Neandertals on the edge of existence

A new paper in PNAS, Patterns of coding variation in the complete exomes of three Neandertals, reiterates what seems to be an emerging fact about ancient northern Eurasian hominins: they were rather inbred. The chart above illustrates it by focusing on regions of the genome that code for proteins. The ratio of benign to deleterious for Neandertals is obviously not optimal. As noted in the paper the authors suggest that this implies small population sizes which could not purge the deleterious variants through negative selection. The operative dynamic here is that when populations are small, drift becomes very strong relative to selection, and can therefore increase frequencies of mutations which would otherwise be swept out of the genome.

And it is clear from a variety of standard population genetic metrics that these Neandertals, separated geographically by the expanse of all of Eurasia, and distinct over ~20,000 years, would be considered quite inbred if they were examined today. This is curious because it is often stated that humans, our own neo-African lineage, are a relatively homogeneous population which expanded rapidly in size recently. We haven’t had that much time to diverge, and non-Africans tend to have genomes suggestive of bottlenecks in the deep past. And yet looking at these Neandertals, as well the Denisovan individual, modern humans seem a positively genetically variegated lot in comparison. But the authors of the above paper also found something else interesting, the Neandertals were themselves very distinct from each other, more distinct in terms of between population variation than modern human lineages. The main caveat is that the individuals were separated in time by thousands of years, in additional to geography, and the authors don’t seem to have corrected for this aspect of drift (though then they’d have to have an explicit demographic model, which they may not have enough information to construct). Additionally I do wonder if pre-Holocene humans also exhibited the same pattern of high genetic distance over small regions, and that only the demographic expansions and admixtures of the past 10,000 years have produced the reduced F_st values that we take for granted as expectations.

Finally, I have to emphasize that looking at hominins on the edge of the human range on the north may not be representative of archaic lineages more broadly. It seems possible that meta-population dynamics characterized by extinctions and expansions, and very low effective population size, were much more normal on the northern fringe of marginal habitation than further south in the core of the hominin range. Imagine taking Amerindians as representative of human variation, as an analogy. Because of the preservation bias in ancient DNA we’ll always over-sample these northern locales, but we should be wary of over-generalizing.

Addendum: I didn’t say much about the functional differences, because I’m not sure that we can say much with the data sets that they had. It is suggestive that they found differences in genes related to skeletal morphology in Neandertals, but that is not surprising. The fact that modern humans may be enriched for differences and evolution in regards to pigmentation seems likely to be a bias of the fact that this area has been well studied in modern humans. Obviously there haven’t been genome-wide associations for this trait in Neandertals, since we don’t have cases and controls (to my knowledge).