 All right, so what we're going to do, we've got three speakers and then we'll have a more detailed discussion. So I'll ask that we only have any clarifying questions during the presentations by the three speakers then we'll have them come up here and as a panel be able to address questions but we are going to leave 15 minutes at the end for any questions that you have. So we'll have some time for interaction with the audience. And I'd like to introduce our first speaker, Jake Byrne, he's a scientist with Ancestry DNA. So go on Jake. I guess I can do that. Well I can say hello while we're getting my slides up. I'd like to thank all the organizers for inviting us a chance to talk to you today. I guess we probably represent the quote unquote bad guys so I'll be excited to sort of hear your thoughts and feelings about what we're doing and take that feedback and try to incorporate it in what we do going forward. Okay, so today I'm going to talk to you a little bit about both what Ancestry DNA is trying to do in terms of supporting genealogical research for African Americans and also about some of the things that I think we will be able to do in the future by synthesizing genetic data along with genealogical information. So first of all let me give you a little bit of background, what's the current state of affairs for Americans doing research and family history? Oh that might make this easier, huh? Okay so here are the top ten non-U.S. source countries contributing ancestors to U.S. customer trees and this is actually a subset of trees that I've collected for individuals who've also taken our Ancestry DNA test. And you can sort of see what the proportional breakdowns are. You've got Germany, England, Canada, etc. And I'm sure this audience, probably the first question you're asking yourself is, why are there no West African countries in this list? Well, there's a number of possible reasons for that. One obvious might be maybe the Ancestry customer base doesn't really include a lot of individuals with West African ancestry. So I looked at this. We don't actually ask anyone to self-report information about race or ethnicity but we can look at what the DNA says for those who've taken our genetic test and of about 150,000 customers who've taken the test to date, over 10,000 of them have at least 5% West African genetic ancestry from what we see. And I'll point out here that there's a huge diversity of the proportion of West African ancestry individuals have. And I think this sort of gets to some of the earlier discussion about the fact that how you self-identify and sort of racial categories in the U.S. may not in any way represent the diversity of what we see genetically. So I think there's a lot of hope for what we might be able to do with incorporating genetic data. Okay, so I hope I've demonstrated to you that we definitely have some customers with West African ancestry. So I sort of put up here another possible explanation that I think probably has a lot more to do with this, that there are few paper records that actually allow African Americans to link back to African ancestors. I think we had a really nice session earlier showing that there are tremendous records and resources if you would like to trace your ancestry within U.S., within U.S. But it's very difficult to make any kind of direct link back to Africa. So now, how can we use genetic analysis to try to enable this connection? Well, there are two sort of major ways. At least this is sort of the way I think of it. There's two major ways that genetics can empower genealogical research. The first is that DNA actually captures a lot of the history of where in the world our ancestors may have lived. And the second is that we can also identify segments of DNA that we share with distant relatives. And these segments might allow us to sort of collectively work on shared history. For the purpose of this talk, I'm going to focus more specifically on this first topic. So what does ancestry DNA do for our customers with West African ancestry today? Well, we do something that's not terribly exciting in our beta panel here. If you have genetic ancestry that comes from West Africa, we point you to this big blob and say you have ancestry from West Africa. Which is probably not terribly exciting for most people in the audience. This is something you already know. Here's what we're getting ready to launch and which I hope will bring more people and sort of get people a little more interested. We have done some work to attempt to sort of partition up West Africa into somewhat distinct genetic groups. Again, I don't want to suggest that I know anything about sort of the historical representations of these populations. But there is distinct genetic signal and I will sort of show you that. So from a genetic representation, we've been able to sort of partition West Africa into six distinct groups. And the hope is that now when you take the test, you might actually be able to link to specific regions of West Africa. You might wonder how the heck are we doing this? Ancestry DNA, I think, did a great job with its launch of building and acquiring a very large DNA sample catalog. And we are just beginning to take advantage of this. It's taking time for us to go in and genotype these samples and add them to our reference collection. A lot of this includes samples from the Sorenson Molecular Genealogy Foundation that collected samples for over a decade. But this gives you sort of a global map of our sample representation. And you can see we have a tremendous number of samples from West Africa. And so now I'm going to focus a little bit on about 400 samples that we've genotyped and added to our reference collection. And here I want to take a second. I'm going to walk you through this slide because I think it's very important. On the left, what I'm showing you is something called principal components analysis. It's basically a 2D representation of genetic variation in the samples that we've genotyped. There's no geography in the left figure, okay? So this is just a plot of sort of a collapse of the information, the genetic variation among the samples that we typed. On the right, I'm putting up a map that includes the countries from which these samples were collected. And what I want to sort of highlight as I sort of click through, I start coloring individuals in this PCA plot on the left based on the country of where these individuals were sampled. And I should also say a lot of these individuals were able to record deep pedigrees that represented at least multiple generations of ancestry from that same location. But as I fast forward through this, you'll notice that there's a clear genetic signal, again exclusively genetic on the left, that also sort of mimics what you see geographically in terms of the population structure. So I know this has come out in the discussion a lot today. Populations are not static. They move, they migrate, they mix. But there is still some genetic signal that does exist that also is sort of representative of the geography of the region. And this is how far we've sort of gone today, but I point out that there is potential future partitioning that we can do. So there's some enticing clusters in here that might suggest we could also pull apart more distinct groups. I don't want to say that I know where in Nigeria these samples are from or which tribal groups they might represent. I don't know that yet. But the idea here is that we might be able to push this a little bit further. So what can individuals expect from this particular development and future developments? Well, first of all, I've