 OK, good afternoon. So thank you, Charles. I find myself sandwiched between two experts who know far more than I in telling you about what ancestry and our individual ancestries and our groups mean to our individual health. So I thought the best way that I could spend my time is to try and emphasize this message and try to tell you the kind of work that we do and others are doing and at least try to find an answer by some independent ways. I want you to note my disclosures. But the general lesson that I'm sure I unfortunately couldn't be in the morning, but I'm sure that this has been emphasized again and again and again, that if you take a number of people from around the world, essentially individuals who are native to their geographies or at least have been native to their geographies for a considerable period of time. And you look at the human genome sequence in a very broad way that it shows highly specific geographic patterns. This is, by the way, a very pioneering study by Luca Cavalli-Sforza. It was done decades ago with much more limited data before the genome was in view. But when it first appeared, and this kind of study has been done with modern methods again and again, and I'm sure the next speaker is going to speak about it, that there was a striking feature to this kind of map. This looks very much unlike a political map that we used to where the borders are sharp and separated from its neighbors. Here's a map. It's a false-colored map of genetic differences, and it shows that there are very few essentially discreet boundaries. In fact, there are no boundaries that this variation is continuous across the globe. Yes, there are distinct colors for the African continent, for parts of Asia, clearly the New World and Australia. But largely, we find continuity in the kinds of genetic variants anyone has, and in fact, their neighbors. Now, this pattern, geneticists have stayed at over the last many decades. Lots of people, both in the room and outside, have contributed to these studies. And the bottom line that we've now known for at least 20 years or more is the pattern that I showed you before is the consequence of our shared history of all humans. We are a very young species. Our history of modern humans is roughly about 100, 150,000 years. And I'm not going to show you the detail, but we started all of our ancestors started out in Africa. They roamed the globe, and some places were colonized earlier than others. Some actually very recently, many Pacific islands have been colonized only in the last several thousand years, this country itself by Native Americans about anywhere from about, say, 12,000 to 20,000 years ago. So one of the questions that modern data, having the kind of tools and technologies we have since the human genome sequence brings up is, well, how well can we reconstruct the total ancestry of humans, of anybody? As all of you know, and you've heard at this meeting that mitochondrial and Y chromosome data are very promising. They're very popular. There are many companies that provide this kind of information to us as individuals. But so far as questions of health are concerned, and I would admit that even questions of total ancestry are concerned, this is insufficient. It is imprecise in the sense it tells you something about our maternal lineage and maybe our paternal lineage. But of course, most of our genome is encoded elsewhere. And so in that sense, it's imprecise that might even be incorrect. And there were last few speakers pointed out the kind of discordancies that might appear to us. So in this day and age, as sequencing costs are falling, we could get genome sequence data from the entire genomes, from any of our genomes, and many have. And there are some technical features that can render the data almost as informative as mitochondrial and Y chromosome DNA, which is to focus on specific regions of the genome that have little to no recombination. I think we are very much there. I think we could do the first and second. And I want to emphasize this last part, that we need samples from many parts of the world. There are problems, and we'll come to them, but we need samples from populations from around the world. And this aspect is extremely crucial. And the resolution with which we can do this analysis depends on this. And we are far from it. And I'm a very big advocate, as most people in genomics are, that these samples have to be publicly available so that everybody can study them and validate them and not be proprietary in any ways. So I think these kinds of public samples have left to not only the understanding of what we call the out-of-Africa hypothesis that's the history of all humans, but of course, the history is far more complex. And I raise this complexity not to say that we cannot infer this, but to say that further study and many other samples are necessary to show that human migration and movement is, in fact, offered in two ways. Our ancestors not only left Africa, but some of them returned. And the modern genomic resolution from the sequence data enables us to do these kinds of analysis. And this is Joe Pickrell and David Reich and others in a pre-published pre-print that, in fact, suggests the scenario. On the other hand, there are other kinds of ancestry we can detect from genome sequence data. The technical word is introgression. And this is, again, a revision of the standard out-of-Africa model that is our ancestors not only left Africa, but they didn't go on to colonize the world just by themselves. They met other sort of proto-humans, the so-called archaic forms, and archaic forms, such as Neanderthals, a one that we know, another