Tip:
Highlight text to annotate it
X
Francis Collins: Well, it's been a long day, I would bet, for
all of you, and there's always the risk of being the last person in a symposium of this
sort, that everything will have said by everybody except the last speaker, who then has to say
it again, and I will try not to make that error. It is at least an opportunity for me
to say something from a personal perspective and not try to summarize the meeting. I've
been asked to do that occasionally at meetings that I haven't attended, and that is truly
a challenging task, so I won't try to do that.
I regret I missed all of the wonderful presentations because of a commitment made many months ago
to be at a meeting in New York, and I was in New York until 1:00 this afternoon, so
thank you, Acela [spelled phonetically], you actually were on time today, even if this
symposium wasn't quite, but that's all right.
So I do want to say a few things about looking back, and looking forward. It's going to be
a pretty high-level view, because I think at this point that's probably what you would
want to hear from the NIH director who formally had a hand in this remarkable experience called
the Human Genome Project. So it will be a little bit of, "where did we come from," and
it will be a little bit of, "where we're going to go." And some of it will be serious, and
some of it, frankly, will be a little bit silly, because at this point you're probably
ready for some of that, too. So as soon as I figure out which of these buttons...
Well, this is what we're here to celebrate, and I did wear my DNA tie, and if nobody looks
too closely, you won't notice that one of the three strands here is actually a left-handed
helix, pay no attention; two out of three isn't bad. And, you know, the people who designed
these things apparently don't understand how critical that is for us to be able to hold
our heads up, or wear our appropriate garb. But here we are, and it is pretty remarkable
to say it's been 10 years since that conclusion, the first time we said we'd finish the Human
Genome Sequence and sort of meant it, April 2003. But, of course, that didn't come about
without a great deal of discussion and deliberation, and David has already referred to the Albert's
panel. And this is the Albert's panel publication there in green, which was very much an important
step in giving this genome project some specificity, some credibility, some scientific milestones,
a course over which one could actually see things go forward and know whether it was
happening.
At that same time, although not nearly as much referred to, the Office of Technology
Assessment came up with a similar report, and that, I think, made a turning point. But
frankly, for those of you who weren't around, even after the publication of this, I would
say the majority of the scientific community was either skeptical, or, actually, frankly,
opposed to this project going forward because of the concern it was technically unfeasible,
it might take away the money for their RO1s, or it was just frankly so boring that nobody
who was a good scientist would want to work on it. There was that one, too, which was,
I thought, rather patently offensive, but there it was. And fortunately, that did not
come true because you could argue, I think, quite convincingly, that one of the reasons
the Genome Project succeeded was because it was so compelling, it was so game-changing,
it was so interdisciplinary, it had such opportunities for creativity from multiple perspectives,
that it attracted some of the best and brightest scientists of our generation. Whether or not
they had their arms twisted by Jim Watson, they all came, and they came enthusiastically,
and they stuck with it, and they didn't worry too much about who was going to get the credit,
because that was obviously going to be a big team enterprise.
So it got underway, roughly, in October of 1990 as far as the NIH part of it, and a couple
of quick milestones from that 13-year enterprise, but I'll only hit upon maybe one or two. I
do think it's appropriate to point out that not only did the Genome Project break new
ground technologically, but it also broke new ground in terms of data release, data
access, openness, giving the data away. In this particular slide, a picture of the people
who attended the Bermuda meeting in 1996, where this decision was made; also includes
a photograph of what was written on the white board by John Sulston and Bob Waterston, leading
a session that came to the conclusion that all the data from the Genome Project should
be released every 24 hours, and there should be no filing of intellectual property claims
upon this information. It ought to be just out there for all the smart people in the
world to begin to figure out how to use it. And again, that has really been, I think,
a seminal moment in that it has changed the dynamic for genomics, and increasingly for
other aspects of biomedical research as well, into one where if you're involved in a large-scale
community project of this sort, it is your obligation to make that data available as
soon as you're sure it's right, even in advance of publication, and I think that has enhanced
progress in ways that cannot be overstated. Oh, well, other milestones to mention. Well,
after a while, we actually did get the sequence put together in a draft form. We had a nice
party on June 26th of 2000, even though John Sulston famously said we were all a bunch
of phonies because we hadn't actually published a paper at that point, and there was a little
bit of a problem there, but we got to that. So, February 2001, this paper describing a
draft version of the Human Genome Sequence with a cover, intentionally designed to convey
the idea that this was DNA, but it was also about humanity, and we had a little fun with
that. If you've stared at that cover closely, you may find there are a few surprises. There's
one of them, just sort of hiding in there in the mosaic, Watson and Crick, whose discovery,
of course, we celebrate today, 60 years after the publication of that April 25, 1953 paper,
which essentially got this whole thing started. And in 2003, 10 years ago, the finished version
of the Human Genome Sequence, possible to put out.