already said this. A higher resolution in West Africa hopefully means customers can gain some sense of specific parts of West Africa that contribute to their genetic makeup. Now most individuals who take this test will probably have multiple regions that light up, and that's not going to surprise anyone in the room. But I also want to say that this is really sort of only the beginning of what's possible with this data. Many of you have probably already seen this map. If not, I think it's quite an interesting representation of census data from 2010. So here this represents a US map where there's a single dot on the map for every single individual who took the census and it's colored based on their response to the question of race. And as you can see, there is still sort of clear population structure. There's still gradients of sort of where different racial groups reside, or at least people who've responded to these racial questions in a specific way. But what I also want to show here is that that structure is really complex. So here's just the map of DC. So if you zoom in on this map, and you look specifically at DC, there is clear sort of structure within where individuals are living from different groups. So again, when I first saw this map, I was really interested in the possibility of could we sort of run this map backward in time? And also, could we define our groups based on genetics, not this question of race? So to get at this question, what we're starting to do, or we're starting to explore sort of on the science side and side ancestry DNA, is what we can do with collective analysis of family trees for individuals who are enriched for a particular genetic ancestry. And here I don't want to say that they're exclusively representative of this genetic ancestry, but that they have an enrichment of it for a particular population. So here's a little schematic of what we're trying to do. The little blocks here on the left are a bunch of individuals. They're colored by some kind of enrichment of a particular genetic ancestry. We go ahead and we look at their family trees as a collective group. And then we look at the trees through time and look for birth locations in the US. We can also do it for birth locations that are abroad, and we sort of get a picture of where their ancestors come from. So this would be our background birth location image. Then what we can do is we can focus on a particular population. So individuals who are enriched for a particular ancestry. And the idea here is that when we look at their trees, we hope that they have lineages that actually represent that specific ancestry. And those individuals probably have birth locations that represent sort of how that piece of genetic material or those pieces of genetic material arrived in the US and migrated through time. So now what we can do is we can compare our sort of background locations to specific birth locations for a given population. We went ahead and did this experiment. And what we did is our background population are individuals who had between 70 and 90% West African genetic ancestry. And our specific group we're looking at are the individuals amongst that population of individuals with West African ancestry who have the most Senegalese ancestry based on that sort of cluster I defined earlier. And so what I'm showing you here is a map where I've colored in green the birth locations for individuals who have enriched Senegalese genetic ancestry, where they also have enriched birth locations. And I was really stunned to see that there is sort of a concentration of individuals' birth locations within South Carolina for those who have Senegalese ancestry. So when I first looked at this, I didn't have a particular story or historical story in mind. And then I started to Google, is there a connection between sort of the northern bit of West Africa and South Carolina? And that's when I remembered I had actually heard this story before. Dr. Kittles gave a presentation a year ago at a session where I was introduced to the story of Black Rice or sort of the Windward Coast story. So coastal South Carolina and Georgia turn out to be perfect locations for the cultivation of rice. However, the 18th century plantation owners who lived in these areas really knew nothing about growing rice. And so they preferentially imported slaves from the Windward Coast, sort of from Senegal, Sierra Leone, Gambia, because these were places where rice was commonly grown. These individuals would have expertise in growing rice. And as, you know, again, probably the audience here knows far better than I do, there's a story about the Gullah people who currently live in coastal South Carolina and Georgia who are thought to sort of descend from this group of slaves that were brought from the Windward Coast. And so I was quite fascinated that taking the family tree information and the genetic signal, I was able to sort of see some kind of suggestion of this historical story in the genetic and family tree data coupled together. So just briefly, conclusions. There's few paper records existing to link African-American individuals back to African ancestors. I think that's known. Genetic analysis hopefully can provide a path to make this connection possible. And AncestryDNA is working hard to help individuals be able to link to specific locations within West Africa. And also sort of this final story that I think using collective information shared across family trees might allow us to sort of highlight important historical stories. And this is something that we're very, very interested in doing. Finally, just acknowledge the team of individuals involved in this work and just say thank you. And I guess I can take a clarification question if need be. And then we can move on to the next speaker. Referred to genetic signatures. What specifically are those signatures comprised of? Are these gene frequencies? What is a genetic specifically signature? In this specific instance, we're looking at allele frequencies. So we're looking at a large panel of sites across the genome. And it's simply the pattern of allele frequencies. And that's not to say, sort of to your point earlier, Dr. Blakey, that these are single isolated markers that individually distinguish a population. So it's sort of the collective information and frequency differences that we see that give us some kind of information about sort of the geographic origins of individuals. Just a comment on the history end of it. You're using late 19th century states. Senegal did not exist at the time of the slave trade. There were kingdoms. And those kingdoms were heterogeneous. People moved in and out of them. There were lots of language groups. Mali is known to have been connected to Morocco and Portugal and Spain. I think you've got to be really careful about locating this stuff ahistorically or in the contemporary period and giving people the idea. I mean, the Gullah people come from Sierra Leone. They're in the same large language family as some of the people in Senegal, but not all. So I think that I'm a little disturbed by sort of going, oh, and you're from the camera room. What does that mean? There wasn't a camera room. Yes, I understand that. And I should say, the individual samples that we have used, what we have as an initial sort of definition of where they're from is simply a current country boundary. And that's absolutely true. So I'm trying to give some context, but you're absolutely right. I need to be careful about how I do that. Yeah. Thanks a lot, Jake. Next I'd like to introduce Joanna Mountain. She's the Senior Scientific Director at 23andMe.