 Hello, and welcome to the eighth session of the National Human Genome Research Institute's 10-part seminar series on bold predictions for human genomics by 2030. These predictions are described in NHGRI's Strategic Vision for Improving Human Health at the Forefront of Genomics, which was published last year. Next slide to see this paper. My name is Sarah Bates. I'm the communications chief for NHGRI. I'm the moderator for today's event. It is being recorded and will be posted later on Genome TV. I'm very excited to give a short introduction before I turn it over to our guest experts for today. Next slide, please. Bold predictions for human genomics by 2030. So we've already been through seven bold predictions during the seminar series, and today is the eighth. Next slide, please. An individual's complete genome sequence, along with informative annotations will if desired, be securely and readily accessible on their smartphone. I don't know about you, but I find this prediction very exciting. The complete human genome sequence could provide a lot of new insight into missing heritability in human disease. Understanding our own sequences could lead to more personalized care, like disease prevention and treatment. After all, you are unique, and so is your genome sequence. Next slide, please. If you have been watching this incredible series, you will notice that we are mixing up the format somewhat. Some of the previous lectures have had two 25-minute talks. Today, we are going to have our two experts first go into the future with us. Now, before we go there, they will be talking for 10 minutes each about what that future looks like. Then we are going to have a time for discussion in Q&A. Then they are going to talk about the research that they are currently doing to get us to that future. So, instead of waiting also for all questions at the end, we encourage you to submit your questions now and throughout the next hour and a half. So, if you are on Zoom, please drop your questions into the Q&A box, and I will be looking for them, and then I will ask on your behalf, so we can get a very engaging conversation going with our two speakers. So, get those questions ready. Joining us today, oh, next slide, please. To present two perspectives on this auditiously bold prediction are doctors Jillian Hooker and Mike Schatz. Their full bios are available on our website. Jillian Hooker is the chief scientific officer of the health IT firm Concert Genetics. She is also the immediate past president of the National Society of Genetic Counselors and an adjunct faculty member with the University Medical Center, where she leads the research arm of the genetic counseling training program. Michael Schatz is the Bloomberg Distinguished Professor of Computer Science and Biology at Johns Hopkins University. His research is at the intersection of computer science, biology, and biotechnology, and focuses on development of novel algorithms and systems for comparative genomics, human genetics, and personalized medicine. It is an intersection of a lot of different things, Mike. So, let's start out by traveling to the year 2030. Whoa, we've had physicists open up this wormhole for us. Now we're going into the future. Whoa, we're there. Oh, look outside your window. It's the year 2030. So, for the next 20 or so minutes, we are going to be in this future, where if you look down at your smartphone, you will see your full complete human genome sequence on it. What does that look like? Who has access to it? What technology has enabled it? So, Mike and Jillian have been incredibly good sports and they're going to be in the future with us for the next 20 or 25 or so minutes. So, Mike, what does this future look like to you? Thanks, sir. What a great introduction and thanks to everyone for being on the call today. Can you advance to the next slide, please? Yep, and go ahead to the next one and the next one. All right, so here we are the year 2030 or 2130, maybe. But I would argue that this notion of carrying around genomes and genome sequencing on our smartphone has been part of our culture for decades now. Right, going back to Dr. Bowens-McCoy here with this tricorder, there's always been sort of this notion that someday, I don't know if it'll be 2030, 2130, 2230, I don't know when. But someday we'll just have this amazing capability to kind of take scans of yourself, take scans of your environment, take scans of others, just sort of immediately read off maybe their genomes, maybe other sort of biological measurements about them. Next slide. Now, the amazing thing is this is actually starting to come true. So, when I got started in genomics some 20 years ago, I was at a research institute called the Institute for Genomics Research Tiger. And at the time, DNA sequencing required massive hardware and technology to be able to sequence genome. This is a picture inside of the sequencing lab. There's this huge facility where there'd be instrument after instrument after instrument, just to sequence one human genome. Next slide. Over the next several years, we saw this technology shrink down. And today it's pretty standard to have a desktop instrument, maybe the size of a refrigerator, that's used for sequencing. Next slide. But there's some very, very cutting edge sequencing that can actually miniaturize this to the point where it's very portable. So, it's said that a tricorder could be possible. So, here's a picture inside the Hopkins Pathology Lab, taken last year, kind of at the start of the COVID-19 viral outbreak. Went for the very first time. We wanted to sequence the virus genomes that were present inside of patients. Next slide. If you zoom in on the table, that's actually the DNA sequencer that was used to actually sequence the viral genomes. Next slide. So, this technology comes from a company called Oxford Nanopore. It's a handheld sequencer. In fact, I have one right here. And amazingly, it actually has the capability to sequence DNA, RNA in real time with a handheld device. The way that it works is nucleic acids DNA or RNA pass through a little tiny hole called a nanopore. As that's happening, there's an electrical signal that is read out, the current that is produced. With the idea is that the different nucleotides passing through the pore will change. The resistance will change the current. And then we can decode that signal into nucleotides, the ACGT that everyone knows. So, at the sort of sequencing side, our tricorder is starting to come true. Next slide. And we've applied this very broadly. We've applied this to human genomes, cancer genomes, agricultural genomes, microbes, fungi, across the tree of life. We've been able to use this handheld sequencer. Next slide. On the other side, we see this sort of miniaturization in the analysis platforms. When I got started in the field, we're using big institutional clusters or just being rack after rack of computers all wired up, all trying to work together to make sense of one human genome. And those platforms, those sort of on-premise clusters still exist. But there are two really important trends that are really emerging, that are really going to change the future. Next slide. The first is computing in the very large. So, I'm one of the co-leads of something called the any stri-amble, which is an analysis platform that runs inside of the cloud, inside of the Google Cloud platform that is designed for scaling and analysis of thousands and thousands, millions and millions of genomes. And the way we do this is through the cloud computing technologies where on-demand, we can tap into data, different data and compute resources and get many thousands of computers all working together once to be able to make sense of millions and millions of genomes. Next slide. On the other side, we have an amazing advance where handheld mobile devices are suddenly becoming incredibly powerful computing devices. Shown here is one of my students, former students, Elastin Plotnik. And in his left hand there, he has the nanopore. And in the right hand, he has an iPhone, just a regular iPhone. And what's amazing, he's been able to develop software called iGenomics that can make sense of sequences that come off the instrument in order to make sense of your genetic profiles. Next slide. So available today in the App Store, on the iOS App Store for any iOS device is a total mobile analysis platform where with raw sequencing data, raw reads coming off these sorts of instruments, we can make sense of them to do genetic profile. Now, to be careful today, it's really suited for viral genomes, small microbial genomes, but we're just at the very beginning. There's no reason why these technologies couldn't tap into cloud platforms, other remote platforms to make sense of human data sets as well. And next slide. And just to kind of highlight towards the future, lots of people are excited about this outside of my own lab. So here, totally independent of us. Here's an educator and researcher, Eric Tong out in Hawaii, where he's working in a classroom with high school students to do nanopore based sequencing and actually using the iGenomics platform on iPads to be able to do the analysis. So I think we're at this really, really exciting time. It's at the bleeding edge, at the leading edge, but in 2030 or maybe just a little bit beyond, I think Genomics will be everywhere and for everyone. Thank you. Thanks, Mike. I think we can go even farther into the future if we need to. And next, Jillian will give us her view of what this feature looks like. Hi, everyone. Next slide, please. It's so wonderful to be here and really, really wonderful to get to talk after Mike and hear about the amazing things we can do and take it into the area that has kept me incredibly preoccupied for much of the last decade or so, which is how do we actually bring it out to people? How do we actually make it all real in the daily lives of people? And so I think that's really the approach I'm gonna take before I take you into my 2030 future. I wanna tell you a little bit about the past and to some extent kind of how I got here, which honestly, I'm still asking myself how I even got invited here today. It is such an incredible honor to be speaking as a part of this seminar series, partly because I am an alum of the NHGRI, Johns Hopkins Genetic Counseling Training Program, a very proud alum where I trained under Barbisa Kerr, my mentor and great friend who's shown here in the first picture where we were at the opening of the Smithsonian exhibit in 2013 for the genome exhibit there together. And coming from the program, having worked with the training program, it just really means a lot to me to be asked today. I also feel a little, I guess, surprised to have been asked because I'm not an academic scientist, or at least not a full-time one. I have a little part-time role in the academic world, but really stand in awe of all of the other speakers in the seminar series to be able to speak alongside them. And then there's a hint of irony to it as well because when