And some people said "Okay, that's good, we're done with that. Now we're in the post genomic-era."
And that's when I started to go nuts about that phrase, and I will continue to for many
years to come, because until we've actually figured out the genome and how it functions,
which I think David has made rather clear we are a long way away from, we're still in
the genomic era, let's be sure. We are happy about that, we're celebrating that, we're
making the most of that. We did, of course, not stop there. The focus on human variation,
a natural place for attention to be turned, and an opportunity also there for a large-scale,
coordinated international effort, and a lot of technological developments in order to
speed up the ability to do SNP genotyping, coming forth as the HapMap project, and moving
quite swiftly, creating an ever-deeper database of human genetic variation, and making it
possible, as you heard, I'm sure, from Nancy *** earlier today, to make a wide variety
of discoveries about common variants that are associated with risks of common disease,
here depicted by this myriad of colored circles, now numbering well in the thousands. And yes,
while many of these do, in fact, convey rather modest odds ratios to say the least, they
are, nonetheless, clearly statistically significant. I should mention that one of the hoped-for
outcomes here has been slow to happen, namely to use this information to try to identify
new possibilities for intervening for those common diseases, and I'm having the experience
right now, finally, after a few years of wheel-spinning, of working closely with 10 pharmaceutical
companies, heads of R&D, to try to figure out how could one apply a filter to this set
to actually identify those GWAS findings that are most likely to be pointing to a drugable
target that we didn't know about. There must be some in here, because the positive controls
turn up on this list. If you're looking for evidence that this might be a good way to
find drug targets, it will probably encourage you to notice that most of the known drug
targets for things like cholesterol or diabetes turned up in this particular survey, without
our having biased the situation to make it so.
So while this has, thus far, been, I think, somewhat disappointing as far as predictions
of individual future risk, because the odds ratios are quite small, and we're still missing
a lot of the heritability and a lot of the environmental influences, it still seems to
me that perhaps the major advantage here is to understand disease pathogenesis and pathways,
and to utilize that information to come up with new ideas about therapeutics. And we
really haven't taken advantage of that as we might have, and it's time.
Of course we needed to know as much as we possibly could about genomes, and what we
could learn in terms of full sequences. And the folks that have been conducting that deserve
a lot of compliments and applause for what's been done. And I guess just today I learned
that the low coverage in exome sequencing of 2,500 samples, which is one of the goals,
has now been placed up there on an FTP site, so that there is even more data than before,
and this is teaching us a huge amount of interesting things about genetic variation across the
world.
Already mentioned, and I'm sure described in much more detail by Levi Garraway, a direct
implication and application of the ability to do high-throughput genome sequencing at
increasingly low cost, has been the application to cancer, and I think we all see those datasets
coming forward at increasing speed, with great hopes that this will lead us to not only better
ideas about how to identify subsets of cancer, which is happening almost immediately when
one has these tools, but also to target therapy more effectively. And we are now somewhere
in the neighborhood of having 11,000 tumors in the pipeline to be completed by the end
of 2014 through the NCI/NHGRI effort, The Cancer Genome Atlas, and more to come through
the international effort.
But I think it would be remiss to simply talk about DNA sequencing as a study of the genome,
as our goal here for genomics in this genomic era. Clearly, I will align myself with a lot
of what David said about the importance of understanding function, and pathways, and
networks, and systems, and I have found the ENCODE project to be enormously inspiring
because of the ability to begin to do some of that, and to develop methods that allow
you to begin to see it is how variation connects to gene expression and many other things as
well. And so ENCODE, now having produced remarkable datasets, including the modENCODE effort,
and I'm sure you've heard more about this, seems to me to stand also as a major contribution
to our understanding of what we should be doing to get the genome's function in front
of us. Of course in 2012, last year, a big outpouring of results from that, 30-plus papers;
so many papers and so much information, in fact, that Nature had to put up this explorer
to allow people to dig through the data and find the connections. If you were looking
for particular kind of information, if you haven't the chance to play with that, certainly
people in my lab found that extremely interesting as we are trying to sort out how it is that
the pancreatic islet utilizes it genome in health and disease to make insulin, or not
enough.
