 Good morning. Welcome to the second day of the workshop on genomics and health equity. My name is Lucia hind or if I'm a program director at NHGRI and the training diversity and health equity office. I have just a brief, a few announcements before we get started with our first talk today. First of all, I'm going to repost our code of conduct for this zoom meeting in the chat for those of us who are joining those of you who are joining for the first time today. We also wanted to give you an update on the plan for the breakout groups this afternoon. We're very excited about the enthusiastic response to this workshop. Due to the due to the large number of registrants, each breakout group will now have its own zoom link. Everyone should have received an outlook calendar invite with your assigned breakout group that includes a direct link to the breakout group. You will join the breakout from this link. The breakout groups will remind you of this process and repost the link. So please watch the chat or consult your invite that you received yesterday. And so now we're ready for the first talk of today. And it will be given by events Bonham who is the NHGRI acting deputy director and associate investigator and the social and behavioral research branch and the internal research program. The research focuses primarily on the social implications of new genomic knowledge, particularly in communities of color. These studies help genomics influences the use of the constructs of race and ethnicity and biomedical research and clinical care. Fence over to you. Great, thank you Lucia. Good morning everyone. Let's see if I can put my slides up here. Again, good morning or good afternoon where you are in the world. I am so pleased that you're able to join us today for the second day of the workshop building a genomic science health equity research agenda. Again, my name is Vince Bonham. I'm the acting deputy director at the National Human Genome Research Institute and I'm just pleased to for this workshop and for this conversation that we're having with the community with regards to work with regards to health equity and genomics. I'm also just pleased to have the level of engagement across the diversity of the communities that we hope to engage with around issues of health equity research. So I want to start out my talk with my thank yous. And I want to just thank everyone who's been involved with this workshop, but particularly the speakers, the panelists, the moderators, but also to the participants for your engagement and active involvement, both in asking questions and engagement within the chat. I encourage you today to continue to do that. I also want to thank our sister Institute centers and offices who have program directors and staff that have been participating in the workshop. This is important as we move forward with regards to our work across the National Institutes of Health with regards to health equity and particularly health equity and genomics. And finally, I want to thank our NHGRI staff. We have a number of staff members who have participated in the workshop. But I particularly want to recognize and thank the planning committee for this workshop, starting with our two co-chairs, Dr. Cho and Dr. Soojin Lee, for their leadership and guidance that they have provided over the last six weeks, seven weeks as we've been working with them to design and develop this workshop. We appreciate the time and the commitment that you've provided to date with regards to these efforts. Within NHGRI, we have a committee, a planning committee, the leads on that committee are Dr. Hendorf and Dr. Matten, but I also want to recognize my other colleagues who are important to this committee. So let's move forward and start to talk about this issue of building a agenda, a research agenda for health equity and genomics. Yesterday, Eric highlighted the strategic vision for the forefront of genomics that was developed by the Institute, and he highlighted the values and principles section. And there he focused on one of the areas that I think is extremely important as we move forward our work as a research institute with regards to what do we do to actually make the field of genomics much more equitable. I think it's worth our time to go back to this specific principle that's included here and highlight that because I think it can help to frame our research agenda to maximize the usability of genomics for all members of the public, including the ability to access genomics and healthcare, engagement, inclusion and understanding the need of diverse medically underserved groups are required. I highlight in yellow to ensure that all members of society benefit equitably from genomic advances with particular attention, given to the equitable use of genomics and healthcare that avoids exasperating and strive toward reducing health disparities. So as I would, from my perspective today and talking about developing a research agenda that we use this principle to help guide the recommendations that are made with regards to what we can do at the National Human Genome Research Institute, but what we can do across NIH to actually to use genomics in an appropriate way to increase equity and reduce inequities that exist in health and healthcare. Yesterday this slide was shown by I think a couple individuals who were speaking as an example of one of the challenging areas that we have within the field of genomic science of the lack of diversity of diverse ancestral populations in the research. This creates clearly inequities that exist that we must address as a community. We are doing that, but we must continue to figure out how do we do that and what are the new questions that we need to be asking that are related to this lack of diversity within genomics research. So I want to just take a minute and talk about establishing an equity lens to our research agenda at the National Human Genome Research Institute. And really to just talk for a minute about what are our research areas with a recognition that as an institute, we cannot do everything that is involving genomics. Now at the National Institutes of Health, across the 27 institutes and centers, genomics has been integrated, and there's targeted focused work on diseases, there's targeted focused work in various areas. But I wanted to just recall for everyone and highlight what is our research agenda, what are our research areas, and so that we think about this from an equity lens of how do we bring an equity lens to our work. So genomic technology development, can we think about what are the questions, what are the issues with regards to equity with regards to genomic technology development. Genome structure and function understanding the basis of the genome, is there an equity lens that we can bring to this research area. There's no genomics, this is an area and we have one of our breakout groups this afternoon, focus on equity and data science that's an important issue in area. How do we think about what are the research questions that are needed to be studied and to explore. And clearly genomic variation population genomics and disease, understanding diversity, ancestral diversity is important with regards to disease, and what can we do as an institute from an equity lens with regards to this area clearly I believe there is a lot that we can do. There's clinical genomics and sequencing, there's work that's already been going on through our CSER program and other initiatives in this area, but how can we advance that work. What are the questions that we need to be asking to create a much more equitable environment for clinical care and genomic medicine implementation and evaluation. What type of implementation research is needed from an equity lens perspective that to advance the work that we're doing. And then finally the ethical, legal and social implications. Here I think this is another area that there's an opportunity to do really great and important work that's advancing on work that's already going on by scholars across the country. There are some thinking about the questions of equity and inequities within the field of genomics, genomic medicine, and what we can do from an ethical, legal and social perspective and studying those and framing how to advance equity. And finally, which we're not focused on this meeting today, but it clearly has come up throughout day one as training and training a diverse workforce to be involved in genomic science research. This is an important part of equity, and it's something that we also have to be exploring as we think about a research agenda. This afternoon, each of us will participate in one breakout group that is focused on what we see as some of the broad areas for genomics and health equity and our research agenda. The social determinants of health and genomic equity. I just want to highlight the NIH is putting much more attention on social determinants of health across the institutes with regards to our research program and research agenda. And the issues that are particular for the field of genomics structural factors for genomics that would include barriers of all type with regards to health care systems, racism, the variety of structural factors that exists. What can we do what are the important questions to ask, and the issues a bench to bedside, what can we do to actually bring appropriate genomic Claire to diverse patient populations, diverse health care systems. Diverse individuals across our country. And then of course data science raises a number of issues with regards to sharing data, data science access information about equity and use that we need to explore and understand with regards to genomics. And finally, health equity research in LC, we need to make sure that we think about what are the kinds of questions. So this afternoon, we are looking forward to each of the breakout groups coming out with some recommendations. So as a group that we can help to highlight what we see is most important. But I added a six bullet of other areas. So I'm asking and encouraging you to identify other areas that we have not framed and thought about that are important for our field of genomics with regards to health equity. So again, I just want to come back to the language from our strategic vision. It's incorporated in our strategic vision, the issues of health equity is a part of our research agenda. And now's the time for us to move that forward. So we look forward to your help and support and guiding us as we move forward and agenda that's impactful, and that will advance the field of genomics and advance genomics and human health. So with that, I just want to come back and say thank you. Thank you to all of the participants today and thank you for all that have been involved in this workshop. I'm looking forward to day two. Thank you. Thank you events for that great talk. We need to move on to the other talks in the next session but please add any questions or comments for events in the chat and we'll try and get to them in the chat. So let's turn it over to Judy Cho, one of our workshop coaches to introduce the next session, Judy. Thanks Lucia and thanks fence. It's really been a privilege to be part of this initiative, the launching of this office and center, and working with the NHGRI team on developing the agenda for this terribly important initiative. It's an distinct pleasure to introduce our next two speakers. First is Dr James Hildreth, who was a renowned immunologist and the 12th president and chief executive officer of the Mary Medical College. It's really privileged that he's agreed to, he's accepted our invitation. He's known for his groundbreaking work with AIDS and HIV. He was the first African American to hold a tenured full professorship in basic research at Johns Hopkins. Dr Hildreth has led the Harry's efforts to ensure that disadvantaged communities have access to COVID-19 testing and vaccines. Graduated from Harvard as a Rhodes Scholar at Oxford obtained a PhD in immunology and an MD at Johns Hopkins. And in terms of our interactions with the Mary team, where he's been a leader in creating summer research programs for underrepresented minorities, and has been active in recruiting undergraduate students for graduate programs. The second speaker in the session is Genevieve Wojciech, who is an assistant professor of epidemiology at the Johns Hopkins Bloomberg School of Public Health in Baltimore. She is a