so now I will turn things over to dr. Andrew berry thank you very much and very nice to see so many people here especially on such a frigid occasion David I'm sure will warm our hearts it is a real honor it really is an honor for me to introduce one of my scientific heroes David Reich David who as I tell my students does the best and coolest science being done today end of story stuff and you're about to discover that the stuff David does is stunning he started off as a physicist right here at Harvard these events are actually normally introduced by melissa franklin some of you will know Melissa he worked in Melissa's lab and she's unfortunately out of town she was rather disappointed I think she wanted to indulge in some form of ritual humiliation David however knew better I moved on to biology I went to Oxford is getting smarter and smarter to do a PhD in statistical genetics then waivered thought of going to mow it did go to medical school at Harvard MIT but then realized there were many more interesting things to do than curing disease and has been on the faculty at Harvard Medical School genetics department since 2003 since 2013 he has been a Howard Hughes Medical Institute investigator which sounds obscure but was is mind blowing lis prestigious he's also Tori of prestige won lots of awards despite his relative youthfulness including I noticed a 20-17 award it's true it was joint but this is a real award it's not just some medal or something this was a million dollars so he's serious three extraordinary achievements one is and they all pertain to the use of ancient DNA which you're about to learn about in some detail one is the deep past the notion that we wonderful exalted us Homo sapiens are not the product of some sort of vector devolution where we got better and better over time know what we are is the product a lot of sort of behind the bush sex with Neanderthals and other close relatives it's a murky twisted history we have to they're not so deep past and that's what David's going to be talking about tonight the lens and the precision that David and his group can bring to bear on human historical and pre historical questions using ancient DNA is extraordinary and finally I'm an evolutionary biologist David is changing the way we're doing our science normally we're interpolating we're trying to understand what happened in the past from the distribution of things in the present now we have a time machine ancient DNA helmed curiously I like this vision little goggles by David Reich which takes us back in the past and we actually see the processes that we've been reconstructing laid out before it's actually happening the the world of ancient DNA is changing science in extraordinary and wonderful ways and it's going to be a privilege to hear some of those ways from David right now thank you to people hear me so thank you for for that amazing introduction and so I'm gonna be talking to you about I'm gonna be trying to in this talk to really leave you with one primary idea if you have if you just take one idea away from it which is that these two sub continents of Eurasia of about equal size and historically about equal numbers of people although relative numbers have fluctuated a bit over time have a remarkably parallel history of of being a mixture of three major major ancestral populations some of which are non-indigenous and some of which have similar sources there's profound differences also but I wanted to highlight those parallel all isms and hopefully you'll come away the talk seeing these two subcontinent's in more similar in ways so I'm gonna begin by talking about what's happened since 2010 which is the coming online of a new scientific instrument so I'd argue that this is a powerful new scientific instrument similar to the way that microscopes were powerful when they first introduced that they made it possible to look into worlds that were never previously looked at and once the measure of a new scientific instrument and how powerful it is is that when you turn it to look at something that's never been looked at before like the world of microbes when people first started using a microscope in the 17th century you realize that there are things in the world you never previously imagined and what happens when we use ancient DNA again and again is that when we look at past peoples who have not yet been looked at before with regard to the question of how they're related to each other and to people living today the stories don't conform to the ones that we thought were most likely before again and again and again so in 2010 what happened was that it began possible for this first time to extract whole genome sequences from ancient humans who lived thousands or even tens of thousands and in some case even hundreds of thousands of years ago prior to that for several decades it had been possible to extract little snippets of DNA but the increased information that comes from sequencing the whole genome which is literally hundreds of thousands of times longer than those snippets that was that were being sequenced before makes it possible to ask and answer questions that are profoundly more different and more precise so typically the way ancient DNA analysis goes these days this we start with the skeletal remain usually I work on humans but people also work on animals and plants this is an ear bone Petrus bone so it's the part of the skull that contains the inner ear region and in particular the cochlea or the otic capsule which contains a key organ of hearing and this turns out as the people in this field have discovered to have about a hundred times more DNA per unit of bone powder than any other part of the skeleton that people have analyzed and so we preferentially study that part if we can obtain that skeletal element so in a clean room where the goal is to protect the sample from the people handling it who have much more DNA of course than the bone that's been in the ground for thousands of years we isolate the part of the bone that we're interested in either by drilling or by using another methodology we grind it into powder or pulverize it in some other way we release the DNA in a watery mix that removes the protein and mineral content and hopefully inhibiting molecules that will prevent the reactions from going forward where we want to use to extract the DNA and we convert the for DNA into a form that can be sequenced since about 2000 the price of DNA sequencing as probably many of you know has dropped by about a factor of a million fold and that combined with the more sensitive ways of extracting DNA and methods for isolating the parts of the DNA that we're interested in for the first time makes it possible to obtain large amounts of DNA from ancient humans there it's kind of a miracle that DNA preserves for it so long but it turns out to be a rather stable molecule and in the right conditions we can often get it out of samples that are thousands or tens of thousands of years old so as a result of this implementation of these series of technologies the amount of the number of human samples with genome-wide data has rapidly rapidly increased from the first samples that were published in 2010 to more than 2000 published samples today and it's a really hyper exponential growth with a more than 100 fold increase since 2013 and so with an increase in the amount of data this dramatic every year or two it becomes possible to ask and answer questions that really couldn't be addressed with a much smaller sample size in this talk I'm going to be telling you about the application of these large sample sizes to two parts of the world to Europe and Central Asia and the boundaries borders of South Asia so I talked about this in the book that I published last year and and this talk is really largely based on chapters 5 which is called the making of modern Europe chapters 6 which is the collision that formed India and maybe I think it's chapter 11 which is called the genomics of inequality but it's considerably updated relative to each of those chapters especially the South Asian chapter so I'm going to start in the first half of my talk about Europe which is the part of the world we now understand best from ancient DNA and that's not because it's more important than other parts of the world but it's because it was in the backyard of the people who developed much of the technology for ancient DNA so still the great majority of the more than 2,000 samples for which there's ancient DNA whole genome data are from