 I want to welcome you all to the first Grand Rounds of this academic year. I hope you all had a bang-up summer like I did, water skiing, in case anybody's interested. As you all probably know, I hope you know that in July we marked the 60th anniversary of the opening of the clinical center. And what we've decided to do is that for this year's Grand Rounds schedule to develop a program that highlights accomplishments that are really current and past that typify what's going on here at the clinical center over these 60 years. And so I hope you'll be excited about the program that we have coming. And I think today's presentation is a remarkable example of the quality of the, and excitement of the science, the clinical science that you're going to hear. So we have two speakers today who are going to talk about genomics in medicine. Our first speaker is Dr. Eric Green, who I think you all know, you should know. He's the director of the National Human Genome Research Institute in a position that he's held since late in 2009. And his talk today is titled The Decade of Human Genome Project, Unraveling the Genetic Basis of Human Disease. Eric received his BS degree in bacteriology from the University of Wisconsin in Madison in 1981, and his MD and PhD degrees from Washington University in 1987. He completed his residency in laboratory medicine in 1992, and also a postdoctoral research fellowship at the University, Washington University, which he did from 1988 to 1992. In 1992, he's appointed assistant professor of pathology and genetics, as well as a co-investigator in the Human Genome Center in Washington University, and in 1994 he moved to Bethesda into the newly established intramural research program at the then National Center for Human Genome Research at the NIH, which later was named the National Human Genome Research Institute, and he was head of the physical mapping section. Later, he also served as the chief of the institute's genome technology branch and director of the NIH Intramural Sequencing Center, and he served as a scientific director for the institute from 2002 to 2009. While directing an independent research program for almost two decades, Dr. Green was at the forefront of the efforts to map sequence and understand eukaryotic genome. His work included significant involvement in the human genome project, and these initial efforts later blossomed into a highly productive program and comparative genomics that provided important insights into genome structure, function, and evolution. He's received many awards, including the Helen Hay Whitney Postdoctoral Fellowship Award, the Alumni Achievement Award from his Alamata, Washington University, and the Cutlow Lectureship Award from the Academy of Clinical Laboratory Physicians and Scientists, as well as the Wallace H. Coulter Lectureship Award from the American Association of Clinical Chemistry, and he's been elected to many societies, including the American Society of Clinical Investigation and the Association of American Physicians, and he's been founding editor of the Journal of Genomic Research and a series editor for genome analysis, a laboratory manual. And now, let's welcome Dr. Eric Green to the podium. Well, thank you, John. It's a pleasure to be here. By the way, Karen and I were deciding who were going to go first, and the deal was I would arm wrestle John, and you can see I broke his arm, and that's why I got to go first. No. In any case, actually, it's a pleasure to be here and kicking off this special series commemorating this important anniversary of the Clinical Center. By way of introduction, I'm an institute director, which means I'm financially boring, and number two, the objectives that we're going to... You can read them as I give a couple of introductory remarks. I've been involved, as you've heard, in genomics for about 25 years now, starting as a recently graduated MD-PhD student in a resident in clinical pathology. But I think the relevance of genomics has really changed, and I would predict that over the course of the next year, when you hear Grand Rounds as part of this year-long celebration, you're going to get a flavor for how much genomics is finding its way into Grand Rounds here at the Clinical Center. You know, when I started in genomics, it was mostly relevant to biomedical researchers, people who were sort of thinking about how to map and sequence the human genome. I think as the genome project advanced, increasingly we imagine and envision clinical applications, and healthcare professionals got interested in genomics. But I think the thing that's happening now, and will absolutely happen before the end of the decade, and you'll hear it at these Grand Rounds, is that genomics is getting relevant to patients and friends and relatives and patients, which means it's all of us. Now what, of course, started all of this was a human genome project, which began in October 1990, and then about 13 years later, ended. In fact, this year, we celebrated the 10th anniversary of completing the human genome project. Now 10 years ago, when the genome project ended, the institute that I now direct and have been at for 19 years, NHGRI, certainly had to think about what they were going to do next from the point of view of having been created by the U.S. Congress to lead the U.S.'s effort in the human genome project, and having been successful actually two years ahead of schedule, what was going to be the main challenges in genomics that the institute pursue, should pursue in its research agenda. And I would say there's many things that I'm going to describe some of them to you. But my view of the world, as it pertains to genomics, especially now as director of the institute, is much of what we should be doing is to facilitate the application of genomics to medicine. The term I like for this is genomic medicine, largely synonymous with personalized medicine, individualized medicine, precision medicine, and so forth. But I regard this as an emerging medical discipline that involves using an individual's genomic information as part of their clinical care. And my view, as my responsibility as the director of NHGRI, is to really chart a path, if you will, leading to the large-scale implementation of genomics as