This week, Science published two papers about the genetics of Neandertals from a team of scientists based at the Max Planck Institute of Evolutionary Anthropology. The first (which is the only one anyone seems to really care about) gives a draft version of the entire Neandertal genome - a whopping 4 billion base pairs of DNA. They use this information to look for genomic regions that may have been affected by positive selection in ancestral modern humans that led to their separation from Neandertals, and found some very interesting ones that include genes involved in metabolism and in cognitive and skeletal development.
But, of course, that's not what everyone is saying about this paper. What everyone is talking about is that this study 'proved' ancient people f*cked Neandertals. Not only were there romantic trysts between these different hominid lineages, they led to human-Neandertal hybrids, the evidence of which is written in the genomes of all people from Europe and Asia.
In the words of the media...
There's the blunt and simple: "Modern man had sex with Neanderthals"
Or the creative reversal: "Neanderthals 'had sex' with modern man"
Or, my personal favorite: "Neanderthal Genome Shows Most Humans Are Cavemen"
But what did the paper actually say, and is the authors' conclusion that modern humans and Neanderthals screwed productively the only explanation?
The Neandertal Genome
How did they get the Neandertal genome anyway? Well, they took small powder samples of otherwise fairly useless bones from three female Neandertals, all of which were about 40,000 years old. Through some complicated methods, they turned these bone fragments into libraries of DNA chunks, each less than 200 base pairs long. They then took these fragments and pieced them together, using human and chimp genomes as road maps.
Now, the really tough part isn't getting and aligning sequences from Neandertal bones (though, that's not to say that it's easy). It's making sure you're sequencing the bones and not something else. Over the years, the bones have been handles by people, are coated in bacteria, and in general, are contaminated. When doing genetics work with many, modern-day animals, such contamination isn't as much of an issue - for the most part, primers are specific to one kind of organism or another, and the odds of samples being contaminated by a similar enough organism to matter are low. In my lab, for example, the odds that the samples I make from fish fin clips are contaminated with another fish's DNA are pretty low, so there's little worry that my primers will amplify something other than the gene I want.
But these guys are trying to amplify DNA from a creature that is very, very much like us. Furthermore, they don't know what to expect in terms of sequence, and have to use primers that are a little more universal to all primate species. This means that if even one skin cell from one of the scientists falls into a PCR tube, they'll be amplifying the wrong DNA to get their genome sequence. Such contamination has been a big issue in the past studies on Neandertal genetics. In this study, the authors went to every extreme to prevent contamination. Still, they estimate that human contamination in their samples may be as high as 0.5% for the mitochondrial genome and 1.53% for the nuclear genome. I'll explain why that matters in a bit.
The Good Stuff: What They Found
So what did they find? Well, for one, the divergence between human and Neandertal genomes was just a hair under 12.7%. Though you may have heard that you're 98% chimp, the actual sequence divergence between humans and chimps is actually closer to 30%, and scientists have estimated that this represents somewhere between 5.6 to 8.3 million years of separated evolution. Using that as a reference, Green and colleagues estimated that humans and Neandertals diverged about 825,000 years ago. But that's not the whole story, as any two people will have a certain amount of divergence between their genomes. The team also looked at the divergence between native people from Africa, Asia and Europe, and found that the overall sequence divergence between any two people was 8.2% to 10.3%. The Neandertal was significantly more divergent from modern humans than lineages of modern humans are from each other, but the new information changed the time line a little. In the end, they estimate that modern humans and Neandertals became distinct between 270,000 and 440,000 years ago
What the scientists really wanted to know, though, was what was different. Were there clues to how human intelligence, or any other traits, evolved, based on the information contained in our closest relatives? A large part of the paper sought to find evidence of positive selection on the human lineage after the split from Neandertals.
They did this two ways. First, they looked for genes that were the same in Neandertals and chimpanzees, orangutans, and rhesus macaques, but were very different in us. There weren't as many differences as you might think. Only 78 of our 20,000 or so protein genes have changed their amino acid sequence since the split from Neanderthls. They also found 268 changes in sequences that don't code for proteins, but instead function other ways, like as promoters, silencers, or microRNAs.
They also looked for stretches of genetic material that are symptomatic of natural selection. How does a set of genes "look" like it's undergone selection? Well, when a single allele (variation of a gene) is strongly favored due to selection, its proportion in the population rises. Sometimes, it becomes so common that the rest of the alleles disappear, and this is what is called a "selective sweep." Scientists can find these sweeps by looking at Single Nucleotide Polymorphisms (SNPs), or areas that differ by a single base pair. Most of the areas that look this way between chimps and people are similar between people and Neandertals, because they occurred after hominids split from primates but before the human-Neandertal break. But, some have occurred since. In all, they found 212 regions that showed evidence of this kind of selection.