statistical geneticist and genetic epidemiologist, and her research focus on method development for diverse populations, specifically understanding the role of genetic ancestry and environment in genetic risk and ad mixed populations. Dr Wojciech integrates epidemiology, sociology and population genetics to better understand existing health disparities in minority populations, as well as underserved populations generally. So the floor is yours. Thank you, Judy. And thank you for the invitation to the organizers to speak this morning. I'm going to try to share my screen and get started. That's not correct. Okay, here we go. So as you heard I'm the 12th president of Mary Medical College and I always like to start by reminding people that Mary's been around for a long time since 1876 and our main focus has been to provide training opportunities to African Americans at a time when they were not not available. But our other mission is to provide health care to those who otherwise wouldn't get it and we embrace the mission. We've been doing this. Dr Hildreth, your PowerPoint is not showing properly. But now looking good. Are we good. You're good to go. Okay. So I'm just going to talk about some current challenges and genomic research and genetic medicine related to health disparities and introduce you to a new partnership that we're really excited about that we think will be very impactful in this, in this area. As many of you know, African Americans and Hispanics have a higher rate of all the major chronic conditions, including cardiovascular disease, hypertension, obesity, diabetes, a number of other disorders as well. This has led to something that's been born witnessed by the whole country in the whole world, which is that for African American and Hispanics. COVID-19 was disproportionately impactful in these groups. These are data from summer of 2020. And as you can see in some places in our country, the ratio of deaths per 100,000 individuals for whites and blacks. There's a tenfold difference if you are a person of color in New Orleans, Detroit and Chicago is quite profound. This has been a constant theme for the, for the pandemic. These disparities as evidenced by the pandemic are best understood, as you probably heard through the social determinants of health. And the social determinants of health, of course, are defined as the conditions in which we are born, live, grow, work, and I would add die. This has to be the five degrees of separation, education, health care, the built environment, the social and community context and economic stability. I would like to add a sixth component to this, which is technology, which brings us to six degrees of separation and technology gap was revealed in the pandemic in the sense that health care organizations had to pivot to telemedicine and for many minority communities, the technologies are not available to allow to allow them to avail themselves of that kind of medicine. Likewise, in schools, schools pivoted to virtual learning and for a lot of students and minority communities, they were not able to do so, because they like the technology, the simple technology such as bandwidth broadband internet to do so. I would like to add that one of the major technologies that is going to be impactful for disparities is going to be genomics, and the power of the genomics, developing algorithms for treatment plans, using GWAS to find disease associations and identify a new diseases and treatments for them. So I think it's quite appropriate that we think about technology as one of those six degrees of separation. So in order to achieve health equity, we need more inclusive genomic research that will lead to improve genomic medicine opportunities for all of us. In other words, as you heard from Mr. Bonham, we've got to make sure that genomic medicine applies to all of us, not just to certain members of our population. Here you've seen these data today around the country around the world actually genetic information is being used routinely to diagnose diseases for prognosis of diseases and treatment of diseases. But there's a huge gap in the use of that information, or the ability to use that information for minority populations compared to people European descent. This slide on on the right shows you that millions of individuals have been involved in GWAS studies, but the vast majority of those individuals are European descent. And only a minority of them are persons of color, whether you're talking about Africans or Hispanics. This has implications for genetic screening for markers of disease and allowing us to intervene early in diseases. It has huge implications for polygenic risk scores and how accurately they can be applied to, to medicine and personalized medicine and pharmacogenomics will not apply to all of us unless we correct this huge gap in genetic information, as it relates to different sub populations. An example of this is shown here when people use the ClinVar resource to look for opportunities to identify risk variants. And what they found was that 50% of the observations that were made came from people European ancestry because the deep sequencing was limited to this population. And since diverse populations like many of those actual observations made in Europeans, that means that the ability to apply those disease polygenic risk scores or to do this for minority populations was limited. Therefore, in order to change that we have to have large scale sequencing of diverse populations to make sure we can actually find the rare variants that apply to those populations. Another example of this is individuals who are studying the UK Biobank found that polygenic risk scores when applied to minority populations were much less accurate for persons of color compared to people from other backgrounds. As you can see, these are only 25% as accurate for African people of African descent compared to whites. And this just again points to the fact that genomic data versus are not very diverse. And if they're not diverse, the outcomes or the ability to apply these strategies to minority populations would be much more limited than the otherwise could be. Another example relates to pharmacogenomics. A few years ago it was identified that 20 genes affect 80 medicines, prescription medicines that are used routinely in people in our country. And it was found that there are certain modifications or variants of some of these markers. That means that certain individuals cannot metabolize drugs, and therefore they do not they're not very helpful in those groups, for example, Plavix. Plavix had a blackmark black box marker added to its to the drug because it didn't work in people of Asian descent, because 75% of them, like the marker genetic polymorphism required to metabolize the pro drug. And likewise a few years later was identified that persons of Alaskan or American natives had a modification of that same marker that made war for and not work as well and those individuals. And these are just a few examples of how powerful pharmacogenomics can be. And again, if we don't have a diverse database to look for pharmacogenomic opportunities, then it's not going to be applicable to the whole population just a few examples of why this is so very important. So, in order to achieve health equity, we need more inclusive genomic research that will lead to genomic opportunities for the whole population, as I said earlier. We need to make sure that there's clinical clinically actionable genetic information available for the whole population, we need to make sure that predictive genetic risk scores are accurate for all of us. And again, pharmacogenomics is a big area where you need to make sure that databases reflect the whole population to make sure we can identify those challenges in drugs that we use, but also other opportunities, maybe to use drugs selectively in certain populations. In order to achieve this, we need to build capacity and training and in our communities diverse communities. We got to make sure that our genomic studies are large scale and include diverse populations. And very importantly, we got to make sure that we engage community and deal with the trust issues that exists for special for African Americans, when it comes to medical research and medicine in general. How do we get there. First of all, we got to make sure we engage communities and partner with them and educate them about the genetic research and the implications of it and what actually means the power of it to address diseases and help us treat diseases better. We've also got to address the realistic and justifiable distrust that populations have, especially African Americans for research, research, medical research generally. And there's a lot of basis for that and we have to acknowledge that the mistrust is legitimate and there's a reasonable basis for it. We've also then got to provide African Americans with the information they need from trusted partners to move forward. In a recent study, an analysis was done of the online discussions being had by African Americans regarding genetic research. What they found was that 20% of people were supportive of it, and excited about it. 50% were neutral, and another 30% were felt negative, had negative feelings about genomics research. What that actually means is there's a great opportunity for us to engage in a conversation with African Americans to educate them about genomics research and move this agenda forward, because a large number of them seem to be willing, at least to learn more about genomics research and what it means. You've actually seen this slide before, numerous times I suspect, showing that the large scale genomics research that has been done so far has been for the most part, vastly applied to people European descent, but only a minor fraction of those data coming from people of African descent, only about 2%. Recently, a consortium of scientists finally produced a complete genome. The genome that was published 20 years ago as a complete genome was actually not complete. And very recently a consortium of investigators have produced finally a complete human genome. It is now time for us to fill the gaps for diversity in genomics, just as these scientists have produced the full genome for the first time. Another challenge that you heard about is that there's a lack of diversity in researchers that do genetic research. Investigators tend to have a personal connection to their ancestry and their country of origin. And this can have a profound effect on how they prioritize to focus with their research. Another major issue of course is lack of funds and funding for capacity building and a skilled workforce, some other major barriers, especially when it relates to minority institutions for doing this work. Another huge challenge we have is diversity of the workforce as you've heard, less than 3% of all the genetic counselors in our country are African American. There are no genetic counseling programs that any of the 107 HPC use and quite this is a really a profoundly disturbing statistic. There are less than 20 African American medical or genetic or clinical geneticists in the whole country in the whole country. So, I'm excited to make you aware of a program that we're leading at Mary, called it together for change initiative. And this is what I think are some of the major challenges of doing the work that needs to be done. First of all, we need to have a collaboration between lots of different organizations, academic private organizations, former companies, etc. We also need to invest in capacity in organizations that are trusted by minority communities. We need to have databases that reflect diversity. So what the together for change initiative is all about is a collaboration between HPC use African American communities. Also African African academics and African African institutions and industry and corporate partners. We have four major goals in this program, and it's all going to be focused around ethics, we're going to start with ethics. Everything would do will be focused on ethics. And at the end of the day is going to got everything that we do but there are four major goals we have established within our program. We're going to establish a K through 12 stem program and HPC communities to expose children to genetics genomics research and careers in genomics. We're going to launch an HPC you genetics counseling and training program. We're