Europe and we have really exquisite profiling of the changes in ancestry and types of ancestry over time I'm going to start there because the information we have is so rich and surprising and it will also set up the questions that we can are beginning to learn about also in South Asia so in this beginning of the talk I'm going to talk about two cultural processes that are documented not from DNA but from other lines of evidence so one of them shown on left is the evidence from archeology from the study of material remains left behind of a very dramatic phenomenon that occurred which is that after 9,000 years ago agriculture for the first time spread into Europe from its homeland in the Near East so farming probably develops twelve to eleven thousand years ago in the eastern part of present-day Turkey or northern Syria the that area and then within a couple of thousand years started spreading in many different directions one of them was into Europe probably from Anatolia in fact we know almost certainly from Western Anatolia present-day Turkey so a question was was that spread of Agriculture which transforms the archaeological record the types of pots and implements people leave behind and is very clearly documented there was an accomplished by movement of people or were people copying the technology as often happens when people copy for example use of cell phones or other technology even though it's not made by people of their own group that's been a big question in archaeology was it pots or people was it the spread of ideas or was it movement of people or some combination of the two another type of cultural evidence that's very relevant to this is the distribution of languages across Eurasia and here I show that the distribution of indo-european languages across Europe these are a closely related family of languages which it consists is responsible for almost all the languages of Europe as well as northern India Iran and Armenia but there are a few exceptions in Europe for example Basque in northwestern Spain Hungarian which is an island within Europe of a different type of family of language and fin and and languages and in Finland and Estonia and Lapland which are of different origin as well so a question is why do people speak such closely related languages Latin Romance languages are only a subset of these languages we speak one of these languages and what process spread these closely related languages which are highly unlikely to descend from a common ancestor much more than eight or ten thousand years ago so in 2014 in 2015 the first whole genome data from ancient Europeans was beginning to be available and it was very clear by that time that what had happened was that the spread of farming into Europe had been achieved by large-scale movement of people so there was data at that time from European hunter-gatherers the people were the exclusive inhabitants of Europe prior to 9,000 years ago and they had a very characteristic ancestry type shown in dark blue in these bar plots each of which is supposed to represent a Singh does represent a single individual we analyze data from and sari represented in green in these plots and after 9,000 years ago there was a large-scale movement into Europe first into Greece and into the Balkans near Greece of this type of ancestry dark blue which is very which is almost a perfect match to the ancestry we have data from from the from Anatolian farmers from Western Anatolia from about eighty five hundred years ago so it's very clear that there was large-scale movement into Europe of people very similar to Western Anatolians and it pinpoints the likely origin of the movement of people that brought this new technology it wasn't a complete replacement and in in the Balkans initially there was no most complete replacement but over the next two or three thousand years the local hunter-gatherer populations mixed a little bit with the farmers and they persisted actually at the fringes of Europe which were not farmable too far north or in mountainous regions and the overall proportion a farmer of hunter-gatherer ancestry settled at around 20 percent across Europe but varying in different places but if you look at Europeans today there's a third component of ancestry which is shown in red here and so that's the situation we were in in 2014 and 15 we knew that in some populations especially in northern Europe about half the ancestry in fact the single largest component of ancestry simply didn't exist in multiple sites more than 5,000 years ago so when did that third ancestry arrive it must have been sometime more than 5,000 or sometime after 5000 years ago so here is going to be a summary of what was found with genetic data in a series of papers beginning in 2015 and from our group and others so what I'm showing here is a principle component analysis plot and I'm going to walk you through this because I'm going to refer to this several times and also tell you about this approach a little bit so what you see here are dots each one corresponds to a present day or recently living person about a thousand dots drawn from the places shown on this map so that's Europe and that's Arabia just to orient you and what the data consists of is the following so we're looking at about 600,000 places in our DNA where people vary so you probably mostly know that our DNA sequences are very similar to each other so copies of copies of copies of copies of copies of copies of copies of copies of copies of copies of copies of copies of copies of your genome one from your mother and one from your father there are 99.9% similar at positions you can line them up so we're genetically extremely similar to each other but our whole genome sequence is about 3 billion DNA letters long and so one in a thousand of that is millions of differences and it's those differences that tell us about history because those differences accumulate over time due to random mutations that have occurred and they tell us how long it's been since any two genomes were separated because the more mutations separating genomes the longer it's been since time that since they share a common ancestor so the data consists of looking at about 600,000 positions where people are known to vary for example some people have an adenine one of the for DNA letters and some people have a cytosine another of the for DNA letters and so we look at these 600,000 physicians and you should think about it as a table with 600,000 rows and we look at the thousand individuals there so you should think about a thousand columns and in each cell if the table is a zero one or two depending on whether you have to add amines and adenine and cytosine that would be one or two cytosines that would be two and then you multiply this table by itself to get an approximately thousand by thousand table which shows how closely related every sample is to every other and you perform a mathematical dimension reduction standard technique on this to most efficiently separate the samples from each other and what shown here the axes are not shown is a plot of the most efficient separation of these samples from each other and the second most efficient the first versus the second principal component and without knowing the labels of where these individuals come there's a remarkable pattern that emerges so europeans almost exclusively splay out along a gradient a line here near Easterners people from here splay out a long line here with very few populations in between these groups in between are mostly Island Mediterranean groups which are have plausible more recent contact between Europe and the Near East as well as other groups with plausible more recent contact between Europe and the Near East in the bottom of this plot tends to be groups from the Mediterranean like Sardinians over here like Levantine populations from Jordan and Israel over here at the top are northern populations like Lithuanians and British over here and like people from the Caucasus over up to there and so this is a remarkable plot I'm now going to turn these dots gray and so I'm going to show you where the ancient samples over time fall relative to the modern samples because once you perform this principal component analysis on the ancient samples you can then see how the modern samples relate to the ancient variation so prior to 8,000 years ago the hunter-gatherers position on the plot here and this is very remarkable because they fall outside the variation of present-day West your agents they are fall beyond Europe