March 28, 2009

Quantifying the extent of possible Neandertal introgression

The absence of Neandertal mtDNA sequences in modern Europeans argues against large-scale admixture between Cro-Magnoids and Neandertals. On the other hand, this could be explained if there was a selective advantage of Cro-Magnoid mtDNA types. Thus, the survival of ancient Cro-Magnoid (but not Neanderthal) mtDNA into modern Europeans does not in itself demonstrate the absence of admixture, although it certainly makes it more likely.

mtDNA is implicated in energy production of the body, and a selective advantage is not inconceivable, perhaps associated with fluctuations in temperature that may have weeded out cold-adapted Neanderthal mitotypes.

Alternatively, Neanderthal mothers may have been at a disadvantage when it came to bearing offspring from Cro-Magnoid fathers, due to some incompatibility of the genomes (or body types) of the two species, a stronger version of what happens in modern humans of different races.

Ultimately, detecting Neandertal introgression will have to involve the detection of a genetic variant in Neandertal remains, that is not found in the earliest Cro-Magnoid remains, but is found in modern humans. Given the very small samples of both species, this may, in fact be quite difficult.

Thus, I don't anticipate the question of Neandertal introgression to be settled any time soon, although the continued absence of a "smoking gun" may ultimately lead scientists to accept the absence (or quantitative insignificance) of any such admixture event.

Heredity doi:10.1038/hdy.2008.103

Comparing models on the genealogical relationships among Neandertal, Cro-Magnoid and modern Europeans by serial coalescent simulations

E M S Belle et al.

Abstract

Populations of anatomically archaic (Neandertal) and early modern (Cro-Magnoid) humans are jointly documented in the European fossil record, in the period between 40 000 and 25 000 years BP, but the large differences between their cultures, morphologies and DNAs suggest that the two groups were not close relatives. However, it is still unclear whether any genealogical continuity between them can be ruled out. Here, we simulated a broad range of demographic scenarios by means of a serial coalescence algorithm in which Neandertals, Cro-Magnoids and modern Europeans were either part of the same mitochondrial genealogy or of two separate genealogies. Mutation rates, population sizes, population structure and demographic growth rates varied across simulations. All models in which anatomically modern (that is, Cro-Magnoid and current) Europeans belong to a distinct genealogy performed better than any model in which the three groups were assigned to the same mitochondrial genealogy. The maximum admissible level of gene flow between Neandertals and the ancestors of current Europeans is 0.001% per generation, one order of magnitude lower than estimated in previous studies not considering genetic data on Cro-Magnoid people.

Link

7 comments:

Ponto said...

I don't think the Cro Magnoids which are dated to only about 30 kya, older remains turn out to be Neanderthal, are in the ancestry of modern Europeans. Modern Europeans descend from people who lived in Europe after the LGM, and those people were not Cro Magnoids. European derived haplogroups of the Near Eastern mtDNA H are H1 and H3, and V are dated to about 11 to 15 kya. Haplogroup U5a is much older but found in mainly in areas of Europe that became habitable after the LGM, so an extra European origin is likely.

The Italian Cro Magnoid remains, one with mtDNA CRS H2b, is unreliable. The remains are dated to 28 kya and mtDNA H2b, CRS to half that age. Doesn't add up.

Jim said...

The Italian Cro Magnoid remains, one with mtDNA CRS H2b, is unreliable. The remains are dated to 28 kya and mtDNA H2b, CRS to half that age. Doesn't add up.

There may be many reasons to doubt any ancient DNA sequences, but dating disagreements should not be one of them. Sequence-based age estimates are at best an educated guess. A 2-fold mis-calculation could be trivial on a molecular dating scale, depending on the method used.

Anne Gilbert said...

Why, I wonder, hasn't anyone seriously considered population disparities in all of this? It appears that Neandertal populations were rather small and (relatively) isolated, which means that admixtures could well have taken place, but if you do the math, the chances of leaving a genetic trace, begins to get pretty small. Bear in mind, too, that since the Neolithic, there have been population turnovers among perfectly modern humans in Europe.
Anne G

Maju said...

