 Second day of Nobel 35, my pleasure to introduce my colleague from the Department of Mathematics and Computer Science, David, David Wolff. My pleasure to introduce Dr. Leroy Hood, a leading researcher and visionary in biology and medicine. Dr. Hood received his undergraduate degree at Caltech, his MD at the Johns Hopkins School of Medicine, and he returned to Caltech for his PhD, which he completed in 1968. He spent much of his career at Caltech, serving as chairman of the Division of Biology from 1980 to 1989. In 1992, he moved to the University of Washington, Seattle, where he serves as the director of the NSF Science and Technology Center for Molecular Biotechnology, as the William Gates III Professor of Molecular Biotechnology. Dr. Hood has received over 50 academic awards and honors and over 100 lectureships, among them the Louis Pasteur Award for Medical Innovation and the Albert Lasker Basic Medical Research Award for Studies of Immune Deficiency. Beginning in the 1970s, Lee Hood's lab has developed instruments for the automated sequencing of DNA and proteins. This work, which has revolutionized the speed with which sequencing can be done, not earned him a number of awards, including the Commonwealth Award for Distinguished Service. More recently, Lee Hood's lab sequenced the 685,000 base-pair T-cell receptor locus. Understanding T-cells is critical to the fight of autoimmune diseases, such as allergies and rheumatoid arthritis. This research exemplifies the potential impact the genome research project will have on health and medicine. Dr. Hood is also concerned with the next generation of scientists. Hood believes that research that is inquiry-based investigation is critical to a complete science education. Along with his colleagues at the University of Washington, he organized a program whereby high school students contribute to the genome project by decoding small fragments of human DNA. Lee Hood's efforts in K-12 education earned him the Distinguished Service Award from the National Association of Biology Teachers in 1998. We are pleased to have Lee Hood here to start off on conversation on this, the closing day of the Nobel conference. Please give a warm welcome to Dr. Lee Roy Hood. It's a real pleasure to be here and continue this discussion of the century of biology, that is the 21st century. I think in many ways will be a century of biology. What I'd like to do today is continue from where Greg Venter left off yesterday, asking the question, once given all of this sequence information, how do we convert it into knowledge? How do we convert it into an understanding of basic mechanisms in biology? How do we convert it into the revolution that is going to come in medicine? I think the answer to this question is contained in the idea that the 21st century, in the 21st century biology will be concerned with the analysis of complex biological systems and networks. I will in the course of this lecture try and give you a general feeling for where this research is going, how it will be carried out and some of the powerful new tools that we have at our command as we move into the next millennium. In 1992, I moved to the University of Washington to start a new department called Molecular Biotechnology and in the course of that move I had an opportunity to meet Bill Gates of Microsoft and have since that time gotten to know him reasonably well. I remember in the first dinner conversation we had together, he made a statement that I found quite interesting in many ways. His feeling was that there were going to be two technologies that were going to dominate the 21st century, information technology and biotechnology, and dominate it both in an intellectual scientific sense as well as in an industrial sense. What I found fascinating about this view is at its heart, biology and biotechnology and medicine really are informational sciences just as informational technology is. And in fact, the really fascinating question we'll explore in this lecture is what exactly do we mean by information and biology? What is clear is that there has been a real revolution in the last 10 years or so in part led by the Human Genome Project. It has given us enormously powerful tools for deciphering basic types of biological information, again, as you heard from Craig Benter yesterday. The challenge for the 21st century very much is going to be able to manipulate, and I mean manipulate in the best sense of the word, this information so that we can understand fundamental mechanisms in biology. And so we can lead to this revolution that's coming of preventive medicine. And that's what I'll talk about in this lecture. So let's talk about the three types of biological information. The first type, of course, is the information that's contained in our genes and our chromosomes. That's a four-letter language. It's digital in nature. It's digital in the same sense that the computer codes are digital, but it has four rather than two letters. The variation in these letters generates information. And as I said yesterday, chromosomes, the long strings of DNA, are marvelous repositories for a multiplicity of different languages. One of the languages is the language of genes. That's the language in some sense we understand best. In humans, there are perhaps 100,000 genes. And it is these genes that allow us to become what we are. We'll describe that in just a few moments. So the first kind of information is a one-dimensional type of information. Now, these units of information that we call genes can be expressed in different cells and in different combinations. So in a sense, each gene is expressed in a quantized fashion. In a cell, it's either on or it's off. And when it's expressed, it is first expressed as, again, a related four-letter language which we call messenger RNA. And the only importance of understanding about messenger RNA is if we want to look at different types of cells, say normal cells and cancer cells, one of the questions we must be able to ask is what types of different genes are expressed in normal prostate cell in a cancer prostate cell? And we have very powerful tools for asking this question, again, at the level of assessing this expressed information that we've talked about here. But that isn't the second type of major biological information. The second type is called a protein and it's generated when this messenger RNA is fed into a specialized cellular machine that causes the synthesis of protein molecules, which again are initially synthesized as long strings. But their language is much more complicated. There are 20 letters in this protein language, rather than the four letters in the DNA language. And the consequence of having 20 letters is really quite profound because the order in which these letters are associated with one another in a string and the letters differ in charge and size and shape cause that string to fold into a three-dimensional molecular machine. So each unique order of those letters generates a unique three-dimensional molecular machine and it is proteins that are the molecular machines of life and they catalyze the chemistry of life and they give the body shape and form and when you look at another individual, virtually everything you see is a protein. So there are two really interesting questions about proteins. One is given this linear order of letters in a protein, can we from first principles predict how it folds in three dimensions to make a molecular machine? That's called the protein folding problem. And it's a problem that requires the union of computer scientists and applied mathematicians with biologists to solve. The second problem about proteins is given a particular protein shape. How do we know what it does? It's very much as if you came down from Mars and you looked at a car for the first time, could you deduce from first principles what a car actually was all about? And the answer is no. You'd have to see the car in action. You'd have to make the molecule, the car, do what it can do and then assess it and so it is with proteins. But again, there is a beautiful blend of computational techniques and experimental techniques that are necessary to solve the structure function problem. Now biology for the last 30 years has been comprised of studying individual genes and individual proteins and it's led to enormous success. The future is really going to focus on this third type of biology, which is the biology of complex systems and networks. So we see here some subset of the 10 to the 12th brain cells, neurons, that are present in the human brain with their 10 to the 15th different connections. And of course, what's interesting about this network are its systems properties. The systems properties are, for example, memory and consciousness and the ability to learn. The important point is these properties emerge as a consequence of all of the elements of that network working together in some concerted fashion, a system operating together. And indeed, if you or I were to take any one of these nerve cells and to study it for 20 years so that we knew virtually everything it could do in response to appropriate input stimuli, it wouldn't tell us one iota more about the systems properties than we already know because, again, to study systems properties, you have to study the elements of the network in concert. So it means to study systems in the 21st century. We have to have what we call global technologies. The ability not to look at one gene or one protein at a time, but to look at all of the genes or all of the proteins that are involved in a particular informational pathway. And we'll say a great deal more about that in just a few moments. Now, as I've