Neanderthals

[Note: you posted the same link twice.]

quote:

---

Abstract from second link:
Human mtDNA shows striking regional variation, traditionally attributed to genetic drift. However, it is not easy to account for the fact that only two mtDNA lineages (M and N) left Africa to colonize Eurasia and that lineages A, C, D, and G show a 5-fold enrichment from central Asia to Siberia. As an alternative to drift, natural selection might have enriched for certain mtDNA lineages as people migrated north into colder climates.

---

These results do not support a multiregional hypothesis. The distribution of mtDNA haplotypes still strongly supports a migration out of Africa as the source of non-African populations. What this paper proposes is that as the migration expanded into colder regions, some variants were at a selective advantage and so became more common. This is completely consistent with an Out of Africa model -- OoA doesn't say anything about whether or not selection occurred during the migration.

quote:

---

Both of these lines of evidence and the anthropological evidence are almost completely aligned with the out-of-africa hypothosis. I'm not in the field but I think that the multiregional hypothosis is falling out of favour with the additional evidence of recent years.

---

There's a strong consensus now that Out of Africa is at least largely correct, although there's still debate about whether there was any genetic input from non-African archaic lineages or not, and lots of debate about the details of the migration (when, where, how many times).

quote:

---

Quote from abstract:

​

"Africans have had a huge genetic impact on humanity," he says. "But my analysis really isn't compatible with complete replacement."

---

There were several significant problems with Templeton's Nature paper, the one he's referring to here. There's good evidence that one of the genes he looked at (MC1R) is under strong purifying selection in Africa and either weak or diversifying selection outside Africa (MC1R controls skin pigmentation), so dating it as if it were a neutral locus isn't right. He redated another locus, increasing its age by an order of magnitude over the age in the original study, but offers no justification for the change. For a more fundamental problem, seeOut of Africa with regional interbreeding? Modern human origins.
Satta Y, Takahata N.
Bioessays. 2002 Oct;24(10):871-5.Which is largely a response to Templeton's paper.

quote:

---

Can't post more now...must...watch...stupid...baseball game.

---

Stupid game? Sounds like you were watching a Red Sox game.

quote:

---

mtDNA can be spread by either migration or gene exchange, it is inconclusive as a means by itself to determine modern origins.

---

Yes, but the paper you quoted was about mtDNA haplotypes that arose outside Africa and then increased in some places outside Africa due to selection. You were citing it as if it were evidence that selection had caused the spread of African mtDNA to replace non-African, which it isn't. You also claimed that selection was inconsistent with an Out of Africa model, which isn't true.

quote:

---

A more serious problem (and one that Svante Paabo and Alan Cooper recently shot themselves in the foot with) is what constitutes an "acceptable" neandertal or cro magnon sequence for example? Paabo made the argument with neandertals but more recently with Cro magnon that if a sequence is retrieved from either and looks modern, there is no way to distinguish it from a modern contaminant. Thus, only divergent sequences are accepted as bona fide neandertal or cro magnon. What the hell kind of science is that then? I know researchers who have pulled out very modern looking sequences from neandertal fossils. The one published Cro magnon sequence is identical (for the portion sequenced) to the first full mtDNA genome published by Anderson and colleagues i.e. plain old normal human mtDNA. These sequences are called into question or outright rejected as valid. So if you can only a priori accept that divergent neandertal or cro magnon (no H. erectus sequence claims have yet been made) sequences are valid the you will a priori determine that they were either a separate species or radically divergent from modern humans...that is just not good science. This was pointed out to Paabo and Cooper by the Italian group that did the Cro magnon sequencing.

---

I don't see how that follows. If the experiment is done well, the test is one-sided but valid. That is, you can get two answers, "Different from modern humans" and "Can't tell". The Neandertal specimans clearly gave results that were different from modern humans, and reproducibly so. And now that multiple individuals have been tested, there are multiple results, all different from modern mtDNA and clustered together. What alternative hypothesis explains these results?

quote:

---

It has been traditionally viewed that any mtDNA did not have any selective advantage, it was only a passenger carried along by autosomal genetics.

---

What do you mean by "traditionally"? Many analyses have considered the possibility of selection in recent mtDNA and Y evolution. Look at the big Takahata paper on testing multiregionality, or Kaessmann et al's 1999 paper on Xq13.3, or Stumpf and Goldstein's review article on the Y. It has been widely recognized that selection is a possibility that needs to be taken into account when drawing conclusions from any locus (Takahata even modeled it), and especially for mtDNA and the Y, since they're nonrecombining and single-copy.As for where I got the idea that you thought OoA and selection were incompatible, it was from this:

quote:

---

The Eve hypothesis postulates that mtDNA was simply carried along with a migrating population. While the MRH states that genes were exchanged and natural selection facilitated the spread of favorable traits found in mtDNA.

---

(I assume you meant OoA rather than Eve, since a mitochondrial Eve is a feature of any theory of human origins.) "Simply carried along" (especially when contrasted with natural selection) sounds like you're making an absence of selection a postulate of the theory.

My comments were based on the assumption that the tests were done well. If specimens were dropped from the papers because they didn't give the expected results, then the tests were not done well.(The only person involved that I'm sure I know is Paabo himself, but I've probably met others at meetings. Since I tend not to remember anyone's name unless I'm married to her(*), I'm unlikely ever to know.(*) I'd include my kids as well, but I sometimes get the younger one mixed up with the guinea pig.)

quote:

---

One other thing is that you would expect Africa to have larger populations of hominids than any of the other regions due to its favorable climate and thus contribute more to the genetics of modern humans in that way.

