The work that I've been involved in has consistently shown that I was wrong in my biases coming into the work, and I've really been almost traumatized by this. Like, again and again I've come into a project with some kind of guess about what the data was showing, and then the data doesn't show that. So, for example, when I got involved in the Neanderthal genome project and helping to analyze data looking at how archaic Neanderthals were related to modern humans, I was part of a group of scientists who had established that non-Africans were a simple subset of African variation, and that there was no evidence at all of Neanderthal interbreeding into the ancestors of modern humans or other archaic interbreeding. Different analyses that I and very much more other people had done made it look like non-African variation was just a subset, a small sample of that in Africa, and that could have fully explained the data. And so, when I was involved in analyzing the Neanderthal DNA sequences, what happened was I found this very strong evidence of Neanderthals being more closely related to non-Africans than to Africans, and it was very surprising and I thought it must be a mistake. I was quite incredulous. I thought it was unlikely to be true because other evidence that had been found before seemed to point in the other direction. And so, I spent several years trying to make these results go away, as did my colleagues, and we just couldn't make the results go away. They just kept getting stronger.
And this experience working on natural selection was the same. So, what we had felt here was that what we were convinced of was that natural selection had been pretty quiescent in our species over the last several hundred thousand years. Therefore, if we look at patterns of variation in non-African people today, or in any people today, we should see not a lot of selection going on. And indeed, the first ancient DNA studies beginning in 2015 with this paper that we were involved in with Ian Mathieson and colleagues, indeed these papers seem to show relatively small numbers of genetic positions associated with natural selection. So, in 2015, we analyzed data from about 200 Europeans and Middle Easterners to try to understand frequency changes over time. And we compared those ancient people who were the sources of modern Europeans to people in Europe today. And we looked at frequency differences that were too extreme to be due to chance. We were very excited to find 12 positions that we were convinced were highly different in frequency between Europeans today and what we would expect based on the history that we and others had identified was the history relating modern to ancient Europeans. And so, some of these were known and some of these were not known and this was very exciting. And we hoped that as the numbers of samples would increase and we would get higher resolution to be able to appreciate differences in frequencies over time, we hoped that this would make it possible to detect far more. And what was quite disappointing over the subsequent decade is that that didn't happen. So, for example, the largest study of that type in 2024 by a group in Copenhagen analyzed the data, much better data than we had in 2015, and found only 21 positions that were highly different in frequency across time. And while that was exciting, it was almost twice as many as we had found in 2015, in a lot of ways it was disappointing because the sample size and data quality had gone up so much and yet this is all that was found. And so, what that suggested is that we might be hitting an asymptote and we might not be able to get beyond where we currently were and that this approach to learning about biology, which was very promising in theory, might actually not produce a high yield. That maybe in fact natural selection was quiescent and in fact the reason we're seeing so few changes is that actually there's not been a lot of adaptive directional selection. So, that was the situation we found ourselves in until just a few years ago when we carried out this study in our research group led by Ali Akbari.
So, what we did is we deployed a few innovations or changes to try to improve our power to detect natural selection. One of them is we just pumped a lot of data into the system. And so, we increased the amount of data by about 14-fold. And the main thing that we do in this study is we report data in this study from about 10,000 individuals with new data. So, this is like a very big increase in the amount of data in the literature. And the total data set size of ancient individuals distributed over the last 18,000 years is about 16,000 people. So, this is a large data set. It's much larger than was previously possible. And when you have more data, you can estimate frequency changes with much more subtlety. And the data comes from only one part of the world, which is Europe and the Middle East. It's not a more important part of the world than other places, but it's the place where maybe 70 or 80% of the data in the ancient DNA literature so far comes from due to historical reasons. And it provides us with a natural laboratory where we can see what happens over one place over time as environments change the genome. It's really interesting to imagine doing this type of analysis in other parts of the world. And the comparative analyses are super important and interesting, but this study right now is about this one place in the world where we have particularly fantastic data. The other thing we did is we developed an entirely new methodology that hadn't been used in this area before. And the methodology is based on a technique that had been developed for finding risk factors for disease in medical studies. And a simple way to explain it is we ask how to predict the genetic type a person has based on its pattern of relatedness to other people. So, we'll have a data set of about 16,000 ancient people and 22,000 people if we include the ancient and modern people. And then we look at how closely related each of these 22,000 people are to each other. And we predict the genetic type at each position in the DNA at 10 million positions based on the pattern of relatedness to all of the other 22,000 people. And then we ask if natural selection blowing the frequency of the mutation in the same direction in all geographic places and at all times predicts the data a little bit better than just knowing the relatedness to all the other samples in the database. So, we're simply asking, the alternative hypothesis is that selection has been blowing in the same direction at all times and we simply ask if that explains the data better. And that's a dumb assumption because of course the truth is that natural selection is going to have changed in frequency over time. But, we're just asking the simplest of questions whether assuming a constant rate of selection explains the data more than not doing so.
