 Yn ymgylchedd yma, rwy'n amser yma yn ymlaen i ddelweddolol y Llyfrgell Pugol Cymru, ond rwy'n credu i fynd i'r projeg yr Llyfrgell Pugol Cymru yn ymlaen i'r iechyd, ac ydych chi'n cywir yn ei bwysig i'r rhan o'r corffwyr. Mae'r angen i'r angen i'r angen i'r angen i'r projeg. Yn ymlaen i'r angen i'r angen i'r angen i'r angen i'r angen i'r angen i'r angen i'r angen i'r rhagwyr cyffreddau that we have compiled in a paper that they have recently published at the end of last year in a journal called Scientific Reports. So if you type this title into the internet you should be able to find the paper if you want to actually have a read of the actual document. If all this is published data, so if you want to take photographs, that's completely fine. A rare occasion when I am presenting. And below are a list of the authors on the I'm the one at the front. So, today I'm going to be talking about the main results of that paper, but first I thought to kind of introduce the concept of population genetics, also the genetics of Ireland as we knew it before we started this work, and then finally talk about the Irish DNA Atlas, and then talk about the main results of the paper for the main brunt of this presentation. So the main results of the paper were split into papers of two broad groups of results, one investigating the population structure of Ireland, which I'll get into what that actually means, and then also we investigated the genetic ancestries that are within the modern Irish, so that's kind of the bones of this presentation. So first of all, just to explain population genetics. So this population genetics is to do with humans, obviously you can use population genetics on lots of different organisms, but I study humans. So the human genome is about six billion letters long, we call them base pairs, and these letters are either AT, CLG, kind of like a binary code, but a base four system instead of base two. The genome can be thought of as a library of the blueprints that house a human being, and this library is organised into, or a book, sorry, and can be organised into chapters, and these chapters are called chromosomes, and they're separate bits of DNA, but you can put all of the DNA together, so if you put one copy of all my DNA in one of the cells, I think it's... I think the length of it is about as high as I am if you just put the entire molecule side by side on its end. And the basics of population genetics within any organism, but also with humans, is that human beings vary. So if you take around two random people, two random humans, and you compare their genomes, about one in a thousand of these letters are going to be different in each person on average. And these are mutations, or variants as we like to call them sometimes. The scale of these mutations can vary, so they can be single letter changes in the genome, or they can be very large scales. They can be whole swathes of a chromosome, duplicated, deleted, inserted, changed, inverted or whatever. And as you can imagine, the consequences of these mutations are also varied. So the vast majority of these mutations are actually quite neutral. They really have an effect on you as an organism. But a smaller fraction of them will be deleterious to you as an organism. So they'll create a genetic disease. So for example, cystic fibrosis. But an even smaller fraction may have a beneficial effect. So they'll give you a selective advantage over your generational peers. So a good example is lactose persistence. So within the Irish there is a mutation that allows you to continue digesting lactose milk into adulthood. And we think that that's to do with vitamin D metabolism and the, yeah, to do with that. So Irish genetics, if you compare Irish genetics to any kind of population within Europe, one of the defining signals of the Irish genome is this signal for lactose persistence, which is quite cool. So population genetics is simply just the study of this variation between groups of humans. So that's what I do as a living. And it's rather interesting. So these mutations, these variants can be inherited in different ways. So obviously from generation to generation you inherit the DNA of your parents. But you have these two types of systems of inheritance, two modes, two streams of the river of inheritance. One's uniparental and one's autosatron. So uniparental are a small minority of the genetic material that you carry on to your children. Uniparental just means from one parent. So it's either the Y chromosome or the mitochondrial chromosome. So the Y chromosome is inherited from father to son and the mitochondrial chromosome you're inherited from your mother. So I inherit my Y chromosome from my mum, but if I had children and if I had any children they wouldn't inherit my Y chromosome. They would inherit the mother of my child's mitochondrial chromosome. But for the Y chromosome I would pass on to my son but not obviously to my daughters. So the Y chromosome and the mitochondrial are very, very small compared to the whole genome. There are about less than 5% of the entire genome. And there's very little what we call recombination. So in the general genome, parts of the genome are swapped and basically shuffled from copy to copy. And this creates new combinations of genetic material and is one of the reasons how genetic variation can happen. Because the Y chromosome is very small for example and doesn't recombine much. Population differences, if you compare the Y chromosomes between two different populations, are exaggerated compared to the genetic variation that you see on the whole genome. So they're quite useful in the kind of an exaggerated viewpoint of population genetics. But to get a more accurate, more nuanced view on population genetics you have to look at the autosome instead. So this is the vast majority of your genomes. This chromosome is 1 to 22. And these have a 