 Great. So I'm going to give a brief report on GM9, which was held this past April here in Silver Spring, and I was on the organizing committee along with Terry and Howard Jacob. And as Terry mentioned, the origination for this meeting was from Genome 8 when one of the discussion items was how to better bring together basic scientists with scientists focused on more clinically relevant work and with clinical care. And so that ended up being the focus of GM9. And that's a very broad topic. So in the end, what we decided to do is to focus it specifically on looking at the clinical significance of variants of uncertain or unknown significance. And of course, this stems from the fact that all these wonderful sequencing projects out there are identifying many, many variants in many different populations, and we really don't know what the biological significance of those and which one of those merit moving from discovery actually into operations in the clinic. And this seemed like one of the areas where it would be really fruitful to have better integration and alignment of basic science activities with clinical activities. And that was the objective, one of the objectives for this meeting. We wanted to look at some examples where there had been very successful interactions between basic scientists and clinical folks working with genome data. How that virtuous cycle of bench to bedside, back to bed, could be facilitated. And how to achieve a better alignment of the research efforts in the basic science community with the clinical application of those insights. And so we came up with a number of specific topic areas that we decided to focus on. But part of the purpose of this slide is to say the things that we took off the table that we didn't really cover in this meeting that we didn't intend to cover and we didn't have representation of the participants or in the audience to cover. So we really didn't tackle at all the issues regarding the regulatory hurdles of getting clinical variants into practice. And we didn't deal at all with issues related to payers. So the focus was really on how do we understand the function of these variants of unknown significance and their relevance to disease mechanisms? What is the state of the art and what do we need in terms of prediction and annotation of genomic variant function? How do we facilitate integration of data and sharing of knowledge in a way that enables this bridging between bench and bedside? And then just the general topic of overall how to facilitate those kinds of interactions. So the meeting itself, we broke down into six sessions. This was a day and a half meeting or almost two full days of meetings. We started out with a couple of talks that really talked about just the magnitude of the of the problem, had a couple talks illustrating some specific issues from the clinical perspective, some talks from basic scientists whose work had been translated into the clinical sphere, and then had some talks about computational methods that are applied or could be applied to understanding variant function. Some some really new technologies that are coming along in terms of experimentally with high throughput, functionalizing variants of unknown significance. So there's some really exciting technology in this area. And a lot of it has been cell-based assays, very high throughput, very high function, but it's only been on a limited number of genes. And so discussions about how to select genes for additional work in this area, how to get these these communities to be talking to one another so that the technologists can be working and aligning their research goals with what's important for the clinicians. That was an exciting session of this thing. And then something near and near to my heart for data integration and data sharing was kind of the state of biomedical phenotype ontologies. And we've heard phenotype come up a couple of times now. Dr. Gibbons mentioned it in his talk. And I think it's a it's a really important thing, something that NHGRI has actually had quite a bit of impact in. So after we step back from from the sessions and from the talks when we were thinking about how to summarize what had been presented, we basically distilled it into three broad areas of variants phenotype and then a general topic of bridging that gap between basic science and clinical practice. In the area of variants, it was really about aligning priorities. So how do you get the communications between these communities going so that the variants and genes that a basic scientist is working on are really in alignment with what a clinical priority might be. So we talked about the need for enhanced transparency for how the algorithms that are being developed for predicting variant function are arrived at and also developing criteria and evidence standards for moving variants into clinical practice. In fact, at the very beginning of the meeting and in sort of setting the stage for the magnitude of the of the problem, Howard gave a great talk about a meeting he went to where he presented or he asked, I think the clinicians to say, how much evidence would you need from model organisms before you would take this clinical variant and actually put it in somebody's medical record? What is the most important information to you? And he stepped through and he asked for raise of hands of people and apparently didn't get very many people raising their hands, which really shows you the magnitude of the problem. You know, intellectually I think we all know that the evidence from model organisms and from cell-based systems could be invaluable in