(air whooshing) (bell rings) (soft electronic music) - [Narrator] We are the paradoxical ape. Bipedal, naked, large-brained, long the master of fire,
tools and language, but still trying to understand ourselves. Aware that death is inevitable, yet filled with optimism, we grow up slowly, we hand down knowledge, we empathize and deceive. We shape the future from
our shared understanding of the past. CARTA brings together experts
from diverse disciplines to exchange insights on who
we are and how we got here, an exploration made possible by the generosity of humans like you. (upbeat music) - Good afternoon. Very nice to be once again
in California, in San Diego at a CARTA meeting. In the course of the
last eight million years, the hominins, the tribe of
primers to which we belong, displayed a remarkable diversity. And what you see on this chart is that for almost any period of the past, there was always several
species of hominins existing on earth, sometime
in the same regions, which is kind of puzzling for us because today there is only
one species of hominins, I would say fortunately, and this species, Homo sapiens, displayed a remarkable
evolutionary success and then adaptive and
reproductive success. It's present everywhere. It expanded replacing, displacing
or partially absorbing all other forms of hominins, and this situation, having a single species
of hominin on earth it's in fact exceptional in many respects and prevail only for the
last, say 40,000 years, which is nothing at the
scale of geological times. And it's probably one
of the most challenging of the main challenges
that paleontology meets is to explain why and how this happen. Of course, we tend to think
that technological progress innovation played a major role, but it's likely that
also social complexity and the ability that we have
to create extended networks of individuals beyond local
groups is essential also in this process. We have many reasons to support the notion that our species originated in Africa. What genetics says is that the
genetic diversity in Africa is much larger than elsewhere on earth, and the humans that we
found outside of Africa seems to be a sort of sub
sample of this African diversity and of course this much as this notion of an out of Africa expansion of our species. We also have fossils and we have a group of fossils mostly just after 200,000 years
that has been found in sub-Saharan Africa and
more precisely in East Africa that are rather close to us anatomically. They are rather close but they are not really exactly like us and we used to call
them early Homo sapiens or early modern humans. I found this term modern
humans a bit problematic because when we say modern
humans for this hominins we mean cladistically modern. It means they belong to our lineage because they were not
really anatomically modern nor behaviorally modern. And that's a bit confusing,
I think, for the future. And so with these discoveries
emerged the notion that there was in the past
somewhere maybe in East Africa a sort of restricted Garden of Eden where in a sort of biblical ways suddenly around 200,000 years ago, a fully human creature like us emerged and then later expanded. And this view that
prevailed for a long time has been challenged by many
discoveries these past two years And I want to take you
very far from East Africa in this landscape where
you see in the back, the Atlas Mountains in Morocco. It's a place called Jebel Irhoud. I think it's difficult to be
further away from East Africa when you are still in Africa. It's very far from Kenya. And we are somewhere between
the town of Marrakesh and the Atlantic coast. And this place of Jebel Irhoud is known by paleoanthropologist and has been known for a long time because in the early '60s, the
exploitation of barite mine revealed an archeological
site, many bones, many lytics, and the workers of the
mine found this skull and brought it to the
medical doctor of the mine. This medical doctor sent
me a letter in the '90s explaining the discovery. The site was passed on to
a professor of paleontology at the university of Rabat. And this fossil I would say did
not really match the picture we had of evolution of hominins in Africa, and it was long considered a
sort of African Neanderthal and it was seen this
way because in the '60s, well, everything was a big brain and a big brow ridge was a Neanderthal and so people would find
Neanderthal anywhere. It's only later that Neanderthal will identify as a Western or Asian clade. I've been always interested
in the Jebel Irhoud site and when I moved to the
Max Planck Society in 2004, immediately I contacted
my colleagues in Morocco to resume works in the site. And initially, our intention was mostly to date the site because the exact age of this hominin has been a mystery for us for many years. And there were several attempts
but rather unsuccessful to have an exact date. So we spent, I would say
years cleaning the site and exploring it and we were lucky to find in a corner of this sort of query, because the site has been
mostly destroyed by the mine and some other excavations. We found an institute with
sequence of about three meter of archeological deposits. And in this sequence, the bottom of the sequence is very rich. There is a lot of artifacts, a lot of remains of fauna, and also a lot of traces of fire activity. And this was a great luck because this fire burned flints and so we could collect a
lot of these burnt flints and burnt flints are essential to implement a method that
we call thermoluminescence to date sites. And so with this burnt
flints, we are able to provide a much more accurate dating
of the site than ever before. Well, of course when
the paleoanthropologist start digging somewhere, he hopes not to find only a burnt flint. So we're very lucky to find
