borneo (Brothwell, 1960; Kennedy, 1977; Barker et al., 2007), Tabon Cave on Palawan Island, Philippines (Fox, 1970; Macintosh, 1978; Dizon et al., 2002), Gua Gunung Runtuh in Peninsular Malaysia (Zuraina, 1994, 2005; Matsumura and Zuraina, 1999) and Moh Khiew Cave in Thailand (Matsumura and Pookajorn, 2005) have provided additional support for the existence of an ‘Australo-Melanesian’ lineage in ancient Southeast Asia".

Onur Dincer said...

Then you will have noticed the proportion of the gene falls off as we move south from its proposed region of origin. This exactly mirrors the gradual change in phenotype as we move south, with a small proportion reaching Austronesian-speaking people in the Pacific. At the margins, though, the simple correlation starts to break down. Some Amerinds have a very high proportion although are not especially 'Mongoloid'. But they must have passed through very cold regions before reaching America and so the gene may have been under positive selection.

It decreases (relative to China) in eastern Siberia and Japan as well, so what? You are splitting hairs. As for the Austronesian-speaking Mongoloids of the Pacific, they live in the margins of the Mongoloid expansion and nearby Australoids and Negritos, therefore it is very normal them to have significant non-Mongoloid admixture.

I agree totally. However East Asians (as a whole) have such a well-defined set of features that it is impossible to believe that the set developed simultaneously over the whole of Eastern Eurasia.

What I am saying is this: different Mongoloid phenotypes likely emerged in different corners of Eastern Eurasia and came together as a result of many thousand years of gene flow in Eastern Eurasia. Holocene expansions in Eastern Eurasia related to agriculture further served the racial homogenization in Eastern Eurasia.

Onur Dincer said...

It decreases (relative to China) in eastern Siberia and Japan as well, so what?

By "it" I mean the frequency of the EDAR370A mutation.

terryt said...

"different Mongoloid phenotypes likely emerged in different corners of Eastern Eurasia"

I can easily accept that the various genes involved may have originated in various regions, even beyond Eastern Eurasia. But they are not going to reach fixation without selection. Without selection the East Eurasian population would have become far more diverse than it is.

"and came together as a result of many thousand years of gene flow in Eastern Eurasia".

But as a combination they must have undergone selection in a particular region. As most of the characteristics look very much adaptations to cold we can surely assume that the region involved was a cold region.

"It decreases (relative to China) in eastern Siberia and Japan as well, so what?"

It (the EDAR370A mutation) decreases as it moves away from its region of origin. That is because it has become progressively mixed into populations that didn't have it until the arrival of the source population.

"As for the Austronesian-speaking Mongoloids of the Pacific, they live in the margins of the Mongoloid expansion and nearby Australoids and Negritos, therefore it is very normal them to have significant non-Mongoloid admixture".

Yes. They live on the margin of the Mongoloid, and the EDAR370A, expansions. As do the Siberians, Japanese, the Munda and the Amerindians. And therefore it is very normal for all these groups to have significant non-Mongoloid admixture'.

"Holocene expansions in Eastern Eurasia related to agriculture further served the racial homogenization in Eastern Eurasia".

I agree with that comment. I would stress 'Holocene' of course. And certainly the southerly portion of that expansion involved the expansion of a Mongoloid phenotype containing the EDAR370A variation over a previous Australoid/Papuan phenotype that lacked the gene. The northward expansion was certainly considerably earlier. Where we disagree is as to any connection between the EDAR370A and Mongoloid expansions. I say they are closely related. You claim no relationship at all. We will have to remain in disagreement on that point.

Onur Dincer said...

Terry,

I did not say there was no connection between the EDAR370A and the Mongoloid expansions, I said the connection is very loose and partial.

We can test our theories. If your theory is correct, then characteristicly Mongoloid traits should be tightly distributed among Mongoloid populations. If my theory is correct, then characteristically Mongoloid traits should be loosely distributed among Mongoloid populations.

terryt said...

One more comment. We can be almost certain that the EDAR variant did not expand from its region of origin as a single gene. Surely it was part of some population movement that contained at least some other genes distinctive to that population, including haplogroups most likely.

terryt said...

"If your theory is correct, then characteristicly Mongoloid traits should be tightly distributed among Mongoloid populations".

Not exactly. If my theory is correct characteristicly Mongoloid traits should become more diffuse as we move away from a centre of expansion. That is exactly what we see as we move south through China, into SE Asia and out into the Pacific. Most accept a genetic difference between South and North China. I believe it is also what we see as we move into America but it seems many disagree.

"If my theory is correct, then characteristically Mongoloid traits should be loosely distributed among Mongoloid populations".

Can you accept, then, that if there is a centre where characteristically Mongoloid traits dominate then your theory would be proved incorrect? As the traits spread out from a centre we would expect tham to become more 'loosely distributed'.

Onur Dincer said...

Terry, you obviously did not understand what I mean by "loosely distributed". By "loosely distributed", I mean the peak of different Mongoloid traits in different regions. If different Mongoloid traits are peaking in different regions, your theory cannot be correct, it instead means that my theory is the correct one.

terryt said...

" If different Mongoloid traits are peaking in different regions, your theory cannot be correct"

So how widespread are the 'different regions' where 'different Mongoloid traits' peak? Which traits very far removed from the northeast Tibet/northern China/Mongolia border region?

Anonymous said...

This topic is only ever discussed when I'm absent from the internet, and regrettably, I have to chip in 67 comments late. I just scanned the first few comments, so maybe this has already been addressed (though, from past experience, I doubt it).

The latest findings (Loh et al.) found very consistent levels of 'ancient North Eurasian', where significant recent admixture is absent, across all studied modern West Eurasian populations. Orcadians even obtained a lower midpoint than Basques, Tuscans and Italians. The French also obtained a lower midpoint than the Tuscan and Italian sample.

In said latest paper, f4 statistics showed Basques and Sardinians to be equally ancient N. Eurasian (around 25%), contrary to earlier estimates based on specific Amerindian populations.

Middle Eastern populations were consistently around the Sardinian midpoint, minus whatever SSA admixture they have.

The stronger signal of this ancient element in Northern Europeans is probably a combination of several factors:

Some N. European populations have moderate to significant recent East Asian admixture. Reich et al's initial claims of 'N. European affinities to Amerindians' were driven by the high scoring of Russians alone. Orcadians were not found to be markedly higher than Tuscans or French.

Second, the figures are considerably affected by SSA admixture. The reason that, say, Portuguese obtained much lower results than the Basques in Dienekes' tests is not that Basques are much more admixed, but simply that, unlike Portuguese, Basques lack SSA admixture.

Thirdly, modern Amerindians are a poor proxy for the 'ancient N. Eurasian' element, having had significant genetic exchange with East Asians after the split between their ancestors and those of modern West Eurasians (see the EDAR gene).

Fourth, drift. The admixture occurred prior to the European/Middle Eastern split. The very negligible surplus amongst mesolithic Europeans (which Dienekes established, but neither Reich nor Loh did) is more reasonably explained by drift.





Anonymous said...

Just a quick correction: 'Loh et al' should be 'Lipson et al', although Loh did work on the paper.

Onur Dincer said...

So how widespread are the 'different regions' where 'different Mongoloid traits' peak? Which traits very far removed from the northeast Tibet/northern China/Mongolia border region?

You tell me. Since it is you who is assertive in theorizing.

Onur Dincer said...

Hamar,

Patterson et al. and Lipson et al. used different softwares for their genetic analyses (the former used ADMIXTOOLS whereas the latter used MixMapper). Different softwares means different algorithms.

One thing is clear though: the results of ADMIXTOOLS and TreeMix are much more consistent with each other than either is with the results of MixMapper. So MixMapper is the odd one in the group.

Anonymous said...

Onur,

No, Lipson et al's paper used MixMapper first, and then confirmed the findings with ADMIXTOOLS. In said paper, MixMapper found Sardinians to be ~25% ANE, and Basques to be ~27% ANE (Orcadians were ~26%). The same paper used ADMIXTOOLS to find that Sardinians and Basques were both ~24-26% ANE, depending on the particular Amerindian population.

TreeMix found the directionality of geneflow to be from modern (Northern) Europeans to modern Amerindians (which is probably false). I also believe the geneflow was specifically from Russians anyway, and I oppose the use of a highly, recently and divergently admixed group such as Russians in trying to gauge any kind of admixture that might be generalised.

ADMIXTOOLS and MixMapper are brought into symphony by rejecting the erroneous assumptions that modern Amerindian = Ancient North Eurasian and instead inferring the existence of an extinct population. This remains erroneous, however, since it ignores the contribution of East-Eurasian-proper geneflow into Amerindians subsequent to the ANE/W. Eurasian split. But, up to now, it's the best we have for getting an understanding on absolute (and relative) proportions.

A further point I've not explicitly explored before is that ADMIXTOOLS can't settle on the position of extra-European West Eurasians. For example, Patterson and Lipson found Adygei to be more admixed than all Europeans (except Russians). Dienekes did not find the same thing. Considering that if you subtract recent admixture from the Adygei, you reach the same ANE base-level as Europeans, even though the Adygei are pred. West Asian. ADMIXTOOLS, then, as employed by Patterson et al. in their first paper, also supports the later finding of Lipson et al: a basic pan-West Eurasian stability of the element, with certain recently-admixed groups being the only significant deviations.

I do, however, maintain that there is a slight excess of the ANE element in mesolithic populations vs. other West Eurasians, explaining the small but consistent surplus of ANE in non-Sardinian Europeans, probably >5%. This explains some trivial fst phenomena, but allows the overall close-clustering of all Europeans (and, accounting for admixture, of European and non-Europeans modern West Eurasians) in global PCAs.


Anonymous said...

Just to anticipate criticism, I checked the TreeMix diagram. The finding was that there was geneflow from a fairly recent ancestor of all Amerindians to Russians (which is irrelevant to all non-recently-admixed Europeans, and also detects admixture from Tuscans, Orcadians and Russians specifically to Mayans (though apparently, the inclusion of flow from Tuscans and above is an error, and that only Orcadians and Russians should have been included). So, basically, it shows flow from Amerindians to Russians.

I should also clarify my clause, "depending on the particular Amerindian population" and add to it "...Amerindian population used to calculate the inferred ANE."

As I say, this ignores the fact of significant input into Amerindian populations after the ANE had already been injected into the ancestors of all modern W. Eurasians, and this is probably why Russians appear wholly 'ANE' and AWE', instead of the correct ANE, AWE and AEE.

Also, when I mentioned the 'split' of ANE and W. Eurasians in the post above, I didn't mean in terms of common ancestry, but in terms of the injection of the element into W. Eurasians. Admixture occurred anciently, and then extra stuff happened to Amerindians (and some W. Eurasians), to put it simply.

Onur Dincer said...

No, Lipson et al's paper used MixMapper first, and then confirmed the findings with ADMIXTOOLS. In said paper, MixMapper found Sardinians to be ~25% ANE, and Basques to be ~27% ANE (Orcadians were ~26%). The same paper used ADMIXTOOLS to find that Sardinians and Basques were both ~24-26% ANE, depending on the particular Amerindian population.

Lipson et al.'s paper did no such thing. They do not even mention ADMIXTOOLS in that paper, let alone experiment with it.

ADMIXTOOLS and MixMapper are brought into symphony by rejecting the erroneous assumptions that modern Amerindian = Ancient North Eurasian and instead inferring the existence of an extinct population. This remains erroneous, however, since it ignores the contribution of East-Eurasian-proper geneflow into Amerindians subsequent to the ANE/W. Eurasian split.

As shown by Dienekes using TreeMix, Northern Europeans are not only more Amerindian-shifted than Southern Europeans, but also more East Asian-shifted than them:

http://dienekes.blogspot.com/2012/08/4-population-test-and-east-eurasian.html

Second, the figures are considerably affected by SSA admixture. The reason that, say, Portuguese obtained much lower results than the Basques in Dienekes' tests is not that Basques are much more admixed, but simply that, unlike Portuguese, Basques lack SSA admixture.

Dienekes scrubbed the Sardinian samples of any potential Negroid admixture, and the Mongoloid-looking element in Northern Europeans persisted despite the reduction in its force:

http://dienekes.blogspot.com/2012/08/scrubbing-sardinians.html

Anonymous said...

Lipson et al.'s paper did no such thing. They do not even mention ADMIXTOOLS in that paper, let alone experiment with it.

They used f4 ratio estimation. If this isn't by use of ADMIXTOOLS, then fine. But in the language of the authors, it constituted a 'confirmation' of MixMapper's results.