also going to start a medical genetics fellowship program at our two institutions. This is going to have a profound effect on the diversity of genetic counselors and medical geneticists in our country. We're going to invest in student and faculty research and training at our two institutions, and let me just say a filter to say that my Harry is leading this, and we've been involved in this work for the last almost two years. In our partnership with Howard University so the two legacy HPC medical institutions Howard and my Harry are parting with the pharma companies to make this possible. One piece of all of this, the thing that I'm most excited about is that we're going to create a database derived from people of African descent of 500,000 genomes or genome exomes people 500,000 individuals, we're going to sequence their genomes and exomes link that to phenotypic data to create the largest database of its kind from people of African descent. And very excitingly for me is this gives us a chance also to do training in data science and artificial intelligence and other things. So the impact on training, our communities, diversity and the fields of genomics and genetics would be found in profoundly influenced by our program. We're really excited about it. We're about ready to to announce the launching of this initiative. As I said, it's been two years and the making, and I'm really excited to be able to get to this point. I'd like to acknowledge that over the past two years, it's been a lot of people involved in this work, but there are two individuals who who weren't special recognition. Dr. Neil Shanker is a senior vice president for research and innovation at me Harry and Dr. London Middell is a senior scientist at Regeneron, and these two individuals have led this relentlessly to bring us to this point. This is going to be historic collaboration between HBCUs inside our science centers and 10 pharma companies to achieve health equity through genomics research, really excited about it. And I want to thank you for allowing me to share this good news with you today. Thank you. Thank you. We'll move right to Genevieve. Sorry, bring up my slides. First I'd like to thank the organizers for having me today it really is an honor to speak to you about things that we're, we're doing with the research that we're doing as well as some thoughts and so I'm going to talk today about some of the challenges we have in genomic research. From a more I think from the further up the pipeline perspective in terms of the data. You know when I when I think about the economic health inequities and the products that we have coming out in terms of polygenic risk scores any kind of genetic testing. You kind of try to figure out where things went wrong and as you delve back further and further into the pipeline. You kind of realize that it sort of it's turtles all the way down and by turtles all the way down what I mean is that there is this Eurocentric bias in terms of what we accept as the default that has permeated every step along the way starting with who and how we sample populations, how we even measure what we accept as the default and sort of how that's built the field. And so when I talk about the default, you know, I think it's important to make distinctions. We all make assumptions when we try to look at human health or assumptions and even the way that we translate these discoveries to the clinic and so I think it really goes to down to I think this isn't news for anybody here, but it goes down to what we accept as the default and sort of how that's built the field. So the default, you know, I think it's important to make distinctions. We all make assumptions when we try to look at human health or assumptions that need to be made for assistive models assumptions to be made for different conclusions we have. But that is different in my mind than what we accept as the default, because the default influences the questions that we ask it influences who we include in our studies and who we allow to do the studies, as well as the systems that we model. So when we think about equity and moving forward with that, I think it's key to think about what we accept as a default and what are the complex systems that we need to model and sort of have the data for to move outside of this sort of box we're in. And this goes to both the point of the actual data itself and then the participants, but also who we include in the research, and how that can inform our research ideas and how we do them. So many speakers today and yesterday have already talked about talked about the lack of representation for diverse populations in tonight's studies. And often we look at the total number of participants but what I like to look at is the mean and median sample size of these published studies and the GWAS catalog. These genome wide association studies sort of the bread and butter, a lot of these trade analysis. And what I'm showing you here is from 2005 2020 so couple years old this point but the trends haven't changed that much. And what you can see on the left hand side is the mean sample size and then past two years skyrocketed gotten really high, mostly because of these really large scale biobanks and sort of these national level efforts. But it's almost entirely in European and European descent populations. The other groups that you see here, kind of our way below in terms of orders of magnitude. On the right hand side looking at the median this sort of helps you look, you know, peak beyond those really large studies that are few far between but then I think the more the bulk of what is actually being published and you see the same trends. What I think is really stark to see in these numbers is that it hasn't changed in the past, you know, decade about in terms of the median sample size. And this is important because the sample size determines of course the physical power along with a lot of other factors. And so it influences what is discovered what is published, and also shows us what this long term investment means so a lack of long term investment in these other groups leads to them being to the smaller sample sizes less well powered, and therefore, not being able to see things that you can see in the European ancestry groups. You know, we can look at the actual numbers as well as in 2018 so a few years old but again it's pretty similar, and you look at these demographics and look at the breakdown. About 78% of individuals in the GWAS catalog or European or European descent, 11% Asian and then only 1.3% Hispanic Latin American, and 2.4% African African American. You know what I really want you to take home from this is that this is not representative United States. It's not representative of global demographic. It's really not representative much at all that's based in reality, and I think that's a major problem. And what I hope to show you in this presentation is sort of what happens, what the downstream the cascade effects of this are, and how it's more than just about discovery, but it's even about how we're even able to model complex systems. So I also want to touch on one thing, I think that that goes to what was brought up yesterday and we talk about accountability or being able to assess equity, that a major barrier for this is actually are in precision of even being able to describe the study populations. When we think about, you know, I'm an epidemiologist, I'm always going to throw this in there that it's really important for us to know these populations and a major challenge to know whether our study populations, and how they relate to the target populations and who we're trying to actually help. And this can include of course race ethnicity that sort of the big big factor we don't like to describe it well we are inconsistent it's hard to look across the basis for this genetic ancestry, but also geographic location social context in general which is a whole you know big big field there and other basic demographics. And so this can create issues and barriers to really achieving this health equity. We can't even describe the populations that we're trying to help and work in. And we can see this in the way that that data is aggregated. This is from the GWAS catalog. They harmonized all of the published GWAS this point, and in order to that they had to find the groups to have some sort of level of you know to count. And what I think it's important here is that to have some accounts, they had to combine numerous constructs that are not exactly the same. So some are overlapping but not exactly the same. And so this is just the definition for European and Europeans with ancestry backgrounds. And what you have here is a combined genetic ancestry something done computationally with a reference population perhaps geography Europe nationalities such as Dutch, as well as race so Caucasian white in this case. And so when you dump all these concepts together you know it's sort of what had to be done at this stage for this sort of for this end goal, but it does sort of muddy the waters and you're looking at trying to bring evidence bases together to better understand what's going on. So, that's our presentation I work a lot right now and polygenic risk scores. And it is nobody's surprise that the lack of diversity for GWAS and for other German studies reaches into project risk for this is a few years old. In terms of the numbers says 19, but let me Duncan at all looked at published PRS and they found that European ancestry studies were vastly over represented. So about 460% compared to the world's population. And so 67% of public studies looked only in European ancestry groups so the whole publication was just on European ancestry groups. What I think is really damning is that if we took all of the published papers that had any individuals included who were African Latino Hispanic or indigenous peoples and they combined all of them. So 3.8% of publications. And so with that you have you know several continents being combined and only 3.8%. So really there is a problem. And I'm not going to go too far into right now the actual problem because I think other speakers have spoken on this already. So if you train your polygenic risk scores and the European ancestry background and you apply it to other groups, you do worse. And so having the majority done in one population, inherently creates a product and the tool, this risk calculation risk estimator that does worse than other groups, which could exacerbate health disparities and Alicia Martin really well put it on this paper in terms of what can happen. Now how badly it does depends on the outcome it depends on a lot of things, but it does worse. And so this is a problem. Now it's just I think for most people we think about this, it's usually chalked up to LD or a little frequencies, but I want to sort of add on to that that it's actually more complicated than just the sort of random factors eugenic drift. So if we look at this is paid this is a project risk score for BMI that we've applied in page the population architecture using genomics and epidemiology study a long running study from NHGRI looking at minority populations in the United States. And what you see here is first of all the performance of this risk score is not great. I also realized that BMI is not the most relevant of an outcome but I think it helps illustrate a point. And what I really want to point out to you here is that you know you have these these groups that are stratified by their self identified race and ethnicity. And they're split out here in terms of you know those groups, and it's important to note that for this R square to the model fit on a continuous scale. There is a lot of different performances you know some people, some people do it doesn't badly and sometimes it does worse. And then if you do it on a more dichotomous way of looking at it so obesity versus not severe obesity versus not you have the AUC for that prediction. But what I want to point out to you here which is sort of a little straight to you how it gets more complicated is that you have the Asian group here. And so you have a really low R square to point 017. And, you know, that's one of the worst in this of all the different groups, but when you dichotomize it goes up it's one of the highest in terms of the AUC. So why is this why is there this difference between if you even included on a continuous scale versus a binary. And