Somewhere, in those 350-550 odd changes, is likely what makes us so different from anything else that has every lived - where, though is still a mystery.
But wait - what about the sex?
Yeah, yeah, I'm getting to it. Geez! Y'all have a one-track mind.
After plowing through most of the paper, one arrives at a section titled "Neandertals are closer to non-Africans than to Africans." In it, the scientists lay out their argument for inter-species sex. What they did to determine this, in their own words was:
"To test whether Neandertals are more closely related to some present-day humans than to others, we identified SNPs by comparing one randomly chosen sequence from each of two present-day humans and asking if the Neandertals match the alleles of the two individuals equally often. If gene flow between Neandertals and modern humans ceased before differentiation between present-day human populations began, this is expected to be the case no matter which present-day humans are compared."
So they looked at how frequently the Neandertal genome matched each of the people from different locations to see if one lineage of modern humans was more similar to the Neandertals than another.
They found that when they compared a European and an Asian to the Neandertals, there was no significant difference. But when they compared an African and any non-African lineage, the Africans were less similar to the Neandertals than other people. They estimate that 1-4% of non-Africans genome is Neandertal genes.
Now, the obvious first question is whether or not contamination explains this. After all, they said the Neandertal nuclear sequences could have as high as 1.53% human contamination - which seeing as the scientists themselves are of Eurasian decent, would certainly skew the Neandertal's genome that direction. But they do give good explanations as to how it's not likely that this effect is due to contamination (which involve some complex math and stats that I just can't see an easy way to break down).
The media stopped reading the paper right about there. Indeed, the authors strongly believe that their data suggest that people and Neandertals interbred. But there are other explanations, even given by the authors, that would cause the patterns they saw.
For example, it's possible that Neandertal's didn't split from all people, but a subset that was on its way to becoming Eurasians. This is known as the "substructure within Africa" hypothesis, where different groups of modern humans actually began splitting before leaving the continent in search of the rest of the world. Carl Zimmer explains this idea rather nicely:
"Cast your mind back 500,000 years, before the populations of humans and Neanderthals had diverged. Imagine that those ancestral Africans were not trading genes freely. Instead imagine that some kind of barrier emerged to keep some gene variants in one part of Africa and other variants in another part.
Now imagine that the ancestors of Neanderthals leave Africa, and then much later the ancestors of Europeans and Asians leave Africa. It's possible that both sets of immigrants came from the same part of Africa. They might have both taken some gene variants with them did not exist in other parts of Africa. Today, some living Africans still lack those variants. This scenario could lead to Europeans and Asians with Neanderthal-like pieces of DNA without a single hybrid baby ever being born."
The substructure in Africa hypothesis is somewhat supported by other genetic and paleontological analyses. Indeed, the authors note that they "cannot currently rule out a scenario in which the ancestral population of present-day non-Africans was more closely related to Neandertals than the ancestral population of present-day Africans due to ancient substructure within Africa"
There's another possible explanation, too, that the authors didn't bring up: selection.
Let's jump back those 500,000 years and take a look at our pre-split hominid population. Let's say, just for argument's sake, that a given region had 10 variations in that group of individuals - we'll label them A, B, C, D, E, F, G, H, I and J. Now, a group of them breaks northward. Due to founder effects, this small group only has A, B, C, D and E to start with. As the rugged north takes its toll, certain variations prove more useful - A and B, for example. Soon enough, genetic drift and selection leaves this population with only A, B and D.
Meanwhile, the original population in Africa has been evolving. They have A. B. C. D. E. F. G. H, I, J and how K, L, M, N and O. A small group from that population breaks off, too, and follows the path taken by the first subgroup hundreds of thousands of years earlier. Again, due to founder effects, this new little group has only a small amount of the variation that is in Africa - they take with them A, B, C, E, F, G, H, I, M, N, and O. As they face the frigid winds and weakened sun, they are whittled down to only those that give them the best advantage in the north - A, B, C, E, F, and O.
Meanwhile, the African population is evolving, too. They find that some of their variations give them a better chance in the African sun, like H through O. Over time, A, B and C just disappear.
So now you have one population with A, B and E, one with A, B, C, E, F and O, and one with E, F, G, H, I, J, K, L, M, N and O. Although the two groups that stayed in Africa longer are more closely related, the group that left first will be more similar to the second split off group because selection favored the same variations in both that were present before the first split occurred. So you see, it's possible that selection could explain some of the similarity between Eurasians and Neandertals, without a single person getting hanky-panky with another hominid.
That's not to say that the hybridization hypothesis is without merit. Indeed, the authors did a damned good job presenting their case, and their reasoning is sound. The point I'm making is that sex isn't the only option. And if hybridization did occur to the extent they predict, we're likely to find more hints at its existence. Analyzing the DNA of some of the suspected hybrid fossils, for example, might settle it once and for all. Or we may never know if the gene variants that are similar between Eurasians and Neandertals are due to sex, selection, or substructure. Time may have simply destroyed too much of the evidence for us to be sure.