also new hominins in the site. In fact, in a couple of years, we tripled the number of fossils
coming from Jebel Irhoud. We found many post-colonial remains, but also we found a partial skull, fragmentary phase, but rather complete. Also an adult mandible
that's fairly well preserved. I must say it's for this part
of the late middle places seen is probably the best
preserved adult mandible that we have in Africa. This mandible is not like my mandible, it's not like your mandible. It does not have a very
prominent chin, for example, but it has many features in its proportion and also in the dental features which have shared evolved
conditions with us, and this is one of the arguments that led us to assign these
fossils from Jebel Irhoud Not to the Neanderthals as it
was believed for a long time but to an early form of Homo sapiens. Regarding the age of the site. With this method of thermoluminescence and using another method called ESR, we're able to provide a series
of dates, very consistent assigning these deposits
where we have hominins to a period of time
around 200,000 years ago, which made these fossils from Jebel Irhoud the oldest Homo sapiens that
we know today in Africa. They are much older than
this East African forms that I showed you, and they are not in the Garden of Eden. They are very far from the Garden of Eden unless you consider all
Africa is the Garden of Eden. From a morphological point of view, one of the most striking aspects of the fossils of Jebel
Irhoud is the face. And this face is very similar to our face. What you have here on the left
is a chart showing you points which are located on this
two dimensional space in relation to their
morphological proximity. It's something we call
geometric morphometrics. So the distance between points express the difference in shape. And what you see here
is a big blue polygon representing the variation
of present-day humans. The pink stars that you
see are either fossils from Jebel Irhoud or different possibilities
of reconstructions of these fragmentary phase
that I showed you, Irhoud 10. And you'll see that they
all fit into this polygon. So it means you could meet in the street the face of Jebel Irhoud today, okay. In the meantime, you see
also that what you have here, Neanderthals and also
middle places in hominins that were long considered to
be ancestral to Homo sapiens are very different. So, about face, what's interesting is that Homo sapiens retained
a number of primitive features that we have in erectus, and these forms, they are very derived compared to these primitive form. This being said, Irhoud and us would not have just an erectus face. There are also some slight difference. We'll discuss that maybe
tomorrow in our workshop. Regarding the brain and the brain case, it's a completely different story because the brain case of
of the Jebel Irhoud hominins is a rather long and broad brain case and so these individuals, they don't have the sort
of rounded globular brain that you find extra in humans. Again, on the geometric
morphometrics analysis, you have extend humans in blue, very different from Neanderthals in red, erectus are at the bottom in green, they showed a primitive condition. And again, you'll see
that Neanderthals evolve in a very different direction
than the Homo sapiens and our fossil hominins from
Africa, those from Irhoud and the other ones from
East Africa that I mentioned are here along an arrow
that's evolving gradually toward the modern condition. Again, middle place was in
forms that were considered to be ancestral to sapiens possible. They grew up with Neanderthals
in very different, very, we'd say, derived situation. So we can zoom in a little
bit into this evolution of the brain in the last
three, 400,000 years. And what you have here is
a more detailed analysis of different groups of Homo
sapiens of different ages. They are represented by blue polygons with numbers in relation to their age. Along the horizontal axis, what is represented is mostly size. And so what you see is
the brain size increase between erectus, Neanderthals and sapiens, but also what this charts shows is that when Homo sapiens reach a brain size which is about the brain size
of the present-day humans, 1400 cc, something like that, morphologically, it started diverging in a completely different direction, almost perpendicular to this trend. And along this axis, what you have is this increase
of globurality of the brain. So you have the parietal
area becoming more salient and importantly, the cerebellum, this part of the endocranium that is at the bottom of
the brain in the back, that becomes bigger and bigger. This is one of the main features
evolving in all species. Cerebellum has been long considered to be mostly involved in
in motor coordination, but we know now that
cerebellum is involved in much more complex tasks. It's connected to many
parts of the neocortex. It's involved in the reward
circuits of the brain. It's involved in language, it's involved in social interaction. And this is one of the
features that we see developing in our species. So, all these discoveries
raise a number of questions and the main one is the
origin of our species because what paleogenetics
tells us is that the split point between the lineage
leading to Neanderthals and leading to Homo sapiens is somewhere around 650,000 years ago. So there is quite a stretch of time between 300,000 and 650,000, and we don't have, I
would say much evidence for this time period. For a long time, these
middle places in hominins that we call Homo heidelbergensis
was considered to be a possible common ancestor of
Neanderthals and homo sapiens. And today, and because of
what I just showed you, the fact that it's derived already on the side of Neanderthals, we see it more like a common
ancestor of the Denisovans and the Neanderthals. So it might be, I would say
