During this special Grand Rounds event, Josh Denny, M.D., M.S., discusses the All of Us Research Program and the future of healthcare in this unique discussion moderated by the associate director of education for the Mayo Clinic Center for Individualized Medicine, Denise M. Dupras, M.D., Ph.D.
Center for Individualized Medicine (CIM) Grand Rounds
Main presenter Josh Denny, M.D., M.S.
And we would like to welcome Doctor Denny. Many of you may know already. But doctor uh Josh Denny has been involved with all of us research program since its inception, he was a member of the advisory committee to the NIH Director of Precision Medicine Initiative working group which developed the program's initial scientific blueprint. He then led the program's initial prototyping project and the ALL of us data and research center. Dr Denny was named as the CEO of the All of us research program in January 2020. The NIH all of us research program is part of an effort to advance individualized health healthcare by enrolling 1 million or more participants to contribute their health health data. Over many years, the program aims to reflect the diversity of the United States and to include participants from groups that have been underrepresented in health research in the past today. This presentation and following conversation will describe the current landscape of precision medicine and genomic research. Identify gaps in data and participation among populations underrepresented in biomedical research and strategies to increase diversity. Describe health related research results in pharmacogenetics and hereditary disease conditions being generated by all of us and potential impact to health care providers and recognize the kinds of data available on the all of us. Researcher workbench and the opportunities available to researchers to leverage this data to advance scientific research including current activities underway at Mayo Clinic. Throughout this talk, please use Slido to submit your questions. We have planned time for Q and A at the end of the session. If you're joining us in online and have not yet claimed continuing education credit. A message will be sent out in that in this chat shortly. If you are in person and would like to uh claim credit for today, please either uh register online or look at the credit information posters uh near the entrance. Now, please join me in welcoming doctor Denny. Well, hello and good afternoon, everyone. It's a pleasure to be here. Mayo has been a wonderful partner for all of us and uh as a program and have many friends here uh from prior work as well. Uh I'm gonna give a brief uh introduction to the all of us research program and then we'll go through a Q and A program uh which I really look forward to, you know, as, as a foundation here. Um uh We have learned so much and medicine has been so transformed by what we've learned from observational longitudinal cohort studies, embedded clinical trials. And these kinds of things have really helped us uh advance knowledge and one that you know, every physician knows I remember learning about in medical school is the Framingham Heart Study, which began in 1948. This is uh some screenshots from its first publication in 1961 and really helped us learn some foundational things about blood pressure, exercise, smoking. Um Combined with other resources have really helped us identify individual risk factors that can help us improve health and concomitantly. With that, we've seen a decrease in cardiovascular mortality. Uh uh 50 70% over the ensuing decades of programs like that. And you know, there's been much that's been said about precision medicine um uh as a practicing internist when I was at Vanderbilt before moving to NIH, uh we did efforts in implementing around pharmacogenomics and some disease risk variants. But one of the real challenges is understanding how diversity applies or sorry, how genetics applies to diverse populations. And you can see this is why the vast vast majority of genomic research that's been done worldwide has been on people of European ancestry with over 90% of individuals uh and all genome wide associations that is coming from European ancestry. And so that has real impacts. The top figure here says that, you know, some of these variants that we think are pathogenic for diseases like uh hypertrophic cardiomyopathy or other conditions such as that may actually be very common in other populations and thus not pathogenic. And also a new thing is these ideas of polygenic risk scores to predict risk for cardio vascular disease, diabetes, other things like this, taking into account up to millions of genetic variants, those don't necessarily work when you translate to other populations as well. And so the mission of the all of us research program is to accelerate health research and medical breakthroughs, enabling individualized prevention treatment and care for all of us. We really want to begin with the idea of nurturing relationships longitudinally with the diverse population of a million or more individuals that we will follow for decades. Um uh that reflect the diversity of the United States. Um uh We want to take this resource that participate and contribute to make it available to as many researchers in a safe and secure way that we can. And then we want to catalyze an ecosystem of researchers, funders and communities to participate and really help make this resource as valuable as we can. We link these to a set of core values that we began the program with. It starts with participation being open to anyone in the United States. We want to reflect the rich diversity United States, importantly, participants are partners and this is something we've said from the beginning is we really don't think of this as you know, research subjects. We think of partnership with participants longitudely with the program and that's so important that we are transparent, we are authentic. Um We give participants access to information and those kinds of things as we think about moving forward, about making their data accessible, making change uh in a secure way. And you know, I want to high up this last one about being a positive catalyst for change and research because I think we see that we don't talk about research subjects nearly as much. And we think about returning information to participants, how we can return value to people um uh in other research projects we do and I think these are becoming part of the fabric of what we want to do as a country. Anyone can join all of us from across the country. Despite no matter how you come in, we share the same in common elements. It's a common consent process includes a sharing of electronic health record information sets of surveys, physical measurements, bio samples which