 So we are getting ready to conclude with the summary session here. I just want to take the time to say thank you to everyone who's contributed whether you gave a talk or asked a question or are critically thinking about these issues with us. I think we heard a lot of really, really good insight from you guys. And our intent for this session is to summarize someone on a cursory basis, some of the recommendations we've heard from each topic, and then we'll have a longer moderated discussion to make sure that we captured everything that we heard and potentially visit some ideas that we didn't get to explore in depth. So I'm going to go through these, Carolyn and I will actually go through these fairly quickly. If you have additions or revisions, please hold them until the end. So starting with the first session on facilitating development of a shared evidence space for healthcare systems, the recommendations we heard were to develop rapid learning systems to capture phenotype and outcomes information. The other ideas there are being to learn from each encounter, continue to build the evidence linking genotypes to phenotypes and encourage sharing in widespread implementation. So we obviously need to learn from sequencing happening outside of CSER as well, facilitate recommendations in disease specific organizations, figure out how to engage families and then define and refine guidelines in populations and individuals, whether these individuals are phenotyped or not. And then the final recommendation from that session was to continue work on legal and regulatory frameworks. In the second session on interpreting variance and actionability, there was a little bit of discussion about how do we come back and improve standards for consistency and variant classification and really making sure that that's transparent in terms of the evidence used. So you can go in and see the justification for the annotation and be able to evaluate it rather than just having an assertion. Also there was a lot of discussion about the importance of really including the case level data in that as part of what needs to be this information as part of the transparency. And then obviously that this would all need to sort of be, databases would have to be improved in order to be able to facilitate all of this type of generation that would really allow this sort of rapid learning, which has sort of certainly come up throughout the day. Then there was also discussion about how this shouldn't be done in a vacuum and the importance of having both phenotypic and clinical context to help inform the evaluation and interpretation of variance and actionability. And really the discussion and Elaine and I were talking about this a little bit in the break of really having a dynamic process between the lab and the clinician and also an understanding between the lab and the clinician about what's going on. There was discussion and Dan had a slide which I've turned into, Dan Rodin had a nice slide he sent around which I've turned into one little two word thing of taking a Bayesian perspective and the importance of how we really are doing that, but maybe to formalize and think about that. And understanding in this context what should we be sequencing and is this, is how much of this is disease specific. And then also can we really think about and maybe this should be a separate bullet really coming into logical studies and sort of questions of how do we elucidinate penetrance. The last two things had to do with clarifying and improving the process for refenotyping and reinterpreting data and really coming up with guidelines in clear ways and clear understanding about how we're doing that. And then also refining the guidelines, addressing key gaps in guidelines and making sure that we move past just guidelines within the genomics community and really involve other communities as well. Okay we have two assessing clinical utility slides. So one recommendation we heard clearly was to improve how we measure clinical utility and specify what we mean by it. So two points there were to clarify what individual or societal benefit we measure because there was a distinction there. For example, what's the value of the diagnosis if it's not clinically actionable to an individual. And also there were a couple of examples of how sequencing can inform population screening. Second recommendation was to adopt multiple approaches to studying clinical utility, building on the diversity of approaches that we see in current CSER. So three ideas here were to encourage comparative effectiveness research. We heard a lot about doing CER including comparison with other approaches in clinical medicine. I think one idea that we heard was perhaps we're holding ourselves to a very high standard, too high of a standard and some of the CER research might help us kind of evaluate that. Second idea here was to deepen evaluation of the outcomes collection including collecting morbidity and mortality measures and then aggregating these measures and then to acknowledge the contributions that RCTs can make. Second set of recommendations, consider protocols for refenotyping. This is a theme that we heard in a couple sessions. What should we be measuring and how do we improve the standards and common elements across studies? We need to develop effective and efficient methods for integrating functional genomic studies with the clinical phenotype data we have. Consider functional research as an important component of the virtuous cycle that's been mentioned a couple of times. And then facilitate feedback between individual patients and high-thrubit functional assays to elucidate the deleteriousness or pathogenicity of a variant. Last recommendation was to improve sequencing of indels, structural variants, and other non-SNV variation. So this will help us identify, improve clinical