Good morning, and welcome to 23andMe's R&D Day webcast. I'm Wade Walke, Vice President of Investor Relations at 23andMe. Thank you for joining us today. For your information, this webcast is being recorded. Today's presentation will last approximately two hours, and at the end of the presentation, there will be a Q&A session. If you would like to submit a question at today's presentation, please use the Q&A feature in the platform. Before we begin, I'd like you to know that you can download a copy of today's presentation from the handout section of the webcast platform or from our investor site at investors.23andme.com. A replay of today's webcast will also be available on our website for a limited time within 24 hours after the event. Please note that certain statements made during this webcast regarding matters that are not historical facts, including but not limited to management's outlook or predictions for future periods, are forward-looking statements. These statements are based solely on information that is now available to us. We encourage you to review the section entitled Forward-Looking Statements in our press release and in our SEC filings, which can be found on our website and at the SEC's website, for a discussion of numerous factors that may impact our future performance. Joining us today are Anne Wojcicki, our Chief Executive Officer and Co-founder, Kenneth Hillan, our Head of Therapeutics, Joe Arron, our Chief Scientific Officer, Adam Auton, Vice President of Human Genetics, Jennifer Low, Head of Therapeutics Development, Adrian Jubb, Senior Clinical Development Fellow of Therapeutics, Paul Johnson, Vice President and General Manager of our Consumer Business, Davis Liu, our Chief Clinical Officer, and Steve Schoch, our Chief Financial Officer, who will be joining us for Q&A. Now I'd like to turn the call over to Anne. Thank you, Wade. What a pleasure to be introducing all of these, my fabulous team today. You'll hear a lot from me and Wade and our CFO, Steve. Today I think is gonna be fabulous because you will get to meet a lot of the other people within the company who are really making the exciting, you know, progress on therapeutics as well as on the consumer side. With that, let's go to the next slide. Two important announcements that we put out this morning that we will also be able to go through more. First, GSK has decided to extend the exclusivity period for a fifth year. That will mean a $50 million payment for 23andMe. I'm happy to be announcing that. Second is that 23andMe is electing our royalty option on the collaboration program, CD96. Jennifer Low will also be talking more about that. That will enable us to have low double-digit royalties if the program is successful. We are very enthused, as always, about that program. Next slide. I want to just go through a couple highlights that we've had in this year. Fiscal year for us ends in March 31. We're a couple of months away from the end, but it has been an incredibly productive year. First and foremost, I just want to highlight our wholly owned immuno-oncology program that is now in phase I that has, you know, recently just started. We recently acquired Lemonaid Health to really get us into primary care and enable our almost 12 million customers to get access to genetic-based primary care, so taking us into a whole new area. We recently received the FDA clearance for the hereditary prostate cancer. FDA reports are still a huge priority for this company, and making sure that we're making meaningful ones that are going to have broad applicability for our customers is important. We did also release 14 new health predisposition reports. Again, making sure our customers are engaged and continuing to mine our dataset, make sure that our customers are benefiting from all the data that we are putting together is also important. We've had a number of important genetic research findings. Again, the heart and soul of 23andMe is our research ability and about giving that back to the world at large as well as also for our customers. That included research progress in areas like COVID-19, Parkinson's disease, and depression. Last always is Ancestry is a very important feature for our customers. I am thrilled that we're continuously making progress in that area and being able to give customers deeper and deeper resolution into their ancestral backgrounds. Next slide. Like I mentioned before, the heart and soul of the company really is about the data. The reason why it's about the data is because that is how we are going to help our customers really benefit from the human genome. By having a huge amount of data, of genetic data, as well as self-reported data, we're able to make these discoveries. When you compare us to everyone else that is out there, what is the huge differentiator with 23andMe is the size of our dataset and the ability for us to recontact our customers and continuously you know, make more discoveries from the community. Like I said, our mission is to help people access, understand, and benefit from the human genome, and the way that we do that is by helping our customers continuously learn and make new discoveries from the dataset. Next slide. The way the company works is this flywheel, and it has been this way since we started the company. We always had this idea that how can we actually empower our customers to learn about genetics themselves and then participate in research, knowing that the more they participate and the more that the dataset grows, the more everyone is going to benefit. We now have almost 12 million genotype customers. Over 80% of them opt into research. We ask them then to take questions or surveys