But, of course, this is, I guess, just one example. All of these things that I'm talking
about, about the proliferation of datasets that we increasingly depend on to understand
biology. And we, at NIH, became increasingly alarmed about whether we had our house in
order as far as the computational aspects of that; and so we invited some experts, including
some people who are here, to advise us about what we should be doing about the so-called
problem of big data, which I don't think of as a problem, it's an opportunity. But it
will be a problem if you don't prepare for it adequately. And this big data explosion
presents us with challenges, both in terms of algorithms, in terms of hardware, and in
terms of training. And NIH received this report from the Data
Informatics working group, now not quite a year ago, and we have acted rather swiftly
to take action upon it. We've put together a set of internal governing bodies that go
across all of the NIH institutes, and started a new trans-NIH initiative, which is called
BD2K, Big Data to Knowledge, which includes the potential of funding centers of excellence,
but also a very major focus on training. We don't want to see NIH-supported trainees coming
out of graduate training or post-doc training without being skilled in computational approaches
to biology. They will simply not be able to be competitive or be able to make the kinds
of insights happen that we're counting on. And we have a new leadership position, the
associate director for Data Science, which you can see is very nicely abbreviated as
ADDS, because it is supposed to be mathematical, and we having an acting ADDS, namely Eric
Green himself, who is both getting this enterprise going, and also co-chairing the search committee
to try to replace himself as soon as possible, because I imagine he's wearing enough hats
already. But anyway, if you have great ideas about who would be the perfect person to step
into this role, I'm sure Eric would like to hear from you.
So all of this has, in terms of the medical applications, resulted in pretty dramatic
consequences. We heard from David that perhaps there are 5,000 Mendelian disorders for which
we have the known molecular basis, and, in fact, this comes out of OMIM, it's just about
that, maybe 4,800; we are all pretty much in the same place here on these numbers. And
look at the way in which that has happened, beginning in the 1990s. That's not by accident,
that's because, first, genetic and physical maps, and then increasingly sequencing abilities
came along and made this kind of effort possible. And, of course, it's accelerating now with
the ability, with even rarer conditions, to be able to use exome sequencing on very small
numbers of families, sometimes even one or two, to be able to find a mutation that is
causative of the disease. And that is going to cause this curve, I suspect, to jump up
a bit more steeply in the next couple of years, and that's a good thing. And in a few instances,
at least, that has resulted in dramatic outcomes in those instances where that discovery in
the circumstance of an undiagnosed disease leads you to the idea of an intervention that
you would not have come up with otherwise. Here on this stage, back in the fall, we had
something called The Celebration of Science, and at the conclusion of that, perhaps sort
of the most impressive example of how this sort of genomic sequencing has changed lives,
we had on the stage the Berry family, and the two kids there, not the one in the middle
who's the older sibling, but the other two, who are twins, came out after their mom and
dad told the story of these two who developed an increasingly awful form of neurologic degeneration
and dystonia, and were found, on sequencing of their genomes, to both be homozygous for
the loss of function in a previously undescribed pathway, which basically meant that they could
not make either dopamine or serotonin. And simply by supplying them with both L-dopa
and dietary supplementation of 5-hydroxytryptamine, these two went under a rather remarkable,
almost miraculous, recovery in the space of weeks.
Now we wish that that would happen all the time, and it doesn't, but those examples do
give one some inspiration that we're on the right track here, at least in some instances.
Obviously, this worked because a lot of people had done a lot of very good biology previously
about the pathways that were involved so that you knew when you found the mutation what
might be the right dietary and pharmaceutical intervention.