I left NHGRI in 2013 to go work for a health IT company that was cataloging the entire marketplace of genetic tests, there were folks who said to me, like, why would you go do that? Everybody's just gonna have their genome sequenced in five years. Who cares about individual genetic tests? That's gonna be a thing of the past. And here we are now, I guess, eight years later, and it's still a very, very complex space. So not to say that I don't buy the prediction. I'll give you a lot of information, a lot of thoughts on why I think it is a solid prediction. And there's some really, really great words in there in particular in the prediction, but there is an irony to being here. There's also a very, very personal meaning for me to be being able to address this question. I got into genetic counseling after my younger brother died of hypertensive cardiomyopathy at the age of 25. And he's shown here with me in this picture. And hypertrophic cardiomyopathy, I believe in the future as a condition we'll be able to detect before it takes young people's lives. And I believe that genetics is a big part of the answer to figuring that out. And so it's because I believe in this future that genomic information will be accessible to people and delivered in a way that they can actually use it in their daily lives, that I do what I do and that I've done what I've done. And then when I think about the 10 year question, I think about my children, where the two people shown in the third picture on the slide here, who 10 years from now will be just entering adulthood. And so this question of what 2030 looks like for me is very much about what do I hope for them in 2030 and how do I envision them being able to use their smartphones to access their genomic information. So next slide. Here are some of the tenants that I really hold dear for 2030 is that I believe if an individual can access their genetic information on their smartphone, that that information should be accessible and affordable and lost my Zoom screen for a minute. Accessible, affordable and understandable to all people who can benefit from it. And what I mean really by accessible, and that's one of the things I think I like the most about this prediction is that the smartphone has the connotation of accessibility. We live in a world where 5.3 billion people have cell phones today. That's two thirds of the entire world population. And so I think inherently by bringing the smartphone into it, we bring into it a future of accessibility. I also think affordability is important. I think we need to get prices and payment models to a place where many, many people can benefit from being able to use technologies like this. And finally, I believe very firmly that the value of the information lies in individuals' ability to understand and use this information. So I think having access to tools and people like genetic counselors to help them understand and make it useful will really be critical. I suspect that in the future, genetic information will be most useful when delivered as one component of a data stream of health information designed to promote decision-making and allowing folks to really make smart decisions about how they allocate their healthcare resources and allocate dollars around healthcare. I think that, for example, familial hypercholesterolyne of genotypes should be delivered alongside cholesterol screening. I think that a priori germline cancer risks need to be factored in when we get to a place where we're screening by cell-free tumor DNA. And so I think we'll need tools that really integrate this information into a data stream that is designed to inform both providers and individuals about managing their health. And then I think the most important use of cell phone technology is really about accessing genetic information or it may not be as much about accessing genetic information as it is about accessing control and power over one's genetic information. And I think the idea that many, many people who have a smartphone in their hands opens up the door to using tools like smartphones to enable advanced mechanisms of control and consent over one's own health information. And so I think a key part of this future that I imagine is one where cell phones are used to enable access to people who need genomic information to help individuals make decisions to their providers when they need it, to researchers when they're doing research that is consistent and in line with the values of the individual whose data it is. I think if the ability to share data could be embedded into a smartphone app and I think there are emerging models of this and also the ability to not share when folks don't want to, I think that that would be a really, really powerful use of this. And I think underlying all of this, we need a more mature infrastructure. Next slide, please, to support that. And so what you're seeing now is an image from a commentary that I wrote earlier this year for a special issue of the American Journal of Medical Genetics on technology at the bedside. And we discussed learnings from the last eight years at Concert Genetics working with hospitals and payers and laboratories and helping them work together to build a network infrastructure to support genetics. And what we envisioned, and this was sort of the end of the paper, usually I put at the end of a talk and not at the beginning of the talk, but what we envisioned about on the timeframe of 10 years from now or about the 2030 timeframe was an integrated network where stakeholders across the genetic space are working together to deliver value. Right now, all of these stakeholders don't work together very smoothly. It's really hard for information to flow through health systems, to laboratories and to payers who in many ways hold the sort of reimbursement models that keeps the flywheel turning. So I think that the same data streams that I was talking about in the previous thread of how do we identify patients? How do we identify markers? How do we take in the data we already have to identify people for whom more information is needed, be it more genetic information be it a different type of test in order to make a particular decision. We need better tools to identify those people and in the future, we have those, right? And then those tools will be used to set providers up to know which questions they need to ask next, which tools they need to select the right tests, integrated into the provider workflow where they can then order those tests, be it an in-house test in their own hospital, a test that needs to be sent out to a different place or ideally and better yet, a test that somebody can take at home without having to come into the healthcare system. And I believe as a part of this model we need to be embracing the idea that part of the goal of these diagnostics is to keep people out of the healthcare system to help them avoid unnecessary care. I think all of these models require the integration of data allowing providers access when results come back into the system. And then we need to consider reimbursement models that value this data according to the decisions that subsequently get made with it that recognize that this test or this analysis or this interpretation is critical and important for the decisions that need to be made subsequent to it. And so getting to the point where payers are not really asking the question so much do I pay for this test or do I not but rather asking the question was the right test performed or was the right interpretation done in order for me to know that this service that's gonna be provided next so that's gonna be paid for is the right service. And then I think finally with a more connected system comes opportunities to collect data to really make truly learning systems which is what we need to get to our systems where the data is coming in in an integrated way that we can learn from it that we can develop better treatments and better models. And in order for that to happen I think the other piece that then comes up that I mentioned on the previous slide is the data control and security and transparency piece that people need to be able to allow their data to be used in a way that's consistent with their values but also have control when it's not. And with that I'll hand it back to Sarah. Thank you, Jillian, that was fantastic. I love both of these futures. So let's take a few minutes to answer a few questions before we go into more research on how it can make this future happen. So please if you're watching and you have not yet dropped the question into the Q&A box, please do so. We are getting some fantastic questions so I will ask those on your behalf. So I think one first question that folks have Jillian and Michael that we might talk about for a few minutes is this issue of privacy. The bill prediction itself, of course, has securely in the prediction because I think lots of people worry about security as far as health goes and especially as far as their smartphones go. So how do we prevent a future where we don't get Google ads based on whatever genome experience we see on our phones? I'll jump into that since I mentioned it. I think we need to think about smartphones as the way to control that, to give individuals transparency into who has their data and when, to allow temporary access to data. I think the reality is right now that in most cases a consent form is signed and then decisions are made subsequent to that without maintaining engagement of individuals from whom that data may have come. So I think it needs to be more about ongoing engagement transparency, the ability to share or the ability not to share. And I think that there are some like cool ways we could think about using smartphone technology to enable that sharing. Yeah, if I could follow up, I mean, someone's genome is their most private and personal information. And it's very precious, right? If someone steals your credit card number, I can give you a new credit card number. If someone steals your genetic data, your medical records, can't really give you a new genome, anything like that. So as researchers, we take utmost care to put all the safeguards in place to make sure that there are no breaches. I will say one advantage of the smartphone technology is it's something you have a lot of control over. You can do the analysis right there. When you're done, you can delete your results. You can have a lot of control. And other systems like the cloud system I mentioned that NHGRI Anvil system, this is, I mean, not to get overly technical, but this is something called a FedRAMP moderate system, where it's the same level of encryption, authentication, logging that you would use for any other high security government system, kind of working constantly in the background on intrusion detection to just make sure that there are no unauthorized breaches. And I should say today of the 300,000 genomes or so that we have in there, there's, in addition to keeping the genetic data, we keep very, very careful control