about themselves. We have a very comprehensive intake survey. We collect, you know, thousands of points a day, thousands of data points a day from our customers, and we have now over 4 billion phenotypic data points on our customers. We combine all of this information then to be able to get insights, and we turn those insights either into novel drug discovery programs or into novel consumer products. The idea, again, is that all of these outcomes from our research are going to go back to our customers and encourage them to be participating even more because the more they participate, the more they are actually learning about themselves and benefiting from the human genome. Next slide. I wanna highlight a little bit about what is coming and some of the things that you're gonna hear about today. First, we have these next generation reports that are the polygenic risk scores, but they're actually gonna start to include additional lifestyle factors that are going to improve your risk estimate. You're gonna hear about that more from Jeff Benton and from the consumer team. I'm super excited because we see that with our customers, is that they're collecting a lot of information about themselves, and they want to understand more and more how all these data points are factoring into a risk score. Second is genetics-based primary care. Like I said, I highlighted, you know, as one of our, you know, one of my highlights from fiscal year 2022 is the fact that we acquired Lemonaid. Lemonaid empowers us to have a phenomenal healthcare platform and also has pharmacy. We now have this ability to really deliver personalized, prevention-oriented, genetics-based healthcare at scale. One of the things that I've noticed over, again, the last 15 years that we've been running the company, is that genetics has been adopted in oncology and in reproductive care but has not been adopted in primary care. This is the wide-open field that I believe that we can play in, where it's not just primary care and it's not just telemedicine, it really is a unique offering where we're going to help our customers truly benefit from the genetic information we are providing and integrate that into their primary care. Last, obviously, one of our highlights for today is talking more and more about the therapeutics pipeline. I'm thrilled about this because, again, as everybody here knows, biotech moves can move slowly. It has been an exciting desire since we started this company to have an impact on therapeutic discoveries. When we started this program originally, we started the therapeutics team originally in 2015, with the hope that we were going to really successfully be able to go through the database and move programs into the clinic. It is a huge thrill that we actually now have two programs that are in phase I, and we have an incredibly robust pipeline. I will hand over now to Kenneth to talk more about therapeutics and where we are going. Thank you, Kenneth. Thank you, Anne. I'm excited to have this opportunity to talk about therapeutics at 23andMe, and also to introduce you to some of the members of the leadership team. Particularly excited to talk about why we believe our database platform and our genetics-based approach really provides us with a strategic advantage in today's pharmaceutical world. The reason it provides us with this advantage is because a core problem at the heart of our industry is that about 90% of the programs that you start will ultimately fail. Only 10% of the product candidates will actually make it across the FDA finish line. On average, it can take seven years to advance from kind of research idea discovery through to moving a program into clinical trials. We believe this is just an area where we can, by leveraging our data, make a big impact. We believe that the combination of both human genetics and then also, as Anne spoke about, this real-time consumer health research just can enable much more efficient and more productive drug discovery and also development. My team will be talking about that and telling you about that today. Next slide, please. At 23andMe, and Adam will talk about this, we are parallel processing existing and emerging data from our growing customer database and from the many hundreds of medically important medical conditions that we collect data on. What this does is, as Adam will explain, it provides us with a really with a roadmap or a playbook that allows you to start with the human genetic drivers of a given disease that you're interested in. By starting with all of the potential causal drivers in that disease, it can really shorten that drug discovery timeline. As one example of this, and Jennifer will be talking about the program, our anti-CD96 program advanced from research concept to the clinic within four years, which is significantly shorter than the industry average. Really importantly, though, particularly if you're managing a portfolio of programs, where most programs will ultimately not be successful, if you can find anything that will increase the probability of technical success, that can have a big impact on a portfolio. The biggest driver of improving ultimate outcomes for a drug is by starting with targets that are validated by human genetics. At 23andMe, by simply starting with all of our targets being validated by human genetics, we really believe that gives us a significant leg up. Since we're managing many programs, we've initiated more than 40 programs from the database. It's important that we also manage our investments very strategically. The way that we prioritize things at 23andMe is based on these terms, power, need, and speed. On the power side, we're