 All right. Well, thank you very much. I'm going to talk a little bit about the baby-seek project, but by way of some of the work we've done over the last 10 years, because I think that the aggregate message is really interesting. And the aggregate message I'd like to talk about today is the case for preventive genomics. So while I'll be thanking everyone, including all of the baby-seek investigators for the data that we will be showing, I won't hold them responsible for some of the opinions that I'm going to express that they may or may not fully agree with. I do want to share my disclosures thanks to the NIH for support of this and other projects and my advisory roles to several companies and a co-founder role for a technology-enabled telemedicine company, Genomemedical. So the context for my remarks, I think, is this notion that genomics, and particularly multi-omics, may be the tip of the spear toward the promise of prevention rather than treating disease. The idea that we could really revolutionize what we mean by health care in a way that prevents and keeps us healthy, prevents disease and keep us healthy. Now genomics is only one small part of this, as is very clear. The whole arenas of proteomics, metabolomics, transcriptomics, and so forth are going to add much big data complexity to this. But genomics is the portion of this that we have the most experience with. Some of that experience is helpful, but some of that experience, I believe, is sent us in the wrong direction. Some of that experience arises out of Huntington disease, being the first disease to which we use genetic testing. Some of that experience arises from the notion that we see a child with a phenotype, and we use a genetic test to make a diagnosis. And as we've heard, some of that experience arises out of a very specific, particular form of public health mandated screening, which we all agree has been successful in terms of newborn screening. What I'd like to suggest is a different kind of model for the way we think about sequencing healthy individuals for the prevention of future disease. I'm not sure what the name of it should be, but it's not quite diagnostic, because they don't have a phenotype. And it's also not public health screening, because we're not imposing it within a public health context. I guess the question to bring it down to sort of our everyday lives is, why isn't your own doctor today using preventive genomics in your daily care? And if you believe that's where we're going, as most of us would, I think, say, we're going there, what is it that's preventing us from getting there sooner rather than later? We're a strange specialty in genetics. We have created quite a lot of barriers for the implementation of our own technology. And some of these have arisen out of the elements I've been talking about. The notion that, for example, genetic information is somehow toxic, that it will cause dramatic distress, that patients and providers will misunderstand genomic information, that genomic information will never change behavior, nothing will change behaviors, or that they will change behaviors with severe iotrogenic consequences, and that there will be unwarranted costs that accrue from the multitudes of genetic information. Now, you've heard that there's lots of cute names for all these studies, and our studies are no different. Our program at Brigham Women's Hospital has helped lead a lot of studies, including reveal PgN, MedSeq, BabySeq, MillSeq, PeopleSeq, and PopSeq. We're going to talk mostly about BabySeq today, but I want to show you a few slides from MedSeq, the first randomized trial of whole genome sequencing in adults to lay the groundwork for that. In fact, I'll start a little bit farther back. Over 10 years ago, maybe 15 years ago now, we asked the question, would receiving genetic information for a severe, untreatable diagnosis like Alzheimer's disease cause unwarranted distress? And remarkably, the answer was no. People who learned that they were at increased probabilistic risk for Alzheimer's disease were glad they had the information on average, were not overly anxious or depressed, and were not made more anxious by the information, and they did do things with it. They tried exercise and vitamins. They changed their insurance purchase patterns, and they did things with their families that they would not otherwise have done. The whole arena of personal utility was born. We've also reduplicated that in the MedSeq project, where we showed that people learning the results of whole genome sequencing did not have increased anxiety or distress. And in fact, there was something we didn't quite expect. The people who received their genomic results were happier, more empowered, and more relieved than the people who wanted them, and were not able to get them because of the randomization. They'd say things like, I'm definitely trying to modify my exercise and eating habits. Now, we all know it's hard to change those things, and maybe they didn't, maybe they didn't. But they were trying. They said it was better to know this than not to, so that they could get their families involved, and they were often relieved when a comprehensive evaluation did not reveal a risk for a future disease. And then there's a question of whether patients and providers will misunderstand, act inappropriately, or fail to act appropriately. And our data from various studies before MedSeq really indicated over and over again that the vast majority of people who sought out some sort of elective testing, whether you're talking about direct-to-consumer testing or even whole genome sequencing, were not grossly misunderstanding what they were giving back, what they were getting back. Not only that, but they were not changing their prescription medications in contradistinction to what their physicians had ordered. And they were reporting that they were going to improve their diet and exercise. Again, I'm not naive about how hard that is and whether or not they succeeded, but they did report that they were choosing to attempt to. Mu and Kuri and others have sort of spoken to some of these results as a teachable moment where clinicians and providers can kind of reinforce different public health messages that we all