 Yeah, okay. We take a bit more serious note for a little bit now. I know that all of you in the audience both remotely and in this room are very complex people. You know you're complex. So we wonder why when it comes to research or to healthcare, we try to boil that complex down of you down to one or two line entries in an electronic medical record or in a research record. So the description of two or three facets of you is probably not enough and not sufficient. Would our research be stronger if we captured a more nuanced and detailed description about you and cover more factors about your physical and social environment so we can see if those have an impact on your health. And that's this charge that was given to the National Academies of Science, Engineering, and Medicine to create a panel that spent a year working on these questions. How can we better, especially in the research context, which was their focus, how can we better describe you and the things that impact your health. And so we're going to move to have invite Vince Bonham, our deputy director, Jen Wojcik and Sandra. Oh, there you are. Sandra to come up to the seats and we're going to have a discussion and Q&A about the panel. As you could see in the program, both Sandra and Jen were members on this quite august panel. So I'm going to turn over the microphone to Vince now. Thanks Larry. So I am so excited to have this conversation with two colleagues and two members of the Consensus Committee. I'm going to start by asking you to tell your story briefly. There are a number of students that are on Zoom or in the room. Can you just briefly tell, you know, what's your career and what was your career path to get to what you're doing currently? Sandra? Sure. As these paths usually go, it's somewhat circuitous. So I started in college at Stanford studying human biology and what I loved about this program is that at Stanford, there's human biology and then there's biology. So you can understand that there's probably one distinction at least that human biologists study the human. But what I loved about this program is that it allowed for students to to study the body in terms of its biological aspects, but also really encouraged and frankly required that students also study it in terms of social, cultural, other aspects. And so I was while I was studying human physiology and genetics, I also was studying courses in psychology, anthropology, sociology. And this interdisciplinary approach actually was very exciting and and I was really taken with the idea that you could answer a question by pulling on all of these different fields and disciplines and understanding the human in context. So from there, I decided that I wanted to study medical anthropology and I did that at UCSF, UC Berkeley, where medical anthropology, it's a field of anthropology, but focused on the experience of illness and disease and really understanding how illness is not only a product of the physiological processes in the body, but also how important it is to think about it in terms of social, cultural context. And so questions about the culture of medicine, the culture of science, how do bi-biotechnologies come into being in a particular place and time? All of those were questions that I was very interested in. I ended up doing fieldwork in Japan and really interested in minoritized identities and how the accumulative effect of discrimination had an effect on the body and and on health. And then I when I finished my dissertation work in my PhD, I I had the opportunity to to do this Rockefeller post-doctoral fellowship in the humanities at University of Wisconsin and it just so happened that there was a year-long seminar on genomics and identity at that institution and and I was really hooked because I was really interested in how differences under the skin were influenced by social, cultural context, at least our understanding. And so from there I applied for an F-32, a post-doctoral fellowship, and was given funding by Genome to study race and ethnicity and how scientists in genomics conceptualized and operationalized ideas around race and genetics. From there I applied for a K01 where I was I studied biobanking and population descriptors both in in the US and outside and then continued to be frankly a beneficiary of Genome's funding, studying bioethical and social questions and genomics. My journey I think looking at my CV seems like it was some great master plan, but really it was just looking right in front of me and going with the best option. The discussion earlier about you know, where were you in 1990 and 2001 and 1990 I was a toddler and 2001 I was starting high school. So the human genome was sort of always there. It sort of was always a given and sort of on the periphery. I was very lucky in high school. I grew up in a college town to have a family friend sort of take me under her wing into her lab and you know, that was a mixed blessing because I spent the summer purifying RNA from rat brains, and so I got some experience, but also all my shirts were bleached by the end of the summer. And so I went to undergrad for genetics at Cornell and while I was there I was also lucky to have exposure to a lot of different genetics research. I spent a summer looking at flies through a microscope and got pink eye, so that was very clear that wet lab was not for me. And also sort of happened to fall upon to work with a professor who was doing more computational approaches. So I say, okay, I've