really leveraging our unique statistical power to identify drivers of disease that other people cannot see. It gives us a competitive time advantage. We've sort of predefined the areas that we believe are the highest areas of unmet medical need where we can make an impact. Then we're particularly looking for opportunities where we can move programs through to proof of concept in the clinic with some reasonable speed, recognizing, of course, that, as Anne said, things don't move quickly all the time in therapeutics. Next slide. Adam Auton, our Vice President of Human Genetics, and Joe Arron, who is our new Chief Scientific Officer, will be walking you through our approach to DNA-based target discovery based on the genome-wide association studies in these diseases of high unmet need. They'll share how our data allows us to interrogate not just all the genes within one disease, but if we have a specific gene that we're interested in, we can look at that across literally 1,500 diseases, you know, with the click of a mouse. These are known as PheWAS studies or phenome-wide association studies, and both Joe and Adam, and actually Jennifer will touch on these. In addition to identifying potential new indications for a given gene target, it also allows us to pick up potential for early on-target toxicity, all before our scientists ever pick up a pipette, and that really helps us to drive this efficiency. Next slide. As I mentioned, we're conducting genetics research literally across hundreds of different diseases, and we're basing our analysis on genotype and phenotype data from both European and non-European populations. The size and scale of the database allows us to broadly mine for targets and diseases that have a greater than 0.1% prevalence in the population. Importantly, as the database grows, we expect that prevalence to drop, also as our computational methods improve, that we believe we'll be able to really dig into rarer diseases. The phenotypes highlighted on the left cover a range of common and rare diseases. You can see the size of the customer bases that we have there, and just highlighting a few of the rare diseases that we capture at the bottom where we have very significant numbers of cases, which enables us to make discoveries. On the right-hand side, we're speaking here about our 2020 COVID-19 study. This really speaks to the rapidity of our ability to do real-time research with this recontactable and highly participatory and engaged research community that we've developed at 23andMe. With that study, we were able to recruit 750,000 participants within 90 days. In total, more than 1.25 million of our customers participated in the COVID study, and it enabled us to identify important genetic risk factors for developing more severe infections and hospitalization. Next slide. We have about 100 scientists in therapeutics and all full capabilities for drug and antibody discovery, as well as the early phases of drug development. We have the major collaboration that Anne spoke about with GSK. It's been highly productive, and together, we've initiated more than 40 programs from the database. We've built what we believe is a very exciting pipeline across multiple therapeutic areas, as you'll see, and the two of the programs that we'll highlight today have now moved into phase I clinical development, including one, which is wholly owned by 23andMe. Jennifer Low, our Head of Development, will walk you through these programs shortly. Before that, I'd like to pass over and introduce you to Dr. Joe Arron, who joined us recently. He's our Chief Scientific Officer in therapeutics, and he'll walk you through our vision for target discovery. Thank you, Joe. Thanks, Kenneth. Thank you all for tuning in today. I'm super excited to be here. As both Anne and Kenneth alluded, you know, some of the things that really attracted me to join 23andMe late last year are the size and scale of our database and the proven ability to discover and develop a diversified pipeline of novel therapeutics. I'm really looking forward to helping to continue to grow this pipeline over the coming years. Next slide, please. A little bit about my background. I completed a combined MD-PhD program at Cornell Medical School and the Rockefeller University. My PhD was in Immunology. I then moved out west to do a postdoc at Stanford, and in 2006, I was recruited to Genentech, actually by Kenneth, among other people. I was one of the founding scientists in a group that was dedicated to biomarker discovery and translational research in inflammatory diseases. Over the next 15 years at Genentech, I gradually took on increasing responsibilities and ultimately was the Vice President and a Senior Fellow in Immunology research, where I led target discovery for inflammatory, fibrotic, and ophthalmic diseases, across over 20 different laboratories. Key to our strategy in that group was really developing both forward and reverse translational strategies across these disease areas. I worked on a whole bunch of different therapeutic programs from discovery through post-marketing and published a lot of papers. Really the experience there at Genentech taught me a lot about some of the key challenges in drug discovery and development. Next slide, please. Unfortunately, as Kenneth mentioned, most projects that are initiated in drug discovery and development do not succeed. We are very enthusiastic about the power of human genetics to increase the probability of success in drug development. I do wanna talk about a couple of reasons for failure that really can be addressed by our translational approach. The first reason is wrong target. You've selected a therapeutic