November 07, 2014

Genome of Kostenki-14, an Upper Paleolithic European (Seguin-Orlando, Korneliussen, Sikora, et al. 2014)

A new paper in Science reports on the genome of Kostenki-14 (K14), an Upper Paleolithic European from Russia. This is now the third oldest Homo sapiens for which we have genetic data, after Ust'-Ishim (Siberia, 45 thousand years), Tianyuan (China, 40 thousand years), and now Kostenki (European part of Russia, 37 thousand years). Of these three genomes, the Ust'-Ishim is both the highest coverage and earliest (Siberia is the gift that keeps on givin'), Tianyuan only has its chromosome 21 known, and K14, a complete 2.42x coverage sequence (and, apparently, good teeth, after all these years; left).

The publication of the Tianyuan genome showed that populations related to East Asians and Oceanians existed in the world 40 thousand years ago. So, models based on modern humans that put the split of East Asians from Europeans to a much more recent time period were basically wrong (more on this a little below). The Ust'-Ishim genome showed that populations basal to both East Asians and Europeans existed in the world 45 thousand years ago. So, either East Asians and Europeans hadn't gone along their different paths yet, or, if they had, Ust'-Ishim happened to be a side branch and not the major East Asian and European lineages.

K14 may not be the older Upper Paleolithic human, but as of this writing it is the only Upper Paleolithic European that has been published so far, the next ones being the Loschbour, Motala, and La Brana Mesolithic Europeans who who have about 1/5 of its age. The new paper shows that K14 was definitely European (or more correctly West Eurasian or Caucasoid), as it was more similar to modern Europeans than to East Asians or other non-West Eurasian populations. Thus, the morphological description of the sample as "Australoid" by some early anthropologists did not reflect its ancestral makeup. Also, this proves that Caucasoids existed 37,000 years ago, which most physical anthropologists would believe, but it is nice to have direct confirmation. This pushes the lower bound from 24,000 years ago (because MA-1 was West Eurasian according to the results of Raghavan et al.). It will be nice to push the lower bound further to the past as there are much older bones (and plenty of teeth) from earlier Upper Paleolithic Europeans.

But there is a slight kink in the story, as K14 also belonged to Y-haplogroup C which is predominantly East Asian/Ocenian/Native American today. So, maybe there is some distant link to these populations in its ancestry. But, there is definitely a link to much more recent Europeans: the tiny percentage of living Europeans who have preserved K14's Y-chromosomal type (some of which were doubtlessly told a few years back that they were descendants of Genghis Khan, before the phylogenetic structure of C was known), the La Brana hunter-gatherer from Mesolithic Spain, as well as Neolithic Europeans from Hungary.

The authors of the current paper also date the date of Neandertal admixture to 54 thousand years. This seems very compatible with the finding of between 50 and 60 thousand years by Fu et al. (2014) based on the Ust'-Ishim genome (which is both earlier and better, so the chunks of Neandertal ancestry in it are probably be longer and more well-defined).

The authors propose the following model for how various populations are related to each other:


This model is not formally tested, but at least it seems to derive Europeans as a 3-way mixture that is basically identical to that of Lazaridis et al., with some relabeling of populations (MHG=WHG and NEOL=EEF).

The model also includes Yeniseian Siberians as a mixture of MHG and East Asians (although it does not include actual East Asians). It's strange that Yeniseians apparently are given no ANE ancestry but only WHG/MHG. Both Raghavan et al. and Lazaridis et al. mentioned that ancestry related to MA-1 in living Siberians is diminished, but none at all?

The major new finding of this paper, however, is that K14 had Basal Eurasian ancestry, which was first proposed for EEF from Germany 7,000 years ago, so now it postulated for Russian hunter-gatherers 37,000 years ago. I don't think many archaeologists would derive European farmers from Russia (Russia is actually one of the last places in Europe that became agricultural). So, maybe the hunter-gatherers from Russia had Basal Eurasian ancestry and this wasn't limited to the ancestors of the EEF? If they did, it's strange that Loschbour, La Brana, MA-1, Ust'-Ishim, Swedish Mesolithic (and maybe KO1?) didn't have it. So, either Kostenki was very unique or there is an alternative explanation for its strangeness.


The evidence for the Basal Eurasian ancestry in K14 is summarized in the figure above in bullet point (b).

  • The statistic D(Mbuti, East Asia; HG, K14) is less than 0. So, there's some link between HG and East Asians. Is this because of Basal Eurasian admixture in K14 or due to some admixture between Caucasoids and Mongoloids after the time of K14? (this might cause the lower dates of European-East Asian splits alluded to above).
  • The statistic D(Mbuti, East Asia; NEOL, K14) is 0. So, East Asians don't "prefer" either Neolithic Europeans (NEOL) or K14. I guess the value of this statistic depends on how much Basal Eurasian the different populations have and what's the relationship between East Asians, K14, and the non-Basal Eurasian part in K14.
  • Finally, "NEOL component for K14 in ADMIXTURE". I think they are referring to the "Middle East" component (right). This may be Basal Eurasian ancestry, or maybe because K14 is so old, it pre-dates the European/Middle Eastern divide and its ancestry isn't attracted to either Europe or the Middle East, so it gets ancestry from both (and many other colors besides).

It is fascinating how many new questions are both answered and raised each time a new genome gets published (and there has been a constant stream of these over the last couple of years).

Science DOI: 10.1126/science.aaa0114

Genomic structure in Europeans dating back at least 36,200 years

Andaine Seguin-Orlando1,*, Thorfinn S. Korneliussen1,*, Martin Sikora1, et al.

The origin of contemporary Europeans remains contentious. We obtain a genome sequence from Kostenki 14 in European Russia dating to 38,700 to 36,200 years ago, one of the oldest fossils of Anatomically Modern Humans from Europe. We find that K14 shares a close ancestry with the 24,000-year-old Mal’ta boy from central Siberia, European Mesolithic hunter-gatherers, some contemporary western Siberians, and many Europeans, but not eastern Asians. Additionally, the Kostenki 14 genome shows evidence of shared ancestry with a population basal to all Eurasians that also relates to later European Neolithic farmers. We find that Kostenki 14 contains more Neandertal DNA that is contained in longer tracts than present Europeans. Our findings reveal the timing of divergence of western Eurasians and East Asians to be more than 36,200 years ago and that European genomic structure today dates back to the Upper Paleolithic and derives from a meta-population that at times stretched from Europe to central Asia.

Link

36 comments:

eurologist said...

I still think that the confusion is mostly over age, i.e., the inability of major researchers to agree and admit to that many "set" points in archaeology are much earlier than previously thought.

Lathdrinor said...

It is unfortunate that we do not have the Y-DNA of Tianyuan for comparison
purposes. These finds of Y-DNA C 'archaic Europeans' indicate that the genetic 'type' of Europoids was set fairly early, even before the male ancestors of 90+% modern day Europeans arrived on the scene. Yet, was C the first modern hominid bearers of this type, or a later arrival?

Grognard said...

eye and hair color known?

eurologist said...