what does that mean. And again, I'm going to get very obnoxious this line because this is my field but this is about the epidemiology of the trait in the population. So what I'm showing you here is the pierce decile and the proportion within that decile that is obese so BMI greater than 30. What you see here is that because the prevalence of obesity is so much lower in the Asian group compared to the other groups that even though the model thing like continuous scale is very poor. It's able to discriminate between obese and not obese just because of low prevalence. This isn't new in terms of, you know, dynamics and relationships between these different metrics. And I just want to show that polygenic risk scores and genetics in general are not fundamentally different from the rest of epidemiology and the way we think about things and therefore it's important to understand the specific epidemiology of the different groups. It's not just about the linkages equilibrium. It's not just about the frequencies, but it's about what's actually going on in those groups which requires expertise. So some other things to think about in this level. I think often we think about tools just performing worse. They're always going to perform worse. But I think it's important to understand why and it's not just a matter of the LD. And we can see this in more clinically relevant outcomes. This is a comparison of some risk scores for coronary artery disease. Looking at different groups and what they show is that there's an elevated risk for this, these outcomes in non European ancestry groups. There's African Americans and Hispanic ethnicity as it's titled here, but there's decreased performance. So again you have this relationship where you're possibly giving folks elevated risk estimates, but the performance is worse. And so what are you actually giving them in terms of a product. You know, I really want to delve in even further and show you sort of how complicated things can get and sort of it's not just about a lack of representation, but also a lack of large scale data sets with their representation that require a breath of information. So, you know, we know things PRS performed differently by race ethnicity, what happens if we look within a group so just within one population, how do their sort of add make sure portions make a difference. And then how also do you have complications with heterogeneity in the environment within that one group. So how can basically how more complicated could have possibly get the answers very complicated. And so for this we're going to look at the Hispanic Latino groups is again in page. This is about 22,000 individuals. And so I think it's important for us to note, when I say we're looking in these Hispanic Latino groups and we think about genetics. What are we actually talking about when it means the genetics of Hispanic Latino groups. And so here I'm showing you a principal components analysis each dot is a person they have all self identified as Hispanic Latino. And what you can see is that there isn't one cluster there's a lot of diversity. And I think what's important to note is that if we just have this dance Latino label and that's it that's all we allow people to self identify as that's where it stops. That same level of detail as other groups perhaps you would fail to see the bigger picture and that bigger picture is that if you love people identify themselves further. You can see substructure and you can see sort of what's going on and the complete heterogeneity and substructure that is included in this one population or is one of identified group. So another way of looking at this is add mixture proportions, each vertical line on this plot is a person. It is colored by the proportion of their genome that is due to a specific continental background. Indigenous to the Americas, you have European and you have African. And so what you see in the middle part here that strip it looking at Hispanic Latino individuals is that there is a wide spectrum. If I, you know, picked one person out of that group, it could be any range of proportions. So if people split themselves up into further you know identifying themselves, you can look at right below, which is the four for page, people could further identify themselves in the Caribbean for Puerto Rican Cuban and Dominican and continent Mexican and then these studies combined Central America and South America, because of the numbers. And so you have this heterogeneity. Now again I want to note that if you know if you picked one person out of this that I could not tell you how they further self identify based on the proportions alone, but there are differences on average that can contribute to differences in terms of the polygenic risk for. And why is that. So if you look at the project risk for is even within one group. This is the project risk for that has been stratified by folks who identify as self identify Hispanic Latino. It has been standardized in terms of you know centered and once in a deviation to normalize there. You see this relationship between ancestry and the score where the folks who have more indigenous ancestry have a higher score. And it's not a real relationship, meaning it's not actually contributing to helping predict anything. It's just for the ancestry, and you explicitly adjust for the specific ancestry that we're talking about here, you do the best in terms of your model fit. And so it's important to note that there can be this sort of incorrect relationship, even within one group that needs to be properly modeled. And you can see here you can adjust for top three PCs, but it really does the best if you do, if you explicitly look at the ancestry. So this you know I wanted to show that really it's about the ranking that changes and again you give these false relationships and estimates of risk for all of this. All right. So, you know I only have a minute left so I'm going to really go through this very quickly is that it gets more complicated so if we look at what this project risk for is actually explaining. Is it explaining the outcome is it explaining the the ancestry, you can get an estimate. What I think it's important for us to note is that if you split people up more granularly, such as Mexico, Puerto Rico, the proportion of that risk for the variance of that risk for that's being explained by different components changes, and it further changes when you think about environmental variables as well. So not only further identified in terms of these groups, but then the environmental differences between different problems is in those groups. And so one thing that we want to think about to for environment that might be very specific to different groups is immigration length of residence United States. If we go back to that BMI what we see is that there's a relationship between the length of residency in the mainland US, and this is consistent with all his answers you know groups. So how would this affect a polygenic risk score. So there are really great people working on this is the study that was led by a losing man as roads and Kristen Mcgardeau, a published last year, and what they found that if you look at the relationship between a project risk for an immigration. First of all, the actual risk for and the model does differently, different on different backgrounds based on enough to adjust things. The fact of the risk for is different again by background it's not a monolithic group there are differences in the background, and then a really complicated factor on top of that is that there is an interaction between immigration and the PRS for this as well so immigration affects the strength of PRS and believe the predict in the Cuban Americans but not in Dominican so there is sort of this again interactions interactions and heterogeneity. And what this means that the systems are very complex you need these diverse data sets it's not enough just to have a very little sort of long data set in terms of numbers, but it really requires a breath of information that I think is lacking which really needs a long term investment for these studies. I'm out of time, so I'm just going to sort of wrap this up but you know I think again I just want to sort of hammer home that it's our main challenge is not just having a lack of data that has these sort of tick boxes for different populations, but to have a rich set of data for all of those different populations that spans not only sort of the genetics, but also more individual factors more structural factors, and that really is I think one of the main challenges we have. To leave us out you know before I told you that the demographics in the GWAS catalog. We're not represented anything based in reality. They're not really based on the population demographics, they're not released to the US population. But they are actually related there they're representative of us the federally funded faculty. And I know that you know, many of us did not get into this for selfish reasons but this goes back to what we accepted the default. So in general, try to have their own lived experiences and if you have a lot of people who have certain lived experience. It's harder to model these systems that are sort of different to them. So I think it's goes to the point about training we have before. And so, thank you again as we move to help, you know precise for whom for us to remember. And thanks for everybody here especially Lindsay and Kristen for the integration stuff and Misa for help with the PRS. Thank you. Thank you Dr. Hildreth and Dr. Wojcik. I have a few minutes to moderate some q&a that have come through in the chat. I'd actually like to start with a comment that was made yesterday. And the idea is that we're looking to advance some really important areas of research, but how do we keep these this research accountable to people. And so maybe we might start with Dr. Hildreth first and then Dr. Wojcik. Yeah, how do we keep. I clarify that a little bit please. Oh sure thing. It was Dr. Clayton's. Yeah, go ahead. You know we've talked a lot about how people are through your slide that showed that only 20% of people thought that research was this research was a good idea, and 27% thought it wasn't and another big chunk didn't didn't have an opinion. So you have to think about how to, and you also showed the cover of the book about Henrietta Locke's. We have a lot of distrust to talk about here, we heard a lot yesterday about how various communities where we're not did not think that this research was desirable or was problematic. And I think that to a situation and I think we really as a community desperately need to do this to figure out how we find out what under what conditions this research would be set acceptable, and how do we make it accountable, because frankly, I think that's a gigantic unmet need. I totally agree with you and I spent a lot of time over the last two years speaking to minority groups about the vaccine because they were hesitant and then trust the, the, how the, how was made and all that and my conclusion from all the work that I've done and my work in HIV is that you have to have trusted and we can have trusted messengers unless there are more people involved in the research who look like me. You know, and I think until we change that. And that's what this together for change initiative is all about that we're doing with the pharma companies is to try to change the, the, the composition or the disposition of people who do this work. And then in genomics research and the insights we're going to gain from it are going to change how we treat a lot of diseases. And we got to make sure that all of us benefited from it equally but in order to do so we have to have more people involved in the work that are trusted in their communities. I call it leading from behind right you got to identify those folks who are trusted in their communities and let them lead the efforts, but you got to empower them with the right information and the right tools to do so. And that's the approach that we're taking so your point is well taken, but it comes down to me to having trusted messengers to lead the effort and that's what we're trying to achieve. Thank you, Dr. Hildreth doctor what chick would you like to comment on that as well. Yeah, I mean I don't really have much to add to that and that really sums it up. The only other thing that could possibly add to this and I