already animated, what has led to an enormous revolution in biology has been the Human Genome Project, this effort to decipher human heredity that started now almost 10 years ago. I was actually fortunate to be at the first meeting ever held on the Human Genome Project. This was in 1985 and Bob Sinchimer then chancellor of the University of Santa Cruz had a $35 million gift and he was wondering whether he should spend that money on an institute to sequence the human genome. So he invited Wally Gilbert and George Church and Charles Cantor and a variety of the other pioneers in the Human Genome Project to consider this whole question over a period of two or three days. And I went, I have to say, slightly skeptical on technical grounds, but I came away convinced that the Human Genome Project was the most single transforming event in biology up to that point in time. And I say that because interestingly enough, it introduced a brand new type of science which I call Discovery Science. So the idea about Discovery Science is that it is a science which is directed at defining all of the elements that are present in an object. So for the Human Genome Project, it was sequencing the three billion nucleotides present in the Human Genome Project. It didn't ask questions. It didn't formulate hypotheses. Rather, it created an infrastructure which is now in the process of actually transforming biology and medicine. The important point was this Discovery Science interplays with the more classic hypothesis-driven science where the investigator poses a question and designs an experiment and attempts to test it. It interfaces with that beautifully because it gives an enormously powerful infrastructure for examining science in a hypothesis-driven manner. And it was interesting in the time from 1985 to 1990. There was enormous hostility toward the Human Genome Project. And it all really focused around a fundamental misunderstanding of the power of this Discovery Science. It was called stamp collecting. It was called big science. It was said that it would take lots of money away from ordinary hypothesis-driven science. There was an enormous antipathy toward it. But what happened in the end is people, rationality won out. A very important committee from the National Academy of Sciences put together half skeptics and half proponents. And it turned out that everybody agreed in the end this was the way to go. So in 1990, the Human Genome Project was generated with the projection of finishing by 2005. And as you heard yesterday, we're obviously going to finish much earlier than that. But let me pose two different aspects of thinking about the genome that has come up in earlier discussions but should be made explicit. In a sense, the Human Genome Project is the most incredible software entity that has ever been devised. It's been devised by molecular evolution. And what this software program does is create human beings starting with a single cell of fertilized egg. And it directs this marvelous chromosomal choreography where in different cells different subsets of genes are expressed. And it is this differential expression that manifests the phenotypic differences, the differences that make a muscle cell different from a brain cell, different from a connective tissue cell. And one of the challenges in contemporary biology that the Human Genome Project is going to enormously aid is deciphering this software program and coming to understand and better how development actually occurs. And in many ways, development and disease are opposite sides of very much the same coin. Now as we heard yesterday, some have argued because of the emergence of genetics in such a powerful way that this famous quote of we used to think our fate resided in the stars, we know now it resides in our genes. And of course, one has to temper this view I think in a very significant way. So here are the left index fingerprints of two identical twins, nine-year-old girls. And as you can see, the fingerprints are entirely different from one another in spite of the fact the genes that make these fingerprints are the same. So either the stochastic processes which lead to the developmental patterns and fingerprints or the interaction, the differential interaction with their environments led to these differences. But the important point this slide makes is for any human trait that we're interested in, we have to ask the question, where in this spectrum of mostly controlled by genes or mostly controlled by the environment does a particular trait reside? And I would argue that we don't have very effective means for ascertaining those answers very effectively at this point in time. Now what exactly is the human genome? It's basically the analysis of these 24 different types of human strings, the 24 different chromosomes that collectively have 3 billion letters of the DNA language. They range in size from 50 million to 250 million. It constitutes obviously an enormous task for the genome project. The genome project is all about four different types of maps for each of these chromosomes. A genetic map that is sprinkling across each of these chromosomes of markers that vary among the human population that can serve as signposts for localizing where genes that control particular traits actually reside. And the human genome project has been marvelously successful in this endeavor. A second type of map is what's called a physical map. That is you take a chromosome and you chop it with these DNA cutting enzymes. And then you attempt to join together as a linear jigsaw puzzle, all of these fragments. So you have in test tubes all of the physical clones that cover each of the 24 different human chromosomes. And ironically enough, even though the human genome project spent probably more than $400 million on this endeavor, this has collectively been an enormous failure. Almost all of that work, for interesting reasons, and Dr. Venner discussed some of them yesterday, has turned out to be irrelevant to the human genome project. The third kind of map is of course a map which allows us to identify the locations and orientations and family associations of all 100,000 or so genes. And the fourth map is this map that allows us to determine the entire nucleotide sequence across each of these chromosomes. In a sense, it's a map at the highest level of resolution. Now, the really critical question we want to ask today is not how are we doing and what is that leading to right now? But what I'd like to ask is if we jump ahead 15 years and we look back on a completed human genome project, let's ask ourselves the question, what were its most fundamental contributions to biology and to medicine? And I believe there were four contributions. And the first of these contributions was a parts list, in a sense, it was creating the periodic table of the elements of life. So just as in the 19th century, the periodic table of the chemical elements which interrelated them one another in very important ways, allowed for a revolution in certain aspects of chemical science. So today, this periodic table of life is going to allow for a revolution in human biology. And what are the elements in this periodic table of life? They are the identification of the 100,000 or so genes. They are the determination of the related sequences that govern when, where and how those genes are turned on and off. This is a second type of language called the regulatory code. They are the ability to take the human genes and the genes from other model organisms and to deconvolute from these genes the basic building block components of these genes, which are called motifs. And an identification of the lexagon, the general lexagon of motifs for all genes and proteins is actually going to be key to solving not only the protein folding problem, but the structure function problem as well. And then this gives us access, as we heard yesterday, to human variation, the variation that makes us different from one another and gives us access to understanding our own unique individualities. It is the key, of course, to deciphering all of these chromosomal languages that we talked about yesterday and we'll talk more about that in just a few moments. But what it also does in the ultimate is it gives us the blueprints for many different organisms, it will ultimately, and it will allow us to do comparative genomics, that is to compare these genomes and to be able to decipher from each of those genomes their logic of life and to understand how the logic of life changes in the course of evolution. And this gives us fundamental insights into how things work as well as how things came to be. Now, the other aspects of the human genome project are this idea of discovery science because an intimate part of being able to do systems biology is having the tools for discovery science. So for systems biology, discovery science and hypothesis-driven science are irrevocably linked if we're really to be successful. It is this periodic table of life that we've talked about. It is the creation of these global tools that will let us look at whole systems worth of elements, and we'll talk about the global tools in just a few moments. But perhaps the most revolutionary thing that the human genome project has done is to catalyze a series of paradigm changes that have revolutionized our whole understanding of biology and medicine and how we are going to go forward with it. So let me talk about a series of these paradigm changes to give you some idea of what they're all about. So the first is this simple idea that biology is an informational science. And we've already discussed how you have major types of information, one-dimensional