I understand that they use "Cro-Magnoids" meaning European UP Homo sapiens in general. This terminology very common in European prehistoric literature in fact, not to mention popular usage. Only when books or papers get deep into anthropometry they use Cro-Magnoid in a restrictive sense. This is also because it was believed till recently that Cro-Magnon man was Aurignacian (only a few years ago was clearly determined he was of Gravettian culture and age).

And yes, I agree with Jim's comment on age estimates not being a reasonable evidence of anything, much less an argument to reject factual aDNA data (that is consistent with other aDNA data, as well as with other TRMCA estimates that Ponto seems to prefer to ignore altogether).

On the main matter I find interesting that the only admissible ammount of Neanderthal admixture could be of (at most) 1 among 100,000 each generation. Considering that the population of Europe at that time was surely of some thousands, it would mean less than 1 succesful hybridation event each 10 or 20 generations within all living Europeans of the time. As the period of co-existence was of some 10 to 15,000 years probably, meaning 400-600 generations (at 25 years per generation). We could have a max. of 40-60 possible individual hybridation cases that might have left any legacy. And this is only the maximum possible figure.

As we have apparently found zero autosomal introgression as well, I suspect that the real ammount was much smaller. It is even possible that the two Homo species could not produce fertile hybrid offspring at all.

Ponto said...

So Jim and Maju, you would rather believe the age estimate of some bone fragments from a cave in Italy than the age estimate of various mtDNA or Y chromosome haplogroups. When R1b was discovered to be common as muck in South Western and Western Littoral Europe, everyone rabidly raved on about the connection of some modern Europeans, Spanish, Frence, British, Irish to Paleolithic Europeans. Now R1b is found to be younger than the LGM in Europe and originated in Southwest/Central Asia, everyone shuts up.

The fact is you want to believe your paradigms, the Paleolithic/Neolithic, European refuges etc because it complies with your belief system and reject all proof. The CRS mtDNA H is Near Eastern in origin, more common in Eastern Europe than Western Europe and is only about 12 kya. The CRS finding in those bone fragments most logically belongs to some museum worker who handled that bone fragment.

It is good that truth always comes out eventually, and wishful thinking of Cro Magnoid or Paleolithic origin is found to be utter invention. The Basques only date to 4000 years ago, and represent nothing but inbreeding and endogamy in some part of France and Spain.

Maju said...

So Jim and Maju, you would rather believe the age estimate of some bone fragments from a cave in Italy than the age estimate of various mtDNA or Y chromosome haplogroups.

Absolutely. C14 and other archaeological methods of datation have many many decades and have been refined and contrasted with alternative datation methods a lot. They do not arise any real controversy except among creationists.

TRMCA age estimates have at most 10 years of age and are based on several assumtions that are under question. It is not even clear if there is any "molecular clock" and, in any case, its precision cannot be compared at all with that of the half-life of the carbone isotope. There is nothing like any DNA isotope that has a well studied half-life of clear relevance, it's just an elaborate estatistical guess on the very approximate rythm of mutations in the human DNA, which can be even somewhat correct (in some lucky cases) but is not clearly validated at all. It's not even as sufficiently well based as the mere stratigraphic datation methods that were the only available for archaeologists in the remote past (i.e. some 50 years ago).

While the efficiency of C14 has been demonstrated once and again and is not the only datation method used, the reality of the molecular clock and/or our guesses that serve as base for further guesses on all that has not yet been demonstrated. And anyhow each author uses his/her own personal variation for these age estimates.

For example, in Karafet 2008, a table was published with her estimates for the major Eurasian Y-DNA lineages an became somewhat popular as it's somethin "backed" by some authority and has a recent date of publication. Whatever the exact method she used (unpublished AFAIK), she clearly estates that there is one underlying assumption: CT (now confusingly called CF by YSOGG) has exactly the convenient but arbitrary age of 70,000 BP.

Is that a well known fact? Not at all. It's a mere conservative assumption based on the coastal migration model (via South Arabia leaving no trace) and some ideas related to the Toba catastrophe epysode.