clearly bleached all my clothes. I've given myself pink eye. I also smelled ether directly at one point. It was a bad match for wet lab. Let's try dry lab, where I'm just in front of a computer. I probably can't hurt anybody or myself at that point. And so I ended up doing, at that point I actually was still wanting to go to medical school because that was sort of the route that was supposed to be there. And then doing a masters in human genetics and genetic epidemiology at Hopkins in the public health school. And I realized I really liked it. I was good at it. I liked it a lot and I just kept on going through the PhD. And so my undergrad was in genetics and biology and genetics specifically. My PhD is in epidemiology. So I was really interested in human health and population scale research. And then a lot of the research that I liked, so my dissertation was on the genetics of enteric infection and polio vaccine in kids and little babies. I needed a more population genetics approach. I needed to understand more about the population and who I was looking at and how that was in that context. So I did a postdoc at Stanford in population genetics in a lab there. And while I was in that lab, I found that I really liked the work that touched upon a lack of diversity in human genetics research. You know, before that, a lot of the work that I had done, you know, was in one population and most the literature that we relied upon came from these European sort of ancestry group populations. And as I went through my postdoc and kept on going through my postdoc, I also started getting a little angry because as you keep on going through it, I am biracial. And so when I do these analyses, I had to drop people like me from analysis over and over and over and over again. And then you get the results and you put them out. And there's always that little asterisk saying, you know, you read the literature and this might not apply to anybody besides this group, which means not to me, not to my loved ones, not to my friends. And so you get more and more and more angry. I'm also from Massachusetts, so we're just naturally more angry. And then, yeah, and so I decided to go the academic routes that I could pursue questions that really mattered to me. So looking at the intersection of genetic ancestry and environment and how that affects human health. And now for several years, I've been at Hopkins as faculty back in happy. So you were two of 16 individuals who are members of this committee. So can you tell me what is a consensus study? What's a consensus committee? What does the National Academies do? Well, so it was a real privilege to serve on this committee. So the consensus study really is trying to leverage the expertise that we have on a particular set of questions. And so, you know, we had disciplines represented all along the spectrum, law, history, population genetics, epidemiology, bioethics. And we convene to try to understand what the evidence is around population descriptors. And what's really nice about this committee is that despite our deep expertise in a particular discipline, we were able to come together and discuss research questions in a very productive way. I will say that many of the committee members had been thinking hard about these questions for a very long time. And so they were coming to the table with a real interest in trying to solve some serious questions and also some experience talking across disciplines, which was very helpful. I don't know if you want to add. Yeah, I mean, it was, I mean, for me, sort of sort of the junior person in the room, it was really a great opportunity to hear sort of the same topic, but from very different perspectives, which I think if you're sort of entrenched in your own research and your own sort of area, you might not consider those. And so I think having us all in the room and talking it out really helped to come to this consensus that incorporated these different fields. So why do we need population descriptors anyway? I mean, okay, so I'm going to say this from the epidemiology standpoint, which is we're very into describing our populations as sort of a whole branch of the field. And there's several reasons why. One of them is that for any study that you do, there's an understanding that the work you're doing is therefore only found and then might only be applicable for those people in the study. And therefore, there needs to be a way to relate what you did and sort of the things that you found in your study to the broader scientific community and sort of the general population and to see what what's happening in the scientific research that they're sort of paying into, especially for NIH work. And so you need a language in order to talk about that. And a language that is guided by, you know, there's these guiding principles we'll talk about I think later on, touch upon, but a way that's a common language that means sort of the same thing for everybody to sort of have that context. Because without the context, there's always a danger in human genetics research to think that any finding must be universal, right? Because we all share 99.9% of our code and therefore anything must be if this, then in everybody. And so having this language is really important to put that into context and realize not only who it's important for, but also what the limitations are. Yeah, and I would just add that this question about population descriptors may be allowing things that are embedded under the