 Well, thank you, Eric, and thank you for your willingness to play leadership role in the discussion of this meeting this evening and tomorrow. It's great to be back for the second time in the month of June with so many of my friends from the genomic community. Thank you to NHGRI for making sure that I continue to have a chance to hang out with my peeps. It is very nice to have a chance to hear some of the discussion this evening. It has been a bit of a memorable day around here because we were preparing for this meeting. That was the only reason. No, actually, I suspect historians might not identify the opening session of sequencing and cohort studies and large sample collections as the thing that happened on June 28, 2012, but something else which did happen at about 10.15 this morning, initially improperly reported by CNN, but ultimately correctly identified as the Supreme Court upholding the Affordable Care Act, was really quite a moment for anybody who lives in this town and especially anybody who works within this government. So we are all feeling pretty relieved that a very large cloud that was hanging over the future of healthcare in this country has at least diminished. It's not gone away, obviously. There are many other things that are potentially in the way of this becoming fully implemented, but it is a memorable day. I do want to thank everybody for coming here to this particular space for this meeting. We have in the process of increasing austerity of government support for virtually everything now pretty much moved into the zone where we try to use our own government facilities for holding meetings so that we don't spend a single dime of anybody's taxpayer money on hotels. So you will notice the surroundings are very nicely appointed. You will notice that the refreshments are particularly impressive that I hope you enjoy your water. It's really good stuff and I'm sorry it's inconvenient for you, but it was very convenient for me. I just had to walk up two flights from my office, which is just right down there in this very building, but I'm glad to be able to hang out with you this evening and hear some of the conversation because I think this is a very important kind of debate that stretches beyond almost any limited view of what medical research needs to do. And it's very much reflective of the direction that science now allows us to go in which we need to be sure we're taking full advantage of. You're probably aware that NIH institutes through their various programs are in the process already of supporting whole exome or whole genome sequencing on roughly 65,000 individuals, and that data will be essentially in hand by the end of this calendar year, with much more to come. So there are already enormous amount of work going on in this space. Would I look at that portfolio of programs that have undergone this kind of whole exome or whole genome sequencing and say this is the ideal? Probably not. These have all been chosen by various perspectives to be important for the understanding of a certain category of diseases, but I don't think anybody would say that the samples themselves now cover the universe of possibilities, not by a long shot. So again, I think there's lots of reasons to focus as is the case here at this meeting on how to make the most of this exceptional circumstance. Now to be clear, several of you did attend the workshop held earlier this month by NHGRI, and it's going to be, I think, risky that we might sort of forget which workshop we're in and slip back into that one, so let me just remind you what that workshop was about if you weren't there, or even if you were. That focus of that effort was to try to make the most of aggregating and analyzing the sequence data that has already been produced or is imminently being produced through those 65,000 samples and others that are becoming publicly accessible. So there the idea was how do we build an infrastructure, some of which exists, some of which doesn't, to allow that kind of data to be assembled at high quality in a fashion that addresses all the consent limitations that may have been applied at the time the samples were obtained that respects privacy of individuals that maintains the highest possible quality of the data and also provides some analytics to allow individuals to be able to learn more from the complete whole of the data than you could from any subset. Because right now I don't think anybody would say that kind of capability is conveniently in hand for lots of people who would like to conduct those kinds of experiments. So recommendations from that workshop are being assembled. Lisa Brooks has ably tried to put that together and I've seen a draft of those recommendations. I don't know if you're actually distributing them, but you did. Okay, good. So you if you haven't seen them, they are there and they were very much worth looking at on page two and three of that document because it's relevant to what we're talking about. But again, I'm partly bringing this up as background and also as an exhortation and not to imagine that we're going to that same workshop. This is a different set of questions. That was really about how do we make the most of data that's being generated or has been generated to make it actionable. Now, here we're much more focused on if you were able going forward to make the wisest choices about sample collections that could be utilized for sequencing. What kinds of criteria would you apply? Again, what Eric said is very much the case. We don't want you to identify specific cohorts with specific numbers of patients of specific sorts that already have a name attached to that study. We want the sort of criteria about what would make a good mix. And again, it will be a mix. It won't be that there's one perfect kind of cohort here. I'm pretty sure that's the case. I appreciate the comments about diversity, which clearly need to be a big part of this conversation. So again, I think one of the areas to think about in terms of applications is going to be the biological opportunities here. And particularly, I would encourage you to think about ways in which this kind of selection of ideal samples would inform the development of new therapeutic ideas. Some of us have been spending a fair amount of time interacting with industry about this very question. And I think some significant progress is being made about the idea of some industry, NIH academic consortium to try to make the most of the current science, particularly when it comes to identifying potential new targets for therapeutic development. And in that regard, one of the things to think about in your discussion is if you are trying to identify, in the course of sequencing carefully chosen samples, individuals that have interesting knockout phenotypes after all the human knockout project has been under way by nature for some time, and we would potentially benefit the whole future of identifying the consequence of antagonists against various proteins that are gene products if we had those natural experiments for comparison to see what the consequence would be. A corollary of that, if we're particularly interested in loss of function mutations and loss of function mutations, ideally in the homozygous state if they are viable as adults, that will play out in your discussions, I think, in terms of which kinds of populations would have the greatest enrichment for such. So a outbred population with lots of genetic diversity will be good for some things, but maybe not so good for finding homozygous loss of function where you would rather be looking at populations with lots of first cousin or uncle niece matings where you might have a much higher likelihood of finding such homozygous loss in a substantial number of important loci. Again, this is sort of the PCSK9ing of the whole issue of how to do target validation. Everybody points to PCSK9 as a wonderful example where human genetics effort started by Helen Hobbs and now resulting in many industrial investments to develop a new treatment for high cholesterol really led to a remarkable series of events pointing to a particular target, namely this PCSK9 protein which was shown by genetics to be deficient in about two or three individuals, two or three percent of individuals in the heterozygous state and with rare homozygotes also available, having very low risks of cardiovascular disease and also apparently no adverse effects. So the nice sort of gold standard of what you'd want to know in terms of validating a drug target from human biology, not from an animal model or from a cell model. So from the industry's perspective, those kinds of discoveries are going to be particularly interesting and ought to be on the list of things to think about as you're considering what criteria to collect samples for. Protective alleles of other sorts, obviously also of great interest. Finding out for instance what is it that actually is different about somebody who's 85 years old and has an ApoE4 allele and who has completely normal mental capacities. What has allowed that kind of outcome to happen? If you can identify such modifier alleles that are protective, well, maybe you're on the road to some pretty interesting therapeutic ideas. And I don't mean to say that we should also limit our thinking to loss of function, although gain of functions a lot harder to see when you look at it, but certainly duplication of a locus is a pretty good indicator. There's likely to be a gain of function that shouldn't get lost in the shuffle. Anyway, I'm just giving a few examples of the kinds of things that might be considered in your array of criteria to consider for the ideal kinds of samples that if we could start now and do everything in exactly the most efficient way you'd want to have included on that list. Obviously another very major one will be whether in fact these samples can be obtained in a way that consent is broad enough to be able to get access to the maximum percentage of the scientific community. That's going to be a very important issue. So I'm looking forward to the discussions about all of this. This is a privilege for me to be here to say a few words at the beginning. Again, I think remind you that the real objectives of this workshop, first, identifying key scientific questions that can be addressed by exome genome sequencing. Second, define the criteria for samples that would best answer those questions. I'm just echoing things that have already been said by Eric Burwinkel. So without further ado, thank you for the chance to say a few words and let me turn this over to Eric Green, who is going to give you his thoughts about why we're all here.