target, but it turns out not to be a critical node in disease pathogenesis and/or there are safety issues associated with hitting that target so that you cannot achieve an adequate therapeutic index. Second reason is the wrong drug. Perhaps you've picked a good target, but the molecule that you're taking into patients does not engage that target in the right way. It may be insufficiently potent, or it may hit other targets, or you may not be able to deliver an adequate dose of the drug to the target tissue. The third reason is really moving kind of translationally and thinking about how do we actually design and interpret proof of concept clinical trials. Many times we may pick a target that is associated with the disease, but we don't understand how it manifests in terms of clinical presentation of that disease. We may be looking at the wrong clinical outcomes in a trial, or the clinical outcome measure that you've selected may not be relevant in the selected trial population. A fourth reason for failure is patient selection. It's become increasingly clear that in these complex chronic diseases, particularly those that 23andMe has very well represented in its database, there's significant heterogeneity, not just in terms of genetic drivers of disease, but in terms of pathological mechanisms and clinical presentation of the disease. If we can really understand meaningful subsets of patients all the way from target biology through to clinical presentation, then we can design better, more targeted clinical trials. It should go without saying that our rich database and our focus on translational research and really trying to understand the biology of human disease has the potential to mitigate these and increase the probability of success. The word success fell off the slide there, but trust me, it's the last word in that green box. Next slide, please. What do I mean by focusing on translational research? Ultimately, what we're trying to do is indicated out at the right side of this slide, is to deliver value to patients in terms of meaningful benefit so that they can feel, function, and survive better. Now, at the first level of abstraction, we try to demonstrate this meaningful benefit to patients by way of doing randomized clinical trials, and we measure certain clinical endpoints. There are a lot of steps, a lot of dots that need to be connected between the actual mechanistic biology of the target and these clinical endpoints, and that's sort of filtered through this black box of pathophysiology. We constantly need to be both working forward from bench to bedside and backwards from bedside back to bench, where we can identify indications and develop outcome measures that are most relevant to biologically compelling targets, but also really think about the clinical disease and develop targets that are most relevant to the evolving unmet needs in the indications that we're pursuing. When we think about this, we really need to understand the heterogeneity in a population, again, not just in terms of the genetics or the expression of a particular target, but how does that actually manifest in terms of disease pathology and clinical presentation. Next slide, please. Our research platform is really set up to yield novel drug targets based on insights from human genetics. Again, as Kenneth mentioned, we have many thousands of genome-wide association study hits that come out of our database, where we can see here is a locus that is statistically significantly associated with a given phenotype that's represented in the database, but that's really just the starting point. There's a number of steps that we have to go through, and to get to the point where we're really confident that we can take a clinical candidate into patients. I'm not going to read through each of these steps here, but suffice it to say, the devil's really in the details of trying to understand how do we move from these thousands and thousands of loci that are implicated to what we believe are meaningfully druggable targets that are significantly associated with disease pathogenesis and which may address unmet medical needs. Next slide, please. Just to give you a quick example of this, and there will be quite a bit more detail coming later from Adam, Jennifer, and Adrian on this. I just want to give an example of how we can leverage our database in unique ways. We established an immuno-oncology signature by taking advantage of the fact that we had really good representation of many different autoimmune and inflammatory diseases, as well as quite a large number of patients that had self-reported skin cancer. We hypothesized that there may be genetic associations that go in one direction with risk of autoinflammatory diseases that may actually be protective for the development of cancer and vice versa, because these patients that have hyperinflammation may also have better immune surveillance of incipient tumors. Indeed, this immuno-oncology signature has identified a number of targets, and one target that we're gonna talk to you about today, which we're very excited about, is the CD200R1 pathway. CD200 is a ligand that's expressed on tumor cells. It can engage CD200R1 on T cells and other immune cells that may be surveilling the tumor. It signals inside the T cell through a number of different intermediates, among them Dok-2, which ultimately lead to an attenuation of an inflammatory response and a reduced anti-tumor response. We were very excited in our IO signature to see that all three of these components that I mentioned, CD200, CD200R1, and Dok-2, all had the same direction of effect, where there were significant associations with