surface some assumptions about groups and how you're defining them actually become surfaced in the sense that people can talk about the ways in which you've described a group in your study explicitly. And then when it's brought to light, you can start to inspect what are the underlying assumptions around those dimensions that tie that group together. So this is all about looking at the history of the human genome project and where we are today. So just even thinking about since the start of the human genome project, can you talk about some of the types of descriptors that are used in genomics research and some of the challenges of some of those descriptors? I mean, so a lot of the descriptors sort of lie along these lines of race, especially in the United States. And I think this is for several reasons. One of them is that's one of the more readily available variables you have in a data set for when it comes to medical records or intake for cohorts is almost always asked in some capacity. And so that's sort of what you have there. And there's also, you know, people tend to categorize people into this sort of this idea that people are inherently different in these lines that sort of perpetuated. And as much as we try to sort of get away from it, those lines still are there. So typically, race is a common way. And then you can sort of rename the races to really say ancestry. So instead of saying black participants, you would say African ancestry participants, which is different words that are supposed to mean something slightly different. But for a lot of intensive purposes, in a lot of studies, they're the same in terms of handling the data. And so there's some of that there's some more computational approaches. But generally, they're on the lines of race, ethnicity that people either self identify or identified for them by either the researcher or the health care provider, or this sort of language around ancestry. Yeah, and I think we could also look at some of the requirements that NIH funded researchers need to abide by. For example, you know, the OMB categories that are used in terms of telling NIH or a funder who has been in your study. Oftentimes, those categories, despite the fact that the OMB explicitly says they should not be used as scientific variables, those categories of race and ethnicity are sometimes used in a way that's inappropriate in terms of answering a research question, but because that information is collected, researchers are asked to sort their participants in that way. There can be some confusion as to whether or not those categories should be used. And others are like geographic and ethnic identity and tribes, a variety of different ways that population descriptors have been used over the history of genomics. And so as we think about where we are today and the report's recommendations, can you talk a little bit about some of the recommendations that the committee made? Well, I think the first kind of overarching idea, there are several recommendations, but the first is really trying to resist typological thinking. This idea that human groups exist in nature as discrete entities, particularly as it relates to race and ethnicity. And so it's important to think through why you're labeling a particular group in a particular way. And when I teach medical students, I use history as a way of trying to understand systems of racialization. So what are the categories that have been used for sometimes hundreds of years? And trying to understand why those categories are used, what are the sociopolitical origins of those categories? And to disabuse people that they're somehow neutral when it comes to race, that they're really systems of difference that are embedded in political social values. And so I think the overarching idea of the recommendations is that we should really be careful and go through the important work of identifying categories that are really addressing the questions we're asking instead of assuming that groups are somehow discrete categories and to resist types. Yeah, I think another sort of theme of the recommendations is really to sort of mean what you say and say what you mean. Right, which is to say that to not rely on these poor proxies that we often have in literature. So if you say that you're using, you know, race in your models as a sort of surrogate for environment, then you should probably measure the environment itself. And then if you, you know, when it comes to, there's a big sort of discussion within the report about genetic ancestry versus genetic similarity. And that's sort of a very similar concept in which when you're, when you have your samples, you're sort of looking at a snapshot in time. And therefore you're not actually looking at the ancestry of those individuals. You have no historical samples you're comparing them to for most cases. And therefore you're not really looking at their genetic ancestry. You're looking at sort of how similar they are to other people. So that's the language sort of get away from the sort of the typological thinking, giving towards saying what exactly you mean. If you want to use a surrogate, make sure it's a good surrogate or measure things directly. That's sort of a big theme I think of a lot of the recommendations. So you're both active researchers. How has this process changed your work? Well, so, so my area of research is in the ethical, legal and social implications of genetics or LC. And so I've been very interested in how population