And just to summarize to make sure I've understood, you're trying to make a model that predicts allele frequency changes over time. It has two different parts. One part is this genetic relatedness matrix which captures how similar different genomes are to each other and that should capture the impact of different bottlenecks and of drift and of population admixtures and all those things which affect the entire genome. And then you have the separate thing which is like, okay, if we look at specific locations, can we just say that oh, this location has been selected at whatever coefficient over time and if we add some coefficient, does it become easier to predict the allele frequency changes than you would have just seen from this other artifact which is just looking at like, oh, if you look at the whole genome, are these guys going through the same bottlenecks, have they been through the same drift, etc. Okay. So what have you learned?
So, when we analyze the data this way, we looked at 10 million positions in the DNA in these 22,000 people, 16,000 of them were ancient, and we looked to see if there was more change in this consistent direction over time than you would expect by chance. And when we analyze the data, we found many, many hundreds of places in the DNA that were changing too much over time in too consistent a way to be explained by chance. Now, there's a bit of a statistical problem in figuring out how many there are because they're so densely packed that they're close to each other and they're interfering with each other. But, when you try to piece them out and say, 'Let's count them only one in each place in the DNA and blank out the others,' we find at least about 479 positions that are all independently pushing in the same way. Those positions are 99% confident that they're real. By another criteria of more than 50% confident that they're real, we think that about 3,800 positions are all pushing in the same direction. So, this is like a crazy number of results given that in our work previously and other people's work, there were at most a couple of dozen discoveries coming from a single scan. So, when we got this result, we were very surprised. We thought it must be wrong, and we spent the next couple of years trying to make the results go away, but they just kept getting stronger. And so, what we were trying to do is to look for some kind of independent type of evidence to tell us whether these positions were real. And we stumbled on something really powerful for this purpose that had not been used in this way before, and it relied on the fact that we had very large numbers of discoveries, like many hundreds of discoveries or even thousands. And so, what we did is we took a completely independent data set, which was the corpus of genome-wide association studies. So, these are studies that people have carried out in hundreds of thousands of people looking for whether particular genetic mutations are more common in people with high blood pressure than with low blood pressure or something like this. So, we took the UK Biobank, which is about 500,000 people from Great Britain who have been measured for hundreds and hundreds of traits. The whole genomes of all these people have been sequenced and for each of these traits, we could look whether each of these 10 million positions are connected to this trait in some way in a convincing way. So, in 10 million positions about 15%, about 1.5 million positions in the DNA are predictive of at least one of these several hundred traits. So, then we could ask a question, is our natural selection signal or statistic, is it related to whether a mutation causes high blood pressure or some other trait. So, we slid our statistic for natural selection upward, you know, to a value of one, a value of two, a value of three, a value of four, a value of five. And as we did that, the enrichment for genetic mutations that affect traits got higher and higher. So, whereas it was only 15% when we didn't use our selection statistic, when we required the selection statistic to be above about five, there was about a fivefold enrichment for mutations that cause traits.