those loci in increased risk for autoimmune disease or decreased risk for cancer, or vice versa. The fact that three components of this signaling pathway were represented in this IO signature gave us quite a bit of confidence that this would be an important node that could be targeted in terms of treating solid tumors. This is just one example. We're applying this type of approach to many diseases that are well represented in our database and not just cancer. Next slide, please. Just sort of extending on that, discovering the target from the genetics is one thing, but there's actually quite a bit of translational work that has to go into moving this target discovery into actual drug development. It's important to have complementary data to show that not just the genetic pathway is associated with cancers, but actually the target is expressed in tumors. Here we're showing an immunohistochemistry slide from a solid tumor, where you can see that the brown staining represents CD200 expression that's very highly expressed in these tumor cells. Then on the right, we've shown schematically, and again you'll see more data coming up in the next few minutes, that we've developed a highly potent antibody against CD200R1, which inhibits the interaction of CD200 and CD200R1, and it releases what we believe will be more potent anti-tumor immune responses driven by T cells in the tumor. Next slide, please. I just wanna finish here, but I hope I've conveyed my enthusiasm for what we're able to do uniquely at 23andMe and come back to this flywheel that that Anne alluded to, where we have this fantastic database and as we continue to add phenotypic data, we're going to continue to develop useful insights that may help drive new therapeutic options and feed back into into this virtuous cycle. With that, I will wrap up, and I will pass the baton to Adam Auton, who's gonna tell us in much more detail about our approach to genetics-based target discovery. Thanks. Hello, everyone, my name is Adam Auton, and, as I think, Joe and Kenneth have both alluded to, we really do believe that genetics offers us a unique opportunity to really improve our ability to identify, novel targets and ultimately bring those targets through to novel therapies. What I want to do today is to tell you a bit about some of the analytical approaches that we take in, trying to identify new targets and try and express some of the reasons why we really do believe that this is a fundamentally powerful way of identifying novel targets. With that, next slide please. An analysis that we rely on very heavily at 23andMe is known as a genome-wide association study. Just by way of level setting, I wanted to explain what a genome-wide association study is and give some insights into why that is a useful tool for identifying novel targets. When we're thinking about genome-wide association studies, we start with these genetic variants in the genome known as single nucleotide polymorphisms, which are shown in the top of this figure here, where you might just have an individual base pair or letter that differs between, say, my genome and your genome. Given that we can measure one of these single nucleotide polymorphisms, what we want to do is we want to ask the question: Is that genetic variant associated with a particular disease? In order to do that, we collect a large number of individuals who have told us that they have a particular disease. Those are our cases. We want to compare that to a large number of individuals without the disease. Those are our controls. Given that we can measure this genetic variant in all of these individuals, we can perform a statistical test asking, is this genetic variant more or less common in the individuals with the disease? Now, of course, we don't just wanna do that at a single genetic variant in the genome. We can actually do that across millions of genetic variants across the genome. If we do that, we get a figure that's shown in the middle here, which is known as a Manhattan plot. It is so-called because if your genome-wide association study is working well, it should resemble the skyline of Manhattan with many skyscrapers peaking up over the top here. What we're doing here is we're scanning along the genome, along the X-axis, testing each of these genetic variants as we go. You can see the chromosomes numbered along the bottom here. At each variant, we're performing this statistical test, and the strength of statistical association is indicated on the Y-axis. For genetic variants where we think there really is overwhelming evidence of a link between the genetic variant and the disease, we're coloring those in red. This is really much a ten thousand foot view. We're looking across the whole genome here, and you can see in this particular disease here, there are dozens of regions showing up as being associated with this particular disease. Really the beauty of this type of analysis is that we can actually zoom in on any one of these association peaks, as we're doing at the bottom here. What we can now see is how are these genetic variants interspersed amongst a number of genes within the region, and we can begin to ask the question, which of those genes is really driving the association with disease? Are any of those genes really representing therapeutic opportunities for intervention or cure of that disease? Next slide, please. Really the key to these types of genome-wide association study is scale. It really is the case that you need very large numbers of individuals who have told you that they have a particular disease in order to be able to identify these types of genetic associations. In order to illustrate that