Last week a paper was published on the DNA sequencing of the complete genome of an approximately 40,000 year old finger bone from Denisova in Southern Siberia (Reich et al. 2010). There are a number of very significant findings in the paper.
1) The results showed that the Denisova individual was more closely related to Neanderthals than to modern humans. However, the Denisovan does not fall within the range of Neanderthal variation. There are a few lines of evidence suggesting that the Denisovan and Neanderthal lineages had separate histories once they diverged and did not form a single population. The Neanderthal genomes sequenced so far have low genetic diversity, indicating that Neanderthals passed through a genetic bottleneck after splitting from the Denisovans. With only one Denisovan genome available as yet, we don’t know how diverse they were.
2) Even more spectacular was the finding that the Denisovan genome appears to have made a genetic contribution of about 4.8% (+/- 0.5%) to the genomes of living Melanesians. Interestingly, they did not contribute to the genomes of modern populations such as Han Chinese and Mongolians which live near Denisova now. The Denisovans obviously interbred with the ancestors of modern Melanesians at some point, but it seems unlikely to have happened at Denisova, which suggests that the Denisovans lived over a considerable area of eastern Asia.
This finding recalls another major discovery made earlier this year, which was that Neanderthals appear to have contributed from 1% to 4% of the genome of all living non-Africans (Green et al. 2010; Panda’s Thumb blogged about it in June). This was not expected, given that earlier mitochondrial DNA (mtDNA) results showed no evidence of any genetic mixing between Neanderthals and modern humans (but did not exclude the possibility either). The new Reich et al paper has improved the precision of this estimate; they calculate that Neanderthals contributed 2.5% (+/- 0.6%) to the genome of modern non-Africans. That means that the Neanderthals and Denisovans together account for an impressive 7.5% of the ancestry of modern Melanesians.
The finger bone, by the way, is the same one that was in the news in March 2010, when another paper (Krause et al. 2010) was published on the bone’s mtDNA. The results of that paper showed that the Denisova mtDNA was more different to both humans and Neanderthals than they were to each other. The Denisova mtDNA shared a common ancestor with both humans and Neanderthals mtDNA about a million years ago, compared to about 500,000 years for the human and Neanderthal mtDNA common ancestor. This is obviously in disagreement with the new findings from the full genome, which are almost certainly correct because they are based on far more complete data. There are a couple of possible explanations for this. The mtDNA lineage may have been introduced into the Denisovan population by interbreeding with another unknown hominid lineage, or randomness due to genetic drift may have allowed a divergent mtDNA sequence to survive in the Denisovan population while being lost in the human and Neanderthal populations.
The finger bone was nicknamed the X-Woman, X for ‘unknown’ and ‘woman’ because mtDNA is maternally inherited, but in fact it didn’t necessarily belong to a female. Males have mtDNA too; they just don’t pass it on to their children. The new analysis, however, shows that the ‘X-Woman’ bone really was female - there are not nearly enough Y-chromosome genetic sequences in it for it to have belonged to a male (the handful of Y-chromosome sequences found would be from contamination).
3) A second bone found at Denisova, a second or third upper molar tooth, has also had its mtDNA sequenced. This mtDNA sequence was very similar to that of the finger bone, indicating that both individuals probably belonged to the same population.
4) The molar tooth from Denisova also gives us some intriguing hints about the anatomy of the Denisovans. It is larger than comparable teeth in both early modern humans and Neanderthals (and also Homo erectus, if it is a third molar). It also has a number of anatomical differences from Neanderthal teeth to which it would be most closely related. Obviously it would be desirable to have more Denisovan fossils and knowledge of more of their anatomy, but the tooth suggests that they are distinctive anatomically as well as genetically from both humans and Neanderthals. The 2nd and 3rd molar teeth of other hominids alive at the time such as Homo heidelbergensis and Homo erectus are not well represented in the fossil record, but the little we have also does not seem to closely resemble the Denisovan tooth.