"So, either Kostenki was very unique or there is an alternative explanation for its strangeness."

Basal Eurasian must have survived somewhere between the Levant and Afghanistan. Not Pakistan or further East, since then it surely would have admixed with other Eurasian groups. We know that the UP appeared very early in the Levant (earlier than in Europe). So, it is rather likely that when these modern people first came (from ~ Pakistan?) they admixed with Basal Eurasians in that general area. I don't know how good the time estimates of West Eurasian admixtures into "NEOL/EEF" are, but part of it could be from the UP.

If a small group of such Basal Eurasian - admixed UP West Eurasians then migrated north and admixed with UP people who took a more northerly route to Europe, the Basal Eurasian signature would have been present but quickly overwhelmed by the larger general West Eurasian population.

Mark Moore (Moderator) said...

I have a hypothesis to explain the mystery of why there was neandertal mixture early on but not later even though they lived near one another for many thousands of years. There has been some evidence for a mystery third hominid which contributed genes to Eurasians. Maybe this species interbred with both Neandertals and our forebears, but our forbears did not directly interbreed with Neandertals. We picked up some of N genes that this unknown group shared with N. Perhaps the descendants of the Skuhl and Qazfeh hominides are this middl group. Once this population went extinct, so did the gene flow between humans and Neandertals.

Gabriella Kadar said...

This skull has huge lateral incisors. Very unusual. Only seen it once in a live human from Hungary.

Tobus said...

@Grognard:

Brown eyes and dark skin: http://news.sciencemag.org/archaeology/2014/11/european-genetic-identity-may-stretch-back-36000-years.

Same as UI and MA-1. They don't say but I'd wager a $50 he had black hair too.


Mark D said...

"Also, this proves that Caucasoids existed 37,000 years ago, which most physical anthropologists would believe, but it is nice to have direct confirmation."

So Carleton Coon wasn't too far off back in 1939 before genetics.

Dospaises said...

@Grognard

The eye and hair color known might be known soon. Check http://www.fc.id.au/ periodically.

German Dziebel said...

"So, either Kostenki was very unique or there is an alternative explanation for its strangeness."

When the authors say "We find that all contemporary non-Africans, except Australo- Melanesians, are closer to either Mal’ta (MA1) or MHGs than to K14 [e.g., Z = -5.3, for D (Mbuti, Han; Loschbour, K14); SOM S9; table S10; fig. S19]. This would suggest a basal position of K14 with respect to MHGs and ancient north Eurasians..." they overlook the fact that Sub-Saharan Africans, including San, too, gravitate toward MA-1 more than they do toward K14, so it's MA-1 that must be basal to all of modern humans. "Basal European" is in reality an incipient African component and it must be derived.

So, yes, there's an alternative explanation. More at http://anthropogenesis.kinshipstudies.org/2014/11/ancient-kostenki-14-markina-gora-dna-a-glimpse-into-a-population-on-its-way-from-america-to-africa/.

Hamar Fox said...

Evidence that Basal Eurasian was a real population seems to be holding up, despite some peculiarities. On PCAs, BE admixture just seems to be 'extreme-West Eurasian', sometimes overshooting Sardinians in that respect. I've not seen PCA evidence that basally admixed ancient genomes pull toward some ghost population close to the root of modern Eurasians. I doubt we'll have resolution until an actual BE genome shows up.

Kostenki-14 and LB1 were both C, and both seem to have had hints of Basal admixture (IIRC, Lazaradis et al. mentioned finding some evidence of it in LB1, but couldn't demonstrate it convincingly).

agiering said...

Wouldn't this data be consistent with a Mesopotamian origin for the EEF/"basal" component? I could see some EEF ancestors making their way north to intermix with the K14 population,perhaps hugging the caspian/black sea on the way. And of course we know agriculture was developed early in Mesopotamia.

agiering said...

Oh, and another point:

It's interesting how they interpret K14's affinity to Mal'ta. According to their model, it wold appear that the affinity is due to their shared "West Eurasian UP" ancestry. But then how do they explain the fact that K14 is equally unrelated to East Asians and Amerindians, while MA1 is so similar to Amerindians against East Asians?

Really, this is only confirming my suspicion that postulating this hypothetical "ANE" is causing more problems that it solves. Why not just assume that Mal'ta is an admixed population, which has a native American Component [that is not shared with K14], but also not-Native American components [that are shared with K14]? No ANE necessary.

blogen said...

So finally these peoples were very archaic Cromagnoids with Australoid like traits based on their genetical background. Then presumably this was the situation in Sungir and the Grimaldi case too, where the racial makeup was similar.

Hartmut Zänder said...

I was curious, how he could look like:

http://regbands.blogspot.de/2014/11/the-smile-of-kostenki.html

Unknown said...

There are some peculiar things about chart at:
http://dienekes.blogspot.com/2014/11/genome-of-kostenki-14-upper-paleolithic.html

1. Isn't "Basal Eurasian" strange terminology if it isn't ancestral to "West Eurasian UP?" If it's "basal" how is it West Eurasian UP isn't descended from it?

2. Lazaridis et al report a "WHG" element in "Early European Farmers." The chart does not. Instead it shows a dotted line from "West Eurasian" and none from "MHG" Mesolithic Hunter Gatherers. So the chart at least at that point seriously disagrees with
Lazaridis et al .

3. Ancient North Eurasian is shown as lying in a direct line to modern European. Yet Lazaridis et al showed that ANE represents 10% or less of the genetic heritage of modern Europeans.

4. The chart seems to form a solid line from ANE towards modern European on the last line towards modern European. Does this refer to some specific aDNA that showed ANE traits after the neolithic? There is a pattern here that once again shouts grant money.

batman said...

Scandinavians are the earliest Europeans:

"From a genetic point of view he's an European," says Professor Eske Willerslev, Director of the Centre for GeoGenetics at the University of Copenhagen, who was involved in the new study, and adds:

“Actually, he is closer to Danes, Swedes, Finns and Russians than to Frenchmen, Spaniards and Germans”

http://sciencenordic.com/scandinavians-are-earliest-europeans

eurologist said...

Hartmut Zänder said...

"I was curious, how he could look like: ..."

No, not at all like that. Think of archaic people in Northern Pakistan, and Uralic people without any East Asian admixture. Think of "super-Caucasians" that are still visible in current-day Baltics and Germans.

Large, robust head, wide cheeks and widely spaced large eyes, visible eye brow bulges, large chin - all also for females, to some extent.

Tobus said...

@German: they overlook the fact that Sub-Saharan Africans, including San, too, gravitate toward MA-1 more than they do toward K14

On the contrary, you overlook that the Z-scores indicate that any difference in this regard is *not* significant, just noise.


@agiering:
But then how do they explain the fact that K14 is equally unrelated to East Asians and Amerindians

Where are you seeing this? Amerindians score higher than East Asians on the f3 and lots of the f4s.

while MA1 is so similar to Amerindians against East Asians?

Because the admixture went from MA1->Amerindian after MA1 and K14 had diverged.

