 Great, so as Eric said, this was the second of these workshops addressing this issue. This one was a little more to give NIH and the various institutes kind of guidance on how best to utilize the very large scale sequencing resources that we have at our command, not just in NHGRI, but as you know, the majority of sequencing supported by NIH goes on outside of NHGRI. So the goal of this was to provide guidance to NIH and the scientific community at large on the utility of sequencing large sample collections. We initially focused on cohorts, prospective cohort studies, but recognized that there was certainly value in clinical trial collections or case control studies or other things. So large sample collections to improve the understanding and treatment, primarily complex diseases, although obviously relevant to Mendelians as well. We initially started with an objective of kind of looking to see what's out there and kind of assessing the field. This rapidly turned into yet another inventory that I got asked to do, but it was too large of really a task for us to be able to handle. So what we did was to collect as much as we could, and actually the planning group felt that that was worth making available. So that is on our website, and I'll show that to you in a moment. But there was also kind of a tendency on the part of some to perceive this workshop as being kind of almost a, you know, the forum for selecting what cohorts might be sequenced. We very much wanted to avoid that. So this was not a beauty contest or anything like that. It was just, you know, here's a list of cohorts. But let's really address two key objectives. What are the main scientific questions that can be addressed by sequencing? Because much as we love sequencing, it doesn't answer every question. And there are some questions that it is very good at answering. And then define the criteria for selecting samples to answer those questions. So fairly straightforward kinds of plans. This is the workshop. These are the participants and the starred people or the planning group. Rick, I'm calling you to comment when I'm finished just to give you a warning. So in terms of the inventory, Eric and Lisa mentioned the trans-NIH sequencing inventory that we put together about a year ago that is available on our website. I think if you just Google NHGRI, you know, trans-NIH sequencing inventory comes up. But if you also go to the Office of Population Genomics website, which will be going away but will be replaced by something even better, it should still be there. This is actually the potential sample collections for sequencing. It's, again, just an Excel spreadsheet, and there's a little bit of a description on it. And then all we did, basically, was to kind of using the people on the planning group who know this field fairly well. Just list as many as we could remember of cohorts and case control studies and other things, just showing one of them here. But there's actually, it starts, I think, with African studies and all around the globe. There are about 130 or so large studies, some of them described in excellent detail because we had people involved who knew them, others not. And they would describe things like, you know, the numbers, the age, the year that they started, how long they'd been followed, primary disease endpoints, anthropometry, diet, you know, whatever might be collected. So that is available, and it's something that we've actually encouraged people whose studies are listed there to contact us. And there is a contact name on this website. If you want to fill in the data, that would be fabulous. But we don't have the resources to be able to go out and get that information. So just in the long story short category, we did come up with objectives and questions that could be addressed by sequencing and cohorts. There are many of them, and they're kind of grouped in these rubrics. The genetic architecture of human diseases across the life course, for example, questions about, you know, what are the rare variants that contribute, what are risk versus protective alleles, you know, something about pleotropy in their effects, gene by environment, interaction, et cetera. So they're about eight or 10 questions after those. Novel drug targets, pharmacogenetics, and non-drug treatments that don't relate to novel drug targets, longitudinal changes in phenotypes and disease outcomes, health disparities, epigenetics, and annotation of the genome. So all of these were sort of areas that the group saw as being potentially contributed to or addressable by sequencing. And then in terms of criteria for selecting the cohorts to answer those questions, there were some that were felt to be a very high priority. And many of them are very obvious, obviously large size for almost every question, not every question, but almost everyone. That's very important. Broad baseline phenotyping just increases the value of the variants that you already have. Multiple disease outcomes is related, but a little bit different, longitudinal data, ongoing contact so that you can actually go back and re-phenotype them if you need to. Broad consent and ability to re-contact, ancestral diversity, availability of both electronic medical record data and other omics data, basically as much data or as many data are available would be great. Family data or focused family studies were also seen as being a major plus in available cohorts and capacity for de-phenotyping when needed again, sort of a little bit related to those kinds of things. We also tried something and we have yet to see how well that will work. But we thought we might do a little matrix kind of crossing these things. So along the left side we have the questions that could be addressed by sequencing and along the top, the criteria. And we've asked the participants in the workshop to basically say, all right, granted, you'd love to have all of these things for all of the studies. But if you had to pick a few places, either a few scientific questions where it was just really critical to have large size or just absolutely essential to have multiple disease outcomes, what would those be? And similarly, if you really, all you wanted to do was to focus on a study of gene by environment, for