information in DNA, three-dimensional information in proteins, and time-variant four-dimensional information in systems. But what one also has to understand from this endeavor is that there are different hierarchical levels of information. So the most primitive level of information is information at the gene, and the next highest level is information at the protein. But what is interesting is their proteins acquire a great deal of additional information that can't a priori be predicted merely from the gene sequence alone. That is, proteins are modified chemically before almost any of them carry out their functions. Proteins are processed in interesting ways. And most important, proteins interact with themselves and other macromolecular components to create the machines of life. And thus, there is an accruement of an enormous additional amount of information going from the gene to the protein. And likewise, in going from the protein to the informational pathway, that is, the informational context within which a protein or a gene operates, that is, the set of interactions that lead to signal transduction that lead to a certain aspect of development that can lead to cancer. These informational pathways are absolutely key to understanding biology and understanding medicine. And of course, as you create informational pathways, you accumulate additional information, not directly predictable from the proteins themselves. And likewise, as you put any particular informational pathway in the context of all the other pathways that interacts with, you accrue yet another level of information. So what systems biology is all about is capturing information at these different levels and then being able to integrate that information together to convert information about biology into knowledge about biology and knowledge about medicine. Again, a point that we'll return to later. Now as a parenthetic afterthought, my own conviction is that the way to teach biology, whether it be to K through 12 kids or whether it be to our cross-disciplinary colleagues or to undergraduates, is to view biology as an informational science. And one can teach it in this manner in an enormously efficient hierarchical way, stripped of this horrible vocabulary that is paralyzing for most students that are starting in biology. And indeed, what I've shown you here on this slide are the manipulations of DNA strings, or chromosomes, that explain all of molecular evolution. So I can, in 15 minutes, tell you, if we had the time, I could make you understand all of the events that transpire in molecular evolution. And in a similar slide, I could put up the manipulation of our digital DNA strings in the context of genetics and explain very simply all that genetics is about. So teaching from an informational point of view I think is going to open up new opportunities not only for the education of the young, but for the mandatory integration of many other disciplines into biology. A point I'm going to be coming to in just a few moments. Now the second major paradigm change is one that I've already hinted at, and that's this idea that we need global tools to be able to interrogate lots of genes or lots of proteins that are involved in particular informational pathways. And I'll give you a couple of examples of the kinds of global tools that we're talking about. That is the ability either to study many elements at one time or the ability to study many elements very quickly in a serial manner. And of course the first of these global tools that is beginning to revolutionize biology is large scale DNA sequencing. And the point I'd like to make before I discuss that is this fascinating interrelationship between technology development and the advancement of biology. They are enormously interrelated because biology should drive the development of new technologies that will let us attack the frontiers here to for unattainable. But once we've developed that new technology these frontiers open and you can mine enormous amounts of new information up until you come again up against a barrier and then once again new tools must be developed. So this interrelationship between technology and biology is a fundamental theme of biology as we move forward into the next century. DNA sequencing, how do we carry out DNA sequencing? So let me give you just a general idea of the thinking that went into the invention of this fluorescent based DNA sequencer almost 15, 20 years ago now. The idea at that time was there was an elegant chemistry for taking a piece of DNA and being able to generate from that DNA a ladder of fragments each of which was successively one letter of the DNA language shorter. And I've shown you such a ladder in this slide. One has the ability to separate these fragments from one another by electrophoresis in a gel which sizes them letting small fragments migrate faster than large fragments. So what we added to that ability was, number one, a way of color coding each of these DNA fragments so we could tell what letter it ended in. And we used four different fluorescent dyes to color chord each of the fragments that ended in a different letter. And then we used basically an argon laser to interrogate those bands as they migrated by the bottom of the gel. So we had chemical ways of labeling all of the C fragments with one color and all of the G fragments with another color and so on. We could separate them in a gel and then the laser just read out the color, fed that into a computer and the colors were equivalent to the DNA sequence. And that was the essence of the fluorescent DNA sequencer. Parenthetically, the first sequencer that we developed was a capillary sequencer. And that is the basis for this latest advance that Dr. Venter talked about yesterday where you have not one capillary sequencer, but you have 96 at one time. We have advanced our ability to sequence DNA from that time 2,000 fold, that is we can do it 2,000 times as fast today as we could roughly 15 years ago. A remarkable advance in one kind of global technology. The second kind of technology I'll talk about are DNA chips. Many of you have heard about DNA chips and this technology was pioneered by a company called Affymetrix that actually designed the chips in exactly the same way you design computer chips. They used a process called photolithography. But Alan Blanchard in our lab actually did it a very different way. He used the little pumps that are present in inkjet printers, pumps that have the capacity to deliver 5,000 droplets per second of a solution that has 40 pika liters in volume. So 1,000 of those droplets would make up one ordinary droplet to give you an idea of the size of this. And we could use a combination of 120 of these pumps to literally spray across a glass chip the size of your thumbnail, the ability to synthesize a series of different DNA fragments. And we're developing, as Affymetrix is, the capacity on a single chip, for example, to have 100,000 different fragments of DNA which represent each of the 100,000 human genes. Or we will be able to do that in another year or so. And what that means is we have the capacity, then, by a process of molecular complementarity. And let me talk about that. That is the ability to take DNA, that is the messenger RNA from a cell, and to purify it, and then to convert it into DNA, and then to actually separate the two strands of those DNA fragments from one another. And let the strands inevitably find their right partner by virtue of molecular complementarity. The idea that pairs of letters in the DNA language always associate with one another down the double-stranded helix, which makes up the background of DNA. The A letter with the T letter, the G letter with the C letter. And the basic idea, then, is we can take, let's see, can you put, now I want to go back one, if I can. What we can do, then, is take, as I'll show you in a few moments, the messenger from a normal prostate cell and the messenger from a cancer prostate cell. And we can actually ask for the first time, what are all of the differences in the genes that come about as a consequence of cancer? So this is truly an example of a fascinating global tool. The third technology, which I'm not going to explain it all, but my colleagues, John Yates and Rudy Aversol, have developed parallel technologies for separating proteins effectively and being able to analyze them using a tool called a mass spectrometer. And what Rudy Aversol is in the process of doing is using microfluidics and microelectronics, again, small chips the size of your thumbnail, to actually etch channels on these chips that can carry out the complicated reactions of protein chemistry necessary not just to analyze one protein, but to analyze 10 at one time, 100 at one time, or even 1,000 at one time, and then feed them into analytic tools such as the mass spectrometer. So these miniature tools are going to revolutionize the study of lots of proteins at one time, an area called proteomics, and they'll be applied to genomics as well. We'll use these small tools to carry out sequencing and genotyping and array analysis in a much more effective fashion in the future. So that gives you an idea of some of these very powerful global tools that let us look at a lot of information together. Now, what exactly do we do with all this information that we collect? How can we begin to assemble it into something that's actually meaningful? It's here that we have to turn to model organisms, because the only way to really effectively study complex informational systems is to perturb these informational systems in these model organisms and study how genes change, or how proteins change, or how cells change in