think that there can be a slight paternalism when it comes to not believing people when they have reservations and I think that a lot of researchers just go ahead with things because like oh well they don't really, they don't actually understand the risk, not that much risk. And I think that's doing everybody a disservice. And there's really no place for that it's our job not to go ahead and do what we think is best, but to work with communities and sort of jointly figure out what's the best for them. Absolutely. Terrific thanks. I'm Dr held with I wanted to point out a number of comments in the chat, congratulating you on your new effort and also pointing out similar challenges and opportunities at some of our other participants institutions such as Hispanic serving institutions. I have a specific question about the plans to create a sample size of about 500,000 Africa ancestry genomics databases. So, given that Africa is extremely diverse how are you sampling and how representative will the data be for all of Africa, and do you think that sampling and making inference for entire Africa could lead to unintended consequences. I don't, I don't think we're inferring or hoping to have a have a sample or data set that reflects all of Africa. But I think that having sites here in the United States and collecting samples from the African continent will make our database much more useful and powerful. We're not trying to necessarily sample all of the African continent we've identified at least five academic institutions. They're going to partner with us to do this work. We think the combination of African Americans and Africans will make our data set along with the phenotypic data, a very powerful indeed. Thank you. And a question for Dr. Wojcik, are the same African African American and Hispanic Latin American people's data used repeatedly or more often across GWAS studies with that tend to reinforce conclusions based on a relatively small number of clinical experiences as well. Yes, they are but I can't speak to whether that's more often than in European ancestry groups because you know of course the big ones are you can buy up and get the same people being used over and over again. So, so yeah, so sort of, it would, it would behoove us to really look at new studies and sort of see where the gaps are beyond these sort of very large all encompassing labels to see where the patches are, you know, within that to see better, what's going on. Thank you. I'm looking there's a there's a couple of other questions about whether, for example, if before we adjust away ancestry from PRS do you think we need a first check of genetic ancestry itself could sometimes be associated with risk of certain outcomes. Yes, yeah, yeah, so I actually, I have slides, but I didn't include the study but yeah they have to know any matter so you know for one of metric trade for BMI indigenous ancestry does not predict anything, but for height it does. And so you can imagine more maybe clinically relevant trades mean the same thing where it again it depends on what your goal is for a risk for whether it's just prediction and you don't care where the power comes from. Then yes, you might not need to adjust with that specific ancestry, but if you really want something that can be portable across populations, and really have separate terms in that prediction model, then maybe you do need to adjust for it. So again, yeah, it all depends. I need to know that better. Thank you Dr. Hildreth one last question for you. How are you going to convince people about genetics where the evidence is much more elusive when it's very difficult to do that with vaccines when the evidence is so compelling. My approach to this has been to not make assumptions about what people will understand or not understand when science is communicated in a way that an ordinary person can get it, they get it. And that's what I did in my role in the mayor's press conferences and my role in the national level is to not make assumptions about people or presume that they won't understand something. There's a certain amount of a cultural humility that's been lacking in what we do. And what we're going to bring to the table and the work that we're doing is that cultural humility to assume that an ordinary person if you explain it to them and where they can understand, they will get it. Genomics has to be something that all of us take part in. It's, it's, it's just has to be that way. And we have to do the work is a heavy lift but we can do it. We can start with a humility to assume that if we do our job correctly and break it down in a way that people can understand it, we're going to be successful and I have no doubt about it. Thank you for that comment I think it reinforces many earlier discussions we had about this kind of going back to relationships among people and partnerships. Thank you both for some very interesting and thought provoking talks I will, I will note that we weren't able to get to all the comments and questions in the chat if you are able to stay on and would like to look through some of them. Please feel free NIH is also going to be capturing the comments in the chat for further consideration so let's give Dr. Wochik and Dr held with a round of applause. Thank you. Thank you very much. Thank you to you both. We are moving on to the next panel on structural factors. And one minute here well I get everything set up on my end. Okay, so this is a panel that is intending to follow up directly on the previous terrific talks that we just heard. Some brief introductions. So first, Kellan Baker is the executive director of the Whitman Walker Institute which is the research policy and education arm of Whitman Walker, a federally qualified community health center in Washington, BC. Renee Begay is a professional research assistant at the Centers for American Indian and Alaska Native Health at the University of Colorado School of Public Health, while studying as a master's of public health and Bloomberg scholar with the Johns Hopkins School of Public Health, focusing on the topic of childhood obesity. Faith Fletcher is assistant professor in the Center for Medical Ethics and Health Policy at Baylor College of Medicine, and a senior advisor to the Hastings Center, a leading bioethics research institute. Neil Rich is the LeMond Family Foundation Distinguished Professor in Human Genetics, founding director of the Institute for Human Genetics, and professor and former chair of the Department of Epidemiology and Biostatistics at the University of California, San Francisco. This panel will be moderated by Carol Horowitz, professor of population health science and policy, professor of medicine, and a practicing general internist at the ICAN school of medicine at Mount Sinai. She is the founding dean for gender equity and science and medicine, and the director of the new Institute for Health Equity Research. Dr. Horowitz, over to you. Hi, it's an honor to be with you all today. We're going to do a brief round Robin where where we're each going to give our reactions to the excellent talks we had and, and then we're going to open up for Q&A. I want to start out by briefly saying that I think what we've learned from our outstanding speakers is that systems are perfectly designed to get the results they get. And that's why we have the problems we have right now. As Dr. Hildre said, we need to uncover impact structures underlying inequities to ensure that the diverse populations who we work with are among the first, not the last to benefit from scientific advances. What about the need to diversify who is in studies, and who is on a research team. We heard some but I want to also highlight that part of dismantling harmful structures is diversifying who gets to ask the questions, what questions we ask, and how the answers are used to inform larger policies, systems and the environment. So I'm just going to give a brief example and then, and then the rest of us will share. The example I'll give is what is a gene environment interaction. And for a lot, you know, from someone newer to genomics, a lot of what I hear is that the environment is something like blood pressure. But you know, outside our genetic circle that's not what people consider the environment and Dr. Rojak shared some other examples of it. So what about the built environment in terms of where people live. We heard yesterday about April one gene variants that are nearly exclusively founded people with African ancestry and they confer an increased risk for kidney failure. This is true. You should also know that residential segregation leads to more black people living in neighborhoods where people breathe more polluted air, and that air is toxic to kidneys. So our community partners were the ones who said to us, wait a minute, this April one kidney disease. What about where people live. And sure enough, when we looked, we found that air pollution April one or so are associated with additive detrimental impact on kidney function. So with open minds and diverse teams, we can change the way we look at things and then the things we're looking at will start to change. So, Dr Baker, would you like to add. Sure. Thank you, Dr. Horowitz. It's a real pleasure to be here today and I very much appreciate all of the perspectives and expertise that has been shared. I'm coming specifically from the perspective of working closely with sexual and gender minority populations, which I think has both some specific considerations in the realm of genomics and also is really emblematic of a lot of the other challenges that we see in the terms of populations that historically have been marginalized, excluded or exploited, and the ways in which we are invisible in genomic medicine genomics research. And to your point, Dr Horowitz, where the structural problems continue to shape the degree to which the stories of different communities are are not heard in terms of some of the structural factors that I think about in relation to how genomic medicine and research is serving or often not serving diverse communities, really thinking about things such as what questions are being asked. The questions that we're asking are literally framing to your point about systems, you know they are designed to get the results that they get and the questions that we ask really do both express our worldviews but also shape our expectations and our understanding of the answers that we get back. What populations are being included are we even thinking about different populations in relation to genomics, or is it sort of, honestly business as usual in scientific research where so many populations, of color, women, children, LGBTQ people, sexual and gender minorities, people with disabilities have simply just been overlooked. And so thinking about ways to include them both as populations that are being partnered with in research efforts and to the point about diversifying the workforce, making sure that folks with lived experience of the concerns of the issues that we're trying to investigate are really being represented among the researchers themselves. Specifically about the experiences of sexual and gender minority populations, one of the biggest structural barriers that we run into is a question of mistrust the degree to which so many members of sexual and gender minority communities which is common across a number of populations experiencing disparities to the point about the experience of Henrietta Lacks, for example, and the exploitation of the data and the experiences and lives of African Americans in genomic research. There is a great deal of mistrust of what the purpose of the scientific enterprise is and how the data will be used, how the data will be communicated back to actually turn into benefit for communities. From a more personal level, just in terms of thinking about our own understanding of who we are, thinking about what is the truth of our lives, what is the truth about who we are and where does it live. So in genomic medicine, we often go back to the genome to try to identify what is it that makes us who we are. And there are some of the ways in which the genome really does tell us a lot of important truths and there are some of the ways that myopic focus on the genome may actually miss a great deal of what makes us human, what makes us who we are. For example, me as a transgender person, if you were to look at my chromosomes I haven't had