(click) (inspirational piano music) - 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) (upbeat music) - Good afternoon. I think I have the longest
title talk of today. So that's something. (audience laughs) 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. In 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 it pretty vociferously. 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 on 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 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 Neanderthal and Denisovan 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 the 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 Denisovan 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 Neanderthals and Denisovan 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 Thousand 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 2500 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 light 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 Asians, 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 laughs) So we can do the same thing
for Denisovan sequences, again, 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, 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 in. 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 the sequences that we inherited from
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? I'm going to 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. In European individuals in blue and East Asian individuals in red. And 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 we stare at it for long times, is that there's a non-uniform distribution of surviving Neanderthal lineages. For example, this region, also highlighted by Sriram this morning, is about a 10 megabase region on chromosome seven that's
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 Denisovan 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's phenotypes, these deserts of archaic
sequence, I think, are a really, a 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 introgression. That is, Neanderthal
and Denisovan sequences were beneficial and rose to high frequency in the population. 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 we're likely influenced
by adaptive introgression tend to fall into a couple 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. We know that pathogens are
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 and Denisovans influenced the trajectory
of human evolution. But in the last few minutes that I have, I want to tell you about
some work that we published just a few weeks ago, actually, in 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 were, there was in East Asians,
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 Population or Project. Purple here represents
African admixed individual, 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 the 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 the major heart
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 something that we really
wasn't on our radar until we, 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 humans 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 admix 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's 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 in 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 not watching right now. Thank you very much. (chuckling) (clapping) (upbeat music)