those particular organisms. And as you'll see in just a moment, those organisms are really the Rosetta stones for doing systems biology. What is obvious is that the Human Genome Project has catalyzed the idea of model systems by suggesting that we sequence four simple organisms, bacteria, yeast, the worm and the fly, and all of those genomes are done now, and a fifth most more complicated organism, the mouse, that rivals our own genomic complexity, and shares interesting phenomena such as immune responses and cancer and things like that. The important point is if you take an organism like the nematode, it has 19,100 genes, their 70% of those genes have their analogs or their homologs in the human. So it means for many human genes, we can find their counterparts in these simple, biologic and genetically manipulable organisms, and we can come to understand not only how the genes work, but how the nature of the informational pathways within which they operate. And that is the essence of why simple model organisms and more complex model organisms are absolutely critical. And there's a wonderful quote by Max Delbrook, any living cell carries with it the experience of a billion years of experimentation by its ancestor. What we now have the ability to do is make explicit these experiences of the ancestors and in doing so, we come to not only understand evolution, we come to understand biology. So in that sense, you can see these model organisms are indeed Rosetta Stone. So just as with the original Rosetta Stone, knowing Greek allowed us to translate the demonic and hieroglyphic languages, knowing the genome of yeast and eventually if mouse is going to be critical to understanding the genome of human beings. And this is a point we'll return to again in just a few moments. A final point in these paradigm changes is the central role computer science and applied mathematics are going to play in this type of systems biology. And it is obvious I think to all of us that computer science and applied math give us our ability to take biological information at all of these different levels and acquire it and store it and analyze it and model it and display it and ultimately distribute it. So in that sense, they are absolutely critical to what we do, but the counterpoint is also true, namely living organisms have had 3.8 billion years to learn how to manipulate their digital strings and in doing so, they've come up with enormously clever tricks that will be useful to computer scientists in thinking about the logic and the strategy for the manipulation of their own digital information. And this is already given rise to algorithms that have been useful, neural nets, genetic algorithms. And even DNA is a computer, it isn't great guns as a computer, but it's made computer scientists think in quite different ways about the manipulation of digital information. What I think is again, the most exciting opportunity we have for the future is this one of comparative genomics we've talked about before, to be able to take a genome, such as the genome of homophilus, which Craig Venter showed us yesterday, to be able to decondilute that genome into its informational pathways and thus delineate the logic of life, and then to be able to compare the informational pathways and the logic of life of many different organisms to understand how evolution has conspired to give us different approaches to dealing with biological information and ultimately to give us very deep understanding into the complexities of human organisms. Now, the final point that I would make is again, just underscoring the importance of systems biology and the central role it is going to play in the next century. And I want to give you just some feeling for, now, can we go back one? I only want for the analysis of biological systems. So most biological systems that we're interested in are very, very complex. The brain has 10 to the 12 cells, so we can't study it in its entirety. What biologists must do is use biology to devise subsystems that are practicable. Subsystems whose properties, at least in part, reflect the systems properties, but can be approached by these global tools that we've talked about. Once we have these subsystems, what we have to do is use these discovery tools to identify the elements present in the system. Once we've identified the elements present in a system, then we have to use model organisms to perturb the system and to capture the nature and flow of information through those informational pathways. To look at how express genes change, to look at how express proteins change, to look at how cellular patterns change, and then we have to use these insights in what is probably the most challenging tasks that faces biologists and mathematicians and computer scientists for the future. That is, to create models that give us the ability to predict two things about the informational pathway. One, the structure of that pathway and ultimately it's associated interactions. And two, the systems properties that emerge from that pathway given a particular kind of perturbation. I would argue this is going to require completely new types of mathematics because we have, it isn't just understanding how express genes or express proteins work, it's understanding how to integrate information at these very different types of levels. And biological information is enormously heterogeneous. So there are enormous opportunities for computer science applied mathematics for people that have good backgrounds in physics to begin thinking about the modeling. The point that I would make that is paramount is this will never work unless there is an intimate interaction between the biologists and the computer scientists and mathematicians because the modeling process is reiterative. We try something, it suggests experiments, we try those experiments, this refines the model and it goes back and forth. And too much of theoretical biology in the past has been idle speculation about the way nature might work in isolation from real biology and to solve the systems problems the mathematicians and computer scientists and physicists will have to join directly with the biologists. Now, let me just give you an example of a system that we're studying to develop all the tools that we're talking about here. It's a system that's very well studied. We're looking in yeast because it's genome is done and we know it has 6,200 genes and we can make a DNA chip with all of those genes and we can actually use proteomics to study many of the proteins at the same time. And we're gonna look at a system where the sugar galactose is metabolized to capture energy. We know that you have to have transport molecules, we know that you have to have enzymes that operate on the galactose and we know that you have to have proteins that control the expression of other proteins that are involved in this pathway. So can we a priori predict the nature and interrelationships of that informational pathway with no a priori assumptions about the nature of it? And so the approach that we're taking it, we're doing it with two computer scientists, a theoretical physicist and several biologists is to initiate activity in this pathway in yeast and then to sample after we've initiated that activity the yeast at differing points of time and then we'll interrogate the nature of the expressed proteins and the nature of the expressed genes and we'll compare them in a semi quantitative fashion one to another and try and understand the biology of what is operating in this pathway. Now if you think about it with the biological perturbation the things that change are not only the things that are involved in that pathway but reflections of that pathway's interactions with other pathways. So how do we actually begin to define the structure of that particular pathway? And it's here that the yeast community has helped us normally because it's created is in the process of creating 6,200 strains of yeast where each of those strains has a different gene destroyed. So we can take all of the genes from the first experiment that changed in their behavior and then we can interrogate them in the appropriate yeast where their activity is destroyed and if you think about that that not only gives us the nature of the information the structure of the informational pathways they're in but it gives us polarity and it begins to give us insight into connection and these together with other experiments will allow us to begin this reiterative process where we look at data and what always happens when you look at data in the first analysis is you have 100 models that are possible. We are developing computational tools that can optimize the next set of experiments we need to minimize and cut out the largest number of those models. So we go through this reiterative process and ultimately hope to be able to arrive at one or at least a very few models that reflect the reality of the structure of that informational pathway and have built into it predictions about the systems properties that actually emerge from that particular pathway. So in this context let me talk about several systems we're studying using this systems approach just to give you a very high level global idea of what we're interested in and the first is the immune response so the immune response as you know is a response where foreign entities such as virus, bacteria or even cancer cells trigger a complicated set of cells to activate themselves and in doing so you create two classes of specific immune cells T cells and B cells and each of them have receptors that play an important role in specifically recognizing the unique molecular nature of the virus or bacteria