a karyotype that if you look at my chromosomes you may well conclude that I should be categorized as female. That's a lot of ways in which that's not actually true. And the question of how do we understand the relationship between sex, gender, my own identity, my interactions with the environment and what my genome says about me is a very serious issue for a lot of people in general gender minority populations that thinking too about the desire to locate race in the genome, the desire to locate and eradicate disability by focusing on the genome, and really thinking about the ways in which we need to make sure that a focus on genomic medicine and research does not become an excuse to look for individual solutions to what are really structural problems. Race is not encoded in the genome. Racism is the problem. Anti LGBT bias is not something that we should be looking for what is the gay gene what is the trans gene, because the subtext of that is how do we get rid of it. We should be looking about the ways, looking at the ways in which the structural factors are making it possible, or making it impossible for communities to access and benefit from genomic medicine and research. Thank you Dr Baker. Thank you and thank you to the organizers for allowing me to be on this panel. Um, so I guess the title of our talk is structural factors and I feel like I struggled figuring out what it is that I wanted to say for this panel. Thank you, I feel like you kind of summoned up everything that I wanted to say. But coming from my tribe Navajo Nation located in Arizona. I grew up on the reservation early on in my life, but I also lived off the reservation as well. So I've had the pleasure and experience of being able to, you know, get running water when I was off the reservation and having to not have electricity and running water when I was on the reservation and engaging in my cultural practices when I was at home. So I have a very background and I also don't have a PhD at this point in time. And I struggle with that I struggle with that power and balance of not having a PhD and everybody, you know, coming to the table and already having a PhD. I also feel like my background gives me insight into what it is that genomics medicine can contribute to my community to indigenous people. And I think that's one of the things that I think is one of the barriers is that there's no diversity within these sessions within these meetings these really important meetings are really happening. Are we really being inclusive of community members and people who have, let's say, just master's degrees or just bachelor's degrees, because it's, it's a long haul to get a PhD. It's a lot of work and a lot of time and money and effort and away from our communities as native people. And to get a PhD, you have to leave all of that behind and sort of water down with who you are your culture. And I think that's one of the things that I've been struggling with in this meeting is that there's not a lot of indigenous representation. But I can. We apologize for the disconnect everybody. Renee doesn't seem to be back. I'm here. We are, we are sorry for the disconnect. Could you that we all we all got, we all got lost. So we're going to ask if you can just give maybe another, you know, 30 seconds to recap your most important things because we heard most of what you said, and then we're going to turn it over to the other two speakers and then we're going to ask some questions. They'll cut a little bit into the next talk but we all deserve a chance to say what we need to say. So Renee, can you finish up please. Thank you. I think to sum up, I recently saw a quote saying that there's never been a lack of talent or expertise or understanding within our tribal communities. Just a lack of resources. And also to say that genomics has been a part of indigenous communities for a long time. I think we've understood genomics and genetics from, you know, holistic point of view. And yeah, we didn't have necessarily the term genomics and we didn't have this fancy technology, but we understand and we, we want to engage in genetics research when it's the right time and when it's the right question that will really help and benefit our communities. Thank you. And I, and I apologize on mis-titling you I guess they're different kinds of doctorates and what you said you have a different kind of doctorate than I do but exactly as valuable. Sure. Thank you for having me here today and I appreciate all the remarks from the panelists before and during this panel around health equity and genomics. So I'll just comment on a few things really aligning my remarks with Dr. Hiltres, ethos that we need a more inclusive genomics research that will lead to improve genomic medicine opportunities that really resonated with me. So in my research I examine ethical and social implications of biospecimen collection and research processes with underrepresented populations. Similar to what Dr. Baker said I also traditionally come from the field of HIV working with black women with multiple intersections overlapping vulnerabilities at the intersections of race, gender, economic inequities as well as sexual orientation in some cases. So my interest in conducting and using my lens is to really understand unique concerns of communities, informational needs, access barriers as well as structural and intersectional barriers experienced by these groups participating in genomics research. So are there issues you meet with individuals that we traditionally view as having overlapping vulnerabilities or intersectional vulnerabilities and how do we use that information to tailor it to protocols and practices. I want to amplify the importance of also grounding health equity in all phases of genomics research to examine these complexities surrounding multiple forms of participant vulnerability rooted in structural inequities. So we don't we know and we've heard over the past two days that in the absence of participant voice, there is the potential for over estimating or underestimating even participant risk and personal agency, which can lead to over exposure to research harms are over participation of limitations and research benefits. So essentially we're hearing about community informed benefits, as well as community informed harms, we want participants to really characterize their perceptions of research harms, benefits and justice and research what does that look like when we think about conceptualizing and operationalizing research justice and research injustice from the perspective of participants. And I'll highlight if we're truly committed to building a trustworthy and equitable genomics research enterprise, we must listen more to the communities that we serve, engage them in partnership, respect their contextual knowledge as my colleague has mentioned. We must reflect inward to improve our structural competency brought from that so enhancing from the field of medicine, and we are all morally and ethically responsible to work collectively to dismantle harmful narratives, harmful frames practices and policies across the research continuum. Thank you and I look forward to questions. Thank you and I hope it's okay with all of you I switched over to first name so that we are all we we can so we're all in equal footing. Neil, if you could give some brief comments please. Hi everyone. This has been a very full conversation last couple days and I don't want to try to avoid repeating a lot of the things that have been said already. What I'm going to try to do is maybe give a little bit more of a framework on this topic, which is about inequities. And I just wrote down some points I wanted to make. First of all disparities and equities are defined in terms of social and not genetic categories lack of genetic information is just one source of lack of information in terms of health. You can't, you can't recruit subjects into studies based on genetic ancestry, but only on social categories which is the ones that matter in the first place. In this in this broad topic it's not really just an issue of genomic health inequity but health inequity is much more broadly and I'm going to discuss that further in the minute. In terms of health equity and equity I see for this was already brought up yesterday, but for areas which I think can be broken down into first prevention diagnosis treatment and outcome. So gaps in equity can occur I feel gaps in an equity in equity can occur for primarily two reasons. First gaps in knowledge, and this is related to gaps and who and has been included in research studies in the first place. And then second gaps and access to the tools and benefits of that knowledge, which is primarily socioeconomic and political and nature, and is most directly related to access to quality medical care. And then third gaps in prevention. So, regarding again this is regarding genomics. Then we're talking about genetic predictors of disease. So, in the Mendelian case, for example, carrier screening and knowledge gap may exist because found mutations and specific race or ethnic groups have not been identified, because they haven't been included in disease sequencing studies. This may also apply to incidental findings that come from exomer genome sequencing studies that are diagnostic. Case equities related to genetic ancestry, for example, we've heard a lot about this already polygenic risk scores based on whites are not always that equally applicable to other race or ethnicity groups. And they have been generated primarily in white populations. This represents a gap in knowledge, but in both cases there can also be gap in access depending on the health provider setting that individuals exist. Next is diagnosis. So, for example, we're just going to ask you just 30 more seconds please because there's a lot of questions coming in. Thanks. Okay, so, sorry. So I'm not sure what to say now so diagnosis. So exome sequencing to diagnose previous and diagnose disease. This is still not a settled issue but if you're at the genetic acid probably plays only a small role at best. So that's something that needs to be considered. But again lack of access to technology would be more likely treatment we already heard about pharmacogenetics there can be knowledge gaps there to but also gaps to access. In terms of outcome, again, large inequities occur just because of the health care setting people live in not because of the technology in and of itself. So on that topic, if I can, I think, you know, when you were just talking about where people live. One of the questions that's coming through the chat and that's coming from NHGRI is, when we think about structural factors. Can you all give a little bit of information about some of that kind of macro level factors that we need to pay attention to. Can I just give my final comment here, which is my bottom line. Thank you. It's critically to actually determine the utility and benefits through evidence based research of each aspect that I talked about above. And I can give you examples of pharmacogenetic tests for which there's the ancestry relationships and so on which not to be ineffectual because you can't lower the dose because because of side effects because if you do, you reduce the effectiveness of the drug. So there was a National Academy report on evidence framework for genetic testing. And before you define what inequities are, I think we better have a discussion about what is actually the real utility of all the genetic testing that we're talking about before we can discuss inequities. All right, I'll stop there. Thank you. So one of the one of the questions has come across is, we're thinking about how to diversify the workforce we're thinking about how to diversify who we're working with the questions we're asking. So these macro level factors that you all are talking about what's underpinning health inequities to begin with structural racism, other kinds of inequities. How do you, how do you guide genetic research is NHGRI in thinking about structural larger macro inequities, who would like to start with that. Faith, you're nodding your head. I don't know if you'd like to begin. It's an affirmative panelist. So yeah, happy to start. I'll say when we're talking about populations and being really sensitive to the needs and complexities and unique concerns and potential barriers and even resiliency is a populations. We have to make sure we're