of the consents. Some of those data are consented for open release. There's great projects like thousand genomes projects and others where the donors were really informed and really consented to make those data available. But there's also a lot of data, a lot of people are just not really comfortable to do so. And I understand, right? It's your own privacy and it's also the privacy of your children and your relatives and your aunts and uncles. There's a lot at stake there. And we're just on the cusp of this transition and we just gotta be really safe about it. So most of the data that we store is consented and only a sort of authorized users, authorized researchers can get at those sensitive data. But nevertheless, there's been millions and millions of people that have had their own genomes as sort of sequenced or genotype partially. And I think this is just the very beginning but we want to move slowly as we move forward. I think the other thing to follow up to something you said was that I think there's a lot we can learn from the financial sector. So I appreciate that you brought the credit card analogy into it. I mean, I think mobile banking has become more and more common thing by a cell phones. And I think a lot of investment has gone into securing that industry. And I think that there's a lot that healthcare can learn from the securities that have been put in place there too. But that is a question we're getting from a lot of folks is what role does industry play in this, what kind of industry and how does the US healthcare system sort of uniquely tackle this development in science, compared to other countries that might have more nationalized healthcare? Jillian, this might be a question for you. Yeah, and I have a slide about this later on too. So, but I'll circle back to it, that right now I think a lot of it is about creating incentives to respect people's autonomy and their right to privacy and their desires to control. And I think a lot of those incentives are likely to best made or brought about with regulatory policy. And I think we've seen some of it already going in this direction through some of the policies within the 21st Century Care Act that was passed a few years ago about enabling or promoting folks access to their own data or their own medical records. And though those may be cumbersome to complete, in some ways they may run counter to the interests of certain industries, right? Industries who are making a lot of money, selling data to other parties, selling medical data or selling genetic data, who may not be as interested in those policies. I think there's a lot of room for regulatory policy that really changes the dynamic and shifts the incentive in that area. And I think it also involves people collectively sort of recognizing their own ownership over the data. And then I think the benefit of that, and I think Mike touched on this, is that it builds trust. It builds trust in the whole system. If we can create better tools to allow people at a minimum just more transparency and to where their data's going and what's being done with it, better yet, more control over it. I think that the net benefit is a lot more trust in the system as a whole, a lot more people willing to participate in research. And then if I could follow up on kind of the other half of that question about the role of industry versus academia versus government. And I think Genomics has always had this partnership, kind of this three-way partnership between government, industry and academia. And they all kind of serve different communities and sort of different stages of the technology. I'd say academia tends to focus on, the really the leading edge, the bleeding edge, developing those brand new basic ideas out of basic science, basic methodologies. But often, it takes more than a single academic lab to put it into people's hands. And there's a great history of commercialization. And I think that's a good thing. I would also comment that some of the big technology companies like Google and Microsoft are suddenly getting really interested in Genomics and setting up whole arms of their sort of research of programs to develop their own expertise, their own technologies. So I think this tradition is just continuing on, but at truly massive scales where it'd be just, the logistics are impractical for a single research lab to be talking to tens of millions of customers, but that's very, very natural to some of these technology customers. And then of course, government plays an enormous role on kind of the funding side, on the regulation side, on the ethics and the policy. These are really complicated issues. I'm being sort of playful when I talk about putting track orders in everyone's hands. And at some level, I think that is our destiny, but we just got to be really careful how this proceeds because we kind of have only one chance to get this right. Yeah, absolutely. And there are a lot of questions about access. Who does get this access? Is it only, I think that also depends a little bit on how much of an optimist you are. Does everyone have equal access to these smartphones with genomic data? Or is it only a few? Only those who can afford it, only those who have a Google account. So I think actually though, before we get into more of what does that future look like, let's get into your research because we're getting a lot of really interesting questions about how do we get there? And I know you both have different visions of how we get there. So, Jillian, if you don't mind going first, talking about your research and how do we get to this future? Absolutely. Next slide, please. So I think that the approach that I took was to think about what are the key areas of challenges? What have I learned from some of my experiences, both as the leader of the National Society of Genetic Counselors last year, through my work with Concert Genetics and Digital Infrastructure and through my hobby interests and policy. What are some of the learnings that I see as challenges to be overcome as we move forward, as we move towards this future, this accessible future of having access to our genomic information on our smartphones? So I'm gonna start with workforce. Next slide, please. And then the next slide. And talk specifically about genetic counseling. I think the workforce issues around having a workforce ready and able to help implement genetics are certainly broader than genetic counseling, but my experience is most heavily concentrated in genetic counseling. So I'm gonna focus on that today. Genetic counseling in the field of genetic counseling is growing very rapidly. This chart was shared with me by the American Board of Genetic Counseling and presents the data as of just like five days ago, September 30th, 2021. They had 5,921 certified genetic counselors. Almost 6,000, I was hoping it was gonna hit 6,000 for this talk. A lot of people took their boards this summer, but we're just shy of that. But I think the last time I gave a talk in an audience like this, it was 5,000. So we are growing very, very quickly. And this is just those certified genetic counselors in the United States who've graduated from an accredited program and taken their board exam. But I think it represents most of the genetic counselors who are currently practicing in the United States. There's definitely growth worldwide at this point. Training opportunities for genetic counselors have expanded dramatically. We've gone from 29 training programs, so we're 10 years ago to 55 accredited programs and a number of other programs under development. There's also significant demand for genetic counseling training. In the last cycle, we had over 2,000 people register for the match to apply for genetic counseling programs. About 1,000 people interviewed for genetic counseling programs and about 500 were granted training slots. So there are more folks out there who are interested in this tap and who are I think largely one of the unifying characteristics of folks who are entering genetic counseling is a passion for genetics and a desire to take genetics to people, to use what they know about genetics and what they love about genetics to serve other people. And so I think that's a really a special set of characteristics that we can be capitalizing on as we build towards this future is how can we take these really innovatively minded people who care a lot about the application and bring them in to helping us build better tools for this future. So I think that's one of the tools that might be smartphone apps one day. And we definitely are seeing across the genetic counseling spectrum that folks are diversifying in their roles going into various specialty settings, working for laboratories, working for payers, working in different parts of the healthcare landscape. The other thing to really acknowledge about the growth of genetic counseling is that a major focus over the last four years or so within the National Society of Genetic Counselors but more broadly as well, has been to address the demographic homogeneity which is very white and very female. And we've recognized that much of this work lies not just in recruiting a more diverse applicant pool into the genetic counseling programs but also in building a much more inclusive community and in providing more equitable paths to becoming genetic counselors. And so, and I think that this really needs to happen to get to a place where we have a more diverse workforce where the diversity of the workforce is something that can be celebrated and nurtured within our profession and such that we can more equitably serve a more diverse population of patients. And so, Sarah, I think that ties back to the question you just punted on but this would be part of my answer. In terms of how do we provide access as we make sure that the workforce, and then this is the entire genetics workforce and I know the strategic plan from NACCR, I actually spent a good bit of time on this in the document as we make sure that the workforce that is driving this is a diverse workforce and that we're building inclusive communities to enable the growth of a diverse workforce. So that they will help us all collectively build a more equitable future. Next slide. Currently, this is from a great, great study that was published recently by researchers from AstraZeneca and Mark who did a geospatial analysis of the distribution of genetic counselors. I think another topic that comes up a lot is this, is there a shortage of genetic counselors? And I actually right now don't perceive that there is a shortage to genetic counselors. I think there are upstream bottlenecks of how folks get to genetic counselors. And then the other thing that's clear from this study certainly is that there's a distribution problem for genetic counselors in that there's widely unequal access to genetic counselors across the country and many people have very limited access. So nationally there were on average 1.49 genetic counselors for a hundred thousand people and it varied between 0.17 genetic counselors and 5.7 for a hundred thousand at the state level. 