 Well, thanks, Eric. That's going to be a difficult introduction to live up to. And needless to say, this is going to be a difficult meeting to summarize. The life of a summarizer is always a precarious one. And I encountered some special hazards at this meeting, particularly the sort of creeping intrusion of really polished rhetoric at the podium. We really had some seriously good, very carefully prepared talks. It's a challenge, of course, for a summarizer to operate in that environment. I was asking Rick Lifton whether he prepared his talk just for us. He said, oh, sure, actually, I just put it together yesterday morning. This reminded me of a joke that was kind of at the antipode of the damn quail jokes. As far as speakers go, involves one of the famous speakers of the 20th century, Winston Churchill. As this story goes, a parliament staff member was wandering through the back halls of the parliament building and went into a conference room. Suddenly encountered the great man himself, by himself, making a vigorous point. And so he tried the impossible, which was to sort of back out quickly without being noticed. But without breaking his train of thought, Churchill said, I was quite all right. I was just practicing my extemporaneities. So I think that's probably what Rick Lifton was doing yesterday morning. We had everything at this meeting from some really polished and deeply thoughtful talks about the future of our endeavor to other activities that were a little closer to a bar room brawl. And I will take a pass at trying to make some sense out of this, but inevitably we'll do so by projecting some of my view of where we've been, where we are, and where we're going. I really can't avoid that. So we had a lot of talk about 20-year timescales and some interesting ways of playing on that theme, which I'll comment on, but I wanted to play my own 20-year game involving looking back. This is a slide that I prepared 20 years ago. And I want to use it to illustrate a variety of points. First of all, we're well aware of how the technology has changed and the technology of slide making is out on the lead. The PowerPoint at least provides some leveling of the playing field between the summarizer and the other speakers. This is a PowerPointized version of a slide which I drew myself with a Leroy lettering set using India Inc. And one of the points I'd like to make about it is there's a principle of human nature that has been pointed out by a variety of futureists that humans tend to overestimate the amount of change that will occur on a relatively short time scale. Let's say on the order of five years and underestimate the amount of change that will occur on a longer time scale, such as 20 years. And this slide I think illustrates that point. When I drew it, I thought of the implementation of this plan as something that could certainly be accomplished in, let's say, two or three years in an organism such as yeast that I was working on and that the extension to larger genomes would be relatively straightforward. It actually took much longer than that. In fact, it was really only in about 1998 that Mark Amara and John McPherson and others actually successfully implemented schemes of this type on the scale of the mammalian genome, so it actually took nearly 20 years, despite really the conspicuous simplicity of this basic plan of action. I think in thinking ahead, this principle of overestimating how different the world will be five years from now and underestimating how different it will be 20 years from now will loom over our prognostications, because although I couldn't have imagined in 1981 that it would take nearly 20 years actually to bring this simple scheme to full fruition, I also couldn't even come close to imagining how much the world of genomics would have changed. So there was this yin-yang aspect, if you had told me at that time, that we would be looking at a complete human genome sequence and detailed genetic maps and a very rapidly growing area that is central to biomedical research in 20 years. I couldn't have imagined that, even though I would have thought we should have been able to build these contigs a little more quickly. So in trying to think about summarizing this meeting, there was so much discussed that it seemed clear to me that I would have to attack at a relatively high level of generalization. And I identified two what I call central tensions that I want to focus on. So I deliberately am using this term tensions because I think they're not choices and not dichotomies. We're going to pursue some mixture of these various options. But I think that they will in many ways dominate the decisions that need to be made by all of us individually and by programs and so forth over the next 20 years. So the two that I chose are the following. One central tension is between what I call reference databases and on-the-fly acquisition of genomic data. I think much of the discussion at this meeting really hinges on the question of when is it the right thing to do to develop a relatively large, well-organized, centralized project to create a reference database that many people will use over a very long period of time and when is it the right thing to do to really promote highly modular, extensible, distributable technology, get it out in the hands of the smallest labs that have the critical mass of expertise to deploy it and encourage them to do so in the context of hypothesis-driven research focused on very specific questions. And its attention, as I said, not a dichotomy. I'm sure we'll be doing both of these things. We've done both in the past and we'll do a lot more of both in the future. But the one cautionary tale or cautionary note that I would like to emphasize is really made at the bottom of this slide. And that is, I don't think we should be too quick to adopt the reference genome model that was at the core of our greatest success. The idea of sequencing reference genomes of intensively studied organisms despite the difficulty, despite the cost, and despite the initial large skepticism of the molecular biological community was a uniquely good idea. I would actually like to impart some scientific content to this claim and use it to contrast with many of the types of genomic activity that were discussed at this meeting. Really at the core of this, the uniqueness of the reference genome idea is the central dogma of molecular biology and partly again to show my age, I've taken a little image from the book that many of us, I suspect, in this room used when we first encountered this very powerful view of how molecules are organized, informational molecules are organized within the cell functionally organized. Of course we've learned a lot since 1965 about the details of this process and there is a whole cottage industry of philosophers of science sort of challenging the extent to which this view even is a remotely adequate summary of the way that cells really work but it actually is a pretty good summary and underlies much of the way we think about these processes. And really the message that I got from Watson's book which is about genomes that is central to our thinking about this subject is that there really are the one finite component of biology. There was always a silly notion promoted in some quarters that they were the key to understanding all of biology. Clearly there is a need for an immense variety of activity. I think all of it and then some discussed at this meeting to surround genome sequences but we still come back to the fact that biology is at its simplest in a germ cell when there is a haploid genome there and from that information cells and organisms must bootstrap themselves with a rather minimum of pre-existing machinery compared to the organism that they will build and maintain. And this finiteness allows the completeness problem to be solved to a good approximation and it gives a database which really is in a sense, I would argue, the mother of all databases in biology because of its universality and its permanent importance in analyzing diverse biological phenomena. So that's the good news. The bad news is obvious and much illustrated by this meeting that biology gets more complex very rapidly once one moves beyond the germline sequence and so we talked at this meeting about understanding epigenetic modification, alternative splicing, developmental regulation, translational control, reversible covalent modifications of proteins and so forth and whereas it is to me as a sort of old genome guy, hardcore genome guy a little frightening to think about the $1,000 genome and extrapolating current production of sequence data out into the stratosphere, at least that's a well-defined activity and bounded by essentially an infinite number of orders of magnitude compared to understanding the epigenetic state of every genome in every cell at every stage of development and so forth for RNA molecules, for proteins and so forth. So I think the completeness problem is huge and the importance of its hugeness is that it will create constant and I think critical tension between the question of when should we go out and measure certain reference databases with the expectation that it will be worth the cost to acquire them, it'll be worth the financial cost, it'll be worth the cost in centralizing activities and producing a database that will really be of sufficiently permanent value to a sufficiently diverse user community for sufficiently diverse purposes to really be meaningful and when should we be doing really science conceptually as usual but doing it with new tools that allow people to ask fairly specific questions and to look at epigenetic modifications and alternative splicing and so forth in their particular cell type, as their particular process plays out while they're asking their particular questions and tuning the technology of that moment to answering them as best they can. So that's one tension. In some ways, a larger struggle is underway and to me was a clear tension in this meeting and will be a major tension for the future and it really is, I think, between essentially two ways of formulating the future of human genomics and related activities. So there are these two ways of formulating the future. We heard both of them in many variations many times at this meeting. We have the human genome sequence and now we should concentrate on delivering the human genome as to health benefits versus we have the human genome sequence, something that we agree upon even though it's not true actually. Now we should concentrate on understanding how it works. So the points were made, I think, in synthesis, very articulately and from many points of view that this is a tension, it's not easily resolved. Many people in this room are not in agreement about the relative proportion of activity that should go in these two different areas. The areas are actually different. They reflect rather different views of what scientists would like to accomplish with their own careers, what they would like to see happen to our collective acquisition and application of knowledge. And so we're going to need to work on a constructive, creative way of managing this tension. I'm going to focus largely on the first point, mostly because I believe it was by far the most novel feature of this meeting compared to many previous ones that I've been at. It was an interesting example. Last night I was at this dinner from hell listening largely to discussions about how to make the mystery breakout groups as effective as possible and when we got to discussing one, which I ultimately went to, sort of on therapeutics, there was an interesting little discussion amongst the various parties there about how, gee, we don't have the right people here to talk about that. This point was absolutely right. The number of people in the audience here who know anything about therapeutics is small and most of the traditional participants in the genome project, like me, know almost nothing about therapeutics and therefore convey very little such information to our students, to the environments that we work in, and so forth. So I did make the one point at this meeting that this is actually quite an interesting situation. That under these historical circumstances, which Francis, I think, very nicely articulated the first night, we really are at a major point of departure in this endeavor. This planning exercise is unlike all previous planning exercises because we are looking at a much blanker slate as to where are we going to go as a community. And we now find that in the year 2001, when we assemble much of the elite of the genome world and a very interesting collection of other people that can contribute to this discussion, that there are very few people there that know much about the development of therapeutics. That is about the details of how one, or even the generalities of how one gets from a genome sequence to health benefits. So I think this actually is a profound point and one that we need to change. Regardless of the way that this tension that I am focusing on plays out, it needs to change. We need to teach our graduate students. We need to inject it into our training programs, into genome meetings. You go to a Cold Spring Harbor genome meeting and there's not even a poster, much less a talk or a session about how to get from genomics to therapeutics. Journals like genomics and genome research really just don't publish this type of work. That's going to change and we've got to make that change because of course all of these processes are really us. So I'm going to focus a little bit despite my very tenuous knowledge of this subject on how it is that this side of this tension might be imagined to play out and what some of the issues are that I heard discussed at this meeting that need to be taken into account. So I was quite taken by my colleague, Bonnie Pagan. I think like most of you, I get to know some of my colleagues mostly on the overnight flights to Dallas and shuttle buses to Erie House and so forth. But she, who for those of you who don't know her, is a very experienced clinical geneticist who has spent her career taking care of largely kids with a variety of genetic defects and has more recently been heavily involved in trying to improve the accessibility of genetic information to the practitioners who need it. But I was taken by her metaphor, which was the metaphor of the narrow squeezed pipeline and it seemed to me to encapsulate a lot of discussion I heard at this meeting and I want with some care to look at this dynamic because I believe some of the issues that are symbolized on this slide are very important to us. So the basic metaphor, as I understood it, is that we have over here genomic data techniques, conceptual paradigms, that is the kind of hard science of genomics, which is really where the successes have been in the last couple of decades. And we have over there what is really largely a dream of better health through novel therapies, prevention and personalized medicine. And not only is the pipeline narrow that connects these two, we have some successes, but the very frequency with which they're mentioned is an indication that the amount of flow through this pipeline is as yet small. And there's no question that the pipeline has been under very close scrutiny because of real or perceived ethical, legal and social risks. And certainly in addition to the theme of this meeting that suddenly we have people like me now in discussions about drug development, the permeation of this meeting with much broadened vision of the ELSI issues was also, I think, new and striking. It struck me and a number of other people have commented on that. And so all of that is in a sense here on this slide. So I want to say just a little more about it. So we, I think, are coming to a pretty clear consensus that a major NHGRI mission of the future could be formulated somewhat in the following way. It's to develop broadly applicable paths from the genome sequence to better health that are consistent with our society's ethical, legal and social values. So mission statements of this type are easy to state and exceedingly difficult, not just to carry out, but it's difficult even to design a sensible path forward that will move in the direction of this large mandate. So I'm going to look at this a little bit programmatically and a little bit scientifically. Let me first take up what I regard to be some essentially programmatic issues. So these are just examples. There are a lot of other examples, but three that stood out for me at this meeting are the ones on this slide of implications for taking up seriously this challenge of delivering on the health promises of the genome. And I've already said that a striking feature of this meeting was an expansion and changing of the dynamic of the discussion between the LC program and the larger enterprises of the NHGRI. I think it's clear that we must define a role for the LC program that is much broader than its current one and more integral to the overall NHGRI mission. I think it is becoming increasingly clear to many people involved in this endeavor from many different points of view that it needs to become literally more integral. It needs to become integrated in many ways into our thinking about resource allocation, about the way the science is actually carried out. And I believe in a subtle way, but potentially very powerful one, a greater awareness of these questions will gradually influence the kind of science that we all do. And that will be the ultimate impact of a broadened LC program is really a different kind of scientific program. And I don't mean one that's more tightly regulated and which has more attention to this, that, or the other constraint on the research, but it involves a bigger thoughts about how the science integrates into the society in which the science is carried out. And I'm going to come back. I'll just anticipate at the end of this talk really to one of the big themes that I've gotten out of this phase of the genome project, and that is the intensely social nature of this endeavor that we're involved in. So that's one set of issues. The second point to some degree applies regardless of where we are on how it is that we're managing this tension between trying to understand biology and trying to improve human health and dealing with the dichotomy, the overlap, and all of the aspects of that tension. But I put it here because I do think that it has a broader dimension in the context of the desire to focus more of our attention on this translation of the genome sequence into better health. So I think it's conspicuous that we must devote an increasing fraction of resources. And the issue really is the fraction of resources because if we have more resources to collect data with, there is going to be an increasing need for the attention to the ways in which we make the data accessible and analyzable and so forth. So devote an increasing fraction of resources to facilitating access and analysis, access to and analysis of existing data as opposed to the acquisition of new data. We're going to be collecting a lot of new data and if we get the thousand dollar genome, we're going to actually start to do something that people erroneously thought was going 20 years ago, which was stressing the hardware capabilities that Moore's law was providing us with. So I'm not arguing that we're making some conversion to theoretical biology here, but an increasing fraction of resources must go into this area and must do so in a much broader way than we've really thought about before. Actually, a lot of progress has been made in this area. Scientists love to complain about things and yes, the genome sequence is difficult to use. Yes, it's not very well annotated. Some of the difficulties are because it's not done and it'll get easier to annotate and easier to use when it is done, but others of the difficulties are just inherent. It's a hard problem, but as an example, I would encourage people who haven't used MapViewer at NCBI, let's say in the last month or two, to do so. It has been a product that has been improving dramatically month by month and I think that's just one example and that's at the sort of narrow base of this problem. We need annotation, annotation, annotation, integration, and so forth. We've been hearing that over and over again and I think it's absolutely true and to the extent that we take seriously this health mission and I think we will be taking seriously this health mission, it adds, of course, a whole additional set of layers of need for access to phenotypic information and ultimately clinical information and so forth. This is a big challenge and has many dimensions, but is going to be a big part of our future. And finally, just following up on my comment that we're envisioning this large-scale, somewhat true believer effort to improve human health through genomics, better livings through gene sequencing. We need a lot of education, not too much because the strength, actually, of genomics from its earliest days is that people who didn't really know what they were doing had sort of a basic view that they could change the way that things are done and it's always people who don't know what they're doing that are the best change agents. You know, Delbrook didn't know a damned thing about viruses, but it's fortunate that we didn't sort of turn that project over to the virologists of the day. So we will be intensely criticized as a community, certainly to the extent that people like I are involved in it, we will be intensely criticized for moving on to all sorts of other people's turf. Fortunately, those of us who've been through a lot of these planning processes and ancillary activities are immune to such criticism. I've altogether lost track of the number of things that I've been criticized for that were of this same general nature. But we succeeded at sequencing the human genome because as a community we learned what we needed to learn. We may not have done everything in the most elegant way and some of our fluorescent dyes, maybe didn't have quite the quantum yield that the best photo physicists could have provided, but they did get the job done. And as a community we learned what we needed to learn and we can do that here too. And we have to do that and it has to be a really major activity. This problem that I stated that our major journals, our major meetings, our training grants really don't engage this issue, has got to change. And it's not going to change tomorrow but it's going to change by an intense sort of cultural process. I think a lot of us are personally committed to this and will be ultimately the most important. That will ultimately be the most important factor for change but there are of course things that can be done, the way of workshops, short courses and so forth. So I want to get back to this mission which is kind of central to my synthesis of this meeting and now take a look at it from a more scientific perspective. And I'm going to do that using a slide that we've seen I think at least twice before. One of the virtues of PowerPoint, I really didn't prepare this talk in advance, but I have on my little disc here every slide that I've ever shown in the last 20 years, 10 years. A lot of my earlier ones were PowerPointized as you've seen. I think this is actually the original version of this figure, the two that we saw were updated versions. But the idea is clear, the updating is mostly about the top of this, sort of modernizing the scientific context in which we arrive at a gene that we already have knowledge because of the way that we got there about genetic knowledge about, but often very little other knowledge. And so in this Shattuck lecture, which I would commend to you because it's really exactly on this topic of how to translate knowledge of the genome sequence into health benefits, these are three branches of translation were articulated. I just want to touch very briefly on some of the things that I heard at this meeting about each of these branches and then I'm going to take the liberty of providing a modest critique of the right-hand branch, which is where I believe most of our bets ultimately lie. So we have a lot of diagnostics and we had a lot of talk at this meeting about the both potential benefits, a few examples of real benefits and various LC issues which are real and have been studied rather intensively about attempting to carry out this branch. We had the sobering reminder, which comes up every time these issues are discussed and should come up every time these issues are discussed, that history isn't terribly encouraging about this preventive medicine, particularly when it involves any change in people's lifestyles and the tobacco example is the mother of all such examples. And there's no question that pharmacogenomics is going to play a role. I think the issue is whether it's going to be a rather niche role, somewhat like other well-defined genetic diseases. There are certainly people out there. We can even recognize some of them now who are at high risk of adverse reactions, sometimes catastrophic adverse reactions to particular drugs because of Mendelian genetic conditions or near Mendelian ones. I think that's not the issue. The issue is, is this really going to be a path broadly towards better healthcare for most people to personalize the medicine? And I think that Rick a bit finessed this issue, but I think in the year 2020 I'm still trying to make up my mind about the impact of diagnostics and preventive medicine on mainstream medicine and gene therapy issues were discussed in a somewhat similar vein. I think the only summary I would make of several discussions I heard about gene therapy is that even amongst some of the people here who are very knowledgeable about this subject and involved in it, I heard no strong claim or any claim that the field is seriously under supported at the moment and that existing mechanisms of funding progress in this area are inadequate. And from that I infer that if gene therapy is to play a substantial component of this translation of the genome sequence into better health, that the NHRI's role will be at really the more novel idea end of this. That is a very basic research because the message I got is that the translational research on the diseases that are under serious study now is progressing and is probably adequately supported. And then on the right-hand branch we have where I think most of our bets are placed. And this was discussed, but I want to provide a slightly contrarian view of the level of challenge in carrying out this agenda and to make a very simple, but I think important point in thinking about the challenge ahead. And I'll do it by contrasting, and this really is a freshman genetics point but one that I think frequently gets lost in these discussions. I'm going to do it by contrasting the middle branch with the right-hand branch. The middle branch has the immense virtue of actually being a logical pathway. That is, you identify in the classic medical tradition what's broken at the beginning of the causal change of disease process and you fix it. So when we find this gene and know what bad effect it has on the patient, then we in principle have a strategy for fixing that defect and providing effective therapy. The right-hand branch typically is an illogical path and struggles against its fundamental illogic. It is, we find what's broken and then hope essentially to fix it by breaking something else. The overwhelming majority of drugs are antagonists rather than agonists and so they in general all of you know that I'm oversimplifying a situation that continues on into graduate level genetics but nonetheless the typical mutation leads to deficiency or absence of protein function and the typical drug has the same effect and when I have given talks to very general audiences on this subject I've used the example of BRCA1 and I've just found to be an effective teaching device is that the immediate pharmacological message from identifying BRCA1 as the gene mutated in this high predisposition form of breast cancer is that the pharmaceutical industry would know how to develop novel carcinogens. It tells us nothing about how to develop anti-cancer agents. Now in slides on meetings of this type we have optimistic talk about bootstrapping around pathways and using the mechanistic insights that come from recognizing these mutations and their phenotypic effects to find drug targets lurking somehow nearby and some examples as I've reviewed this literature mostly rather tenuous ones show how this might work out. I think the point I'm trying to make here is we don't have a general approach here. We have a belief that understanding a lot of mechanistic issues in pathophysiology is an essential step in providing good treatment but we certainly don't have a general way of moving down this pathway and we're never likely to because of the fundamental logic of the situation and the fact that this biochemical study that is the critical intermediary is going to be inherently idiosyncratic in every particular case depending on the particular nature of the defect. So I injected this scientific point really not to sound a note of pessimism but to emphasize that we are taking on here an immense challenge. It's a challenge that has immense social components. It also has immense scientific components and I'm going also to return shortly to the question that one of the reasons that we need more of this public education that was much discussed at this meeting is to undo some fairly serious damage that has been done from many sides in exaggerating the extent to which there is a simple path down this diagram to improve therapy. This is the sort of gene-to-day gone-tomorrow theme which I have seen used in settings that are poignant. It is disease interest groups focused on what by current science are absolutely intractable developmental genetic defects in which using the tools that we know so much about how to deploy we've gotten started on this diagram. That is we have the gene today but these diseases are not going to be gone tomorrow and some of these diseases are never going to be gone. This is a message that I have not heard made even one time in a public environment in which the broad subject is discussed every day. So there are challenges ahead. I want to move now to a somewhat facetious look at the aspect of Rick Lifton's time travel. So this exercise reveals something that was actually deeply personally rewarding to me is that I discovered that those of us scientifically active in 2001 are truly members of the greatest generation of genome researchers. The reason I had this rare insight is that I found that in 2020 every health advance owed to genomics was firmly rooted in research agendas which are already clearly formulated underway and just need good execution. So we can feel really good about having gotten this very hard problem almost exactly right. But we should also of course feel a little uncomfortable. These reports from the future may not be entirely infallible. The way I would sort of frame that question is in this way. Do we risk adopting policies now that sell the generation let's say that we'll be hitting its stride in 2010 short? And what I mean by this is that given a proper chance they are likely to devise broadly applicable pathways from the genome sequence to improved health that differs substantially from those that we presently envision. And this is partly a speculation which has overwhelming historical support given the difficulty of prognosticating particularly prognosticating the future. But it also has policy implications which I really think we need to take seriously. So let me amplify slightly on this point. The worry that I have is that we have so much confidence now after two decades of success that we know how to carry out this mission that we are going to lock up resources ways of doing the science ways of thinking about the science in expensive, very long-term commitments which will preclude the type of innovation that I think that we're going to need. And I'm not actually talking here about the serendipitous scientific discovery which is really a more relevant issue on the biology side of this tension which is not where I'm focusing. I'm really talking here not about new discoveries but just about new ways of approaching the problem ways of adding new branches to Francis's diagram. And what I mean by a broadly applicable pathway is illustrated somewhat. All analogies are treacherous but it is illustrated somewhat by the positional cloning phenomenon. I would describe positional cloning as having been a new, broadly applicable pathway from genetic observation to an understanding of molecular defects and genetic disease. There were actually no really new ideas behind positional cloning by the time that it was being seriously promoted in about 1980. It was rooted in aspects of human genetics and recombinant DNA technology that were by that time well understood. And the feasibility actually was in no serious doubt. But there was a large job to formulate the potential advantages of this strategy, of course, to work through a large number of implementation details to educate a whole community that wasn't accustomed to thinking in this way. But all of these things did happen and in the course of really just say 10 or 15 years certainly the medical genetics textbooks were rewritten from beginning to end and rewritten in a way that a biochemist could understand and that's, as we've seen, an essential step toward better treatments although not an infallible one. But my point is to illustrate what I mean by a broadly applicable pathway. And the question that I'd like to focus on just very briefly at the risk of indulging is one of my own personal interests is to try to illustrate what I mean by this. And my argument here is not that the particular idea that I'm going to propose is a certain winner or even a likely winner but it is to illustrate with a scientific example the way I think about this problem and it does also connect to a substantial amount of discussion that occurred at this meeting about the relative priority that should be given to primate sequencing and how we would work with that information. So these are our closest relatives and it's easy to pick the human out here without being a primatologist. He's actually the photographer and of course there's a major message in the fact that he's taking pictures of them instead of them taking pictures of him. That is, there are major cultural and behavioral differences between these different closely related primates. But there is a more fundamental point and that is that it isn't just anthropocentrism to look at this picture or to walk around a primate house in a zoo and to recognize that humans are the novel offshoot of a generally fairly conservative evolutionary lineage. One of the great problems in evolutionary biology of course is to understand how such lineages arise. Generally stable evolutionary lineages rarely but occasionally launch altogether new experiments in a way that no one could have predicted five million years ago even with the benefit in another time travel sense of sort of modern cognitive capabilities and even education say in logical thinking and the empirical method could have predicted how this was all going to play out but it play out it did and we have now the conspicuous outlier in the primate world and it is us. So I want to connect what I readily acknowledge is simply personal interest in this to my broader theme by illustrating how this way of looking at the comparative genomic problem with our closest relatives might represent a broadly applicable pathway which we are not pursuing in any significant way now toward connecting genomes and a better human health. It's based on a speculative evolutionary model for the evolution of novelty and one that potentially is applicable to many such situations and has a number of textbook features to it but also I think some new ideas. We start with abrupt environmental change leading to greatly reduced competition in a rich habitat. This tends to be a characteristic as near as one can tell from most situations in which rapid evolution occurs the simplest example being the Galapagos type 1 in which what previously non-existent terrestrial habitat in a fairly favorable climate and with reasonable soil chemistry pops up in the ocean and there is a greatly relaxed competition for the first migrants to arrive and rapid evolution ensues with considerable novelty given the time scale and the small populations in otherwise stable lineages. So in more realistic situations let's say such as the Savannah rainforest boundary clearly there do need to be actual molecular novelties that allow one species to exploit this new habitat but what's different than many views of this problem in this model this speculative model is the idea that following this genetic opportunity to exploit a rich new habitat that there is a cascade of loss of function mutations that compensate for the intrinsically poor adaptation of the species to the habitat. We know as geneticists know that this is what happens most readily when you put yeast cells on a plate that let's say a pH that yeast can barely tolerate so you don't suddenly evolve new proton pumps you actually break ones that you've already got and it seems a reasonable conjecture and not a new one H.J. Mueller actually said that humans are hastily made over apes that's the essential idea behind this model I'll skip the point about extinction The point medically is that it may well be that we need to understand this process what were these compromises and see this is an altogether different focus than the question of why can we compose music and so forth and what were those few mutations that somehow enabled language actually focusing on the downside of being relatively early members of a novel evolutionary lineage evolving in a constantly changing environment to the endpoint that's illustrated on this humorous cover and of course symbolizes the core of the medical problem in so-called advanced countries so it may be that to some degree our obsessive focus with phenotype variations within the human population is overdrawn relative to trying to understand the broader spectrum of health vulnerabilities that are an intrinsic feature of this aspect of the human evolutionary context so I just hope that you'll keep now clearly in mind my purpose here which is to illustrate that we don't necessarily just want to keep doing more efficiently, better organized duic style management to do what we already have clearly in mind and we particularly, let's be cautious in not locking up all of our resources in these lines of activity which we now have in mind it reminds me of the current debate about the B-2 bomber which basically all components of the military industrial complex want to build a lot of B-2 bombers except actually the Air Force doesn't want them and the Air Force doesn't want them because at a time when the demands on the Air Force to be able to provide air power in highly flexible tactically responsive ways are ever increasing the Air Force High Command understands that a commitment to building 50 new B-2 bombers will preclude a lot of more innovative activities in their somewhat distressing mission we should take some caution about that same principle so to really summarize my summary it is that I think we're all coming to more of an appreciation perhaps many of us had before that the human genome project was a social event science is an intensely social activity it was a social event at every level the social interactions in this kind of group in this community and of course the relationship to the larger society I can't resist joking about this and that typically the social events I go to at most I suffer one hangover and this party which is now coming near to a close I can't count the number of hangovers that I've contributed to this really heroic effort that I think all the participants should take pride in but should also reflect rather deeply about where we go from here so I will close with one last slide in which I cast my vote on a topic that we were charged to think about but I actually heard no discussion about during the meeting and that has to do with whether or not we should retire our jersey and my vote is that we should let me skip this slide that I think at the appropriate moment let's say that the human genome project is over and one of its legacies was the founding of an exciting, eclectic and powerful new field of science and the final charge I would just make with my emphasis on enabling our successors is that we should put a lot of our energy into inspiring a new generation of scientists mostly younger than the people in this room because they're the people that are going to have to make this claim real. Thank you. We do have time for a few questions. Ewan. Speaking of young guys. I have to ask a question. I very much like to talk. One of the things that I always find confusing is that people never seem to say let's just do the simple stuff well. So we have a lot of incredible information and a lot of what we've got to do is do the simple stuff very, very well about delivering it to a whole bunch of people who can use it very, very simply. They stopped doing degenerate PCR. They stopped doing Southerns. They stopped having to worry about how they collect their markers. They stopped having to worry about how they collect their SNPs. And I just find it confusing that we miss out that why don't we do the simple stuff very, very well. I mean, that should be somehow an emphasis somewhere. No, I appreciate your comment. And I think people have seen me in this role before know that I sort of made my living by arguing that we should do the simple stuff well. I think that to some degree that was an implication that I intended by my sort of short-term, long-term issue. In the next five years, the world's going to change less than I think many of these predictions. Despite my admiration for my former graduate students, technological research, I'm skeptical that we'll even achieve one burq in five years. But I am not about to predict how many burqs we're going to achieve in 50 years or 20 years. And the implication, the way I connect this to your question, is just that the world I think won't change as much as many people think in the next five years, and that's always an argument for doing the simple things well, the things that can be done on that timescale and which have a relatively predictable payoff. Maynard, do you really think that we can use the genome for developing therapeutics without finding out how it works? Yes. Well, that's what I wasn't certain you really thought that, but I wanted you to confirm it. Yes, I do think that. The problem of trying to understand all of biology is a wonderful problem and trying to understand how the genome works. And it is, I tried to emphasize an aspect of this attention, but I think that if you look at the whole history of developing therapies, and let's even limit our attention to the relatively knowledge-based therapies, as opposed to the ones that were just altogether accidentally discovered almost always, they were discovered with very limited information. There's still a lot of highly effective drugs that we know very little, actually, about the detailed mechanism of action, and we certainly know nowhere near enough about all of the things they interact with to make any kind of predictive judgment that this drug would have the effect it does. But nonetheless, it's an area where very much a little bit of knowledge often goes a long ways. That was really a beautiful synthesis of a lot of complicated stuff. And one of the points that you made that I would like to emphasize and come back to is the notion that understanding pathophysiology doesn't immediately translate into therapy. And obviously we have examples that go back to the 1950s, such as sickle cell anemia, where we've known in great detail to impact meaningfully on the disease, at least today. However, I would submit that we are at a time when now that we have genome sequence, we really have the opportunity to do the first part, which is to figure out the basic pathophysiology of disease. And that's the part that I'm most enthusiastic about because I do see a path to be able to do that. I'm quite agnostic about how that's going to be translated into therapy because I'm not smart enough to know that. I do begin to have the tools to really try to systematically figure out the pathogenesis of common diseases. And that to me is the galvanizing aspect of what NHGRI ought to be able to focus on and motivate for the next 20 years. Right. Well, I basically do agree with that and didn't mean to imply otherwise. Rick Lifton, who was once one of my star freshman chemistry students, characterized himself the other day as someone who, despite his charming 20-year look into the future, someone who mostly thinks about what to do that afternoon and his track record indicates that he thinks about it very clearly and very well. And there are... this is something we can do, and we should do, and it's fantastic. I mean, what's been learned is fantastic. You take the genetic diseases, particularly the ones that are fairly intensively studied, where we've had the gene for a decade or so, and it's astonishing the amount of pathophysiological information that we've learned. And I think my only point... it's the issue of which way to lean in the boat. And my only point is that I wouldn't want to see the NHGRI's mission defined as understanding pathophysiology. When understanding pathophysiology is the most effective path toward improved health, then we should do it. But we shouldn't regard it as our central mission. It was a wonderful talk. I did have, however, some small concern about the slide concerning the pipeline, and the LC as the pipeline seemed to be presented almost like the rate-limiting factor in taking basic science and translating it into health interventions. I don't know if that's how it was intended, but it could give that impression. And of course, it's not possible, I don't think, to think of LC as the rate-limiting factor, because whether or not there's an actual demand or need for such kinds of services, whether or not there's any incentive on the part of anybody to take basic science and translate it into those services are equally important. And indeed, on some occasions, there sure can slow things down in a kind of pipeline narrowing way. In some cases, might speed it up by alleviating certain problems that might otherwise arise in that pipeline. So the slide, I know you didn't intend it to be one that created a kind of adversarial relationship between LC and basic science or LC and technology development, but I think it could give that impression. I think it would be a misimpression. It is certainly a misimpression, but I appreciate your comment and I think it actually was one of the most interesting aspects of this meeting. Both Bonnie Pagan and Rick Lifton, for example, told me that they had quite a number of comments about their talks yesterday from people involved in LC activities along these lines that statements they made were perceived as having put, you know, an unfair amount of responsibility in the LC process as inhibiting progress. What I think was very constructive about this meeting is that I believe this type of dialogue was engaged at a level of candor and realism that I haven't heard before. It is absolutely a misimpression about this slide, the reason for my rather careful attention to my view of the scientific challenge, the reason why the scientific challenge is severe is to indicate that I believe that pipeline is narrow because we don't really know how to do this scientifically. I absolutely do not believe that we are right on the path toward tremendous health benefits if we could just get the paperwork out of the way. In fact, the slide that I skipped actually was an advertisement from Agilent as it happens, it shows a sort of colorful double helix here, and the attention-getting claim that at the top of this ladder is a world without disease and that we're going to get to that ladder essentially by better execution technically of the types of laboratory activities that were the dominant theme of this meeting. I'd like to have more scientists out there with me on the lecture circuit saying that it is morally wrong for companies to run advertisements of this type. Keith Crandall from Brigham Young. I agree with that, and I want to stay on this theme of that slide, your dumbbell slide, with the narrow line in between. I want to ask the question, we don't even have this letter yet, it's expected April 2003, right? And presumably, there must be some expectation of a time lag from having that basic information to going to useful therapeutics. What, in your mind, is a reasonable timeframe for that, I mean, to me, it's unrealistic to expect that before we even have the basic data, and how long do we expect, what's that timeframe for that expectation to become realized? So the way to answer that, just referring back to something I said earlier, is that the first thing I'd just like to say and that I'm very uncomfortable with the talk about the diagnostic therapeutic gap, and I heard a fair amount of talk about it at this meeting, that's a common phrase. This gap is going to extend forever in the case of a number of catastrophic human conditions. That's the reality. It is not imaginable with our current view of biology as to how it is that we're going to alleviate the catastrophic effects of certain types of mutations that lead to viable, the birth of viable