example, what are the criteria that you'd want to have that you really want to have for the cohorts that you select to do that? So this is kind of a work in progress, there's a much larger matrix than it shows here, but we're getting input from the various workshop participants. And then this will be, is put into a draft manuscript that will be revised and submitted hopefully very soon to a prominent journal. So, Carlos. So in the previous slide, you had sort of all the things that could be given priority. And I'm just wondering about ancestral diversity as to how that fell out and why broad ethnic representation wasn't viewed as a particular. It was viewed as important. It wasn't unanimous that it was a high priority criterion. And I think you and I have both dealt with the fact that it is not a unanimous view that that is a high, high priority and it's something that we're trying to deal with. But I think that the feeling was there are advantages and disadvantages to it and some of these others were more important. I don't personally agree with that. Yeah. And is that now sort of set policy or is there some opportunity to kind of revisit that or how to think about that? Yeah, I mean, this is a draft workshop report. Now the workshop report really does have to represent what the workshop participants feel. But I think we can push them on that a little bit and see if we can't get that a higher priority. Carlos, can I ask you what you mean by this? That you want more diversity in any given cohort or you want the cohorts to be diverse? Because if you have a lot of diversity within a cohort, of course you lose power, right? So are you referring to the other? I was actually asking, so my personal view should be that we need to broaden ethnic representation and that might include funding more diverse cohorts or specifically targeting cohorts that have representation that hasn't been included so far. But I was more curious about in the workshop how come that didn't sort of float up as an issue or how that was discussed or what were the issues around it. Carlos, it was an issue. That's why it's on the slide. But I think if you see that look at the three things there with the asterisk, decade priority, those are sort of non-starter issues, right? So if you can't address issues like that, you just don't even get going. And I think it also depends on what cohort you're talking about and what you're going to hope to learn from cohort slash study. I agree. No, I was just, how did we designate? Yeah, so I think in the next slide there are like four things that were on that. Those are just examples. That's the little upper right-hand corner of the whole thing. Of the whole thing, the table goes on forever. I wasn't sure because it mirrored exactly the three. And so I wasn't, OK. Yeah, so that's what that is. Got it, OK. And I didn't have a chance to call on Rick. So Rick, did you want to add anything to this? It was a really good workshop. And it was gratifying, I think, to see the amount of institutes that were represented and the resources that they brought to the mix. We also discovered that if you hold these workshops within 100 feet of Francis' office, that he will actually come and participate for part of the meeting. Which was good, I think, because he had some useful things to add to it. So I think this is a really great opportunity for NHGRI because this sort of initiative, if we may call it that, touches a lot of things that we've been talking about over the last several council meetings and that we've talked about earlier today in terms of, I think, key pillars of NHGRI's mission. So it was a good meeting. There was a lot of good ideas brought up. And I hope we can continue to push forward on this. Thanks. Other comments? Oh, I'm sorry, Mike. I was just going to suggest one additional criterion. And that is not just broad consent, but also the ease of getting these cohorts to deposit data. And promises are great. But past behavior is also a useful indicator of that. And so I think, honestly, this is really important. Because if this initiative is to be as successful as we'd like, that that's fundamental to making it successful. Excellent point. And I think, Rick, correct me if I'm wrong, but the broad consent and ability to recontact, this broad consent was not just for diseases, it was for data deposition. So we can make that more prominent. Do you recall that? No, that's right. But I also would second Mike's point with regard to phenotype data and sharing. And beyond that, behavior. Yes. Behavior of investigators, not behavioral data, but behavior of investigators past behavior in terms of depositing data. You don't just have to have consents for the subjects. You have to have investigators behaving properly. That's close. I would also kind of echo at that point that once a cohort begins to contribute to this, it would be great to be able to not only aggregate the data so that you didn't need, obviously, separate DB gap for the GWAS data, and then for the exome chip data, and then for the exome sequencing data, and then for the 2x coverage data, and then a separate one for the full x coverage data. But also that once they go down this road that you begin to think about creating a kind of repository of samples with phenotypes that you can just kind of keep going back to and that you don't need to continue to negotiate that at each step, because I think that that's part of the issues that come up. You're kind of going back to the same set of samples, but each time you have a separate set of. It's like it's a new day. It's like a new day, exactly. So I don't know if we have a plan for going how to view that going forward, or whether it's going to be continued investment in the cohort that I've already been studied, or whether it's going to be more kind of clinical EMR data. But it would be good to get some kind of clarity of that in terms of a policy. Not so much from the workshop, but thinking as a group, I think. Other comments? OK. Adam, you're the dig meeting. Is that the dig meeting, sir? I'm a cardiovascular type. So then Adam's going to wrap this up with the third of three workshop reports on integrating functional data for connecting genotype to phenotype.