are you'll hear are stored here at Mayo as our bio bank and building one of the world's largest bio banks. Uh And then things like mobile technologies, you can come in through the web. We have smartphone apps. This is a screenshot of um the app on my phone. And uh you can link in your electronic health records through patient portals. Electronic health records are also donated by some of our uh health care provider organizations. On behalf of the participant, we have launched nationally. In 2018, we have over 630,000 participants that have joined the program. People contribute in different ways some have things like FITBITS that they share into the program, others uh will link in electronic health records. Um uh We collect by samples and if we haven't always been able to reach the entire country to collect things like bio samples, and now we can, during COVID, we created technologies, we made new partnerships that let us really reach the entire country to donate bio specimens in different ways. We've met really focused on engaging diverse populations. And you can see the race and ethnicity of our participants there, about half of our participants identify as nonwhite or non-hispanic. Um uh And uh if you think about the larger categorization of underrepresented and biomedical research, we really think of it much broader than just race and ethnicity. It's things like age of consent, it's where you live. So rural location um is uh underrepresented category, uh gender, identity, sexual orientation, educational attainment and one of the ones we added most recently was disability. So, you know, this gets at the longitudinal nature of our program because we didn't initially ask those questions. Um So we introduced them later in the program and we did a catch up module to identify people that have disabilities. And we expect that we'll continue to do this over time. We began the program with engagement studios across the country asking questions like what would be valuable for you? What does precision medicine mean? And found out that precision medicine doesn't necessarily mean itself. Uh a lot to people, things like individualized medicine. And you know, our name, all of us come from some of the engagement we did with our participant partners, what they told us they went back and what would be valuable to them. First and foremost was genetic information. But then there's lots, lots of other kinds of things that they're interested in. Uh The majority of people said health is much bigger than things like ancestry and non health traits. Uh when you think about genetics, but we return all these kinds of information to our participants. Um And uh you know, talk some about the genetics piece. Um uh We uh return e essentially uh health uh sorry disease risk, genetic reports and also pharmacogenetics reports which we called medicine, your DNH or participants. About 96% of participants will have an actual finding in a pharmacogenetics report if they of course, have exposure to the medication of interest. Um About 2 to 3% of people have actionable results in hereditary disease risk reports, uh much like you would expect from other populations. These numbers seem fairly consistent. Uh It's early in days of this, but across different populations as well. We recently released a new uh data set in April of this year. This includes over 200 or right around 250,000 whole genome sequences making us the currently the largest uh whole genome uh resource available. Um uh in the world that's broadly accessible for research. Of course, this data set is 45% of people not from European ancestry. So that really adds to the global diversity as an as an available resource. And you can see the other kinds of things in here. We've been increasing our Fitbit data there we released here. Uh Sleep data, 15,000 may seem like a small number compared to the magnitudes of other things we're talking about. But that's actually a very large data set and we'll come at the end a little bit to that as well as we're starting to generate long read sequences as well, which um uh get a better idea of some of the more difficult areas to genotype in the genome. And some of you are very familiar with that. I'm sure um researchers use the resource through a cloud based uh um uh uh resource. This is the researcher workbench, it's a passport access model. So researchers from institutions that have um uh that are signed on to the program can rapidly get access to the program, go through trainings on human subjects and uh how to use all of us. And it's currently open to any uh us based academic or health care or nonprofit um institution. This gives you some sense of the use of the researcher workbench. We have over 5300 researchers using it now. Um since launch in 2020 the growth of that has really been uh fast compared to what we expected it to be, which is great for us. Um The publications lag as, as you can imagine, but we already have over 100 and 60 publications. And you can see that we have a number of researchers from Mayo Clinic already using it. And let me use this opportunity to encourage more of you um uh to come in and use the research uh resource. Um This is a open resource. When people come on, you get initial free credits to use the resource if you're a new researcher there, and you can see a number of different kinds of things that are being studied. Uh on the platform. We have over 500 organizations that have signed up. Um We really focus just like with participants on under-resourced institutions. So, non R ones, uh H BC US, Hispanic serving institutions, other minority serving institutions. We also have seen a number of nonprofits join including uh patient advocacy group, things such as that which bring um uh other communities that may not typically engage in research. Uh You can see it. Mayo there's 42 researchers, researchers, I hope, you know, through the next week or so, we can increase that number. Um 50% are underrepresented biomedical uh workforce. And you can see in general um uh that we are looking at the race and ethnicity of our researchers just like we were participants and we're doing programs specifically to help uh on ramps for researchers into the program and to learn how to use it. So I want to give you one example of, of the importance of diversity here and, and, and what it could look like. And many of you probably are familiar with the A one variants that are um of African ancestry that are responsible for up to 70% of kidney disease and people of African ancestry. And so if you look at the UK ban, which is a huge resource and do um uh surveys for associations on that, you can see that really nothing is, is particularly strong in that association. Nothing there is crossing the bond from any correction. But