For now, the population to which these two fossils belonged is being called ‘Denisovans’. A decision on how to classify them will probably be deferred until we know more of their anatomy. The fact that Denisovans and Neanderthals both interbred with humans does not necessarily mean they are all the one species - there are many cases of modern species able to interbreed. Animal species are more usually defined by whether they ordinarily form a single interbreeding population, rather than by whether they are merely capable of interbreeding. This is obviously a fuzzy criterion, but the fact that the Neanderthal and Denisovan contributions to the human genome appear to have been limited events means these hominids could still end up being classified as three species. (Or two. Or one.)
The evidence that the Denisovans contributed genes to modern Melanesians did not come completely out of left field. In April, a group of researchers reported evidence of humans interbreeding with other hominids in the Middle East about 60,000 years ago, and in eastern Asia about 45,000 years ago. That matched up with the discovery of Neanderthal genes in modern non-Africans, and there was speculation then that the Denisovan fossil, whose mtDNA was published around the same time, was in the right time and place to account for the 2nd interbreeding event. And so it came to pass.
Incidentally, these discoveries would seem to be the final nail in the coffin for the multiregional model. There are two competing schools of thought about the origin of modern humans. The ‘Out of Africa’ model claims that the ancestors of modern humans left Africa in the last 100,000 years, totally replacing all other hominids around the world. The multiregional model holds that humans evolved towards modern humans in synchrony throughout Africa, Europe and Asia, with genetic mixing between neighbouring populations ensuring that they did not diverge into multiple species. The Out of Africa model isn’t completely correct, but could be modified to include two episodes of genetic mixing with other hominids. The multiregional model, on the other hand, looks to be completely dead in the water.
Implications for Creationism
Young-earth creationists such as Answers in Genesis will doubtless claim that this research supports their claims that humans, Neanderthals, and other archaic hominids all form one species. However, it’s a lot harder to see how all the necessary population events can be squeezed into 10,000 years. Starting from Adam and Eve, humans apparently populated Africa, Asia and Europe, then some of them left Africa, picked up some Neanderthal genes from the Middle East, then populated the world again, with some of them picking up more genes from the Denisovans and going on to populate Melanesia. Somehow, this emigrating group was also able to cause all the other humans to become extinct. At some point a flood occurred, killing all but 8 humans and removing most of the genetic variability. It would be tempting to assume the flood took care of removing the Neanderthals and Denisovans, but that would leave the problem of explaining how their genetic contributions made it into the modern world. Supposing one person on the ark had Neanderthal genes, and another Denisovan genes. The Arkers then had children who would have married each other. How could it happen that Africa ends up with the greatest genetic diversity, and yet none of the Neanderthal and Denisovan genes are found there? Meanwhile, the Neanderthal genes managed to find their way into all non-Africans, while the Denisovan genes found their way into the Melanesian population, but nowhere else (if 8 people populate the world, how can one of those people account for 5% of the genome of 0.15% of the world’s population?). This scenario seems, to put it mildly, hopelessly improbable if not completely impossible. I await with interest a creationist population model to explain all this activity in a young-earth time frame.
See also: The Denisova Genome FAQ, by John Hawks
References
Dalton 2010: Neanderthals may have interbred with humans. Nature Online, Apr 20, 2010.
Green et al. 2010: A draft sequence of the Neanderthal genome. Science, 328:710.
Krause, Good, Viola et al. 2010: The complete mitochondrial DNA genome of an unknown hominin from southern Siberia. Nature 464:894.
Reich, Green, Kircher et al. 2010: Genetic history of an archaic hominin group from Denisova Cave in Siberia. Nature 468:1053.
Repost from the older Denisovan thread:
This is fascinating. A third subspecies of human and we just now found them. I read John Hawks take on it, linked above.
1. We don’t know what they looked like. All they have is one finger bone and a tooth.
2. There seems to have been a widespread population based on where the genes ended up, Melanesia. A prediction is that more fossils should turn up around south Siberia and south Asia.
3. If the population models are correct, Denisovan genes should also be found in New Guinea and Australian Aborigines. IIRC, the Australian natives are part of the same migration through Melanesia and New Guinea.
