High-coverage Neanderthal genome and gene flow with Denisovans

Two new preprints on Neanderthals, A high-coverage Neandertal genome from Chagyrskaya Cave and 100,000 years of gene flow between Neandertals and Denisovans in the Altai mountains.

The first preprint is an empirical one focused on a new high-coverage Neanderthal genome, which allows for more powerful inferences. To me, the most interesting insight is that Neanderthals seem to have been a highly structured population. This is something we knew from other research, but they used the fact that European Neanderthals seem to have medium-length runs of homozygosity to get a better sense of it. One modern group with lots of medium length runs of homozygosity are Ashkenazi Jews, who went through a very modest bottleneck and were highly endogamous. This suggests that Neanderthal subpopulations did now have much gene flow between them and that their demes were rather viscous.

Additionally, the contribution of Neanderthal ancestry to modern humans seems to come from one particular subpopulation. This reiterates that even if there were multiple admixtures (there were), in non-African humans the vast majority of the Neanderthal admixture comes from a single pulse.

The second paper focuses more on broad population genetic questions and theory. Using an HHM-based framework the author looks at various Neanderthal and Denisovan genomes and infers long-lasting and repeated instances of reciprocal gene flow between the two populations in the Altai. This is entirely expected. But, the author notes that Denisovan ancestry isn’t present in European Neanderthals. This is in keeping with the earlier preprint, which highlighted a strong degree of population structure. Presumably the Altai was a zone of contact between these two Eurasian hominin populations, but it did not serve as a mediator of gene flow due to high structure in both groups, and repeated local extinction (later Altai Neanderthals are very different from earlier Altai Neanderthals).


The telos of modern humans

Credit: Luke Jostins

Next week’s episode of The Insight is going to be on Denisovans. It’s a long one because so much has come out in the last few months on the specific topic, as well as the broader framing issues (e.g., the discovery of a new human species on Luzon).

One of the major points Spencer and I discussed is how important it is to understand general trends in the hominin lineage, that is, humans, before the great expansion ~60,000 years ago. For example, Neanderthals and Denisovans were very different in their paleoecology and biogeography. Neanderthals seem relatively homogeneous (probably due to repeated mass die-offs). In contrast, the “Denisovans” look to have been very deeply diverged within their clade. If the latest work is correct, and some Denisovan lineages split more than 400,000 years ago and persisted down to >100,000 years ago, then the differences between Denisovans may have been considerably greater than between any modern human lineages. For example, the Khoisan diverged from all other humans ~200,000 years ago, and there are possible deeper lineages, but not that much deeper.

Right now what you know about the Denisovans are from genomes in the Altai region. Imagine if we extrapolated to all modern humans from Altaians? It seems entirely likely that the Denisovan lineage was very diverse because it occupied very diverse territory geographically.

But diversity aside, one of the things I like to point out to people, is that there was an overall trend of encephalization among hominins. Neanderthal brains were growing larger too. We need to understand the natural history of all human lineages to understand what happened 60,000 years ago. I am coming to the conclusion that it wasn’t some incredible miracle of a behavioral big bang, but the inevitable outcome of systemic forces in hairless ape evolution that started ~2 million years ago.


Denisovans’ biogeography is very different from Neanderthals’

I want to elaborate on my earlier post, Deep Denisovan Population Structure. Though I don’t put much stock in any particular result, including the most recent ones reported at a conference, I think that biogeography tells us a lot about what we should expect in the future.

First, notice that Neanderthals, Denisovans, and European hunter-gatherers have all exhibited evidence of being subject to massive bottlenecks and very low effective population sizes. Things changed in Europe with the arrival of Neolithic farmers, whose higher genetic diversity is in the range of modern people. I think this is indicative of the fact that abiotic factors, climatic shocks, drove down population sizes periodically consistently with all these northern hominins. I’m not saying that population sizes couldn’t get large periodically, but the fact that later Altai Neanderthals are genetically closer to Neanderthals from Croatia than earlier Altai Neanderthals indicate lots of local population extinction. It applied to Neanderthals, it applied to northern Denisovans, and, it applied to northern modern humans. Only with a change in the mode of production were higher long-term population sizes feasible.

Read More


Deep Denisovan population structure

The Denisovan session at the American Society of Physical Anthropology meeting was very interesting. In Science Anne Gibbons reports on the findings, Our mysterious cousins—the Denisovans—may have mated with modern humans as recently as 15,000 years ago:

The elusive Denisovans, the extinct cousins of Neanderthals, are known from only the scraps of bone they left in Siberia’s Denisova Cave in Russia and the genetic legacy they bequeathed to living people across Asia. A new study of that legacy in people from New Guinea now suggests that, far from being a single group, these mysterious humans were so diverse that their populations were as distantly related to each other as they were to Neanderthals.

The finding of two Denisovan lineages in Southeast Asia adds to results reported in Cell last year by Sharon Browning of the University of Washington in Seattle and her colleagues. They had suggested that New Guineans had a separate source of Denisovan DNA than people in East Asia, suggesting at least two mixing events.

“I’m skeptical,” added Cosimo Posth of the Max Planck Institute for the Science of Human History in Jena, Germany. He suggests the hints of a late mating could reflect an encounter of previously isolated modern populations rather than of moderns and Denisovans. In this scenario, modern humans mated with Denisovans, then the modern populations diverged, with each branch retaining a different set of Denisovan genes. The moderns then reconnected, mixing the two sets of Denisovan DNA together again.

Here is one thing that I think is important to remember: unlike Western Eurasia parts of Eastern Eurasia were better insulated from extreme climatic events. Neanderthals show strong evidence of repeated die-offs and population expansion so that in general Neanderthal relatedness is more a function of time than location (i.e., Neanderthals tended to go extinct in much of their range periodically, to be repopulated from refuges).

In contrast, the Denisovan range likely went far into Southeast Asia. It is not surprising that this is a highly structured population, with deep lineages. This is exactly what we see in Africa for the same time period. Tropical Southeast Asia is not as extensive as Africa, but it was more expansive during the Pleistocene due to lower sea levels. Hominins with low population densities occupying a huge range of territory almost certainly had developed local lineages and traditions.

As should be clear in the quotes we shouldn’t take these presented results as definitive. But they are suggestive and align well with earlier work that there were several Densiovan admixtures across Eurasia.