potential one called the Denisovans, and that our ancestors mixed with them. And we know this because resolution of the human genome sequence has shown in work by Svante Pavo and David Reich and many others. And some here that our current genomic sequences, in fact, carry the imprints of these DNIs that perhaps go back to 30,000 years ago or more. So when we talk about ancestry, it's not only recent ancestry that there are these imprints of much more recent, much more remote ancestry. I think, to me, much of genetics has been a surprise because there are things that I would not have anticipated. And perhaps that's the greatest part of any new and dynamic science. And these are data, again, published from the lab of our next speaker, Carlos Bustamante, and done by a very bright young disciple of his called John November, that in these analysis suggests that genes mirror geography. There's a study done in Europe. But I'm sure when the equivalent samples come from other parts of the world, the same basic story might be true that if, in fact, you did collect genetic material at very high density from around the world, you could place people, current people, into their geographies from which they most likely came to remarkable resolution. You can't see the bar chart at the bottom in blue, but it suggests that many individuals can be assigned to a region just based on their genetic marker status to under 400 kilometers. Well, we have Americans. The word kilometer is sort of strange to us. That's of the order of a couple of hundred miles. And I'm actually not quite so sure. I would want some of my relatives that close, but nevertheless. And the question is that, are these simply just so stories? Well, we are becoming better at detecting trends as the question came up during the discussion about what is the provenance of ancestry of some African Americans from Sierra Leone. We know that there are gaps in the story. That when we detect admixture, it says we know something about the specific identity of the populations that were mixed. The source populations will, of course, we don't have the source populations. They may be sometimes very good guesses, but they're guesses. Perhaps better sampling is going to remove some of the guesswork. That there has always been significant population movements. I pointed out a few of the very well-known cases. And clearly, the current resident of many individuals, residents of many individuals, doesn't say that their ancestors were precisely at the same time and place. In fact, often we know that that's not the case. But perhaps the most important part of all of human history, including that within Africa and its residents elsewhere, is that we've gone from a story of human evolution from being evolution in populations that were relatively intact, and you might use the word homogeneous, to almost all of human evolution being the history of admixture. This is true within Europe, roughly from the data that I showed you. It's clearly true in Africa. It's true in India, and it's true in the Americas as well. Our ancestors have been meeting others and intermixing for a very long period of time. Clearly within the past 5,000 years, perhaps as much as 10,000 years, not to mention the admixture between archaic and modern forms. So if we can increase the limit of this kind of resolution, as Esteban also said, the question is how much is it going to improve our understanding of human disease and its causes? And the basic hypothesis is that if the current distribution of human genetic variation is nothing more than our cumulative past history, which it is, then the same disease variants that have been carried in our genomes, of course, they not only lead to influence on our diseases, as the last speaker noted, but of course, it is going to be completely confounded or largely confounded with our ancestry. And so unraveling that is actually a very important task. This has been known in many studies. Esteban showed an example of doing spirometry in individuals and looking at their proportion of African admixture. This is similar kind of data on blood pressure, which is one of the traits that our moderator Charles Routini and I have been involved in for a very long period of time. What the graph shows on the left are individuals who have been grouped by their average blood pressure in units of 10 millimeters of mercury. And what the y-axis shows is the degree of African admixture. And what you find is as individuals, if you pick individuals in the population, these are all African-American women in a fairly large study, then what you find is individuals with higher blood pressure also have higher degrees of African admixture. This is similar to the last speaker's story. But this can't be all of it. It cannot be all genes. Also work done by Richard Cooper and others. This is sort of represented in a very large study that he and others, and I think Charles had been involved in, is to look at the prevalence of defined hypertension among six populations of West African origin. And what you find is on the left, the smallest bar are individuals studied in Nigeria. And on the right, are individuals studied in Maywood, Illinois, outside Chicago, with a number of studies from the Cameroon and the Caribbean in between, that the prevalence or the percent of hypertension in these groups vary by almost two-fold. Even though if you looked at the average amount of admixture, it clearly doesn't vary by two-fold. It varies just by a few percent. So it cannot be, although admixture or ancestry and prevalence of human disease or a disease burden is confounded, genes cannot be the only cause. So I'm going