um uh a a um postdoc um uh in my research lab uh did the same analysis in the old data set from all those 98,000 people of which 24,000 were of African ancestry. And you can see dramatic signals of, of, you know, no failure like you expect to see. Um And so, you know, it's just the a lot of the population not being so, uh commonly present in some of the big resources can lead for very strong effects, not being seen if they're population specific. So the number of publications uh are coming out in a routine way. Um uh I will um uh these are available, this is one from the Mayo Clinic looking at a phenom wide association study looking at different categorizations of smoking use. And so, in this case, if you look at, you know, uh younger age smokers on the far right, compared to uh longer term smokers and, and panel, a, you see some slightly different panels of uh associations, you see commonalities, but the strength of some of those based on age uh based on uh whether you quit or not um reveals, can reveal some things about, you know, the nature of smoking um in this kind of approach. And uh similarly, with this is a study looking at the Fitbit data and looking at diseases that are associated over time with number of steps per day. So incident diagnoses uh following a time varying way, those number of steps that you take. And so as we are not surprised, walking more or running those kinds of physical activity are protective of many diseases as a primary cure doc, one of the things I love about it on the right is that you could see that, you know, you don't necessarily have to get to 10,000 uh going from 1000 to 2000, 2000 to 4000. And you know, uh that curve bends around 8000 steps per day is protective against many different diseases. And so having this re rich resource that gets into things like activities, monitors and electronic health records and genomics allows us to get into different kinds of uh analysis. So, right, now, over 457,000 people have bio specimens stored at the Mayo Clinic bio bank that maps onto 3 million tubes and nearly 12 million aliquots. Um On the way to more than 30 million II, I hear um uh as we finished out this cohort and uh you know, I say finish because I don't know that we will necessarily ever stop. I hope this keeps going and building over time, but that will get us to our million. And we recently launched a large ancillary study on top of the program called Nutrition for Precision Health, which is storing other kinds of bio specimens into the mayo clinic as well. So uh with that, I'm ready to move on to our next thing. Um Thank you so much for that intro. Thank you, Doctor Danny for an excellent beginning to today's grand rounds. So we will now move on to the second part of today's event. And we would like to welcome Doctor Dupra to this stage who will be moderating the final uh conversation and Q and A. This conversation will be included in the podcast series, Genes and Your Health, which is also moderated by Doctor Dupra. Doctor Dupra is a consultant in community internal medicine and the Associate Director for Education for uh for the Center of Individualized Medicine at Mayo Clinic, Rochester Minnesota. Again, as a reminder to our audience, uh we will be using slider for the Q and A uh and you can find the slider link uh on your Zoom uh link again. Thank you, Doctor Denny. And uh we would like to welcome Doctor Dupra. The, the floor is yours? Thank you very much. Um uh Doctor Danny, I wanna welcome you as well. Um This is a real pleasure for me to be able to sit across from you and talk about this really cool stuff. So, as mentioned, I do moderate the podcast Genes and your health. And it's fascinating to me. My training is as a uh MD internal medicine doctor and as a pharmacologist. And so my initial entree to this was in pharmacogenomics. And um I've learned an awful lot as I've gone through this about the whole field of genomics in general. So I gave Doctor Danny a few questions ahead of time and he covered a lot of the information in his talk, but I'd like to go ahead and explore a little bit more about that. So, one of the big criticisms about genomic research and individualized medicine has always been that it's been very homogenous Northern European. I think your slide showed greater than 90% and your work is really critical in terms of learning more. So can you talk more about how you're diversified and how are you getting all of these diverse individuals to participate? Thank you so much and again, thank you for this. It's wonderful to have a chance to talk with you. And continue this dialogue. You know, it starts with engagement. And uh we, before we recruited anyone before we wrote a consent, before we uh decided what our questionnaires would be and certainly, what are the program name would be? Um We began with Engagement Studios with uh 16 different populations that we identified that were underrepresented and did community engagement studios in 17 different cities. Um uh 77 total uh of those. And then we did surveys as well complemented with ongoing involvement with participants in every level of governance and every committee uh sub-committee. Uh We have boards, you know, uh across the program, we have a lot of participant uh input or, you know, potential participant input um uh of to help guide us what we were doing. We thought, you know, in the beginning when we were writing this and, and thinking about this even back to 2014 and 15, um it really was a scientific imperative um uh and a social imperative to diversify uh this research. And, you know, uh you've probably noticed in following this as well. And many of you in this field, the numbers actually haven't gotten better over time. In some cases, they've actually gotten worse. Um uh because of the availability of such big rich resources that are not diverse that have been transformative for science. And so we wanted to complement those with one that would be diverse and really think broadly about what kinds of diversity. We want to capture race and ethnicity as, as one proxy for ancestry and genetic background is one. But then there's so many other components of that that we thought were so key to help get a more complete picture of things like social determines of health and and ongoing reassessments that because obviously help changes over time, all these things will be really critical. There's a related question that came in from the audience as well, which is so you mentioned the importance of race and ethnicity, not only of the patients, but also the researchers. Can you talk a bit more about why this is so important? I'm so glad you asked that question. Uh You know, part of it is almost a promise to our participants that um uh that they want to see uh their communities