As shown by Dienekes using TreeMix, Northern Europeans are not only more Amerindian-shifted than Southern Europeans, but also more East Asian-shifted than them:

See below on the effect of other admixtures on the creation/exaggeration of clines where previously near-equality between populations in an element prevailed.

Dienekes scrubbed the Sardinian samples of any potential Negroid admixture, and the Mongoloid-looking element in Northern Europeans persisted despite the reduction in its force:

http://dienekes.blogspot.com/2012/08/scrubbing-sardinians.html


You missed the point the last time I made the same comment a few months ago as well. I wasn't talking about the population used as an AWE proxy (Sardinians); I was talking about the individual figures calculated for each population. Scrubbing Sardinians does not eliminate the fact that SSA-shifted populations receive a lower D-statistic relative to non-SSA-shifted populations, by virtue of mentioned SSA shift. (See D's last few sentences in the post you linked me to.)

The reason a clear, logical cline exists only in Western Europe (Basques, British Isles, French, Dutch etc.) is that these populations are the only modern West Eurasian populations uncomplicated by additional, more recent admixture.

That said, Patterson et al. did not establish a universal, unambiguous North/South European cline, despite the language of its sensationalisation.

The midpoints from that paper:

Italian - 9.5%
French - 11%
Tuscan - 11.75%
Orcadian - 12.3%
Russian - 19.15%
Adygei - 20.5%

Dienekes' results are more heavily affected by various 'other population' shifts than Patterson et al's and Lipson et al's. were, which creates the impression of clines where no clines, or much weaker clines, exist.

This is key to understanding why Patterson's Adygei received a score so high, while in D's experiments, they did not. Even though they're not SSA-admixed, they're still SSA-shifted in PCA's. This affects D-statistics.

And again,

the Mongoloid-looking element in Northern Europeans

No mention has been made, AFAIK, either in an official paper, nor by Dienekes, about the physical characteristics introduced by this element.

Anonymous said...

Ah, D's explanation is in a different article. My revised response:

Dienekes scrubbed the Sardinian samples of any potential Negroid admixture, and the Mongoloid-looking element in Northern Europeans persisted despite the reduction in its force

We're speaking of separate phenomena. Read this article to understand my earlier points:

http://dienekes.blogspot.co.uk/2012/09/quantifying-karitiana-like-admixture-in.html

Onur Dincer said...

They used f4 ratio estimation. If this isn't by use of ADMIXTOOLS, then fine. But in the language of the authors, it constituted a 'confirmation' of MixMapper's results.

It is not clear how much they confirmed MixMapper's results.

That said, Patterson et al. did not establish a universal, unambiguous North/South European cline, despite the language of its sensationalisation.

The midpoints from that paper:

Italian - 9.5%
French - 11%
Tuscan - 11.75%
Orcadian - 12.3%
Russian - 19.15%
Adygei - 20.5%

Dienekes' results are more heavily affected by various 'other population' shifts than Patterson et al's and Lipson et al's. were, which creates the impression of clines where no clines, or much weaker clines, exist.


How did you calculate those percentages?

Even though they're not SSA-admixed, they're still SSA-shifted in PCA's.

I did not understand what you meant by this sentence.

No mention has been made, AFAIK, either in an official paper, nor by Dienekes, about the physical characteristics introduced by this element.

By "Mongoloid-looking" I meant genetically Mongoloid-looking, I did not say anything about physical characteristics.

We're speaking of separate phenomena. Read this article to understand my earlier points:

Could you be more specific? I do not have much time to re-read those articles and comments and try to figure out what you mean.

Anonymous said...

If I might be permitted, a response to Eurologist:

In another thread we had some fun with facial merging software and merged extreme Caucasoids (approximating Aurignacians) with Mongoloid faces, and the results looked like modern Caucasoids (think Maria Höfl-Riesch + Mongolian women --> attractive French lady ;)).

This isn't exactly true. For whatever reason, I used the exact same images as Matt, merged in the same proportion (75% Euro, 25% East Asian), and received a much more Mongoloid composite than Matt. I have no explanation for this. But it's not really debatable that if a ~25% introgression into Europe (or, more accurately, all of West eurasia) of full, or even partial Mongoloids occurred, such features would be easily detectable, as they are in NE Europe as a result of smaller-scale N movements.

Clearly, then, Mongoloid influences are lacking in Western Europe and were brought to parts of Europe via various Y-haplo N associated movements.

Also, Maria Höfl-Riesch isn't more Caucasoid than the average French woman, indeed she's less, and the French population as a whole displays no tendency toward physical Asiaticness whatsoever.

Anonymous said...

It is not clear how much they confirmed MixMapper's results.

They were very close. MixMapper's midpoints: Sardinians = 25%, Basques= 27.3% (Orcadians were 25.3%). The f4 ratio estimates found both Basques and Sardinians to be between 23% and 25%, depending on the Amerindian reference. It seems that, for whatever reason, the chasm between Sardinians and Basques (and also between Sardinians and all Europeans) is significantly closed when the ancient population that likely introduced this signal is considered, rather than basing the figures directly on modern Amerindians.

How did you calculate those percentages?

I calculated the midpoint on the basis of alpha levels provided in table 1 of that paper.

I did not understand what you meant by this sentence.

I mean that even though Adygei have a zero percent median score on Dodecad ADMIXTURE runs, they are still African-shifted on PCA plots:

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedC1XUjJMNnFBRnRvLXhpUFdRSWp0bGc&authkey=CJ7C6ugB&hl=en&authkey=CJ7C6ugB#gid=0

They're even more African-shifted than the Portuguese. It can't be argued that 'lack of Asian-shift' is the reason they're closer to Africans than Europeans are, since Adygei are also Asian-shifted relative to most European populations.

I'm not arguing that the Adygei are SSA-admixed, but only that their general African-shift may have a similar effect as the real admixture that causes a similar shift in, say, the Portuguese. The Portuguese have a lower estimation of this ANE in Dienekes' runs because of actual SSA admixture, but perhaps Adygei have lower levels because of their similar shift. This might explain the differences between Patterson et al's and D's results.

Could you be more specific? I do not have much time to re-read those articles and comments and try to figure out what you mean.

"So, the Karitiana-like admixture of populations such as Spanish_D (est. 1.2%) is lower than their actual such admixture, because Spanish_D includes African admixture. For the Portuguese_D (est. -3.3%) where African admixture is even more significant, the effect is even stronger, and a nonsensical negative admixture score appears."

"It will be interesting to derive corrected African admixture estimates after taking into account that CEU have Amerindian-like admixture, and, covnersely, corrected Karitiana-like admixture estimates after taking into account African admixture in some populations."



Onur Dincer said...

They were very close. MixMapper's midpoints: Sardinians = 25%, Basques= 27.3% (Orcadians were 25.3%). The f4 ratio estimates found both Basques and Sardinians to be between 23% and 25%, depending on the Amerindian reference. It seems that, for whatever reason, the chasm between Sardinians and Basques (and also between Sardinians and all Europeans) is significantly closed when the ancient population that likely introduced this signal is considered, rather than basing the figures directly on modern Amerindians.

They applied f4 ratio estimation only to the Sardinians and Basques, and did not bother to check the other European and West Asian populations' MixMapper results with f4 ratio estimation.

Another important point: midpoint does not mean average, midpoint just denotes the middle point between the highest estimated ANE ("ancient northern Eurasian") admixture level and the lowest estimated ANE admixture level in a population and does not take into account the weights of the estimated ANE admixture levels in that population. So, for instance, the average of the estimated ANE admixture levels of the Sardinians may be much closer to the lowest estimated ANE admixture level in the Sardinians than to the highest estimated ANE admixture level in them.

I calculated the midpoint on the basis of alpha levels provided in table 1 of that paper.

Could you explain how you did that with an example on one of the populations?

I mean that even though Adygei have a zero percent median score on Dodecad ADMIXTURE runs, they are still African-shifted on PCA plots:

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedC1XUjJMNnFBRnRvLXhpUFdRSWp0bGc&authkey=CJ7C6ugB&hl=en&authkey=CJ7C6ugB#gid=0

They're even more African-shifted than the Portuguese. It can't be argued that 'lack of Asian-shift' is the reason they're closer to Africans than Europeans are, since Adygei are also Asian-shifted relative to most European populations.

I'm not arguing that the Adygei are SSA-admixed, but only that their general African-shift may have a similar effect as the real admixture that causes a similar shift in, say, the Portuguese. The Portuguese have a lower estimation of this ANE in Dienekes' runs because of actual SSA admixture, but perhaps Adygei have lower levels because of their similar shift. This might explain the differences between Patterson et al's and D's results.


Make a decision: do the Adygei have Negroid admixture (whether ancient or recent) or not?

"So, the Karitiana-like admixture of populations such as Spanish_D (est. 1.2%) is lower than their actual such admixture, because Spanish_D includes African admixture. For the Portuguese_D (est. -3.3%) where African admixture is even more significant, the effect is even stronger, and a nonsensical negative admixture score appears."

"It will be interesting to derive corrected African admixture estimates after taking into account that CEU have Amerindian-like admixture, and, covnersely, corrected Karitiana-like admixture estimates after taking into account African admixture in some populations."


Apparently, as you say, you and I were speaking of separate phenomena. I now see that you were just referring to the small Negroid admixture in the Portuguese and non-Basque Spaniards (which is something obvious from the genetics of both populations, especially in the Portuguese's case) and probably also southern Italians (including Sicilians) and its negative effects on the estimation of the ANE admixture levels of those populations. I agree with you on that issue.

Anonymous said...

They applied f4 ratio estimation only to the Sardinians and Basques, and did not bother to check the other European and West Asian populations' MixMapper results with f4 ratio estimation.

Correct, and that was an inexplicable decision on their part. However, it clearly did confirm the severe underestimation of the signal in previous analyses, confirmed Sardinian and Basque parity, and confirmed the pan-European nature of the admixture (since it showed that the least admixed populations were still ~25%)

All reasonable inferences lead to my conclusions, i.e. that the findings were confirmed. However, Polako-like characters like to exploit slight ambiguities in research to maintain various unrealistic delusions. You wouldn't want to appear to be such a character, would you?

Another important point: midpoint does not mean average, midpoint just denotes the middle point between the highest estimated ANE ("ancient northern Eurasian") admixture level and the lowest estimated ANE admixture level in a population and does not take into account the weights of the estimated ANE admixture levels in that population. So, for instance, the average of the estimated ANE admixture levels of the Sardinians may be much closer to the lowest estimated ANE admixture level in the Sardinians than to the highest estimated ANE admixture level in them.

Then it's a good job my conclusions have the endorsement of the researches themselves:

"The high score for the Russian and Adygei is likely to be partially confounded with the
effect discussed in the section on flow from Asia into Europe (below)."

The midpoint is an inferior substitute for an average, but it's far from meaningless, and, of course, from using the same calculation on, for example, Uygur, we can see the midpoint method produces a figure consistent with the average for said group.

Considering ancient admixture, midpoints and averages should be even closer together, since all members of an ethnicity should have it in approx. the same proportion, and differences in detected levels would owe to the age of the signal and difficulties in detecting it and not in an individual's significant excess of it. If pops are similar in terms of upper and lower boundaries, then they are similar overall.

Make a decision: do the Adygei have Negroid admixture (whether ancient or recent) or not?

I've no real opinion. Maybe Europeans are just more drifted from Africans, or maybe Adygei have more ancient SSA ancestry than Europeans, whether or not it's detectable. Or maybe SSA Africans have W. Eurasian ancestry more related to Andygei than to Europeans.

At any rate, it's irrelevant, since Adygei African-shift is the most logical explanation for their differing relative position to Europeans in various analyses.

Apparently, as you say, you and I were speaking of separate phenomena. I now see that you were just referring to the small Negroid admixture in the Portuguese and non-Basque Spaniards (which is something obvious from the genetics of both populations, especially in the Portuguese's case) and probably also southern Italians (including Sicilians) and its negative effects on the estimation of the ANE admixture levels of those populations. I agree with you on that issue.

Even minuscule amounts of SSA admixture have a significant effect on the ANE estimation. The minuscule amount in the CEU sample decreases their score relative to Britons by 3% in some runs. If African shift, and not just detectable African admixture, is considered, then the entire ranking of modern W. Eurasians re. their ANE component needs revision.

Also, I predicted the results of Lipson et al. before the publication of that paper and before I had any knowledge of it, so my logic seems to be sound.

Onur Dincer said...