point, on the left-hand side here, I'm just showing how one of these genome-wide association studies has varied as the database has scaled at 23andMe. In this case, we're looking at osteoarthritis. Back in 2016, when I think our database in total was about 1 million people, if we looked in osteoarthritis, we could just identify a single region of the genome associated with that particular trait. If osteoarthritis was an area that we were interested in for therapeutic development, we really wouldn't have many opportunities for identifying novel targets within this area. If we looked under this genetic association peak and there was a gene that was targetable, that would be great, but more likely than not, there wouldn't be. Likely this would limit our opportunity for prosecuting osteoarthritis as a therapeutic area. Fast-forward to 2021, our database is over 11 million people, and you can now see this same pattern that you're really seeing dozens of regions associated with osteoarthritis. Now we really have the opportunity to understand what the genetics is telling us about the underlying biology of osteoarthritis and ask the question, are any of these associations pointing towards particular targets that present a really unique opportunity for therapeutic intervention? This is a pattern that we see not only in osteoarthritis, but across a whole range of phenotypes within the 23andMe database. It really is the case that the scale of the database drives our ability to identify genetic associations. Just on the right-hand side, you can see just how the number of independent associations within the 23andMe database is scaling as the database continues to grow. Next slide, please. A question we get a lot at 23andMe is, well, why don't you sequence everybody within your database? I'd just like to take a few minutes to address that particular point. I think there's a key insight in human genetics that actually the genetic variants within your genome tend to be correlated with each other. If I know what a genetic variant is at a particular region in your genome, I can likely predict with a high degree of accuracy what the nearby genetic variants are also doing. Just to give some sort of intuition as to how that type of approach might work, I'm giving a kind of example here where I've spelled out a sentence with a number of letters blanked out. Now, I'm sure if you look at this sentence, you can very quickly see that the sentence is, "The quick brown fox jumps over the lazy dog." But very much the same principle applies within genetics. By understanding the context of the genome, we can actually fill in the gaps. In this process that we utilize at 23andMe is known as genotype imputation, where we can essentially take the genetic variants that we have typed using our microarray and fill in the gaps using our knowledge of the genetic variants within the genome. Next slide, please. Just to emphasize this point, we type roughly 650,000 variants using our genotyping array. Using this imputation process, we're actually able to impute well over 35 million genetic variants across the genome, which actually represents nearly all of the genetic variants that are present within the human genome for which we could detect an association with disease. This process is really just much more cost-effective than utilizing large scale sequencing across the whole genome. Whereas I can impute a genetic sample for under a penny, if I want to do a whole genome sequencing experiment, that's likely to cost me roughly $1,000 per sample. Or if I want to exome sequence, that's likely to cost roughly $400 a sample. If I want to apply that across the millions of customers within the 23andMe database, it quickly becomes quite prohibitive. We really do believe this imputation process is a much more efficient way to go. That said, there are situations where we do think sequencing will yield us a novel benefit. For example, say in rare disease, we may want to utilize sequencing, and a large number of our customers actually consent for biobanking. If we do want to go back to the DNA sample and utilize sequencing, that's absolutely a capability that we both have and we do utilize in specific situations. Next slide, please. Just to sort of emphasize this point that we really do believe that this is a much more efficient and powerful way of going, I wanted to show a representation here of the number of associations we're actually able to identify in the 23andMe database using this imputation and GWAS approach and compare that to the number of genes that we would identify using an exome sequencing approach. There was a recent paper published from the U.K. Biobank that identified genes from exome sequencing, and it's shown as this blue circle here, representing the number of genes that they were able to identify within that dataset. If you look to scale at just the number of associations we're able to identify from our approach, you can see it's really night and day. I think this really should emphasize the fact that a type of GWAS and imputation approach is really driven by the scale of the database, and we really do think is a much more efficient way to go than trying to utilize sequencing at scale across the whole database. Next slide, please. What does this mean in practice for a scientist working at 23andMe? Well, really at the click of a mouse, a scientist coming to work on a Monday will likely have access to the best available genome-wide association study in really hundreds or thousands of distinct diseases. Just as an example here, I'm showing a representation of these GWAS studies across a whole host of phenotypic areas, including, say, cardiovascular