50-50, roughly 50-50 inheritance from your mum and your dad. And the majority of the genetic variation are located on these chromosomes and they recombine. Because they recombine, because they're so large, i.e. nearly 3 billion base pairs long, they're a very complex system to study. And we haven't begun, and we haven't studied the vast majority of the amount of mutation on variations in the genomes, like the bottom of the oceans. We know some stuff that's in there, but a lot of it's kind of unknown. So the data that I'm going to be presenting today is based on the autosome. So we haven't actually generated uniparental data on using the average DNA adverse yet. So, as well as most of the inheritance and population genetics in general, I'm going to be talking a lot about genetic structure. So genetic structure is simply the stratification of individuals based on their genetics. And this is because humans tend to group together to the exclusion of other groups. This may be for sociological reasons, or it may simply be because of geographic reasons. For the vast majority of human history in Europe, you're more likely to have a child with someone that also lives in Europe compared to someone that lives in Asia, for example. And because people tend to keep within their own groups at a continental scale or even at a country scale, the variety of groups varies because they keep the same genetic material and the same genetic variation within that group. And this is genetic structure, essentially, the different types and variations. So you might have the same mutation that are in all the populations in the world, but it's the frequency of that variation or that mutation itself might be the difference between populations. So this is genetic structure. And genetic structure is what we essentially detect around these analyses. This is a really good and very famous example of the genetic structure that's found in Europe. It's from a paper in nature 10 years ago. So before I started my undergraduate, let's ignore how young I am. So this is an xy plot through... So you've sequenced about 1,000 European individuals and you sequenced their genome so you've described their genome in about 600,000 different positions for 1,000 individuals. So this is quite a complex dataset and is very multi-dimensional. I.e. that you could describe 600,000 times whether they have a mutation at that point or whether they have a mutation at that point so on and so forth. So PCA is just a simplification algorithm. And what this does is output principal components but just simply just dimensions. And what happens is in a single dimension, individuals that are genetically related are on the same area of that dimension. So they cluster together and individuals that are genetically far apart cluster away from each other. And you can see that on this plot. So each letter is an individual and they're called by the country of origin. So this is an INNE, so this is an island, ENS is Spain, IT is Italy, and you can see that the peninsula starts to appear. So you've got the Iberian peninsula down here, you've got the Italian peninsula down here, the Balkans and Spain down here, France up here, and you can see Britain and Ireland over here. And this basically shows the main determinants of where your genetic material comes from is where you're born. For the vast majority of history within Europe or at least modern histories in the last couple of thousand years. So this is really, really cool. This has obviously made nature back in 2008. But nowadays we're looking at more in detail on these individual populations. So we want to know what the equivalent of this structure is within Ireland, for example, or within Spain. And this is essentially what I've been doing as part of my PhD. Looking at fine scale structure, so to quickly summarise the genetic work in Ireland. So to understand the genetic of Ireland you have to kind of understand the geographic location of Ireland. So Ireland, if you've got the map of Europe, Ireland is all the way over here. So to get to France you can go from Spain, to go for Italy, to go for Germany or wherever. So France is kind of quite admixtures. The mixture of lots of different populations from different areas in Europe. But to get to Ireland you usually have to either take a sea route up here, go through Britain, then to Scotland and then do a small hop across the Irish Sea to Ireland. And therefore, relatively speaking Ireland's quite isolated from Europe. So it's kind of like a reservoir of ancestry. And so this, and despite its geographic close to Britain you might expect it to be distinct from Britain because that has been a level of isolation between Britain and Ireland historically. So, using very simple methods back in the 60s groups in Trinity looked at the frequencies of ABO blood groups within Ireland for example. You start to see some population structure, i.e. kind of the stratification of genetic material. So if we just focus on the O blood group and the A blood group, the different shadings of hatching means progressively more frequent with the greater frequency of these blood types. So you can see that O is more frequent in the west of Ireland and A is more frequent in the south-east of Ireland. This is the first real evidence that there was some level of genetic structure within Ireland. Just to get forward about 50 years suddenly we're working with wider chromosomes. And we know from these papers we know several things about the population genetics of Ireland. For example, Ireland and Scotland has a degree associated with this wipe chromosome haplotype M269 and the haplotype is essentially just a type of wipe chromosome. There's a subset of that M269 called the Irish modal haplotype and it's associated in Ulster. This is what I'm showing here on the right. This is kind of a heat map of where this wipe chromosome is found and you can