Why not just assume that Mal'ta is an admixed population, which has a native American Component [that is not shared with K14], but also not-Native American components [that are shared with K14]? No ANE necessary.

Because MA-1 has next to zero East Asian affinity. If MA-1 had Amerindian admixture then he's be drawn closer to East Asians as well since Amerindians and East Asians share much of their non-European DNA. The only way to model MA-1 as admixed is to invent scenarios in the Amerindian/East Asian relationship that are less than parsimonious. Those first TreeMix runs in Lazaridis show that complicated as it may appear, an ANE population is the best model that fits the data.

On this point, the Kostenki paper provides more evidence that MA-1 didn't receive Amerindian admixture. Table SI10 shows that MA-1 is closer Amerindians than K14 is, meaning the admixture, whichever way is went, happened after K14's time and before MA-1 - if it happened before K14 then whichever way it went, K14 would be closer to Amerindians than MA1 is. With that fact in place, K14 shows no significant difference between French/Karitiana in Table SI 9, yet a strong difference between French/Han (and similarly less difference with Karitiana than Han in SI 7 and 8). Since the admixture event happened *after* K14, he should have zero pull to Karitiana if the gene flow was Amerindian->MA-1 - he was too early to get any of it, just like a white grandad doesn't get any African affinity if one of his children marries an African and he has mixed-race grandchildren. Since K14 *does* show an increased Amerindian affinity, it's shows that the gene flow went MA-1->Amerindian, pulling Amerindians towards all the ancestors and descendants of MA-1's lineage.

German Dziebel said...

@Tobus

"On the contrary, you overlook that the Z-scores indicate that any difference in this regard is *not* significant, just noise."

When the same pattern appears twice (Amerindians are less Ust-Ishim-like than East Asians and more European-like than Ust-Ishim) it's not noise. The statistical support weakens the older the sample is.

"Because MA-1 has next to zero East Asian affinity. If MA-1 had Amerindian admixture then he's be drawn closer to East Asians as well since Amerindians and East Asians share much of their non-European DNA. "

Precisely what one would expect if Europeans and Asians split from Amerindians and didn't mix much with each other thereafter. All of East Asians share their alleles with Amerindians, all of Europeans (including Stuttgart and K14) share theirs with Amerindians. A perfect pattern of a population split into two daughter populations. East Asians are closer to modern Amerindians than modern Europeans are because there likely was a secondary late Pleistocene-early Holocene (post-Malta times) migration out of North America to East Asia only. Hence northern Amerindians are closer to East Asians than southern Amerindians who are closer to Europeans. If there was gene flow from MA-1 to America after the founding East Asian gene flow into the Americas (as Raghavan and you believe), northern Amerindians would have been closer to Europeans. But they are not.

BTW, parsimony is irrelevant here: history is not parsimonious. Only myths are.

"On this point, the Kostenki paper provides more evidence that MA-1 didn't receive Amerindian admixture. Table SI10 shows that MA-1 is closer Amerindians than K14 is, meaning the admixture, whichever way is went, happened after K14's time and before MA-1 - if it happened before K14 then whichever way it went, K14 would be closer to Amerindians than MA1 is."

No, it doesn't. K14 is more geographically removed from America than MA-1, hence more genetically different from Amerindians. What's most important is that K14, just like MA-1 and modern Europeans, is closer to Amerindians than to East Asians.

"Since K14 *does* show an increased Amerindian affinity, it's shows that the gene flow went MA-1->Amerindian, pulling Amerindians towards all the ancestors and descendants of MA-1's lineage."

Since K14 *does* show an increased Amerindian affinity, it shows that the gene flow went Amerindian > K-14 (even earlier), pulling all Europeans, ancient and modern, toward the Amerindian lineage.

Tobus said...

@German:
When the same pattern appears twice

Patterns are used to extrapolate where we have no data, in cases where we do have data, just use the data.

Amerindians are less Ust-Ishim-like than East Asians

You're ignoring the Z-scores again... (and aren't we discussing K14, not UI?)

and more European-like than Ust-Ishim

Where are you seeing this? None of the UI D-stats compare UI/Karitiana against Europeans/Outgroup.

The statistical support weakens the older the sample is.

Yet lots of other populations have no trouble getting a significant result. The low Z-scores have nothing to do with age and everything to do with a lack of genuine affinity.

Precisely what one would expect if Europeans and Asians split from Amerindians and didn't mix much with each other thereafter

Incorrect. We'd expect whichever split first to be equidistant to the other two. Draw a tree and compare the branch lengths.

Hence northern Amerindians are closer to East Asians than southern Amerindians who are closer to European

You're doing your ambiguous "closer to X than Y" thing again, not specifying "than to Y" or "than Y are", and it seems to be confusing you. To clarify what you are trying to say:

Of the 4 populations, North Amerindians and South Amerindians are by far the closest to each other, indicating they were the most recent split. East Asians are closer to both North and South Amerindians than any of the 3 are to Europeans, indicating that they were next most recent split and Europeans were the first to diverge, so: Eu---[{NAm-SAm}--EAs].

On top of this we have secondary affinities that indicate post-divergence admixture: EAs and NAm share an affinity to the exclusion of both Eu and SAm, indicating admixture after the NAm/SAm split. Eu has affinity to both NAm and SAm to the exclusion of EAs, indicating an admixture event after the EA split but before the NAm/SAm split.

northern Amerindians would have been closer to Europeans. But they are not.

Not sure why you'd think that German, they appear European-shifted on the Mal'ta PCAs, the Anzick f3's and every other plot I can find them on.

No, it doesn't. K14 is more geographically removed from America than MA-1, hence more genetically different from Amerindians.

Physical location doesn't determine affinity, ancestry does. As already explained, if the admixture happened before K14 then K14 would be closer to the event than MA-1 is (and hence contain more of it). Since it's MA-1, not K14, who has the highest affinity, then MA-1 must be the closer sample to the admixture event. Draw a tree and compare the branch lengths - admixture after K14 is the only way that MA-1 can be the closest.

Since K14 *does* show an increased Amerindian affinity, it shows that the gene flow went Amerindian > K-14 (even earlier)

Taken in isolation it could mean the admixture went either way. Taken with fact that the admixture happened after K14, there's only one way the gene flow could have gone and still have an effect on K14. This direction is independently confirmed by the lack of paleo-East Asian affinity which would have accompanied any paleo-Indian gene flow into MA-1 or K14.

German Dziebel said...

@Tobus

"Patterns are used to extrapolate where we have no data, in cases where we do have data, just use the data."

There are no patterns without data. There are always gaps in the data. You are trying to imitate scientific reasoning but every time you're just misjudging what science is.

"You're ignoring the Z-scores again... (and aren't we discussing K14, not UI?)"

I'm looking at D stats. Plus Ust-Ishim seems to belong to Y-DNA hg X that leads to NO but not to QR (P), so it all fits together - UI is closer to East Eurasians than to Amerindians and/or West Eurasians. What exactly in Z stats contradicts D stats? You need to be specific.