in the direction of European differentiation from the Near East in the correct interpretation of that we can show is that Europeans today are a mixture of groups like this in groups like this that is Europeans today are a mixture of hunter-gatherers and people from the Near East who arrived with farming and also through other processes that I'll tell you about shortly the first farmers move dramatically on this plot over on top of present-day Sardinians and here are the Anatolian farmers and Sardinians today at least some Sardinians we now think of them as a relatively isolated population that descends from some of the first farmers who spread into Europe without much impact of later movements that affected the rest of Europe but you at this time 8000 years - 5,000 years ago just don't see people like the bulk of Europeans today meanwhile about 5000 years ago in the steppe north of the Black and Caspian Sea this group forms and I'll tell you more about them shortly and they fall on the extreme top of the European gradient but people like present-day Europeans are still not there 5000 years ago but after 5000 years ago they suddenly appear and the correct interpretation is they're a mix of people like this people like this and people like this forming a gradient over here and ever afterward Europeans look like this they fall in that area over there so what happens so a summary of what happens is that after 9,000 years ago there was a large-scale movement of people spreading into Europe and bringing with them farming technology and that this spread was accomplished to a large extent by movement of people although with important elements of exchange in terms of culture and other elements with the local hunter-gatherers especially at the periphery of Europe where people could not immediately farm so there was a long period of interaction when the hunter gatherers and farmers learned from each other at this after this time this created a continent full of people with different proportions of first farmer in hunter-gatherer ancestry but mostly first farmer ancestry and then there was a later impact from the steppes north of the Black and Caspian Seas and we can really show that this group that these yellow samples that formed at the top of that gradient about 5,000 years ago are the source of this movement and in fact are the single largest source of ancestry in many Europeans today so there's two large-scale movements into Europe one from the Near East via Anatolia and the other from the steppe north of the Black and Caspian Sea about five thousand years ago and this is sufficiently late and sufficiently large scale that it's very difficult to imagine that this movement from the steppe north of the in Caspian Seas was not a source of important languages that are spoken in Europe today and is due to linguistic arguments which already favor the step is the source of indo-european languages at least some of the indo-european languages in Europe this makes the step spread the most likely explanation for the source of at least some of the indo-european languages spoken in Europe today so who are these people who were those yellow dots falling on the top of the gradient well here's a skeleton from one of the individuals we analyzed from Far Eastern Europe from the Samara Oblast in present-day Russia and so this individual had a copper mace it's just before bronze began to be used and these are amazing archaeological culture so the yam nya were the first people who used pastoralism and spread it out into the open steppe lands prior to the Yama there were many scattered groups that lived in the river valleys and they had different cultures in different part of the steppe lands in this region but nobody or almost nobody was in the very dry areas between the rivers so these people took advantage of two very powerful inventions that had been invented just before that not necessarily by them but they quickly adopted them one was the invention of the wheel which had happened just before and the other was the domestication of a horse which possibly occurred by these people or certainly by some others in Central Asia and these people hitched their horses to wagons with wheels and used it to bring supplies out into the open steppe and they were able to exploit the grasslands and the resources of the steppe in a way they had not exploited them before and as a result or pus were somehow these people spread very dramatically all the way from Hungary in Europe in Central Europe all the way to the LT melting mountains on the boundary of Mongolia today and replaced many of these very disparate cultures before so it's a very dramatic spread in the archaeological record not just using more of the landscape that they occupied but also spreading geographically and replacing or displacing or living with other groups right so I'm now going to talk about how this ancestry not only got to Central Europe to Germany and to hungry but actually got all the way to the extremes of Atlantic Europe in a case study so I'm gonna first talk about Britain so this is AIT's in the past so Britain farming got to Britain rather late so it gets to Greece about eighty five hundred years ago but it gets to Britain only about 6,000 years ago prior to that there are hunter-gatherers like the rest of Europe and then around 6,000 years ago there's at least a 99% replacement of the local hunter-gatherer population by farmers coming from mainland Europe and it was a very dramatic event and here's a bunch of samples we reported data on at the beginning of 2018 from first farmers from Europe down to about 4,500 years ago and you see that none of them have ancestry from the Yamaha and then suddenly Stonehenge the big stones at Stonehenge go up just at the end of this period built by people with entirely first farmer ancestry then about 45 - 450 400 years ago the reason we have such precise date is because of the incredible power of radiocarbon dating you suddenly see ancestry like this so there's a dramatic shift with a 90 percent population replacement in Britain a second dramatic population replacement so it's a minimum of 90 percent population replacement from people from the continent who replace the people who built Stonehenge almost with very little local mixture and so what this is telling you is that there's a large scale movement not just once but twice in the history of this island and these people are genetically extremely similar to people in Britain today but these people are hardly related to people in Britain at all today so now I'm going to give you another example which we're trying to work into publish in a couple of weeks in Liberia so here's the pattern in Liberia it's the same time period shown so from 6,000 years ago all first farmers and then about the same time you see for the first time people bringing ancestry from the east from the Yamaha but here in iberia we detect a period of several hundred years of coexistence of these two groups and then a population mixture event where instead of a 90 percent population replacement it's only a 40 percent population replacement what that means is that if you take people living here and you ask who are their ancestors where were they living through ten generations ago a few hundred years ago that means 40% of those ancestors on the maternal and paternal lines would have been on not in iberia probably and 60 percent would have them at been Iberia but if you look at the coloring of these dots that's and it tells you something else so the coloring of the dots corresponds to the males and the individuals so the open circles are females but on the male's we can sequence their Y chromosome which men inherit from their fathers and we can determine whether it's characteristic of the first farmers of Europe or of people from the steppe and what you see in both Britain and especially impressively in Iberia is that essentially all the y chromosomes after about 4200 years ago are from the steppe so that means that while only 40% to the percent of the percent of the percent of the percent of the percent of the percent of the percent to the percent of the y chromosomes do and that's telling you that there's a social inequality that's associated with the movement of these new people into this region where the males are contributing much more to the next generations than the local males the incoming males then to the local males so this stems the first part of my talk and I'm going to now come to the second part of my talk