So that's the good news in terms of the discovery of disorders at the molecular basis. The part
that I think we should not feel so good about is this: That as of right now, of those 5,000
disorders or so, there are only 250 of them that have a therapy that is considered to
be appropriately beneficial; a huge gap. Now we all know that gap is a tough one to cross
because there's so many steps involved, and many of these are very rare diseases where,
in fact, the expectation of any real serious private sector interest has got to be muted
by the lack of anything like a very large market for financial purposes. But I don't
think it's appropriate, though, for us to just shake our heads and go, "Well, that's
the way it is." I do think there are things we can do to speed up the process of going
from gene discovery to therapeutics. That long, 14-year typical pipeline has some aspects
of it that could be susceptible to new scientific ideas, and are, in fact, being tackled now
in systematic ways, in partnerships between public and private, that no company would
undertake on their own. And that was, in fact, the motivation to start,
for the first time in quite a while, a new center at NIH, The National Center for Advancing
Translational Sciences, which has already, I think, made a significant impact in this
space, albeit it with very modest funding. The only new funding for this effort that
was not already there from other things that were cobbled together is something in the
neighborhood of 0.2 percent of the NIH budget. But it has allowed us to focus on things like
how to do better a job of drug toxicology testing than the tried and true, and often
not so true, testing of animals. It's also allowed us to set up a program to repurpose
compounds that had failed for one application, but might turn out to be just the thing for
a rare disease where you've just discovered the molecular basis, and a few other things
as well. So I think it is an interesting development,
and one, which while a very modest form of new financial contributions, potentially might
get us closer to closing that yawning gap between what we know about diseases, and what
we can do about them. And, of course, this was a center that attracted some controversy.
Anytime we do something at NIH that sounds like we're building a new enterprise, people
are worried about "Okay, where's that coming from?" And, "What kind of consequences will
there be for funding of other things?" It seems that we're in that space again here
in the last month or so because we have a new kid on the block here in terms of NIH
initiatives, namely The BRAIN Initiative, Brain Research through Advancing Innovative
Neurotechnologies. This is an enterprise announced by the President on April 2nd, which is very
basic science. This is an effort to try to build very fundamental ideas about how circuits
in the brain work, starting with model organisms, because the human brain is way too complicated
with its 86 billion neurons. But we don't really have, at the present time, a very good
understanding of how to record simultaneously from tens of thousands of neurons, and be
able to assess how it is, in real time, that some complex function is being carried out.
How do the emergent properties of the brain actually emerge? You'll never really learn
that by studying one neuron at a time. You've got to get into the complexity business.
And that was the motivation for this. Again, a modest scale, $40 million dedicated to this
next year for NIH and FY '14. If you're doing the math, you'll know that's slightly over
0.1 percent of the budget, but just slightly. But potentially also an opportunity to bring
together scientists from multiple disciplines, nanotechnologists, engineers, neuroscientists,
to see if we could, kind of in a Genome Project-like way, learn to talk to each other, and develop
new ideas about what to do. This project clearly needs just like the Albert's
panel did, a blueprint of what exactly are going to be the steps, and that is something
that is now under construction, being led by a very capable team, which Corey Bartman
[spelled phonetically] and Bill Newsom [spelled phonetically] are co-chairing, and which,
by the summer, will have some general ideas of what we should be doing first, and by the
summer of 2014, a more Albert's-like panel. So there are parallels here to the Genome
Project, but we shouldn't overstate them because the Genome Project had a much clearer endpoint.
I don't think we're going to be done with studying the brain anytime in my lifetime
or yours. Nonetheless, it was an exciting opportunity to think about sort of the next,
as the President called it, the next great American science project, to try to see what
we could learn about that most complicated structure in biology that we know of in the
universe, the human brain, but to work up to it gradually just like we did with the
Human Genome. It did attract some attention and some anxieties.
So, yes, there seem to be some similarities here between things that happened in The White
House and what people say about it before or after. So I thought it would be fun to
give you a little quiz here, which appeared in Wired magazine. This came just about a
month ago, where there were quotes put forward by Wired, and they asked you to read the quote,
and to answer the question, "Was this a quote that was made in about 1988 about the Genome
Project, or is it a quote made in 2013 about the Brain Project?" So here we go. Okay, you're
going to get the answer here. I'm going to tell you, is it brain or is it genome? And
we've got a little animation so that you'll get to enjoy that part, too, when you see
the answer, so here we go. So here's a quote: "The whatever-it-is project
is bad science, its un-thought-out science, its hyped science." Is it brain or is it genome?