to end with a very small story again to give you what is it that we know of blood pressure, which is clearly a clinical problem of very large importance to many, many communities, including those of African-Americans in this country. There's a very large study published a couple of years ago, and I just need to point out this is taken from a meeting just over the last several days in England, and I was lazy and just picked the slides, is to say this has been a very difficult disorder and a very difficult phenotype, the phenotype of blood pressure to study. We are now making headway in identifying genetic variants that are repeatedly shown to be involved in raising blood pressure in many communities and many studies, but these are studies done in individuals of European ancestry. And much more important than the last slide is what our current evidence collectively shows. If we do have variants that we think, genes, if you will, at this level of definition, if they do impact blood pressure, then the question is that does it impact all of the other clinical sequelae, that all of the things that we worry about what elevated blood pressure does. I'm not going to go over the details, but on the top are measures of target organ damage, in this case, in the heart. This is cardiac damage. And when the studies have been done, these are studies done from the studies we did, but compared in independent studies, there's a highly significant effect. If you take even the 29 variants we know, it is a miniscule amount of the blood pressure disease burden, but it clearly has major cardiovascular sort of effect. If you look at the much more serious consequences of stroke, or paureotary disease, and heart failure, it also has major effects, even the little bit that we understand. But here is why we do science, and we do not conflate the two sources of variation that I told you. It is well known that renal disease, kidney failure, is one of the severest causes of elevated blood pressure. And in fact, it's one of the largest reasons of kidney dialysis in this country. In all individuals, including in African-Americans, and what we find is the blood pressure variants that we have explain no bit of the renal damage that individuals have. At least so far suggesting, this has now been replicated, but it's so far suggesting, that in this case, whatever the ancestry effects are, or the blood through blood pressure, at least these genes don't affect kidney damage or renal failure in a direct way, that it is elevation of blood pressure that can happen from the renal failure. Now the results, as the previous speaker clearly mentioned, the vast majority of the studies have been done in individuals of European ancestry. That's true. And so there's at least some attempt here to try and see what the results mean for other communities. I'm not going to go through the detailed result, but what the panel on the top shows is that when we take our results and look at independent communities of European ancestry, the results are highly significant. But more importantly, when we take the same genetic factors and look at East Asians or South Asians or we look at both Africans and African Americans, the results are also highly significant. So this suggests that at least from what we know today, the majority of the genetic risk factors are probably likely comparable and shared between at least many kinds of population groups and ethnicities. But because we have a lack of molecular resolution down to understanding the specific genetic variant that leads to disease, we cannot distinguish whether the differences are due to genetic or other causes. And I'm going to end with the slide that speaks to an issue that all of you know. Sickle Salonemia is a disease largely of Africa, but not exclusively. To Africa, it's found in many Mediterranean countries. It's from widely in the Saudi Arabian Peninsula. And it's found, in fact, very widely is found in one of the most common single gene defect in the Indian subcontinent. This genetic variant, although so closely allied with the history of Africa, didn't arise only once. It's the same molecular variant. It arose at least two times, if not three times, in Africa. And it likely has, or it appears very likely to have had a completely independent origin within the Arabian Peninsula, which later its carriers migrated into India. So the population distribution of sickle sal, here ancestry has helped little. It's helped in a very global way, but not locally, because there are mutations that look the same, but arose differently with different genetic properties and properties of severity. But clearly knowing its molecular nature has helped us sort out the physical identity, its conflation with ancestry, and in fact, the specific effects on severity of sickle salinemia and its relationship to malaria. So I'm going to leave you with the message that because ancestry and disease burdens are confounded, that we have to understand disease not only at the level of genes, but we really have to do so at the sequence level. And we have to do so not using surrogates of race or ethnicity as we have. We really need to identify the specific sequence changes in human disease, not simply as markers, but in aspects where we understand the path of physiology of the disease. Thank you. I want to say amen to that conclusion, because it's really where we need to be getting the right sequence. And let's speak, and final speaker for this section is Professor Carlos Bustumante from the Department of Genetics, Stanford University School of Medicine. He's going to talk to us about population genetics in the personal genome era.