essentially in the Asian group. But the problem is that
we also have it in Africa, and this is kind of puzzling. We have at least two fossils that we can assign to
this group in Africa. And we have indication
that it could have survived much later than we used to think. So that's, you know, again, it's something we have to consider. An African continent
with different groups, the ancestors of Homo sapiens
that we still have to identify This sort of Neanderthal-like creatures that might have survived
for a while, Homo naledi. So in other words, a much
more diverse situation that we imagine in a continent with some kind of mechanism of isolation for all these groups. The Jebel Irhoud specimen,
they are associated to what we call Middle Stone Age assemblage. A lot of flakes and points, not any more end axes. And it's one of the earliest
form of this Middle Stone Age. We actually find that soon after 300,000 in South Africa and in East Africa, we find early forms of Middle Stone Age, and it's very tempting
to relate the expansion of our species at the
scale of the continent with the development of this
new kind of technologies. Of course we would like
to have more fossils but we don't have that many. If we look at the fossil
record in Africa after 300, we have mostly two fossils
that could be something like the Irhoud specimens but
they are very poorly dated. And what's really interesting
with the Middle Stone Age is that among this Middle
Stone Age assemblages that we see gradually emerging
more and more complexity, cultural complexity with the
development of bone industries. A lot of points. Also the development of
non utilitarian objects like these beads that has been found in
different parts of Africa and also a lot of
diversification on the continent. This development did not occur
just in sub-Saharan Africa. Just in one site in Morocco, we have more than 300 of these beads which are around 100,000 years old. So the new picture emerging about the origin of our species in Africa is first of all, a much more
mature older coalescence point for all the lineages
belonged to our species probably beyond 300,000. possibly structured off of this population in
Africa for a long time. This is something difficult to test based on the fossil evidence because we don't have much fossils, and it's quite possible also
that there's different branches that would have existed in
different parts of Africa, could have been reshuffled
by environmental changes. As a matter of fact, Africa witness a lot of climatic changes during this time period and in particular, we
had occasionally the, the monsoon coming from the Gulf of Guinea going much further North than today. And so we have a series of what we call Green Sahara Episodes. By the way, one of them occurring just before the time of Jebel Irhoud. And a large region today covered by desert and it's very large, was
peopled by hunter gatherers living in the Savannah with rivers, with lakes, some of
them as big as Germany. And so we have to keep this in mind when we speak about the evolution
of our species in Africa. In other words, looking at the location of present-day populations for example, to reconstruct the past
might be very misleading because it's likely that the
descendants of all these people who left all these points
all over the Sahara, maybe today in West
Africa or in East Africa. So we think that these episodes might be episode during
which advantages mutations or innovation would have passed
from one group to another. And this may explain why at the
end of the Middle Stone Age, we have these kind of
beads that I showed you, which are made from a very
special type of shells, can be found from South Africa to Morocco, and from Morocco to The Levant. I thank you very much and I want to express my acknowledgements to all my colleagues who
participated in the project, and to the Max Planck Society
for its very generous funding, and the INSAP in Morocco for
letting us work in Jebel Irhoud and supporting us. Thank you. (audience applauding) - Good afternoon. I think I have the longest
title talk of today. So that's something. Thank you very much to the organizers for inviting me to come and speak. Hope you feel better soon, Sarah. So we've heard throughout the afternoon that modern humans
overlapped in time and space with multiple hominin lineages. And one topic that has been
of enduring interest is, was there any admixture that happened between modern humans and
these other groups of humans? And for a long time the answer was, well, maybe yes and maybe no. And people debated that
pretty vociferous way. And that was largely because the data to answer that question didn't exist. And it wasn't until more recently
that Svante Paabo's group in Leipzig developed and pioneered tools for studying ancient DNA, and they produced the first
Neanderthal genome sequence and we finally had the tools
to be able to say definitively whether admixture occurred or not. And indeed as Sriram talked about earlier, all non-Africans derive about 2% of their ancestry from Neanderthals. And what was even more
interesting is a few years later, this same group published another paper where they sequenced ancient DNA isolated from a small fragment of a pinky bone thinking that it was perhaps Neanderthal or maybe modern human. And it turned out to be this
entire new branch of humanity that we now call the Denisovans. And this is in fact the first species to be entirely described by DNA. And so ancient DNA has
transformed our understanding of human history over the past decade. And we've learned many things
like what the distribution of Neanderthal ancestry is in
populations across the world. And again, we saw this morning, but we can see that on average, individuals outside of Africa can trace about 2% of their genomes
back to Neanderthal ancestors. And strikingly, we find