or cancer cell that initiated this immune reaction. So how do we and I might say that there are 10 to the 12th of these immune cells present in each of our bodies so how do we go about carrying out this systems analysis and creating appropriate subsystems? Well we with Slasher Rodensky in the Department of Immunology now have a T cell that interestingly enough can reflect four of the major systems properties in the immune response. This T cell in culture can be triggered to go to immunity, it can be triggered to become paralyzed that is to react with nothing, it can be triggered to kill itself or in special circumstances you can actually start it inducing an autoimmune reaction. So we have the capacity then in this model system to interrogate the change in flow of information as we trigger all of these systems kinds of events and what is really striking is molecular immunology is probably the most advanced of all the molecular disciplines because the cells and genes for a variety of reasons were so accessible yet because it's been studied in the one gene and one protein at a time mode we don't begin to understand anything very deep about any of those systems properties and that will change obviously very quickly. Now as you heard from the introduction one of the things we've done is sequence the families of genes that are involved in T cell receptors 700,000 in one case a million in another case 650,000 in another case and the only point that I make here is these contiguous sequences of DNA from a particular human chromosome give us a marvelous opportunity to begin identifying the nature of the languages that exist on chromosomes. So we can identify all of the genes that participate in this immune recognition process in subtler ways we can begin to recognize the sequences that are important for controlling how these genes are expressed we can actually look at these sequences and we can deduce the nature of the evolutionary events that have modified this locus say since the divergence of humans and mice we can actually identify all of these repeat sequences that are scattered all the way across this locus that plays such an important role in evolution and in some cases we can capture the role they're playing in the evolution and change of particular aspects of this locus and finally we can identify the variation in this locus and we're beginning to correlate the variation that occurs with certain very interesting types of autoimmune disease so knowing the sequence opens up our ability to decipher many different languages now a systems approach to disease is I would argue really going to revolutionize medicine in interesting ways and let me just give you a thumbnail sketch of how we've approached it with prostate cancer because it lets you look at cancer in a very different way the idea is prostate cancer arises as the result of a perturbation of a particular information pathway and by that we mean the mutation of a particular gene now this mutation may be either inherited from your family or it may arise during your own lifetime what is almost certainly true now is it's not the perturbation of one pathway but the perturbation of several pathways and that means we have to be able to stratify prostate cancer into its different types for the different types have very different diagnostic implications for the future so how can we then get diagnostic, therapeutic and even cancer relevant genes by using these global tools of genomics well the idea is for example to use these DNA arrays that we've talked about so here is an array that has 1500 genes expressed in the prostate that has been reacted against messenger RNA from a normal prostate and here is the same 1500 genes reacted against messenger RNA from a cancer prostate cell and we've developed computational methods that can look at these two images and then give us on a idealized image all the genes that are either overexpressed in the cancer cell or overexpressed in the normal cell and it is from these genes that come the candidates to actually look for the diagnostic markers we need for stratification that is the identification of which type of prostate cancer you have or another parameter for diagnostic analysis is progression each prostate cancer progresses down a series of stages and we have markers that can begin to identify the different stages as well therapeutic targets there are unique markers that occur in some cases only on cancer cells and if they're on the cell surface with appropriate types of immunologic intervention there are very powerful new ways of thinking about brand new therapeutic approaches and this is based all on this discovery science and identifying markers that have significant potential for the induction of particular immune responses and we've identified a whole series of genes that may be fundamental to the biology of how the prostate cancer process actually arises Now scientifically the final thing I'd like to talk about is the revolution that I think the genomics and this new systems biology is going to have on medicine and I think it's really going to be profound and I suspect it's gonna come much more rapidly than any of us think about so here's Willie Shoemaker and here's Will Chamberlain those are individuals that lie within quote normal variation genes almost certainly are the dominant thing that make them different from one another I doubt very much that there are large environmental influences in this particular case and likewise so it is with disease the breast cancer one gene was discovered now about four or five years ago and it's a gene which when present in one bad copy in women as the 70% probability of giving you breast cancer by the time you're 60 years of age now the process is very complicated but I don't think we need to talk about is it causing or does it predispose or does it help whatever you wanna say this has a absolutely profound impact on families of women that have this particular defected gene now what is interesting about this predictive medicine is the idea that it isn't inevitable in this case there's a 70% chance and so the really intriguing question is why do 30% of the women not get this disease so one idea is that there are is a primary predisposing gene but there need to be other modifying genes that have been changed that operate in concert with it and this almost certainly is going to be the explanation at least in part for a lot of other diseases and of course the other idea is to say that there are environmental influences that operate on that particular gene and that's certainly something that has to be investigated as well but as we study particular genes such as Alzheimer's disease we can begin to see that there is this process of stratification that I've talked about where we've identified dominant genes within families that cause differing forms of Alzheimer's disease and presumably there will be associated with each these modifying genes that are necessary for actually realizing this disease process and it's my prediction and you can call it 20 years, 30 years whatever you like that we will have identified hundreds of genes that predispose in a general way to many of the common diseases today certainly cancer certainly cardiovascular diseases many of the immunologic diseases metabolic and certainly I would guess a lot of the neurologic diseases as we design more and more effective assays for looking at them but we will as Craig Venter showed yesterday be able to put this information into a computer and read out disease susceptibilities that is the probability that you are likely to get a particular disease and of course the real key is that for each of those we want to have a preventive measure that allows you to circumvent the limitation of your gene and these preventive measures will be the manipulation of biological information at all of these different levels at the DNA level, at the protein level and even at the level of systems and the key to being able to effectively intervene is going to be this systems approach to disease thinking that a gene that causes cystic fibrosis is of itself going to give you any insights into how to treat that disease as we've seen for the last 11 years is not an effective way to approach it you have to put it into the context of an informational pathway and in doing so you gain many other alternatives for manipulating and circumventing again the limitations of whatever genes that we might be talking about so this will move us eventually toward a medicine where a very large focus is going to be on preventing disease identifying early the potential for disease and being able to design regimens that will circumvent these types of limitations and this, the whole focus then is going to be on keeping people well this so-called preventive medicine and as you can understand that's going to have really profound implications on medicine itself how do we deal with individuals that can live significantly longer but not only live will be creative and physically alert able to contribute actively to society we don't treat elder people very well at this point in time one of my favorite questions is to ask physicians what their job description will be in the year 2020 and most of them don't come very close to the enormous revolutions that we see will occur how do we educate society to the things that are going on and I think with the web that probably isn't going to be a problem a bigger problem is going to be dealing with physicians that are made insecure by patients that know more about their disease than the physicians do and of course there are as we heard yesterday a whole series of these ethical, legal and social questions