71% of all US residents live within a 30 minute drive time to a genetic counselor and 82% of people in metropolitan areas are within 30 minutes of a genetic counselor. In contrast, only about 6% of people living in non-metro areas live within 30 minutes of a genetic counselors. So there's really, really desperate access and what you see here is you see on the map the location and the black dots of genetic counselors across the country and then a heat map of the drive time for individuals, for the people across the country to the nearest genetic counselors. So you can see there are big parts of the country where there's a pretty significant drive time to get to the genetic counselor. And geography is just one of demographic variable. I think across many other demographic variables you also see market disparities. I think models that are emerging that can address this would be the telehealth models where, excuse me, far more people can access genetic counselors in ways that are convenient to them and continuing with the theme of today's talk on their smartphone. So I think that that's something to consider too as a part of this whole prediction. One of the biggest barriers right now to telehealth services and clinical services in general is reimbursement and genetic counselors having financially sustainable models of clinical practice. And here I'm showing data from a Stanford master's student I worked with, Eric Chaku, who's now at Geisinger where he studied across the country. Next slide, please, I forgot my slide prompt. Here's the data and here's Eric on the slide. He looked at billing and reimbursement. So he took data from a survey that is conducted annually by NSGC of genetic counselors, the NSGC professional status survey where they ask a number of questions about professional practice for genetic counselors across the country. Among them questions about billing and reimbursement practices. And he asked folks just about whether or not they were billing or the survey, sorry, the PSS asks folks whether or not they're billing for their services. And he took this into a state level average sort of percent of genetic counselors who are billing for their services. And on the graph, the heat map is heat map from 0% to 100% the percent of genetic counselors who report that they are billing for their services across the country. And there, again, there is really wide variability state by state as to whether genetic counselors are billing and getting reimbursed for their services that they provide for their clinical services. And he dug a little bit into the why of the reimbursement. What were the predictors by state of reimbursement? He looked at licensure, like just having licensure in and of itself was not a predictor, although I think there's other variables we could look at there, like how long the state had licensure that we didn't look at. He looked at payer policies, which is do payers pay for genetic counseling in those states or not? And in those, we didn't detect any relationships there, but where there was a really strong relationship was among the type of genetic counseling that was provided. And we found that prenatal genetic counselors were much more likely to report that they were billing for their services, followed by cancer counselors and followed by pediatric and general genetics counselors. And I think one of the hypotheses that came out of this is that going back again to the theme of regulatory policy is that because cancer counselors and pediatric genetic counselors tend to see patient populations that are more likely to be publicly funded in particular in cancer, the Medicare population is much higher. I think many folks, because they can't bill for a good majority of their patients are not billing for their services. And so we go to the next slide. Solution, what we put on the table for this would be the access to genetic counselor services act. So I just put in a plug here for a bill that has now been introduced into both the house and the Senate, which would add genetic counselors as providers under Medicare and allow them to bill Medicare beneficiaries for their services. Genetic counseling services are actually already a covered benefit under Medicare. It's just not that genetic counselors are not recognized yet as providers. So this bill would do that. It's gaining significant momentum. I'm pleased to report with bipartisan support. 23 co-sponsors on the House side or on the Senate side and over 300 different organizations, provider organizations, advocacy organizations, hospital organizations supporting the bill. So I think the passage of this bill will be a big step towards promoting access to genetic counseling. Okay, now I'm gonna move on to the next area of challenges. We go forward a couple slides. Talk about digital infrastructure. So when we think about the infrastructure behind genetic counseling, I think there's a number of pieces that really need to come into place. And here, as well, I start with a figure from the commentary in the American Journal of Medical Genetics that we published earlier this year. When we really laid out the idea that ordering a genetic test in today's day and age can be thought of as a value chain. Value chain is a term that gets used a lot in manufacturing to describe all the parts and processes that need to happen together in order to get something of value at the end. And within genetic testing, there are many different stakeholders who need to be collaborating together in order to really get value from the genetic tests in the end. And there's a number of places where this value chain is falling down. And it really starts with patient identification. We're missing a lot of patients who could benefit from genetic testing and we're requiring a lot of places, people in a lot of places to have to self-identify in order to get access to services. We are now developing models where we could be identifying people in a much more automated way who could benefit from either genetic testing or as we move towards the future particular genetic analyses that would ultimately help their healthcare. Right now, we still live in a day where very few genetic tests are actually ordered within a physician's workflow within the EHR. In most cases, they're going out to outside portals. In some cases, they're still using fax machines to order genetic tests. I think many people in this audience may not have heard of a fax machine. It's a very outdated technology that sends a piece of paper digitally to another place but that's still used in many areas of healthcare. As long as I have been in genetics, we've been talking about the need for clinical decision support. And I think for the last 20 years, we really haven't been able to get past the pilot stage of clinical decision support. And part of that is because the market is turning over so quickly that by the time we build any new genetic sort of test decision support model, that genetic test may not even be in vogue anymore or be used. And so we need to build more scalable, more sustainable sort of systems-based decision support approaches that can account for many different genetic testing approaches. The majority of genetic tests today are performed in laboratories outside of the hospital system or the healthcare environment where the patient is seen. So mostly they get sent out to a lab that may or may not have access to the patient's clinical data. And so this is an area where we need better data flow. The people who are interpreting genomic information need to know the clinical context of the patient and have sufficient information. And we need to have that information flowing more freely. Results still generally come back as PDFs into the system. There are very few emerging, there are more and more, but very few hospital systems that are managing structured data results around genetic testing. And even those that are are rarely doing it for all of their genetic tests. And I think this is a tremendous opportunity. If we're really gonna get to a point where we can integrate multiple data streams, then having structured data that can be integrated is gonna be key. And then having structured data that actually plugs into the what to do next, how to manage the care. I think many people, in particular people who've had predictive testing or predisposition testing, and even those who learn that they have very, maybe a very high risk for a particular condition, a condition like hypertrophic cardiomyopathy or hereditary cancer syndrome are left on their own to manage their own high risk care. There aren't many integrated cohesive models that support people along that path and keep the guidelines up to date. There are some specialty centers they may go to, but across the country, I don't think we have many options for most people. And then finally to the billing and reimbursement process, which is honestly where I have spent the most time. And if anybody wants to talk to me after for a much longer conversation, you can grab me for that. I'm not gonna get too deep into CPT codes today, but the coding and billing and policy situation for the coverage of genetic tests is really complex. And we live in a world where you can't, a patient can't just ask their doctor, is this test gonna be covered for me without getting a very complex answer, which usually routes them to the health plan, which usually sends them back to the lab, which sometimes sends them back to the provider before they go back to their health insurance company again, just to get an answer to a yes or no question and what should be a yes or no question in the future. So next slide, please. I think a lot of building this infrastructure, and this has certainly been the approach that we've taken at concert is about cataloging the range of things that can be ordered and building a system, a taxonomy around all of the tests that might be ordered. We've seen in the time that we've been cataloging the market, concert genetics maintains a database of all the tests sold across the country. We've seen tremendous growth in the range of things that could be tested for. Also in the number of labs who test for similar things. So right now we have a database of 165,000 genetic tests marketed in the United States that are available for clinicians to order for their patients. And that's been part of the challenge of integrating it into the EHR is how do you keep up with that kind of a growth? And so we need systems and tools and taxonomies to track these sort of order sets that a provider might order and be able to reference back to, to know what was ordered, what was found and what was ruled out at any given point. Next slide. I think because of some of the infrastructural problems we see really wide variability and utilization of genetic testing across the country. And what I'm showing you now is data from a white paper that we put together in collaboration with the Precision Medicine Coalition, the