see that it's really centred around the north-west of Ireland. There's also another separate haplotype that's associated with the north-south of the island monster. This shows that there's also genetic structure if we look using white chromosomes. All these different methods are kind of looking at the structuring of the genetics of Ireland through a number of different lenses. We're seeing through independent lenses that there does seem to be this stratification of individuals based on their genetics in Ireland. We're seeing the genome and we're not getting the complete picture which is what we could get if we used autosomal analysis. There was also work done investigating white chromosomes and as an aside they didn't find any evidence of north's white chromosomes associated with north's surnames in Ireland. So we skipped forward another couple of years and we're now looking at autosomal data with Ireland and we did our lab ten, nine, eight years ago and did some of this work. The preliminary work in Ireland suggested that there was a slightly lower genetic diversity in Ireland if you compare mainland Europe which is what you'd expect considering geographic isolation from Europe. Genetically, Ireland is distinct from Britain but its closest genetic relatives are from Scotland. We should kind of expect given the history between Ireland and Scotland. There was also evidence found of a structure within the south, west and north of Ireland and this came from a paper in ancient communications last year. So these are three different kind of groups of individuals based on their genetics and you can see that one is centered around the north, one is centered around the south and one is centered around the west. Lastly there's been work using ancient DNAs. Ancient DNAs really exploded in the last five or six years and this is because of a variety of different reasons but suddenly we have hundreds of ancient genomes whereas 20 years ago this was like a Jurassic Park kind of territory. So at the moment there are four published Irish genomes, one in Neolithic which is the first farmers of Ireland and three Bronze Age individuals. So we know that the modern Irish genome was established during the Bronze Age so if you look at the Bronze Age this is kind of a heat map of genetic affinity to modern populations. You can see the Bronze Age has the most affinity to Ireland, Scotland and Wales. But if you look at the Neolithic actually has the most affinity with Iberia as well as Sardinia. And this is saying that this is showing that the modern Irish when they're not they're not descended mostly from the Neolithic they're mostly descended from Bronze Age. So to kind of wrap all of that up quickly so we're seeing evidence of genetic structure in Ireland through different lenses. But what we don't know is using the autism what the extent of that structure is in a comprehensive review of the genetic in Ireland. But we also don't know is the genetic impact of historical migrations into Ireland. So I've said that Ireland's quite isolated geographically and historically but that's not totally the case there have been historical migrations into Ireland and those have brought people and they might have brought genes with them as well. But we don't know the historical impact of that so we thought decide to investigate that using the atlas as well. So the Irish DNA atlas is a DNA cobalt of individuals with regional Irish ancestry that's essentially it. What that means though is that when we recruit individuals they have to have all eight of their great grandparents born within the same region of Ireland ideally within 50 or 30 kilometres. And this is a result of the as a collaboration between a Royal College of Surgeons and the Genial of Society of Ireland that really is quite unique from around the world. There's no other cobalt that combines this level of genealogical data as well as genetic data together. On the right you can see a map of the individuals that we've currently recruited and each dot is the average latitude and the longitude of that individual's eight great grandparents' birthplace. So essentially where their ancestry is coming from. And you can see that we've got coverage across the island of Ireland. We've got some areas that are kind of like gaps. But for the most part we've got even coverage across Ireland. So we've got both North Ireland, Northern Ireland and the Republic as well. So when we collect samples for the Atlas we collect genealogical data so name, birthplace, birth year and that's for the individual their parents, their grandparents and their great grandparents. We also collect basic health questions in case we need to in case we're investigating genetic diseases within Ireland and of course we take informed consent. So each individual knows what we're going to be using their DNA sample for and how we're going to be using it. And then also we anonymise each individual. So each individual is given a code and then for the most part that code is used for that individual. Only a subset of people are allowed to have access to the data that could allow you to join those two parts of the data together. Of the samples, so each individual provides a survivor sample to us and they combine it with a preservative that allows the sample to just remain at room temperature for about 18 months I think. So we just ship them off in the post. We've got samples from Australia I think. We've got a sample coming in from New Zealand. Hopefully, fingers crossed. So yeah, it's really this kind of like that technology is really useful for us. So the DNA is extracted at RCN site. We do some quality control just to make sure that there's actually DNA in the sample and then we start the DNA on site for various analyses. So at the moment we've got about 265 individuals recruited. The average area between great-grandparents is about 30 kilometres and the average date of birth of those great-grandparents is 1849. Which means that each individual on average represents the genetic diversity around Ireland in the mid 19th century. We've got kind of a snapshot of Irish genetics at that time for any major modern population movement. And then on top of that we've been genotyping. So genotyping is the collection of genetic data on individuals running chemistry on their DNA and getting the results from that. So we've done two rounds of this genotyping and we've generated autosomal data using those rounds of genotyping on a chip called a sniffery chip. It basically describes each individual 700,000 times. And we've currently genotyped 194 of those individuals. 