What is certain is that sex sells, which is why the only thing the media is talking about when it comes to this paper is that ancient people may have shagged their evolutionary siblings. It's just so damned frustrating because the sexual exploits of early humans is only the tiniest piece of this huge discovery. Oh, the things we may learn from this genome about our own evolution, and our closest relatives! Whether we had sex with Neandertals or not, the work this team has done will change forever our understanding of hominid evolutionary history. The impact this genome will have on the science of human evolution is huge. The breakthrough science, the future implications of this work - that's what the media should be talking about - not ancient sex scandals.
Green, R., Krause, et al. (2010). A Draft Sequence of the Neandertal Genome Science, 328 (5979), 710-722 DOI: 10.1126/science.1188021
I saw these articles popping up and as soon as I did I started wondering what the research actually said. I'm so glad you did a write up on it.
Great explanation of the article. Thank you. In my opinion, the proximity of humans and Neanderthals virtually guarantees interbreeding. Given the opportunity, when have humans ever failed to kill and rape groups of people living nearby? Assuming that there was no such interaction assumes that prehistoric humans behaved completely different than at any time during recorded history.
That was a great read. Thanks!
Let me see if I understand the "substructure within Africa" hypothesis properly. If we were to arrange a phylogenic tree, it would basically look like this.
__________African Homo sapiens
......\____Non-African Homo sapiens
As a result, Homo sapiens as it has been come to be defined is polyphyletic, and thus a "bad taxon." One alternative, which is splitting Africans and non-Africans into different species, is clearly wrong, as modern humans of all races interbreed freely. Therefore, Neandertals are Homo sapiens. I don't even think you could excuse them being a subspecies any more, unless you also proclaimed there were two modern human subspecies, one for Africans, and one for everyone else, and I believe the general consensus is human racial differences do not warrant subspecies.
Practically speaking, it's also hard to defend interbreeding not happening under this scenario, given two more genetically dissimilar populations interbreed gladly. I suppose you could argue Neandertals had a unique mutation like changing the chromosome count which made fertile hybrids impossible. Things like a novel change in courtship structure wouldn't cut it, given how...indiscriminate...male humans can sometimes be, particularly regarding consent.
I'm unaware if there's anything in nature showing polyphyletic species which regularly interbreeds which as a monophyletic second species "nested within" the taxa. There are of course some examples of artificially-selected populations which almost fit (e.g., domesticated dogs). I'd be happy to be proven there is one though.
Though you may have heard that you're 98% chimp, the actual sequence divergence between humans and chimps is actually closer to 30%
Does this mean that humans and chimpanzees are common over 98% of their entire DNA, junk and all, but that the bits that actually code for anything are only 70% common?
@5 No, the 98% figure references a different method of determining genetic similarity, not based on comparing sequences. Something to do with dissolving bits of DNA from two species in solution and seeing how well the different strands stick together chemically. (I am a mathematician, not a biologist or chemist, so I don't know the details.)
I hesitate to try to correct somebody of your stature, but I believe the folowing diagram better illustrates the "substructure within Africa" hypothesis. And it does away with the "nested" taxa problem.
..........____Non-African Homo sapiens
______/____African Homo sapiens
I love the article, Christie. It's very informative. Thanks.
Still, the title got me to read this post though I was in a rush and didn't have time to read everything in my reader. The whole post was fascinating and I was just as interested in the alternative explanations. Thanks for putting it all so clearly for a non-scientist.
Now I know why I am so smart. I have TWO species running around in my DNA! So, take that One DNA'ers!
I may be a layman (not that you'd have any way of knowing that offhand besides not having a web site), but I'm fairly well read in paleoanthropology and genetics for someone with no formal scientific background. There's no reason to belittle me.
In the original post, Christie quotes the authors of the paper as saying they "cannot currently rule out a scenario in which the ancestral population of present-day non-Africans was more closely related to Neandertals than the ancestral population of present-day Africans due to ancient substructure within Africa." Unless I misunderstand the importance of the caveats of "ancestral" in this case, the meaning seems as clear as day - Non-Africans may be more closely related to neandertals than Africans, as they shared a more recent common ancestor. The longer quote from Zimmer says much the same thing. Thus, Homo sapiens is polyphyletic.
I suppose you could turn the argument on its head, and say that modern Africans have unique genetic signals due to novelty, while neandertals and modern Non-Africans kept a more "basal" genetic structure. This would be parsimonious anyway. But this isn't what the "substructure within Africa" is presented as here.
FWIW, I think the morphological evidence heavily leans toward your more conventional tree being a more accurate description.
Well, I do notice that you didn't title your piece "350-550 Exciting Ways Your Genome Differs From Neandertal's" :) (maybe that one will be used by the editors of Redbook).