a very different picture for Denisovan ancestry. So here, we really only
find Denisovan ancestry in parts of the world down here, oops, that's not working. So I'll just talk. In populations of Melanesian and Australian Aboriginal origin. So we have a very different
geographic pattern of surviving ancestry. And that's great. We can describe global
ancestry proportions, but studying ancient DNA is still hard. So my interest in this area can be traced back to a few
years ago where we had this idea that, well, if modern
individuals interbred with Neanderthals and Denisovans, then maybe we don't have
to excavate ancient DNA directly from fossils, but
we could indirectly isolate Neanderthals and Denisovans sequences from the genomes of modern humans. And so we call this molecular excavations. And I borrowed this slide
from a CARTA meeting a few years ago, which I think is a really
beautiful representation of this idea of molecular excavations. And so literally what we're trying to do is develop computational
or statistical models and walk along somebody's genome and pull out the bits that were inherited from Neanderthals or Denisovans. And molecular excavations
are really powerful because they enable us to identify the specific DNA sequences that were inherited from
Neanderthal or Denisovan ancestors. So it's one thing to say
something about a proportion, but when you can actually
identify the sequences, you can do a lot of
interesting things with it. So you can test evolutionary hypotheses and you can even start thinking about, well, what's the influence of Neanderthal and Denisovans sequences
on traits and diseases in present-day populations. And I'll try to touch on
all of this in my talk. So we've discovered methods to identify Neanderthal
and Denisovans sequences and we've applied them to
geographically diverse populations We've largely looked
at around 2,500 genomes that are part of a
publicly available project called the 1000 Genomes Project. But we've also worked with
colleagues in some cases to sample populations from
particular regions of the world. For instance, Melanesia where
we expect Denisovan ancestry to be the highest. And so how much of the
Neanderthal and Denisovan genome persists in modern individuals? So if we just represent
the Neanderthal genome as this circle, when we look across all 2,500 people, we actually recover about 41%
of the Neanderthal genome. And that's pretty striking, right? That we're not actually
sequencing a Neanderthal, but we're stringing together
these bits and pieces that survive in modern individuals. And by doing that, we can find almost half
of the Neanderthal genome. And that might seem
surprising, especially in that each of us only carries a little
bit of Neanderthal ancestry But the reason this works is that the 2% of Neanderthal sequence that I have might be a little bit different
than the 2% that you have. And when we look collectively across large numbers of individuals, we can recover a substantial amount of the Neanderthal genome. And on an individual basis, non-African individuals have
about 55 million base pairs of Neanderthal sequence per individual, and this is pretty similar
across populations. So East Asian, South Asians, Europeans and American individuals. There's a little variation,
but it's fairly consistent. And incidentally, if you get your 23andMe
report and they tell you you either have the most
Neanderthal ancestry or the least Neanderthal
ancestry, what it's really saying is that if you have the least amount, you have about 40 megabases of sequence, and if you have the most, you
have 60 megabases of sequence, and whether that's interesting or not, that's entirely up to you. (audience laughing) So we can do the same thing
for Denisovan sequences. Again, we represent the
Denisovan genome as this circle. And here, we don't do quite as well. So, but we still recover
10% of the genome, which is a substantial amount. And the reason we don't
recover quite as much is that Denisovan ancestry
is largely confined to Melanesian populations. So in fact, the Melanesians
have about 40 megabases of Denisovan sequence per individual. And you find very little
Denisovan sequence in other populations. And in fact, this 10% number
is actually pretty good because we only have a
sample size of 35 individuals compared to the 2,500 individuals that we're looking for
Neanderthal sequence. So in fact, there's a lot
more of the Denisovan genome to be found. So that's interesting. We can identify introgress sequence. But really what we're interested in is understanding whether admixture was just an interesting
side note to human history or was it something more significant? And in particular, did these
sequences that we inherited from the Neanderthals and Denisovans, did they have negative
fitness consequences? That means, were some of
these sequences deleterious? Were some of the sequences advantageous and confer an advantage to our ancestors? And then ultimately, we'd like to know what are the phenotypic
consequences of hybridization? And we're gonna focus mainly
on these two issues today. So this is an overwhelming slide showing the distribution
of Neanderthal sequence that we can find in modern individuals, and European individuals in blue and East Asian individuals in red. In each place we find
Neanderthal sequence, in one of these populations, we put a tick mark on the chromosome. The gray regions are
just parts of the genome that are too structurally
complex to analyze, so we just ignore them, and the black circles are centromeres. And one thing that you
might be able to see if you stare at this long enough, and let me stare at it for a long time, is that there's a nonuniform distribution of surviving Neanderthal lineages. For example, this region, also highlighted by Sriram this morning, is about a 10 megabase