"Where are you seeing this?"

Fig. 3 in Fu et al. 2014. You'll see that Karitiana is less UI shifted than Han or Onge but more West Eurasian-shifted (modern French) than Han or Onge. MA-1 and La Brana show the same pattern, but stronger than Karitiana supporting the inference that the Karitiana's shift away from UI-rooted East Asians is not noise.

"Of the 4 populations, North Amerindians and South Amerindians are by far the closest to each other, indicating they were the most recent split. East Asians are closer to both North and South Amerindians than any of the 3 are to Europeans, indicating that they were next most recent split and Europeans were the first to diverge, so: Eu---[{NAm-SAm}--EAs]."

Wrong reading of the data coupled with a limiting interpretative framework.

"Not sure why you'd think that German"

Looking at the data makes thinking easier, my creationist friend. Look at Raghavan et al. 2013, Fig. 3. And they specifically voice this conclusion. Anzick is closer to South Amerindians and has stronger affinity to MA-1.

"Physical location doesn't determine affinity, ancestry does."

This is called "tautology."

"if the admixture happened before K14 then K14 would be closer to the event than MA-1 is (and hence contain more of it). Since it's MA-1, not K14, who has the highest affinity, then MA-1 must be the closer sample to the admixture event. Draw a tree and compare the branch lengths - admixture after K14 is the only way that MA-1 can be the closest."

Your problem is that you keep using rigid and abstract models to fortune-tell what happened in history. MA-1 can still be closer to Amerindians than K14 if K14 went through a bottleneck in shared MA-1-Amerindian loci, while MA-1 didn't. Geographic distance increases chances for a bottleneck. K14 is further away from Amerindians but still recognizable genetically related to MA-1/Amerindians, hence it must be the bottleneck in the history of K14 and not admixture from MA-1 to Amerindians that best explains the pattern. The data pattern.

"Taken with fact that the admixture happened after K14, there's only one way the gene flow could have gone and still have an effect on K14. This direction is independently confirmed by the lack of paleo-East Asian affinity which would have accompanied any paleo-Indian gene flow into MA-1 or K14."

You've again provided a bunch of prejudices and called them "facts". It's not a fact that admixture happened after K14. Even Raghavan voiced the idea that Amerindian admixture may underlie all West Eurasians down to West Asians, hence a pre-K14 date for this admixture is plausible. Why in the world would there be paleo-Asian signature in Europeans if we already have a paleo-American signature in them? And we have a paleo-American signature in East Asians to the exclusion of Europeans.

Tobus said...

@German:


I'm sure there are, but on this particular point we *do* have data: the D-stats in Table S10 show that sub-Saharan Africans do *not* gravitate more to MA-1 than they do to K14.

What exactly in Z stats contradicts D stats?

Z-scores (not "Z-stats") don't "contradict" D-stats, they indicate to what degree the "D-score" part of the result is significantly different to that expected from chance. In table S11.1 of the UI paper, the D(Karitiana/Han;UI/Chimp) Z-score is -0.8 - WAY below significance level. This means there is no difference between Karitiana and Han affinity to UI.

I still don't know why this is relevant - we're talking about Africans not being more like MA-1 than like K14, and this data is about UI...?

Fig. 3 in Fu et al. 2014....supporting the inference that the Karitiana's shift away from UI-rooted East Asians is not noise.

Great, but I asked where you saw that Amerindians are "more European-like than UI"... I guess you just made it up.

Wrong reading of the data coupled with a limiting interpretative framework

Funny, you agreed with it last time. Are you saying that North and South Amerindians are *not* the closest two of those pops? Or that Europeans *aren't* the most distant?

Looking at the data makes thinking easier, my creationist friend. Look at Raghavan et al. 2013, Fig. 3.

"Creationist"?!? That's rich coming from you. Raghavan's Fig 3 confirms that North Amerindians are indeed shifted towards MA-1 (and hence Europeans), just like South Amerindians. The logic you are using to dismiss MA-1 gene flow into Amerindians is based on a false premise.

And they specifically voice this conclusion. Anzick is closer to South Amerindians and has stronger affinity to MA-1.

Great, but I was asking you why you are stating North Americans don't show a European shift when they plainly do. You want to quote me the text where Raghavan et al supposedly "voice this conclusion"?

This is called "tautology."

... and yet Polynesians are closer geographically to the Incas, yet more distant genetically to them, compared to the Han in Beijing. How you could confuse physical distance and genetic distance as being the same thing is beyond me.

MA-1 can still be closer to Amerindians than K14 if K14 went through a bottleneck in shared MA-1-Amerindian loci, while MA-1 didn't.

Your problem is that you just "make shit up" when the facts don't fit your theory. Your theory already requires that North/South Amerindians are in fact the most divergent population in the world but subsequent admixture has magically "hidden" this fact; that UI is really closer to Amerindians but archaic admixture has magically "hidden" this and made him look ancestral to all Eurasians; and now you are claiming that K14 is actually closer to Amerindians than MA-1 is but a bottleneck has "hidden" this fact. If you have evidence than K14 underwent a bottleneck that MA-1 didn't then bring it, otherwise cut the circumstantial reasoning crap.

Raghavan voiced the idea that Amerindian admixture may underlie all West Eurasians down to West Asians

Raghavan also voiced that the data completely and utterly rejects this idea, funny that you'd forget to mention that! A pre-K14 date for this admixture is also rejected by the data... if your theory needs it to work then your theory is wrong and you should change it.

Ebizur said...

If I am not mistaken, the Ust'-Ishim specimen does not belong to haplogroup X, but rather to a basal branch of haplogroup K2a(xNO'X). In other words, haplogroup NO-M214 and the Telugu haplogroup X are (slightly) more closely related to each other than either is related to the Ust'-Ishim specimen's Y-DNA.

Likewise, haplogroup R1-M173 and haplogroup R2-M479 are more closely related to each other than either is related to the Mal'ta specimen's Y-DNA.

It seems that neither the Ust'-Ishim specimen's patrilineage nor the Mal'ta specimen's patrilineage has produced any direct male descendant in extant human populations. However, a fairly close patrilineal relative of the Ust'-Ishim specimen has produced both haplogroup X and haplogroup NO, and a fairly close patrilineal relative of the Mal'ta specimen has produced both haplogroup R1 and haplogroup R2.

In the Mal'ta specimen's case, that relative who has produced R1 and R2 seems to have lived somewhat earlier than the Mal'ta specimen himself; in other words, both pre-R1 and pre-R2 probably already existed as separate lineages at the time when the Mal'ta specimen lived, but the pre-R1 individual and the pre-R2 individual shared a more recent common patrilineal ancestor with each other than either shared with the Mal'ta specimen.

As for the Ust'-Ishim specimen and haplogroup NO'X, the MRCA of NO and X seems to have been roughly contemporaneous with the Ust'-Ishim specimen, but again, the Ust'-Ishim specimen belongs to a different (though closely related) patrilineage.

German Dziebel said...