But, yeah, point well taken -- the implications of this research are far bigger than possible proving interbreeding. In fact, given modern human's seeming inability *not* to have sex with just about anything they can fit to their genitalia, I'd say the chances of human-on-neandertal action have always been and are still, uh, a slam-dunk.
So, anthropologically the whole sex scandal angle is pretty much a yawn. But just imagine the possibilities for new museum dioramas!
Karl and Bill:
Actually, you can have Neandertals be more closely related to Non-Africans *without* having a polyphyletic Homo sapiens. The trouble you're having is that substructure in Africa may be more complicated than a bifurcating tree can show. For example, let's say that some kind of natural disaster suddenly creates a river in a valley, separating an endemic species there into two populations - a large population A and a smaller one, B. For awhile, this river persists, and the two populations begin to accumulate some genetic differences. A small subset of the already small population B, then, climbs out of the valley, and goes onto become a whole new species (C) somewhere else. Meanwhile, the river dries up, and the two populations in the valley can mix, because they haven't changed enough during that time apart to speciate from each other. But they don't just merge together - instead, some from population A join up with B. Then the group of B (with a little A) leaves the valley, too, leaving the rest of A behind. At the point B (with a little A) leaves, how do you draw the tree? C will be more closely related to B than A, but A and B are still the same species. That's the idea of the substructure within Africa - that gene flow was restricted (particularly in one direction), but not blocked, between two (or more) populations of humans, and from one of those, Neandertals diverged. There's no easy way to draw that in a strict bifurcating tree.
Christie doesn't that still mean Neandertals are still Homo sapiens? From John Hawk:
"Does this mean that Neandertals belong in our species, Homo sapiens?
Interbreeding with fertile offspring in nature. That's the biological species concept.
Now, some paleontologists might still disagree -- maintaining that species are units that can be distinguished morphologically, or by one or more derived features, or any number of other definitions. That's fine with me, as long as they're clear. But understand: It does define all non-Africans today as an interspecific hybrid population.
So maybe they want to rethink that one?"
Thanks for the reply. It makes a lot of sense, and shows how cladistics can have weaknesses at the micro level. It also explains pretty neatly how mitochondria and Y chromosomes have a more recent common ancestor than non-Y nuclear genes presumably do in some cases.
That said, it still seems to me the substructure theory puts the burden of proof on showing Neandertals were a distinct species, instead of showing they were not.
I guess a useful parallel is polar bears. Genetic studies have shown they originated in a brown bear population roughly 150,000 years ago. They seem to have originated from an isolated population of brown bears during the ice ages. They are, genetically speaking, closer related to some brown bears than all brown bears are to each other. They are morphologically quite distinct, and occupy a quite different ecological niche. That said, they can produce fertile offspring with brown bears, and do so in the wild on occasion, thus it's arguable if they're actually a true species.
Karl, last I heard Neanderthals were a subspecies of Homo sapiens.
Interbreeding in nature and producing fertile offspring is not necessarily evidence that the interbreeders belong to the same species. It does make one wonder, however. The real question is how much gene flow, introgression, occurs between the interbreeders. There are a number of situations in fishes where there is production of fertile interspecific hybrids but little or no gene flow between the two species.
Karl, last I heard Neanderthals were a subspecies of Homo sapiens.
Katherine depends on who you talk to. Some maintain Neandertals are a separate species. Since most people learn a simplified version of species (including those within the biological sciences), the possibility of sex between "modern humans" and another species is thus viewed as very risque and is the reason for the hype we are seeing.
It becomes a less exciting story to the general public if it is sex between different groups of humans.
Separating neandetals from Homo sapiens also plays into our tendency of playing up our "uniqueness" and differences with other organisms. Really should us humans have our own genus? Or is that an artifact of our desire to be unique?
It has been suggested that Pan should be synonomized with Homo, given the close similarities at the DNA level. If I were looking at two fish sister groups with as much morphological difference as there is between Homo and Pan, I would maintain them as separate genera. Should Homo be expanded to include some of the fossil forms now placed in other genera? I don't know enough about it to have an opinion.
"the divergence between human and Neandertal genomes was just a hair under 12.7%. Though you may have heard that you're 98% chimp, the actual sequence divergence between humans and chimps is actually closer to 30%"
12.7% and 30% of what? 98% of what? are these numbers even comparable?!
If it is really hybridization, then Neanderthals may have given us most of the geniuses. Just sayin'. Check this out: http://rdos.net/eng/asperger.htm
What about blood type in comparison to modern humans? I remember reading an article in New Scientist back in January 09 that said Neanderthals would have made good blood donors as they had the 'universal donor' type. As far as I know (from my limited scientific knowledge), that would mean Neanderthals were mostly type O rhesus negative. I'm guessing if they have now mapped out the Neanderthal DNA they can say this for sure?