which is about genetic formation of South Asian populations in light of ancient DNA data so in 2009 in one of the first paper that mics my colleagues and I were involved in publishing on South Asia out of a whole series we analyzed variation from 25 very diverse groups in South Asia that we were collaborating on generating genome-wide data on with our colleagues in India so no ancient DNA data here but I'm actually showing you a plot from our paper into thousand and nine so this again is a principle component analysis again about 600,000 variable places in the genome here and I'll label these individuals so these are Europeans and central and and these are European populations these are East Asian populations as well as some a couple of groups in far northeastern India which are genetically similar to Chinese or somewhat similar to Chinese and most of the Indian groups fall here so what we immediately noticed is that many of the groups in India form a gradient of different levels of proximity to Europeans and Central Asians and URIs Turner's and there were exceptions to this for example people who speak austroasiatic languages like Kazi or mundari languages and of course tibeto-burman languages over here but we were focusing on the people who spoke or the vast majority of the population of India and both North and South India who speak into European languages related to the languages spoken in Europe and Armenia and Iran and Dravidian languages which are the languages of much of the south of India and we noticed this gradient and we developed a lot of new statistical technology to understand what this gradient was in the we were able to show that this gradient was driven by population mixture that is that the people on this gradient with different levels of proximity to European Central Asians and URI Turner's were the result of that least one mixture event between groups with different levels of proximity to Europe so we call these the ancestral North Indians and the ancestral South Indians and the ancestral North Indians had some kind of relationship not so deeply in time only maybe the many thousands of years rather than maybe tens of thousands of years to European Central Asians in the Easterners and the ancestral South Indians the closest natural we could find them or Chinese or two indigenous people at the end amine islands in the Bay of Bengal which are near Sumatra in Indonesia and India and Myanmar and people in India today are mixtures in different proportions of these two ancestral populations which are different from each other as Europeans in East Asians today so we documented this in 2009 and we could sign proportions of mixture of ancestral North Indian related ancestry which ranged this is not the full extent of the range but from about 80% to about 20% and really remarkably it's correlated in South Asia and in India too to cultural phenomena one is the speaking of languages so people who have relatively higher proportions of ancestral North Indian ancestry tend to speak into European languages oops more than Dravidian languages and even within states of India people of traditionally higher status in the traditional caste system tend to have relatively more ancestral North Indian ancestry than ancestral South Indian ancestry so what that shows is that the process of language spread and the process of the definition of this traditional social status are both related to the mixture process that occurred in some way so in the series of papers that we wrote subsequently we try to understand more about what happened and we want to know in particular what's the history behind the ani ASI mixture process and we had three ideas in our head and of course the truth is probably none of these three ideas but we have three hypotheses so the three hypotheses were that what we're seeing is the effects of populations coming together during post iceage migrations which the big the big melting happened after 18 and especially after 14,000 years ago and even though South Asia was never glaciated the landscape changed very dramatically and perhaps that pushed new populations into contact but had not been pushed before another possibility is that what we're seeing is what happened in Europe the spread of farmers to this region for the first time from the core region of farming in the Near East instead of Western westward as in Europe eastward into South Asia and it's very clear the archeology documents that the last possibility is that what you're seeing is events related to the collapse of the Indus Valley Civilisation and the rise of archaeological intensification of archaeological cultures around the game Ganges after about 4000 years ago so these were three temporarily very distinct events and we want to see if the genetic data could provide any information about whether the mixture correlated better to one of these times than to others so in order to answer this we developed new technology new statistical technology for estimating dates of mixture and it works like this so here is supposed to represent two chromosomes so you add every part of your genome you have 23 pairs of chromosomes these are the packages on what your DNA letters are stored and so this is a cartoon of the two chromosomes that you might get on two copies of chromosome 3 colored by whether they're a si or a ni and so if you look at the great-great grandparents of someone today on their chromosome 3 after their mixture of Na a and I and a si you might have a first-generation mix of them every generation when you produce an egg or a sperm you create one or two breaks in those chromosomes per generation splice them together and send a mixed chromosome down to your offspring and you get one or two of these breaks every generation until today you have chunks of ancestry that reflects several generations of breaks and by measuring the average size of these chunks you know how long it's been since the mixture process initiated you're basically it's like you're chopping a salad at a constant rate and you measure the sizes of the chunks the dice size and so that's what we did we looked at the average dice size and we estimated based on this that across South aged or the mixture happened between two to four thousand years ago on average depending on the group which very strongly suggested that the mixture happened in the last scenario after the collapse of the Indus Valley Civilisation and the beginning of the spreads of a rise of new archaeological cultures and urban intensification and the Gangetic plain so the summary of hat population history before ancient DNA is that South Asians many South Asians are a mixture of at least two ancestral populations and Susteren North Indians and ancestral South Indians they mixed convulsively after Ford between four to two thousand years ago in a way that's related to the institution of caste and the spread of indo-european languages what happened next so here's a paper that we're now trying to bring to publication which is a very large-scale study of ancient samples shown in red here Co analyzed with many many modern populations shown in dots here and so we are trying to make sense not just of the history oops not just of the history of the these populations shown in red but also their relationship to the populations in South Asia you see there's very little ancient DNA so far from the South Asia the only data we have is from the Swat Valley and present-day northern Pakistan oh that's very interesting data so there's lots of stuff in this study and I've been a focus really on just one group of samples from here the from Turkmenistan and Uzbekistan which is particularly revealing as well as here from far eastern Iran so this is a site from about 4,000 years ago called gone or Tet Bay in Turkmenistan is a very impressive huge town walled town it was one of several built at that time as part of what's called the bacterial margiana archaeological complex which is one of one of the greats of one of the civilizations of the ancient world which was really only discovered by archaeologists in the 1970s and when we performed principal component analysis on the data from more than a hundred individuals from backyard margiana archaeological site complex sites from gone or in three other sites we got a plot like this and what's very powerful about this plot because we have so many samples is that we not only observe a main cluster of samples but we observe clear groups of outlier individuals so these are people's buried in the