All reasonable inferences lead to my conclusions, i.e. that the findings were confirmed. However, Polako-like characters like to exploit slight ambiguities in research to maintain various unrealistic delusions. You wouldn't want to appear to be such a character, would you?

You make conclusions that Lipson et al. do not make. It is premature to make such conclusions.

Then it's a good job my conclusions have the endorsement of the researches themselves:

"The high score for the Russian and Adygei is likely to be partially confounded with the
effect discussed in the section on flow from Asia into Europe (below)."

The midpoint is an inferior substitute for an average, but it's far from meaningless, and, of course, from using the same calculation on, for example, Uygur, we can see the midpoint method produces a figure consistent with the average for said group.

Considering ancient admixture, midpoints and averages should be even closer together, since all members of an ethnicity should have it in approx. the same proportion, and differences in detected levels would owe to the age of the signal and difficulties in detecting it and not in an individual's significant excess of it. If pops are similar in terms of upper and lower boundaries, then they are similar overall.


Lipson et al. refrain from such premature and unproven conclusions such as yours. I already told you that midpoint does not mean average. Average can be very different from midpoint even in fairly homogeneous populations. Also, the estimated ANE level intervals are too large; MixMapper lacks precision when it comes to estimating ANE levels.

BTW, in case you have not noticed, in my previous post I requested you to show with an example how you did your Patterson et al. population midpoint calculations.

Anonymous said...

You make conclusions that Lipson et al. do not make. It is premature to make such conclusions.

All conclusions I drew were explicit in the paper. Example:

"First, we
found that Mozabite, Bedouin, Palestinian, and Druze, in decreasing order of African ancestry,
are all optimally represented as a mixture between an admixed western Eurasian population
(not necessarily European) related to Sardinian and an African population (Table 3)."

Of course, if I were to start talking about Mongoloid this, Mongoloid that, you'd be correct in designating my conclusions premature.

Lipson et al. refrain from such premature and unproven conclusions such as yours. I already told you that midpoint does not mean average. Average can be very different from midpoint even in fairly homogeneous populations.

Again, we see a confirmation of my logic in the f4 statistics, which match the midpoint values of tested populations.

Also, the estimated ANE level intervals are too large; MixMapper lacks precision when it comes to estimating ANE levels.

Also the case with Patterson's paper. It certainly seems strange that you wish to reject a paper whose conclusions you dislike by criticising features you fully accepted in a paper whose conclusions you do like. It almost seems as though you lack an intellectual conscience. I'm sure I'm mistaken, though.

BTW, in case you have not noticed, in my previous post I requested you to show with an example how you did your Patterson et al. population midpoint calculations.

I thought it would be a waste of my time to answer, since obviously I calculated the intermediate value between the lower and upper alpha levels.

Onur Dincer said...

All conclusions I drew were explicit in the paper. Example:

"First, we
found that Mozabite, Bedouin, Palestinian, and Druze, in decreasing order of African ancestry,
are all optimally represented as a mixture between an admixed western Eurasian population
(not necessarily European) related to Sardinian and an African population (Table 3)."


What does that to do with your conclusions?

Of course, if I were to start talking about Mongoloid this, Mongoloid that, you'd be correct in designating my conclusions premature.

Amerindians are Mongoloid, so do indigenous Siberians except the westernmost ones, who are Caucasoid-Mongoloid hybrids. If a study finds an Amerindian-like genetic element in all Caucasoids that comes from the ancestors of Amerindians after their split from Asian Mongoloids thus from a time when the ancestors of Amerindians were already Mongoloid, it is inevitable to make conclusions about the partial Mongoloid ancestry of Caucasoids. The Caucasoid race itself seems to be the product of the admixture of racially undifferentiated Proto-Eurasians and racially Mongoloid Proto-Amerindians according to Lipson et al.'s results. It is the Proto-Eurasian/Proto-Amerindians ancestry proportions of the various Caucasoid populations which is more in dispute (that surely requires a more advanced and precise method of genetic study to unveil).

Again, we see a confirmation of my logic in the f4 statistics, which match the midpoint values of tested populations.

Don't make me repeat myself, but where are the f4 ratio results of Caucasoid populations besides Sardinians and Basques?

Also the case with Patterson's paper. It certainly seems strange that you wish to reject a paper whose conclusions you dislike by criticising features you fully accepted in a paper whose conclusions you do like. It almost seems as though you lack an intellectual conscience. I'm sure I'm mistaken, though.

Not sure what you are talking about.

I thought it would be a waste of my time to answer, since obviously I calculated the intermediate value between the lower and upper alpha levels.

Please don't get me wrong, I am not questioning the integrity of your calculations. I just wanted to see a small example of your calculations.

Onur Dincer said...

By the way, the formation of the Caucasoid race must have happened when Proto-Amerindians were still living in Siberia during Paleolithic times. Amerindians seem to much better preserve the genetic structure of Paleolithic Siberians than modern-day indigenous Siberians. Modern-day indigenous Siberians, as Lipson et al. show, seem to be the result of the Holocene-era admixture of Paleolithic-era indigenous Siberians, who seem to be genetically Amerindian-like, and the Asian Mongoloid new-comers from the south.

terryt said...

"Modern-day indigenous Siberians, as Lipson et al. show, seem to be the result of the Holocene-era admixture of Paleolithic-era indigenous Siberians, who seem to be genetically Amerindian-like, and the Asian Mongoloid new-comers from the south".

Exactly. The Amerindians can in no way be called 'Proto-Mongoloid'. They are mixed.

"By the way, the formation of the Caucasoid race must have happened when Proto-Amerindians were still living in Siberia during Paleolithic times".

It could be older than that even. As people moved north they would hardly have done so in a single advancing wave. A group from one region of the northern margin would develop the technology necessary to move somewhat further north. They would then have moved along the northern margin, mixing with groups to the south. Then another group would develop a yet more sophisticated technology enabling them in turn to move further north. In the Y-DNA Q looks to have been first out of the blocks from somewhere near the Altai. They carried a possibly pre-existing Caucasoid phenotype. Next was C3 with a somewhat Mongoloid phenotype. Last would have been N with a very Mongoloid phenotype which failed to reach America.

"If a study finds an Amerindian-like genetic element in all Caucasoids that comes from the ancestors of Amerindians after their split from Asian Mongoloids"

I don't see how they could be said to have 'split from Asian Mongoloids'. Mixed with them, yes.

Elsewhere you asked what is the point of calling 1/16 mixture a hybrid. That proportion comes in at 6.25%, which I seem to remember is the level of eastern genetic element many Europeans have. That level of admixture is of course equivalent to one Mongoloid great great parent, which I think you will agree is reasonably substantial.

Anonymous said...

My reply failed to register, it seems. I'll have another go.

What does that to do with your conclusions?

It depends which conclusions we're both talking about. The quote I offered supports my conclusion that ANE admixture is pan-West Eurasian.

Amerindians are Mongoloid, so do indigenous Siberians except the westernmost ones, who are Caucasoid-Mongoloid hybrids. If a study finds an Amerindian-like genetic element in all Caucasoids that comes from the ancestors of Amerindians after their split from Asian Mongoloids thus from a time when the ancestors of Amerindians were already Mongoloid, it is inevitable to make conclusions about the partial Mongoloid ancestry of Caucasoids.

The EDAR discovery shows, if nothing else, that substantial geneflow occurred between East Asians and 'Amerindians' (whether still in Asia at the time or in the Americas) that Europeans weren't party to.

Also consider the extreme physical diversity of American natives. Australoid tendencies are fairly evident in South America, and the 'Amerindian' signal in Europeans is strongest in them. But my independently concluding and dogmatically asserting that ANEs were proto-Australoid would surely be unfounded, do you not agree? This is why it's best to stick to the facts and the language of the researchers in question, no matter how convinced we are of our own infallibility.

Don't make me repeat myself, but where are the f4 ratio results of Caucasoid populations besides Sardinians and Basques?

I said that those results confirmed the accuracy of MixMapper. The developers of those methods know what constitutes confirmation. The midpoint method is also shown to be sufficiently accurate by the same results discussed.

Also note how the upper and lower alphas for each population are higher or lower relative to the upper and lower alphas of other populations in a way perfectly consistent with the midpoint. Indeed, this follows from the fact that all populations consistently have internal variation at levels of around 20%. This strongly suggests symmetrical patterns of variation in the signal for all populations.

Not sure what you are talking about.

The intra-population variance is high in both papers. If you discount one, discount the other.

Please don't get me wrong, I am not questioning the integrity of your calculations. I just wanted to see a small example of your calculations.

Adygei alpha = 0.254-0.461

Add together, divide by 2.

Onur Dincer said...

Exactly. The Amerindians can in no way be called 'Proto-Mongoloid'. They are mixed.

I did not say Amerindians are mixed, I said modern-day Siberians are a mixture of an Amerindian-like northern Mongoloid and an East Asian-like southern Mongoloid population. Both Amerindians and East Asians are full Mongoloids. Amerindians are closer to the Proto-Mongoloid form while East Asians are more derived.

In the Y-DNA Q looks to have been first out of the blocks from somewhere near the Altai. They carried a possibly pre-existing Caucasoid phenotype. Next was C3 with a somewhat Mongoloid phenotype. Last would have been N with a very Mongoloid phenotype which failed to reach America.

We have no information about the initial phenotypes of the carriers of those haplogroups. Haplogroups and especially paternal haplogroups and especially for such early periods are not a trustable source of information about racial makeup. Look at the autosomal genetics instead.

I don't see how they could be said to have 'split from Asian Mongoloids'. Mixed with them, yes.

See the trees of the Lipson et al. paper. There you will see that Amerindians and Asian Mongoloids form a clade to the exclusion of all other populations. They are both full blood sons of Proto-Mongoloids and have equal right to the throne.

Elsewhere you asked what is the point of calling 1/16 mixture a hybrid. That proportion comes in at 6.25%, which I seem to remember is the level of eastern genetic element many Europeans have. That level of admixture is of course equivalent to one Mongoloid great great parent, which I think you will agree is reasonably substantial.

As I already told you at that thread, racial hybridity is not a matter of mathematics, it is a matter of phenotype.

Onur Dincer said...

It depends which conclusions we're both talking about. The quote I offered supports my conclusion that ANE admixture is pan-West Eurasian.

I am not questioning its pan-West Eurasianness. It is in its distribution among individual West Eurasian populations where we seem to have disagreements. You claim that it is unifomly distributed in West Eurasian populations with no recent (relatively speaking) non-Caucasoid admixture, while I say that it may not be. You are the one who is making strong assertions. I am just saying that the evidence is not conclusive in either direction and that more study is needed.

The EDAR discovery shows, if nothing else, that substantial geneflow occurred between East Asians and 'Amerindians' (whether still in Asia at the time or in the Americas) that Europeans weren't party to.

The EDAR370A mutation is so widespread and in high frequency in both Amerindians and Asian Mongoloids that it seems sure to be a Proto-Mongoloid trait (note that Amerindians and Asian Mongoloids are equally-descended from Proto-Mongoloids). As for its absence in West Eurasians, it may well have been deselected during the formation of the Caucasoid race (=the AWE-ANE admixture process) since it is the bearer of a salient trait.

Also consider the extreme physical diversity of American natives. Australoid tendencies are fairly evident in South America, and the 'Amerindian' signal in Europeans is strongest in them. But my independently concluding and dogmatically asserting that ANEs were proto-Australoid would surely be unfounded, do you not agree? This is why it's best to stick to the facts and the language of the researchers in question, no matter how convinced we are of our own infallibility.

Amerindians are Mongoloid, not Australoid.

I said that those results confirmed the accuracy of MixMapper. The developers of those methods know what constitutes confirmation. The midpoint method is also shown to be sufficiently accurate by the same results discussed.

Also note how the upper and lower alphas for each population are higher or lower relative to the upper and lower alphas of other populations in a way perfectly consistent with the midpoint. Indeed, this follows from the fact that all populations consistently have internal variation at levels of around 20%. This strongly suggests symmetrical patterns of variation in the signal for all populations.

The intra-population variance is high in both papers. If you discount one, discount the other.


See my reply at the top of this post.

Adygei alpha = 0.254-0.461

Add together, divide by 2.


Those are the Lipson et al. values, not the Patterson et al. values. I did not ask you how you calculated the midpoints of the Lipson et al. paper, since I already calculated them myself. I asked you how you calculated the midpoints of the Patterson et al. paper, since you claimed to have calculated the midpoint values of the Patterson et al. paper and even listed their values:

"Italian - 9.5%
French - 11%
Tuscan - 11.75%
Orcadian - 12.3%
Russian - 19.15%
Adygei - 20.5%"

terryt said...