functionalizing these variants, but yet how are we going to do it in a way that the clinical community is going to accept that evidence and actually then make decisions based on that evidence? So I think that pretty much summed up the magnitude of the problem very well. And although we haven't come up with a definitive answer to that, I think a lot of things have come out of this meeting that could actually end up fostering the kinds of communications and research alignment that are needed to actually address the magnitude of the problem. So with respect to phenotype, there was a lot of discussion on phenotype and you know, phenotype and phenotype annotation has come up over and over again. And a lot of it is phenotype being that is what you collect on patients and then a lot of it is on phenotype descriptions. So how you annotate what you're observing in a way that you can compare it across individuals, that you can compare it from human to model organisms. These are big, big thorny issues, but they're ones that are really at the heart of being able to bridge all of this information to integrate it, to make sense of it both biologically and from the standpoint of clinic. So there's a lot of barriers to doing that right now, even though there's a lot of individual groups working on these. And so this is one of the very, I think fruitful areas of discussion that we had at this meeting. And then the final big picture sort of GM9 in a nutshell was just bridging that gap. And a lot of this had to do with the stuff that's always very, it's actually very difficult to do, which is you go to a model organism database or a model organism based research conference and you don't really have anybody there who's talking about clinical work and you go to ASHG and you often don't have a whole lot of people speaking in sessions from the model organism or cell based assay community. So fostering those communications in part means asking people to step outside the usual boundaries. And we talked about fostering opportunities for workshops and speakers from different communities as being one of the ways that we need to start kind of a new interaction between basic scientists and clinicians. This is also important for identifying existing standards and resources that could be very instrumental in bridging this gap as well. So what were the outcomes? Besides a really fun meeting, we have been working on writing a manuscript to capture kind of the things that I've put up here as GM9 in a nutshell. And so that manuscript is well underway. The title is going to be Bedside Back to Bench, Building Bridges Between Basic and Clinical Genomics Research. The manuscript is in the final stages of review and revision. The wonderful thing about this manuscript is that the co-authors have been extremely engaged in making comments and suggestions. And the reason we haven't really met our deadline is because our colleagues have been very engaged in making suggestions. And so we're especially Terry has done an amazing job of integrating all of that stuff and we're very close. This actually is directly from the manuscript. It's kind of summarizes the purpose or the goal of the manuscript. It really highlights the value and the need for that basic clinical science integration. But we actually do come up with some practical recommendations too. Some of them being identifying categories or sources for identifying clinically relevant genes as priorities for functional studies. Some ideas on developing larger reference variant databases with links to phenotypes, adopting standards for phenotype description that will facilitate data sharing, and then promoting cross-disciplinary understanding. One of the things we're struggling with right now in preparing this manuscript is actually summarizing kind of the sources of clinically relevant genes for functional studies and then criteria for prioritizing genes or variants for functional studies. This is this goes directly to trying to find a way to very practically align interests of basic scientists with clinicians by providing lists summarized lists like this so that somebody has a place to go to to get started. I don't know if it would be worth going through all these now. I mean these are just examples of what folks have been coming up with as sources. So there's the 56 genes recommended for return of incidental or secondary findings from ACMG. There's 33 genes with variants affecting drug response from CPIC. A lot of these genes aren't going to have their biological function fully spelled out and having more research into function and some of these genes could then go back and help make the use of these genes even better in a clinical context and that was the idea here. And then you know you can't do everything. There's a lot of genes that we need research on so how do you prioritize them and folks came up with a number of interesting ideas for different ways to prioritize those genes as well. So stay tuned. So the meeting like all other NHGRI initiatives is really meant to try to make this information the presentations and stuff as broadly available as possible and so this is on the NHGRI website. The information about the meeting and the presentations are all there. This is a list of the acknowledges of the folks who have been working on the manuscript and we're either the session leaders or presenters at GM9 and then a special thanks to the folks that helped us coordinate the meeting and especially to the Genomic Medicine Working Group and I'll close there and be happy to take any questions. I don't know Terry or Howard if you wanted to add anything