descriptors that are used in genomics inform the way in which we make decisions around who to recruit. And also how the findings from genomics research are then related to the public. And so I'm going to talk about that last piece in terms of public communication. I mean, for many years, many, including myself, have been concerned about the ways in which genetic research gets taken up in the public domain. And some of the challenges actually that the last panel described in terms of the interpretation that sometimes suggests that there are differences between groups when perhaps those are overstated or misunderstood, but are used for different purposes by groups out in the public. And so to the extent that this report goes through describing the history of racialization in our country and as well in biomedical research, it's useful to understand how that translation to the public in terms of genomics research really does require that we're careful with language and that we are being very careful with the kinds of population descriptors that we're using in research to begin with, so that we can try to mitigate misinterpretation of racial or ethnic differences from genetic research. Yeah, I mean, I think it's definitely influenced my research program heavily. And part of it is that the report is on population descriptors and how you describe, which makes it seem like it's mostly about language and which language and words you use. But in reality, I think it's more than that. It's sort of how you define those populations, how do you form those groups to begin with. And I'm sort of more on the analytical end of things, doing a lot of still analyses on mine since I'm still a junior, so I still do all my own analyses. Most part is that there's set pipelines for often in these large-scale genetic studies and there's lots of steps in which you just do them because you've always done them that way. And that's sort of the standard that's always been in the field. And the incentive structure is to publish, you know, often and to keep on going. And so for my research program, it's sort of been pushing against a lot of those assumptions. Well, why do we do that? Do we actually need to do it that way? Is there a better way of doing it? An example is with polygenic risk scores. It's typical right now to stratify by these ancestry groups, or these, if they call them, or different populations as you well as you may define them. But why? Why do we do that? And why do we use the lines that we do and sort of pushing against that to find a different way in which we don't rely on these sort of faulty ways of going through things to sort of stop perpetuating the concerns? And then the other side that I really benefited from the report and the committee is having all these different opinions. And I always think about this line from Jurassic Park in which they say, like, you know, you scientists, like you, I'm paraphrasing it here, but you never, you know, you never stop to think like just because you can doesn't mean you should, right? And so thinking back when you have a study, okay, I have a data set, I have these variables, but like, is it really relevant for me to ask this question? Is this an important question? What's going to happen with this question after I answer this question and it sort of sent out into the world? And having those considerations from start to finish was really sort of, it was great hearing other perspectives from researchers and academics and scientists at all sort of stages to sort of see how that would play out to better inform what I do in my lab. So there were 13 recommendations, but there was also a set of guiding principles. Can you talk about the guiding principles that were part of the report? Yeah, so the committee from the very beginning wanted to set up a framework for the recommendations. And the idea is that there are principles that we want to adhere to as researchers and folks who are in the research ecosystem. And to also anticipate that we could make recommendations about the kind of research that we know about currently, but there may be research in the future that we didn't account for. And so in those cases we would want researchers to refer back to these guiding principles. So there are five, respect, beneficence, equity and justice, that's one that we treated as one, validity and reproducibility, and transparency and replicability. So I'll start us off and then Jen I'll hand over to you. But respect probably is very familiar to you all as a hallmark bioethical principle. We've done a great job as a society thinking about it in terms of the individual, but less so about groups in terms of autonomy and the ability for groups and communities to think about how they want to be described, in this case, population descriptors. And so the recommendations really encourage researchers to reach out to communities in terms of asking and engaging with them on how they would like their samples to be described for their community to be described. In terms of beneficence, also a hallmark principle. Here we're really thinking about as the researchers being conducted and population descriptors are being chosen for researchers to think through the balance of benefits and harms. History shows us that stigmatization, discrimination, some of these harms to groups are real. And so on the other side of the equation we need to think hard about what are the benefits of the research for not only individuals but groups. Equity