Okay. Genome, they say. Ah, you've got it right. All right, let's see. "I believe the
scientific paradigm underlying this project is, at best, out of date, and at worst, simply
wrong." I hear brain. Okay, you're doing pretty well here. They aren't all alternating. It's
not going to be that easy. [laughter]
"Concentrating hundreds of millions of dollars on this one megaproject in the era of budget
cuts is sure to starve hundreds of small, more promising biomedical research projects."
Genome, yes, Got you. "In contrast to some areas of physics, which require extremely
expensive facilities, biology does not have an obvious need for big science. Our country's
spectacular success in this area has depended, in large part, on the wide support of independent
investigator-initiated peer-reviewed research." Genome, oh, yeah. I even remember who said
that one. [laughter]
I don't know if he does, but we do. [laughter]
"Creative science is bottom-up, not top-down. Are we talking about central planning inside
the beltway?" Brain. That was actually said by Corey Bartman, who's now the co-chair of
the group that's making the plan. So you see, you criticize, this is what happens, we put
you to work. "It's going to do absolutely no good to develop tools for a new generation
of scientists if we, in the process, seriously damage that same generation of scientists." Uh-oh. Well, we didn't get
the answer? Oh, no, it didn't rotate. Well, I guess I don't know.
[laughter] To be continued, yes. Sorry leaving you hanging
there. Now, "Arguments are made that the project will give birth to a new generation of technologies.
What good will that do in the absence of individuals trained and capable of applying these technologies?"
Genome. Genome. Male Speaker:
That was a real concern.
Francis Collins: That was a real concern, and maybe it still
it is. [laughs]
Male Speaker: We're still there.
Francis Collins: We're still there. "The amount of money we
ask to accomplish the task, $200 million a year," which has been floated for brain, as
well as genome, "is commensurate with the projects role in the fight against many serious
health problems." All right, a positive, yes. And "Everybody I talk to thinks this is an
incredibly bad idea."
[laughter]
Female Speaker: Both of them.
Francis Collins: [laughs] That was genome. That was genome.
Okay, so maybe there's a pattern here, and maybe those of us who start these kinds of
enterprises or start talking about them remember that. But of course it doesn't mean just because
genome turned out pretty well, that the next big science project will turn out pretty well,
so you should all be watching carefully here and holding us accountable.
Okay, well briefly, where is this whole thing going? I'm going to finish up with just a
few brief remarks about the future, and again, I'm very much, I think, pointing you towards
this issue of Nature, which included a really wonderful summary by a lot of people, but
Eric and Mark led this enterprise. And I like the graphic that they put forward here about
the ways in which this enterprise is going to move forward as sort of as a comet does,
but with a very broad reach across all kinds of aspects, from sequencing on up to medical
applications, and that all of those things ultimately will move a little bit to the right,
but not so fast, because, clearly, we have a lot of work to do on the left side of this
as well. And the kinds of things that that paper points out: In terms of genomic medicine,
genomics-based diagnostics need to become routine, define the genetic components of
disease, comprehensive characterization of cancer genomes. We're in the middle of all
these things. Devising practical systems for clinical genomic informatics, uncovering the
role of the human microbiome; fascinating, and I'm sure Claire talked about that today.
And to get to all of these goals by 2023, these authors lay out some fairly ambitious
milestones: Understanding the biology of genomes at the level that we don't currently; bioinformatics
in computational biology; education and training; and, of course, the important aspect of keeping
track of the implications of all this for society beyond the scientific and medical
issues.
So let me finish with a little bit of a silly futuristic view here about where this will
all go, assuming that it goes well, and this is going to be a made-up story about personalized
medicine over the next, shall we say, 100 years. So this is about a baby named Hope,
who was born on April 14, 2003, 10 years ago. Family heard news that the Human Genome Project
was completed because that was the day we had a big party at the Library of Congress,
and said we had achieved all of the goals of the HGP. Well, okay, fast forward 10 years
to now; we're having a 10th anniversary celebration, and Hope is turning 10 years old and having
a nice party with a lot of balloons.
Okay, another 10 years forward, Hope attends a funeral of her aunt who died much too young,
at age 53, causing the whole family to be quite in grief and wondering what happened,
and that led to an interest on the part of the family in actually trying to look at their
own family history and say, "Are there other people who might be at risk, and what should
we do about that?" At the point where we are right now, the surgeon general's family medical
history tool would be a likely place to go. No doubt it'll be much better in 2023, and,
in addition, there would be some genome analysis potential, which maybe by 2023, will have
somewhat better abilities to make predictions, especially if you combine it with the family
history, that free genetic test, maybe will be able to say something actually fairly useful
for somebody in this situation.