region on chromosome 7 that significantly depleted
of Neanderthal sequence. It's also significantly
depleted of Denisovan sequence. And what this suggests is that there once probably was
Neanderthal and Denisovans sequence in this region, but it was deleterious in modern humans and eliminated by natural selection. And as Sriram pointed out, right in the middle of this
region is the gene FOXP2 that's been implicated
in speech and language. So if we're interested
in the genetic substrates of uniquely modern human phenotypes, these deserts of archaic sequence I think are a really good starting point. But not all sequences that we inherited from Neanderthals or
Denisovans were deleterious, some in fact were advantageous. And we know that there's somewhere on the order of 50 to
100 places in the genome where there's examples
of adaptive integration, that is Neanderthal and Denisovan
sequences were beneficial and rose to high frequency
in the population. And we can find examples of this in all of the populations that we look at. And this is pretty fascinating because as modern humans are dispersing
into these new environments, they're admixing and picking
up beneficial copies of genes from species or group of
populations that have been there for hundreds of thousands
of years before them. And so this is a pretty
efficient way to adapt to new environmental conditions. And you can sort of generally say that the phenotypes that
were likely influenced by adaptive integration tend to fall into a couple of categories. So things that influence
our ability to adapt to new environments like
high altitude, for example, vast majority of adaptive
introgression genes are involved in pathogen defense. And we know that pathogens are one of the strongest
selective pressures in humans. And then there's a set of genes that we don't really fully understand that are involved in skin and hair biology and they too show a very strong signature of adaptive introgression. So we'd like to continue to understand how hybridizing or
mating with Neanderthals, with Neanderthals and Denisovans influence the trajectory
of human evolution. But in the last few minutes that I have, I wanna tell you about some work that we published just a
few weeks ago, actually, and which we developed a new method that reveals a new twist
in our understanding of human history and
mixing with Neanderthals. And one thing that you might
have noticed earlier in my talk is that when I talked about
patterns of Neanderthal ancestry I exclusively focused on
non-African populations. So I showed you how much
Neanderthal sequence there was in East Asian, South Asians, Europeans and American populations, but didn't say anything about individuals of African ancestry. And that's because all of the
methods up until this point have assumed that Neanderthal
ancestry in Africa was either very little or non-existent. And so we recently developed a new method that didn't make this assumption. And so we were excited to apply it to individuals of African ancestry. And to our surprise, we actually found substantial amounts of Neanderthal sequence
in African individuals. And these were the five populations that were available for analysis from the 1000 Genomes Project. Purple here represents
African admixed individuals, so largely African-Americans. But even in these African populations from the 1000 Genomes Project,
we find about 17 megabases of Neanderthal sequence per individual. And just as a comparison, when we look at sort
of the same individuals and call Neanderthal sequence
using previous methods that we developed, that make this assumption that there's little
Neanderthal ancestry in Africa, we only call maybe 500 kilobases. So like two orders of magnitude less. So this was a really strong signal and it was very surprising. So we do see Neanderthal
ancestry using this new method. But what explains this signal? Well, to make a long story short, there's really two primary explanations. So the first is that there
were migrations back to Africa. So people left Africa in the major all of Africa dispersal, hybridized or admixed with Neanderthals, and some returned back to Africa carrying the Neanderthal
sequence with them. And our results show that
the amount of back migration has probably been much larger
than we've previously thought. So that's one part of the signal. The second part actually
is really fascinating and it's something that we
really wasn't on our radar until we got this result. And that is that part of the
signal of Neanderthal ancestry in Africa is due to an early
out of Africa dispersal and gene flow from
humans into Neanderthals. And so let me unpack that
a little bit for you. So this is a simple phylogeny showing the relationship
between Neanderthals and three modern human populations. So Africans, Europeans and East Asians. And so the bottom here
represents the present. And we go into the past
as we go towards the top. And this hatch mark just is to indicate that the times aren't going
to be drawn proportionally. So we know that Neanderthals
and modern human split around 600,000 years ago, and what our data shows is that not only was there this out of Africa dispersal that happened 80,000 years ago that resulted in the
peopling of the world, but there was also a much
earlier dispersal of humans out of Africa around 200,000 years ago, and they encountered Neanderthals
and admixed with them. So in fact, some of the sequence that we call as Neanderthal, it's not Neanderthal
sequence in modern humans, it's that Neanderthals
have modern human sequence. And so this adds a further twist to sort of this complex pattern
of admixture and gene flow and arrows pointing in every direction. So in conclusion, there
is substantial amounts of the Neanderthal and Denisovan genome that remain in modern individuals. There were fitness