that I might have missed. Yeah Dan. I wanted to add something that I thought maybe didn't come across quite as robustly as it could have. I want Howard to talk a little bit about his example because it was one of those you know when you had the story in front of you and you're the clinician it was very compelling to sort of make a decision that you were going to do X or Y and you know the details I can't remember them and it also involved some pre-clinical assays as I recall and it's interesting that the clinicians will say, not the clinician, will say yeah you know that's just a mouse thing or that's just the yeast thing and I'm not interested until it's their patient or until it's their mother and then suddenly you know oh boy whatever you can give me you know an IPS cell fabulous a yeast cell thing so it there is this sort of emotional tension around how to use all those pre-clinical data and Howard may want to expand on that because it comes back to this business of model organisms as well. I'm you know every day I get asked you know what do you think of this variant can't you study it in your cells or in your this or in your that and it so clinicians want the information whether they want to know the answer whether they like the answer that's a separate story. Well in a nutshell it's really about evidence-based medicine is largely premised now on randomized case control studies and so many many physicians will not use that information unless there's guidelines associated around that and so or well until it's her mother but that's that's a harder one to get to and so I've simply been going around in clinical meetings taking a shroom 3 is the gene that has been in 11 GWASs for renal disease it's across all ethnic groups and there was no known function and we've now figured out what the function is and I walk through rat data mouse data zebrafish data cell data putting the actual rescuing the phenotype in the zebrafish from the human not making the exact human mutation and I get somewhere between one and ten percent of the clinical audience saying that they would use that information so it's a real challenge if we can't get to randomized case control studies and and you know it comes back to the epidemiology the challenge is is that our case control studies that we use we make a lot of assumptions and by sequencing genomes we're going to identify that we're all individual and how do you bring the individuality back into epidemiology and it's a big challenge and so I was just in a meeting this week in Europe same thing a few more than I got in the states but it's it's really a very very difficult challenge other other questions are anybody on the phone wants to chime in okay I'll wait one more gail sounding really stupid but in the CSER meeting approached this with so which such a different lens and I think one of the questions although not it didn't hold sway but one of the questions that that the some people came to was you know should v us's ever be returned as and I know in incidental findings no no but even in diagnostic findings the you know the information that that people are you know are providing and gosh I hope I would like Sharon to tell me what was wrong with that comment from CSER but CSER I'm just struck this is fascinating because it's just so totally different than the dialogue at CSER well so first of all that was one person's comment um no and and many of the ethicists in fact some of the ethics were coming out of CSER have argued particularly for children that you could make it and Larry McCullough and he was part of the committee also you can make a strong ethical argument for not reporting v us's but that that's not like the consortium viewpoint I would say but no no I just meant I think it's it's one approach is to say if you can't this gets back to the evidence-based kind of thing if you don't have evidence why are you even telling me I do think it is important to separate two overlapping issues and I think what you were just describing is this issue of where we don't even know if the gene is real um and so we have variants in genes that we may have found through exomes or genomes and we don't what what are we using to really define as the evidence that that gene plays a role in disease and that those variants do then there's the ongoing issue um and Larry Brody who's in the back has done a lot of work on this of genes we know a lot about just all of the rare variants and there on the American College of Medical Genetics and IRC and several national or international organizations have tried to define criteria for classifying variants one certain significance so that was going to be my question I wasn't completely clear from the meeting description how much you're trying to build on the existing professional standards or where or whether you're really talking about these genes where we don't even know yet that they play a role in disease because there are standards for classifying variants from professional organizations in known disease genes and I'd hate to see us developing two separate set of standards so I think part of the challenge is is that you know most of those guidelines it's it's strongly human driven and so the basic part of the meeting uh that we had set up on was when we first were putting this together was thinking about the basic research and when we were in the middle of doing this meeting I had happened to be at a clinical meeting and I tried this out and I was surprised how you know that the American society in nephrology I had three investigators stand up so first of all no one out of 500 people raised their