Blue Cross Blue Shield Association and Illumina looking at commercial claims data and looking at utilization of genetic tests across the country. And then you're seeing heat maps with utilization rates per million members plotted state by state across the country. I like to NIPT, exome and cancer genome profiling. And NIPT, we saw the scale ranges from zero to about 6,000 per million members. And just for reference, the birth rate in the United States is about 11 per thousand people. And so if you've got six per thousand people which would be NIPT, there are some states that are as high as about 50% of pregnancies getting an NIPT test right now. That would be on the really high end. Many, many states are far lower than that right now. In exome sequencing, we saw obviously for rare disease, far lower numbers, 97 per million members was the highest across the states that we saw. And when we extrapolated the nationwide utilization rate out to population data, we calculated that it would amount to about five or 6,000 tests per year exome testing performed and billed through payers. This is a key caveat here nationwide. And this really is less than even a conservative estimate of the frequency of rare and undiagnosed disease across the country. We also looked at cancer genome profiling. So tumor, NGS testing, which as with the other two varies really widely across the country in its utilization. If you go to the next slide, you'll see one of the other things we looked at and one of the questions we asked was whether it relates to coverage policies and how particular coverage policies from commercial payers may be influencing this data. We didn't actually see much of a relationship between the specific payer policies and states with higher or lower utilization. Although I think there are some local trends in some areas. We do see overall that payers are expanding their coverage of genetic testing. So there's a trend towards broader coverage. We developed a rubric and we pulled a bunch of commercial payer policies from the major commercial payers in every state. Each test had a slightly different rubric. So for NIPT, we looked at like all pregnancies versus just women over 35 and it came out somewhere in the middle wasn't really changing between 2018 and 2019. For exome, we looked at like multiple congenital anomalies versus broader indications for coverage. And we saw that it was becoming broader between 2018 and 2019. And then in cancer genome profiling, we looked at, did they cover it at all? Was it, if it was no coverage at all, it was a zero. Five was they cover it for lung cancer only and 10 was they cover it beyond that. And we definitely saw indications for coverage expanding over this time period. So I think that opens some doors, but I think we need, and it's not, none of these things are gonna happen due to one factor alone. I think they need to be moving together. I think coverage policies need to move along with a more mature reimbursement infrastructure. And I think to get back to the very specific question of how do we get to a point where a genome on smartphone is accessible to many people? I think we need very new reimbursement models to get there. And that will include reimbursing for interpretation services. So when the genome already exists, but you're gonna analyze a part of the genome, we're gonna need a way to code for that. And right now in the billing code, CPT code set, there really are, for most places and most types of tests, great codes for doing that. We're also gonna need reimbursement models that support this idea of data integration and data stream integration and interpretation of multiple pieces of data coming in together to support reimbursement. And perhaps we even need things like subscription models, be it through a health plan or some sort of modernized version of a health insurance company that allows you to subscribe to ongoing diagnostic feeds with the idea that perhaps that could, as I said earlier, actually keep you out of the healthcare system, which I think is part of the goal. Okay, so now onto the third part to circle all the way back around to our discussion earlier about regulatory policy. Next slide, please. And then onto the next slide. I need to be a little bit cautious because I am definitely not an expert in policy, an expert in privacy, an expert in data security, nor am I a bioethicist. But I love these fields dearly. And just to put this one slide in here in preparation for this talk, I spent a lot of time reading. So I don't have a lot more content than this, but it just felt like such a critical part of this discussion to me was to include a space around how regulatory policy could really enable this future, this 2030 future. And I think in order to create the learning systems, we need to really rapidly accelerate our understanding of how to use genomic information to make healthcare better. We need to get away from the current models of data ownership and commoditization of data and turn towards one of partnership and permissioning of data. And I think regulatory policy can play a role in really setting up the right incentive structures to enable this future, to enable more ethical and equitable practices. And I think we're starting to see, as I mentioned earlier, policies roll out that really are pushing us in this direction. And there remains much more room to build more. I also think we're seeing some commercial data ownership models with companies like Luna DNA and Nebula Genomics that we can learn a lot from as they develop their models of engaging people, putting them in a position with more power to control when and how and where their data was shared. And I really liked in particular, this paper from Ellen Clayton and colleagues at Vanderbilt, this one line that we really need to start asking very nuanced questions and even in the consent process, be having more nuanced ongoing conversations about how data will be used, under what conditions and what are the trade-offs and helping folks to understand when they're doing it for their own personal benefit and healthcare when they're doing it out of an altruistic desire to drive research forward and helping make those things very clear to people. So I do think one of the most promising ideas that's really bundled into this bold prediction is the idea that smartphones could be a vehicle for control and transparency. And we talked about that at the break. How can we build tools that really enable that? And maybe in the end it doesn't matter so much whether your genome is on your smartphone, but it might matter far more whether there's a consent button on your smartphone and then the button that you would use to permit sharing of your information. That's my last point I wanted to make. Other than, certainly not, next slide please, to thank everyone that I've had the pleasure of working with throughout the last few years with whom I think many of the thoughts in this talk were put together. My team at Concert, Eric Crippu at Geisinger, Deepti Babu who is the lead writer on the white paper I mentioned with PMC and the folks at Illumina and with the Blue Cross Push Up Association. So with that, I will stop talking and pass it back to Mike to inspire us with all that we can do. All right, thanks very much. You brought up so many really important issues there. So let's go ahead and talk a little bit more about this. Next slide please. So a few years ago, myself and a few of my colleagues we try to do this analysis. It's actually pretty tricky to get an exact number but I think this is sort of the right order of magnitude. But the analysis that we try to do is we looked around the world at where all the sequencing instruments have been installed and we just try to tally up how many instruments are out there, what capacity did it have and then we kind of also looked in the publication record to talk about what sort of analysis has been done. When we put together this projection about how much sequencing had been done to date and then looking forward, how much do we expect there to be in the coming years? So we published this in 2016 or 2015 but I think the model is basically still holds to this day. So at the time that we did it, there are several hundred thousand genomes that have been sequenced, maybe approaching a million. Now when we look around, we think there are easily millions of genomes that have been sequenced. I think, and this is again a hard number to get, I think the worldwide sequencing capacity is something like a hundred petabytes per year. So this is a very weird number. So your laptop maybe has a few hundred gigabytes, maybe a terabyte, so a thousand terabytes, makes one petabyte. We think there's maybe a hundred of those in the world every year, just constantly churning out data and churning out data. Amazing thing is the trajectory on is on one of these exponential growths curves, you may have heard of like Moore's Law, where it doubles pretty quickly. Moore's Law doubles every 18 months, Illumina themselves estimates it doubling every year but historically it doubles every seven or eight months or so. So there's maybe a hundred petabytes now and then next year 200 petabytes, then 400, then 800. Suddenly we're into the exabyte range by 2030 and there's just kind of an infinite appetite for just more and more and more sequencing. It's just truly astonishing what's going on there. Next slide. So that's exciting, but kind of really, at least the question, what are we gonna do with all this data? What sort of things are people interested to do? And here's kind of a few examples of some of the things that people are excited to do and I think they're just gonna grow more and more in time. Next slide. If you look at kind of the number one users of genome sequencing or genotyping today, it's actually around ancestry. And this, and earlier we were talking about how there's these like academic industry partnerships. I think this is a great example. So for eons now in sort of an academic setting, there's been various systems been proposed throughout the years. So if you get someone's genome information to actually look and sort of determine computationally the ancestry of the individual. So here's sort of a map on the left there where it's an analysis done of someone's genome using this machine learning technique called principal components analysis. And kind of just by first glance you get this immediate recognition that just by someone's genome we can pretty accurately place someone in terms of continental Europe. Now as these databases grow and methods become more sophisticated we get better and better resolution about ancestry. And if you look in sort of the commercial sphere that's exactly one of the key drivers. So I try to look this up. I don't know if this is totally true but I think it's the right word of magnitude. My heritage reports more than 50 million customers. Ancestry DNA reports more than 50 million customers. 