consequences to hybridization, both good and bad. Humans Neanderthals and Denisovans
have mixed multiple times likely and multiple places. And that there were multiple dispersals both in and out of Africa. And I think this last point is something that is really important in genetics, is that we often have a simple models of how humans dispersed around the world. And that the more data we look at, the more complex these models become and that it's important
to take into account the dispersals both out of
Africa and back into Africa to really understand patterns
of Neanderthal ancestry. So I would like to acknowledge
my lab in particular, Ben Vernot who did a lot of the early work on finding Neanderthal sequence, my collaborators and my two boys who I'm sure are watching right now. Thank you very much. (laughing) (audience applauding) - Well, I can tell you
about using ancient DNA to try and understand
evolution and natural selection actually among humans today. So we've heard a lot
about using ancient DNA to understand archaic humans,
understand the relationship between different archaic populations and the interactions between archaic and
modern human populations. But most of the work in ancient DNA is actually focused on
more recent populations, particularly populations movements in the last 10 to 15,000 years. And that's partly because
of sample availability and preservation of DNA, but it's partly because
that's a time period when there are large number
of population movements, if you will, migrations. And I'm just showing you here some of the recent population movements that have been investigated
in depth using ancient DNA. Now what I'm interested
in is trying to understand not just how and when these
population movements happened, but as people moved into new environments, how they adapted, in particular
how they adapted genetically to different environments. Now you see there are a lot
of arrows here in Europe and there's been a lot of focus on Europe and I'm going to focus largely
on Europe in Western Eurasia. That's not because it's necessarily
the most important place or the most interesting
place in the world, but it's because it's really
the only place as of today that we have the large
samples that we need to study the kind of questions
we're going to look at. So what does the genetic
structure of Europe look like today? Well, this is a rather famous figure, this shows you what genetic
variation looks like in Europe. So here in the larger plot is a principal components
analysis of genetic data, you can think of this as a genetic map, people are close together on the map if they are more genetically similar, and further away, if they're
genetically more distant. Now each individual here is colored according to their country of origin. And if you look at the inset map, you'll see the geographic structure. And basically the point here is if you look at the genetic map and turn your head and
squint a little bit, it looks like the geographic map, okay. So the genetic structure of Europe recapitulates the
geographic structure, okay. So by looking at genetic data, we can learn about geographic structure. And so if I gave you
an individual's genome, you would be able to
tell with some accuracy which part of Europe they were from, or at least which parts of Europe their recent ancestors were from. So this is what ancestry
testing companies do in order to tell you
what your ancestry is. Now, the idea is that by looking
at this kind of genetic map we can learn not only about
a geographic structure, but also about historical structure. And this is an idea that really goes back to the work of Cavalli-Sforza in the '60s and perhaps even earlier than that. But there's a problem with trying to learn about historical demography
from this kind of data. And that's that present-day data really gives us a very
small window into the past. So if I show you the same kind of plot, including ancient Europeans, you get a very different picture. So now present-day Europeans
in this plot are in gray and ancient Europeans
dating back 45,000 years. So this is essentially the entire time that anatomically modern
humans have been in Europe. And the first thing you notice are that the ancient Europeans are well outside the range of
present-day genetic variation. And it's not until the Bronze Age and even really the Late Bronze Age that you see ancient Europeans
who look genetically similar to present-day people. So you can immediately see
it's going to be very difficult to make inference about
these past populations if we're just trying to sort
of look through the tiny window that we get from present-day data. Now, the other thing
we get from ancient DNA that's very powerful is a
very clear sense of time. So here I'm showing you a time
series of ancient Europeans over the last 15,000 years. So x-axis's time, each column
is an ancient individual and they're colored according
to how much ancestry they have from each of three inferred
source populations. And you can immediately see, just visually looking at this plot, the genetic impact of events like the introduction of agriculture and the transition from this
red Hunter gatherer ancestry to ancestry that's associated with early farming populations in blue. You can see those two
groups continue to mix over the next few thousand years. And then around 4,500 years ago, we see the appearance of this
green ancestry component, which is associated with
population movements from the Eurasian Steppe, okay. So in some sense, this is why people think
of ancient DNA as cheating because you can see these
population movements directly without having to infer them using complicated statistical
methods from present-day data. Now what I'm showing you here
is a genome-wide average. Okay, so this is looking at if
we every average the ancestor of each individual over
