After the evolutionary revolution

By Razib Khan | January 27, 2011 10:27 am


Image credit:
Luna04

My post The paradigm is dead, long live the paradigm! expressed to some extent my befuddlement at the current state of human evolutionary genetics and paleoanthropology. After the review of the paper of possible elevated admixture with Neandertals on the dystrophin locus a friend emailed, “Remember when we thought everything would be so simple once we could finally see this stuff?” Indeed I do remember. The fact that things aren’t simple is very exhilarating, but it is also a major quash on theoretical clarity. Science is after all not a collection of facts, but it is in part facts which one can sieve through a analytic framework.

In hindsight with the relative robustness of ancient DNA results we can make some assessments about the role of human bias within particular heuristic frameworks over the past generation. From the mid-1980s up until 2000 it was victory after victory for the Out-of-Africa with total replacement model. The rise of mtDNA and Y chromosomal lineage studies seemed to buttress the idea of common descent from neo-Africans within the last 100-200,000 years for all human populations. There wasn’t much of a perturbation from this march toward paradigm ascendancy in the aughts, except that there were now also now a trickle of papers which claimed to phylogenetic “long branches” in the human genome. The 2006 Evans et al. paper, Evidence that the adaptive allele of the brain size gene microcephalin introgressed into Homo sapiens from an archaic Homo lineage, was probably the one that made the biggest media splash. But these were inferences. Subsequent analysis of the draft Neandertal genome seems to suggest that in fact the microcephalin allele in question did not introgress.

Case closed? Obviously not. Now we’re in a different era. The Evans et al. paper may have wrong in the specifics, but its general framework seems to likely have been validated: there are genetic lineages in the modern human genome which are not derived from the neo-Africans. But, let us remember that the overwhelming majority of the human genome is neo-African. A reasonable interval for non-Africans is 90-99% neo-African. But, a non-trivial minority has introgressed or admixed from other lineages. Out-of-Africa is mostly correct, but in some ways so is Multiregionalism. But how do we describe this? “Weighted multiregionalism”? “Mostly Out-of-Africa?” The old terms were nice because they were punchy and precise. If you look at Multiregionalism or Out-of-Africa in Wikipedia the newest results are noted, but it doesn’t seem that they’ve been integrated into the analytic narrative. Yet.

  • SteveF

    There’s a piece by Ann Gibbons in Science today that provides an overview of all this, framed around Chris Stringer and Milford Wolpoff. Ends with quite a nice quote from Chris:

    As for Stringer and Wolpoff, both now in their 60s, their battle has mellowed. Their views, while still distinct, have converged somewhat, and they shared a beer at a Neandertal meeting last year. “The reason we get on well now,” says Stringer, “is we both think we’ve been proved right.”

  • http://washparkprophet.blogspot.com ohwilleke

    Perhaps it ought to be called “Syncretic Evolution” since it involves the blending of previously well defined separate lineages.

  • http://sciencepolice2010.com John Jackson

    I’m using ‘out of Africa plus mixing’.

    I’ve never understood why anyone ever thought that mtDNA and the Y chromosome could veto very small archaic contributions, particularly advantageous ones (e.g. cold-adaptations). That possibility should have remained open in everyone’s mind, shouldn’t it? I’m not driven to become an expert in e.g. mtDNA analysis, but I’d love to know if it was me or so many others who were wrong over this.

  • http://www.kinshipstudies.org German Dziebel

    “The reason we get on well now,” says Stringer, “is we both think we’ve been proved right.”

    “Perhaps it ought to be called “Syncretic Evolution” since it involves the blending of previously well defined separate lineages.”

    As Alan Templeton pointed out, the Multiregional and out of Africa intellectual lineages were never properly separated analytically from each other in the first place. One theory was just a special case of the other. No wonder both of them are “right.” It’s a good example of how tautological, scholastic and medieval the science of human origins is. A couple of high priests getting together to congratulate each other on being right.

  • onur

    I’m using ‘out of Africa plus mixing’.

    That is a more appropriate term than “Syncretic Evolution” in light of the available evidence IMO.

  • http://washparkprophet.blogspot.com ohwilleke

    ” I’m not driven to become an expert in e.g. mtDNA analysis, but I’d love to know if it was me or so many others who were wrong over this.”