194 of those individuals are included in the analyses that I'm going to present to you on. Speaking of which. So first of all I'm going to be talking about the population structure within Ireland. Describing just the underpinnings of the variety within Ireland. So to do this we used a combined data sets. We combined Irish DNA atlas with a few more Irish individuals that are power to detect Irish genetics. And then we combined them with a data set of British individuals called the People of the British Isle Study. And this is a similar study to the Irish DNA atlas where they collected genealogical data but they only took grand parental data. Not great grand parental data. And so we used this combined data set to investigate structure both within Ireland and in Britain. So Ireland and Britain are quite interconnected so it makes sense to analyse them and them together on glass down. So a structure in Britain we already know, it's been published two years ago. Or nearly three years ago. And so this is the People of the British Isle Study. So each of these dots represents an individual placed where that ancestry comes from. And the colour and the shape of that dot is a genetic cluster that they've identified. So they've modelled these individuals as I think 17 groups of individuals. So within that group it shares more genetic data with individuals in that group than they do with individuals outside that group. And you can see that these clusters are geographically bound if I can get them out to work in. So you've got Cornwall Denner down here North and South Wales, England, Scotland, North and South West Scotland. But you can see that there's a big gap of Ireland which we sought to fill in. So we took our combined data set of 2,000 odd individuals which included about 500 Irish individuals. And we ran some kind of some basic analyses just to check where the data set was working. So there's printable component analysis that I've mentioned before which kind of generates this kind of cloud plot of dots. So individuals that are genetically related kind of fall in with each other and if they're not they kind of fall apart from each other. And you can see that here. So we've got England in orange Wales in red Scotland in blue Ireland in green and Orkney in purple. So Orkneys in north are one of the islands or sets of islands up in the north of Scotland and they're quite isolated. So that's the reason they fall away from everyone else quite so distinctly in this analysis. So what we saw was Ireland, we recapitulated what we already knew essentially. Ireland's quite close to Britain or that's despite that it's distinct and it's close as relatives of the other Celtic populations in Scotland and Wales. This data used quite simple genetic data i.e. it recorded the allele frequencies. The frequencies of these mutations in all of the individuals. But mutations aren't inherited independently. They're usually inherited as chunks that we call hypertypes. So these chunks are regions of the genome across all the auto zones like chromosome 1-22 and they're all inherited together. So if you can utilise this information how the mutations are linked together and inherited together that's more data that you can use. If you've got more data you've got more statistical power to be able to detect genetic differences between people. So we used instead haplified based data and ran the same sort of analysis. And now you start seeing groups of individuals clustering away from each other at a finer scale. So for example Wales is no longer related to or so related to Scotland and Ireland in this analysis and they've fallen down over here. You can see two broad groups starting to emerge that might be north and south ways. England is over here and then Scotland is separated into these two broad groups over here that you've mostly found over here but you've got these Irish individuals that fall over to the Scottish or English genetic space which is quite interesting. So we took this haplified data and then instead ran clustering analysis so similar to what the POV guys did we tried to model them as discrete genetic groups. So in this clustering analysis we identified 30 Irish and British clusters of these 30, 10 included significant proportions of Irish individuals ignoring the odd cluster that had one Irish individual in it. Seven of these 10 we called Gaelic Irish clusters because 99% of the individuals in there were Irish in ancestry but three of these were mixed Irish and British membership. So they were predominantly a mix of Irish, English and Scottish to varying different proportions. So those three mixed clusters are up here. We call them North Ireland clusters 1, 2 and 3 and then the Gaelic clusters are down here. So what this tree is is like clustering the individuals we cluster the clusters and we put them on branches. It's not an evolutionary tree it's just simply a demonstration of the genetic relationships between each of the clusters. But you can see that all the Gaelic Irish clusters putatively Gaelic Irish clusters performed their own branch down here. So you've got North and South Munster that fall away from everyone else, then Ulster and then