---

Yes, that is the multiregional explanation, but I really don't see how it can be made to work quantitatively. Based on either the genetic diversity or on the amount of linkage disequilibrium, non-Africans look like a very small population -- their effective population size is no more than a few thousand. How could that small a population be spread across all of Eurasia and form viable subpopulations in multiple regions, with enough population density between them to support substantial gene flow? Especially since that kind of structured population preserves more diversity than a single population of the same size.Add the fact that European and Asian populations show signs of having passed through a substantial bottleneck, and the fact that their common alleles are, the great majority of the time, a subset of African alleles, and the weight of the genetic evidence is heavily on the side of African replacement. Whether it was a complete replacement or not is not clear, but it seems to be have been pretty thorough-going.

quote:

---

There's the sum of it, the more people take an honest, objective look at all of the evidence, the more Out of Africa looks like MRH. It doesn't matter how big the populations were, since they remained viable for hundreds of thousands of years in their respective regions.

---

Of course it matters how big the ancestral populations of non-Africans were: if they were really small, then they weren't the archaic non-African populations that were in place, since they had to be larger.To clarify my previous statement: at present, the evidence from human genetics, taken as a whole, indicates a complete replacement of non-African archaic populations by African ones. It is possible that some admixture did occur and the evidence has either been lost or lies in loci that haven't been examined yet, but at this point there is no such evidence.
That's genetics. What other evidence suggests I am no competent to comment on.

quote:

---

And I personally don't care whether multi-regionalism or out of Africa is correct or whether or not neandertals and H. sapiens interbred.

---

If I had to choose one as being more interesting, it would be neandertal/H. sap interbreeding, but so far nature hasn't asked my opinion. I have very little stake in the answer, as it happens, since I was a physicist while most of the debate was going on and wasn't paying any attention. Also, my institution does primarily medical genetics and has no position to defend, which is nice. (Although it does mean that some population geneticists sneer at us when we do population genetics. For some reason population geneticists spend a lot of time sneering at each other.)

quote:

---

Why would they have to be larger? Larger than the Africans or larger than the regional populations? How can it be if they were really small, that they were then the archaic non-Africans?

​

What I'm saying is that the populations may have been small, but that they were able to migrate, exchange genetic information and continue to exist for a long time well before the African genetic migration.

---

The reason there's a problem is that the effective population size is the size of the entire population, not the size of regional subpopulations. The entire human population has an effective size of about 10-20 thousand; the entire non-African population has an effective population size of well below 10,000. The small size is reasonable for a group or collection of groups moving out of Africa and then expanding, but it is not reasonable for a population spread across many thousands of miles and split into many smaller subpopulations. I could understand an archaic population in China of a few thousand, but I can't understand a population of a few hundred in China, a few hundred in southeast Asia, a few hundred in Iberia, and so on. MRH requires that every one of the populations was very small, and at the same time that there was a high enough density across the entire range to maintain substantial gene flow. That just doesn't seem possible.This is not based on mtDNA, by the way. Y, X and autosomal diversities are all consistently low, and autosomal linkage disequilibrium is very high outside Africa. Similarly, X and autosomal loci, as well as the Y and mtDNA, tend to have their deepest branches in Africa, not elsewhere.

quote:

---

No lie! And then they all get together over beers and sneer at everybody who ISN'T a population geneticist.

---

My theory about this phenomenon is based on my background in physics. Like population geneticists, theoretical physicists work in a solitary, abstruse field with lots of nifty equations. Unlike pop gens, however, theoretical physicists get a lot of respect from their experimentalist colleagues. Population geneticists don't get a lot of respect, or even awareness of their existence, from other geneticists, and that makes them cranky.(I didn't say it was a good theory.)

quote:

---

It would mean having an effective breeding population of around 10,000 spread over two continents. This would not be that big of a stretch at all. [...]
So we would see populations from demes with effective breeding populations of around 500 and then periodic mating between demes from adventurous youths traveling fairly great distances to find a suitable mate.

---

That model fails for two reasons, I think. First, a population of 10,000 with such a high degree of isolation between subpopulations would have an effective population size much higher than 10,000. Once again, the diversity would be high than we see. Second, the genetic distances between geographically distant subpopulations would be large. The largest genetic distances we see are pretty modest, with Fst in the range of 0.1 to 0.15.

quote:

---

Hmmm...still stunned from the game tonight. Feeling funkier than usual.

---

I lost interest in the season at the same time Pedro Martinez lost his control in the seventh game of the ALCS.

quote:

---

I'm saying that the isolation was not that large. I'm also saying that the inbreeding would lead to a significant loss of genetic diversity.

---

Small populations lead to loss of a lot of genetic diversity, but you lose much less diversity if your population is subdivided, since different subpopulations retain different subsets of diversity. If I ever get time I'll try simulating your toy model, to see what kind of migration rate would be needed to maintain small genetic distances across the entire range. (Note, by the way, that the distance between demes in your model is something like 500 miles, and that's if they're strung out in a line from Iberia to Java. Scattered about, they'd be even farther apart. That's a long way to walk for a wife.)[Note added later: "toy model" is not disparaging -- I look at toy models all the time.][This message has been edited by sfs, 10-21-2003]