know we should be following. Now when we would find rare diseases in MedSeq, we actually gave those directly back to primary care doctors and rated them with these colors, you can see green, orange and red to denote whether they made mistakes in communicating them or taking action. And indeed, very few mistakes were made with only a few hours of orientation. So this notion that you have to have an army of geneticists, you have to have an army of genetic counselors before you can unleash this information on the populace, just may not actually be true. And when you think about the medical benefits and costs, one of the things we did was we made the decision from the very beginning of both MedSeq and BabySeq to analyze all the genes that we possibly could that were scientifically valid, meaning they were scientifically well evidenced to show association with a particular condition. We didn't worry about actionability. And with some exceptions, we didn't worry about age of onset. We just said, is the association real? When we did that, we were examining between 3,000 and 6,000 genes. Some of them, of course, for extremely rare conditions. And in MedSeq, we found that 21% of us, that's 20% of the people in this room, are walking around with a Mendelian disease variant, either a monogenic dominant condition, also function mutation, or a biallelic recessive. That's incredible. I almost can't believe it, except that we validated this as you'll see in BabySeq and other studies. 92% of people carrying a recessive carrier variant, and 80% of people carrying an atypical response to at least one set of pharmacogenomic medications. And when we circled back, and this is very relevant to what we also saw in BabySeq, when we circled back, we could see that in some of these unanticipated monogenic disease variants, there were fragments of the syndrome that had never been recognized as genetic. We even did polygenic risk scores and found that the doctors were more likely to order things for polygenic risks than they were for monogenic risks. So how does this all relate to BabySeq, the reason that we're here today? Well, Francis Collins, back in 2012, said whether you like it or not, a complete sequencing of newborns is not far away. And indeed, as we've heard, consumer-facing companies, and just the natural curiosity of information-seeking parents as they encounter new technologies that may provide health benefits to their children, is leading to the notion that if sequencing reveals health risks at any point in life, and that if that's good, then it's better to do it early. Because there's more time to capture the track that you can help a child or an adolescent before the conditions of those disease take place. Now as with the other in-site trials, we had to go through quite a gauntlet of our two IRBs and the FDA oversight of the BabySeq project. And I mentioned this not to feel sorry for us, but because there is a certain kind of bias that gets injected into clinical trials, when you have to over-elaborate the risks in an over-elaborated consent process, and I think that's really important to think about as we're trying to ask the question, how do these technologies play out with regular people in regular situations? You make people climb the mountain in order to get into a trial, you're only going to get the people in the trial who are capable of climbing mountains. And that played out in our recruitment issues for the BabySeq project. If you look at the light blue colors here, you'll see that the vast majority of people who declined participation had just had babies. They were a little concerned about fatigue and logistics and getting back for the follow-ups portion of the study. Or they just said, look, I just had a baby, I don't want to have anything to do with research. And that was kind of an interesting experiment in itself in terms of almost a population-based query as we walked from room to room to room on the newborn baby unit, saying, would you like to participate in the research study, and before we could get the next words of our mouth, they would say no, or would you like to participate in a research study about in which your child might get their entire DNA sequence tested? And we did get a chance to sit down and talk to them about the studies you can see from the lighter blue bars, I'm sorry, the darker blue bars. You can see that insurance discrimination and privacy issues in general weighed heavily on their minds. I think this is in part because those are legitimate concerns, but it's also in part this notion that if you lay on concerns within the consent process too heavily, you actually engender these concerns in people to a degree that may make it difficult, more difficult for you to enroll in important clinical trials. But here's the guts of the medical benefit from BabySeq. Out of 163 newborn sequins, we found again a surprisingly higher percentage of unanticipated monogenic findings, 11%. Not as high as the 20% in adults because we had cut the list down in deference to our IRB to include mostly child and adolescent onset conditions. So you would expect the percentage to go down. But still, would you ever imagine that 11% of babies would have a monogenic result, but they're putting them at risk for a monogenic disease? And when you dug down into that 11%, we found that several cases actually explained observations that were already been made but had not been associated with a genetic condition, and several actually were clues to babies with phenotypes that had never been diagnosed. So in particular, a child with a partial biotinidase deficiency was discovered and treated, and a child with a supervalvular aortic stenosis which had never been suspected was diagnosed through this DNA-first entry point. We also found a number of cases where the discovery of the mutation actually prompted discovery of an enriched family history. Family history has been held up as the gateway to genetic testing, but family history is notoriously imperfect. People are poor historians, families are small, people forget. Just the reminder, oh, you've got a cardiac risk variant here. Are you sure you told us everything about heart disease in your family? Gave us an opportunity to get an enriched