offspring. And now, that's harsh, but I, again, am looking for the right place to lean in the boat. I think that more candor on this subject is required. The length of this gap, of course, it's hard to estimate, even in the ideal cases, that it is going to vary from a lot shorter than we think, perhaps a few years, in which we discover that an existing FDA-approved drug actually has some way-off label used, maybe in combination with some other intervention to affect major disease processes, to potentially forever. The human condition is not going to be, ultimately, the tragedy of the aspects of the human condition is simply not, ultimately, going to be attenuated by genomics or anything else. So that's a hard problem. Jeff Deweyck said a fairly standard estimate is that it's about 12 years from having kind of a lead compound or at least sort of a serious plan of attack in developing a pharmaceutical and bringing a drug to market. And that's in the presence of adequate and huge amounts of capital. So that's a long time. And to the extent that we're going to be looking, for examples, of mainstream drugs that are really affecting human health in a broad way, based on the genome, the human genome sequence, we would expect that it's going to be approaching 2020 before we have such examples, even in those cases that go extremely well. Penn Law School. Someone who's a complete outsider to all of this, one thought that I had when I saw the dumbbell slide and also more generally over the last couple of days is that one of the things, I don't know whether this is appropriately categorized in the LC group or if there's another place for it, but it seems to me that there's a lot of institutional input that should be gotten from economists and biopharmaceutical or pharmaceutical executives. I mean, they talk all the time about these drug development issues. I mean, they're obsessed with this, for better or for worse, and economists are obsessed with these questions. So the pipeline question and how to make the pipeline wider is fundamentally one that does involve some of the traditional ethical and legal and social questions, but is really also much more fundamentally a regulatory economic question that a lot of people who spend a lot of time thinking about drug economics have a lot of knowledge on. Right. This is exactly the type of thing I envision in trying to expand the LC program and make it more integral to everything we do because those are, to me, our critical LC issues and we should get engaged with them and broaden the community and put the resources in that it takes to do that well. What we can bring to the table, I think is not so much a lot of policy activity and so forth is serious scholarship on these issues. I mean, I've gotten into a few of these fights. I like fights and sometimes I'm just appalled at the lack of really solid information. Just because these big pharma guys get together and talk about these things all the time, it doesn't mean that they're doing it on the basis of any solid information. I think you're absolutely right. I mean, the fight to catalyze universities is doing engage in this research so there is a bit of a back and forth. There's a lot of back and forth but I'd like to see this process is extremely important for our society and I'd like to see this process studied with serious scholarship and debated and open forums and I'm not just looking for help in some of the arguments that I like to make. I mean, at the midst of, let's say, the Solero Wars I would have been delighted to have a well-informed scholar out there explaining the case for privatization because a rational argument, even one that's the complete antithesis of my views, would have been a lot easier to deal with than what we were up against. We're going to do two more. Kurt and then David. Those of us who interact with patients who have currently untreatable genetic diseases appreciate the importance of hope. Well, it's clearly, on this slide, it's clearly hazardous to overdose people with hope. It's also important not to deprive them of it and there's maybe a narrow therapeutic window there. I think you have to give people the right amount of hope without overdoing it. The point's very well taken. Maynard, I too really enjoyed your talk and I appreciate the emphasis on treatment of genetic disease and I would just like to point out to people in the audience that there is a metric, one of the problems with evaluating treatment of genetic disease is that the literature is largely anecdotal. Let us say a particular investigator reports on their success with one particular genetic disease over a rather, you often over a rather short period of time and so you don't hear so much about those diseases which people have tried and there is no, that doesn't get into the equation. So I would just point out that there is a metric out there that people can use and that was a set of studies done by Charles Greiber and Barton Childs which were done first in the early 80s with one set of, with a large collection of diseases, all Mendelian, selected at random and then a subset of those selected with a known defect and then that large collection of diseases looked at in the literature in terms of what was the outcome of therapy on a variety of different parameters of quality of life. The same set of diseases were then looked at again in the mid-90s, interestingly, the fraction which had no response at all to treatment was absolutely constant over that interval and there was a small decrement in the fraction of diseases or there was a small increment in the fraction of diseases which had partial response and they looked again in 97 and it hadn't changed much but I think it will be fascinating to look at that well-studied set of diseases and ask what's going to happen over the next decade and I would say with the genome sequence available and with the advances in particularly, I think, mouse models where pathology can be studied in a whole animal and the advances in sort of noninvasive measuring of pathophysiology in patients that there is some hope for improvement in the slope of this line. The only comment I'll make there is something absolutely consistent with where I think we should go. The one thing I had intended to emphasize a little more strongly than I did, I'm just reminded of by your question, is that I do think that there is a kind of special responsibility of the genomics community to engage itself in the better treatment of genetic diseases simply because of the nature of what we do scientifically and the causality of those diseases but I also think that genetic diseases will not always provide the best opportunity for carrying out this mission of improving health generally and that I would predict that in 2020 starting to see a big impact of genomic-based methods on really mainstream medicine where much in the LDL receptor story that was alluded to in a couple of talks where a study of rare genetic diseases is informative about how to treat the general population. Okay, well, once again, thank you, Maynard. You make the organizers look good. Thank you.