And Hope turns out like her aunt, to have a higher risk of heart attack. And so she
decides, "Okay, I'm not going to smoke, and I'm actually going to watch my weight and
exercise, and try to do something to reduce that risk." All sounds fine, 30 years go by,
Hope is celebrating her 50th birthday party in 2053, and she's also taking advantage of
new developments in terms of personal health maintenance, wearing a Smart shirt, which
keeps track of all kinds of bodily functions, to make sure that she's doing well. And it's
also time for another party because the human genome is 50 years old, and Hope and her family
are very attached to this, too, I can assure you. But, of course, Hope's getting older,
and in 2071, now in her late 60s, Hope feels some tightness in her while she's gardening,
she thinks maybe it's a pulled muscle, but her Smart shirt comes to the rescue, says
this is not good, picked up some arrhythmia there; calls the emergency responders, who
arrive. She is actually treated quickly, and because her information is readily accessible,
she gets the right treatment for her, and ends up surviving what otherwise might have
been a very bad episode. And all goes well; she celebrates her 100th birthday with a night
of dancing in 2103. I hope by then the clothes will have changed, but oh, well, Hope and
her husband are still living in the past as far as their wardrobe.
So that's a happy outcome, right? And that's not too implausible to imagine it might happen.
I don't think we stretched too many scientific or social developments. But it didn't have
to happen that way, so now the darker side of this. Could we count on that kind of outcome
for Hope? Well, maybe not. Hope's story gone wrong. In 2023, her aunt dies of a heart attack.
This is sort of the teachable moment, but there's no online tool left for family medical
history available. Physicians have not learned how to use family history or genetic information,
genome analysis is dismissed, and so nothing really is provided back to Hope.
She makes some unhealthy life choices, gains weight, maybe smokes occasionally, and what
do you know, by 2038, she's developed high blood pressure. She basically could have been
tested to see what would be the right drug for her hypertension, but that wasn't actually
available, or hadn't been developed, or wasn't paid for, so instead she got a random drug
that causes a hypersensitivity reaction, which was quite unpleasant, and she said these doctors
don't know what the hell they're doing, and stopped the treatment. So she continued, then,
eating an unhealthy diet, gaining more weight, uncontrolled hypertension, and now, in 2053,
she feels tightness in her arm while gardening. There's no Smart shirt in this scene. Doctor
dismisses it as a pulled muscle. She's taken to the E.R. three hours later in cardiogenic
shock, given standard therapy with a pro-drug, but you know what, her particular metabolism
was very slow, so she didn't actually metabolize it into the active form. Hope dies in the
E.R., at just 50 years old, having missed out on fully one half of her potential life
span.
So that's not good. And one of the goals, I guess, as we think about the future, is
to try to identify the kinds of actions that we have to take as scientists, but also as
people who care about the health care system, to make sure that we give a better opportunity
for this -- not to have this bad outcome. So, basically, the essential goal of genomic
medicine, if you want to ask it, what should it be, and to get it down to three words,
it is to keep Hope alive.
[laughter]
As you knew that was coming, didn't you? And Hope, in many ways, is all of us. Hope is
you, Hope is me, Hope is our families, our loved ones. We have the promise here, with
all the knowledge that we are gaining about the genome, and how it functions, and how
it plays out in health and disease, to turn these coming years into really dramatic advances,
for medicine as well as for science. But it isn't automatic that that will happen. And
things such as this really awful event that we are living through right now called the
sequesters, which slow down our momentum, are to be regretted, and to be pushed back
against, and we should do everything in our power to continue to make the case that we
are engaged in an enterprise which has the potential of enormous advances in human health,
and actually stimulate the economy, which is something that people pay attention to.
And furthermore, that genomics, and all of what we do in medical research, is not just
a random activity, it is a noble enterprise. It is one of those things that I think we
are all fortunate to be called to be part of, to take the time that we have here on
this planet to spend it in the laboratory or in the clinic, making discoveries about
things that have mystified people for all of human history, and then trying to apply
them to the betterment of the human condition. What a privilege that is. So I want to thank
you all for being here and for putting up with this final silly presentation, and I
hope this day will linger in your memory. I know it will in mine. Thank you all very
much.
[applause]