their whole genome, we see these kinds of systematic changes. And what we want to know is, are there parts of the genome that don't follow this pattern? Because if there are
parts of the genome which either change more rapidly or
in some cases less rapidly, those are the parts of the genome which we think are likely
to have been targeted by natural selection. So the evolution has
been driven by selection rather than, in this
case, drift and admixture. So if we do that, we get a map like this. So what I'm showing you here is a scan for selection across the
genome using ancient DNA. So on the x-axis is chromosomal position. Each point is a particular genetic marker. And the y-axis is showing you
the evidence that that marker is not following this
genome-wide average pattern that I showed on the previous slide. So this is called a Manhattan plots. And when you see a skyscraper with a lot of variants lined up, that means that that part of the genome has a large number of markers which are not behaving neutrally and are likely to have
been targeted by selection. Now often when you do this kind
of scan, all of these hits, all of these signals are in
obscure regions of the genome and you can't figure out what they do and it's very hard to interpret. But fortunately when we did this analysis, it turned out that almost
all of these lay in genes that were well annotated and which we can make a pretty
good guess of the function. Okay, so that gives us some confidence that these are real signals of selection. So I'm showing you here the gene names. If we ask what do these genes do? It turns out that they fall
into three main categories. So there are genes associated with diets, genes associated with pigmentation, particularly skin pigmentation, and genes associated
with the immune system. So this kind of makes sense. This is the time which
agriculture is introduced. People's diets change dramatically, that people are moving
from lower latitudes into Northern Europe. We think a selection for
lighter skin pigmentation is related to vitamin D availability. And of course people are
moving into new environments but also into denser populations and living in close proximity
to domesticated animals. We expect them to experience
new pathogen challenges and so see selection on the immune system. So again, this set of genes is
in some sense what we expect and gives us some confidence that we are finding real signals. Now as I mentioned, one of the things we
can do with ancient DNA that's very powerful is to get a very clear sense of the timing of selection on some of these. So if we look at the
lactase persistence locus, which is the strongest signal of selection in the entire human genome, we can see exactly when it
starts to become common. So many of you are familiar with this. Most adults in the world
today do not have the ability to digest lactose as adults. There are a number of mutations in different parts of the world which allow this ability
to persist into adulthood. And what I'm showing you
here is the real frequency of the mutation, which today
is relatively common in Europe, in different parts of Europe. And what you can see is
that 5,000 years ago, no one or almost no one in
Europe had this mutation, so no one had lactase persistence. And that's interesting because that's many thousands of years after the domestication of cattle, goats, and actually after archeological
evidence of dairying and use of milk products. You see that in Britain and
Ireland and in central Europe, also in Scandinavia, not shown here, the mutation starts to become
in three to 4,000 years ago, increases rapidly and frequency
before actually leveling off maybe around 1,000 years ago. On the other hand, in
other parts of West Eurasia where this mutation is common, in Iberia, and in the Indus Valley, it is not selected until much later, perhaps one to 2,000 years ago. It increases rapidly and frequency and doesn't apparently
seem to have leveled off. This data stops maybe a
few hundred years ago. Now, the idea is by looking at these kinds of very detailed temporal data and correlating with
archeological and cultural data, we can start to try and understand why this particular mutation
would be so strongly selected. As I said, it was the strongest
signal of natural selection in the entire human genome. We still don't really
understand the exact mechanism of selection pressure. Other interesting loci
are also related to diet. So one of my favorite genes is FADS1. There are a number of
genes that the FADS locus is associated with fatty acid metabolism, and we think that the derived
allele here is an adaptation to a diet rich in plant fats
as opposed to animal fats. And you can see here that
Mesolithic hunter-gatherers do not carry this allele. It appears in Europe
and with early farmers. And then it's rapidly selected
actually in the Bronze Age. We think likely related to the advent of intensive cereal agriculture. One of the interesting
things about this allele is that today in medical genetic studies, this is one of the strongest
associations in the genome with LDL cholesterol and triglycerides. So this is an example where
our evolutionary history, in fact, our recent evolutionary history has a big effect on our health today. Now, some of the things that
are actually very interesting are not the things that
are under selection, but the things that are
not under selection. And one example of that
is salivary amylase. So the salivary amylase gene allows you to start breaking down starch in saliva. It seems like the sort of thing that would be an adaptation
to an agricultural diet based around starchy products. Today, most humans have
many copies of this gene. In the top row here you can
see the distribution of copies. In present-day Europeans, you see most people have on