@Tobus

"sub-Saharan Africans do *not* gravitate more to MA-1 than they do to K14."

They do. See here, with tables, http://anthropogenesis.kinshipstudies.org/2014/11/ancient-kostenki-14-markina-gora-dna-a-glimpse-into-a-population-on-its-way-from-america-to-africa/.

So far it looks like that the so-called "ANE" is the earliest modern human lineage whose frequencies decrease with geographic distance from America.

"where you saw that Amerindians are "more European-like than UI"

Modern Europeans are closer to Amerindians than they are to ancient and modern East Asians (Raghavan's data provided above). UI is closer to modern East Asians than to modern Europeans (Fu et al.'s data). This means that Amerindians are more "European-like" than UI. The data from Fu et al. 2014 (D stats in Fig. 3) supports this inference by showing that UI shares more alleles with Han and Onge than with Karitiana. Karitiana is consistently less Onge-like and less Han-like in its relationship to UI. Also, UI is also slightly more French-like when compared with Karitiana than when compared with Onge or Han.

"North Amerindians are indeed shifted towards MA-1 (and hence Europeans), just like South Amerindians. The logic you are using to dismiss MA-1 gene flow into Amerindians is based on a false premise."

yes, all Amerindians have a special affinity with MA-1 on a worldwide scale and with West Eurasians when the latter are also compared with East Eurasians . MA-1 is the closest to Amerindians out of all modern human populations. Northern Amerindians tend to be less MA-1-like than are Southern Amerindians. East Asians are, on the contrary, closer to Northern Amerindians than to Southern Amerindians.

"why you are stating North Americans don't show a European shift when they plainly do."

See above. Not sure when exactly you began to misinterpret what I've always stated.

"Your problem is that you just "make shit up" when the facts don't fit your theory."

My theory is borne of facts. I don't need to make shit up. Not all pieces of the puzzle are in place but all available facts are consistent with a version of out of America.

"Your theory already requires that North/South Amerindians are in fact the most divergent population in the world but subsequent admixture has magically "hidden" this fact."

That's what historical reconstructions are all about: one first needs to sort out later admixture effects before building theories about ancient common descent. Geneticists work in reverse and hence arrive at pseudo-historical ideas. Yes, Amerindians are the most divergent among modern human populations once admixture with archaic hominins in Papuans and Africans is excluded. Since there were no archaic hominins in America, the divergence of modern Amerindians is fully due to modern human divergence, while Papuans and Africans happen to be on "archaic steroids." yes, Amerindians would have been even more divergent if Northern and Southern Amerindian demes didn't expand and mix beginning around 12,000 years ago but even with that mixture having taken place, Fst among Amerindian groups is absolutely highest among modern human populations, while heterozygosity is absolutely lowest suggesting that if Amerindians are admixed, they are a product of a continent-internal mix of Southern and Northern Amerindians and not of East Eurasians and West Eurasians, which have a completely different, more "evolved" demographic profile.

German Dziebel said...

@Tobus (contd.)

"UI is really closer to Amerindians but archaic admixture has magically "hidden" this and made him look ancestral to all Eurasians"

Again, this is just from the data: Amerindians have retained the longest Neandertal chunks after UI (with 45,000 years separating them). But since UI is so heterozygous, while Amerindians and Neandertals are so homozygous, it must be that UI is a mix of ancient East Eurasian and West Eurasian populations that got formed after they had expanded out of the Americas.

"now you are claiming that K14 is actually closer to Amerindians than MA-1 is but a bottleneck has "hidden" this fact."

the other way around. MA-1 is closer to Amerindians than K14 which is consistent with K14 being further removed geographically from the New World source of MA-1. K14 must have undergone a bottleneck in those loci shared between MA-1, K14 and Amerindians. K14 is just chronologically attested earlier than MA-1 but genetically it's more derived than MA-1 when it comes to the founding Amerindian loci.

"and yet Polynesians are closer geographically to the Incas, yet more distant genetically to them, compared to the Han in Beijing. How you could confuse physical distance and genetic distance as being the same thing is beyond me."

Many things from the world of science are beyond you. We are talking about geographies that have accommodated plausible routes of human migrations not random distances. With geographic distance, the frequencies of the original pool of alleles decay because of successive bottlenecks. At the same time, other alleles may get absorbed into an advancing population if the latter mixes and replaces pre-existing populations.

Tobus said...

See here, with tables, http://anthropogenesis....

That blog is written by a crackpot fringe theorist - I recommend you take anything it says with a large grain of salt.

Modern Europeans are closer ....

... so you didn't "see" it in the data, you "inferred" it (ie "made it up").

The data from Fu et al. 2014 (D stats in Fig. 3) supports this inference by showing that UI shares more alleles with Han and Onge than with Karitiana.

The source of this chart (see S11) shows that there is no significant difference in the amount of alleles that UI shares with Han, Onge and Karitiana (|z-scores| < 2).

yes, all Amerindians have a special affinity with MA-1 on a worldwide scale
Not sure when exactly you began to misinterpret what I've always stated.

When you said "If there was gene flow from MA-1 to America... northern Amerindians would have been closer to Europeans. But they are not." They are in fact European-shifted, as you just agreed.

My theory is borne of facts. I don't need to make shit up.

And yet you do... where is the *factual* basis for K14 undergoing a bottleneck? Where is the *factual* basis for UI having additional archaic admixture? Where is the *factual* basis for deep divergence between North and South Amerindians? None of these ideas are "borne of facts", they are all circumstantial extrapolations you had to make up to keep your theory from falling to pieces.

all available facts are consistent with a version of out of America.

But a version of out of America that requires a whole bunch of other "facts" that are both implausible and non-falsifiable, and hence pseudo-scientific.

Consider the statement "pigs can fly, they just don't do it when people are looking". It might be 100% "consistent" with "all available facts" (no-one's ever seen a pig fly!), but it requires a highly implausible and non-falsifiable premise to make it work. Just like when psychics say that their talents are genuine but the "negative energy" of scientific testing conditions stops them working - again, consistent with the facts but pseudoscience because it's highly implausible and non-falsifiable. The same "logic" is used by your pet creationists - God *could* have created the world 6,000 years ago and planted "fossils" that make it look much older, but that's highly implausible and non-falsifiable, and hence has no place in a rational discussion.

You are doing the same thing - it's patently obvious that any MA-1/Amerindian admixture happened after K14, but you invent an implausible and non-falsifiable premise to avoid that fact. I ask you honestly, are you serious about having rational discussion of your ideas? Or do have a set belief that you will hold until you die regardless of any evidence against it? If the former then please specify your falsification criteria - what evidence would cause you to reject Out Of America?

Yes, Amerindians are the most divergent among modern human populations

No they are not - there is more divergence between Europeans and East Asians (for example) than there is between North and South Americans. You are talking about a different kind of divergence, "stupid!".

Fst among Amerindian groups is absolutely highest among modern human populations..

Proving that there has been no significant admixture among them.

...while heterozygosity is bsolutely lowest

Proving that their divergence is the most recent.