utility by identifying the current holes and how to fill them. And then the last idea here was to consider what we mean by the clinically meaningful variation. Okay, this one is mine, too. All right, so from the Patient-Centered Research Group, we also had two slides here of recommendations. The first one is, I think we did hear a very clear consensus to incorporate patient-centered and family-centered approaches. I think we had some very compelling anecdotes as well as experience from people with N-cesar and external diseases that this is really critical for us to think about, and we'll do so going forward. Acknowledge the value of having a diagnosis as a patient-centered outcome. Acknowledge differences and the meaning of actionable to the patients and their families versus the physician. And then leverage resources to connect patients with researchers and foster a community. So I think this is something that we think of more generally in terms of outreach, but in terms of what you all are discussing today, it sounded a bit more like outreach on steroids. So we need to really leverage the Internet, for example, and make sure that we connect with the disease communities and so forth. And then for patients who are not comfortable with sharing their data on the Internet, identify alternative approaches for building this community. The second slide for the patient-centered research group was included recommendations to continue to study the impact of diagnosis and subsequent outcomes. Pay attention how we package results to patients, especially variants of uncertain significance. So this is something that we are exploring in C-Zar-1, but heard a call to do more critically from a patient-centered perspective and research going forward, and then use flexible approaches to the patient-centered, I think we maybe could have worded this better, be flexible in considering patient-centered approaches that can change as technology changes, and then the number of populations and ethnic groups increases. So then in the session of increasing ancestral socioeconomic and clinical diversity, which as Pilar mentioned really focused on the first of those more than the three, I think the things that came about first was to really focus into these communities, you have to implement targeted methodologies and approaches, and you need to do this work both to recruit and to sustain people and sort of thinking about how we come in and do that. There was a great discussion about the focus on trust and how some of that can also be facilitated through making inroads in community based institutions. And as we come in and do this to be sensitive to the barriers, some of the ones that were highlighted in particular were loss of work, transportation costs, et cetera. There was also a discussion about the scientific value, and the social justice, I didn't put that on the slide, but under there and thinking about making sure that the studies that we design achieve our scientific goals, you know, and in terms of the sufficient sample size it may not just be proportional or it may not be proportional to what's in an area, and you may need to actually over sample or over represent in order to have the numbers you need. And then I thought there was also a really good discussion and a really good recommendation about how including more diversity is going to both enhance and improve work in LC areas. Then there was also the discussion that even though there are other concerns in terms of the major health concerns in some of these populations, we need to make sure that the gains in genomic medicine are equitably distributed, and we need to make sure that where there is value it's being made equitable to everyone. And then I think there was also, Janita brought up the very good recommendation about integrating social determinants with the sequencing enterprise and not considering them as two separate things or in a vacuum. Finally, there was a lot of discussion that's caught up in the spellet of broadening all aspects of diversity, both in again talking about socio-anomic and clinical diversity in addition to ancestral, raising that we need to think about, you know, Asian populations for example that may not be considered under, represented by NIH standards, but certainly need to be, are understudied in terms of genomic medicine. And then Debbie will be glad to see, as am I, that the increase the diversity of our workforce recommendation is on this slide as well. So the final session on healthcare utilization, economics and value was a really, really rich discussion. So we tried to distill it down into a few points so you can give us some feedback as to whether we missed anything. So the first recommendation we clearly heard was to involve payers, for example, establishing a payer advisory board. Second, invest not only in methods for better tools, but also methods for better healthcare delivery. The third to be cognizant of regulatory issues as work in this area goes forward. And then fourth to study the cascade effects, and we discussed sort of two cascade effects, one stemming from the consequences of having secondary findings identified, and then one in terms of having the diagnostic odyssey ended and sort of truncating the cascade. So I think at this point we will turn it over to Bob and Lucila. Bob Nassbaum and Lucila Ono Machado, who are our council members, are going to be moderating this final discussion. I think for the next, for the rest of our time today, it would be helpful for you guys to imagine yourselves as our sort of thought partners in terms of figuring out how best to take advantage of all the feedback that we've heard today. I think we encourage, I think, prioritization and sort of clear consensus where there is clear consensus in terms of the feedback that you provide. So I will turn it over to Bob and Lucila.