23 and me reports more than 12 million customers. So in terms of like putting genomes into people's smartphones and in their hands for these millions and millions of customers people it already is. There's already applications that can do this. I think it starts with ancestry. I think everyone's kind of curious where they come from kind of their own personal history but you'll notice that these companies are becoming more and more involved in genetic association. So in the case of 23 and me in addition to detailed ancestry reports they're starting to really look at association. Some of these associations are kind of interesting kind of maybe a little silly things about I don't know your genetic contributions to things like how tall you are or your hair color or other things but it's increasingly becoming more and more focused on health considerations. So this can be perhaps recommendations for diet recommendations for what sort of pharmacogenomics sort of associations there might be and I think we're just at the very beginning. I think we're just at the very beginning. I think we're gonna see those types of services grow and grow in 2030 and far beyond. Next slide. So at the DNA level the obvious significance for ancestry, genetic associations but the thing to keep one thing to keep in mind is inside of cells. There's a lot of other sort of biological activity that could take place. On the left there we have this very famous plot looking at say gene expression inside of breast cancer tumors. And then from those genetic expression profiles we've really come to appreciate that breast cancer is not one disease, right? It's one organ where there's many, there's a few different subtypes to it that need to be treated with entirely different chemotherapies and one correct chemotherapies are applied. The health outcomes can really be improved. A famous example is in the case of her two amplified breast cancers, Herceptin is really a powerful drug to treat that but if it is not a her two amplified breast cancer it just brings with it a horrible mess of complications and side effects and very little benefit. So we see it started in academic research where we can see these readouts at a sort of a level of gene expression and now we're starting to see these be commercialized. So here I just have logos from a few companies that offer products that try to guide patients. I mean, these are really, really challenging times for patients when they're first detected with these cancers. And now we have to really empower them to present to them this really powerful data about what sort of health treatments they should get. Next slide. So we have kind of your own DNA, you have your own expression. Another important aspect to the future is gonna be looking at your interactions between your genome and the genomes of other species around us, right? So maybe think about it all the time but on your body, in your body, there's trillions and trillions of microbes that are growing and living there. In some ways they outnumber us even though obviously we're much bigger than them one at a time, the microbes inside of us actually outnumber us. In fact, if you took out the sort of microbial cells inside of your body and sort of added it up and you kind of weighed it, it would be approximately the same mass as your own human brain. And what we're coming to appreciate is these microbes play a really important aspect of human health. Certain compounds, certain drugs that you might be given, you really need the microbes to be present in order to sort of break them down. There's a lot of really interesting associations now between the microbes in your gut and your own sort of mental states. I think we're really at just the beginning now but in a nutshell, these microbes are pumping out neurotransmitters and they're interacting with our cells and they're interacting with our gene expression and it's really fascinating to think about them. And again, a few companies are leading the way to sort of mind those associations. And in the first phase, it's sort of diagnostic looking and seeing what's there. I think in time these are gonna become more and more prognostic. What sorts of microbes, what sort of probiotics should we be promoting inside our health? Next slide. And then finally, if you kind of think beyond sort of the healthy microbes and microbes in your environment, you think about an infectious disease study. I mean, here in COVID era, there's no greater example of the importance of being able to do worldwide monitoring than all the tracing that has gone into SARS-CoV-2. Here's a map I took last night using something called NextStrain which is this worldwide database and analysis suite that is mapping out all of the viral genomes that have been sequenced and deposited in the databases about 4 million viral genomes that have been sequenced and this builds up maps and trees of transmissions. Now, this is how we know about things like the Delta variant and how sort of how easy it is to spread and how deadly it can be. Incredibly important that we build these resources and keep them maintained so we can stay abreast of new mutations as they arise. Next slide. So there's lots of applications from the DNA inside of you, other molecular activity like expression or maybe other epigenetic marks, the DNA and RNA of other sorts of species all around us, but there's some really massive challenges in order to sort of really realize this vision over the next many years. Next slide. So the way that I think about this, this is actually an idea that I had with Adam Philippe, he's an NHGRI intermural investigator, also a great friend of mine, is that the way we think about genomics in the future is somewhat analogous to the way we think about, say maybe the weather forecasting that we do now. If you wanna know and wanna sort of anticipate what the weather is gonna be like in your own backyard tomorrow, a really powerful tool to being able to recognize that is to look at maybe a few hundred miles west or a few hundred miles south to see those storm fronts approaching. In a similar way, if we're gonna make sense of our kind of genomics future, especially in the case of say infectious disease, it's really, really important that we have ubiquitous sensing technologies all around the world. I mean, this already exists to some degree in the case of say influenza and now say COVID where when they're designing the next year's influenza vaccine, they look to see, well, what are the prominent strains and say Southeast Asia? That's often a very good indicator, although not perfect as to what's coming here in the United States. Next slide. There are several sort of technical challenges associated with this kind of at the top of the list. And in some ways the easiest, the one that we have the most control over is fast, accurate and ubiquitous sequencing. And now thanks to these technologies, we're really at the forefront of genomics. We're able to do things like completely sequence an entire human genome for the very first time. I think there was an earlier bold prediction by Karen Miga and Evan Eichler and others on discussing this earlier. And with these new technologies, we can actually as close to perfect as we can possibly imagine actually realize entire genomes. This sort of sensing technology is starting to move into other essays, other applications. I'm really exciting about the combination of genetics data with maybe gene expression, with healthcare data, imaging data, measurements of exposures, environmental. Also things like wearables are gonna be really important in the future as this continuous physician's test. It'd be very hard to notice that, oh, I'm only 10% off on my behavior, on my activity, on my pulse. It'd be really hard to recognize 10% changes as yourself, but a continuous physician's measurement, constantly remaking these recordings, a 10% change can be an incredibly important signal that there's something major going on. And in terms of making it ubiquitous, these trends of mineralization that I spoke about, they continue. On the right there is a tweet, a picture of a tweet from Clive Brown. He's the CTO of Oxford Manifor, the company that manufactures these devices, showing off something they call the smidge ion, which is a further miniaturized sequencing device that actually plugs into your iPhone, enabling that tricorder. So I think we're there, or at least we're on the right trajectory in terms of the technologies to be able to take these measurements at a global scale. Next slide. Next sort of major challenge is this is gonna be a global phenomenon. We wanna sort of support everyone around the world, especially a premier example of this is infectious disease. You can't just be isolated. You can't be siloed looking very narrowly. You have to look globally in order to anticipate where these challenges are gonna be. Now part of the solution will be smartphones and there's a lot you can do on them and they're incredibly powerful. What the reality is, is these data become more and more powerful when we can integrate them together and we can aggregate together across huge data sets. Today in human genetics, it's very common to have studies with tens of thousands, hundreds of thousands, millions of millions of genomes. And as powerful as these devices are, it's just not practical to do so anytime in the near future at that sort of scale. So I think a big part of the future is gonna be centered around cloud computing. Maybe it's your smartphone that accesses it, but it's cloud computing. And we're already actually very familiar with this. If you ever use, I don't know, Facebook or Instagram or Amazon or any of these sort of tech services, yes, you have a browser, maybe you have a smartphone, but you're really tacking into this enormous resource behind the scenes. They can do incredible analysis at scale. In genomics, like I mentioned in any Shuri, there's a big project called Anvil, which is to provide this cloud-scale platform that can bring a lot of tremendous computing power. And the beauty of it is it's available for anyone, anyone that has a smartphone or a laptop can tap into it. In the case of the telomere to telomere consortium, we're routinely using 10,000 computers together in order to do this analysis very, very quickly. So it provides equal access, it scales up, it scales down, but there's gonna be some challenges if I'm really honest with myself, right? There's never gonna be a single cloud, a single entry point that will do everything. Part of this is by design, there's different legal privacy considerations that restrict data to certain facilities. You know, it starts to look at international laws, say the laws for privacy in Europe are quite ahead of their time in terms of what needs to be disclosed and the safeguards that are in place, they're really, really proactive about preventing unauthorized usage. And I really applaud them for how that worked. Now, furthermore, different types of studies, inevitably are gonna require different types of analysis. Platform that you might use for infectious