cemeteries on around this town but they're genetically not homogeneous this would be like digging up a cemetery here you would find perhaps a main cluster corresponding to the majority group but you would find other groups and it would be telling you about the groups with which the main group was in cultural interaction so prior to 4,000 broadly this group is similar to ancient Iranians that we also have data from from around the same time and then when you see the first individuals who are outliers from the screw prior to four thousand years ago there's a group of them that clearly are ad mixed mixed with people related to siberian hunter-gatherers and are presumably the hunter-gatherers of Kazakhstan and Central Asia who are known archaeologically but we didn't have direct sampling from but we think that what we're seeing is local hunter-gatherers being buried in these cities and coming into contact with the urban people of towns like this after four thousand years ago you see a new group appearing only then and these people have young eye a steppe pastoralist ancestry the same type of ancestry that gets into Europe after five thousand years ago and we see it or appearing not just in the boundaries the cemeteries of this town but in multiple places in Kazakhstan in other parts of Central Asia just around this time so it's clearly hitting this region at this time but not before and then finally we find these three outlier individuals with South Asian admixture these are we first found a girl buried in a pot who had a bit of South Asian ancestry we found her about a year and a half ago and we noticed you had South Asian a mixture was clearly coming up from the south and she was buried along with this main group and it's known that these people in Goa or in these other BMAC dr. and Lauren Gianna towns were in cultural contact with the civilization to the south known as the Indus Valley Civilisation so we thought we might have an immigrant from the Indus Valley Civilisation or other points in the Indus Valley so we found this a year and a half ago and we found we collected more data because we thought that this was so interesting and we found two other individuals and then we found another eight individuals from a town called Irish okhta in far eastern Iran so we now have eight individuals with ancestry like this which are clearly migrants from South Asia and the rest of their ancestry is Iranian related but it's not of this type so they're not just South Asians mixing with the local people these are migrants are descendants of early generation migrants from somewhere else so here's the summary of what we think the data are showing so these individuals from gone or from the bacterial margiana complex and sorry shota and for Easter Ron are a mixture of two source populations and ancient Iranian related group in an ancient South Asian related group without any step ancestry not no ancestry related to step pastoralists this is the first layer of gradient that we see in South Asia but this is a gradient that existed four or five thousand years ago not like the one today after four thousand years ago and especially after three thousand years ago from our samples from sub from Pakistan we see a new gradient forming and this gradient is a mixture of individuals sorry this line should actually point all the way out there rather than hitting the triangle here this gradient is a group that's a mixture of different proportions of ancestry from this indus valley related or in disparate free-climb gradient and the step pastoralists but you still don't see people like South Asians today the ancestral north indians fit well in modelling along this gradient and then the ancestral south indians form as an extreme point on this Indus Valley Cline more closely related to the groups with less iranian ancestry and people in South Asia today are mixtures of these two mixed populations and we're here's where the Indian groups fall if you look at the proportion of ancestry coming from steppe pastoralists genome-wide verses on the y chromosome you see today groups in India tend to be above the line which tells you that similar to what happened in Iberia more of the ancestry of these groups is coming along the y chromosome then is coming up from the RET from the maternal line and so it's telling you that there's some degree of a socially unequal process coming occurring between these two groups as they mixed another phenomenon that's very interesting is that if you actually look at the groups that deviate from this mixed model that has too much steppe pastoralist young er related ancestry compared to this model they tend to be groups that are in the Brahmin or Bumi har groups Brahmins are groups that are traditional custodians of the sacred the indo-european texts and this provides a new line of evidence that those texts and that indo-european culture is connected with step ancestry in particular and that the bringing in of this step ancestry ultimately derived from the yam nya via Late Bronze Age middle to late bronze age groups that descend from them in part is the source of these languages in Europe in in in South Asia that also spread to Europe this also explains and probably explains why linguistically the languages of South Asia and Iran have a close cousin in both of Slavic languages like Lithuanian because genetically we can see that the middle to late bronze age groups are genetically very similar in these two regions so the conclusion of this is that there's two parallel sub hunt subcontinent's of your age Europe and India farming forms in the near East eleven to twelve thousand years ago it moves explosively both east and west after nine thousand years ago mhm and then it moves across these two sub continents over the next few thousand years for people with far more related ancestry mix with local hunter-gatherers moving slowly because it takes them time to adapt culturally perhaps and ecologically foreshore to the new environments and to which people are moving meanwhile these people in the steppe north of the Black and Caspian Sea spread to the boundaries of both region and then formed mixed populations and mixtures of these mixed populations drive the primary gradients in both regions today so I'm going to conclude just a little bit by talking about the extraordinarily high population substructure in South Asia today Indian groups are much more genetically differentiated from each other groups today than European groups we notice this for the first time in 2009 there had been just a paper that claimed the opposite by studying actually not Indian groups from India but Indian groups from the US and categorizing them by state and there were it's very little differentiation but when we looked at the sampling by based on traditional social status groups in different villages in India there was huge differentiation for are four times greater than the average differentiation between groups separated by much larger distances in Europe and this didn't go away when we controlled by length for language or traditional social status or by state of India and the truth in India is that actually India today is not a large population like Han Chinese are a large population where people easily mix across different places and boundaries in fact in the India is a large number of very small populations there's at least about 5,000 well-defined groups and some people count many more where there's very little mixture across groups what's actually happening is that many groups in India today our founder events with a relatively small number of individuals founding a large number of descendants today anywhere from thousands to tens of thousands hundreds thousands or even millions but rarely more than a few tens of millions an example is the vice EF from Andhra Pradesh which is around Hyderabad today in the middle of the peninsular India and in this group we were able to show using these chunks of DNA that are today inherited from these founders who gave rise to a large fraction of the several million people in this group today that the founder event is at least two to three thousand years old that is a group two to three thousand years ago that was very successful in having offspring and their kids were very successful in having offspring and today they have many descendants today so this is a really interesting finding because for example Nicholas Dirk's in this argument in this in this book cast of mind argued that caste in modern India is an