and justice, you've heard already about the SKU in terms of who participates in genetics research and the samples that we have in our repositories. How can we ensure that not only study participants are representative of the diversity of the world that we live in, but also researchers, the research workforce, how can we ensure that the diversity exists and that the questions that are meaningful to underrepresented communities and groups are making their way into genetics research and specifically in the selection of population descriptors. Can I hand off? Yes, this was validity and reducibility. And so this is sort of saying that what you say you're looking at with your group or your population is actually what you are looking at. So I think one of the sort of extreme examples of this is that let's say that I was going to do work in Asian populations. And so I did my first work to say, okay, there's this group in the United States that has a higher rate of a certain outcome and they're Asian, identified Asian-Americans and therefore I'm going to do this work in that group. And then I'm going to reproduce this work. I want to go to the UK Biobank and I'm going to use their Asian group, right? Now, when you do that, you assume that Asian means the same thing. You're looking at the same sort of construct between the two groups. When in reality, because of history and sort of what the context is, race is a social construct and therefore it can mean different things in different social contexts, you'd be looking at completely different parts of the globe in terms of the ancestry. You'd be looking at predominantly East Asian in the United States and South Asian in the UK. And therefore, if I have a study that says I looked for this because I thought it'd be at a higher rate or this little frequency would be higher, the frequency of a certain variant in these Asian groups, it's not really valid because what you're looking for is not what you did, right, in those groups. And therefore it's not reproducible, right, because you're saying you're going through these different groups and saying, well, it should be found here in all Asian groups if you would use that descriptor in the population definition, but that's not the case because it's different constructs you're looking at. And the next one was sort of the transparency, recability. And so for this, again, it's sort of this underlying concept of you have to say what you did. And so often in literature right now, when you have a population and you're defining it, you have to go into the weeds when it comes to how people defined their groups. And often it's very subjective. So you can have, let's say, an African ancestry group in two different papers, and one of them defines it as anybody who identified as African-American or black, and then the other one says they had to have over 20% of this African component, in which case if that information wasn't there and you tried to reproduce the analysis, you wouldn't be able to do it, right, because not all the details are there. So having all the details of what you did, sort of what references you used, with the idea of being that then people could reproduce this or it could be recletable. And there's a transparency in the scientific process to really show again these potential biases and how you define your groups for the wider community. So we're coming to an end. So if there's questions, please come to the mics. So I want to ask a question about social identity. So what impact does this report have on how people perceive themselves or how people communicate information about identity? Does this have an influence on me as an individual, not as a researcher? So I would say that the report is not meant to say the way that you identify yourself is not valid, the way that you identify yourself needs to be changed. It's more, I think, on the researcher end to capture all the ways that a participant identifies themselves. On the other hand, there's also a language around genetic ancestry testing. And in that case, it does have some relevance for how somebody might see themselves and interpret the results, because often the language around that is saying, you know, okay, I'm this percent, you know, Scottish. With the idea of being that 20% of my ancestors might have come from Scotland. When reality is saying 20% of your genome looks like people who we've sampled from Scotland right now, right? And so that might have some bearing on how you identify yourself. And in sort of the vast majority of folks who are not looking at the sort of results, no, I think it's more on the researcher end of things. All right. We have a couple of questions. We'll take some questions from the audience over here. I think she was here first, actually. All right. Can you hear me or no? Yes. Okay. So full disclosure, I have a transgender child assigned to female at birth. So if you asked me 10 years ago what the sex of my child was, I would tell you female. Right now they actually don't classify themselves as male or female. They're non-binary. And I'm wondering, so my loaded question, I guess, is as generations sort of evolve over time in the categorization of themselves, how do you think that's going to affect population genetics? And so, yeah, so I can talk to her personal experience about this as well, sort of how you identify. I mean, it wasn't until 2000 that people could check more than one box for race, right? And so, and still to this day, the vast