by Greg Mayer
Alert reader daveau has drawn my attention to a manuscript (abstract only; BBC story here) posted on Science’s website by Laurent Abi-Rached and others on the genetic evidence for interbreeding between anatomically modern Homo sapiens and earlier Eurasian Homo (Neanderthal man, and a related group, the Denisovans.) Jerry and I have addressed this issue earlier here at WEIT here and here.
The story so far is that study of ancient DNA has shown that Neanderthals contributed a few percent to the nuclear genome of modern Eurasian populations (but not African), and that a previously unknown population of archaic humans, the Denisovans from Siberia, has contributed a slightly higher percentage of the genome of Melanesians. There are two new developments, both discussed by Ann Gibbons in a news article (abstract only) in Science, in which she reports on a conference hosted by Russian researchers held at the fossil sites (including Denisova) in southern Siberia. First, Australian Aborigines, like Melanesians, derive about 5% of their genome from the Denisovans. Second, Abi-Rached and colleagues have looked at three loci involved in the immune response (HLA loci), and found a much higher proportion of Neanderthal/Denisovan contribution at these loci. Over half the alleles had an archaic origin, and they reached frequencies of over 50% in some modern populations.
Abi-Rached et al. attribute this to interbreeding between anatomically modern humans coming from Africa, and the resident archaic populations (Neanderthals/Denisovans) of Eurasia. Why would the immune loci show much stronger influence than the genome as a whole? Immune loci are highly functional, and subject to strong selection. Diversity of alleles is selectively advantageous at immune loci, so that heterozygotes resulting from interbreeding would be favored, and the particular alleles of archaic humans, who had been in the Eurasian environment for a long time, might also be favored.
John Hawks of the University of Wisconsin, Madison, finds this a plausible interpretation, and favors it himself. But he does raise some cautions. Because these loci are subject to strong selection, some of the alleles may have been maintained in human populations for a long time, without any need for interbreeding to introduce them into anatomically modern humans. And one of the alleles that Abi-Rached et al. think shows archaic influence, HLA-B*73, wasn’t actually found in any of the archaic individuals studied. (This is not as bad as it sounds– HLA-B*73 is closely associated with linked alleles that are in the Denisovan.) He has some other cautions, and his full account is well worth reading.
So where does this leave us? Mostly wanting more data, especially from more archaic individuals, but the conclusion Jerry and I (and John Hawks) came to still stands, and is in fact now slightly firmer: Neanderthals, and Denisovans, are us.
(As an aside, it is unclear to me if the posting by Abi-Rached will eventually appear as a paper in Science or not. It is not one of the posted-just-before-publishing posts, since it has not been formatted for publication, and at nine pages it’s longer than what Science usually publishes. There are also 45 pages of supplemental text, with 26 extra figures– if they think they have a monograph, they should publish a monograph! As John Hawks dryly put it, “bibliographic information not yet available“. Science has had some bad experience lately with posting things they intended to publish, but then things didn’t work out as well as they hoped. I’m not suggesting there’s anything wrong here, just that this odd state of not knowing if or when a manuscript will be published is not something to be encouraged.)