"modern-day Siberians are a mixture of an Amerindian-like northern Mongoloid and an East Asian-like southern Mongoloid population"

That is an impossible scenario. The region now occupied by what you like to call 'southern Mongoloid' has only become 'Mongoloid' in the last three or four thousand years. Until then people in the region looked more like Papuans or Australian Aborigines than anything like Mongoloids. Any ancient contribution to Siberia from that region would certainly not look Mongoloid in any way at all. Southern China, and especially Southeast Asia, provide a huge amount of evidence for a relatively recent southward moving Mongoloid phenotype. I suggest that the more Mongoloid look of Siberians in relation to Amerindians shows the movement was not restricted to just a southward direction.

"Amerindians are closer to the Proto-Mongoloid form while East Asians are more derived".

What the hell is a 'Proto-Mongoloid'? Or are you claiming that some selection pressure leading to a Mongoloid phenotype continued acting on a Eurasian population but ceased to act on the ancestors of the Amerindians? And where did such a postulated population originally live?

"The EDAR370A mutation is so widespread and in high frequency in both Amerindians and Asian Mongoloids that it seems sure to be a Proto-Mongoloid trait"

The EDAR370A mutation is a Mongoloid trait. I agree it has spread (unevenly) with the expansion of the Mongoloid phenotype.

"As for its absence in West Eurasians, it may well have been deselected during the formation of the Caucasoid race (=the AWE-ANE admixture process) since it is the bearer of a salient trait".

And yet the mutation continued present as a population carried it into SE Asia, and even into Munda-speaking people of India. I find your proposed scenario difficult to accept. I would say it is far more likely that the EDAR mutation was never present in West Eurasians.

"Amerindians and Asian Mongoloids are equally-descended from Proto-Mongoloids"

That statememnt is simply not true.

"Both Amerindians and East Asians are full Mongoloids".

Most Amerindians are easily distinguished from East Asian populations.

"There you will see that Amerindians and Asian Mongoloids form a clade to the exclusion of all other populations".

They cluster together because they share a substantial proportion of common ancestry but Amerindians are not descended totally from Mongoloids. In fact we have utosomal studies showing they have genes in common with Europeans. After all the subject of the post is 'Amerindian-like admixture in northern Europe' not 'Mongoloid-like admixture in Northern Europe'.

Onur Dincer said...

That is an impossible scenario. The region now occupied by what you like to call 'southern Mongoloid' has only become 'Mongoloid' in the last three or four thousand years. Until then people in the region looked more like Papuans or Australian Aborigines than anything like Mongoloids. Any ancient contribution to Siberia from that region would certainly not look Mongoloid in any way at all. Southern China, and especially Southeast Asia, provide a huge amount of evidence for a relatively recent southward moving Mongoloid phenotype. I suggest that the more Mongoloid look of Siberians in relation to Amerindians shows the movement was not restricted to just a southward direction.

I said "East Asian-like southern Mongoloid" not "Southeast Asian-like southern Mongoloid". The southernness in question is relative to the territory of Proto-Amerindians (=Siberia).

What the hell is a 'Proto-Mongoloid'?

Mongoloids prior to the diversication into the Proto-Amerindian and Proto-Asian (Proto-East Asian to be more correct) branches.

Or are you claiming that some selection pressure leading to a Mongoloid phenotype continued acting on a Eurasian population but ceased to act on the ancestors of the Amerindians?

A selection pressure leading to a characteristically Asian Mongoloid phenotype that seems to have ceased to act on the ancestors of Amerindians.

And where did such a postulated population originally live?

Proto-Mongoloids? Anywhere in the habitable zone of eastern Eurasia north of the Australoid territory. It is a huge territory, and the physical traits of Proto-Mongoloids probably evolved independently in different parts of that territory and came together through admixture to form the umbrella population that we call Proto-Mongoloid.

The EDAR370A mutation is a Mongoloid trait. I agree it has spread (unevenly) with the expansion of the Mongoloid phenotype.

There is no such thing as a single Mongoloid phenotype. Mongoloid is an umbrella term for the conglomeration of certain diverse phenotypes that came together in varying proportions depending on space and time.

And yet the mutation continued present as a population carried it into SE Asia, and even into Munda-speaking people of India. I find your proposed scenario difficult to accept. I would say it is far more likely that the EDAR mutation was never present in West Eurasians.

My scenario is the most plausible one in light of the results of the Lipson et al. paper.

That statememnt is simply not true.

Genetic studies say the opposite.

Most Amerindians are easily distinguished from East Asian populations.

So what? East Asians are not the paragons of Mongoloidness.

They cluster together because they share a substantial proportion of common ancestry but Amerindians are not descended totally from Mongoloids. In fact we have utosomal studies showing they have genes in common with Europeans.

Only at very small K levels and not always. In some autosomal analyses East Asians appear closer to Caucasoids instead. So there is no consistency. This confirms neighbor joining trees, which show Asian Mongoloids and Amerindians as equally-descended from Proto-Mongoloids. BTW, the fact that Asian Mongoloids are more derived may also be influencing autosomal results by showing Asian Mongoloids as more distant from Caucasoids than Amerindians are.

After all the subject of the post is 'Amerindian-like admixture in northern Europe' not 'Mongoloid-like admixture in Northern Europe'.

But Amerindians are Mongoloid. You cannot disassociate them from Mongoloidness.

Anonymous said...

I am not questioning its pan-West Eurasianness. It is in its distribution among individual West Eurasian populations where we seem to have disagreements. You claim that it is unifomly distributed in West Eurasian populations with no recent (relatively speaking) non-Caucasoid admixture, while I say that it may not be. You are the one who is making strong assertions. I am just saying that the evidence is not conclusive in either direction and that more study is needed.

I agree that more study would be nice, but in all likelihood, Lipson et al's paper was the 'final word' from that department, at least for a while.

I've never held the position that ANE is exactly equal in all West Eurasian populations. I've always said that mesolithic Europeans likely had a slight surplus of it, and this explains some insignificant minutiae of fst distances within non-admixed (relatively speaking) Europeans. However, the difference within Europe, best exemplified by comparisons between Basques (whom we now have reason to consider equal in ANE to the Sardinians, because Sardinian 'uniqueness' seems to have been exaggerated in earlier papers) and NW Europeans, is quite minor, and that's even before we take into account the ~25% pan-West Eurasian baseline.

It appears to have been a single admixture event in the history of all modern W. Eurasians, so an essentially equal level is logical. It's possible that this admixed population then bred back into an extant, still-unadmixed W. Eurasian population to differing degrees, and in this way created different levels in modern populations while still explaining the similar admixture times, but Lipson's paper doesn't show much evidence for this.

Basically, if you subtract the ~10% recent admixture from Russians, the ~7% recent admixture from the Adygei, and account for SSA in the Druze, Mozabite etc., all populations seem to end up within 5% of each other.

And, of course, the West Asian and North European components are almost equal in their relation to Amerindians (and South Asians, since I remember your previous argument).

The EDAR370A mutation is so widespread and in high frequency in both Amerindians and Asian Mongoloids that it seems sure to be a Proto-Mongoloid trait (note that Amerindians and Asian Mongoloids are equally-descended from Proto-Mongoloids). As for its absence in West Eurasians, it may well have been deselected during the formation of the Caucasoid race (=the AWE-ANE admixture process) since it is the bearer of a salient trait.

Are there any parallel circumstances of a trait being so extremely selectively favoured by one human group and so extremely selectively disfavoured by another? Doesn't Ockham's razor demand that for now we assume West Eurasians never had the mutation to either select or deselect?

Amerindians are Mongoloid, not Australoid.

You know I didn't say that they were. I said they could possibly be part Australoid:

http://culturakaritiana.blogspot.co.uk/

Maybe such traits were adaptations, maybe they were the result of admixture with a more indigenous population. The only point I'm making here is that Amerindian anthropology is far from clear-cut.

Those are the Lipson et al. values, not the Patterson et al. values. I did not ask you how you calculated the midpoints of the Lipson et al. paper, since I already calculated them myself. I asked you how you calculated the midpoints of the Patterson et al. paper, since you claimed to have calculated the midpoint values of the Patterson et al. paper and even listed their values:

Ah, okay. I now understand why you asked me how I did it. The Patterson score gives you the 'Sardinian' proportion, so you work that out by adding the lower and upper alphas, and then count the difference between that midpoint and 100.

Onur Dincer said...

I agree that more study would be nice, but in all likelihood, Lipson et al's paper was the 'final word' from that department, at least for a while.

I've never held the position that ANE is exactly equal in all West Eurasian populations. I've always said that mesolithic Europeans likely had a slight surplus of it, and this explains some insignificant minutiae of fst distances within non-admixed (relatively speaking) Europeans. However, the difference within Europe, best exemplified by comparisons between Basques (whom we now have reason to consider equal in ANE to the Sardinians, because Sardinian 'uniqueness' seems to have been exaggerated in earlier papers) and NW Europeans, is quite minor, and that's even before we take into account the ~25% pan-West Eurasian baseline.

It appears to have been a single admixture event in the history of all modern W. Eurasians, so an essentially equal level is logical. It's possible that this admixed population then bred back into an extant, still-unadmixed W. Eurasian population to differing degrees, and in this way created different levels in modern populations while still explaining the similar admixture times, but Lipson's paper doesn't show much evidence for this.

Basically, if you subtract the ~10% recent admixture from Russians, the ~7% recent admixture from the Adygei, and account for SSA in the Druze, Mozabite etc., all populations seem to end up within 5% of each other.


The recent Mongoloid admixture levels are lower in Vologda Russians and Adygei than those. As for your theory of a relatively uniform ANE admixture in non-recent-non-Caucasoid-admixed Caucasoids, we will wait and see whether it will be confirmed in future studies.

And, of course, the West Asian and North European components are almost equal in their relation to Amerindians (and South Asians, since I remember your previous argument).

Which admixture analysis are you referring to?

Are there any parallel circumstances of a trait being so extremely selectively favoured by one human group and so extremely selectively disfavoured by another? Doesn't Ockham's razor demand that for now we assume West Eurasians never had the mutation to either select or deselect?

I know it sounds weird, but after seeing the results of the Lipson et al. paper I can see no other theory more plausible than that.

You know I didn't say that they were. I said they could possibly be part Australoid:

http://culturakaritiana.blogspot.co.uk/

Maybe such traits were adaptations, maybe they were the result of admixture with a more indigenous population. The only point I'm making here is that Amerindian anthropology is far from clear-cut.


Amerindians cannot be more Australoid-admixed than East Asians, since East Asians were - and still are - closer to the Australoid territories. If you want to see Mongoloids admixed with Australoids, you should look at Oceania and Southeast Asia.

Ah, okay. I now understand why you asked me how I did it. The Patterson score gives you the 'Sardinian' proportion, so you work that out by adding the lower and upper alphas, and then count the difference between that midpoint and 100.

Thanks for the info.

terryt said...

"A selection pressure leading to a characteristically Asian Mongoloid phenotype that seems to have ceased to act on the ancestors of Amerindians".

It is extremely difficult to imagine any selection scenario that would give that result. For some time the ancestors of Amerindians must have lived as far north as any East Asians and so any environmental selection would actually be more extreme on them than it would be on East Asians as a whole, surely.

"Mongoloids prior to the diversication into the Proto-Amerindian and Proto-Asian (Proto-East Asian to be more correct) branches".

With the information presented by Dienekes here it is obvious that there was no simple split between Proto-Amerindians and Proto-East Asians. Amerindians have a genetic element not present in East Asians. Unless you're going to postulate the entire elimination of a set of genes in East Asians simultaneously across a vast region.

"There is no such thing as a single Mongoloid phenotype. Mongoloid is an umbrella term for the conglomeration of certain diverse phenotypes that came together in varying proportions depending on space and time".

I disagree entirely. THe EDAR gene is simply the only one identified so far but I'm certain time will reveal other genes. East Asians are immediately identifiable and that through phenotype. I do agree that the various genes involved may have originated across a considerable region but that region is nowhere near as huge as you imagine:

"Anywhere in the habitable zone of eastern Eurasia north of the Australoid territory".

We can eliminate most, if not all, of South China for a start. Amerindians did not arrive in that continent until some 15,000 years ago and presumably lived somewhat south of Beringia until soon before they reached it. That narrows the region considerably.

"So what? East Asians are not the paragons of Mongoloidness".

But Amerindians are certainly not. So that takes us back to East Asia.

"I said they could possibly be part Australoid"

Very possibly so. Here is a paper on teeth:

http://ejournal.anu.edu.au/index.php/bippa/article/viewFile/213/203

It specifically draws a connection between Australian and Jomon dentition. Perhaps that is a reflection of the extreme northward extent of the Australoid phenotype before the expansion of the Mongoloid phenotype. Amerindians may include an element of this Australoid population.

"I know it sounds weird, but after seeing the results of the Lipson et al. paper I can see no other theory more plausible than that".

Only because you are not prepared to accept the possibility of hybrid formation between differing genetic pools. Surely such hybrid formation would explain everything.

terryt said...

"genetiker is not open to even the most moderate forms of criticism. He refuses to publish my post on the 'Amerindians have both Mongoloid and Caucasoid physical features' thread"

Have you changed your mind?

Onur Dincer said...

It is extremely difficult to imagine any selection scenario that would give that result. For some time the ancestors of Amerindians must have lived as far north as any East Asians and so any environmental selection would actually be more extreme on them than it would be on East Asians as a whole, surely.

What the hell do we really know about the selection presures acting on ancestors of Mongoloids, be it the Amerindian variety or the East Asian variety?

With the information presented by Dienekes here it is obvious that there was no simple split between Proto-Amerindians and Proto-East Asians. Amerindians have a genetic element not present in East Asians. Unless you're going to postulate the entire elimination of a set of genes in East Asians simultaneously across a vast region.

Even if you are right, it now seems clear from advanced genomewide autosomal analyses that Amerindians are overwhelmingly, if not fully, Mongoloid. It is possible that some of the genetic divergence between Amerindians and East Asians is really from non-Mongoloid admixture in Amerindians, but that non-Mongoloid admixture cannot be in the levels that make it plausible to envision a Proto-Amerindian population (the population that formed at least about a quarter of the ancestry of the Caucasoid race according to Lipson et al.) that is not overwhelmingly Mongoloid. What is more, probably it is precisely that Proto-Amerindian admixture in Caucasoids that is making Amerindians seem as "Caucasoid-admixed" at the the lowest levels of ADMIXTURE analyses. Since ADMIXTURE is not a formal test for admixture, it has limits in distinguishing between source and destination populations for genomic sections. To make accurate inferences about source and destination populations and direction of gene flow, you need formal tests for admixture such as those of MixMapper and ADMIXTOOLS. It is thanks to them that we are now inferring that the gene flow was from Proto-Amerindians to what would become the Caucasoid race, and not the other way around.

I disagree entirely. THe EDAR gene is simply the only one identified so far but I'm certain time will reveal other genes. East Asians are immediately identifiable and that through phenotype. I do agree that the various genes involved may have originated across a considerable region but that region is nowhere near as huge as you imagine:

You are again very much in the speculative zone. I advice you not to be so certain in your personal opinions.

We can eliminate most, if not all, of South China for a start. Amerindians did not arrive in that continent until some 15,000 years ago and presumably lived somewhat south of Beringia until soon before they reached it. That narrows the region considerably.

That "narrowed" region is huge enough for the formation of a lot of different physical traits in different corners.

But Amerindians are certainly not. So that takes us back to East Asia.

I answered this in detail above.

Very possibly so. Here is a paper on teeth:

http://ejournal.anu.edu.au/index.php/bippa/article/viewFile/213/203

It specifically draws a connection between Australian and Jomon dentition. Perhaps that is a reflection of the extreme northward extent of the Australoid phenotype before the expansion of the Mongoloid phenotype. Amerindians may include an element of this Australoid population.


Amerindians are not included in that analysis. Plus, Japan is a pretty restricted and isolated region (it is an archipelago!), you cannot draw conclusions about Eastern Eurasia in general based on Japan.

Only because you are not prepared to accept the possibility of hybrid formation between differing genetic pools. Surely such hybrid formation would explain everything.

Answered this too.

Onur Dincer said...

Have you changed your mind?

What do you mean?

terryt said...

"What do you mean?"

You seem to maintain two contradictory positions. One that Amerindians are only of Mongoloid, or East Asian, origin, the other that they have a non-East Asian Caucasoid element. The latter postition is one I can easily agree with.

"What the hell do we really know about the selection presures acting on ancestors of Mongoloids, be it the Amerindian variety or the East Asian variety?"

Surely it has to be some sort of environmental pressure. Most Mongoloid characters seem adaptations to strongly reflective, cold environments. I've seen 'sexual selection' proposed but that makes no sense at all.

"It is possible that some of the genetic divergence between Amerindians and East Asians is really from non-Mongoloid admixture in Amerindians"

Not 'possible' but 'extremely likely'. The Amerindian mixture in Europe seems independent of, and more ancient than, Mongoloid admixture.

"the gene flow was from Proto-Amerindians to what would become the Caucasoid race, and not the other way around".

Absolutely. But the Caucasoid and Mongoloid 'races' have less in common than do Amerindian and Caucasoid.

"I advice you not to be so certain in your personal opinions".

I think anyone could tell immediately and Amerindian from most East Asians, especially those from Northern China and points north in Eastern Eurasia, so it is hardly 'personal opinions'.

"Amerindians are not included in that analysis. Plus, Japan is a pretty restricted and isolated region (it is an archipelago!), you cannot draw conclusions about Eastern Eurasia in general based on Japan".

No, Amerindians are not included in the paper but ancient Japanese are. A connection between Japan and points south is indicated in the paper along with a strong indication that sinodonty is a later presence in Japan itself. The dentition suggests an ancient connection between Australia (including SE Asia) and Japan. And there is a possibility that at least some genetic element of Amerindians may have come via Japan rather than via northeast Eurasia. Mitochondrial DNA B for example is not a north Eurasian haplogroup.

Onur Dincer said...

You seem to maintain two contradictory positions. One that Amerindians are only of Mongoloid, or East Asian, origin, the other that they have a non-East Asian Caucasoid element. The latter postition is one I can easily agree with.

I have never claimed that Amerindians have a non-East Asian Caucasoid element. Where did you get that idea?

Surely it has to be some sort of environmental pressure. Most Mongoloid characters seem adaptations to strongly reflective, cold environments. I've seen 'sexual selection' proposed but that makes no sense at all.

The possible time period and area for the formation of the Mongoloid race are so broad and the physical traits in question are so diverse (also only a couple of them are potentially cold adaptations) and, apparently, genetically independent from each other that there is no reason to favor a spatiotemporally restricted and simple scenario such as yours for the formation of the Mongoloid race. The real world is far more complex than you imagine. Even the cold adaptations do not need to be related to each other.

Not 'possible' but 'extremely likely'. The Amerindian mixture in Europe seems independent of, and more ancient than, Mongoloid admixture.

Genetics refutes your scenario of "later Mongoloidization" of Proto-Amerindians. So we cannot think of Proto-Amerindians without Mongoloidness. Mongoloidness is inseparable from them.

Absolutely. But the Caucasoid and Mongoloid 'races' have less in common than do Amerindian and Caucasoid.

Incorrect. Amerindians are racially Mongoloid, while the Caucasoid race is only partially-descended from the Mongoloid race (much more from its Amerindian-like varieties than from its East Asian-like varieties).

I think anyone could tell immediately and Amerindian from most East Asians, especially those from Northern China and points north in Eastern Eurasia, so it is hardly 'personal opinions'.

My reply "I advice you not to be so certain in your personal opinions" was directed at your statements "the EDAR gene is simply the only one identified so far but I'm certain time will reveal other genes" and "but that region is nowhere near as huge as you imagine", not at your intermediate statement "East Asians are immediately identifiable and that through phenotype". Sorry for the ambiguity.

No, Amerindians are not included in the paper but ancient Japanese are. A connection between Japan and points south is indicated in the paper along with a strong indication that sinodonty is a later presence in Japan itself. The dentition suggests an ancient connection between Australia (including SE Asia) and Japan. And there is a possibility that at least some genetic element of Amerindians may have come via Japan rather than via northeast Eurasia. Mitochondrial DNA B for example is not a north Eurasian haplogroup.

From the abstract of the recent paper about the genomewide ancient DNA autosomal genetic analysis of a Jomon individual from Sanganji, Fukushima, Japan, whose results will be presented at the SMBE 2013 conference:

"Sanganji Jomon was very similar with modern East Asians when we compared the worldwide populations in the PCA plot"

So, completely in line with the results of the previous studies of the Jomon people, it demonstrates that the Jomon people were racially Mongoloid. But to which varieties of the Mongoloid race they belonged is still waiting to be discovered.

terryt said...

"I have never claimed that Amerindians have a non-East Asian Caucasoid element. Where did you get that idea?"

You said in the first comment, 'Amerindians have both Mongoloid and Caucasoid physical features'. Or am I misinterpreting your viewpoint?

"The real world is far more complex than you imagine".

It is you who seems to believe in a unidirectional human movement out of Africa. I believe the movement has been much more complicated than that, with groups becoming isolated and developing in their own direction before meeting up with other groups, forming hybrids and then expanding further.

"the physical traits in question are so diverse (also only a couple of them are potentially cold adaptations)"

Are you certain of that? What adaptations do you believe are not cold adaptations?

"Genetics refutes your scenario of 'later Mongoloidization' of Proto-Amerindians".

I'm not claiming 'later Mongoloidization' for Amerindians. I'm claiming hybridization before they ever came near America, although I accept that later movements into America appear to be more 'Mongoloid'.

"the Caucasoid race is only partially-descended from the Mongoloid race (much more from its Amerindian-like varieties than from its East Asian-like varieties)".

The very characteristics in Amerindians that are not specifically 'Mongoloid'.

"My reply 'I advice you not to be so certain in your personal opinions' was directed at your statements"

My comments on the EDAR gene were entirely based on the map the authors supplied showing the small region where the gene had first reached fixation before expanding from that region. If you know of any other genes that give rise to elements of the Mongoloid phenotype please let us all know.

"Sanganji Jomon was very similar with modern East Asians when we compared the worldwide populations in the PCA plot"

Yes. From the abstract, 'we obtained 20 million base pairs of genomic DNA from a ~4,000-year-old Jomon male tooth'. That is stunningly close to the time when people carrying the Mongoloid phenotype first appeared in SE Asia and Taiwan so it is hardly surprising that it had started moving into Japan as well.

terryt said...

I accept that Wikipedia is not the most reliable of sources but here are a couple of links with quotes:

http://en.wikipedia.org/wiki/Sinodonty_and_Sundadonty

"the Sinodont patterns of the Paleo-Amerindians identify their ancestral homeland as north-east Asia. Some later studies have questioned this and found Sundadont features in some American peoples. For example, in 1996, Rebecca Haydenblit of the Hominid Evolutionary Biology Research Group at Cambridge University did a study on the dentition of four pre-Columbian Mesoamerican populations and compared their data to 'other Mongoloid populations'.[3] She found that 'Tlatilco', 'Cuicuilco', 'Monte Albán' and 'Cholula' populations followed an overall 'Sundadont' dental pattern 'characteristic of Southeast Asia' rather than a 'Sinodont' dental pattern 'characteristic of Northeast Asia'".[3]

http://en.wikipedia.org/wiki/Mongoloid

"Professor of anthropology, Akazawa Takeru (Japanese:赤沢威) of the International Research Center for Japanese Studies in Kyoto Japan, said Mongoloid features are an adaption to the cold of the Mammoth Steppe.[29] He mentions the Lewis waves of warm blood cyclical vasodilation and vasoconstriction of the peripheral capillaries in Mongoloids as an adaption to the cold.[29] He lists the short limbs, short noses, flat faces, epicanthic fold and lower surface to mass ratio as further Mongoloid adaptions to cold"

So let's hear what Mongoloid features are not cold adaptations.

"Mongoloid /ˈmɒŋ.ɡə.lɔɪd/[1][2] refers to populations that share certain phenotypic traits such as epicanthic fold and shovel-shaped incisors and other physical traits common in most of Asia (including Central Asia, South Asia, North Asia, etc.), the Arctic, the Americas and most of the Pacific Islands".

So, the expression 'Mongoloid' includes populations we can be certain are hybrid populations. And hybrid populations in the west are readily accepted as being hybrid:

"Keane said the following peoples are mixed Mongolo-Caucasic varieties: Anatolian Turks, Uzbegs, and Tajiks of Turkestan.[16] Keane said the Kazaks are intermediate between the Túrki and Mongolian races.[16] Keane said the Mongolian race is best represented by the Buriats".

"Archaeologist Peter Bellwood claims that the vast majority of people in Southeast Asia, the region he calls the 'clinal Mongoloid-Australoid zone', are Southern Mongoloids but have a high degree of Australoid admixture".

"In 1998, Jack D. Forbes, professor of Native American Studies and Anthropology at the University of California, Davis, said that the racial type of the indigenous people of the Americas does not fall into the Mongoloid racial category"

"The human fossil remains of the Ordos Man from Salawusu site in Inner Mongolia dated between 50,000 and 35,000 BCE show strong Mongoloid features, specifically on the fore-tooth and occipital bone".

"Theodore G. Schurr of the Department of Anthropology at University of Pennsylvania said Mongoloid traits emerged from Transbaikalia, central and eastern regions of Mongolia, and several regions of Northern China.[44] Schurr said that studies of cranio-facial variation in Mongolia suggest that the region of modern-day Mongolians is the origin of the Mongoloid racial type".

Onur Dincer said...

You said in the first comment, 'Amerindians have both Mongoloid and Caucasoid physical features'. Or am I misinterpreting your viewpoint?

Learn to read. This is what I wrote in my first post:

"genetiker is not open to even the most moderate forms of criticism. He refuses to publish my post on the "Amerindians have both Mongoloid and Caucasoid physical features" thread despite the fact that I do not openly criticize him in that post."

"Amerindians have both Mongoloid and Caucasoid physical features" is the name of genetiker's blog thread to which I sent a critical response but which genetiker censored. It is genetiker who thinks that Amerindians have both Mongoloid and Caucasoid physical features (he explains why he think so at that thread), not me.

It is you who seems to believe in a unidirectional human movement out of Africa. I believe the movement has been much more complicated than that, with groups becoming isolated and developing in their own direction before meeting up with other groups, forming hybrids and then expanding further.

Where did I claim or imply a unidirectional movement out of Africa?

Are you certain of that? What adaptations do you believe are not cold adaptations?

I am not so certain which are cold adaptations and which are not. It is you who seems to be certain on that issue and trying to impose his thoughts on others as if they are established facts.

I'm not claiming 'later Mongoloidization' for Amerindians. I'm claiming hybridization before they ever came near America, although I accept that later movements into America appear to be more 'Mongoloid'.

What is your genetic evidence for such a hybridization event? The relevant genetic data do not support such a scenario as far as I can see.

The very characteristics in Amerindians that are not specifically 'Mongoloid'.

What do you mean by "the very characteristics in Amerindians that are not specifically Mongoloid"? And what evidence do you have that it is only "those characteristics" with which Proto-Amerindians contributed to the formation of the Caucasoid race?

My comments on the EDAR gene were entirely based on the map the authors supplied showing the small region where the gene had first reached fixation before expanding from that region. If you know of any other genes that give rise to elements of the Mongoloid phenotype please let us all know.

It is you who claims that the Mongoloid traits are closely interrelated in their origin and spread and treats Mongoloidness as if it is a single phenotype, but you provide the distribution map of only one gene that is related to any physical trait that is characteristic of Mongoloids and that gene has a very limited effect on the totality of the physical traits that are characteristic of the Mongoloid race. So by not providing the distribution maps of the other genes that are related to any physical trait that is characteristic of the Mongoloid race or the distribution map of any physical trait that is characteristic of Mongoloids, for that matter, you are doing no good to your credibility.

Yes. From the abstract, 'we obtained 20 million base pairs of genomic DNA from a ~4,000-year-old Jomon male tooth'. That is stunningly close to the time when people carrying the Mongoloid phenotype first appeared in SE Asia and Taiwan so it is hardly surprising that it had started moving into Japan as well.

You are still treating Mongoloidness as if it is a single phenotype.

Onur Dincer said...

"the Sinodont patterns of the Paleo-Amerindians identify their ancestral homeland as north-east Asia. Some later studies have questioned this and found Sundadont features in some American peoples. For example, in 1996, Rebecca Haydenblit of the Hominid Evolutionary Biology Research Group at Cambridge University did a study on the dentition of four pre-Columbian Mesoamerican populations and compared their data to 'other Mongoloid populations'.[3] She found that 'Tlatilco', 'Cuicuilco', 'Monte Albán' and 'Cholula' populations followed an overall 'Sundadont' dental pattern 'characteristic of Southeast Asia' rather than a 'Sinodont' dental pattern 'characteristic of Northeast Asia'".[3]

Proto-Mongoloids were either Sundadonts or a mixed community of Sundadonts and Sinodonts. There is no indication that Sundadonty is a later intrusion to Mongoloids.

"Professor of anthropology, Akazawa Takeru (Japanese:赤沢威) of the International Research Center for Japanese Studies in Kyoto Japan, said Mongoloid features are an adaption to the cold of the Mammoth Steppe.[29] He mentions the Lewis waves of warm blood cyclical vasodilation and vasoconstriction of the peripheral capillaries in Mongoloids as an adaption to the cold.[29] He lists the short limbs, short noses, flat faces, epicanthic fold and lower surface to mass ratio as further Mongoloid adaptions to cold"

What else does he say according to that Wikipedia article:

"Professor of anthropology, Akazawa Takeru (Japanese:赤沢威) at the International Research Center for Japanese Studies, Kyoto, said that there are Neo-Mongoloids and Paleo-Mongoloids.[29] Akazawa said Neo-Mongoloids have "extreme Mongoloid, cold-adapted features" and they include the Chinese, Buryats, Eskimo and Chukchi.[29] In contrast, Akazawa said Paleo-Mongoloids are less Mongoloid and less cold-adapted.[29] He said Burmese, Filipinos, Polynesians, Jōmon and the indigenous peoples of the Americas were Paleo-Mongoloid.[29]"

So, according to him Paleo-Mongoloids are closer to original Mongoloids. Why else would he call them "Paleo-Mongoloid"?

Onur Dincer said...
This comment has been removed by the author.
Onur Dincer said...

"Keane said the following peoples are mixed Mongolo-Caucasic varieties: Anatolian Turks, Uzbegs, and Tajiks of Turkestan.[16] Keane said the Kazaks are intermediate between the Túrki and Mongolian races.[16] Keane said the Mongolian race is best represented by the Buriats".

Augustus Henry Keane (1833–1912) was a linguist, not a physical anthropologist, so he is in no way an authority on racial matters. Besides, according to the Wikipedia article on Keane:

"Keane was out of step with the anthropology of the time, preferring linguistic data to that of physical anthropology and came to occupy a marginal position in the emerging scientific discipline.[12] On the other hand his efforts at popularising anthropology were praised by Sir Harry Johnston.[11]"

So, he seems to have racially classified people based more on language than physical traits. The fact that he counts Anatolian Turks among Mongolo-Caucasic varieties clearly supports that. He seems to be no different in this regard from some of the other writers of his time who classified Hungarians, Anatolian Turks, Finns, etc. as Mongoloids or Caucaso-Mongoloid hybrids purely based on linguistics.

"In 1998, Jack D. Forbes, professor of Native American Studies and Anthropology at the University of California, Davis, said that the racial type of the indigenous people of the Americas does not fall into the Mongoloid racial category"

According to that Wikipedia article:

"Forbes said that due to the various physical traits indigenous Americans exhibit, some with "head shapes which seem hardly distinct from many Europeans", indigenous Americans must have either been formed from a mixture of Mongoloid and Caucasoid races or they descend from the ancestral, common type of both Mongoloid and Caucasoid races.[22]"

Genetics supports the theory of the less derivedness of the Mongoloidness of Amerindians.

terryt said...

"but you provide the distribution map of only one gene that is related to any physical trait that is characteristic of Mongoloids and that gene has a very limited effect on the totality of the physical traits that are characteristic of the Mongoloid race".

The EDAR mutation is the only gene discovered so far responsible for any element of the Mongoloid phenotype. We await further discoveries. And the EDAR mutation can hardly be said to be responsible for a very limited proportion of that phenotype. Sinodonty is one of its products, as well a the hair elements.

"Proto-Mongoloids were either Sundadonts or a mixed community of Sundadonts and Sinodonts. There is no indication that Sundadonty is a later intrusion to Mongoloids".

On the other hand Sinodonty IS invasive into regions of Sundadonty in SE Asia and South China. Until Sinodonty had developed in some population no population anywhere could be called 'Proto-Mongoloid'. Once it had developed the population containing it could then be called 'Mongoloid'. Simple? Just like animal breeding.

"Genetics supports the theory of the less derivedness of the Mongoloidness of Amerindians".

That viewpoint is based an an assumption which I suggest is not be justified. The 'less derivedness' is just as likely to be the product of hybrid with some other population.

"What is your genetic evidence for such a hybridization event? [hybridization before they ever came near America, although I accept that later movements into America appear to be more 'Mongoloid']The relevant genetic data do not support such a scenario as far as I can see".

The presence of Y-DNA Q and mt-DNA X, both of which are far more likely to have originally been west Eurasian haplogroups than eastern ones (related to Y-DNA R and the western mt-DNA Ns), and the limited reach of the East Asian Y-DNA C3. The work on aDNA has been complicated by the inclusion of SE Asians as 'Mongoloids' which gives an exaggerated view of the diversity of the Mongoloid phenotype.

"You are still treating Mongoloidness as if it is a single phenotype".

It basically is if you extract the obviously hybrid SE Asians and southern Chinese from your overall consideration.

"according to him Paleo-Mongoloids are closer to original Mongoloids. Why else would he call them 'Paleo-Mongoloid'?"

His Paleo-Mongoloids are an artificial construction on his part. He says they are 'less Mongoloid and less cold-adapted' and then goes on to list a string of hybrid populations, several of which are recently formed and so cannot be 'Paleo-' anything: 'Burmese, Filipinos, Polynesians, Jōmon and the indigenous peoples of the Americas'. All of these are hybrids, although you disagree as to that position for Amerindians.

"What do you mean by 'the very characteristics in Amerindians that are not specifically Mongoloid'?"

You wrote, 'the Caucasoid race is only partially-descended from the Mongoloid race (much more from its Amerindian-like varieties than from its East Asian-like varieties)' so appparently you do concede that Amerindians have a genetic component that is not Mongoloid. Caucasoids mostly do not share a genetic component with East Asians but most do share a component with Amerindians. Surely that component cannot come from East Asians but from some other source.

terryt said...

"Genetics supports the theory of the less derivedness of the Mongoloidness of Amerindians".

I have discovered that you have a timing problem:

http://en.wikipedia.org/wiki/Ordos_culture

Quote:

"The human fossil remains of the Ordos Man from Salawusu site dated between 50,000 and 35,000 BCE show strong Mongoloid features, specifically on the fore-tooth and occipital bone".

'Strong Mongoloid features' at more than 35,000 years ago? That is long before humans arrived in America. If Amerindians truly were a representative sample of East Asians from the period they show a remarkably small amount of the Mongoloid phenotype that had developed in the region some long period before they left. Certainly 'Proto-Mongoloids' (of whatever sort you believe in) had become reasonable 'fully Mongoloid' by the time of Amerindian departure.

Onur Dincer said...

The EDAR mutation is the only gene discovered so far responsible for any element of the Mongoloid phenotype. We await further discoveries. And the EDAR mutation can hardly be said to be responsible for a very limited proportion of that phenotype. Sinodonty is one of its products, as well a the hair elements.

That is a very limited proportion of the physical traits that are characteristic of the Mongoloid race.

On the other hand Sinodonty IS invasive into regions of Sundadonty in SE Asia and South China. Until Sinodonty had developed in some population no population anywhere could be called 'Proto-Mongoloid'. Once it had developed the population containing it could then be called 'Mongoloid'. Simple? Just like animal breeding.

Races are not defined the way you define them. Your viewpoint is way too simplistic and does not reflect the real world. The real world is far more chaotic than the artificial world of animal breeding. Thus, the terminology of animal breeding cannot be applied to the real world.

That viewpoint is based an an assumption which I suggest is not be justified. The 'less derivedness' is just as likely to be the product of hybrid with some other population.

You have no evidence for your hybrid theory.

The presence of Y-DNA Q and mt-DNA X, both of which are far more likely to have originally been west Eurasian haplogroups than eastern ones (related to Y-DNA R and the western mt-DNA Ns), and the limited reach of the East Asian Y-DNA C3. The work on aDNA has been complicated by the inclusion of SE Asians as 'Mongoloids' which gives an exaggerated view of the diversity of the Mongoloid phenotype.

We do not know the original phenotypes of any of those haplogroups. This is especially the case with the Y-DNA hg Q. As for the mtDNA hg X, it constitutes only a very small and geographically pretty restricted part of the mtDNA haplogroup distribution of Amerindians.