part of medicine, a path that will begin with the completion of the Human Genome Project ten years ago, and eventually will result in the realization of genomic medicine broadly defined. And it's going to be a marathon. It's going to be a long journey that's going to involve many, many steps. I don't even know what all those steps are. I think this journey started quite optimistically, having been wildly successful, completing the Human Genome Project, and simply believing we're going to be successful in finding important ways genomics can be used for improving health care. And when we are successful at this, I think we'll fulfill the promise of why we sequenced the Human Genome in the first place. So in thinking about this journey, from base pairs to bedside, or if you prefer the metaphor, from helix to health, what NHGRI did the day the genome project ended was to publish a strategic vision of what was going to be needed. It was actually a good strategic vision for 2003, but in fact, things happened far faster than any of us anticipated, even some of the most optimistic of us. And in fact, this didn't even last ten years, the strategic vision. We found by about 2010 we needed to publish a new one. So by 2011, we published a new strategic vision that sort of laid out a blueprint of what needed to happen in going from the base pairs of the genome project to the bedside of patients. So what I want to do in relatively rapid form is to give you some updates of what this journey has been like over the last ten years that in many ways is bringing genomic medicine into focus, which is sort of a theme of my talk. And I'm going to talk about advances in interpreting the human genome sequence, a first step, developing new DNA sequencing methods as a key second step, cataloging variation that exists among all of our genomes, and then applying that information to understanding the genomic basis disease. And I think these four areas of advancement really capture the major things that have happened over the last ten years that are putting us on a very nice trajectory towards realizing genomic medicine. And I think it will nicely set up the second talk where you hear about somehow some of this knowledge then leads to new opportunities to unravel the molecular and cellular basis of a particular genetic disease. So starting with the interpretation of the human genome sequence, let me remind you that what the genome project produced was this, just simply the sequence of the roughly three billion letters that make up the human genetic blueprint. It didn't involve an interpretation of that. We knew that would take decades of additional work. A lot has happened in the first ten years of having in front of us these letters. We've done a pretty good job at cataloging all of the human genes, the protein coding genes, roughly 20,000 genes we now know, and we know their locations, and we can highlight those within the sequence. But what was a bit of a surprise, and there's been many surprises in the early years of interpreting the human genome sequence, is that if you take all the letters that directly code for protein, they're a minority of all the letters that are functionally important. In fact, there's an additional set of letters. All the protein coding genes comprise maybe about 1.5% of the three billion letters. We now know that an additional 3.5% are conserved to the same degree as protein coding genes and are conserved across virtually all mammalian species and function in ways other than directly coding for proteins. Many of these are involved in regulating where, when, and how much genes are turned on, regulatory sequences, and others are involved doing other functional things that don't involve directly coding for proteins. There have been a lot of surprises as we've cataloged all of those functional sequences. Another surprise is the recognition, in particular over the last five or six years, that there's a lot of biology and a lot of information encoded in addition to the primary sequence of our genome. Also, the various decorations that occur to our DNA, these epigenomic changes, various groups that get attached or associated with the DNA that confer a lot of biological information. Some of you might have heard about a very large project we've had ongoing for a number of years called ENCODE, that NHGRI runs as an international collaboration short for encyclopedia of DNA elements that roughly aims to catalog and eventually better understand all the functional elements within the human genome, both gene regions and functional non-coding regions as well. Where are we 10 years into the first interpretation, if you will, of the human genome sequence? Well, I would tell you at best, we certainly understand a lot, but at best, it's a cliff-notes view of the genome. If those of you who remember cliff-notes, a little bit superficial, maybe get you ready for the first test, but it's not a deep line of understanding that we're eventually going to need. We will be interpreting and reinterpreting the human genome sequence for decades to come, like fine literature. It's going to take a lot of refinement of our understanding. So that was the first initial step, but we also recognize that we weren't just interested in having in hand one sequence. We're going to want to have lots of genome sequences. And in fact, we need powerful new methods for doing the kinds of studies one can imagine to fully understand the human genome and how it works, and then understand how differences in our genomes dictate various features that are important for human health. And so developing new methods for sequencing, of course, became a priority area. The way we sequenced the genome in the first place were with factories that looked something like this. And while they got the job done the first time, they were not exactly the kind of clinical laboratory one might imagine that would be needed for doing a clinical test that would involve doing genome interrogation of individual patients. We needed some fancy new method graphically shown here or in an iconic way that would somehow allow us to vary inexpensively sequence genomes at the cost that would allow it to be a diagnostic test. In fact, 10 years ago when the strategic