majority of forums, you can't check more than one. So it's changing. And so in the context of sort of multiracial individuals in which there's sort of a plurality of identities as well, what's interesting is that it doesn't go one to one. And so a recent Pew report showed that individuals who were multiracial given their parents, a large proportion of them didn't actually identify as multiracial. And who did identify as multiracial depended on sort of who their parents were as well. And so it's a growing sort of demographic on that end. So I do think that it is changing. And the report is very clear that any sort of framework needs to be flexible to how things will change in the future. That there's no one right way to do it, which is why the guiding principles are so important. With the idea that moving forward, there will be changes and that sort of any kind of scientific question or the framework that you use needs to be flexible to those things. Thank you. Yeah, no, I think that's a perfect question for underscoring the importance of engagement and the flexibility in which people think about these categories. No, it's great. Thanks for the discussion. This is really interesting. Thanks for writing the report. My question is about the audience. So the report and what you've described sounds like it's sort of four scientists speaking to scientists in a lot of cases. If you're doing a study, you're writing a paper, you're using a cohort. How do you think about the cohort as you're designing your study, as you're performing your study, as you're reporting your study? My question is in a similar context, but when the audience is everybody else, sort of the community, the layperson, the general public, how do you think these recommendations apply? There's just like a lot of nuance and a lot of detail that doesn't fit in a tweet. And so I'm curious what you think, how you think people should think about communicating to the public. Yeah, so a unique aspect of the report in our discussions is how we are thinking about researchers, but you'll notice at the end of the report, if you read it, we are also talking about stakeholders that are just generally in the research ecosystem. So funders, journalists, other groups, professional organizations, they all have a stake in this, and they all have a role in terms of accountability and responsibilities to help us move the needle in terms of population descriptor use. I think that as a layperson, you know, asking key questions when you pick up a paper and you read something about the latest genetics research findings that have to do with groups, asking questions about, you know, did they explain what they meant by these population descriptors? Even going and looking as if you're a student to the original research, looking at the methods section and asking, did they explain how they assign these population descriptors to groups? What can I infer in terms of what they did in the methods versus what you see in the findings? And I would say that, you know, as journal editors and newspaper editors, we all have a role in terms of asking these key questions and really interrogating the assumptions that are being made about population descriptors. And we go into some detail in the report on ways to do that. Yeah, and I would just sort of want to emphasize that there is some language about sort of journalists in particular because for most people, the exposure they'll have to our research is maybe a tweet or two, as you said, or maybe a headline that comes up when they open a new tab or something that might grab their attention. And so there's always, you know, again, the incentive structure is to have this sort of flashy headline that might grab interest, but really the call for journalists to have that nuance and to sort of make sure that it comes through at every sort of level, including, you know, the headline and the sort of intro there and the actual article to make sure that things aren't sensationalized because we know that it can be both sort of willfully and sort of from ignorant and sort of misinterpreted and so to have that nuance throughout. Thank you. So we have a question online. Yes. The question is, do you find that in addition to intakes, questionnaires, there are other barriers faced in trying to obtain slash recruit more representative genomes slash populations in these populations studies, i.e. cultural barriers? Well, the report really focuses on the trustworthiness of the research and feels that population descriptors and practices around the selection and use of population descriptors has a role to play in terms of increasing trustworthiness. And so this dogged question of how do we ensure that underrepresented groups are able to participate in research, really kind of hones in on that question of whether or not the research is trustworthy and whether they feel that they are being engaged in a way that their interests and their identities are being well represented. And so I think these principles, I hope, are going to go a long way in terms of trying to augment the trustworthiness of the research and creating relationships with communities in particular that may feel very distant and apart from the research enterprise. Yeah, I mean, I think another thing that could address this is sort of allowing people to identify themselves. Often there's missing data on this for some and then maybe you can't use that data as much. And that's because you've tried to make people say