disease, autoimmunity, and allergy. Because of the database, really at the click of the mouse, a scientist will be able to just drill into one of these and likely have novel insights that are available to 23andMe that just aren't available anywhere else within the community. Just to highlight as a way of example, at the bottom here, I'm pulling out an individual phenotype here known as non-alcoholic fatty liver disease. In this situation, when we made this slide, we had about 48,000 cases within the 23andMe database, and we compared those to roughly 2.5 million controls. Those numbers are notable because they're roughly an order of magnitude larger than the largest published study within this particular disease. When we run an analysis here, we find 104 regions of the genome that are associated with NAFLD, each of which may represent a unique insight that could give us a potential therapeutic opportunity within this particular phenotype. I think a really important point is the 23andMe database is not static. It continues to grow. We continue to collect new data from our research participants, and so we actually refresh this type of analysis on a regular basis. Actually, of those 104 associations that we identified, roughly 44 of them were identified within the last six months, really highlighting the fact that we are able to generate new insights continuously as we continue to collect an increasing amount of data from our research participants. Next slide, please. Now I want to talk about a second type of analysis that we utilize at 23andMe, and it's been alluded to by both Kenneth and Joe, namely a phenome-wide association study or PheWAS for short. As opposed to a genome-wide association study where we're considering one disease and looking at all the genetic variants across the genome in that disease, in this situation, we're actually gonna consider one or a small number of genetic variants and look at all of the different phenotypes that are associated with that particular genetic variant. Just as way of motivating example on the right-hand side here, I'm showing a PheWAS for an individual genetic variant that is associated with your ABO blood group. Specifically, this genetic variant determines whether you are likely to have the type O blood group or you're likely to have either the A or B blood group. What you can see at the top here is that we have a number of diseases within our database that where you have increased risk for developing that disease if you have type O blood. Likewise, there's a number of diseases moving in the opposite direction, where you have decreased risk from having the type O blood group. This is just a motivating example, but in disease relevant genes or genetic variants, this can actually be a tremendously useful tool when thinking about therapeutic target discovery and target development. Specifically, this type of analysis can do a couple of things that are very, very valuable. First of all, it could indicate potential indication expansions. I may have discovered my genetic association within a particular phenotype, but using this PheWAS approach, I can now look and ask, is there a better opportunity that we should be considering, for therapeutic development, where maybe there are, other reasons such as competition that we think we may have an advantage over what is currently being done? Likewise, this type of analysis can really at the click of a mouse indicate whether there are potentially unwanted effects from targeting a particular, gene. For example, you may see in this type of PheWAS analysis an association with cardiovascular disease, and maybe that presents a risk for your therapeutic program in a different area. You can ask the question, is this a risk for this particular program? Is this a risk that we can discharge? We can ask all of those questions really at the very beginning of the program before anyone's picked up a pipette. Next slide, please. This type of PheWAS approach has actually been very, very interesting, particularly within the immuno-oncology space. As Joe was mentioning earlier, you can actually identify situations in the genome where you can identify a genetic variant that both appears to be activating the immune system and at the same time suppressing risk for the development of cancer. We first noticed this effect by considering targets that have been progressed within immuno-oncology. I'm just on the left-hand side here, I'm showing a particular example for CTLA-4, which is a well-known target in immuno-oncology. You can see on the left-hand side here, you see a number of immune phenotypes that are associated with this genetic variant within this particular gene, all of which seem to represent an activation of the immune system, and then there's a corresponding reduction in the risk for developing cancer. Having observed that we see this type of signature around immuno-oncology targets, we wanted to generalize that approach. What we've done is we've developed a classifier that can essentially scan across the genome, testing each gene individually and asking, "Well, is there evidence for this immuno-oncology signature?" In doing so, we can actually scan through and identify which genes actually have the same signature that we observed within existing immuno-oncology targets. This has actually proven to be a particularly fruitful way for identifying novel targets, one of which has come from this type of approach is our CD200R1 program, which I'm now going to hand over to Jennifer to tell you more about. With that, I'll say thank you and hand you over to Jennifer Low. Thank you. Hi. Thank you, Adam. I'm Jennifer Low, and I