represented in the researchers. But also I think it's just really important for science. We know better science is done by diverse populations. The more diverse members of the team, you can even look at things like citation uh per journal as well. There's so many different ways of looking at this and studies have been done to show that we do better science and uh have more impact with diverse teams um and ask better and more questions of the data. So, you know, it's, it's a um the access of the data is a limitless resource essentially in terms of it's just compute time on a resource. And so, you know, when we can do that in a secure and safe way and then educate, we will, you know, broaden hopefully the biomedical workforce and be part of that uh greater question. Uh because, you know, in general, science is not very diverse um in this country and we would like to help um uh broaden that. So I'm a clinician and I have patients who say I signed up for this study, they wanna know what information are they gonna get. So when you're getting whole genome, what kind of reports and what kind of genomic information are they gonna get back as participants? And is it automatic, do they sign up for it? What happens on the end of these participants who engage in the research? It is uh we have a couple different layers of information that people can get back from genetics. The first report that we launched was more of an engagement resource. It's ancestry and non health related traits. So you can see continental and subcontinental ancestry. Uh You can see things like uh whether you can taste cilantro or what kind of ear wax you have the kind of engaging uh traits that we're familiar with um from some commercial companies. And then health related was one that we built based on the whole genome sequences that delivers on uh health disease risk of 59 genes and common conditions such as hereditary cancers, uh some cardiovascular conditions. Um and uh things like hyperlipidemia. Um uh And then also another report for pharmacogenetics, both of these are reports that we've carefully developed and tested with diverse audiences. Um uh Talk thinking about things like reading level worked with the FDA. We have a um investigational device exemption with the FDA and we worked with them as being the first program really of its kind to do this kind of thing at the scale and, and the way it's being delivered directly to participants, participants get to choose whether they want these reports back. Um We really felt like, and participants told us, uh they wanted to get the report back and they wouldn't have a choice and, and you know, we see patients that uh participants that don't want the reports back um as well. And uh uh and then we have genetic counselor support for everything, uh native English and Spanish speakers and then language line support for 200 languages um uh in the process. And so those genetic counselors available, if you just want to talk about your genetic ancestry, um as well as, you know, your pharmacogenetics reports. And I know that genetic counselors really have been um uh uh uh a limit and I know that there's a shortage nationwide of genetic counselors, which has been a problem. Uh oftentimes when you get to have a patient in your office and there is a concern. Um I'm a strong believer that genetic counselors should be meeting with my patients if they wanna get their own genetic testing because it's not a yes or no. It's a probability. And that's a hard thing for a primary care clinician to talk about with the patient. You know, your test is positive or negative, but that doesn't mean you won't get cancer and that's foreign for a lot of them. So if you get a test back and the patient comes back and says yes or it says no. Is that something then that they need to come to me as a clinician and get verified by a male test or are they done our test by virtue of working our, our relationship with the FDA our research results? And they say that very much so on the report, if you have a hereditary disease risk positive uh report, meaning that you have one of these 59 genes that you have a pathogenic variant. And uh we uh if you want it, we have available and the program pays for a confirmatory clinical test. And so you can take that test directly to um uh a provider uh on the pharmacia next side, we don't have that available. Um uh And, and you of course, could decide to do um uh a separate test um in that process. Thank you. So, from a patient standpoint, what does an individual patient get out of it? Truly, the individualized medicine? But then on a bigger scale, what's the good for the population health. Those are, you know, wide ranging questions, but they really are. Um, you know, when we started this and we asked, what is the value to you part to participants at that point before the program began? So potential future participants, you know, 91% said it really was around the good that we could discover for communities like theirs. Um And so most people have that is a big part of why they're joined that they would join because of the value, the future discovery, the medications that will, you know, sort of uh be relevant or tested or discoveries that could be made to genetic interpretations that could be made, et cetera on populations like theirs. Um and understanding of that. But then it's also a about the value we can return and things like these genetic test results, these would be very expensive tests clinically. Um uh if you wanted to get them and they're not necessarily widely available. Um And we have genetic counselor support and all that kind of stuff to help walk you through it. That is a potential, very real health benefit to someone, you know, my own uh pharmacogenetics results. Uh uh I have variants in um S two C 19 which processes clopidogrel of her Plavix and other drugs as well as Slco one B one which um is involved with uh simvastatin. And you have a, a great much greater risk of toxicity with simvastatin with those variants. And, you know, those are two things. Actually, I spent a lot of time on studying um at Vanderbilt. So it's, it's very personally relevant. Obviously, I didn't know that beforehand. Um And that's, you know, new knowledge from my uh test. So, you know, those stories can be things that become very relevant um uh if you encounter those health trajectories, but we started a research program and that's why most people join um uh is as a research program. Right? Then Kay told me the same thing about the A gene in that extreme and simvastatin myalgia. So I get that that becomes near and dear to your heart when you can't go up the stairs very easily. Um So, one of the questions that came up is so what are the biggest challenges that the program faces as it aims to continue to add more ancillary studies, more participants to the already rich, rich resource? We um uh you know, our biggest challenges ahead of us and, and one of our primary goals is continue to grow the resource. And one thing I didn't talk about, but a really big goal that we're working on now is launching a pediatric program. So right now, you have to be 18 or over to join the program. Now, interestingly because people share medical records, we have more than 20,000 virtual kids, kids, uh people that have shared medical records sometimes all the way to birth uh that, you know, we just part of their, their electronic health records and people have already published uh pediatric stories based on those. Uh but, you know, we wanna build uh at least one of the world's largest, if not the largest pediatric cohort as well to answer these same kinds of questions in a diverse population. So building the pediatric ward is a big thing out of us. And then we are really starting these diversity of, of co um sort of ancillary studies which build on us as a platform. Nutrition for precision health, which I mentioned is, is looking at diet in three controlled um uh we'll have controlled feeding studies um where people take prescribed diets to study in randomized way, what the impact on diet is and things like the microbiome metabol OIC, other things like that, applying machine learning approaches new ways to assess diet which could be then applied from, you know, the small development cohort, which is actually a huge cohort for this kind of study of 10,000 people to our full 1 million. Uh as we develop ways to take a picture of your meal and you know, know something about what's in that. So um those are things that are really exciting, but, you know, uh we are still developing how to build that as a capacity. And uh and, and right now, we've just generated genetics, we have lots of other bio samples. How do we use those. So we have a lot ahead of us. We are, we are still in our beginnings. So privacy is the largest concern that one of the folks asked about regarding participating in this study. So what is the program doing to ensure that participant data is kept top secret privacy is really job one, we have to maintain trust with our participants and we have to make sure that participants know that their data will be used wisely. Um uh They also want us, they join because they want us to discover. And so we have to do that. But you know what's cool is by having a platform like ours, we can actually make access safer for participants and actually easier and faster and more democratized for researchers. So let me say why? So one of the things is so we bring all the data through a data and research center, the data are deidentified. Uh Those data are put into this resource and then people use the resource only within our cloud. So uh they can be audited, they can be monitored, there are alerts that they try to do things like download. Um uh All those things are, you know, it's a safer environment. In addition of course, being an approved researcher and you know what the projects are, and then we try to be transparent with our participants because they can at any point see what research projects are going on. And we actually get scenarios where participants ask a question on a research study and we have a resource access board that monitors the process. In addition, there's also the, just the general information security part of that, we have uh third party providers. We have a white hat hackers to test the system and try to break in and help us, you know, look things to make it safer. We continuously monitor um and really have best in class um uh security that we try to put in in place. And then just be honest, you know, we want to make sure that people know if something does happen and we don't, we don't ever promise, but we certainly are gonna do everything we can do to protect their data as best we can. So as a research study, then the data, genomic data then does not feed and populate their EMR because one of the cautions I've heard from patients particularly is I don't wanna get that data because I don't want it in my EMR because of the implications for future insurability for long term care insurance for additional life insurance policies. So how does that work with participants in the study? So the first thing is when people go through and decide whether they, they uh want a genetic result back or not is we do educate on this, what are the potential risks? And we talk about things like Gina um which protects health, uh protects you in terms of like medical health insurance but does not protect against long term care insurance or life insurance and use of genetic information. So, um, so we educate in that process and then it is the participant, you know, who decides to get the information. Uh, and then they can decide, of course, whether or not they, uh share that with their physician. We try to make that as easy as possible. And, um, for instance, uh there are ways they could, you know, put their information in and send a report, uh they can print a report. A PDF, there's a provider version as well as a patient version um that they can have. So we try to facilitate if they want to another question. Historically, studies like these have been used to harm the same communities you're targeting, how are you assuaging these fears to bring these participants in? It goes back to engagement, it's really being authentic. Um And listening, we get new ideas from people all the time. Uh Part of it is having researchers that are diverse. Well, and we talked about that earlier um to try to make sure what happens. And then uh we want to make sure that we watch out for potentially stigmatizing research. That's a uh when someone creates a new project, we ask a series of questions that helps guide someone through what might be potentially stigmatized. Uh an educational process. A series of questions that help flag for us if it might be, we look at those in a different way. Uh We also randomly audit and we searched the things that could look potentially stigmatizing. And again, participants sometimes flag uh these public uh reports of research project and says, hey, tell me a little bit more about this. Um And then we uh that's a, a flag for our resource access board to think about that project with the researchers as well. A lot of what you're talking about is really the critical importance of the engagement with the participants. And one of the questions that's come in is that what efforts have occurred to ensure genomic health literacy for all of us participants. We have tools available in our program and we have taken the uh reports that we produce and tested them with diverse audiences. And then there are the genetic counselors and we also monitor the quality of that people can rate them uh after the reaction so far, all the feedback we've gotten has been incredibly positive um across uh diverse populations that are interacting with our genetic counselors. So um uh we try to educate in the process and, and again, if someone wants to talk to somebody, even before they get a report back, they can. So we try to support people as much as we can through the whole process, you mentioned the fitbits and had 15,000 participants, I guess with Fitbit data um which are becoming popular for personal and health use. What are the data implications of using and transmitting this data? And I guess I would add on to that. What about the rest of us who might have the other devices who will be nameless? Um Any efforts to include other types of wearables that collect my health and pickleball calories and steps and things like that. You'll have to talk to me later and teach me about pickleball. It's, you know, I keep hearing it. How long are you here? I got an extra paddle. See you tonight. Uh The um So um uh you know, we actually collected anything that comes in through Apple Health Kit as well. Um I if a participant connects their device uh in that way, so if they connect it, if they authorize Apple Health Kit data, we can get some things from other kinds of devices as well. In that way, we have not harmonized, cleaned and uh deidentified that data to make it released yet. Um So we've been focused first on fitbit and this partially because we actually have a program where we're actively giving out fitbits as well. So we're growing that population, it started with the sort of bring your own device and then we're actually giving out active device uh as well. I love the 10,000 steps though. I use it with my patients all the time. So I'm pulling the article and I'm gonna print out the pictures and say see, because many people tell me they're very active. And I'll say, do you have a device? And they'll say, yes, they'll say how many steps you have? 00 my God. Doctor Duba, 2000 steps a day. And then I put up my device and I go, I had a 3500 this morning playing pickleball and I'll get another uh 7000 tonight. And they go, oh, and I'll say it all helps. So I, I love your point though about every little bit helps. And, and that's why I try to assure my patients that inactive is not good, active is better all the time. So, um that, that's a really important message I think in primary care is that sedentary is bad. Any way you look at it, get the little thing under your desk kind of stuff. OK. So next question. So what is the process of updating participant health information, new diagnoses um look like especially for the pathogenicity of variants? Two things there that are really uh important. One is we are a longitudinal study typically uh studies like framing him that's been done by bringing participants back and reassessing them, you know, with a uh a visit. Uh you know, a really important part is our electronic health records. Uh A lot of my research work before all of us began was in that space and really showing that it's a very powerful way to capture a lot of information to make genomic studies discoveries. And mayo has been a big part of that as well. Of course. Um uh So we have ongoing collections of those kinds of data. We have ongoing surveys that we release to our participants. We are in the process of designing a reassessment module to catch up the key things. And because we have electronic health records, we're able to focus those kinds of questions around things that we don't get from the electronic health record, social determinants of health, you know, the ideas of optimism and, and uh stress and loneliness and like impact of COVID, those kinds of questions that, you know, we don't get other places. Um uh we can assess through uh and spend that time, survey time instead of asking, did you have diabetes? And what was your last A one C and you know, how often do what medicines you take because we get that some other way. Um uh So, so that's how we get the data to form a resource. But in terms of participants, you bring up a great question and clinically actually, there's not a great answer for this. And, you know, when I get a lot of my, you know, I didn't order that many genetic tests as a, as a clinical provider. But when I did, you know, it's often a PDF that gets scanned into the medical record, right? That's not very actionable for future updates. In fact, it's not hard to find if you didn't know it was there, right? And so in our case, um uh you know, we're doing something that hopefully becomes easier in all other circles as well. So as we have three clinical validation labs that call and interpret the variant, uh there's one system electronically that, you know, harmonizes the repo uh variant calls and then that participant is recorded with that variant. And so if time that variant is reclassified, we essentially immediately know everyone who's been called with the prior variant. So we could update. So we know we're gonna have to regularly do this. We don't quite know what the frequency and how, you know what the exact process will be on this. But we will regularly think about this as very interpretation changes over time, potentially. Uh you know, probably more from the unknown variants towards ones that might be pathogenic as opposed to pathogenic to benign. But um but we do even educate participants that you'll get new results on these same test over time. And we're actually gonna test for more genes over time as we learn more about uh what we can test. So, uh it will be an ongoing conversation with our participants. It'll be fascinating to find out as you build the diversity and what we learn about the nonwhite genomic variation to find out. Are there other things we should be looking for? So, one of the questions that came in is um will we be able to sign up for all of us in Rochester because one participant had to drive to Winona. So we're storing the samples, right? You're storing the samples, you know, m and they, so where's our corner shop? I mean, because I got, I'm sending my patients there. I got to tell you all of us is going, that's awesome. Everyone, you know, it's all of us is for all of us, right? So anyone can join, um join all dot org. Um uh And um uh with that process, you can also directly see where your nearest locations are. We can do saliva everywhere. Um We have clinics and you can see on the map where clinics are located. Um And we're always thinking and, and where we can do more and create partnerships to get more closer to people where they're located. Well, and I would think that, you know, especially as you are looking to increase the diversity. I hate to