what they are within these three, four, or five boxes that you've given to them and they don't feel that's helpful for them. And so you have this large other box that sometimes comes up which is a mixed bag in terms of who is there and so allowing people to actually tell you who they are I think would also help them feel engaged in the process and allow you to actually use even the data that you already have in a better way. Question from online from Sylvia Lin. Based upon your interactions with clinicians, what are your thoughts on how clinical genetic practice can learn from this discussion of population descriptor use and research? Well, I have to say that that was outside of our charge in terms of thinking about population descriptors and clinical care. But I do think there's some lessons learned particularly in terms of how we use terms and define terms as race, ethnicity, genetic ancestry to the extent that that's being used. And I guess the engagement but also the transparency. So oftentimes I know that a patient may be described in a particular way in the EHR. Not sure how that descriptor ends up in the medical record. So some scrutiny in terms of that process. How are patient populations being described? What is the assumptions about the descriptors that are being used? Are those transparent? Are they valid? Same principles I think. I think one of the lessons could be that a lot of times in clinical care there's just these drop-down menus or buttons people press to sort of have these variables. And then clinicians sort of treat them as like, well, that's, you know, that means something sort of big. When in reality it might not. And so one example is, you know, when I went through pregnancy with my first kid, you know, there was a lot of prenatal testing and every single test they would do it. And then they would sort of like wink at me and be like, well, it's not really for you, but we did it. And so just take it with a grain of salt. And you're like, well, that's not helpful for me. I think it's still pretty useful. Like I'm sure it's still have some information. And so the idea that these clinicians think that if you're not in those boxes, then none of it pertains to you or that you have to be in one of these boxes. And that also, you know, goes into the insurance companies only paying for certain things, et cetera. But sort of push against that and say what you're really trying to get is the frequencies of these variants, maybe. And therefore if it's there, it's relevant. If it's not there, then that's relevant as well. It really doesn't matter what sort of, you know, drop down menu or button you push to sort of categorize the person. So our last question from the audience. Hi. Thanks so much for the presentation. I wondered, I have not read the guidelines, but thinking about clinical research, I mean, a lot of times I work at the FDA and a lot of times we're just looking to categorize so we can tell is this drug effective or unsafe in one gender or another gender. And I think there are some, and same with rates and ethnicity, I guess. And so those analyses rely on having enough of the population to be able to do a comparison. So I guess when I hear this, I think it's great that people can represent themselves on a form. And then I think, you know, for analysis when you have a clinical trial with 400 people, if all of them are self-identifying, it's going to be very hard to make any assumptions about, you know, if the drug is effective or not for gender or race, for example. Yeah, I mean, I would sort of respond to that and sort of saying, I mean, there's always reasons why you categorize it by different ways and to sort of say, well, why do you think that categorization is really relevant in that case? Which is true. There's always sort of this balance between sort of this precision and power in terms of your sample sizes and the granularity you look at. And I think, you know, there's important reasons to, for example, stratify by race when it comes to accountability. But if it's more about you're working about, worried about pharmacogenetic sort of effects, and then maybe it's not as important to stratify as much. And so to sort of look further upstream in the report has sort of these flow charts to help guide researchers and anybody interested parties through sort of what concepts and what constructs would be relevant for the particular question you have. And so, yeah, it does acknowledge there are questions about numbers, but seeing if maybe, you know, you can pool based on different constructs that are more relevant to the question at hand. Yeah, I think it really does matter which dimensions are of interest. And so I imagine you and I might identify as Asian American, depending on the question or the form. But, you know, so myself being of Korean ancestry, but born in the States, born in Dayton, Ohio, and then moving to the Pacific Northwest. I mean, there are many different dimensions to our lived experience. So it may be easy to just kind of lump us as Asian American, but it really depends on the questions and the dimensions of interest. So with that, I just want to say thank you. First, thank you for your work. And over the 18 months period that you've worked on this report, I think it's important for the scientific community. And I want to thank you for coming today. I encourage everyone to scan and get the report, read the report. It is really accessible, so I encourage that. So thank you. Thank you.