    The key point is that in order to completely wipe out any Y-DNA or mtDNA traces of a population that leaves a small but non-negligable autosomal DNA trace, you need both the Y-DNA and mtDNA to either be eliminated in the first few generations, or you need both the Non-recombining Y-DNA and the mtDNA to be subject to strong selective pressures that aren’t observed otherwise. If you had Neanderthal Y-DNA or Neanderthal mtDNA still in the gene pool five or ten generations out, it would be almost invisibly diluted and mixed in with the rest of the gene pool and dumping that from the gene pool at that point would be like trying to unmix a beach made of two different colored sands – almost impossible without some instrument so blunt that it simply wipes out the entire admixed population taking the autosomal DNA along with the Y-DNA and mtDNA with it.

    To wipe out an entire Y-DNA or mtDNA lineage from the gene pool entirely after that point while retaining the autosomal DNA takes a really, really small population bottleneck, or some astoundingly improbable random genetic drift, or a really complicated multi-step process — for example, first with a large share of a conquered population’s Y-DNA being replaced by a conquering population, and then with men from that population taking mates from another population while the source population is wiped out.

    There are really only two scenarios that make sense.

    One is that most of the Neanderthal genome was really maladaptive, causing Y-DNA and mtDNA and Neanderthal autosomal DNA to all become very scarce in the gene pool until the selectively advantageous genes are separate from the disadvantageous genese in someone who happens to have neither Neanderthal Y-DNA or Neanderthal mtDNA, perhaps a bottle neck of a few dozen people, at which point, the selectively advantageous genes freed of their maladaptive counterparts, balloon in frequency. Given how much longer Neanderthals had to adapt to their local environment than AMHs this seems implausible, but perhaps it could make sense if the maladapative trait were cognition related and the adaptive traits were not.

    The other scenario, which I think is more plausible, is that there are some powerful gender biased dynamics going on in the first few generations. In my view, the reason we lack Neanderthal mtDNA is probably because the Neanderthal hybrids who were born into and accepted in human tribes probably had modern human mothers and Neanderthal fathers, while the hybrids with Neanderthal mothers probably ended up in Neanderthal tribes and their descendants probably died out. The reasons we lack Neanderthal Y-DNA, in my view, is that Haldane’s rule provides that hybrid children should generally be female rather than male as is the case in other hybrids (actually XX rather than XY — in some species the equivalent of a human XX is male and the equivalent of a human XY is female). Thus, in one generation, both the mtDNA and Y-DNA are absent in children who have 50% Neanderthal autosomal DNA when can then be diluted to whatever frequency made sense.

    This scenario also explains why X linked Neanderthal genes ought to be more common in modern non-Africans than other genes.

  • gcochran

    You’re wrong. Assume neutrality. If the effective human population was small, and the original admixture of archaic mtDNA or Y-chromosome was 2.5%, the level can fluctuate to zero (and then stay there, of course), either early or late. Not recently, since by the Holocene population sizes were much bigger. If neutral, the average percentage of Neanderthal alleles does not change over time for the genome as a whole, but any one locus fluctuates – could easily be zero or some number considerably higher than 2.5% today.

    Y-chromosomes and mtdna are not linked to other genes, so any selective advantage or disadvantage does not really have anything to do with the overall goodness or badness of the Neanderthal genome. Of course autosomal genes _are_ linked to their neighbors, and it is possible that some Neanderthal alleles with selective advantages did not get to show their stuff until recombination had freed them from a disadvantageous neighbor, which could have taken a long time.

    That said, it is also perfectly possible that Neanderthal Y-chromosomes or mtDNA had selective disadvantage on an AMH background. Maybe especially likely because the
    Neanderthals seem to have had a small effective population size, which makes selection less efficient.

    As for the likelihood of more Neanderthal alleles on the X chromosome, you might want to look at the online supplement to Green’s Neanderthal paper, page 135, where there is a chromosome-by-chromosome admixture breakdown. Not so in Europeans, mebbe so in east Asians.

  • http://washparkprophet.blogspot.com ohwilleke

    ” If the effective human population was small”

    The key point is how small the effective population has to be to disappear by random genetic drift.

    If the effective population is 80, so that 2.5% is 2, random genetic drift will very easily remove it from the gene pool by random chance. One gets killed by a sabre tooth and the other has only girls and Neanderthal Y-DNA is gone.

    If the effective population is 1000, so that 2.5% is 25, the amount of random genetic drift necessary to get to zero is pretty great and it is quite unlikely, but not impossible. A lot of independent event that differ from the average have to take place to eliminate it from the gene pool.