tell you as I look around the audience, this is not a really diverse place to put up shop. It's more diverse than when I came here 40 years ago. I can tell you that, but we got a ways to go, but I do want to highlight one thing is diversity is broad. So one of the things that we are after um uh you know, broadly is people who live in rural places, people that live on farms and other sorts of um uh you know, are not in a big city and, and that's another population has traditionally been underrepresented by medical research. So, um there are many kinds of diversity and that is true. I mean, we have a very rural population here and actually there's a fair amount of diversity on the outlying areas for Rochester. Um, people that come in traditionally for different types of work. Um I haven't asked much about research but research is clearly a huge area within the all of us organization. So you mentioned that there's a group of male people who are doing research. But how does somebody get in? How does that work? Somebody has a great question and wants to access this. How does that happen? You go to research all of us.org and the process uh guides you through how to sign up and go through that mayo um is already an established institution with all of us. And so new researchers can come on uh from Mayo and, and authenticate that they're a mayo researcher and can really start doing new research and as short as a couple hours um of going through the trainings and those procedures. So um it really is easy to get on as a new researcher. We really wanted to facilitate that and I'll just give you a quick anecdote. So uh my last venerable phd student defended in uh I think it was March of 2021 and he presented his work about three years of, you know, mining the electronic health record and finding, um, combining it with, um, uh, some other publicly available resources to find medications that would lower, um, uh, cholesterol and, uh, predicted across, you know, all medications and then clinically have a tool to do this, um, in rapid fashion across the hr and so the question was asked not by me, um, uh, in the audience. Well, you've done this at Vanderbilt and we see a lot of things coming out of Vanderbilt like this, you know, uh can you do this anywhere else? And, and, you know, we had been live for about a year at that point. And so I just said, hey, Patrick, why don't you do this in all of us? And he had never applied to all of us before he uh signed up for all of us. And in a week sent me a figure that basically was the final figure, you know, with a few tweaks that was in his paper where he's able to replicate those results in all of us. So first application on the website to, you know, getting in, getting training, taking his, you know, hundreds of lines of code from the Vanderbilt system and then putting him into the all of us and going across, you know, 60 plus D hr s as opposed to one and seeing a result which, you know, replicated what he found at Vanderbilt. Wow. So one of the questions came up and I don't know how familiar you are with the tapestry study that's going on here. But one of our colleagues asked is tapestry ST the same as all of us. And if I signed up for tapestry, uh should I sign up for all of us? Is it the same? Is it redundant? How are they different? So, uh I have definitely heard of tapestry. I don't remember the details of tapestry, but what I do know for sure is they're not the same study. So um uh uh that I do know. So uh sign up for all of us as well um as tapestry and um uh we would welcome you and all of us. Uh just like I'm sure tapestry is, is very excited to have you also. Absolutely. So, one of the questions that came up is the data set is obviously wide and rich. So there will be a lot of interest from pharmaceutical companies, from data companies, from other companies. Is this gonna be a problem? And how are you gonna protect against this kind of, or these groups perhaps from wanting to mine the data? Is it an issue? You know, we think more in general, more research in safe and protected ways that are designed to further discovery uh around health um will be, you know, good things in general. And so, um but what we do is people have to use the resource again, within our model, we have a number of protections in it. So, you know, we are thinking about how we can broad researcher access over time in a safe, secure way that is engaging with our participants and, and can communicate well, um what they do. Um uh you know, it's uh I like using the example and thinking about uh the UK ban is completely uh is a very open resource for approved researchers. And we see uh pharmaceutical companies that have um you know, uh green lead are, are stopped medication development, you know, based on drug targets that they're finding in a essentially non diverse population. And so I think there's a real value to creating a resource that, you know, can understand the pharmacogenetics in our diverse population can just sort of drug targets just like the latest, you know, P CS K nine inhibitors found in a small population of individuals, African Americans in Dallas. Um uh that's basically not really present in um uh people like me. And so, so how do we um uh you know, sort of do that in that safe communicative fashion that brings, you know, uh discoveries that will benefit science and benefit people of all backgrounds uh equitably. So, combining genomic data, electronic health record data and patient reported data has to be complex but important in paint, painting a holistic picture. So how is this work going on with all of us? You know, we are building a platform. Our job is to pull those data together, uh harmonize them, put them in common data models, make them accessible, build tools for researchers to use them. And then it is really uh up to researchers to bring hypotheses. You know, we will get more science done uh by more diverse people with more diverse questions um by making that available than trying to centrally resource it. And so our job has always been to build a platform and have scientists come in and do that research. But to collect those different data streams, I didn't talk about the bio specimens we captured, but it feels like this question gets a little bit to that. So we get a DNA which we talked a lot about uh RN A cell free DNA. So that's, you know, free floating DNA, which um uh can potentially be used for all sorts of things like diagnosis of uh potential cancers in the future. Um uh We get uh serum and plasma and urine. And so uh things like NPH, our nutrition for precision health are also capturing microbiome data, uh Metabol mix data, you know, we could generate other data sets and we'll generate other data sets over time, which will provide another way