that we have to consider the privacy of genetic information is an area that there is legislation now being winding its way through Congress some of it very very poor I might add the idea of genetic counseling how do you explain to people that don't understand probability what it means that you have a 50% chance of getting this disease and how do you help them to decide to do what is necessary to limit these diseases and I think the answer in some ways is going to be very very simple regimens because if you look at smoking you can see that even the best campaigns haven't been completely successful if we know all the human genes we certainly can diagnose and utero any simple genetic disease and should there be limitations as to the nature of therapeutic abortions that can ensue from this knowledge germline engineering as we discussed yesterday with some of you in many ways it's much easier than the kind of somatic engineering that's being discussed as therapies for cancer and a variety of other diseases and it's simple because it operates on just a single cell the fertilized egg so and there are two aspects to germline engineering do we want to repair defects that are present in particular families or do we want to enhance human traits and again for enhancement I would argue the things that we really value are complex we're not gonna understand the systems that affect them probably for our lifetimes and that means we can't certainly do this in a capricious manner but even more interesting is the question of how do we make decisions about all of this I mean ultimately that's going to require an informed and educated public thinking in a rational analytic manner about these things cloning humans, genes that control behavior should there be areas of forbidden knowledge are there things we shouldn't experiment in depending on whether republicans or democrats are in office we can do fetal research or we can't do fetal research and is this a good idea is this a rational type of idea and then a question that came up yesterday is how far in systems biology will reductionism take us I think a really interesting question to think about my own feeling is that scientists have an enormous obligation for educating the public in the department I started I argued that all of us should spend five or even 10% of our time in this endeavor the most effective way of educating the public that we found is through K through 12 science education and let me tell you just a bit about several of the programs that we've developed there our philosophy is to use a strategic approach to it so we teach hands-on inquiry based science to teachers we do it with lead teachers and scientists and workshops and in-service training and a critical point of the strategy is to have the community become involved and embrace the kinds of changes that are occurring we believe the changes should be systemic so virtually every program we operate on changes whole school districts you don't do one classroom or even one school for example the first program that we started the elementary program has 66 schools 1400 teachers and 23,000 kids and we're almost finished with that program which I'll describe in just a moment we think you should start with elementary and move to middle school and ultimately to high school in a sequential manner so that you capture and keep motivated the kids that have been educated at these earlier levels and it's all for naught if you can't create an environment of sustainability and the key to sustainability is convincing your public partners that this is essential whether it's business, whether it's the industrial community whether it's the general community at large and we spent a lot of time doing this in Seattle and of course this program by definition has this important societal role that it captures kids at the very beginning hopefully before there's been much stratification into the haves and have-nots so we have underway now a local systemic change program in Seattle for elementary schools that's in our fourth year and it's been an enormous success in revolutionizing science the interesting thing about elementary teachers is only 4% of them have had science so it means almost none of them teach science there was initial hesitation in the programs that we generated I can say now there's enormous enthusiasm as they see how these programs can revolutionize not only what we can do for the kids but what we can do for the teachers themselves in terms of their understanding we've just started a middle school program now and we're just putting in place a high school program that once again will be a local systemic change initiative so to conclude let me make just a few points about where this biology is taking us Gordon Moore in 1970 made a prediction that for every 18 months the number of transistors you could put on a computer chip would double that single change has driven the revolution in communication and information technology that the world has seen in the last 30 years a remarkable revolution if we look at DNA sequence information we see if anything even a sharper exponential and the critical question for biology for biotechnology and for medicine is how do we convert this information into knowledge about biology or knowledge about medicine and I would argue we do it through systems biology we do it through the integration of genomic information and proteomic information and cellular information we do it through the perturbations of model organisms and so forth to create these systems approaches toward the complexities of biology and toward the complexities of disease and indeed if you think about it there are a whole series of hierarchical levels of information that I haven't even discussed cells to tissues, tissues to systems but then systems to individuals individuals to populations and populations to ecologies and one of the fascinating questions is the extent to which we can from this bottom-up molecular approach encompass successively higher levels of these hierarchical systems of information and that represents an incredibly exciting challenge for all of us and what is going to be critical for that is the deployment of the most powerful sophisticated tools micro fabrication, nanotechnology, computational biology and on and on and on we are going to have in very effective ways begin to integrate into biology these other disciplines of math and physics and chemistry and computer science to give us the people that can turn these leading edge technologies toward the enormous challenges of systems biology and indeed in that regard together with our department we've started an institute at the University of that is independent from the University of Washington that is going to be dedicated to this systems biology we've started the institute because these things have to be done on a scale that are not consistent with academic ventures at this point in time and they require a flexibility and entrepreneurial spirit that is very difficult to achieve in large organizations but the idea of the institute is to have half the faculty be biologists that do systems biology and the other be these cross-disciplinary faculties that are going to create facilities and indeed create the global technologies of the future but the essence of what we hope to do in the institute is to create novel, unique and powerful partnerships with academia, with industry and with society in some of the ways that we've talked about here so as an opportunity for students from many different disciplines I would suggest this represents an exciting vision for the kinds of things that could be done but the bottom line is we can begin to think about taking on these most challenging problems of human development and human disease and we can begin to see at least a way that we can begin chipping away at some of the most challenging, fascinating of systems problems. How complex will they be? How far will they take us? We don't know but the important point is they will clearly revolutionize biology and revolutionize medicine. Thank you. Thank you, thank you. Ladies and gentlemen, those of you who purchased tickets for the Nobel bag lunch, please pick up your lunch at the forum which is the space like this only on the north end of the building. Please, however, observe what it is you're eating because diet is as important as genes. Heresy. Okay, as we did yesterday, ushers will be in the aisles to collect questions that you can write up to send up here to our participants or more specifically, Dr. Hood. And while we're waiting for those questions to come forward, I'll ask anyone here and Dr. Venner has already volunteered to address comments or questions to Dr. Hood. We'll take a moment for things to quiet down a little bit before we get going here. Right, you're supposed to do something like such you know, is there something like that? Dr. Venner. So, Lee, I thought in your terrific talk, one of the things that's obviously key to the future of funding science is the nice delineation you made between discovery science and hypothesis-driven science. How are we gonna change the funding agencies to recognize the importance of discovery-driven science which is probably gonna drive most of the discoveries for the next several decades? So, the question of how to change the federal funding agencies to fund discovery research is an enormously challenging one. What we see now is our incremental moves of a number of the institutes at NIH toward funding small discovery projects. So, the National Institute for Allergy and Infectious Diseases funding projects on the genomes of pathogenic organisms at a low level. The National Cancer Institute is