It basically is if you extract the obviously hybrid SE Asians and southern Chinese from your overall consideration.

His Paleo-Mongoloids are an artificial construction on his part. He says they are 'less Mongoloid and less cold-adapted' and then goes on to list a string of hybrid populations, several of which are recently formed and so cannot be 'Paleo-' anything: 'Burmese, Filipinos, Polynesians, Jōmon and the indigenous peoples of the Americas'. All of these are hybrids, although you disagree as to that position for Amerindians.


You are exaggerating the non-Mongoloid admixture in SE Asia. The vast majority of populations of SE Asia are racially Mongoloid, as whatever non-Mongoloid admixture they have is quite limited. As for Amerindians, you already know what I think about them.

If something is artificial and distorted, it is your view of the Mongoloid race.

You wrote, 'the Caucasoid race is only partially-descended from the Mongoloid race (much more from its Amerindian-like varieties than from its East Asian-like varieties)' so appparently you do concede that Amerindians have a genetic component that is not Mongoloid.

On the contrary, I am saying that Caucasoids have an ancient Mongoloid component that pulls them towards both Amerindians and East Asians, but more towards Amerindians due to the fact that the relevant Mongoloid component is more Amerindian-like than East Asian-like.

Caucasoids mostly do not share a genetic component with East Asians but most do share a component with Amerindians. Surely that component cannot come from East Asians but from some other source.

Wishful thinking on your part.

Onur Dincer said...

I have discovered that you have a timing problem:

http://en.wikipedia.org/wiki/Ordos_culture

Quote:

"The human fossil remains of the Ordos Man from Salawusu site dated between 50,000 and 35,000 BCE show strong Mongoloid features, specifically on the fore-tooth and occipital bone".

'Strong Mongoloid features' at more than 35,000 years ago? That is long before humans arrived in America. If Amerindians truly were a representative sample of East Asians from the period they show a remarkably small amount of the Mongoloid phenotype that had developed in the region some long period before they left. Certainly 'Proto-Mongoloids' (of whatever sort you believe in) had become reasonable 'fully Mongoloid' by the time of Amerindian departure.


The link they provide for that claim is broken. Plus, that Wikipedia article is poorly edited.

Onur Dincer said...

Thus, the terminology of animal breeding cannot be applied to the real world.

and its concepts

Onur Dincer said...

Hamar,

Now I understand that your estimate is of recent (=post-AWE/ANE admixture) Mongoloid + ASI admixture rather than just recent Mongoloid admixture. The average recent Mongoloid admixture in the Adygei seems to be about 4% or 5% depending on the ADMIXTURE analysis. The ASI admixture of the Adygei is harder to quantify due to the fact that ASI is closer to the Caucasoid race than the Mongoloid race is and that as a result an exclusively ASI component never emerges in ADMIXTURE analyses yielding instead "South Asian" components that are partially Caucasoid. So I won't put forward a number for the average ASI admixture of Adygei and will only say that their average ASI admixture is clearly lower than their average recent Mongoloid admixture. The average recent Mongoloid + ASI admixture of Adygei may well be lower than your estimate. You may be correct in your estimate of Vologda Russians, but I am not sure, as ADMIXTURE is not a too reliable tool for such delicate recent admixture estimations.

The component I had in mind was actually Atlantic-Baltic. Compare West Asian and Atlantic-Baltic distances from East Asians, South Asians, and Siberians here ( I could cite other calculators too, if you want):

https://docs.google.com/spreadsheet/ccc?key=0ArAJcY18g2GadHZ6SHpiLTNTa3lsUmZJY2pQblVRR2c#gid=1


Populations are not components. You should notice that Northern Europeans have noticeably higher "Atlantic_Baltic" + "West_Asian" component ratio than Southern Europeans and West Asians, who have more "Southern" component ratio instead.

terryt said...

"Races are not defined the way you define them".

How do you define them? Surely physically distinct characteristics are what we use for all other species.

"The real world is far more chaotic than the artificial world of animal breeding".

Exactly the same rules apply, whether to domestic or wild animals.

"You are exaggerating the non-Mongoloid admixture in SE Asia. The vast majority of populations of SE Asia are racially Mongoloid, as whatever non-Mongoloid admixture they have is quite limited".

In keeping with the genetic scale I would place southern China at 7/8 to 3/4 Mongoloid, the remainder Australoid; Malaysia, the Philippines and most of Indonesia at 3/4 to 5/8 Mongoloid; Polynesians at 1/2 of each; Timorese at 1/2 to 3/8 Mongoloid; Australia, New Guinea and Melanesia at virtually nil Mongoloid. The cline is very steep.

"We do not know the original phenotypes of any of those haplogroups. This is especially the case with the Y-DNA hg Q. As for the mtDNA hg X, it constitutes only a very small and geographically pretty restricted part of the mtDNA haplogroup distribution of Amerindians".

Yes. We don't know phenotype of Q and I'll come back to that, but we can be fairly sure it was not Mongoloid. The Mongoloid element in Amerindians comes from their morthers' side: virtually all East Aasian haplogroups.

"You have no evidence for your hybrid theory".

The haplogroups. To mount an at all convincing case that Q is 'eastern' in origin it is first necessary to explain its almost complete absence in China. The small proportion of Q that is present is confined to the north and is entirely the derived Q1a1a1-M120. However we know the Upper Paleolithic first entered East Asia from the north. A further factor to take into account is the currently accepted phylogeny. Q splits into two groups, Q1a and Q1b. Q1b is certainly 'western', although it is possible it is a straggler there from the east. Q1a itself splits into two: Q1a1 and, the big expander, Q1a2. Q1a1 is again largely 'western' although in this case it could easily be an eastern straggler. Q1a2 has become spread from America (Q1a2a) to the Yemenite Jews (Q1a2b). Any information as to where Q1a2b is found? Ket/Selkup?

And of course although Q is descended from SE Asian MNOPS its closest relation is R. R2 is basically South Asian although also present in Central Asia. R1a and R1b (but especially the former) seem to be early Eurasian steppe haplogroups as, probably, is Q. The Mongoloid presence west of the Hindu Kush/Altai mountains is far too recent for R and Q to have been originally Mongoloid.

"The link they provide for that claim is broken. Plus, that Wikipedia article is poorly edited".

Interestingly the EDAR mutation was claimed as entering the modern human population at 35,000 years ago and undergoing a rapid selective sweep.

Anonymous said...

The ASI admixture of the Adygei is harder to quantify due to the fact that ASI is closer to the Caucasoid race than the Mongoloid race is and that as a result an exclusively ASI component never emerges in ADMIXTURE analyses yielding instead "South Asian" components that are partially Caucasoid.

As far as I know, the reason is rather that there's no pure ASI population on which to base that component. 'South Asian' in Dodecad analyses can range from 50/50 West/East Eurasian to predominantly West Eurasian. I cross-referenced the relationship of the 'South Asian' component in each Dodecad calculator with the Reich estimates of ASI is modern South Asians and Dienekes' estimates that I linked to in the last post, and found that the K=4 'Asian' component is the closest match.

So I won't put forward a number for the average ASI admixture of Adygei and will only say that their average ASI admixture is clearly lower than their average recent Mongoloid admixture. The average recent Mongoloid + ASI admixture of Adygei may well be lower than your estimate. You may be correct in your estimate of Vologda Russians, but I am not sure, as ADMIXTURE is not a too reliable tool for such delicate recent admixture estimations.

The main reason for ADMIXTURE's fluctuations in admixture estimates (assuming the same number and diversity of samples are employed each time) is the number of components involved. Admixture estimates are always going to be lower at K=12 than at K=3 (unless the admixture is *really* recent), because some of the admixture visible at K=3 has disappeared into other components by the time we reach K=12. So I tend to base my estimates of admixture on the lowest reasonable K levels. Estimates of admixture in Russians and Finns tend to be fairly consistent if one keeps this fact in mind, though the Adygei's admixture estimates tend to be more unstable.

But if the Adygei's admixture really is lower than my estimate, this would suggest a higher West Asian ANE base-level (which isn't a position I hold, btw).

Populations are not components. You should notice that Northern Europeans have noticeably higher "Atlantic_Baltic" + "West_Asian" component ratio than Southern Europeans and West Asians, who have more "Southern" component ratio instead.

True, but there are other calculators where West Asians are interpreted as mixtures of 'Caucasus' and 'West Asian' components, both of which are as East Eurasian-shifted as the 'North European' component. In addition, in the same calculators, Europeans can have more of the not-Asian-shifted 'Atlantic-Med' component than West Asians have of the not-Asian-shifted 'SW Asian' component:

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedEY4Y3lTUVBaaFp0bC1zZlBDcTZEYlE#gid=1

I don't necessarily think too much stock should be put into such inconsistent results for individual populations, but what should be kept in mind is that there are strong commonalities between the primary components in both Europe and West Asia, which suggests a common history of non-recent admixture between both regions. The components are consistent even if their application to individual populations are not.

The real question posed by ADMIXTURE is why the 'Southern' and 'Mediterranean' components are more distant from East Eurasians, even though Patterson detected a roughly proportionate signal even in the Mozabite sample, who consist almost entirely of not-Asian-shifted (relatively speaking) West Eurasians and SSAs. Likewise with Sardinians and Basques: their ANE is equal, even though, according to ADMIXTURE, Sardinians are more Southern/Mediterranean:

https://docs.google.com/spreadsheet/ccc?key=0ArAJcY18g2GadC1kRjhxcHNfSGhPYlUxbEI0VVZPR0E#gid=0

Note also in the above that the balance of Iranians in terms of Asian-shifted and non-Asian-shifted components is almost identical to the balance found in NW Europeans.

Onur Dincer said...

How do you define them? Surely physically distinct characteristics are what we use for all other species.

We discussed that issue before (more than once in fact) and I already put forward my definition of race. So I won't reiterate.

Exactly the same rules apply, whether to domestic or wild animals.

Mathematical definitions of race are impractical in wild animals (including humans).

In keeping with the genetic scale I would place southern China at 7/8 to 3/4 Mongoloid, the remainder Australoid; Malaysia, the Philippines and most of Indonesia at 3/4 to 5/8 Mongoloid; Polynesians at 1/2 of each; Timorese at 1/2 to 3/8 Mongoloid; Australia, New Guinea and Melanesia at virtually nil Mongoloid. The cline is very steep.

What is the basis of these quantifications?

Yes. We don't know phenotype of Q and I'll come back to that, but we can be fairly sure it was not Mongoloid. The Mongoloid element in Amerindians comes from their morthers' side: virtually all East Aasian haplogroups.

We cannot be sure at the moment whether the original Y-DNA hg Q carriers were non-Mongoloid. In any case, if they were really non-Mongoloid, they must have had very little, if any, impact on the autosomal genetics and phenotype of Amerindians.

The haplogroups. To mount an at all convincing case that Q is 'eastern' in origin it is first necessary to explain its almost complete absence in China. The small proportion of Q that is present is confined to the north and is entirely the derived Q1a1a1-M120. However we know the Upper Paleolithic first entered East Asia from the north. A further factor to take into account is the currently accepted phylogeny. Q splits into two groups, Q1a and Q1b. Q1b is certainly 'western', although it is possible it is a straggler there from the east. Q1a itself splits into two: Q1a1 and, the big expander, Q1a2. Q1a1 is again largely 'western' although in this case it could easily be an eastern straggler. Q1a2 has become spread from America (Q1a2a) to the Yemenite Jews (Q1a2b). Any information as to where Q1a2b is found? Ket/Selkup?

And of course although Q is descended from SE Asian MNOPS its closest relation is R. R2 is basically South Asian although also present in Central Asia. R1a and R1b (but especially the former) seem to be early Eurasian steppe haplogroups as, probably, is Q. The Mongoloid presence west of the Hindu Kush/Altai mountains is far too recent for R and Q to have been originally Mongoloid.


Your scenario is based on your erroneous Northern China/Mongolia-centric view of the Mongoloid race.

Interestingly the EDAR mutation was claimed as entering the modern human population at 35,000 years ago and undergoing a rapid selective sweep.

So what? EDAR is just a single mutation. In any case, I already declared my opinion on the poor quality of that specific Wikipedia article. It is not a good source to use as evidence of anything.

terryt said...

"What is the basis of these quantifications?"

Observation and knowledge of hybrid ratios in dabbling ducks, cattle, dogs and sheep. To believe that humans somehow a obey a diiferent set of biological rules from every other species is not justified in any way at all.

"We cannot be sure at the moment whether the original Y-DNA hg Q carriers were non-Mongoloid".

