 I think every meeting that I can remember, certainly everyone that was involved in organizing, we asked, who should we get to sum up? And the answer has always been Maynard Olson, no question, and we did it again, and this time once again Maynard agreed, and we're really happy to have him here. Maynard? Well, thanks Mark. It's tough to be a closer, you just throw a few pitches and then go back and sit in the dugout. This is a really challenging occasion because genomics has clearly come of age. I'm really just going to make a few relatively philosophical comments about how I see the path ahead in genomics and Amy actually laid the groundwork for my comments because just to anticipate where I'm headed, I believe that looking long term, I'm really not talking about what should be done in 2014, but looking out 5, 10, 20 years which is the timescale that the new strategic plan adopted, I believe that genomics must move toward what I would call radical integration with medicine on the one hand and areas of basic science, basic biomedical research on the other. So you've all seen this graph repeatedly all I'll really say is that we are at an inflection point and for me the message I take from this enormous uptake in our capacity to acquire sequence data and related genomic data is that the potential is finally here to move genomics out into the real world and that's the challenge that I want to address. I'd like to root genomics as a quote from George Gamow's 1954 paper and I believe this was the beginning of genomics as an explicit challenge for biomedical research. This paper is mostly known for having an incorrect model for the genetic code, but it was the first paper commenting on the Watson Crick discovery that pointed out quite clearly that biological information is digital and that the hereditary properties of any given organism can be characterized as a long base four number. Indeed if the chemists had been in a slightly more subsidiary role and the bioinformaticians more ascendant in that time we would probably read DNA sequence the way it is on the second line as opposed to the first. The Human Genome Project was about finding out what our number is and what the numbers are for other organisms and we've heard today that tremendous progress has been built on that to look at variation in these numbers and so forth. But I just want to reflect briefly as a foundation for my later comments on what for lack of a better term I'll call the platonic idealism of this view of the genome. Biology is messy that tissues are squishy, the cells all look odd, they change as you're looking at them. It's hard to study. Who would have guessed 60 years ago that underlying this sort of messiest edifice of complexity in the whole natural world lay digital information. You go and look at the literature of that day and the hypothesis is not even there. This was a discovery that had simply not been anticipated. Now we've heard about epigenetics for example and all of the complexities that are quickly piled upon this nice clean message. So someone will tell us well it's not really digital information there are asterisks on some of the seas or metal groups or whatever that the situation gets more complex. Of course yes it gets more complex but that's the point is that genomics in its first generation its sort of basic driver head behind it the most powerful idea in biology and that is that biological information is digital. As we move on from there genomics must look more like the rest of biology and the rest of medicine. No longer can it operate in a bubble of this platonic idealism. I would argue that genomics over its first decades really has largely been operating in a bubble. Now this activity did as Amy pointed out to us impact on a lot of people's lives. But a lot of people are really a very select group that fell into quaternary care centers and were able to benefit usually because they have quite rare conditions. The challenges have had quite obviously deal with mainstream medicine the millions the tens of millions of people caught up in the economics of medicine and also it's essentially bottom with biological complexity. It's easy to find if you look about certainly if you were taking careful notes today or look about in the literature indications that genomicists have been inside this bubble perhaps already a bit too long. I don't even identify the individual who gave this quote in a recent tenure retrospective because my point is not to pick up on a particular person. Comments of this type are ubiquitous in the scientific literature and certainly in the public communications that come from our community. So epigenetics we're going to talk in a minute about genetic exceptionalism but for the moment we're now talking about epigenetic exceptionalism. Epigenetics may provide hope that we are more than just a sequence of our genes certainly we all hope that indeed we all know that and that our destiny and that of our children can be shaped to some extent by our lifestyle and our environment. What do we think that our lives are shaped by? If not by our lifestyle our environment the choices that we make as human beings we need to move away from the genetic exceptionalism which has driven this field and I would actually argue served it reasonably well in its infancy. That's the past we're talking now about the future. So genomics in the real world this is going to be an uncomfortable place in which to do genomics but let's get used to it and learn how to do it. What will it involve? Well a few years ago I got in some difficulty with my LC colleagues by showing this slide about what I referred to as the narrow squeezed pipeline between basically the science of genomics and sort of better health novel therapies and so forth. I understand the backlash that I wanted to defend the slide. There's a comma between narrow and squeezed. This pipeline is narrow that's the fundamental point about it. It is also squeezed by many concerns often rooted in genetic exceptionalist ideas we need to broaden this pipeline dramatically and squeeze it less. The real acts are going to be difficult but they really are the challenge in front of us. Our message to the public at present is almost completely wrong everywhere you look. Yes we need to get out and talk to people, real people but we also need to start thinking more carefully about what we're going to say to them and again my point here is not to pick on Vanderbilt. It's actually a relatively benign advertisement. There is underneath it actually some scientific findings