disease may or may not have what you need to do, say genetic medicine. So of course there's gonna be specialization to provide all of the appropriate tools for the appropriate questions. So the real need is, yes, these systems are being built, they're being built at global scales, but what we really need is continued focus on those interoperability technologies so that data, analysis, users, patients can move easily from platform to platform, aggregate the right data together so we can actually take advantage of them and make those health discoveries that we need so badly. And then number three, next slide, kind of the third major challenge that I'd like to call out here is we really, really need to focus on genomics equality for all. A lot of what's current research uses artificial intelligence, machine learning, statistics, they're incredibly important. They can tease out these really subtle relationships, these really subtle patterns that you would just be not able to really evaluate by eye, but they're only as good as the data that they're working with. They're only as good as the training data that are available. There's been several really high profile studies although not enough. They're like really calling out that if you look at a lot of the data that have been generated so far in the case of genetic data, it's really skewed in terms of the ancestry, the makeups of the people that are involved. On the case of something called the GWAS catalog, which is this big database of genetic associations, more than three quarters of the patients involved are of European ancestry. And with that, the results are gonna be skewed and findings that you might find for one population may or may not carry over to other populations. So these life or death genomics decisions just may not generalize and we may not know that until it's too late. The way we're gonna address this, I think, and this is gonna be an ongoing problem for years, is we just need multiple interventions at several levels to improve this quality. At the data level, we need to fix the disparity in the sequencing that's been done. We need to broaden our outreach. We need to look at other communities, engage with them, make sure that we have the right data to work with this. At the software level, there are things you can do in terms of your training, in terms of your validation, that can sort of detect when certain biases might arise. We have to be really mindful of that. It has to be essential to how those software packages are developed. At the experimental design, I think we are challenged. I'm a genomicist, I love it, but I think we're also challenged to move beyond just near statistical associations and think about mechanism. That's the ultimate safeguard. If I have an association, that can be many different things, but if I understand the actual mechanism of how we go from a genetic variant into some sort of phenotype into some sort of disease state, wow, that gives us a lot of power to make sure our results are going to vary with us. And then finally, this is hard. We need training, we need education, we need trust, we need consent. We need to work together as this global community to make progress on all of these different challenges that are present here. So next slide, please, just to kind of sum up. I'm excited. 2030, maybe 2040, my prediction is someday, and I think it'll be in my lifetime. I'm gonna carry around a tricorder. I'm gonna carry around a professional DNA sequencer in my pocket that I'm gonna be using, I don't know if I'll use it every day, but I'll be using it all the time, especially with little kids at home. Wow, I would love it when they come down with a runny nose to tell them right away, is this something I need to be worried about or do they just need a good night's sleep? So I think this is coming. I think the technologies are coming into place. I think it's gonna be transformative in the same way that, I don't know, we all carry professional cameras in our pockets now that were unimaginable 20 or 30 years ago. And we all do it now and all the applications that we have now. I mean, you can shoot professional cinematography on an iPhone that exceeds anything you could have done in Hollywood just a few years ago. There's endless applications, people are just super interested in themselves. We wanna improve our health, we wanna improve our lives, we wanna study the world around us. There's just an endless set of needs and desires and interests to do there. But it's hard, but it's hard. I saw the very first question is, is our ability to understand genomes really gonna keep up with the technology? Sadly, no. I think in a thousand years, 10,000 years forever, there's gonna be new and interesting results to be made. And that's because biology, life science, health is so dependent on your environment to your actual life. And we can make associations, we can make some predictions, but sometimes it really matters what your day-to-day interactions were. It really matters what your exposures were. It really matters what other microbes, viruses are around you, what you're eating. These things are really hard to predict. A lot of those associations only present themselves if the right set of features present themselves. And that's very hard to predict. Today, we have some clear-cut examples, but in a thousand years, I assume we're still gonna be working on this. It's very hard, it's out of reach. There's a lot of jargon, there's a lot of confusing legislation. I'm really, really excited. Jillian and others are really thinking deeply about genocostallers, especially for these critical diseases where life decisions need to be made. I want someone that I can call and ask for help. Because these things are beyond even me. And we really, really have to take immediate action to ensure equality for all. It's really important that everyone can participate and it's really, really important that this is done at a global scale because that is how genomics, that is how the world is. It's always at a genomic scale. So just to kind of wrap up here, on the next slide, I have lots and lots of people in my own group, really wonderful students, postdocs, analysts, great collaborators around Hopkins, great collaborators at Cold Spring Hour, Bellet Roads of Medicine, Data Farber, Tilly American Consortium, really appreciate all the funding and support from ASHA, NCI, the Mark Foundation, NSF, and Hopkins Unfunded through the Bloomberg Professors Program. I'd like to thank you. And I think we're going to open it up now for questions from the audience. Thank you, Mike. That was amazing. I did, I did know about the number of microbes that we have. I did not know that they were the same mass as their human brain. So that was something that will always stay with me forever. In addition to a lot of what you both have said. So thank you all. We still have about 15 minutes left for questions for Q&A. So please, if you've not already submitted a question in the Q&A part of Zoom, please do so. We have a lot of questions to get to. So we'll try to tackle themes. One question that I had just to start us out and I think this gets to something that you had both touched on because we have sort of limitations when it comes to both support for the human element, you know, with genetic counselors across the country and limitations in terms of access to healthcare and resources. But also we have all these quickly evolving technologies, especially artificial intelligence and machine learning. Are there ways that you see those things sort of dovetailing at some point to try to address needs in each of your areas in ways that we may not expect? Yeah, yeah, I think they have to, right? I think they have to. I mean, kind of the way I see it is both will be essential moving forward where we're gonna have to have great AI systems and we're also gonna have to have great counselors and others that can help with the interpretation, right? The scale of this is so enormous that we're just not gonna be able to manually look at every single example. There's just too much out there. There's just too many variants, too many potential associations. So let's use the software, let's use the automated methods to do the best that they possibly can. And then we get into those really tricky situations, getting those really complicated scenarios. That's where we can really focus the attention of the person to help out and make sense of it. I think a good analogy is like, a self-driving Tesla, it's not really self-driving, right? It's like a super autopilot that does all the easy stuff. On the highway, it'll engage the cruise control, but then as it gets into a tricky situation, it alerts the driver, oh, you need to take over, you need to bring us in, make these really critical decisions. I think that's the future. Let's keep the counselors where we really need them, but they're gonna be really essential in those tricky situations. Yeah, I would wholeheartedly agree with that. I think, and I think too, when I think about the future roles of genetic counselors, I think about the roles for genetic counselors who will be building the technologies to power chatbots. And this is something that's happening now to build triage models so that the people who need to be seen can be seen in models that are accessible, telehealth type models, or who knows what other models we may have in the future that may work as well. And the people who need a more straightforward answer to their question or sort of the simple stuff in the self-driving car model have tools that are readily accessible when they need them at their disposal. And I think the other thing I would comment, and I was thinking a lot as Mike was talking about this, is as much as I said, I am not really an academic scientist, I definitely spent a while there before moving over so that I have this history of being an academic scientist. And some days I think I miss the days that were more about sort of the ideas of the future and the power of the technology and where it can go. And it was really like a love of genomics that brought me into this. And then somehow like got myself a little jaded when I opened the Pandora's box of the US healthcare system and got really obsessed with that. And then, and started to become more of a cynic. Like, are we really gonna have sequencers in our home? Are we really gonna have a sequencer in our pocket? Like, what is my talking about here? But I think honestly, the last two years and the pandemic have changed my opinion on that dramatically and really like blown my mind in the ways that technologies have so rapidly iterated to move towards at home models of testing in community models of testing for COVID. The fact that like, you will pretty soon have a number of at home PCR, one shot PCR machines you can buy and bring home with you. Like, this is a future I really didn't imagine but it makes perfect sense to me now that that's something you will do and the model of your kid is sick in the morning and has a