actually funding a number of discovery projects that have to do with EST and so forth. But if you look collectively at the funds that are directed toward discovery science, it's amazingly small. And my own feeling is that discovery science is going to be driven by private institutes that raise money, as you did with Tiger, and by companies perhaps much more than it is by academia until the funding agencies get it. And I think there are long ways from getting it. Other questions or comments? I have a question. I was really fascinated by this educational program that you're doing. And I think it's fantastic that you got the entire school system involved. I wonder if you could tell us a little bit, first, about how that's funded and who's supporting it. Because obviously it's a big effort in the schools and these are public schools. And second of all, how you deal with issues like, for example, in Kansas, even teaching about evolution now is controversial and whether or not you've encountered any political resistance or how you would deal with any political issues in the life. So for the funding, for example, of the elementary program, we have a grant from NSF for $4.5 million over the five-year period. And all of that, all that pays for are the salaries for the teachers that go to the workshops. So the community has raised a matching level of funds that is approximately a couple of million dollars. And we do that, for example, the university contributes an enormous number of graduate students and post-docs who are the scientists that help with this program. We have, in our department, 10 full-time people working on outreach and we contribute quite a significant fraction of several of their times to this endeavor. And we've been very successful at going to industry to get support for this program. But in the long run, what is absolutely critical for sustainability is that you convince the school board and the school district that they have to pay for these things because we can't go on raising this kind of money forever. And the simple fact is, for most schools, they do have enough money to do things right. They just spend it in the wrong ways. And getting them to spend it in the right ways is an enormous challenge because there are very powerful vested interests, obviously. As to the question of creationism and or other forms of anti-science, we haven't seen that. I think Seattle is kind of an unusual, technologically-oriented city and probably, I would guess, we're not going to see this kind of difficulty. Our ambition is to use this program in Seattle as a model and spread it to the rest of the state. And I will say the eastern half of Washington is enormously conservative and has a large number of kids from creation homes that are self-taught. And I think that represents a challenge for the future. And how you deal with that is, I don't know exactly what we are gonna be able to do there. But the only hesitation we've seen on the part of teachers is a hesitation that comes from insecurity, knowing they've never had science before and can they do it. And we've set up a program that's really accommodating to bring them in and make them feel comfortable as they learn the science and then assure that they have success and appropriate support in that first year of teaching and everything. Anyone else? Can I ask a quick one? I want for a minute just to play scientifically Dean's talk yesterday where we had a gene, apparently, which had, by many standards, a major impact on a complex phenotype at the end of a very long chain and your delineation of the multiple layers that intervene between the DNA and the phenotype. And could you speculate for a minute about the justifiability of trying to make the kinds of connections that one sees in a lot of academic science at the moment, which is, in a sense, very simple in mind, which says here's a gene, here's an outcome. Let's see whether that gene has any impact on a complex outcome. Do you see those incompatible? Not at all. So let me tell you about what I think, in many ways, is the real revolution that's going on now. In the mid-19th century, we knew about sound waves and we knew about light waves and we knew about all of these types of phenomena from the point of view of theoretical physics. What's happened in the last 100 years is engineering, electrical engineering and so forth, has taken those phenomena and has in a descriptive, imprecise way, harnessed them to make all of the things that we know about today. Now, they don't begin to understand in any very sophisticated detail all of the kinds of interactions and so forth, but through engineering approximations, they've gotten a lot done. I think biology stands poised at exactly that position today. So I will say, I think everything's terribly complicated, but on the other hand, I think there are genes that will have enormous consequences and these are genes that you can use to do interesting things even in the light of not completely understanding all the systems, biology and all the interactions that go on. So how we bring biology to this applied area, how we can start creating these mathematical models of systems and so forth that are never gonna be precise. They're gonna be somewhat descriptive, but they will be useful even if we don't understand the detailed molecular mechanisms that transpire between the individual components in an informational pathway. And they'll be useful because we can make predictions about what we need to do to modify the pathway. Can I deepen the question a bit? Sure. Because I think really I was asking you to comment on the nature of the system where you have multiple steps in which there are probabilities and uncertainties and opportunities for stochastic intervention that would allow us to, would make it indeed conceivable that Lee could, that Dean could see the finding that he saw. Can you sort of comment on how that, whereabouts that gene must be working in that very long sequence and how it must be working to do, to have a simple detectable effect on a complex phenotype? You know, I think we can't. And I think we can't because we just don't understand the systems. But again, I'll reiterate, if you have a gene that has a really dominant effect, whatever its explanation is, it's conceivable it can be used to do interesting things if you can manipulate it or if you can use it. But I think particularly with the brain, I mean, sorting out, I mean, what makes the brain so difficult to study in my view are two things. One, the bewildering complexity of different kinds of cell types. And two, the fact that they're fused together and we don't have any way of getting homogeneous cells of type X to look at. I think that is going to be one of the most fundamentally important things in neurobiology that we'll have to figure out how to do. And there are some great ways to think about how to do that. What has revolutionized molecular immunology was the ability to get homogeneous cells and homogeneous proteins. And we could come to understand in the one gene, one protein era, a lot about how these things actually operated. But the real barrier in the nervous system is how can we dissect out the components so we can look at it? I mean, you can do an EST project, you can take a whole brain and you can be assured that you're probably looking at a gamish of genes expressed and you tell me, 1000, 10,000 different cell types or something like that. So how we get to the, I mean, what was interesting in the prostate cancer studies that we did, I didn't have time to talk about it, is we actually dissected in the best way we possibly could with a terrific biologist, pathologist, prostate tumors, typical prostate tumors that we got out. And we found that less than 25% of the cells in these highly dissected things were prostate cancer cells. They were stromal cells, they were a variety of other things. So what we did is we used cell sorting techniques and we now have absolutely homogeneous populations of tumor cells separated from all these other cells. And we get quite different answers in some regards when we look at those populations. So that is one of the challenges that I think is really, so what we can do that's marvelous is genetics doesn't require this kind of fractionation, but to learn systems, you've got to be able to move in that direction some way. Okay, Dr. Fawkes-Killer. I want to raise again in another form, a question I tried to raise yesterday and let me put it in a little bit of context. There is, many people have worried, have spent a lot of time worrying about the ways in which one guarantees a consonance between the course of scientific research and the interests of the society funding it. And this problem, the ideal of scientific research being funded by the federal government is a step toward guaranteeing this convergence but it's by no means perfect. As our scientific research becomes more and more dependent on the private investment, particularly in the kind of discovery science that you're talking about, where's the form for asking about the choices that are made, the directions of the research is taken. Obviously industry will fund, needs a motivation for funding this discovery research. What's in it for them? And how do you maintain some semblance of democratic control or regulation over the course of scientific growth? So I think that's really an excellent question. So my own feeling is the funding at the federal level is going to continue to expand and keeping with the economy and so forth. And I think that funding will be subject to the classic peer review, broad discussion, kind of constraints that we've talked about for a long time. And