plan following the completion of the genome project was published, NHGRI called for the development of new technologies that would allow you to sequence a human genome for $1,000. It sounded like a good clinical test, $1,000. And we thought one day we would reach it. I think we were all surprised at how effective the program that we put out at our Institute for Developing Technologies would be met by private sector investment and how quickly good ideas would come. And some of you might be familiar with a graph that we frequently show, which simply depicts the cost of sequencing a human genome on a logarithmic scale. And you can see when these new technologies that started to become available right around there started to be used, the cost of sequencing DNA precipitously dropped, and in fact, exceeded Moore's law, which is the law of the computer industry that says that everything doubles about every 24 months. And in fact, the cost went down at a faster pace than that. What does that mean in terms of real dollars and in terms of how quickly you can sequence a genome? Well, let me remind you, we sequenced that first human genome, and it took about six to eight years, of thousands of investigators around the world working in factory-like form to sequence that first human genome, and it cost something like a billion dollars. And we were talking about wanting to reduce this down to $1,000. Well, in fact, the day the human genome project ended, we estimated, back of the envelope calculation, that if we went to sequence a second human genome, we could probably do it with those factories in about three to four months. But it would still cost $10 to $50 million. Fast forward to today, there's little devices, such as that shown here. And I have one in my hand on little chips that go into little desktop or benchtop instruments in a matter of two or three days, and probably by the end of this calendar year, it'll be down to about a day. And for a cost that's not quite $1,000, but it's under $10,000, you can get a whole genome sequence. And there's various shortcuts that could just get you the Kodi regions for under $1,000. So we are very, very close to getting down to $1,000 genome. In fact, I don't think that's the grand challenge anymore in genomics by any means. It's not the cost of generating sequence data. It's actually the cost of actually interpreting all of it. So that's been another success story. First, we have better interpretation of the human genome, and now we have better technologies for reading out the human genomes. The third major advance that's worth emphasizing is our ability to catalog human genomic variation. And the reason this is so important is we're not just interested in how a hypothetical human genome works, as we made a reference sequence by the human genome project, but rather, we're very interested in how our genomes and our patient's genomes individually work. And in fact, of course, just thinking about it, each of us has two genomes in us, one for mom, one for dad. It's about 6 billion letters that make up our genome. And any one of us differs compared to any another person in this room. About one out of 1,000 bases will be a different letter at that particular point. So you can decorate the DNA. And if you add that up, it's about 3 to 5 million variants, single nucleotide variants in any patient's genome, or any person's genome. Across our 6 billion, they differ about 3 to 5 million places. In addition, there's several tens of thousands of places where there's some structural changes, where maybe there's a deletion or insertion or rearrangement. That'll be very relevant for the second talk you're going to hear about, a very specific region of the genome that has structural rearrangements associated with disease. But the vast majority of these variants, especially the single nucleotide ones, have really no phenotypic or clinical consequences. But a small subset of them do. Some of those have detrimental consequences, such as conferring risk for a disease. And some of them might have good consequences, such as protecting more from a disease or another positive attribute. So a big priority in genomics over the last decade was to catalog as many of these variants, especially the more common variants that exist in multiple people in this room, and have those available for scientists to then study to figure out if they're good variants or bad variants or whatever kind of variants or meaningless variants. So you might have heard about projects like the SNP consortium, Single Nucleotide Polymorphism Consortium, which gave rise to the HATMAP project, which became an international effort then to catalog some of the most common variants across the world. And that then gave rise when new technology became available to the 1000 Genomes Project, which initially aimed to study a thousand people, but actually has blossomed into about 2,500 people, all selected from different populations across the globe to get as much global genomic diversity as possible into public databases. And in a series of publications, and another one will come out this year, more and more of these genomes have been analyzed. And with that, we've gained tremendous insights about the most common variants that exist. So for example, when the Genome Project began, we knew of about 5,000 variants that existed somewhere in the population. Now that number is over 50 million and growing. And so some of these are the most common variants that exist in different human populations, and that allows investigators to now develop large-scale studies to figure out which of those variants might be associated or causative with specific diseases. And so that third level of accomplishment then set up this fourth step, if you will. And that's now getting to the more disease-oriented research, elucidating the genomic basis for human disease. Now, to put this in context so that you appreciate what accomplishments have happened over the last 10 years, let me remind you that virtually all diseases have either a virginity cause, or at least a genetic influence, but they differ with respect to frequency and they differ with respect to the sort of the underlying causation. And you can divide it into two very broad categories. On