of resourcing the data we need to get environmental data is one of the things we really focused on expanding. And so these different ways of answering and really thinking about the different dimensions around uh individualized health and, and it's not just genetics, it's not just, you know, sort of what's in your EMR, it's all these other things because most of the time it's, of course, fortunately, um not in the health care system. And so we want to understand those things and we want to understand not just about disease but health trajectory. And I think those will be important parts of the question as well. Quick question, how are sex and gender information collected? And will the gender sex be provided in all of us? So we have a multipart question that gets that sexual orientation and gender identity. Um We have worked with groups like pride net to really think of what these questions should be. And of course, with all of our questions, we ask participants, make sure they understand, make sure these are, you know, received well that we're asking the right questions. And so that's, that's how our questions have been designed and we've actually changed them a little bit since they started those data are available in the work. Be uh you can look through them. Uh You can actually uh anyone can just see what they look like right now without login data browser um dot Research, all of us.org, um you can even look at it on your phone and you can look at like the breakdowns on how people answer those questions. Um See what the question answers are and things like that. So self reported data. Sorry. Yes. Self reported data is, I guess maybe the key part of that. Yes. Ok. So do you anticipate long term results will help direct health policy in the United States? I hope so. And wouldn't it be great if, um, us PST F uh uh guidelines, uh started regularly thinking about real world data from things like all of us and maybe even, you know, those were active kind of investigations as part of this. So, um uh yes, I, I certainly hope that data and discoveries that are made in all of us help inform future um uh things and, and funding and guidelines and everything else about what we pay for health care, et cetera. This is a longitudinal study. How long are you gonna follow participants? And how are you gonna figure out what happens? How will death information be collected? Well, you know, the last time I was asked on this on stage, uh framing him was 74 years old. So I said 75 years love it at least. So that's, do you part? What's that? Yeah, exactly. So, you know, I want this program to keep going. I don't really see it. I don't see any reason necessarily ended and um to just follow people over time and um uh you know, we will uh we're going to continue to grow with our resources, one of which of course will be linking to death and disease um to try to bring those data. And so um uh there's actually an active award that's in consideration if you looked at our website that's actually building on things and trying to, you know, bring in more data like that. So what do you think is the biggest value to researchers? And then do you use research, researcher feedback to improve the data set and workbench tools? We do um we have uh weekly office hours that partic uh researchers call into and we get all sorts of feedback from that you can uh as you're using the workbench, there's a place to provide feedback at any point uh in the workbench and tell us the things that you find. We do. I mean, people find things all the time. You know, I'm a real believer in used resources get better. So instead of uh trying to get everything perfect and then release, you know, we get things out there kind of quick. I mean, if you look at other, some other platforms out there, you know, they release data when the chords been recruited or, you know, a longer period of time. And, you know, we release in 2020 uh our first resource two years after our national launch. And so we want to keep that kind of pace up with things. Of course, we're gonna be really good about things like uh doing the privacy uh technology and cross of it. But we're gonna release all the lab data we have um uh after some initial harmonization steps, even though some are gonna be really clean and some we gonna be much noisier. And over time we'll get that more noise. Uh You know, the, the noisier ones cleaner and the participant community, the researcher, participant community, I should say will help clean that for us. And then you can share notebooks. That's the great thing too is as people develop algorithms, those algorithms can put in our our library of tutorial uh workspaces and they can be shared, we can adopt centrally different algorithms people could develop. And so the idea is that petition, the the resource by being used will continue to get better and more usable. Well, I think you've answered this at least partially but will insurers and payers have access to information and could this impact what conditions would be covered for patients in the future? Uh Insurers and payers have no access to any individual uh information from the program. Everything in the research researcher workbench is uh has identifiers removed. And so you're always using it from a, you know, deidentified viewpoint. Um uh And so, you know, no one can and, and, and trying to reify someone is one of those things that's strictly prohibited um in the agreement. And again, it's all within our platform. So, you know, uh we can audit a workspace at any time. Um uh We can actually, we're actually looking at ways of computational looking for patterns that could look like someone trying to reid someone. We have um uh uh rules about how you report the information you discover as well. So all those things lead to, you know, protection of identity and make it really, um, not possible for an insurance company to know something about individual unless, unless they want to tell the insurance company we're running short on time. But I think my last question is Doctor Denny, I'm your patient. So, what's the downside for me to join all of us? There's no downside. Of course, it's all upside on that. I think I'd like to thank you so very much. I'm glad I get off easy. Thank you. Absolutely. Absolutely. Um, it looks like I've exhausted the questions here that I can read or understand. Um, I wanna thank you very much on the behalf of all of us for the Center for Individualized Medicine for coming to Rochester. We, uh, decided not to have a snowstorm for you. So you didn't have to drive in some god awful weather. And, uh, thank you for participating today and thank you so much for participating in our Genes and your health podcast. Well, thank you very much. Real pleasure to be here. Thank you.