I think what's going to happen is slowly the federal government agencies are going to come to realize that there are other and newer kinds of options and they're going to start thinking more in the discovery level. So I think it's really a matter of timing. What some of us would like to do now, Craig did it eight years ago and we're kind of doing it now, is have the freedom to explore some avenues that we think are particularly, particularly opportune. And it really I think is going to be a question of timing because there's no question that success brings conversion and it just takes a while to get there. I remember the discussions back when you'd know better than I in the 50s and 60s when the biochemists were utterly convinced that this molecular biology was a passing fad that would absolutely go nowhere. And it took a while to get through the power of these kind of opportunities and it was basically done by doing science that convinced people that these were opportunities. So I think the federal programs will continue to fund in accordance with peer reviewed principles in accordance with convincing Congress that their programs are appropriate and the like. But I think as some of these outliers succeed, they'll raise new options that will be taken perhaps more seriously than they have been now. I remember that resistance for the human genome project was absolutely enormous. I mean, I spent five years, one of the most depressing experiences I ever had. I went to Woods Hole in 1987 and I gave a lecture on the human genome project. And the first question after this lecture was a student got up and said, you said, automated this and automated that. He said, where's the humanity in your science? And it went downhill from there. It was one of the toughest. And so I think new ideas are really hard to get across. So I don't think they have to be a prisoner to the bureaucracy understanding the power of these new ideas. I think we'll go to a couple of questions from the audience. The first one here is can you suggest how this new set of genetic tools can help us understand systems above the level of the individual organism? For example, interactions among individuals, populations, et cetera. You know, that's really a good question and it's one I haven't spent a lot of time thinking about, but I think what is clear as we look at the genomes of hosts and genomes of parasites, we can begin to understand the nature of that kind of ecological interaction occurring. And as we look at the express genes and the changes in protein patterns and so forth, exactly the same as true. So at a simplistic level, I think we can begin to see how at least limited interactions between different kind of organisms can be studied. In the global sense, I have always thought one of the really interesting things to do would be to have the capacity to skin, do an enormous scan of microbial organisms that are present under the ground or at these volcanic vents that Craig talked about yesterday. That is, again, just to capture a lot of information and begin deconvoluting it into this logic and strategy of life. As we gain more and more power to do that with genomes, then we can think more and more in the context of how these logics of life are responses to the environments that these organisms actually live in. Now, as many of you know, there are some really interesting approaches to ecological shifts in population using chaos theory and things like that. And of course, that's really at very high level. And the challenge is where we can reach across and bridge these gaps. And in the brain again, I think that one of the bridges that's gonna be critical is the bridge to be able to get homogeneous cell populations. We have in the Department of Zoology at the University of Washington some terrific ecologists and we're just beginning to talk with them and learn more about how they approach things. My feeling is we can give ecologists really powerful tools for asking more molecularly oriented questions that might get us more quickly to the bridges that'll form between the low level and the high level. Here's a question that several people might wanna address. Please comment on the Icelandic sale of its genetic sample store to a drug company. It's comparable likelihood for good as ethical sacrifice of standards. So the Icelandic company is called Decode. It was founded by an Icelander who was at University of Chicago and then at Harvard. He went back to, so the Iceland genetic resource is absolutely an incredible resource. So Iceland was founded about in the 900s and they have lineage histories all the way back to the beginning, okay? Absolutely spectacular. And in fact, in the 30s, they started keeping pathologic samples from many, many different patients. So they have an enormous wealth of material. The population is small, it's about 270,000. People are more than willing for the most part to be participants in this study. What is so critical about this population is one, it's genetic homogeneity and two, your ability to stratify the population. So let me give you an example and then I'll talk about the consequences. Initially, when Decode went there, they looked at a whole series of families that had a history for preeclampsia and when they did these genetic analyses, they got nothing clear. There were no good law scores or anything. So what they did is they went to the genealogies and they went back and saw in the 1600s, these families that they looked at really came from three branches. So then they did the analyses on the families present in each of these three branches and genes popped out all over the place. So the ability to stratify in relatively homogeneous populations where you can stratify by genealogy is an incredible resource that, there are a few other places that kind of have it but none has it this powerfully. So it's a really valuable resource. So what happened is Decode went back and got set up, made an agreement with Hoffman La Roche and then they approached the Icelandic government with a proposition that they'd give the government a hundred million dollars if they had unique access to the database of genealogies and so forth. And this was something that had to go through Congress and what was interesting is the president of Iceland actually had to come out to Seattle and I had dinner with him and invited a bunch of other people and all of us to a person, biotech people, academic scientists, people interested in ethics, all of us to a person argued that look you can't do this and the reason you can't do it is it's an incredible treasure and any company can only explore some very small fraction of all the potential that's there so why would you lock all this away? Leave it as an open resource. If you want give this company Decode special rights on certain very specific things but don't make this blanket promise and unfortunately the government did pass the law, they did agree to this kind of constraint but what's happened that's very interesting is a second genetic engineering company has been started there recently by a friend of mine down in San Diego and he really thinks that the laws can be circumvented and this is not exactly a stranglehold so it'll be interesting to see what happens but it is a really unusual resource and I think locking up any resource to any one entity like that is really a national tragedy. We tried to argue in the end, I mean the Icelandic government was enormously attracted to a hundred million dollars which isn't very much to us but we argued in the end the rewards will be much richer if you don't restrict the information. One more? Okay, I think we have time for one more. This question refers to the nematode genome that you spoke about today. If there's 70% homology between nematodes and human genetics, do we now know where the similarities lie? That is, where is the worm within us? Gosh, that's a great question. I guess it's scattered around throughout our entire genome. Worms diverged from humans, probably, gosh I would guess 600 million years ago, probably at the Cambrian explosion of life, the most fascinating period in the evolution of life period. So what we've retained in that 650 million years is the reflections of the information that are important both for nematodes and for humans. What we haven't retained are blocks of organizational similarity and things like that. Interestingly enough, in vertebrates, most vertebrates we do see these blocks of similarity that even though divergence there could have occurred 400 million years ago, still have been maintained. And one of the really interesting questions about blocks of similar genes living together is the question of whether they work together or not. And that's one of the major unanswered questions that we have no idea about. So all we see reflected from the worm and from the fly and some of these other invertebrates are the genes themselves and their organization has been certainly rearranged many, many times over. But as we go higher up the scale, the blocks of sharing information really, really becomes striking. The difference between humans and mice, for example, is basically as if we took our chromosomes and randomly snipped them into 150 pieces and then reassembled them back into the 21 chromosomes present in mice. So it's just big blocks of information that have been assembled in differing combinations. And these so-called centenic relationships give us really, really wonderful insights into organization and will give us insights into how genes are regulated as well. I wanna remind you that music will start at 1.15 and at 1.30 we'll have Dr. Blackburn's talk and let's thank Dr. Hood one more time.