the one hand, there are rare genetic diseases, like the one you'll hear about in the next talk, but also cystic fibrosis, sickle cell disease, hunting disease, and so forth. These are rare, but they're genetically simple because they involve just defects in a single gene. They're the dominant risk for getting the disease. These are known as Mendelian disorders. Now, it might be that other variants somehow influence the disease and maybe the environment contributes a little, but by and large, defects in a single gene cause the disease. But these are rare, and they're not what fell hospitals and clinics around the world. What fell hospitals and clinics around the world are common diseases, diseases like hypertension, asthma, diabetes, Alzheimer's, autism, and so forth. Common diseases are important for the reason that they're common, but unfortunately, they're genetically complex. They're also known as complex diseases because it's not a single variant that confers the risk, but it's typically a series of variants that conspire together what is typically a greater contribution of the environment to providing overall disease risk. The question was, and always was expected, how would advances in knowledge of the genome, how it works, and knowledge of genetic variants help us unravel the genetic basis of rare diseases and common diseases? There was great optimism here, and there was lots of questions here. What's transpired in the past 10 years and even before that? Well, let's start with simple, single gene disorders and graphed here as a histogram showing genes or monogenic diseases and traits for which the gene was found, the genomic basis was defined. Let me remind you, right there is where the genome project began. The day the genome project started, there were only 61 diseases or traits for which we knew what the specific gene underlined that disease or trait was. 61, and you can't argue that as soon as the genome project began, we increased the pace at which we discovered the genetic basis of disease, and you can see as of the end of last year that number was almost up to 5,000. And in fact, I will now tell you a major priority area in genomics and human genetics is using these new methods of genome sequencing to study rare diseases. In particular, the 4,800 that I showed on the previous slide, this is now all of the diseases for which we know the genomic basis, and that's great, that's the glass half full. It turns out there's still about another 2,000 diseases for which we don't yet know the genomic basis and another 2,000 beyond that, that we think a single gene is involved and don't yet know the genomic basis. Filling in this pie chart is a high priority area in human genomics. It's a reason why at NHGRI we launched a program a little over, just shy of two years ago where we are now funding several centers to industrialize, if you will, the process of taking these remaining rare diseases for which we don't yet know the defective gene and sequence, sequence, sequence, and interpret that data and try to figure out exactly what the underlying genomic cause of it. This program is off and running. I looked this morning. There's about 600 diseases that they're now studying across these three centers, over 100 success stories so far where they figured out a defective gene and off they're going and we hope to fill in the rest of the pie chart along with other groups around the world who are similarly pushing the accelerator, if you will, to understand the molecular basis of these remaining rare diseases. I would also point out, I can't imagine standing in this auditorium and not talking about the undiagnosed diseases program which similarly is taking patients in here with undiagnosed disorders and in many cases discovering using genomic approaches, new diseases, and great credit to Bill Gaul and all of his folks who have led that program which is now actually scaling up because of its success and all of these will help fill in that pie chart, if you will. And you'll hear in the next talk about just when you find that genomic basis for a given disease, it's really once again just the starting line because it doesn't tell you everything you need to know and in fact the story you'll hear after mine is how just knowing the genomic base of a disease then sets up all sorts of important studies to really get at the underlying mechanism that leads to the disease process. So those are rare diseases. What about common diseases? Well here, this is a story in and of itself and I don't have time to describe in detail but I hopefully many of you have heard about genome-wide association studies where basically you take large sets of patients with and without a genetic disease like hypertension or diabetes and you use knowledge about common variants that have been cataloged by step number three I introduced you to earlier and you scan across all these people's genomes, thousands of them for these common variants and simply look for correlations between the inheritance of certain variants and getting the disorder compared to control groups. And you do this in a genome-wide basis and you're looking for statistical association and if you're fortunate you find peaks such as here that point to regions of the genome that seem to have a variant that confers risk for getting the disease. And what I will tell you is there was lots of questions about whether these were gonna be successful or not, whether we'd be able to drill down from the complexities of these common diseases and get down to specific variants and be able to show how they confer risk for disease. And in 2008 was the first publication of successful genome-wide association study and as of today, 1,900 papers have described this and we now have multiple, literally hundreds of regions of the human genome have now been implicated in literally dozens and dozens of the most important common diseases that are out there. And needless to say we are now going to the next stage of going from having regions of the genome associated with a particular disease like autism or Alzheimer's or diabetes or asthma indicated here by these little colored circles to now drilling down in much larger studies to figure out which variants sit in each of those