Lentster, Central Ireland, Dublin and Cullan. And these names we gave to them after this analysis. This analysis is completely blind geographies, it's just purely based on genetics which sometimes you've got to keep reminding yourself because the detail that sometimes kicks out is just incredible. So this cluster these clusters are quite it's just a tree, it's not very it's not very visual, you can't really see a lot of the genetics in here. So what we did was take those individuals and map them according to where their ancestors are from. So you can see within Britain we kind of re-describe the same sort of structure that you see on the previous map of the Pobyample. We've now filled in Ireland. So you can see these different clusters and they're geographically bound as well. Predominantly these blue triangles and green circles, this is north and south Munster and they perfectly describe the boundaries of the province of Munster. You've got Ulster up in purple over here you've got Lentster down in red Central Ireland is kind of a belt cluster Dublin over here in green and Cullan in orange and then also we've got the north island clusters. The largest of which are North Ireland 2 where's my mouse? North Ireland 2 are these blue crosses and they're shown and they are shared across both the south west of Scotland but as well as the north of England over here. So suddenly we're starting to see both these clusters that are purely Irish and these clusters that are there between Ireland and Britain and they're not found just randomly across Ireland or scattered randomly, they are in specific regions and this is telling us that people in Ireland historically have been preferentially breeding with people that are proximate. There's some interesting kind of features that are hidden in here with an island so for example the county of Clare that used to be part of Cullan apparently if I've done my research right but now it's part of Munster and you see that reflected in the genote here so there are a mixture between cluster and clusters that are found in Munster as well. The level of detail on this map was just incredible when I first did it it's pretty good. It's a good day at work. So by way of showing so this all looks very dramatic so we're able to show the different regions of Ireland how different they are the differences are very very tiny the only reason we're able to detect these is because we're using really powerful techniques and we've chosen our samples really carefully so we're using techniques so your laptop may have like four processors we're running analyses that use like 64 processors over the course of a day or two there's kind of like the computational power that's required for the sort of analysis so by way of a comparison so I've put these numbers these are genetic distances measured by FST if you're interested in the estimator so for example the average FST between Ireland and England is 0.0008 so the smaller the number the closer these two populations would be that seems like just a random number by way of a comparison in Ireland, so within our Gaelic Irish cluster it's 0.002 it's nearly an order of magnitude smaller than a genetic difference between England if we compare our Gaelic Irish clusters to our mixed Irish clusters that's 0.0038 so slightly larger but still quite small so these Irish British individuals are not that genetically distinct from Gaelic Irish the difference is between Gaelic Irish but there's still, you can still take them and lastly by way of a comparison if you take the FST between Ireland and Spain that's 0.0037 so an order of magnitude larger than the difference between Ireland and England so just by way of a comparison so the structure that we see within the British Ireland is incredibly fine in the grand scheme of things but we're able still to be able to detect it so I thought it might be quite interesting to show that to show the genetic differences of the largest genetic groups and sequentially the most finest scale of differences within Ireland as a kind of animation so over here we have all our Irish individuals greyed out and they haven't been assigned a cluster yet but if we assign the individuals that are the most genetically distinct first these are the north Irish clusters then the second largest difference in Ireland is Munster away from everyone else and then after that it's Ulster and then after that it's this kind of Lentster sort of Central Ireland cluster after that we're splitting the north Ireland clusters up into their constituent parts into their individual three clusters and then we assign the last remaining individuals the Cynonix and the Dublin clusters so now we've assigned all our Irish individuals to a group so after this it's kind of just the splitting hairs part the really fine level details within Ireland so we identify the difference between north and south Munster and then we split Central Ireland and Lentster apart so maybe wondering realising that some of these clusters look awfully like sociopolitical boundaries in Irish history so I've taken 800 AD just as a by way of comparison but you can take other time periods that are available so you can see the kingdom of Munster down here kind of somewhat sorry the kingdom of Munster over here kind of mirrors the genetics of Munster over here the same with Ulster so this is kind of just some position it's not statistically significant at all but it does produce a really nice light so I thought to show so what this suggests possibly is that this genetic structure we haven't been able to date this genetic structure but this genetic structure could be as a result or could be an echo of these political groups in Irish history so that's the population genetics in Ireland or the population structure should I say in Ireland so it's how Ireland has stratified and ordered itself genetically speaking but those actual genetic