centering diverse research teams. And part of what my colleague Dr. Baker mentioned is that if you're thinking about working with black women living with HIV or other sexual minorities, you should also think about your research teams. Who's in leadership positions. This goes beyond having project coordinators and even recruiters who are from sexual minority community so we need this representation. Not only as a matter of justice but also this representation matters for our science. We know that diversity improves the quality of the science so I'll just say that's one I guess policy recommendation in terms of ensuring and holding people accountable to having different teams with expertise as well as social identities and I think we can do that through our scores are NIH impact scores. And that's one way to evaluate it. Thank you. Callan did you have something you wanted to add here about this or one of the other topics. One of the things that really comes to my mind is this question of mistrust because we talk about it a lot, you know, I brought it up it's something that comes up quite often, you know, communities do not trust researchers. And I really want to emphasize that that places the burden back on researchers. This is an adaptive response from communities to decades or centuries of exploitation, abuse, and violence. And so it is not appropriate to say, well those populations are hard to reach. I tried but I couldn't. That is our problem. That is a problem for researchers to think through and solve it is not incumbent upon communities to change to accommodate our desires to talk to them about often very you know, sort of can be complicated can be sensitive all of this information related to genomic medicine and research and so I really think this, we've seen a lot of this around coven right the the conversations about communities that are hard to reach or just aren't participating or we think they won't participate. But really it often is that people simply aren't asking researchers simply aren't asking in ways that resonate with and benefit communities. And so I think the that that structural factor of mistrust that is deeply rooted in structural racism is deeply rooted again in centuries of violence and exploitation. And it is incumbent upon us to figure out how to reach out to communities and communicate the benefits and the reasons that we are doing this research is not incumbent upon communities to simply say oh that seems reasonable short why don't you mistrust is adaptive. And so I think that it is very important for us to be thinking about real strategies to communicate the benefits and to meet communities where they are. And it's interesting how you're talking about the structure being a structure that, you know, who's the onus on so you know, try, you know mistrust or in skepticism. It's on us. Faith when you're talking about who gets to decide who's on study sections, who gets to decide what's a priority in a grant or the priority RFA is. So what you're doing is you're you're kind of not only talking about structures like pollution or residential segregation or racism but also structures of power. Renee you want to say something and then you'll. Yeah, um, I think living within or near tribal communities another way to gain trust but also really learn the reasons behind why maybe genomics research is such a difficult topic to talk about or researchers are hard to engage with I think it's important that you know researchers sort of gain that cultural humility and humility and awareness and really learn from the community because they have something to say. And if you ever engage with, you know, older folks within, you know, indigenous communities, you know they talk a lot and they have lessons to teach and bestow upon you know researchers or just anybody within the community. And so I think, you know, also the funding needs to be there to put an emphasis on community engagement and let us do community engagement for five, 10 years and be able to fund people to come to the table and fund community members to be at the table and and to speak their mind because it's not going to happen overnight it's not going to happen. Oh, you have one year to engage and then do your research. And I've been through that cycle before and it doesn't work. So, yeah, I just wanted to see. So also, you know, who gets to control the timeline of research, who gets included and when, really, really important. Yeah, would you like to add to the structure. Sure. Yeah, I just follow up on this in the comments when they just made we talked about trust as we've heard and what what engenders trust and what disengenders trust being honest with the people that you're working with and listening to them also in terms of their needs. And I'm going to give one example. We know that the Pima, the Pima population of Arizona has a very high rate of diabetes and obesity. There's a lot of genetic studies done there. But the one thing I didn't mention before is by and large genetics, genetics are not modifiable risk factors, but the diabetes that they're that they have is modifiable potentially by other interventions. And when you go to a community and start talking about genetics, you know, you have to be forthright and honest in terms of how much this actually going to benefit them in terms of the actual what they're living with currently which is a very high rate of risk. You know, you have to be very honest about this. You know, the research is nice and can lead to conclusions but I don't think polygenic risk scores per se are going to solve a diabetes problem among the Pima who live in Arizona. It's much more complicated than that. And we've heard about gene environment interaction. We cannot give short shrift to the environment part because often that is modifiable risk factor and often it's related to poverty. Right. So, so this idea that, you know, that genetics is important but it's not everything in the environment is a lot of different things is really important. Okay, so we have we've just done maybe 1520 seconds to go around to the four of you one more time. And we'll go faith kill and Renee Neil, and just if you have one message you could give the NHGRI about a direction on the forefront of genomics when it comes to structure and structural racism, what would you recommend. I have a question. I would say that we're talking about benefits to communities as well as harms to communities but I would also highlight it's important and I've heard this throughout that we think about benefits and harms to investigators doing this work so if we want to be in a diverse workforce, we really have to think about who's benefiting as we're saying earlier with these structures, especially related to large scale studies and large sample size, the ability to publish and high impact journals to disseminate the work, how that leads to promotion and tenure decisions and we see disparities in terms of who's authoring those papers and who's a part of those groups. So thinking about ways to evaluate that actually developing metrics to evaluate benefit to investigators, as well as thinking about how many diversity supplements and other training grants are coming out of some of these large scale grants if we're thinking about building the pipeline to increase diversity in the genomics workforce. Excellent point. Thank you. I want to similarly underscore the diversity of the genomics workforce, I would note there's an excellent paper out by Dr. L let and co-authors that talks about health equity tourism, and the degree to which there are, there's a lot of interest in health equity and that's wonderful and there's a great deal of expertise in communities, there's a great deal of expertise among researchers who come from those communities who historically have often been marginalized as many of my fellow panelists have noted, who have been marginalized in science and so thinking about how do we support both the cultivation and the sharing of community expertise and community knowledge and the experience of researchers who can actually speak directly to the back community knowledge. I think it's just such an important way of keeping the balance between the focus on individual issues, precision medicine, for example, but having that broader structural perspective it's never just your scientific question. Well said and we'll try to put the health equity tourism article in the chat in a few minutes. Thank you, Renee. Yeah, I think I have a couple. So it's the onus of the researcher to do their homework to engage with the community properly to live within and around the community to really understand them. And to realize that if native communities don't want to engage in genomics research and they are looking for something different like getting clean water getting access to water, getting affordable healthy foods, having access to talk to your education things like that, then you need to be able to say okay I need to back off and do you know take my research elsewhere because indigenous communities know what they need and if they say no then they say no. Thank you. Very important meal. Last comment. Yeah, these are all great comments and everything faith Renee and telling said I totally echo I don't need to repeat them. But again, I think it's really important to focus on the actual benefits. We've talked a lot about the harms but I think to give a fair and accurate assessment of the benefits. You know some people can understand them as you know as we heard earlier today from Dr. Hildreth people can understand it but we have to communicate, you know accurately correctly and honestly. All right, thank you everybody we're a few minutes over but that makes up for our loss of connection earlier on. And please take a look at the chat there's a lot of interesting discussion going on and I will echo gratitude for all of you and for everybody who organized this and turn it back to Lucia. Thank you so much to all of the panelists and to you Carol for moderating a very stimulating discussion. It's now time for us to go into breakout groups I'm going to ask it to come on in just a second and tell us a little bit more about the logistics but first I want to say to all of you that this is really everyone's chance people who have gotten a chance to speak and those who haven't to help come up with recommendations for NHGRI to develop a research agenda and genomics and health equity. So we have five breakout groups and we're looking for each of the five to come up with up to two recommendations to recommend to NHGRI so please keep your, your comments and your discussion focused on that goal that will help us to prioritize all of these really great ideas into things that NHGRI can do and things that we need to partner with other people to do. So with that I will ask Gerald to come on and review the logistics of the breakout groups. Okay, let's see. Anyone from IT. Oh, yes. Thank you. All right ladies and gentlemen. Sorry about the technical difficulties today, but in your chat you will actually see all the five breakout rooms. Okay. If you've heard or you are assigned one of the breakout rooms, if you click on the link, actually, you will take you to the breakout. Let me actually share this. Whoops. Okay, so there's two ways to get the easiest way is to look in the chat, and then click on that link of your breakout room that you're assigned to discussion group, and that will take you to your breakout room. AV supported they're ready and waiting for you. And the leaders will get there as well too. The other thing is earlier today you were sent a calendar invite. You can actually look at that, and in there will actually be the link for getting into the meeting. And I will stay online here if there are any technical difficulties. Okay, so just to recap people are going to click on a separate zoom link to join their breakout rooms and then we need everybody back at 130 Eastern to start the recap so breakout group leaders please help facilitate that if you can. If people have questions, please ask Gerald, who will stay on to answer any questions but other than that we will send you off for some great discussion. Thank you everyone see you back at 130. Okay, we'll come back everybody. We are starting to see people come back. Thank you all for your stimulating discussions in the various breakout rooms. This is our time to do the recap of the recommendations