family history from folks. As we saw before, there was a tremendous, tremendous number of carrier findings. Babies had, 88% of these babies were heterozygous for one or more recessive alleles, an average of two per infant range of zero to seven. And the anecdotes that came out of this were quite interesting. Many people saw and learned about these variants in their children and realized that they, its parents, had been rolling the dice because the only preconception testing that they had been exposed to was not really preconception at all. It was during pregnancy, by their OBGYN, if anything, and it was for a few variants or at most if they were Ashkenazi, for a few more. In fact, we had one family who decided to get both parents tested on the basis of a single heterozygous variant in the baby for a very rare genetic condition found by crazy coincidence that both were carriers, that there was nothing in the baby's genome that actually suggested that. They just went looking and then used reproductive technology to avoid having a child with that condition all through the exposure of the recessive carrier trait in their infant. And the behavioral conditions were presented earlier in this meeting, but I'll just go over them briefly for the audience that hasn't seen these yet. This is where the randomized clinical trial format for baby-seek really came in handy because as we saw from the recruitment diagram, the people who jump into this study are not characteristic of people in general. So the control group is really important when you're trying to ask questions about behavior. And in our randomized controlled trial, there was no difference in perception of child vulnerability by randomization arm, no difference in perception of child vulnerability even among those with the monogenic disease risk. There was no evidence of parent-child bond disruption by randomization arm and no evidence of parent-child bond disruption even among those with a monogenic disease risk finding. Now costs are important and we're fortunate to have Dr. Kirk Christensen, a young faculty member who is developing international expertise in this sort of new field of econogenomics. And his analysis of the baby-seek project suggested that while there was slightly higher costs associated with one of these monogenic findings, it was really not as high as you might have imagined. This is only a short period of time, we're not sort of extrapolating this out over a lifetime and that's a fair question. But it does suggest that quite a lot of the concerns about exorbitant explosion of downstream healthcare costs might be overblown. And even if there are increased costs after using a kind of broad, comprehensive, elective screening of infants, the real question is not whether it costs us more money, it's whether what we're able to find and what we're able to prevent in those children and subsequently those adults is worth the money that we would spend to find it. And you know, he's been able to do some projections, decision modeling of lifetime benefits and costs and all you really have to do is move historically from $10 per gene to $7.5 per gene to $5 per gene down to two and a half per gene or even lower to see that you do eventually cross this threshold, which is a generally accepted value, good value for what is considered to be life years saved in terms of children or adults. Now there are a lot of assumptions in models like these that many people can take issue with, but the point is if we all believe that sequencing and its interpretation is getting cheaper and better, there will be a point at which these thresholds are crossed. But it's really the individual stories that emerge out of this multi-domain kind of testing that grab you and make you wonder, make you really uneasy with a situation where people aren't at least having the option to do this. This little girl is the girl I mentioned with the partial biotinidase deficiency. And what we know of this biotinidase deficiency is that her enzyme levels were less than 50%, more like 30 to 40% of normal. We don't know that she would have ever had a full-blown syndrome, but we could imagine that she would have some sort of reduced IQ. And now she has a simple vitamin supplement that could prevent that. In another case reported by Ingrid Holm, we do look for a limited number of adult onset conditions in the baby seat project. We found a child with a BRCA2 mutation, and that child is way down here in this corner. And what was striking about this was that prior to the discovery of that mutation, we had taken a three-generation genetic counselor family history and not seen a single one of these cases of pancreatic cancer, ovarian cancer, breast cancer, or ovarian cancer. But once we discovered this and changed our protocol to allow us to return it to the mother, she was able to go explore her family history, realize that these cases were there, and then take her own precautions for creating a prophylaxis plan for herself. And there are unexpected consequences and unexpected benefits that are discoverable in the genome. A young investigator named Bill Lane has worked with us in MedSeq, BabySeq, and MillSeq, in which he has discovered and created an entire algorithm for predicting the 300 red blood cell antigens and 33 platelet antigens that differ between individuals and that contribute to serious and sometimes fatal transfusion reactions. And so he's developed an algorithm and published this now in a variety of places that can do this rapidly and accurately from whole genome sequencing alone. So to conclude, I think there's much more aggregate value in the genome than we've been thinking about, in part because we balkanize our medical care into OB-GYN, into pediatrics, into adult onset, into cardiology and so forth. But I think if you try to bring those elements close together, you realize that the potential for reproductive health alone is extraordinary. Why when we have the historical examples of Tay-Sachs disease in the Jewish community, when we have the historical example of different blood diseases in the Mediterranean countries that have been dramatically improved by screening programs and preconception testing, why are we not moving more rapidly to investigate