the order of six to 12 copies. Now, looking at the ancient
populations going down the list, you see that actually most
of the ancient populations we look at have copy numbers, which are not significantly different from present-day Europeans. In particular, Mesolithic hunter-gatherers do not have a less copy number
than present-day Europeans. So it seems like in this case, what seems like it should be
an adaptation to agriculture actually predates the introduction
of agriculture per se. Finally, all of the variants
I've talked about so far have been relatively recent mutations. But actually, some of these mutations turn out to be very old. And in particular, some of
the mutations associated with the immune system turn out to have introgressed from archaic humans. So here I'm showing you
two figures from two papers which demonstrate that
the selected alleles at two of these immune loci are actually clustering
with the Neanderthal loci, and in fact, appear to be introgressed. What's interesting here
is that we're talking about selection in the past 10,000 years, Neanderthal integration
happened 50,000 years ago. So these are not variants that
were immediately adaptive, they're are variants that have contributed to immune diversity and
become selected much later. So ancient DNA allows us to
identify individual variants on a very strong selection. And one of the interesting
things we discover is that actually genetic
adaptation to agriculture doesn't seem to coincide very closely with the introduction
of agriculture itself. Now, all the things
I've talked about so far have been relatively simple traits. If you have the lactase
persistence mutation, you have lactase persistence. If not, you don't. But actually most human traits, certainly most morphological
traits and most disease traits are not like that, okay. So the canonical example here is height. So I'm showing you here
another Manhattan plots. This time I'm not looking
for evidence of selection, I'm looking for evidence of
association with heights. And it turns out that
there are many thousands, in fact probably tens of
thousands of genetic variants across the genome that are all significantly
associated with heights. They all contribute to
variation in height. Almost every complex trait
that's been studied in any detail has this kind of architecture. Of course, the effect of each
of these variants are tiny, so the average effect size is on the order of plus
or minus one millimeter. But if we sum up the effect
of all of these variants, we can actually start to build relatively accurate predictions of height. For example, we can explain on the order of 20% of the phenotypic variance. Now, the other reason
we like studying heights is because it's one of the few phenotypes we can actually measure
in skeletons, okay. So if we do that, we see very interesting
changes throughout Europe. So this is data collected by
our collaborator, Chris Ruff. So the first thing you notice is that actually in the
early upper paleolithic, these early humans in
Europe are relatively tall, actually almost within the
range of present-day variation. Okay. There's a dramatic decrease
in height over the LGM, which lasts really until the Neolithic before an increase, again,
going into the Bronze Age. These data, and about 1,000 years ago, of course in the 20th century, there's a dramatic increase in heights, but we think that's largely
not genetic in origin. So if we take our ancient samples and what we know about the
genetic basis of height today, and we try and predict, make genetic predictions of
height of these ancient samples, it turns out we can largely
recover this pattern. So what I'm showing here is
genetic predictions of heights based on something like
1,200 ancient Europeans. And you can see that we do predict that early upper paleolithic
populations are taller. The decrease continue
as into the Neolithic and then an increase
going into the Bronze Age. If I show you the statue
data on the same model, you can see at least they're qualitatively
consistent technical reasons. We don't necessarily expect
them to be exactly the same. Another interesting observation is that if we do this
analysis for standing heights, sorry for sitting heights, it turns out we see very little change. Okay, so sitting height and
the skeletons doesn't change to genetic prediction. It's a very, we predict
a significant change, but it's actually very,
very small in magnitude. So what this is telling us
is that the known observation that this change in height is driven by change in limb length
actually has a genetic basis. And we think this is likely
an adaptation to climate following Allen's Rule, reduced limb length reduces heat loss in colder environments. So ancient DNA allows us to track not only the evolution of simple traits, but actually of polygenic
complex traits as well. And potentially, we can use this ability to actually predict the
evolution of traits, which we can't measure
directly from skeletons. Now, in some sense, we can
use this information to tell when phenotypes are changing
because of evolution rather than plastic
developmental responses. And to separate out those two effects, I'll end with a caveat
though, which is that it feels like many of these
changes should be adaptive, but actually it turns
out to be very difficult to provide genetic evidence for that. Okay, and I think particularly
even in the case of heights, which is probably the
best studied example, we don't have solid evidence whether this is actually
driven by natural selection or whether this is genetic drift. And I think that's one
of the big questions that we're trying to challenge
over the next few years. So I'd like to thank my lab at Penn, collaborators elsewhere, our funding sources, and I'd like to thank you for listening. Thank you. (audience applauding) (upbeat music)