hand that they would accept that information and this was their field and I showed them all the data and I showed them that it was potosite effacement that we could demonstrate in a zebrafish and the only three guys that stood up said yeah but that's a fish okay but all I would say but so but but so part of the challenge is is that is that if it's not human we don't trust it and that's that's the general statement but that's a big pushback that we're going to have in the clinical community and that's a problem right I would just say that um there there are criteria for including functional data which includes animal models in the ACMG classification scheme and one thing clingent has done is develop a whole framework for kind of scoring the weight of model organisms and mouse models and things like that um which is part of the the gene validity matrix and it might be interesting to kind of share that because that I believe is online um as one way to think about scoring that but I completely agree if you then go to nephrologists or cardiologists they they won't use that data directly they want to see it incorporated by a clinical lab in a report that says do you believe this or not let me just have one more thing about the CSER some of the I think really really unique parts of CSER studies that that I know about and that is that the way the the way the results travel is being tracked and I and um described identified etc so the ways that you know the lab identifies you know v us's and other things um then gets translated um just very simply when we've observed clinicians and how they talk about v us's to participants and then participants later what they thought and it's really clear that v us's are not presented as one thing I mean they are presented in a variety of ways at least three categories that well you know this really might be something this really but we're not sure but you know we that there's the promissory um view of the future um this you know this is really a v us and we just don't know and then there's a this just barely squeaked by in our v us classification so all right so when you've got that kind of communication going between the the clinician researcher and participant you know um I think then you've got an additional layer of complexity and that that's what I think is so great about what the CSER program has been doing and we're just hearing these results now after all these years okay I'm going to go in there Rudy ready for okay so um it's my pleasure to introduce um Eric Dishman um to members of our council let me I'll give a brief introduction also to set the context for his talk needless to say um from what you have heard already about the precision medicine initiative it is a complicated endeavor has to bring together a lot of different things and would imagine somebody who's going to lead the cohort program and would need to have a lot of um a lot of skills to do so and indeed I think NIH was very fortunate that we were able to convince Eric Dishman to come to NIH to lead this cohort program those you don't know of Eric's background it is perhaps not conventional for most institute and center directors the point of view he comes at this as a social scientist his his area especially was around telehealth and personal health records and independent living technologies for seniors but his career experiences also include being a business leader and entrepreneur but he's also at a personal level he's been a patient and a patient advocate and through that he's become a very strong policy advocate in a number of areas that are highly relevant to what we're going to talk about um and he's also has emerged as a thought leader in many of the areas that were very relevant for the precision medicine initiative which therefore made it not very surprising he was invited to some of the earliest workshops and then was later invited to be on the external working group that formulated the plan for the the cohort program in particular um right before joining NIH which has only been like a couple of months or something he was previously Vice President and Intel Fellow of Intel Corporation Health and Life Sciences Group and I'll just read from the press release that was put out when when he was hired from whether 15 years he had led Intel's healthcare strategy and research including the creation and distribution of Intel's open platforms and open source tools to help researchers accelerate scientific discovery ranging from wearables and in-home technologies for movement disorders to big data platforms for cancer genomics and so again it certainly you can immediately see where some of the things that the precision medicine initiative cohort program is going to need expertise Eric certainly brings um and before I turn this over to him I just was going to point out one other thing I often point out there's not a whole lot yet officially written about precision medicine initiative um out in the literature to read and and yet I think these things will start to emerge and last week um in stats you can probably find this online if you just go stat and put an Eric Dishman I'm sure it'll come up um Eric authored a piece that was written that describes um uh his perspective and some of the background on the national precision medicine program that he is now at the helm of so with that I'll turn this over to Eric and I'm looking forward to hearing what he has to say. Well sounds great well it's uh it's a great honor to be here I'm one of these people that needs to move it's the only way I get my steps if I move during talks the Fitbit is a cruel master so