4. Where do the Flores Hobbits fit in? Could they be derived from an early form of Denisovans?
I did do some google anthropology. It looks like Melanesians, New Guinea natives, and Australian Aborigines are all part of the same migration. So they should all have some Denisovan sequences.
Apparently there are so few Australian native sequences that it is hard to do DNA phylogenies.
“I await with interest a creationist population model to explain all this activity in a young-earth time frame.”
And I await with interest any convincing creationist explanation for all of the evidence of an ancient earth: tree ring data; ice core data; magnetic reversals; continental drift; grand canyon; Hawaii; etc. I also await with interest any convincing creationist explanation for all of the genetic and developmental data consistent with descent with modification. But then again, I eagerly await any creationist calculation of specified complexity as well. Come to think of it, I eagerly await any convincing creationist for anything. Sure haven’t heard one yet.
This may give the wrong impression. IIRC the mtDNA calculations did not rule out low percentages of Neanderthal contribution to present-day humans, but mtDNA showed no positive evidence for it. It had an estimate of 0% but with upper confidence limits that did not contradict 4% or 2.5%.
A good point, have made a clarification.
New Guineans/Papuans are Melanesians and do have Denisovan genes (a Papuan was sequenced for the paper). Australian aborigines not necessarily; they may have migrated at a different time or by a different route. If the aborigines were part of the same migration, that’s news to me.
Yet it must frustrate them that it still won’t convince OECs and IDers. While some OECs might challenge them on it, IDers will make sure to neither agree nor disagree, and that silence will speak more volumes to me than any predictable YEC spin.
This is really cool stuff. It is an exciting time just to be a fly on the wall watching the discoveries in hominin evolution I can only imagine how cool it would be to be actively engaged in this research!
DS: I hope you’re patient because I think you’ll have a long wait.
I’m guessing that AiG will probably claim that the Denisovan finger or tooth belonged to Ham. He was a randy bugger and got around y’know. How close is Denisova to Ararat? :)
Regarding the Denisovan population being widespread, is this hypothesis stating a widespread contemporaneous population or just that they covered a lot of territory during the species’/subspecies’ existence but perhaps consisting of a small population that moved around a lot? It occurred to me that based upon the lack of Denisovan genes in other Asian populations near Denisova it may be equally parsimonious to suggest that the Denisovan’s were driven south and east across the continent by other hominin populations where eventually the Desisovan’s may have come into contact with the ancestors of Melanesians.
I did spend some time on google trying to find the genetic affinities of Australian natives. This 2007 article using some sequence data has them as part of the same migration that settled New Guinea.
That occurred to me too. I think (but am not sure) that the eastern pacific islands were only populated in the last 5,000-10,000 years, which would mean the interbreeding had to occur way before the group now known as Melanesians left the mainland.
All I know about Aborigine affinities comes from google searches. They seem to be related to Melanesians and Papuans. If so, they should also have Denisovan DNA contributions.
Yeah, IIRC the Melanesian migrations were supposed to be after about 12KYA so the mixing would’ve been a long time ago. But then Denisova is an awful long way from the southeast Asian coast with some awfully big mountains in the way!
Not a lot of work has been done on Australian Aborigine genomes. At least one group is related to Melanesians.
They are a diverse group and the question has been whether they diverged or whether more than one migration settled the continent.
Aboriginal Australians are perhaps the oldest, and most well-isolated, group of modern humans descended from the diaspora of modern humans out of Africa about 60,000 years ago.
Primarily Haplogroup C4 (Y-STR), their ancestors left DNA along the coastal route from Africa to India to Australia. http://en.wikipedia.org/wiki/Haplogroup_C_(Y-DNA)
Spencer Wells, of the Genographic Project, popularized genetic DNA in his Journey of Man (PBS/National Geographic Channel - CINE Golden Eagle award; available on-line by clicking the reference at http://en.wikipedia.org/wiki/Spencer_Wells ) and highlights the discovery of ancestral Aboriginal DNA on the Indian sub-continent.