groups within Ireland what are they made of so first we looked into these mixed clusters these clusters of a shared Irish and British ancestry because they were of shared Irish and British ancestry and they were in the north of Ireland so they're obviously quite interesting so they're mainly found in Ulster although some of them are found in Dublin so there are several hypotheses on how these clusters could occur they could just be a reflection of just passive gene flow between Ireland and Britain so over the millennia people have just taken a vote over the ROC or both sides of the ROC kind of mixed with the different populations there and what we're seeing here is just a reflection of a subset of individuals with Irish ancestry that are mixed with British and Irish ancestry or it could be due to specific events in history one single migration that came over from Britain or into Ireland and mixed with vice versa and then suddenly you have this ad mixed population this mixture of both Ireland and British or it could be a combination of both so we investigated that population model of ad mixture point so we used a programme called Globe Trotter and we investigated this in all three of the mixed clusters but I'm just going to discuss the results of the largest of these clusters North Ireland 2 simply because they're a little bit more robust and it was 100 individuals in North Ireland 2 compared to 30 odd individuals in Ireland North Ireland 1 and 3 so we found statistically significant results or evidence that there was an ad mixture event sometime into the north of Ireland predominantly from the north of England and this was sometime in the early 17th century more interestingly what we found was evidence of prior gene flow i.e. a mixture between Ireland and Britain into the north of Ireland so what we're seeing is that the genetics of the mixing between Ireland and Britain is very complicated you can detect these ad mixture points these single pulses of ad mixture but you have to take into account on top of that you've just got low level gene flow across Ireland and Britain historically speaking and that muddies the water so this is basically the detail that we're able to detect and past that and waters are so muddy that it was rather hard to detect anything else so another way to describe all those results just to look at the surnames we've got genealogical data so what we did was all of the great grandparents surnames and classify them according to some broad surname origins Gaelic, Gallaglar, Scottish, Anglo-Norman English, Welsh, Manx, Scandinavian and even Breton and then we found their frequency in each of the 10 clusters so the three north Ireland clusters at the top and then the Gaelic clusters at the bottom and you can see the total of the surnames that we show in each of the clusters over here so you can see that the Gaelic clusters are predominantly made up of surnames that are Gaelic in origin although interestingly Dublin has a slightly lower percentage of that slightly elevated levels of both Scottish and English surnames although you'd kind of expect that and then in north Ireland 3, 2 and 1 you have various proportions of English, Scottish and Irish surnames and interestingly that mirrors the genetic the genetic proportions of Irish, British Irish, Scottish and English ancestry so north Ireland 3 has the highest amount of Gaelic ancestry and that has has also the highest proportion of Gaelic surnames someone with north Ireland 2 and north Ireland 1 so the surnames unsurprisingly are mirroring sorry the genetic variety that we find in these clusters and these significant differences are significant so we did a simple student test student test student test it is a simple student test and the p-value is very significant basically what this means is that if you are by genetics, if you are assigned to a north Ireland cluster you are six more times likely to have an English surname somewhere in your ancestry at the great-grandparental level and if you are with the Scottish surnames roughly the same p-value well exactly the same p-value but you are 25 times more likely to have a Scottish surname in your eight great-grandparents which is remarkable so we also looked into ancestry from ancient Ireland into our modern Irish clusters so you are seeing genetic differences in different regions in Ireland but that might be because different regions in Ireland have different proportions of ancestry from different periods in Irish history so for example Munster could be distinct because they are more early farmer than Bronze Age so we used two samples which is a representative of the Irish Neolithic and the Rathen 1 which is a representative of Bronze Age so first of all the Irish farmer basically see no statistical significance between Ireland so this plot shows the higher the value, the higher amount of ancestry that we can model at the bottom Irish farmer as contributing to these populations so these values are kind of relative rather than empirical but you can see that and the Irish farmer actually shows roughly the same proportion across the British Isles which would make sense considering those heat plots that I showed right at the beginning of this presentation where in fact the Irish farmer has far more in common with the Iberian Peninsula and Sardinia than it does with Ireland within the Bronze Age we show a similar sort of pattern in the fact within Ireland so there is no real statistical difference between different regions in Ireland the populations within Britain and Ireland show the most amount of ancestry from Bronze Age which again reflects previous results so what this is showing us is that the population structure that we observe within modern Ireland isn't due to differential ancestry from different regions in Irish history this post states the Bronze Age and then finally we looked at European ancestry so we wanted to see whether different regions in Ireland or Ireland or Brooklyn have different affinities to regions within Europe so for example if a load of Spanish dropped shore ages ago into Munster we would probably see that and detect that using this analysis with a massive spike in Spanish ancestry in Munster this is kind of the idea so we modelled Ireland and Britain as a mixture of various European sources these sources are shown on this map here so these pie plots are shown in different collection centres across Europe and where they're coloured in is a particular DNA source from a genetic group in Europe so for example in France we can broadly separate it out into this kind of yellow, orange, red red, orange doesn't really come out on this just trust me orange, red and yellow which is different from this yellow in Belgium so the yellow is more the south of France the red is more north east of France and this orange is more north north-west of France Britain it yet and the same with Norway, Germany and whatever so we reported the sources the European sources that contribute at least 2.5% ancestry to any one Irish or British cluster so this means that we're reporting the major determinants of ancestry under this model so that's shown here so along the y axis you can see each individual European source that we cherry pick along the x axis each individual Irish or British cluster and they're ordered according to population so you can see this France 1 cluster contributes a lot of ancestry under this model and this France 1's cluster is that northwest France region and we think that this is just reflective of a common Celtic signature between Celtic populations of Cornwall, Wales and Scotland and Ireland in fact you can see Cornwall in this analysis because it's that one looking distinct from the rest of the English populations and as you would expect Ireland has the largest proportion of this Celtic ancestry I use Celtic as a describer not as anarchaeological what you can see also is that within these German German populations within green there's particularly low levels within Ireland compared to say England there's been more Germanic influence within England compared to Ireland interestingly we found this high level of Norwegian ancestry in Ireland particularly from using this Norway 10 signature and this signature was really similar to the signature to the finding of Orkney that Orkney has a well-documented history of Viking activity at the cultural level as well as the genetic level so we started to investigate further in this Norwegian signature and we found that there is evidence of a Norwegian signature in Ireland and we can date it it dates around the turn of the first and second millennia around the time of the Norse Vikings and actually a separate group in Trinity have confirmed these results since which is always really nice so really surprisingly compared to the considering the microemzone work earlier we found a significant proportion of Norwegian ancestry within Ireland which is something that we weren't really expecting to be honest from this analysis and so with that the most exciting results out of our paper let's just finish by summarising some additional work that we're doing at the moment a nice end of chapter of using the Irish DNA Alice but we're going to continue using the Irish DNA Alice for analysis solely using it or using it in conjunction with other data sets and other pipelines so for example using Irish DNA Alice to investigate the links between Ireland and Iceland there's only one constant between them so the Irish Atlas was used in an investigation that looked into the founding of Iceland using modern source populations as well as ancient Icelandic can't show any data because it's not published yet and so we provided the modern Irish references but what was an interesting work for the further work that we might be doing in the future is that which regions in Ireland did went over to Iceland was it from any region in Ireland or were they from a specific region in Ireland so that's what we're going to be having a look at we're always using the Atlas investigating Irish travel genetics so a load of work that I also do is on the genetics of Irish travellers and a genetic isolate found within Ireland so we've been using the Atlas for example to investigate ancestry within the Irish travellers so as a quick explanation we've at the moment managed to group the Irish travellers into four different genetic groups like those genetic clusters that I've been explaining with the Atlas and we've been comparing their ancestry to two different regions in Ireland and we find that these two clusters travel one and two predominantly have ancestry predominantly from the north and west of Ireland and these clusters have ancestry predominantly from the south and the east of Ireland and actually if you look at the languages spoken by these two groups or two groups of two their languages are associated with those different regions as well which is really cool so that's kind of what we're doing as well I hope I haven't bored you too much I'd like to acknowledge the following people JP and Sean who are my supervisors part of my PhD Seamus, Michael and Darren as well as to everyone else at the GSI that have helped with this cohort since its inception to Sophie who was a summer student that was working on the traveller data Jim, Dan, Desi who have variously helped and advised me in different analysis techniques they are a centre behind computing so that's where we do all these analysis on their big supercomputer otherwise this would have taken like 12 years and finally to Science Foundation Ireland no, not to Science Foundation Ireland for funding my PhD otherwise I would be here and then also finally to the RHD and the others participants because we won't be able to do this without thank you very much for listening