If Amerindians had deep divergence and a lot of admixture we'd expected to see high heterozygosity and low between-group Fst... a bit like what we see in Africa :)

K14 is just chronologically attested earlier than MA-1 but genetically it's more derived than MA-1 when it comes to the founding Amerindian loci.

A contradiction that disappears completely if you approach it without a "founding Amerindian loci" assumption.

German Dziebel said...

@Tobus

"That blog is written by a crackpot fringe theorist - I recommend you take anything it says with a large grain of salt."

I wrote it having two doctorates and two books at hand. I don't care if I'm fringe. I like it here. I don't receive millions of emails from ne'er-do-wells like you asking me to tell them how they originated. I'll make dozens of big mainstream folks fringe at some point. And now I'm having a good time exposing nameless creationists such as yourself. Before calling me a crackpot, look at a "science" that derives "truth" from a handful of randomly found, broken skulls. I think I'm going to enrich your native English language with a new term "crackskull": crackpots who divine human origins on the basis of cracked skulls. Go check Urban dictionary. It's already there!

"The source of this chart (see S11) shows that there is no significant difference in the amount of alleles that UI shares with Han, Onge and Karitiana (|z-scores| < 2)."

When we are talking about 45,000 year old specimen, even little things matter. That method becomes less and less reliable as time goes by, so don't expect much from it. But appreciate a little.

"When you said "If there was gene flow from MA-1 to America... northern Amerindians would have been closer to Europeans. But they are not." They are in fact European-shifted, as you just agreed."

northern Amerindians are less European-shifted than southern Amerindians. Just study the basics and don't create confusion!

"And yet you do... where is the *factual* basis for K14 undergoing a bottleneck? Where is the *factual* basis for UI having additional archaic admixture? Where is the *factual* basis for deep divergence between North and South Amerindians? None of these ideas are "borne of facts", they are all circumstantial extrapolations you had to make up to keep your theory from falling to pieces."

1. K14 has less ANE/Amerindian than MA-1. It's further removed from America geographically. Geographical distance increases chances of bottlenecks; 2) what additional archaic admixture?; 3) North and South Amerindians are more divergent from Eurasians than East Asians are from Europeans. If they are not as divergent from each other as East Asians are from Europeans, it's because during the same amount of time they have been admixing with each other, while East Asians and Europeans have not. It's very simple. But apparently not for you!

"Proving that there has been no significant admixture among them."

No, proving that there was no admixture between East Asians and Europeans to generate Amerindians. Also proving that there was admixture within America, otherwise Fst would have been around N/D values.

"Proving that their divergence is the most recent."

No, proving that there were no archaic hominins in America for modern humans to admix with - this is fully consistent with paleoanthropological data. After 1492 heterozygosities in America spiked - so by your pseudoscientific logic humans originated in the Americas because some populations there now have higher or at least the same heterozygosities as Sub-Saharan Africans. Plus you are the one who argued that heterozygosities don't mean anything when it comes to population descent. More pseudoscience from Tobus!

"A contradiction that disappears completely if you approach it without a "founding Amerindian loci" assumption."

There's no contradiction. Just get used to the fact that archaeological finds are random. MA-1 was found in Siberia and not in America, K14 was found in Europe and not in Siberia. Look into the underlying genetics, not the dates!

Tobus said...

@German:
Before calling me a crackpot,look at a "science" that derives "truth" from a handful of randomly found, broken skulls

As opposed to a "science" that derives "truth" from a handful of random, broken skulls that *HAVEN'T* been found?!? Madness!

That method becomes less and less reliable as time goes by, so don't expect much from it.

It was invented to compare Neanderthal samples just as old as UI and was also used on the Denisovan samples that are even older than UI... no problem in reliability there, and no problem with reliability with UI's results for every other modern population. Age is not an issue, I think you're just going to have to grow a pair and admit you made a mistake German - nothing to be ashamed of, they plotted the non-significant samples along with the significant ones and you're not the only one to make the same error.

northern Amerindians are less European-shifted than southern Amerindians

Ahh I see, ambiguous wording - I can understand how *no* European shift would contra-indicate MA-1 gene flow, but how does only being "less European-shifted" prohibit them from having MA-1 gene flow? The "less" is very slight (Anzick is still their closest ancestor) and we know there's been post-Anzick Saqqaq/Siberian/East Asian-like gene flow into North America which would obviously have ameliorated the MA-1 component to some degree.

1. ... Geographical distance increases chances of bottlenecks

I asked for "factual basis" - the "chance of a bottleneck" is not a fact, it's supposition. Is there any *actual empirical data* that suggests K14 really has undergone a bottleneck - a significantly lower heterozygosity than MA-1 perhaps - or did you just make it up because your theory requires it?

2) what additional archaic admixture?

You said UI is central to all Eastern Eurasians because "it's admixed between ancient East Asians like Tianyuan.. ancient West Eurasians... and possibly Neandertals". TreeMix draws a very different picture however. Is there any *factual* evidence behind your alternative proposal - a different TreeMix run for example - or did you just make it up to fit your theory?

3) North and South Amerindians are more divergent from Eurasians than East Asians are from Europeans

That's not true (all Amerindians are closer to East Asians than Europeans are), but it's also not what I asked - an "Amerindian admixture into MA-1" theory requires that the divergence between North and South Amerindians is as great or greater than the divergence between MA-1 and East Asians which it clearly isn't. You cover this up by saying they have since admixed and obliterated any such divergence. My point is that there is no data that suggests North and South Amerindians where once the most diverged populations on the planet (not to mention Fst/heterozygosity results contradict it), it's something you just made up to fit your theory.

This kind of argument is never-ending - you can always invent some new implausible and non-falsifiable theory to explain away data inconsistencies. You ignored the most important question in my last post, so I'll ask it again: Is there any possible evidence that would cause you to reject Out of America, or is it a belief you will carry to your grave, regardless of the evidence?

There's no contradiction

No? The sample that's supposedly *closer* to Amerindians in ancestry has the
*lower* affinity to them? If that's reasonable then the whole concept of using DNA to detect ancestry just went out the window - we can all just make up any old associations we like and then claim a bottleneck made it look like something else.

Just get used to the fact that archaeological finds are random

Random selection tends to the mean. Until we get a subsequent sample that can help calibrate it, it's unreasonable to assume a random sample represents anything but the average.

German Dziebel said...

@Tobus

"As opposed to a "science" that derives "truth" from a handful of random, broken skulls that *HAVEN'T* been found?!? Madness!"

Lie! I specifically do not derive "truth" from skulls whether existing or not. Did I ever say that the absence of skulls in America proves that humans originated there?

"It was invented to compare Neanderthal samples just as old as UI and was also used on the Denisovan samples that are even older than UI... no problem in reliability there, and no problem with reliability with UI's results for every other modern population. "

Irrelevant objection. The older the sample, the more chances subsequent bottlenecks in the history of those individual sampled populations diluted the results. Recall: Amerindians are closer to Europeans than east Asians are, UI is closer to East Asians than to Europeans. UI must be less UI than East Asians and more shifted toward Europeans. The D stat - significant statistically or not - illustrates the inference well. You can throw away the D stat pattern if you believe in the power of SI to differentiate between signal and noise in all cases, but the former pattern still holds.