We can be fairly sure they originated west of the Altai/Hindu Kush mountains and there is no evidence of any 'eastern' element in the population at any reasonable time postulated for Q's expansion.

"In any case, if they were really non-Mongoloid, they must have had very little, if any, impact on the autosomal genetics and phenotype of Amerindians".

Around 7% according to this:

http://dienekes.blogspot.co.nz/2012/10/admixture-tracks-amerindian-like.html

And the Amerindian element is distinct from any 'East Asian' element. Although I note in the comments there that you maintain it is a Mongoloid element that is showing up.

"So what? EDAR is just a single mutation. In any case, I already declared my opinion on the poor quality of that specific Wikipedia article".

The EDAR information is not from Wiki. Try this in relation to Amerindian admixture in Europeans, once again:

http://eurogenes.blogspot.co.nz/2013/02/ancient-amerindian-like-admixture-in.html

Quote:

"Then again, maybe Europeans don't carry this mutation because by and large we don't have East Asian admixture? In other words, perhaps the mysterious eastern component shared by Europeans, Amerindians and East Asians came from a now extinct population in Siberia? If so, this component might have been picked up by the (EDAR370A carrying) ancestors of Amerindians as they moved into the Americas. In fact, perhaps whoever carried this component blocked the entry of any (EDAR370A carrying) East Eurasians proper into Europe at a critical time when population densities in West Eurasia were very low? So the other thing I'm getting at is that Amerindians and East Asians might be ancient hybrid groups, just like Europeans and South Asians".

"Your scenario is based on your erroneous Northern China/Mongolia-centric view of the Mongoloid race".

Erroneous? What makes you so sure? I am certain furture research will confirm a Northern China/Mongolia region of origin.

Onur Dincer said...

Observation and knowledge of hybrid ratios in dabbling ducks, cattle, dogs and sheep. To believe that humans somehow a obey a diiferent set of biological rules from every other species is not justified in any way at all.

It is practically impossible to make such precise admixture estimations just based on observation. In any case, that shows how weakly-grounded your admixture estimations are. Lastly, hybrid ratios in ducks, cattle, dogs and sheep have nothing to do with hybrid ratios in Southeast Asians.

We can be fairly sure they originated west of the Altai/Hindu Kush mountains and there is no evidence of any 'eastern' element in the population at any reasonable time postulated for Q's expansion.

With the current ancient DNA data we cannot be sure of anything.

Around 7% according to this:

http://dienekes.blogspot.co.nz/2012/10/admixture-tracks-amerindian-like.html


That 7% is the estimated level of Amerindian-like admixture in NW Europeans according ot that analysis, not any non-Mongoloid admixture in Amerindians.

And the Amerindian element is distinct from any 'East Asian' element. Although I note in the comments there that you maintain it is a Mongoloid element that is showing up.

The Amerindian element is distinct from the East Asian element, but not as distinct as you seem to assume:

http://dienekes.blogspot.com/2012/08/4-population-test-and-east-eurasian.html

It is clear from that analysis that Caucasoids in general, and Northern Europeans in particular, possess a Mongoloid element. From analyses involving Amerindians we know that it is primarily Amerindian-like rather than East Asian-like.

The EDAR information is not from Wiki.

By "poor quality Wikipedia article", I was referring to the Wikipedia article titled "Ordos culture" that contained information about fossil remains from Ordos that supposedly possess strong Mongoloid features, not anything to do with the EDAR mutation. It is you who sees an association between the information in that Wikipedia article and the EDAR mutation.

Try this in relation to Amerindian admixture in Europeans, once again:

http://eurogenes.blogspot.co.nz/2013/02/ancient-amerindian-like-admixture-in.html


I already offered my opinion on that issue.

Erroneous? What makes you so sure? I am certain furture research will confirm a Northern China/Mongolia region of origin.

Because it has no solid basis.

Anonymous said...

It is clear from that analysis that Caucasoids in general, and Northern Europeans in particular, possess a Mongoloid element.

The analysis didn't conclude that the nature of the admixture was 'Mongoloid' or anything else. In the terms of the Reich/Patterson/Lipson team, expressed over the course of both relevant papers, modern West Eurasians at base level are a mixture of a population closer to the roots of the original OoA population and a second population relatively closer to the ancestors of modern East Eurasians, in particular Amerindians.

Which point brings me to a fallacy you presented in the first comment to this section:

Note that the population used as the East Asian representative in the above 4-population test is the Han Chinese, which is one of the purest Mongoloid populations in the world. So it seems that northern Europeans – even excluding northeastern Europeans – are more Mongoloid-admixed than southern Europeans

The fallacy consists in not taking into account that the ANE component is more related in a general sense to all modern East Eurasians than is the AWE component. It's also the case, then, that the same test would show (and does show) the closer affinity to Australasians, ASI etc. of some non-recently admixed W. Eurasian populations than of others. But this doesn't mean that the admixture is Australoid any more than an affinity to a modern Mongoloid population means that this ancient admixture was Mongoloid. It was a component (partially?) ancestral to a plurality of modern phenotypes.

Also, Basques are a much better proxy for AWEs than are Sardinians, who are much too drifted/otherwise unique to yield realistic values for other populations (like Russians being ~50% ANE, based on their proportion relative to Sardinians plus the 25% baseline already present in Sardinians), as Lipson et al. demonstrated (and as was easily inferrable anyway from the close clustering of all Europeans in PCA analyses).

Anonymous said...

A bit of clarification: I meant that Lipson et al paper found (suspectly, IMHO) that the ANE component merged with the AWE shortly after the split of all modern E. Eurasian populations from their common ancestor, meaning, if we take the conclusion at face value, that the ANE we're dealing with had properties similar to those of a common ancestor to all groups, creating the effect of a general E. Eurasian affinity with a few peaks in affinity in the populations to which the ancient population was directly ancestral.

Onur Dincer said...

Hamar,

I already dealt with the arguments in your latest posts, so I think there is no need to write a new reply. Our correspondences have bacome rather repetitive lately.

Anonymous said...

Hamar,

I already dealt with the arguments in your latest posts, so I think there is no need to write a new reply. Our correspondences have bacome rather repetitive lately.


You didn't 'deal with' my arguments if by that you mean that you refuted them. You did offer weak responses to them, and when I demolished your weak responses, you ignored my arguments and continued to peddle the same fallacies I'd refuted many times previously. I'm sure someone explaining that two and two equals four is repetitive. However, so is the person he's responding to who insists that the answer is 17.

Keep in mind that I don't make posts only for you, but also for any readers who may be mislead by the many inaccuracies contained in your posts. Only I was able to predict the findings of Lipson et al., and the reason for this is that my reasoning was superior to yours. You have no justification for condescending to me when I'm the only person (I'm aware of) who has made accurate predicitions in this matter.

And if you continue to commit errors in pursuit of some agenda, then it is my duty to counterbalance these errors for the sake knowledge. Of course, if you admit that you've been totally wrong on the whole issue and my prediction of Lipson's paper was pure, unadulterated genius of which you're in awe, then there will be no need for future repetition :P

Meanwhile, I'll explore in more detail how your quoted argument was an egregious fallacy and bury it deep along with the rest of your points. Note that if you use this argument again, it will be direct proof of intellectual dishonesty on your part:

Take East Africans. They can be shown to be a mixture of Yoruba and any OoA population. So here are some specious arguments and their conclusions:

East African population X is a mixture of a YRI-related component and a Japanese-related component (when only these two populations are considered). Therefore East African population X is a mixture of Negroid and Mongoloid.

East African population X is a mixture of a YRI-related component and a Pauan-related component (when only these two populations are considered). therefore East African population X is a mixture of Negroid and Australoid.

Repeat with any other OoA population.

Now you'll no doubt dishonestly retort that, despite referring to Han and Dai in isolation (in your first comment to this thread and would have left that error untouched if not called out on it), and hence validating my above analogy, it can be shown that the signal is strongest in Amerindians by comparing the strength of all signals.

Again, it remains a fallacy, and here is why: Taking East Africans again as an example, suppose Western Eurasians no longer existed. The default closest modern match would therefore be whichever other population had the strongest signal of W. Eurasian ancestry; say the Uygur. Needless to say, erroneous conclusions would be drawn by such as yourself as to the physical make-up of the constributing population to East African population X.





Onur Dincer said...

There is nothing new in Hamar's latest post. He is basically repeating the same old arguments that I dealt with and/or refuted in my earlier posts (at this and other threads).

Anonymous said...

There is nothing new in Hamar's latest post. He is basically repeating the same old arguments that I dealt with and/or refuted in my earlier posts (at this and other threads).

What a child you are. Anyway, the evidence is available for each reader to assess for him or herself. Fin.

terryt said...

"Lastly, hybrid ratios in ducks, cattle, dogs and sheep have nothing to do with hybrid ratios in Southeast Asians".

In the case of ducks I am not considering 'breeds' but 'species'. The indigenous grey duck and the introduced mallard are fully interfertile and so should probably be considered regioanl variations of a single species. That is exactly comparable to the different human geographical variations except that the clines are steeper in the mallard duck species than in humans. That is presumably because technology and culture have allowed humans to cross the geographic boundaries to a large extent.

"With the current ancient DNA data we cannot be sure of anything".

I think Q's phylogeny is fairly well-established now. Under the current nomencalture Q1a1a1 and Q1a2a are the only two haplogroups found east of the Altai/Hindu Kush mountains. That is a problem for those proposing an eastern origin for the haplogroup.

"It is clear from that analysis that Caucasoids in general, and Northern Europeans in particular, possess a Mongoloid element. From analyses involving Amerindians we know that it is primarily Amerindian-like rather than East Asian-like".

How do you account for the Amerindian-like element being different from an East Asian element if Amerindians are completely Mongoloid? The argumant that East Asians have lost a basal genetic signature whereas others have kept it doesn't hold up to any scrutiny. A genetic signature could not be lost across multiple populations across a wide region. It could only be lost in a genetically isolated population. The fact that the lack is widespread supports a recent expansion of the East Asian phenotype, surely.

"It is you who sees an association between the information in that Wikipedia article and the EDAR mutation"

The link between the presence of the EDAR mutation and the presence of a Mongoloid phenotype is surely to complete to ignore, except by those who don't wish to see.

Onur Dincer said...

What a child you are. Anyway, the evidence is available for each reader to assess for him or herself. Fin.

Hamar, currently I do not have the extra time to deal with the same old arguments that I already dealt with.

@terry,

Ditto for you.

Anonymous said...

Hamar, currently I do not have the extra time to deal with the same old arguments that I already dealt with.

With respect, I'm not sure you fully understand the nuances of each new argument I present. It's possible you speed read my comments, identify some key words, and then dismiss it as the same point reiterated.

So, if you'll bear with me and not speed read this, then I'll explain how you haven't addressed the point I presented.

If I remember correctly, you stated previously that you assess the physical nature of the ANE population at the time of introgression into the WE population by the dominant features of the modern population in which the signal of admixture is strongest.

However, even though that in itself is a fallacy (explained above), you committed an additional fallacy by adducing the Han as representative of the ANE population on the basis that, using that population, a signal of admixture can be found in most W. Eurasians relative to Sardinians.

Why is it an additional fallacy? Because the signal is not particularly strong in the Han. Indeed, from Dienekes' calculations, the signal of this admixture doubles when an Amerindian population is used instead of an East Asian population. Therefore, the features of the Han or Dai are (doubly) irrelevant, as irrelevant as are the features of any other East Eurasian population (which, in the context of the admixture's age, we should define as simply 'OoA that is not AWE'), since a signal for WEs relative to Sardinians is reproducible also if using them.

I do not recall you ever addressing this. Also, I understand that the argument you used may have had some relevance in relation to genetiker's specific claims. However, you used it as a self-contained argument and later used it in a context outside of anything related to genetiker's argument.

The second point I broached in the comment you dismissed was this: If you believe the earlier proportions relative to Sardinians are correct, that Sardinians themselves are about 25% ANE, and that ANE is Mongoloid, then do you maintain that North Russians are ~50% Mongoloid?

All this said, the discussion has gone on a little too long for me, so I don't mind if you don't answer my remarks at this point.

Onur Dincer said...

Hamar,

As I pointed out, I currently have no time to redeal with your usual arguments. I agree that this discussion has gone on too long and in addition think that it has got into a vicious circle.

Simon_W said...

I know who that guy is. Last time I communicated with him, he revealed to me that he was going to learn Chinese, because he was very fond of Chinese girls. Apparently, something went wrong... (Considering that he regards the claim of Mongoloid-like admixture in Nordics as a slur.)