sniffle and can't go to school because they just failed a criteria checklist of reasons kids can't go to school these days and you have a machine to tell you how safe or dangerous they may be to send them out into the community. I think that that's really, really powerful and it feels like a much more realistic future today than it did to me two years ago. Yeah, that's a great lead into another question I had, Jillian which is, we sort of had this arbitrary 10 year time frame I like that Mike went into a thousand years in the future. How much do we, are we unable to control for sort of world changing events such as a pandemic that really accelerates technology or accelerates the gathering of knowledge? I mean, how much are these barriers dependent upon things like the amount of funding or the amount of time it just takes to do science or are there ways in which that some parts of these streams could be accelerated to get to this future of faster? Yeah, I mean, I can weigh in but I'm very curious to have Mike's answer too. I think, there are a number of technologies that were absolutely in development prior to the pandemic that went pretty rapidly into real time and real life. And then the other area that I've really been struck by that I think isn't getting as much attention is the human communication part of it. So for anyone out there who's done an at home COVID test to me, one of the most impressive parts of that whole package that box you buy are the instructions on how to do the test where incredible background, I'm sure people who are experts in health literacy developed that tools to help people understand how to do the test and how to interpret it when it came back taking multiple methods, multiple visual methods, written methods, these approaches that are very evidence-based that I think a lot of research has gone into over the last 20 years in health policy and applying them to build something that is scalable and sustainable, which I think ties back to your earlier question as well. Like how do we support people? But it's also about this question now of how do we build the component parts that are gonna be broadly applicable for any range of twists or turns that the world may take? Yeah, if I could comment, I think humans, people are just really clever, especially when it comes to engineering. And a great example, this is so-called Moore's Law, which is this sort of observation, this historical observation about, well, the specifics are about the number of transistors, but what it's really speaking to is for decades and decades, people have been able to, engineers have been able to make computers go faster and be more powerful. And that's been going on about 70 years now, something like that, 60 years. And every year people are concerned, oh, is this gonna be the last year of Moore's Law? Is this gonna be the last year of Moore's Law? And amazingly, we've never hit it, we've never hit it. I mean, the metrics have changed and the types of architectures have changed, but that drumpy keeps marching on. So I think in genomics, we're at this really, really special time where we have kind of a few major things all coming together all at the same time. Now one inevitably, of course, is the hardware that is just amazing. I can control individual molecules of DNA with a handheld device, it's just like incredible to me to think that that could happen. And that's been through improved engineering and just incredible also biophysics, being able to work with these molecules at such scales. Another powerful trend has just been computing. Just in general is such a place where the scale of the genome is that the computers that are available can actually make sense of it. I would argue that the human genome could not have actually been a sequence and assembled earlier than the year 2000 because the computers did not exist any earlier in that to actually do the assembly. So you had to have this sort of perfect storm of the computing power and the sequencing power all coming together at exactly the same moment. And then the third dimension, and I'm obviously heavily biased, is the rise of these AI systems that are incredibly powerful at recognizing these really subtle patterns and it's just sort of highlighting, again, in this sort of autopilot mode where they can just highlight to a person, to a researcher, a clinician, to a patient, oh, maybe there's something going on here that you need to pay attention to. And those systems are incredibly powerful and there's this sort of virtuous cycle is now that we have them, they're doing lots of things. That inspires others to do more innovation. There's a lot of sort of public support. And also if I'm cynical and I look at, I don't know, the Googles and the Microsofts of the world, they also see dollar signs that there's a lot of customers which creates more virtuous cycles that there's investments back into these technologies. And wow, we're just at this incredible moment in time. So I'm optimistic. I think we're just at the beginning of this. And I think there's a lot of headroom to improve on them in the future. I wanna respond to that too. I think there's also a trend on the vein of cynicism but that idea does worry me and that's the overall cost of health care in the United States and the fact that our health care system is like driving us off of a fiscal cliff. And so we do think like there's room within the story for optimism. You cite Moore's law and you look and see what we did with technology in the way we drove the cost of sequencing, the cost of goods sold for sequencing down. I would say a lot of those cost savings have not yet been passed on to the rates. Sometimes they get charged for whole genome, whole exome sequences to payers right now. And so there's definitely room for those costs savings in the health care system. But I think it's really critical that we think about genetic and I mentioned this in one of my slides. It's really critical we think about genetic technology as a way to keep people out of the health care system and not just drive them towards more healthcare. Because I think there's a real difference in those approaches and in those mindsets. And I think it's really critical that we be thinking about ways to do what we did with the cost of genome sequencing in other parts of the healthcare system to make costs cheaper, to make drug development cheaper and better using computational methods, using better methodologies to take costs out of the system. We only have about five minutes left. This is gone by so quickly and we continue to get a lot of really thoughtful questions. I think one that I'm seeing a sort of a trend toward is if there is sort of one barrier that you could wish away or overcome. I know we've heard about a lot of barriers today. Which one would you identify as sort of the highest priority in a wishful way? Or if it is something that which should be a priority for the community to really consider. I think we're all very aware that NHGRI is putting on this seminar series in part to hear from the community to see, to listen to all the nuances that are inherent in all of these productions. And what needs to be done to make them come true or to make them come true in a way that is the most ethical, the most socially responsible way to do so. So, Jillian, do you want to go first? So, very broadly speaking, I would say just access to care. If you pushed me to be more specific, I would say racism and systemic bias in the healthcare system, especially if you're giving me the power to wish something away. That would be top of my list, hands down. I think that differential care that folks get, particularly around access to genetics and genetic medicine right now is something I would really love to see fixed. And so that's where, to me, the real allure of this bold prediction is the idea of accessibility and the idea that cell phones are something that many, many people around the world have. And so connecting something that many, many people around the world have access to, not all, so we got a third of the world that doesn't, but still two thirds of the world is a tremendous lot of people. And tying that to genetics feels particularly bold and particularly aspirational. The single thing I'm most concerned about is overestimating our confidence in the genetic results and the genetic associations. In the sort of early 2000s, when the first time ever was like practical, not to do full genome sequencing, but to do genotyping over like hundreds of people. You know, there was sort of study after study where they found this genetic association and it turned out not to really be true. It was just like a spurious result. So I really worry as we move into larger and larger cohorts and just moving to, you know, in the race to commercialize, in the race to have the best system, there's gonna be this desire consciously or unconsciously to sort of have a lot of confidence in our results. And the reality is that confidence comes, but it has to be done carefully, right? You have to have the right controls, you have to have right validation. Again, that's why I brought up mechanism. You know, to me, that is really, really important that we can really demonstrate, yes, that these associations are true. So I think there's enormous potential, enormous progress to be made, but I just wanna be careful about how it's executed so that we don't sort of trick ourselves into seeing associations that are not real. All right, thank you both. So we have three more minutes left. So I think that might be a good time, just if there's anything else that you either, one of you wanted to add that you didn't get to or one of your points that you really wanted the audience to remember. You know, if there's one thing that they can take away from this, what is it that you would like them to take away? Well, just at this incredibly special moment in time, and I just like to thank, you know, you, Sarah, and N.A. Schreif organizing the series. I think it'd be really fun to come back in 2030 and see how much progress has been made. I had the exact same thought. I said, are you gonna let us come back in 10 years and play this for us and shame us for anything. We totally were way off on or missed and applaud us for anything we got right. But I really do feel so honored to have been a part of this today. It's been really, really special discussion and really excited, Mike, to speak alongside you. And thank you very much for inviting me. Well, we'll put it on the calendar 10 years from now, October 4th, 2031. You guys, you're gonna come back and we'll say, gosh, you know, I have my cell phone here and it has all of the world's genomes completely on it. And then we can go from there. Well, thank you both. It's been such an honor and a pleasure to talk to you both and learn about your research. And we're so thrilled to have you part of this series. And I know it'll be part of our ongoing conversation. And thank you all for joining us. If you have more questions, please send them in.