but it's an advertisement. Watch Mad Men. What does someone take away from this? They're intended to take away the idea that in your genome is the secret to protection against heart attacks and stroke. The reality being that we have almost nothing to offer in this area. We have underlie this particular application are essentially trivial in their clinical impact. Haven't been tested actually to see if they're efficacious but even if so play such a minor role in the management of patients who are at risk for heart attacks and stroke which are essentially all of us. But the message here is just not the one that we need to go forward with. So I'm actually going to draw my main inspiration here perhaps repair my relationships with the LC community by drawing on an idea which I think the LC community can genuinely take credit for. It certainly didn't come from the scientists. When I first heard LC researchers talking about genetic exceptionalism I thought that they were off the mark but that was then and this is now. I've changed but the field has also changed dramatically. As near as I can tell the phrase genetic exceptionalism was first used in the literature in this chapter by Tom Murray in a book edited by Mark Rothstein. But it bubbled up out of many discussions going on at that time. I know Glenn McGee for example at about the same time and I think independently I began writing about genetic exceptionalism. And there is now actually quite a rich literature on the subject. Not all of which is particularly aligned with the position I'm going to take today but that's what a rich dialogue is about. A complicated issue. So Tom knew why there were special concerns about genetic information. It's essentially the idea that for example in a medical record there's something really special about genetic data as opposed to all the other information that's in there. And that we must be particularly cautious, have particularly strong protections and so forth. So why did people think this? Well Tom explained why people think this. It wasn't actually irrational idea. In this paper he points out that the core roots of genetic exceptionalism were concerned about genetic prophecy. We've heard a lot about that. Jim didn't want to know about the APOE4 locust. Concerned for PIN. Genetic information has implications not just for the person who's been genotype but for close relatives. Concerned about genetic discrimination. It's a dominant issue throughout the genome project. A major focus of the LC program has been addressed to some degree by legislation. But indeed some of that very legislation is going to make the task going forward more difficult. Which was what Tom and others were warning us about. The more that we put genetic information in some special category and start enacting laws differently than we treat other clinical information, the more difficult it is going to be to achieve the integration that we now need. And there's a concern about enabling certain identification of individuals. An awkward phrase but something that came up repeatedly. And the last point I will just say that I think this is actually the most serious issue. On a time scale of 10 or 20 or 30 years we need to think a bit about this. And I'll just say in passing that I think the core problem here involves whether it will be technically feasible to separate forensic and health care information about individuals. This was rather easy to do when the FBI was off genotyping a few selected mutation driven loci and medically oriented geneticists were looking at other things. But obviously in the era of whole genome sequencing and that is the era that is relevant to my discussion there are going to be challenges there. It may be the most serious challenge. But as Tom and many others such as this I think also very early pioneering paper by Gaston and Hodge pointed out that we need to weigh these concerns stated on the previous slide with a concern for the public good of science. What Amy is telling us is that if we expect this enterprise to prosper and to sustain the generally good will that the public cares for it at the moment we need to address the challenge of how is it that we're going to integrate the information that we can now so readily gather into the real world of medicine. So a standard answer to that question which I would argue is a kind of 20th century view of the problem. Not a bad view one that certainly has had major benefits for medicine. I said well we're going to do a lot of prospective studies of this and that with tens hundreds thousands of individuals to look at the predictive value of known markers and so forth. I think the briefest reflection on this idea shows that whatever we're going to do it's not going to be this. Think about the number of different genotypes that are potentially of interest for their predictive value in medicine. It is simply infeasible to imagine that we're going to take every proposed genetic test whether it's mobilized by a direct-to-consumer company or integrated into health care or whatever and do some kind of prospective longitudinal study to see how wise it is to use that test return results to patients and affect clinical decision making with it. We need some other approach, a more 21st century approach. It's going to be our challenge to develop it. In the strategic plan one quote that really caught me I will say that I believe the strategic plan was a far more realistic document on these points than any I've read before. The particular quote interested me. Correlation of genomic information with high quality phenotypic data will benefit greatly from the linkage of genomic information to data gathered in the course of actual clinical care such as in electronic medical health records. I think there is no question that this is our future. The question is how long is it going to take to get there what is the path, what is the appropriate role for the NIH and for all of us. Lurching in our characteristic American way as we may be we are gradually headed for electronic medical records that one can actually compute on and transfer across institutions and so forth. We are a long ways from that goal as anyone who looks into this issue will tell you but it is a goal and indeed was clearly recognized during the long health care reform debates that if actually were ever going to drive down the cost of medical care without compromising its quality we need dramatically better outcomes research on all of the procedures that we do and this is the only way that we are going to do it because in the real world of medicine the prospective clinical trial also plays an important but very narrow role you just simply can't particularly in a world of rapidly changing care practices and context for medicine retesting everything that way. You do retrospective or observational outcomes research and if we are really going to do it in our population we are going to have to compute across medical records that's true whether we have any genetic data in these records or not however it doesn't take much sort of economic analysis to recognize the implications of the collapsing cost of gathering let's say whole genome sequence data. In this context what we compare it against is the cost of clinical care you know not the GMS budget or the NHGRI budget but the cost of clinical care we are rapidly approaching the time when gathering a huge amount of molecular data will add trivially to the cost of clinical care. We cannot be the ones out of research budgets that are paying for endless clinical coordinators and people to call up patients on the telephone and remind them to come in and someone to draw blood and so forth. This is going on already on a massive scale and if we are going to succeed in integrating medicine and genomics we are going to have to tap into that all of the difficulties of doing so notwithstanding. So my message here really is that to capture this bright future we need a new message and we need some path out of the bubble and into the real world. I've touched on a few of the obstacles to getting there. Many of them are right in the heart of the ELSI agenda and yes we need to get out of the lab and talk to real people but we also need closer engagement with the ELSI community of the type that Amy McGuire discussed because this is now a partnership which is going to succeed or fail together. So I just want to close with a word about basic science focused on translational activities of major focus of the strategic plan that doesn't tend to go deeply into the issues surrounding basic science applications of genomics. I'll just share with you an insight that I had as I thought about what I would say on this topic if I had more time and really it comes back to genetic exceptionalism it's essentially the same problem. Biology is complicated, it's extremely complicated genomics and it's inside the bubble sort of platonic form relatively simple. I'd like to say that sequencing the genome gave biologists their first taste of complete knowledge. It's also going to be their last taste of complete knowledge. Biology just isn't like that. If you read the history of biology you'll see that it's a history of successive waves of frontal attacks, brute force attacks on biological complexity. It's been going on for centuries. There was a Linnaean attack, hierarchical classification, the way of making sense of life forms. There was a comparative embryological attack careful comparison of early embryos all across the phylogenies that the Linnaeans had constructed was going to sort of tell us how development worked. When I was a young scientist there was an entomological attack. Cold rooms were proliferating and everywhere we were going to purify every enzyme and figure out its active sites and its allosteric regulation and so forth. I think it goes without saying that these frontal attacks all failed. Biology itself moved on dramatically but always by changing the conversation that it was engaged in. Not by adding more zeros to a strategy that had already gathered most of its hanging fruit. With basic science my main message would be that let's be cautious of views of the future which basically add more zeros to things that we've already done. There is some scaling up here and there that needs to be done and again I emphasize I'm not talking about the next few years. I'm trying to look to the future but I'm also not talking about the next century. It better not be that long. So you look at a basic science paper a very good one. It's a 131st author as my close friend Bob Waterston. So that's an actual count. What is the future of this kind of work? In the past we've already heard a lot about today an enormous amount has been accomplished and I have no intent of denigrating that. My concern has to do with the future. Many proposals for learning about the modification state of every lysine and every histone and every nucleosome at every stage of development and so forth are too reminiscent for comfort of these past frontal attacks on biological complexity. I think if you reflect just briefly on today's talks the ones that were just the most exciting they were the ones that were operating outside the bubble. Genomic methods, genomic perspectives and so forth were tightly integrated with many other modes of investigation including clinical ones. Dan Kastner talked about particular patients with particular phenotypes which driven his career. Rick Lifton's example was fascinating of going into the Framingham study. The Framingham study no less and genotyping a lot of people so that he could acquire a few by genotype that were worthy of a particular study and tap into the fact that there was already extensive clinical information about them, longitudinal information. So that's the world that I'm sort of looking out at is one at which eventually the whole population is going to be the theater in which we do such integration. But on the basic science front I think that we will increasingly find also that rationalist sort of views that we finally have found the key to cancer, to development, to this, to that are going to disappoint. The key is to take these new tools and to integrate them tightly with the whole variety of basic science approaches hypothesis generation and all that have always driven real success in those fields. Now the same passing that in terms of Amy's charge that we need to get out of the lab and go and talk to real people the broad agenda that I sort of laid out with very broad brushes would actually guarantee that that happens. I guarantee that you start getting molecular data into health records not into some artificial study for a few very heavily counseled people. There's going to be a tremendous dialogue started with the clinical community, with patients and with society at large and I believe that in basic science we may find a similar phenomenon with microbiologists and biochemists and cell biologists and so forth. I'm not saying that this doesn't go on now, I'm just saying that the future of genomics is to make these ties and then the future will be bright. Thank you.