    If the effective population is 10,000 so that 2.5% is 250, the amount of random genetic drift necessary to get to zero is immense.

    If the effective population is 100,000 so that 2.5% is 2,500, it is almost impossible to remove it from the population with random genetic drift.

    The absolute size of the subpopulation matters more than the size of the percentage when you are looking at the likelihood that it can be removed from the population after reaching fixation by random genetic drift. (The math of this is a lot like the math of statistical sample size where margin of error declines dramatically with addition individuals up to a few hundred and then improves much much more slowly as more individuals are added to the sample.)

    And, it takes surprising few generations for a trait to reach fixation. For example, the European admixture percentage in the African-American population of the United States probably reached close to fixation from a very uneven distribution in the overall African-American population in something on the order of a century.

    For NYR or mtDNA, the main way that they drop out of the gene pool is from overall random mortality, or from random variation in male/female ratios at birth. This happens when those events happen at non-average frequencies in the affected subpopulation. The larger the population, and the more generation it is in the population, the more intensely the law of averages (which says that as the number of events get larger, the percentage deviation from the average number of events will fall) catches up with you.

    The effective population size of Neanderthals in Europe has been estimated based on ancient DNA at 1,500 to 4,500 give or take. The low end of the Eurasian founder population estimate is similar, although others in the literature put that number ten times as high. This implies something on the order of 40-500 individuals with Neanderthal Y-DNA or mtDNA that have to have their lineages disappear from the gene pool through random genetic drift if they’ve managed to stay in the gene pool long enough for Neanderthal DNA to reach fixation which is roughly ten generations, but comes reasonably close in about five generations.

    At the low end of that range, random genetic drift eliminating the line is unlikely but not impossible. At the high end of that range it is extremely unlikely.

    A scenario with 0.5% Neanderthal mtDNA and 0.5% Neanderthal Y-DNA and 2.5% autosomal DNA isn’t hard with random genetic drift, but to get to zero which is what the data seem to show so far, is much harder unless it happens in a small number of generations when the absolute number of individuals with those genes is well under a hundred.

    On the other hand, even slight selective impacts, with a low base percentage can cause a trait to disappear from the gene pool even in a much larger gene pool in which a gene has reached fixation, and this too can happen much faster than a typical person’s intuition would think that it would.

    Also, all of this assumes a static effective population size. It is much harder for a gene in an expanding population to drop out of the gene pool. Thus, a genetic drift theory is much less plausible in a modern human expansion into virgin territory or rapid replacement of Neanderthal population scenarion than it is in a scenario with populations co-existing in stable equilibirum for long periods of time.

  • gcochran

    You are mistaken.

  • http://sciencepolice2010.com John Jackson

    Ah yes – traces of esoteric mtDNA/Y-DNA in the _modern_ population. I can see how lack of such traces might be given a meaning; I was thinking about the significance of such single-parent traces in DNA from just a few fossils, specifically their lack of significance. (BTW, thanks ohwilleke & gcochran for airing the concepts – and to onur for your approval :-) .)

    You’ve shown that many things could affect the survival of esoteric mtDNA/Y-DNA; the first scenario I imagined though, was of a repeated or steady stream from Africa, against which Neanderthal etc DNA had to swim just to stay level. At a first guess, I’d say neutral mtDNA/Y-DNA had no chance in that scenario, whereas advantageous nuclear DNA would probably survive… and indeed even a little neutral recombining DNA would have some chance.

    I think though that it’s the main message of the original posting that is the biggest story: surprise at how poor the”experts’ ” predictions were. If Stringer and Wolpoff has been just any blokes in the pub, we could say they’d both been right enough. But as they’d been reasonably paid and trusted professional scientists for decades, we were entitled to expect them to to include and mention that scenario which we currently see as being the most likely, instead of prematurely obliterating it from the picture. They were both wrong; their level of skill in basic scientific thought was and is still dire, and shows that philosophy of science is a valuable subject in itself. We need more philosophers more than we need more fossils. Stringer is still holding us back by saying: “…molecular estimates of the divergence time between humans and chimpanzees presently converge on approximately 5 – 7 Ma…” (Phil. Trans. R. Soc. B (2010) 365, 3265–3266). I’m not sure what the _average_ split time given by recent reports is, but know 4mys hasn’t been rejected by the studies, even though it clearly has by Stringer and so many others. Premature rejection of possible solutions is not only a – perhaps _the_ – major source of scientific error and muddle, but is the trademark of far too many famous figures in palaeontology.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »