Razib Khan One-stop-shopping for all of my content

August 31, 2012

Deep dive into the Denisovans

Filed under: denisovan,Genomics,Human Genetics,Human Genomics — Razib Khan @ 12:25 am

By now you have probably seen the new Denisovan paper in the media. John Hawks has an excellent overview, as you’d expect. The only thing I will add is to reiterate that I think population movements in near and far prehistory significantly obscure our comprehension of the patterns of past genetic variation. One reason that the Denisova hominin presents conundrums (e.g., how did Australians and Melanesians admix with a population whose only remains are found in Siberia) is that we’re viewing it through the lens of the present. What other lens can we view it through? We’re not time travelers. But we should be perhaps more conscious of the filter which that imposes upon our perception and model of the world. This is probably a time when it is best to have only a modest confidence in any given proposition about the prehistoric past.

Also, I don’t know why, but I much like this tree:

January 30, 2012

Out of who knows where

In The New York Times, DNA Turning Human Story Into a Tell-All:

The tip of a girl’s 40,000-year-old pinky finger found in a cold Siberian cave, paired with faster and cheaper genetic sequencing technology, is helping scientists draw a surprisingly complex new picture of human origins.

The new view is fast supplanting the traditional idea that modern humans triumphantly marched out of Africa about 50,000 years ago, replacing all other types that had gone before.

Instead, the genetic analysis shows, modern humans encountered and bred with at least two groups of ancient humans in relatively recent times: the Neanderthals, who lived in Europe and Asia, dying out roughly 30,000 years ago, and a mysterious group known as the Denisovans, who lived in Asia and most likely vanished around the same time.

Their DNA lives on in us even though they are extinct. “In a sense, we are a hybrid species,” Chris Stringer, a paleoanthropologist who is the research leader in human origins at the Natural History Museum in London, said in an interview.

First, for reasons of novelty we are emphasizing the exotic tendrils of the human family tree. Even Chris Stringer, the modern paleontological father of “Out of Africa,” is claiming we’re hybrids! But let’s not forget that non-Africans are the product of a very rapid radiation out of the margins of the Afrotropic ecozone within the last ~50-100,000 years. I am not entirely sure that this is as true of Africans (recall how extremely basal Bushmen are to the rest of humanity; they seem to have diverge well before the “Out of Africa” pulse).


Second, the old model was way easier to write about, even if there were confusions like the idea that mtDNA Eve was our only female ancestor from 200,000 years ago in the past. The new paradigm leaves one with awkward and unhelpful turns of phrase. For example:

But Dr. Reich and his team have determined through the patterns of archaic DNA replications that a small number of half-Neanderthal, half-modern human hybrids walked the earth between 46,000 and 67,000 years ago, he said in an interview. The half-Denisovan, half-modern humans that contributed to our DNA were more recent.

How to make sense of this gibberish? I suspect that the author didn’t have a good idea how to translate a particular population genetic statistic, and its importance to assessing time since admixture, into plainer prose. I have no idea either!

In other news, i09 has an interesting interview up with Rebecca Cann and Mark Stoneking. These two were heavily involved in the mtDNA Eve controversies of the 1980s. Nice capstone to an era. Like Stringer, even they admit the likelihood of a necessity to modify the simple “Out of Africa” with replacement model.

January 21, 2012

The erectus within?

The Pith: More caveman admixture in modern humans, especially Melanesians!

A new paper on archaic adaptive introgression among Melanesians has been discussed elsewhere. But I think it is worth reviewing, because it’s probably a foretaste of what’s to come. Researchers are combing through the human genome, as more and more genomes come on line, in the search of weird and unexpected variation. The paper is in Molecular Biology and Evolution, and is titled Global genetic variation at OAS1 provides evidence of archaic admixture in Melanesian populations (why is it that this journal doesn’t even allow supplemental information to be free to the public?). The two primary figures from this paper do a good job of illustrating the main result.

The first figure is a phylogenetic tree of haplotypes at the OAS1 locus, with pie charts showing the proportion of individuals from a set of populations which contribute to the total number for that haplotye. So you see above that the “deep lineage” is relatively distant from a cluster of other haplotypes (as measured by mutational differences which are proportional to depth of common ancestry), and, that deep linage is exclusively found in Papuans in this set. The second figure shows the frequency of the deep lineage haplotype over a larger set of populations. I cut off the section which shows that Africans are at zero percent. The haplotype is found almost exclusively in Melanesian populations, except for the fact out of over 200 South Asians they sampled, 3 of them carried it (2 Pakistanis, 1 Sri Lankan). There is aspect though not evident in the figures above, but which is clear in the abstract that you need to know:

Recent analysis of DNA extracted from two Eurasian forms of archaic human show that more genetic variants are shared with humans currently living in Eurasia than with anatomically modern humans in sub-Saharan Africa. While these genome-wide average measures of genetic similarity are consistent with the hypothesis of archaic admixture in Eurasia, analyses of individual loci exhibiting the signal of archaic introgression are needed to test alternative hypotheses and investigate the admixture process. Here, we provide a detailed sequence analysis of the innate immune gene, OAS1, a locus with a divergent Melanesian haplotype that is very similar to the Denisova sequence from the Altai region of Siberia. We re-sequenced a 7 kb region encompassing the OAS1gene in 88 individuals from 6 Old World populations (San, Biaka, Mandenka, French Basque, Han Chinese, and Papua New Guineans) and discovered previously unknown and ancient genetic variation. The 5′ region of this gene has unusual patterns of diversity, including 1) higher levels of nucleotide diversity in Papuans than in sub-Saharan Africans, 2) very deep ancestry with an estimated time to the most recent common ancestor of >3 million years, and 3) a basal branching pattern with Papuan individuals on either side of the rooted network. A global geographic survey of >1500 individuals showed that the divergent Papuan haplotype is nearly restricted to populations from eastern Indonesia and Melanesia. Polymorphic sites within this haplotype are shared with the draft Denisova genome over a span of ∼90 kb and are associated with an extended block of linkage disequilibrium, supporting the hypothesis that this haplotype introgressed from an archaic source that likely lived in Eurasia.

There there is “more genetic diversity within Africa” is a cliche rooted in reality. But, this is not true at all genes. For example, at MC1R Africans have less genetic diversity than Europeans. Why? MC1R is implicated in pigmentation, and this locus is subject to strong functional constraint at low latitudes. In other words, there is more at work than just demographic history. When you see more diversity at a locus outside of Africa than within there is a strong suspicion that natural selection or admixture may be at play, because your null expectation is that the dominant “Out of Africa” event will imply more Africa diversity or modern humans. So at OSA1 you have a genetic variation which is very diverse, and very divergent, from the modal human variant. When you take away the deep lineage you also see a pattern which is constant with genome wide expectations. Africans are distributed across the unrooted tree, while non-Africans seem to be nested within a subset of nodes.

But if this was published in 2008 it might not be as notable, because the human genome is big, and there are going to be random patterns here and there. This might have been dismissed. The key is that the authors matched this divergent haplotype to the variant found in the draft Denisovan genome. Naturally it’s going to be harder to dismiss as a statistical fluke when you actually have concrete evidence in this form that the ancient lineage was shared with an archaic hominin group.

There are two scientific points that jump out at me. First, the authors don’t discuss adaptation or selection in very much detail (except to dismiss balancing selection). But if this is due to archaic admixture its fraction in Melanesians is far higher than the genome wide average. Again, some loci will naturally deviate from expected values, but those that do are excellent candidates for being targets of adaptation. And, this gene has a clear functional role related to immune response. Second, the presence of this haplotype in South Asians is strongly suggestive of the location of the admixture event between archaic humans and the ancestors of modern Melanesians. This sort of information needs to be synthesized with the two papers last fall that came out on Australian evolutionary genomics.  One of the interesting aspects of the Denisovan admixture analyses is that it doesn’t seem that any South Asian group, including Andaman Islanders, exhibit it. And yet a few South Asians here carry a haplotype similar to the Denisovan variant. Interestingly, the authors present a rather unbelievable large value for the common ancestor between the deep lineage and other modern haplotypes, ~3 million years, which is an order of magnitude more than the divergence of Denisovans from modern humans.

How to resolve this confused situation? In the conclusion they point to the possibility that this haplotype may have introgressed into both the Denisovans and modern humans from H. erectus! The most recent genomics on Melanesians implies that their own history is relatively complex. On the one hand they may be some of the earliest distinct migrants out of Africa, and secondarily, they themselves may be successive compounds between those early migrants, archaic humans, and a second wave of Eurasians. All that being said, I think there is some hope in the combination of full genome sequencing of modern populations as well as the same of ancient populations via DNA from subfossils. The main qualification is that I doubt we’ll ever get good samples of ancient DNA from the tropics.

Citation: Fernando L. Mendez, Joseph C. Watkins, and Michael F. Hammer,
Global genetic variation at OAS1 provides evidence of archaic admixture in Melanesian populations, Mol Biol, doi:10.1093/molbev/msr301

Image credit: Wikipedia

October 31, 2011

The unbearable thinness of Denisovan

Filed under: Anthroplogy,denisovan,Human Evolution — Razib Khan @ 10:54 pm

A new paper in PNAS, Archaic human ancestry in East Asia: “These results suggest admixture between Denisovans or a Denisova-related population and the ancestors of East Asians, and that the history of anatomically modern and archaic humans might be more complex than previously proposed.” It’s open access, so do go read it. John Hawks has a long rumination. My main thought is that I’m starting to think that people are squeezing this orange too much. I wouldn’t be surprised if the broad conclusions here are correct, and in fact I’d lean in that direction. But is the discovery of relatively trace ancestry all that earthshaking? The reality is that a little over a year ago the interpretative framework of science in this area shifted. That was because of the concreteness of ancient DNA, which allowed for a direct comparison, instead of statistical sifting through the genomes of extant populations. Remember, before 2010 there were plenty of papers utilizing subtle statistics and computational muscle which “proved” and “confirmed” an Out-of-Africa with replacement model. The power and precision of these techniques tended to overshadow the reality of a margin of error, and uncertainty in their conclusions. We need to be cautious when the machinery turns itself in the opposite direction, gleaning glimpses of what we now know is likely there….

August 25, 2011

The divisibility of human ancestry

The class human or H. sapiens refers to a set of individuals. On the grand scale it’s really not all that clear and distinct. When do “archaic” humans become “modern” humans? Taking into account human variation, what is a “human universal”? A set of organisms are given a name which denotes the reality that they may share common ancestry, and interact behaviorally, and are potential mates. But many of these phenomenon are fuzzy on the margins. Many of the same issues which emerge in the “species concept” debates are rather general up and down the scales of natural complexity. A similar problem crops up when we conflate the history of genes with the history of populations. Such a conflation has value and utility to a first approximation. The story of mitochondrial Eve was actually the history of one particular locus, the mitochondrial genome. But it did tell us quite a bit about the history of the human species, even if in hindsight it looks as if some scientists overinterpreted those findings. One of the major issues I’ve noticed over the past year, with the heightened likelihood of ...

Powered by WordPress