developed by each of the breakout groups and as a reminder we asked breakout group chairs to share up to two per group so that's going to be 10 recommendations. We're going to go through them very very quickly so basically no more than like a couple of minutes per group. So I'm going to go through these I apologize, I, I'm going to go through them in the order that I see them, and I don't have the names of the breakout group chairs so hopefully you all remember. Let's go ahead and start with structural factors. Who's reporting back from that group. Do we have anybody from structural factors at Malia is that your. Oh structural factors sorry. Okay, I thought there was one for us. Do you want me to, I thought there was one more before us, but if you would like me to go first I'm happy to. Okay, apologies I'm going from a list that I see here it might have might have been ordered in a different order. Are you are you prepared to go or should I skip and come back to you. Okay, technical issues so maybe you can. It's ready. Yeah, we're ready social determinants. Okay, let's. Thank you so much for the terminus was who I thought was going for you. Okay, very good. Thank you so much and thanks to the group who made the breakout discussion and identification of recommendations really easy so our recommendations are included here and not about who facilitated with me will help me in discussing the second one. You know, something that must come first before these recommendations are actualized is developing a working group to take the short list of metrics that we began to brainstorm on in our breakout metrics related to social determinants and health and genetics of genomics research and to decide on the metrics that matter. And that that brain trust should include researchers it should include policy folks and most importantly it should include community and patient leaders, and identifying what SDOH metrics and measures matter. So once that happens, our first recommendation is that an HDR require investigators to include training on using SDOH measures, and a plan towards how they should be used right. So we know that policies and systems matter. And if there is a requirement of investigators to learn about SDOH measures, and then with this new repository of metrics related to SDOH that have been approved and agreed upon. Then this first recommendation can really be realized and investigators would know what to do with them and where to go to find them. Not about. So we also our group also recognize that many scientists and geneticists may not be super familiar with what social determinants of health social determinants of health might be and what kinds of measures those might be. So my recommendation was to provide initiatives using mechanisms that support long term training opportunities to enable researchers, geneticists and other people who are funded by NHGRI to study underrepresented populations using these social determinants of health measures. And so these long term training opportunities can really enable researchers to understand what these SDOH measures are and use them in a savvy way. Okay, thank you. So apologies I think I have found the list now so I think structural factors is next is, are you ready. Yeah, I am. What I'm a little unsure of am I to share my screen or is it someone from NHGRI who has access to the slides I can share my screen, but I just, I don't want to jump in if I'm not. I believe Jen has the slides and I think she's in the middle of thunderstorms so maybe without slides you can present or well I can I can bring the slides up. Great. Yeah, thanks it'll be better if you do my internet keeps going out. Right. Okay. So, so yes, hi everyone Malia Fullerton University of Washington I was co co leading this group would need a limby from UAB unfortunately Anita had an emergency and was unable to join and so the group was stuck with me. We still nevertheless had a really tremendous conversation where a number of very important points were brought up, and we landed on three primary recommendations there probably would have been more but these are the ones that we ended up with and many many thanks for Jen for taking notes and come helping us to get to this as quickly as we did. The point was, there was a consensus in the group that we need sufficient time and funding to allow for appropriate community engagement with appropriate review of and progress reporting and criterion and then and that funding needs to be allocated equitably and to include community and the resources and so this is a really big you know it from a thinking about structural limitations at the highest level. And then there was an interesting suggestion, perhaps, or their discussion around this this is one way of possibly approaching it required training for researchers on diversity sensitivity and inclusion, and a particular role, perhaps for participant community members both in individual research studies as well as an NIH review and NHGRI decision making. So that was a very interesting thought and recommendation. And then finally, there was an interesting call for important call for collecting data at the earliest stages of research on wider abilities, demographics. So things that could be cast as resiliency, as well as barriers to participation vulnerabilities as a standard reporting element, and really prior to the initiation of a research study and to uses as part of education and improvement process. So not simply to kind of just do this in a checkbox kind of way but use this to actually bring into the research process to help people really design research that will be more broadly inclusive. And so, so those are the recommendations that our group came up with. And I will stop sharing my screen. Thank you. Thank you. And now we'll move on to implementation. Denise and Elizabeth. Yes, trying to share my screen here. I'm not sure I know how to share my screen. You would think I would know this by now in the pandemic, but Denise, I think I can, I think I can share my screen. Let's see. Make sure I share the right one. Thank you. So are the first recommendation. Since this is focused on implementation science is that research should always include a robust understanding of contextual variables so where the research occurs. And emphasize diverse settings, engaging diverse stakeholders in the design and outcome. But there was also discussion that there also should be shared metrics across studies so we can assess whether or not health equity is being achieved. So it's important to understand, for example, that a research project may occur in New York City in the limitations of that knowledge, but that we should also be really trying to get out to, you know, rural sites that have that have fewer resources and really making sure that we're getting to all of those settings. Right. Thank you. So, I think there were definitely, as Denise shared a lot of conversation around, you know, the communities and the practice, the practices that are likely to have impact based on the health condition. And included in that was discussion around ensuring that the questions are relevant. And of course as a part of that I think we heard earlier today who is asking the question. And so involving point of care clinicians, as well as other members of the team in understanding and communicating genomic research using a health equity lens and of course the group also had discussion about how some of this language and literacy communication should be structured in a way that that information, including the outcomes of the research can be shared. And so again, it's drilling down to ensure that the clinicians that are frontline or point of care are involved as much as possible, either in participating and or participating and communicating, especially actionable genomics information. Thank you. Thank you. And then I think our next breakout is from data science. Jeff and Valentina. Great. Thanks. We. This is Jeff leak. Oh, they've somebody has changed my name for me which I appreciate I was barring Valentina's link for a minute. It's the group for a really wonderful conversation and a very in depth both synchronous and asynchronous combination we we also came up with three recommendations. The first is to diversify the genomics workforce in data science by targeting HBC use MSI tribal colleges and community colleges, and thinking about increasing funding and support to these programs to build infrastructure and capacity faculty and leadership, and acknowledging the unique circumstances for faculty of these institutions who may not have the same infrastructure that exists at large R one research institutions. Second was to develop funding and support for community based research where participants are both contributors and beneficiaries to create a more diverse collection of data and acknowledging the needs of these communities in terms of data ownership and participation. And third was about focusing on privileging participants and their perspective in order to support a more diverse data collection. And then lastly to engage students earlier in the educational pipeline to attract students to computer and data science exposing them to diverse scientists and high schools, based on previously successful programs, getting people excited about science and high schools, and then supporting them throughout their career process through graduate schools and on into faculty roles and data science. It was a really rich conversation that was had around, you know, this is obviously a quick summary of a much richer conversation but it was an outstanding set of recommendations by the group. Thank you very much. And then I think that the remaining breakout group is the LC group. Who was reporting back. This is Catherine hammock ever and my fantastic co leader, been will fund. So we had a very robust discussion. We came up with eyeballing like eight or nine different recommendations. Those down to two that we thought were the most pressing and the most important. So the first was diversifying what LC sort of means who we are, and what it is that we do. We also talked about engaging and outreach to different sort of fields that would contribute a lot of valuable insight, specifically, you know, help economics, the, you know, humanity is social sciences. We, we focused on sort of expanding and adapting the scope of what LC actually means in this particular context. And that leads me to my, our second point, which is sort of this. What do we mean by phenomenon that we kept coming back to and so one of our recommendations is developing. Hopefully in it with a data driven sort of evidence based systematic process to identify and define what exactly it is that we mean by insert term. So what do we mean by health disparities what do we mean by underrepresented and marginalized what do we mean by even equity. So, and talking on to that there was this sort of this idea that over focusing on outcomes could actually exacerbate rather than mitigate certain health disparities and so being intentional about how we're defining and talking about and thinking about things to be sure we're all on the same page, and also just being really precise in what it is that we do. I don't know if did you want to add, they capture that really well, I think the only thing I'll add very quick, very quickly. I don't think I made it to our top two list but it's one that I'm here to my heart's I'm going to take the chance to say it, which is really the particular focus on disability issues as one of the areas that is that great risk of adversely impacted by advances genomics unless we really elevate that persistently. I worry a lot about that. Didn't make our top two but it was okay, bring it up. Okay, so thank you does that wrap it up for LC. Okay, thank you so much I again I want to thank everyone for their thoughtful participation and especially to the breakout group leads for leading this fast paced discussion. To let everybody know we're about to go into break and then come back at the top of the hour for two informational presentations on the Unite initiative and the all of us presentation so you'll get to learn a little bit more about what else is going on around the NIH. And in the area of health equity. And so, after that we will come back for a polling and prioritization. So please be back at the top of the hour and whatever times when you're at and we will see you then.