Another issue with the way that the Denisova results are presented above is that the inferred genealogy is shown as a tree. It may or may not be, particularly if there is gene flow among populations (for example, among the Neanderthals or among the modern humans). Individual loci will have coalescent trees and they will tend to be different from locus to locus.
Making “individual-based phylogenies” is a bad idea in general. Actually, I had two parents, they each had two parents, and so on. Going back, my ancestry does not converge as a single tree, at least not in terms of the copies of genes that are ancestral to me.
Thus the Neighbor-Joining tree shown above is not necessarily a good representation of the genealogies.
I believe that this is more an issue with this blog post than with the original Reich et al. paper.
First it’s nailing down Denisovans as a unique population and now there’s ambiguous but intriguing fossil teeth found with 3-400kya tools in an Israeli cave that might shed light on Neanderthal or archaic Homo sapiens evolution and migration (though not to the extent that the popular press have inflated the story).
This is an exciting month for the study of early humans, and really helps drive home the point that several species of people were already around at the time fully modern H. sapiens arrived on the scene.
I can appreciate (if not completely understand the details of) how one might determine the relatedness of two genomes, based on certain assumptions about genetic drift and the accumulation of mutations.
But how does one decide that X.Y% of one genome was “contributed by” another genome? How would you distinguish the case of “‘A’ gave ‘B’ gene ‘W’” from the case of “‘A’ and ‘B’ shared a common ancestor that had gene ‘W’”?
Put another way, let’s say that Group ‘A’ and Group ‘B’ share a common ancestor, Group ‘W’. At some point in time, they separate and their lineages start to diverge. At a later point in time, the two groups meet again, intermingle, and produce Group ‘B-prime’, which is mostly ‘B’ with a little ‘A’ thrown in. Then Group ‘A’ dies out, leaving us with Group ‘B-prime’. Looking at Group ‘B-prime’, how can we tell the difference between a gene that Group ‘W’ had, and a gene that had mutated in Group ‘A’ but was later contributed to Group ‘B-prime’?
That would be hard (but not impossible, actually). But rather than explain the abstruse issues there, I will instead note that we aren’t faced with that, but with having present-day B-prime samples plus some fossil samples of A. In that case when we consider the times of coalescence of ancestry of individual gene copies, most of the loci will show the B-prime samples coalescing with each other long “before” (going back in time) they join up with the samples from the same locus in A. But a minority will show B-prime samples coalescing with some of the copies from A “before” they coalesce with the rest. It is those differences in gene genealogies, locus by locus, that have the signal.
Fascinating stuff. But why bother with “implications for Creationism”? It adds nothing to the argument by responding to the possible criticisms of a ridiculous and outdated theory. Wait until the blithering idiots start babbling on about where exactly their magic man intervened, and then slap them down. Until then, we can get along quite well without reference to risible nonsense.
BBC Science has an article by the Director of the Gibraltar on the Denisovans arguing that Denisovans, Neandertals, and modern humans are all members of a single polytypic species, Homo sapiens. He writes
and
http://www.icr.org/article/ancient-[…]man-species/
“Precisely.”
This sounds like Nebraska Man all over again. Who knows what this bone comes from. Evolutionists just assume it is an ape-man because it fits into some indeterminate, intermediate category between ape and human. We might find out it’s the bone of a chicken next week
I’m perfectly OK with thinking of them as more of a polytypic species. The more interesting thing is that we have gotten to the point in studying hominid fossils that we can make such a claim with clarity. Regardless of which side of the line one falls on, we have at least 4 distinctive groups (Cro-magnon, modern humans, Neanderthals, and Denisovans that basically sit right on that fuzzy area of the species definition and how far we have come in understanding since 1859. The level of detail at which we can look at human evolution is just breath taking. In particular the amount of variation these groups encompass is quite something whether it is skull morphology, brain size, height, robustness… I would argue that humans are one of our best examples of how evolution happens, a rather impressive change from Darwin’s deliberately vague statement in the first edition.
Update