Substack cometh, and lo it is good. (Pricing)

Human evolutionary history as a cryptic tangle

SHUpdate Ignore stuff on mutation rate. Confused mutation rate per year with per generation. Moral of story: read more closely! End update

The genetic data from the Sima de los Huesos (SH) hominins has finally been published in Nature. Nuclear DNA sequences from the Middle Pleistocene Sima de los Huesos hominins:

A unique assemblage of 28 hominin individuals, found in Sima de los Huesos in the Sierra de Atapuerca in Spain, has recently been dated to approximately 430,000 years ago…While the Sima de los Huesos hominins share some derived morphological features with Neanderthals, the mitochondrial genome retrieved from one individual from Sima de los Huesos is more closely related to the mitochondrial DNA of Denisovans than to that of Neanderthals…Here we recover nuclear DNA sequences from two specimens, which show that the Sima de los Huesos hominins were related to Neanderthals rather than to Denisovans, indicating that the population divergence between Neanderthals and Denisovans predates 430,000 years ago. A mitochondrial DNA recovered from one of the specimens shares the previously described relationship to Denisovan mitochondrial DNAs, suggesting, among other possibilities, that the mitochondrial DNA gene pool of Neanderthals turned over later in their history.

I don’t want to denigrate mtDNA, but the whole field of phylogeography and inference of evolutionary relationships in non-human organisms has been confused by the fact that many scientists drew a straight-line between the genealogy of a single locus, mtDNA, and the genealogy of a whole species. In light of results from molecular ecology, which long relied on mtDNA, but is finally moving toward genome-wide marker sets because of new technologies, I’m absolutely not shocked that mtDNA is not particularly predictive of whole genome relatedness. Recall that the Denisovan hominin had an mtDNA lineage which was more distant from Neanderthals than the whole genome turned out to be.

More importantly for this paper is that they find that whole genome inferences suggest that the SH hominins are genetically closer to Neanderthals than they were to Denisovans. I’ll skip over their quality control of the ancient samples…the group that published this paper is the best in the business, and it’s broadly persuasive to me…though I don’t know much about this topic, so you shouldn’t put much weight on my opinion. Because of the small amount of data retrieved they used a rather simple method to establish relatedness: shared derived mutations. Basically these are unique mutations which distinguish particular hominin lineages from each other and the outgroup population (other primates).

The above figure shows that the the SH hominin samples share much more with Neanderthals than with Denisovans, and more with Denisovans than with modern humans. From that one might reasonably infer then that:

1) The ancestry of modern humans diverged before the SH-Neanderthal-Denisovan clade.

2) But Denisovans diverged rather quickly from the SH-Neanderthal clade.

3) Since the SH hominin site seems to date to ~400,000 years before the present, and is found in part of the traditional range of Neanderthals, and these individuals share morphological characteristics with this group, the SH hominins may be the ancestral population, or related to the ancestral population, of Neanderthals.

How robust is this date? I don’t know enough geology or paleontology to judge, so I won’t follow the citations. But a lot of the discussion and novelty of these results seems to hinge on this date. From the paper:

This age is compatible with the population split time of 381,000–473,000 years ago estimated for Neanderthals and Denisovans on the basis of their nuclear genome sequences and using the human mutation rate of 0.5 × 10−9 per base pair per year…This mutation rate also suggests that the population split between archaic and modern humans occurred between 550,000 and 765,000 years ago.

A more conventional mutation rate I’ve seen is closer to 1 to 2 × 10−8, or two to four times faster than the one above. But the mutation rate literature is rather confusing for me, and does not totally align well together, possibly due to variation in rate over time.

The picture above is often derived from a model where we fit the diversification of the human lineage to a tree. But within the paper they suggest Neanderthal mtDNA might have been African, and that there may have been an earlier migration out of Africa before the one that occurred ~50-100,000 years ago. I don’t see a priori why this couldn’t be so, but I’m also not clear how we’re going to get a good grasp of the details of the dynamics at play.

Perhaps the dominance of the African “Out of Africa” lineage over the last 100,000 years is an aberration. It may be that much of human history was characterized by the sorts of meta-population dynamics described by classical multi-regionalists. The possibility that Neanderthal-Denisovan divergence might be as old as ~450,000 years before the present suggests to me that the massive replacement and assimilation we saw ~50-100,000 years ago was somehow atypical in terms of how it disrupted deep regional population structure….

Posted in Uncategorized

Comments are closed.