"how does only being "less European-shifted" prohibit them from having MA-1 gene flow? "

It was postulated that MA-1 gene flow postdated East Asian divergence. So North America should be more MA-1 shifted than South America. The observed pattern is the opposite from this.

"I asked for "factual basis" - the "chance of a bottleneck" is not a fact, it's supposition." K14 is less ANE like than MA-1 and it's further removed from America geographically. It's factual enough. You can't have science without inference.

"TreeMix draws a very different picture however."

Figure S10.5. in Fu et al. shows UI between Europeans and East Asians with an additional contribution from Neandertals. What's not compatible with my admixture hypothesis?

"That's not true (all Amerindians are closer to East Asians than Europeans are)..."

Which PCA are you using? Mind you, there was a "Mongoloid" migration out of the Americas in the last 12,000 years which brought Amerindians and (north) East Asians closer together on some PCA axes (another example of why geographic distance is a good predictor of genetic distance - more chance of gene flow).

"an "Amerindian admixture into MA-1" theory requires that the divergence between North and South Amerindians is as great or greater than the divergence between MA-1 and East Asians which it clearly isn't."

No, a continental population that stayed behind and kept internally admixing would look less divergent than two populations that diverged from that common source and obviously from each other but did not subsequently admix with each other. You keep using greater divergence as a sign of antiquity, which is a flawed assumption.

"My point is that there is no data that suggests North and South Amerindians where once the most diverged populations on the planet (not to mention Fst/heterozygosity results contradict it),"

Fst clearly shows that Amerindian populations are already the most divergent populations. The hypothesis of an admixture between two large regional groups (seen in mtDNA and Y-DNA) logically implies that prior to this event of admixture the differences must have been even greater. N/D heterozygosity data clearly shows the baseline from which Amerindians have evolved.

German Dziebel said...

@Tobus (contd.)

"This kind of argument is never-ending - you can always invent some new implausible and non-falsifiable theory to explain away data inconsistencies. "

No, this applies to out of Africa and into the Americas. Remember how your story goes: Amerindians split from East Asians and went through a bottleneck. They they ran into a wayward band of West Eurasians in Siberia, admixed with them and grew in size, then went through another bottleneck while trapped in an Atlantis-like sunken continent of Beringia. Every next step in the story is supposed to obliterate the lack of factual support for a preceding one using the absence of a different kind of fact.

"Is there any possible evidence that would cause you to reject Out of America, or is it a belief you will carry to your grave, regardless of the evidence?"

It's too premature for this. We need to first marshall objective support for out-of-America to be able to match the support that has been marshaled for out-of-Africa. I've done most of it but there are still some gaps. Once we agree that both models work under specified sets of assumptions, we can formulate falsifiability criteria for both. Out-of-Africa hasn't generated any for itself.

"The sample that's supposedly *closer* to Amerindians in ancestry has the *lower* affinity to them?"

Which one are you talking about?

"Random selection tends to the mean. Until we get a subsequent sample that can help calibrate it, it's unreasonable to assume a random sample represents anything but the average."

How does it apply here when we are talking about K14 being older than MA-1 chronologically without being ancestral to it genetically?

Tobus said...

@German:
Did I ever say that the absence of skulls in America proves that humans originated there?

You've missed the point German - your theory does not come from *any* archaeological or genetic facts, just a pet theory and your imagination.

Amerindians are closer to Europeans than east Asians are, UI is closer to East Asians than to Europeans. UI [typo! assuming Amerindians] must be less UI than East Asians and more shifted toward Europeans.

That's bad logic - the "must" is a non sequitur. Consider the same logic with different pops: South Amerindians are closer to Europeans than East Asians are. North Amerindians are closer to East Asians than Europeans are. South Amerindians "must" be less North Amerindian than East Asians???!?

A good sanity check when using logic is that if your conclusion contradicts the data, then you made a mistake.

It was postulated that MA-1 gene flow postdated East Asian divergence. So North America should be more MA-1 shifted than South America.

Sorry I don't follow - wouldn't North/South Amerindians be expected to have the *same* MA-1 in that scenario?

Which PCA are you using?

All of them, plus D-stats, ADMIXTURE and TreeMix.

No, a continental population that stayed behind and kept internally admixing would look less divergent than two populations that diverged from that common source

Yes, that's the theory you made up - now show me the *empirical data* that suggests it *actually* happened in America. With only modern data to go on this is going to be a self-fulfilling scenario, the admixture has destroyed all the evidence! So you see my point? It's an unnecessary implausible and non-falsifiable proposition, and as such is unreasonable on the level of "we all see aliens but they wipe our memory so we forget". Fine if you want a quasi-religious belief to keep you happy till you die, but not acceptable in a rational discussion.

N/D heterozygosity data clearly shows the baseline from which Amerindians have evolved.

N/D heterozygosity shows how low an inbred population can go before it dies out. Typical archaic homo heterozygosity was undoubtedly much higher than this.

No, this applies to out of Africa and into the Americas.

Every bottleneck postulated in OOAfrica comes from an observed drop in the heterozygosity data, whereas you've just made yours up.

It's too premature for this. We need to first marshall objective support for out-of-America

An expectedly non-committal excuse. OOAmerica is already falsified in any rational person's mind, so you'll be waiting forever to marshall your "objective support"... I guess that suits you though.

Once we agree that both models work under specified sets of assumptions, we can formulate falsifiability criteria for both

OK, I agree that OOAmerica will work under a specific set of (false!) assumptions... so what's your falsifiability criteria?

Out-of-Africa hasn't generated any for itself.

If a random 100-200kya skull (broken or not) turned up in America, Australia or Antarctica tomorrow and it clustered closer to modern humans than any African skull, I can guarantee that OOAfrica would be dropped like hot rocks.

Which one are you talking about?

You are suggesting that between MA-1 and K14, K14 is *closer* to Amerindians ancestrally yet has the *lower* affinity to them - contradicting every inference of ancestry from affinity ever made in the entire history of genetics. PCA, f3, ADMIXTURE, TreeMix, D-stat, Paternity Tests etc. all work off an assumption that higher affinity = closer ancestry.

The only reasonable interpretation of the data is that MA-1 is closer ancestrally to Amerindians, and hence the admixture between Amerindians and MA-1 happened after K14's time.

German Dziebel said...

@Tobus

"your theory does not come from *any* archaeological or genetic facts, just a pet theory and your imagination."

It's based on facts and logic. Not on broken skulls. But to a creationist like you ancient relics will always matter more.

"A good sanity check when using logic is that if your conclusion contradicts the data, then you made a mistake."

Your counterexample made no sense.

"Sorry I don't follow - wouldn't North/South Amerindians be expected to have the *same* MA-1 in that scenario? "

You give a question to a question, instead of an explanation to a problem. You can't even communicate normally.