invention of colonialism in the sense that it became more rigid under colonial rule with the British using it as a way to govern India and I'm sure that's true to an extent but what he argued is that it really in many places effectively didn't exist prior to this time however groups like the VISAA have remained endogenous with essentially no genetic flow into them from nearby groups that they were living very close to for two to three thousand years and what that means is that the strong endogamy rules in india today at least in many groups have been maintained as I say genetics shows that many current distinctions are ancient and strong and Dogme groups must have shaped marriage patterns for thousands of years and what this is telling you is that genetics documen to cultural change after 5,000 years ago as I mentioned to you before there's this convulsive mixture process happening in South Asia and then there's a profound switch to endogamy where groups have hardly mixed with each other for thousands of years in some cases so what that's telling you it's sort of talking to linguists into textual scholars the earliest Vedic texts sacred texts in India like the Rig Veda describe a multi-ethnic society where for example people without indo-european names are kings and poets in that society but at the end in appendix that's thought to be added later it describes the beginning of the system of social stratification and within a few hundred years another text the law code of Manu describes a very extensive system of social stratification related to the current caste system and maybe we're seeing in textual evidence a parallel to this locking in of endogamy after a period of profound mixture that corresponds in time to the dates that we're talking about I wanted to add a personal note just before I finish my talk that I myself come from an ancient caste Europeans don't have many castes but there are a few of them one of them is Ashkenazi Jews who have traditionally fulfilled the definition of a caste so one definition of a caste is a group that interacts economically but not socially with groups with which it lives among for example their dietary rules or body modification strategies that prevent people from mixing with the groups they live among but they have certain economic functions for example money lending or selling alcohol or other things and that defines as historically defined Ashkenazi Jews within Europe at least some parts of Europe so while I'm very much an outsider in India this perspective very much is informed and some provides some kind of sympathy I'm definitely acknowledged being a far outsider of my work in India so in the Jewish community in the Ashkenazi Jewish community this is another founder event 10 to 20 million people descend from a relatively small number of founders four or five hundred years ago who had many descendants today and as a result people who are the product of a marriage between two Ashkenazi Jews are an elevated rate for diseases that do to one at the reign mutations carried in those founders we all carry a few dozen mutations that would be lethal if we carried them in two copies but if the founder for the mutations carried by the founder unfortunately they got jacked up to high frequency because they had lots of kids and if their descendants marry each other 20 generations later you're in trouble so in the Ashkenazi Jewish religious community for example some of my first cousins there's a great deal of arranged marriage and there's this community testing service called Doria taurine which tests hundreds of high schools every for the standard set of known genetic Jewish mutations and in the arranged marriage simply don't match up people who both carry the same mutation they call them incompatible and as a result in Ashkenazi Jewish communities in Israel in the United States the rates of these diseases like tasty ox and Gauchos diseases plummet had gone to almost zero in these communities so in India this seems like another medical opportunity except instead of a one sort of relatively minor population like Ashkenazi Jews the pup that numbers here are very very different so what we found in a paper we published in 2017 that is a problem here approximately one-third of Indians are from groups with clinically significant founder events stronger than Ashkenazi Jews are Finn's so here is a rank ordering of 250 Indian groups in a survey that we did by the strength of their founder events here's where Jews are and a third of them have stronger founder events the now she cannot accuse are Finn's here's the Vice yet which is about the same strength as Ashkenazi Jews or Finn's here's many groups that are much stronger including groups with millions of people in them and very very strong founder events which would result in proportionately many more recessive diseases so there's every opportunity to do the same thing in South Asia as in Doria serine where for example where there's also a lot of arranged marriage in South Asia as well as you could do just testing in utero testing there's epical issues about doing all this but I just want to highlight this as an opportunity literally probably more than 100,000 lethal recessive diseases are in India every year do this process and right now with the ability to discover what these are is it's really trivial right now with modern genetic technology it's not even expensive but there's not a lot of attention to this opportunity in South Asia instead genetic attention is more toward common things that are common across South Asia not specific to each group like specific types of diabetes or melon metabolic syndrome or heart disease which of course are also very important so there's a lot of opportunity to do that so I wanted to conclude by thanking you I think that I want to highlight the parallel history of South Asia and Europe tell you that there were these similarities the movement of farming or farming or of people influenced by farming into both regions mixing with local hunter-gatherers then the impact of people descended from steppe pastoralists who spread after five thousand years ago of course very important profound differences as well and in general these are just two instances of what's being revealed again and again by ancient DNA which is that mixing of groups is extremely different from each other is a common feature of human nature you might think you live in unusual times with the people coming across the Atlantic and Africans Europeans and Native Americans and East Asians all mixing together and the Americas but in fact we're not in an unusual time the equally profound events have occurred again and again in our past a form dozen or in all of our histories and there's really no no groups that are pure or have their ancestors have really been in the same place forever but I think we should learn from that and really feel more connected from that rather than more separate thank you [Applause] so now we're going to have some time for Q&A and I'm just gonna run microphones around the room for people I have a question from your research thousands for thousands of years stemming probing back into human history in modern times in the last couple of thousand years humans have been very warlike and it very seem to be very violent do you have any idea from your research that humans were less engaged with warfare than they are today or is there no research so the question is the question the question is humans there's a lot of history of war and conflict in the last thousands of years that we know about from the historical record and does the research provide any information about that I think the best research on this topic is from archaeology we are really you can actually try to understand whether there's evidence of war and I think that archaeologists are quite consistent in showing that sort of war in the form that we know it with these organized large groups fighting each other is a very recent phenomenon just of the last few thousand years because there aren't state scale societies prior to that however conflict is common in the past and what you can see in the genetic data are mixtures between groups that differently involve males and females and what that's telling you is that you're seeing unequal mixtures of groups that's what I discuss as one of the topics in this chapter of my book called the genomics of inequality where surprisingly genetics is giving you evidence of asymmetry in the mixture process between groups so I think