How ancient DNA proved human origin theory wrong

What looked like a simple story of migration turned into a record of repeated mixing, vanished populations, and scientific shocks—including the discovery that many people outside Africa carry Neanderthal DNA, and that a little girl’s pinky bone once exposed an entirely unknown human lineage. Harvard geneticist David Reich argues that every new sample forces scientists to rethink the past, and that the biggest revelations about our ancestry may still be ahead.

DAVID REICH: The world is full of secrets about the past and how people are related to each other. And all my presuppositions and all my biases are probably wrong, because every time we get data from somewhere that we didn't have data from before, we turn out to be shown to be wrong.

The truth is that the world is a mystery, and that we've carved out areas of clarity. In 2007, it had just become cheap enough to generate genome scale data from lots of people living today. Today, this is routine, but at the time it was at just the price threshold had just been crossed and it made it possible to do this. So we were very excited to be also just studying how people today related to each other.

I was working mostly on medical genetics and trying to learn about risk factors for disease, and especially in African-Americans and Latino populations who our populations have recently mixed ancestry from Native Americans and Europeans and West Africans in the last handful of generations.

And then I got invited to participate in the analysis of the DNA that was being produced from ancient Neanderthals, ancient, archaic humans. So we got brought on to that consortium, which was this sort of amazing group of people working together to analyze this very, very special data. I thought, this is the best data in the world, and I want to spend my own time analyzing it.

So I changed everything I did and was completely clear that this data was revolutionary. We would be able to see how people who lived, 70,000 years ago, 50,000 years ago related to people living today. And these were very special people. These were, people whose skeletons were very different from any people living today. And we could ask all sorts of questions about when they separated from each other and maybe, maybe did they even mix with people as modern humans spread into the territory these people were living in? So it was an incredibly exciting thing to do.

And the results were even more exciting than we could have anticipated, because when we came into this work, we had this bias. I've had so many biases entering the different projects I've worked on. We had this bias that modern humans, as they spread out of Africa, had not mixed with the archaic humans. They encountered, like the Neanderthals. That had been the approximate finding from analysis of modern DNA. There wasn't really very strong evidence that there was an archaic component in Non-africans little hints here or there, but mostly it looked like a simple out of Africa spread. And so nobody really thought that there was good evidence for, archaic humans having mixed with non-africans during the spread.

But then when we looked at the data, there was very, very clear evidence that Neanderthals that we were sequencing from Europe were more closely related to people outside of Africa today than people inside Africa today. And then we kept trying to make this result try to go away, and it just wouldn't go away. And we looked at it in several different new ways with completely different types of ways of analysis. And it just got stronger and stronger. And it became absolutely clear that as modern humans moved out of Africa, they mixed with the local archaic population. And about 2% of the DNA from Non-africans today comes from these archaic humans.

And that was just the first of many, many surprises that comes out of this ancient DNA revolution. Really. In 2010, the same year that, we published this paper on the Neanderthals, we got DNA from another archaic human in Siberia that we completely had not expected at all. It was a pinky bone from a little girl, and it had no other skull or big morphologically informative parts to it. And we got a high quality genome sequence from it, and we could see how it was related to Neanderthals and to modern humans. And it was not either. It was something else.

And you look at it and like, bang! It's related in a special way to people from New Guinea on the other side of the world. Right. A very tropical area, even though this bone is found in Siberia. And what's that about? And you sort of are confronted this data and you say, I don't know anything anymore, right? Like the world is full of secrets about the past and how people are related to each other. And all my presuppositions and all my biases are probably wrong, because every time we get data from somewhere that we didn't have data from before, we turn out to be shown to be wrong.

And then the subsequent data just kept showing example after example. After that, Malawi and Cameroon and the Southwest Pacific and multiple examples and Europe and different groups had Russia, multiple groups in the Middle East, you know, the different groups in North Africa, you know, multiple groups in the Americas. Just example after example, after example, after example. It just seems to be the pattern. You really just get sort of clubbed over the head with the data, and you just have to change your approach to the data.

Instead, DNA is not more important than these other ways of learning about the past, like based on the tools people make or in later times, the languages people speak. But it's a quite different type of information. What's really happened is the introduction of a new scientific instrument that is able to measure things that has not been, that have not been measured before.

So there have been other scientific instruments that have been revolutionary in their area, arguably more important ones, like the telescope or the microscope. So when the telescope was first invented and people peered through it, they could see planets around Jupiter, right? Or they could microscope. You can see whole worlds of microorganisms you didn't even know about before that were not even invented before all these tiny creatures of like, protests and fungi and all of these things with amazing structures that were not even envisioned and people's philosophies and imagination.

And in the same way, when you can take DNA from people who lived in the past and compare it to each other and to people living today, you see things that you just didn't even imagine before the step migrations, the expansion of our pottery culture, people of East Asian ancestry to the Southwest Pacific. Every time we get DNA from Africa for more than 3000 years ago, it's a huge surprise. You look at Malawi and south central Africa, the people there 3000 years ago, nothing like the people there today. They're a population that simply doesn't exist anymore.

You look at Cameroon, which is supposed to be the homeland of the people who speak Bantu languages, which are spoken in the Congo area or in much of East Africa, or related to the languages that live in much of the West Africa. You would think, oh, this is just going to look like Nigerians or something, because that's what it should look like. Given that Nigerians look like most Kenyans and look look like most people from Zimbabwe. But no, it's like a completely different population that we didn't know about before.

This amazing study just came out from Libya of people who lived about 7000 years ago in the Sahara, a completely different population that we didn't know about before. So every time we get data, it's just surprising.

One of the puzzles, the great puzzles, in my opinion, that has been thrown up by the ancient DNA revolution, has been the one of how modern humans and archaic humans are related to each other. So we have sequences now from three very different groups of humans, from modern humans. They like people from sub-Saharan Africa for example, and their non-African descendants, and also Denisovans and Neanderthals. We have high quality genome sequences from multiple individuals, from each of these groups.

As a fruit of the amazing research that's been done, and through a lot of struggle, we know about how they're related to each other. We know when these lineages split from each other, mostly, and we've identified, in addition to those splits, some important mixture events, when these lineages interacted with each other at different periods of time.

But the standard model, we have right now is really almost hard to believe. So the standard model right now is that modern humans are a distant cousin of Neanderthals and Denisovans, who are most closely related. Neanderthals and Denisovans stemmed from a common ancestor maybe half a million years ago, which separated earlier from the ancestors of modern humans.

But actually, maybe there's some alternative way of thinking about what happened that can really change our vision of the relationships amongst these groups and make it sort of the Copernican readjustment, where we say, well, maybe, maybe the maybe the solar system revolves around the sun, and the Earth is a satellite of that. And maybe, maybe that's how it looks.

And so I kind of feel that we're missing a trick and that there's something there, and that maybe if we come up with a different model, it will be much more plausible. And so that's one of the exciting, frontiers that I think that we're like, hopefully that maybe in the coming years, would we as a community would make progress on.