regions and which of those might be the ones actually conferring risk for disease. We're early days in this, but I would predict between now and the end of the decade we will learn a lot more about and it will be complicated. I'm sure it will take many years beyond that but we will learn a lot more about variants that are specifically conferring risk for some of the most important common diseases that exists. There's one other thing I should point out about this before I start to wrap up is that I don't wanna in any way portray that any of this is simpler that it's not even gonna get more complicated. I'll give you two examples. For example, what we have learned increasingly as we've understood more about the genomic architecture of genetic diseases is that the great majority of the time when you have a single gene defect causing a rare disease it is a genetic variant that disrupts the coding portions of the genome, the protein coding portions, the yellow stuff, the stuff we understand pretty well because we understand how DNA makes RNA and how RNA makes protein. The exact opposite is turning out to be the case for common diseases where not all of them but the majority of cases in these genome wide association studies, the regions that are statistically implicated as having variants conferring risk for things like diabetes and hypertension and so forth are falling in portions of the genome that don't directly code for protein. Probably involved in regulating genes and doing other things but it's the part of the genome we don't understand that well yet and we need a lot more to learn but it's actually the regions that are gonna be the most clinically important. And the other sobering thing I should tell you because I don't want my optimism to in any way convey the impression that all this is simple and it's just a straight forward. There are so many other bottlenecks we're facing. You know the truth of the matter is we take these fancy sequencing methods and in a day or two I could sequence any one of your genomes and I could even go through all the computational steps which aren't simple to at least catalog all of your three to five million variants but there's not a genome scientist on this planet right now that could look at most of those variants and really feel confident they know variant by variant which of those is relevant and which one is not. And yes, we're sequencing lots of genomes even in the clinical center we're sequencing genomes of patients including patients in the undiagnosed diseases program but the truth of the matter is when we look at those genome sequences for much of many of the variants that we see we round on the patients we're not totally sure yet what it all means. Our ability to go from genotype to phenotype is still sort of very early days and it's still very hard. Many things we're thinking about how to improve it but for most variants we don't instantly know which ones are biologically relevant let alone clinically relevant. So in closing let me just tell you that I think what's happening and maybe it's why I was invited to give this opening talk for this series is genomic medicine truly is getting into focus. I think when the genome project began we all said that someday this would be clinically relevant but I think that vision was quite a bit blurry. Maybe when the genome project ended it started to get a little more clear we were able to say oh yeah, yeah we got the genome sequence now we could make this clinically relevant and even when we published our strategic vision in 2011 I think it was much clearer but still far from perfectly clear but I am optimistic. I think for many specific examples the clarity will come by the end of the decade as we start to really use these powerful new genomic approaches to improve healthcare. And lastly, since I have at least one minute I wanna put in a shameless plug but it is relevant because one of the things we're gonna have to do if genomic medicine is gonna be a reality is we need to make sure all of you understand aspects of genomics as practicing healthcare professionals but the other thing we're gonna need to do is we're gonna need to ready the general public to genomics and the language that might be important as you talk to them about these applications and so genomic literacy assigned we also care about at NHGRI and have various programs to try to improve genomic literacy but one of the things we did it was one step was to create in partnership with the Smithsonian a new genomics exhibition we did this to commemorate both the 60th anniversary of Watson Crick's discovery of the double helical structure of DNA that's we're celebrating this year and the 10th anniversary of the completion of the human genome project and so we worked for two years with the Smithsonian and then earlier this summer we opened this exhibition and the National Museum of Natural History it's a nice sized exhibition it's an easy remember it's Hall 23 we have 23 pairs of chromosomes it's easy to remember right and actually it's also easy the number one visited thing at the national the largest natural history museum in the world most visited is the Hope Diamond four million people see the Hope Diamond every year and if you just go to the Hope Diamond and take a left you come right into our exhibition so hopefully you get lost and move away from the Hope Diamond and come see us it's gonna be there for a year so you have from now until next September 1st and if you don't get down there and see it and then you have to go somewhere else in North America because it's gonna tour then for four to five years for two years we worked very hard in designing this a wonderful partnership between members of our institute and members of the Smithsonian and we came up with beautiful looking graphics of what this was gonna look like and then earlier this summer it became a reality and this is just some eye candy that hopefully entices you to please come and visit this exhibition and bring your children and bring your neighbors and bring all people that you know are gonna be patient someday because it'll be relevant for them so in closing I'll just say please go down there and visit and hopefully you will enjoy it thank you for your attention