that it's tempting to think when you see the example of the young ayah of an horseback riders coming from the steppes kind of pillaging Europe or something like this but that's an anachronistic picture these people possibly could probably couldn't ride horses they didn't have stirrups they couldn't shoot a bow and arrow from a horse they didn't work in organized bands of that scale however they nevertheless did move across Europe and their descendants sort of produced these very asymmetric mixture events that we observe so I think it provides important observations data points that provide the archaeologists with new information to contextualise within the archaeological record and to see how groups in the past might relate to groups today in terms of bang gear thank you this is sort of a technical question but I didn't understand something you were saying about the groups moving into Iberia with the the past or the step pastoralists moving in you said that a hundred percent of the male ancestors were the steppe pastoralists but only forty percent of the overall ancestors were step by store or less but aren't half of the ancestors of any individual male so am I just misunderstanding something about the the earth might take there okay the question is it's a great question so the question is what can you you're basically asking me to explain what I mean by difference between the whole genome ancestry coming from steppe pastoralists and the y-chromosome ancestry and I think the right way to say this is to ask what fraction of your ancestors were living outside Iberia several hundred years ago before this mixture event occurred and the answer is on any random genealogical line say your mother's father's mother's mother's mother's father forty percent only live outside Iberia but on one particular line your father's father's father's father's father's what percent of those ancestors live outside Iberia and the answer to that is a hundred percent in fact the event is more extreme than I told you because um because that forty percent is the number of people who live outside of Iberia but actually the people who came in Tiberio were already half farmer by that time and so actually it's only 20 percent step ancestry that makes it in Tiberio but it's nearly a hundred percent on the y chromosome on the entire male line so what you're seeing is the result of a very asymmetric process where males from the SAP are successfully propagating their genes event on the entirely male line in Tiberio you see phenomena like this also in more recent times for example five to ten percent of the Y chromosomes in large parts of East Asia descend from a single common ancestor or maybe eight hundred nine hundred a thousand years ago and what that's telling you is that some individual at that time fathered males who fathered males who fathered nails who are responsible for literally you know you know hundreds of millions of people and you know that corresponds to the Mongols spread and you know the time maybe of Genghis Khan when there's known document it's very extreme social stratification and we see other events in the genetic literature which are much more impressive than that for example if you actually look at reconstructed population sizes on the entire flee female line and I talked about this in that chapter I'm mentioning on the mitochondrial DNA you actually and track the expansion of human populations populations are pretty small before fifty thousand years ago then after ten thousand years ago they grow a lot that's sociated by the time of spread of Agriculture and then they keep growing after that time with no obvious interruption if you do the same thing on the y chromosome you see the same 50 thousand years ago a population expansion and you see the same expansion with agriculture but there's a huge crash about five thousand years ago not seen on the mitochondrial DNA and what that's telling you is that individual males four five six thousand years ago are successfully propagating their gleans much more effectively than other males and what that's telling this is not differences between the sexes but rather inequalities among males in terms of ability to propagate their genes and this period corresponds to the first period in the archaeological record with extreme wealth concentration where the first time you see treasures and graves for the first time you see great inequalities of wealth and what this seems to be telling you is that that's accompanied also by extreme inequities in which males get to reproduce possibly there's also extremely inequities and female wealth but you don't see that as much because females don't have the same dynamic range of number of offspring as males do this is fascinating you'll have enough to keep you busy for a long time but looking into the further future how much further in the past can you look in other words what sets if you know the fundamental limits on how long DNA can last the question is about ancient DNA preservation the oldest ancient DNA of substantial quality is about a seven hundred thousand year old horse from permafrost from Alaska the oldest human DNA which is much poorer quality is about four or five hundred thousand years ago from Spain it's an early Neanderthal lineage individual that was sequenced a couple of years ago and that's probably would probably be unlikely most people think that DNA will preserve much more than a million years and substantial amounts so we're really not going to learn about probably dinosaurs or you know Maya seeing evolution except to the extent that those lineages have left lineages that we can sample with ancient DNA that are now extinct the question there I think you're forgetting about Jurassic Park maybe we will learn about dinosaurs No so you drew some attention in the popular press about and maybe I'm getting this not exactly correct but something about like there's more variation within races versus across races I guess obviously the definition of race could be tricky at times but this analysis would tell me that there actually is more variation within quote-unquote races than that I think popular notion you were trying to refute did I say that properly or did I get that the question was how I'm gonna try to restate your question and tell me if it's not not a restating bigger is what in the direction what you want I think you're asking about on the one hand how this relates to conventional notions of quote races that's part of your question and the other question is what this data and what genetic findings mean about the degree of variation amongst you populations and how much better portions and groups versus individuals and so so so the first thing to say about that is race is clearly a social concept people it's the categories are different in different countries in the world and it's really a description of how people see each other and categorize each other in a group and while it correlates in some cases to genetic groupings in other cases for example in the Latinos it doesn't which includes people of entirely European ancestry mostly African ancestry Native American and European ancestry with very little African ancestry so it's a heterogeneous genetic grouping so it's a social idea that explains a lot of cultural variation but it has and sometimes correlated to genetic variation but not a perfect one-to-one matching we geneticists met in my work I never use this in my papers because it's so loaded with cultural and historical baggage that it makes an imprecise term and we just don't use that term we use terms like ancestry instead not as they use euphemism for race but in an attempt to describe concepts of genetic differentiation across individuals and groups without all that baggage in that loading so the second question is about the degree of variation within and across groups and I'm incompletely agreement with what's been determined from genetics for almost the last 50 years beginning with a paper and by Richard Lewontin who worked here in 1972 I think who pointed out that the degree of genetic variation between any two individuals within this room is on average six times greater than the average variation between the average but between two groups across the work room so most of the differences that you can appreciate amongst people are much larger than the average differences across groups I think that perhaps the issues you're referring to are related to the work the the writing I was doing to try to emphasize that while those