So what happens whe you get the Karafet estimates and give CT 105,000 years of age (a reasonable assumption considering the human fossil record in Asia). That all ages are stretched by +50%.

But this is still an assumption and there are other underlying assumptions in every single step of the genetic age estimates. It's less precise and theoretically much more feeble than the classical archaeological method of stratigraphy, that so much confusion has caused in many cases.

When R1b was discovered to be common as muck in South Western and Western Littoral Europe, everyone rabidly raved on about the connection of some modern Europeans, Spanish, Frence, British, Irish to Paleolithic Europeans. Now R1b is found to be younger than the LGM in Europe and originated in Southwest/Central Asia, everyone shuts up.

Well, I do not shut up because the evidence in favor of the Paleolithic continuity model is much more stronger than some highly especulative "age estimates". When authors first thought that R1b was Paleolithic they had a lot of other evidence behind to back up that claim: archaeology, other genetics (like the famous Rh-, but also mtDNA and some autosomal data). It was only logical to make such claim.

The ironical thing is that it is still very logical to make such claim but that some noisy people armed with obscure elaborate equations based on nothing but thin speculative air are raising some questions.

I welcome those questions, they are even interesting: part of the dialectic that keeps human science in continuous progress, but I cannot accept that mere speculative hunches can challenge anything, much less everything else on their own right. Science is not based on empty speculations but on hard demonstrable data.

So I challenge you to demonstrate those estimates you are so fond of with strong clear evidence and a solid theoretical and practical model. I know I will be waiting without an answer forever because there is much dogmatism and little solid science in such speculative estimates. It is not pseudoscience if taken with due precautions but it becomes pseudoscience when wielded as supposd "proof" of anything, much less against hardly questionable strong facts like the finding of aDNA H2 in Gravettian Italy.

It is obviously a lot cheaper to write down some equations and throw a hunch than to go and actually analyze the aDNA of the extant human fossils. But it cannot ever be more solid.

When C14 dating was developed (and took some 11 years to get it finished for standard use), it was solidly proven by detecting the actual age of ancient artifacts. Instead the TRMCA "dating" methods, at least the usual ones, is being disproven precisley by the same kind of hard data: it doesn't seem to be able to effectively predict the age of anything with any respectable accuracy.

It is possible, I admit, that some TRMCA models are more accurate than others or even that they can provide better predictions for certain strains than for others. But overall the whole system seems heavily limited and even probably biased by the many assumptions (subjective judgements of very uncertain variables) that people has to do in the process.

The fact is you want to believe your paradigms, the Paleolithic/Neolithic, European refuges etc because it complies with your belief system and reject all proof.

There is a lot of evidence. And furthermore the "recentist" alternative proposals fail to provide any workable model at all.

The CRS mtDNA H is Near Eastern in origin, more common in Eastern Europe than Western Europe and is only about 12 kya.

That is only an estimate. Obviously there is a contradiction between the the theory (your claim of "only 12kya") and the facts: the presence of mtDNA H2 in ancient Gravettians and the strongly dominant presence of this haplogroup in Epipaleolithic Portuguese, some 10,000 years ago, as well as in their Neolithic descendants.

If you claim that plants must be black because you have developed a theoretical model that shows that black absorbs more solar radiation than mere green or orange shades, I would obviously ask you to take a walk by your nearest park, right? Same with your feeble age estimates.

The CRS finding in those bone fragments most logically belongs to some museum worker who handled that bone fragment.

The very title of the relevant paper says: "A 28,000 Years Old Cro-Magnon mtDNA Sequence Differs from All Potentially Contaminating Modern Sequences". Please!

It is good that truth always comes out eventually, and wishful thinking of Cro Magnoid or Paleolithic origin is found to be utter invention. The Basques only date to 4000 years ago, and represent nothing but inbreeding and endogamy in some part of France and Spain.

Because you say so, amen!

eurologist said...

Well said.

I would even go a step further and suggest that any mutation rates that are not calibrated against known and well-dated events (e.g., migrations) are useless, at this point.

Of course, to some that sounds like putting the cart in front of the horse - but it really isn't, since no mathematical models can ever be trusted just by themselves, without firm grounding in established science.