preconception testing to our entire population of people who are going to have children? We've talked about somewhere between 10 and 20% of people having a Mendelian risk variant. Now we talk about this in terms of screening and the language quickly goes into prior predictive value, prior probabilities and predictive value positives, but I don't think that's the right analogy. I don't think that we should consider this as a screening test for a diagnosis at a point in time. I think we should think of this as some sort of risk stratification that moves people into different layers of surveillance, just like we would consider a family history, just like we would consider a physical complaint. In fact, you could almost think of examining your DNA as almost like part of the physical exam rather than a laboratory test, and it makes a better analogy for I think the ways that we're going to end up using it. About 80% of people have an atypical pharmacogenomic reaction, and with all of the recent publications about polygenic risk scores, even if you just took the top 2% of people as some of the recent RFAs have encouraged us, and you multiplied it by multiple conditions and found only the top 2% of people who are at highest risk, for many of these conditions you're in a category where it's three or four or five times the risk of the general population, and anything that gets us to three or four or five times the risk of type 2 diabetes or heart disease or multiple sclerosis or atrial fibrillation, certainly we should be paying attention to and trying to prevent. There are blockers within our current healthcare system that are kind of working together to make this difficult. High prices for genetic testing have historically restricted physician ordering behavior, therefore there haven't been that many, particularly in healthy individuals, that have allowed us to collect evidence to create practice guidelines without practice guidelines. It's very hard to get insurers to cover, and so this cycle goes round and round. But I do believe we're at a tipping point, and it's partly simply because decreasing costs for both sequencing and interpretation, and of course for the better quality interpretation that all of our speakers today are contributing to. And the more and more people who can start to afford this, even in a self-payway, creates pathways for physicians to order sequencing, even for healthy individuals, which then allows for the collection of evidence that could contribute to practice guidelines. And we're already seeing that, Amy, with ACMG committees and so forth, which can start to prompt insurer coverage, we've got lots of interest now from insurers, but also almost as a sort of orthogonal direction, we have more and more employers who are bypassing the whole medical care system with insurer coverage in order to provide genetic benefits to their employees. And as they do this, and they are, of course, the ultimate customer for the insurers, this is going to move the insurer industry to include these policies. So finally, I'd just like to conclude with this slide where I remind us that I think our data as well as data from many other directions and many other investigators suggest that what we have always seen as genetic disease is very, very much a small percentage of the true genetic disease burden that is contributing to disease and morbidity. And the genetic information, though, characterized for decades as somehow rare, even precious, highly complex, is far more relevant to public health than was previously understood. And I don't think, and this isn't really my opinion, but I don't think diagnosis or screening is the correct context for risk identification and preventive genomics. I think we actually have to invent a different language to talk about this. And partly that's because the more genes you analyze, the more risks you discover. Partly it's because actionability, despite the heroic efforts of the groups we just heard from, is still often somewhat subjective in that what one clinician believes is actionable, maybe different from what another clinician believes, and is definitely different from what different non-medical professionals believe is useful to them. And indeed, this whole concept of actionability, which is sort of top-down, expert-driven notion, is somewhat in tension with evolving concepts of patient autonomy, because we see over and over and over again that patients and customers want as much genomic information as we can give them. It may be that our only responsible course is to give them whatever information they want and just make sure that it's accurate. And finally, we need a new vocabulary and a new way to examine benefit and cost effectiveness, because both the benefits and costs aggregate over multiple domains, which do not often work together, multiple generations. All these tests are the gateway to family cascade testing, and multiple decades, because it's not so much that a cardiomyopathy mutation is a false positive. It's that it's a risk factor for the next 30 years that you may develop cardiomyopathy. That's different than being a false positive. So, as I say, the opinions are mine, but the data is a team effort, and I want to thank the rest of the BabySeq team here, this extensive BabySeq team, as well as the teams that have contributed to our other research. We are starting one of the first preventive genomics clinics in the country, where individuals can come in under an appropriate medical supervision, have comprehensive whole genome sequencing with in-depth interpretation of all of the scientifically relevant and medically relevant variants they want without any kind of limitation due to our perceptions of actionability. And looking forward, we do have an NIH grant to study the outcomes in persons who elect this, take this kind of elective sequencing, and I'll look forward to sharing that with you in future sessions. Thank you. Thank you, Robert. And that concludes the first portion of our public session today. So, folks can join us back here again in half an hour at 3.30 Eastern, and we will start up again with one more presentation from the Insight programs, and then we'll move into our panel discussion. So, thanks, everyone.