"Yes, that's the theory you made up - now show me the *empirical data* that suggests it *actually* happened in America."

There's both modern and ancient DNA data to support it. And I gave it to you. I know it's hard to swallow.

"All of them, plus D-stats, ADMIXTURE and TreeMix."

Let's start with a PCA. Which one is it?

"Typical archaic homo heterozygosity was undoubtedly much higher than this."

Good job making stuff up.

"Every bottleneck postulated in OOAfrica comes from an observed drop in the heterozygosity data"

There's no observed drop in heterozygosity. There's heterozygosity gain progressively out-of-America. Just like there's heterozygosity gain out of Taiwan for Austronesian-speakers.

"OOAmerica is already falsified in any rational person's mind"

I'm happy to see you admit that "falsification" is a subjective process in your case.

"If a random 100-200kya skull (broken or not) turned up in America, Australia or Antarctica tomorrow and it clustered closer to modern humans than any African skull, I can guarantee that OOAfrica would be dropped like hot rocks."

You can keep waiting for the second coming of Jesus. I prefer to do science now.

"I agree that OOAmerica will work under a specific set of (false!) assumptions"

OK, so you admit that OOAmerica will work... And we just established that your criteria of falsifiability are purely subjective. As such they can be dismissed as a voice of God in your head. What's left is that "OOAmerica will work..."

"You are suggesting that between MA-1 and K14, K14 is *closer* to Amerindians ancestrally yet has the *lower* affinity to them - contradicting every inference of ancestry from affinity ever made in the entire history of genetics."

No, AFAIR, I said that K14 is derived from Amerindians just like MA-1 is. The reason for the fact that K14 is less Amerindian than MA-1 is because it's more geographically removed from America. So it lost some of its Amerindian affinity through a bottleneck and replaced it with new alleles.

"The only reasonable interpretation of the data is that MA-1 is closer ancestrally to Amerindians, and hence the admixture between Amerindians and MA-1 happened after K14's time."

No, it's a matter of geography, not time. Why would a K14-like population carry Amerindian alleles in the first place and then progressively grow them on the way through Siberia (as MA-1) to America? You clearly carry in your mind the biblical pictures of Jews seeking a promised land. But I have a piece of news for you, boy: Sunday school is over.

Tobus said...

@German:
It's based on facts and logic

Things you just make up aren't "facts" German.

But to a creationist like you ancient relics will always matter more

Oxymoron.

Your counterexample made no sense.

That's the point - the logic was identical to yours.

You give a question to a question, instead of an explanation to a problem. You can't even communicate normally.

Don't be daft - in a scenario where MA-1 gave DNA to Amerindians before North/South diverged, both would have the same degree of MA-1 affinity.

There's both modern and ancient DNA data to support it

MA-1 DNA shows no divergence, while Anzick shows a little bit that's still present between modern samples... the DNA suggests North/South diverged sometime after MA-1 and haven't come back together since.

You can keep waiting for the second coming of Jesus. I prefer to do science now.

Really? You're going to say the theory is falsified *before* the falsification criteria are met?!?! Nice work Einstein.

OK, so you admit that OOAmerica will work... And we just established that your criteria of falsifiability are purely subjective.

Don't be a dick, just answer the question. Is there any evidence that would cause you to rethink Out Of America, or are you committed to believing it until you die regardless of the evidence against it?

No, AFAIR, I said that K14 is derived from Amerindians just like MA-1 is.

K14 is some 15,000 years earlier, so "just like" is really "15,000 years closer"... and yet the DNA says the exact opposite - MA-1 is closer. Rather than accept the obvious, you make up some implausible (but ultimately non-falsifiable) bottleneck that only reduces Amerindian affinity and so makes K-14 look less Amerindian than his ancestry would suggest... a theory that totally undermines the whole basis of determining ancestry via genetic affinity in the first place.

Why would a K14-like population carry Amerindian alleles in the first place

It wouldn't - it'd carry West Eurasian alleles that an MA-like population gave to paleo-Amerindians some 15,000 years later. This is the only reasonable way to interpret Amerindians having more affinity to MA-1 than to K14... your interpretation requires a suspension of disbelief on the order of psychics and aliens.

German Dziebel said...

@Tobus

"Things you just make up aren't "facts" German."

You are the one who makes things up. That's what creationists do.

"K14 is some 15,000 years earlier, so "just like" is really "15,000 years closer"... and yet the DNA says the exact opposite - MA-1 is closer. Rather than accept the obvious, you make up some implausible (but ultimately non-falsifiable) bottleneck that only reduces Amerindian affinity and so makes K-14 look less Amerindian than his ancestry would suggest... a theory that totally undermines the whole basis of determining ancestry via genetic affinity in the first place."

Kostenki is older but further removed geographically from America than MA-1. Time and geography combined generate the observed reality.

" Is there any evidence that would cause you to rethink Out Of America, or are you committed to believing it until you die regardless of the evidence against it?"

Typical creationist question. Is there anything that would make you drop your religion and join mine? Get lost.

"It wouldn't - it'd carry West Eurasian alleles that an MA-like population gave to paleo-Amerindians some 15,000 years later. This is the only reasonable way to interpret Amerindians having more affinity to MA-1 than to K14... your interpretation requires a suspension of disbelief on the order of psychics and aliens."

Karitiana is the most MA-1-like population out of all modern human populations. So what you observe in MA-1 is Amerindian admixture. You want me to believe that Eurasians carried this Amerindianness in a dropper across Eurasia? Keep reading the Bible for inspiration to interpret human prehistory.

Tobus said...

@German:
Kostenki is older but further removed geographically from America than MA-1. Time and geography combined generate the observed reality.

The "observed reality" is that MA-1 is closer to Amerindians than K14 is, and thus any Amerindian/West Eurasian admixture event took place closer to 24kya than to 36kya.

The unobserved fantasy is that this admixture event actually took place *before* 36kya and a bottleneck just makes it look like it happened later.

You can't expect any rational person to accept the second scenario when the first scenario is a perfect fit for the data. Apart from it conflicting with your predetermined beliefs, do you have a rational reason (with empirical evidence!) to reject the first scenario?

Is there anything that would make you drop your religion and join mine?

As I said before, if a 200+kya skeleton turned up in America that clustered closer to modern humans than African fossils do, then I'd have no choice but to change my mind.

I take it that your continued refusal to provide falsifiability for your theory means that it really is quasi-religious and there's really no point me trying to talk sense to you - you're never going to question it regardless of the facts.

Karitiana is the most MA-1-like population out of all modern human populations. So what you observe in MA-1 is Amerindian admixture.

On the surface you might think so - but MA-1 doesn't have the East Asian affinity that Amerindian admixture would have given him, and K14 has a slight "Amerindian shift" that wouldn't be there if MA-1 received the admixture, so it can't be the case.

Think outside your mental constraints for a bit and consider that what you see as "Amerindianness" might really "paleo-Eurasianness" that proto-Amerindians picked up before they got to America.