differences are small relative to the great differences you see amongst people in this room they're not nothing and we need to actually develop a language to actually think about how to accommodate those differences in our public discourse because if we don't the geneticists who are going to find these differences anyway in words I think that we need to develop a robust way to talk about these small differences that do exist y-you talked a couple times about replacement events do we have like any way of knowing what that word means in terms of like was it violent or just like disease yeah the question is replacement events so that's a possibly problematic word so in the case of Britain 6,000 years ago the population that you sample after 6,000 years ago and before 6,000 years ago 99% of the ancestors of people 6,000 years ago or more can't have lived in Britain based on the sampling that's been done so what does that mean it doesn't mean that the new people came in and killed the local people does it mean that they crowded them out because they had used resources more efficiently and those people just were absorbed perhaps through mixture or just we're not as successful at reproducing because their lives were less efficient or they were displaced from their lands we just don't know what the genetic data it does is it provides facts about movements of people changes and ancestry relationships amongst groups but we are not the experts and we are not really able to really describe what happened and you know maybe we'll never be possible but I think that needs to be done in the context of the archaeology as using the genetic data as one source of information to try to learn what might have happened so really I've been trying to use words that really don't prejudge what happened again in Britain at about 44 to 4,500 years ago there's this new massive influx of people coming in within a period that can at most be a couple of hundred years what does that mean in terms of the facts on the ground I don't know but it actually happened so we have to deal with it and think about it somehow hi question the original farmers that went to India and Europe how closely related were they you mentioned in your book there was a huge difference between the Zagros Iranian early farmers in the 1115 peoples yeah the Anatolian spitted the Western so farming one of the findings of our 2016 work is that farming 11 to 12 thousand years ago developed in a multi-ethnic Society world where different types of crops like wheat and barley or goat domestication happened in different parts of the Near East all the way from western Iran to eastern Turkey to the Levant and those groups genetically are extremely different from each others so ancient iranian early farmers and herders are as different from eastern Anatolians and western Anatolians as europeans and east asians are from each other and in fact ten thousand years ago in this region there were multiple groups in western Eurasia where today there's a lot of homogeneity that were as different from each other as Europeans and East Asians so farming was developed in this multi-ethnic society and in the subsequent few thousand years these groups all expanded in different directions and mixed with each other and glommed together to form the region of relatively relative homogeneity that we see today the ancestry that moved into Europe was from the western end of that gradient the Anatolian farmer related ends and was very different than the type that moved into South Asia or that developed or spread in South Asia which is more related to Iranian farmers and so what you're actually seeing is not the same population but different groups that are quite differentiated from each other from each other within the core region of the Near East so the Levantine group is actually probably is it yet again quite different and actually what you see in terms of the spread into africa and mixture with local groups is related to the Levantine groups we don't know if it's from the Levant itself or from a North African population that was related to it but in any case the West Eurasian related ancestry that you see ad mixing after five thousand years ago in East Africa is more closely related to the Levant team into the Italian or the Iranian groups so we have time for about one or two more questions hi I'm I was curious that I assume that population migration was largely happening along waterways but that the domestication of the horse obviously played a critical role to the movement of populations and I'm wondering if there is another research group that is studying the ancient DNA of horses and whether perhaps what they're seeing supports the movements that you're seeing with the human DNA was question what is genetic say about horses I've just talked about humans and arguably much more interesting things to be said about other animals there's actually a particularly amazing research group based in France right now that by Ludovic Orlando that's been studying the ancient DNA of horses and of all the ancient DNA work going on you know it's very impressive work maybe more impressive than a lot of the work that I've told you about we still don't know where the domestication of the horses that are all around today occurred the first horse domestication occurred in Central Asia in a culture that's been called the bowtie culture about forty fifty five hundred years ago in present-day Kazakhstan and that ancient DNA of those horses in a published paper from Orlando's group last year showed that these groups are not related to modern horses these first domesticated horse there instead related to professional Wolski's horses which are the wild horses of Mongolia today that were thought to be wild horses but that work showed that in fact for Jaworski's horse or feral domesticated bowtie horses that went wild and in fact there are no purely wild horses left anymore in fact they're all domesticated or feral eyes domesticated horses almost for sure it's a possible that it's a question where the first horses were domesticated a possibility is that it was the yemaja itself and that the horses they used and widely used spread and I think that will be a really exciting thing to see in the coming years but horses cattle goats wild animals extinctions and tried to gain some insight in wine to the mammoths and the different megaphone and when extinct is all something that canon is being addressed with ancient DNA thank you hi so I thought it was very interesting when you showed the map of South Asia like particularly I think in the Swat Valley the percentage of the I guess at the indo-european ancestry you saw was really high so I'm just curious that did you see any consistent like gradation like as you move eastward that the like to populate like the Indian populations who live like more on the eastern part of the country have more of the South Indian ancestry concessional South Indians compared to like the like the north western part of the subcontinent so the question is what of the gradients correlate to geography and I think the and I think that broadly it's a Northwest to southeast earn gradient but really the smallest proportions of steppe pastoralist ancestry are from the south of India and especially in groups that are traditionally outside the caste system some tribal groups that speak for vidiian languages in eastern India in what's called the tribal belt there are groups that have even less steppe and Iranian farmer related ancestry than the groups in southern India and these groups often speak all strategy attic languages and are particularly interesting and revealing for what's going on because it's clear that they derive it's part of their ancestry from the deep hunter-gatherer related component of South Asian ancestry that is most Indian groups is only present in form mixed with Iranian farmer related ancestry and so what that tells you is that at the time that Austria ionic languages came into South Asia they almost certainly came from the East likely associated with the spread of rice farming from the east and after 5000 years ago these people when they spread in were encountering groups that had not yet become the ASI who we know actually formed through mixture after 4000 years ago and so in the east of India you actually do see these groups that are Sam from the press previous variation without this impact of Iranian former related ancestry alright I think that wraps up our program tonight thank you [Applause]