joined 23andMe three years ago to build a clinical development organization. I'll be joined in this part of the presentation by my colleague, Dr. Adrian Jubb, who is heading our oncology franchise at 23andMe. A couple of weeks ago, we announced that we have started our first 23andMe-sponsored clinical trial with 23ME-00610, an antibody against the CD200R1 immune target in oncology patients. Adrian and I will be telling you more about this target and about the clinical trial, but in the meantime, we're gonna go back to the immuno-oncology signature in the next slide. The immuno-oncology drug development space has the potential to treat a vast variety of cancers, with the market having been expected to exceed $41 billion in sales last year. In 2020, there were almost 5,000 agents against over 500 targets in development. Why would 23andMe try to compete in this space? Well, Adam already showed us how we are using our genotype and phenotype data to shine the spotlight on immune targets that are genetically driven to predispose to autoimmune disease and cancer. Using our immuno-oncology signature, we're able to identify the most interesting targets to pursue in this space. Next slide. Broadly speaking, our immuno-oncology signature identifies genes that may be more active in people who have developed autoimmune disease but are less active in people who have developed cancer or vice versa. This allows us to identify genes that may play a more causal role in patients' immune disease and developing cancer, and these targets may be more responsive to modulate with therapeutics. In this slide, on the left, we have a heat map for the immuno-oncology signature of CTLA-4, which is the target for the drug ipilimumab. The blue immune and autoimmune phenotypes indicate an activity direction that's different from the cancer phenotype shown in red. On the right, three components of the CD200R1 pathway show a similar opposing phenotype with the CD200R1 receptor and its downstream signaling protein Dok-2 in one direction and for the ligand CD200 which, as expected, is showing directionality in the opposite direction. Because of this, we're really enthusiastic about CD200R1 as a potential target. I'm now gonna turn the presentation over to Adrian Jubb, who is gonna tell us more about this pathway. Adrian. Thank you, Jennifer. It's my pleasure this morning to represent the CD200R1 team and tell you a little bit about the program that we've just brought into the clinic. CD200R1 is a known immune inhibitory receptor. Next slide, please. It's expressed on T cells and on cells from the myeloid lineage in the tumor microenvironment, and it has a very well-known linear canonical signaling pathway, which is shown on the right-hand side of this slide. The ligand CD200 is expressed by tumor cells or other cells in the microenvironment and engages CD200R1 on the surface of immune cells, activating a downstream signaling pathway that ultimately leads to a reduced anti-tumor immune response. In addition to the really powerful human genetics that identified many components of this pathway as promising immuno-oncology targets, the right targets, as Joe talked about in his presentation earlier, there is actually additional genetic evidence that supports CD200 and CD200R1 in human tumors. That comes from oncogenic viruses that have co-opted a homolog of CD200 and use it to manipulate the human immune system to allow the cell types that they infect to grow more efficiently and ultimately to form human tumors. In addition to human genetics, there's actually viral genetics too that point to this as being a potentially important pathway that could have a therapeutic effect for patients. Next slide, please. 23ME-00610 is an extremely potent antibody that binds to and inhibits CD200R1. It has a very high affinity, and it's able not only to block the interaction of the ligand and the receptor but also to displace pre-bound ligand receptor complexes that are often found in the tumor microenvironment. In doing so, it's able to restore T cell activity either directly or indirectly through its effect on myeloid cells, and thereby activate T cells to attack the human tumor cells. Our biology group has been able to demonstrate evidence for this using multiple model systems with human immune cells, giving us confidence that the genetics that we alluded to earlier will translate into a promising therapy for patients. In addition, in our IND-enabling studies, we did not observe any concerning toxic effects of the drug, indicating that we anticipate a wide therapeutic window to be able to explore the potential of this drug as it moves into the clinic. Next slide, please. This is just a snapshot of the biology data that the team has generated for 23ME-00610. A broader presentation of both the genetics and the biology will be presented at a medical conference later in 2022. For today, I'm just going to take you through this slide to give you an indication about some of the properties of this drug and the pathway that excite us so much in treating cancer patients. The data I'm showing here today is an in vitro assay, a cell-based assay in a dish using human immune cells from patients with cancer. These patients have many different types of cancer, and that's shown on the X-axis in the graph on the right-hand side of the figure. Those cells are initially treated with a very nonspecific stimulant of immune cell activity, SEB. In addition, we've treated the SEB-stimulated cells with a variety of different antibodies, including an isotype control antibody, a negative control, 23ME-00610 inhibiting the CD200R1 pathway, and also, anti-PD-1, which is, as many of you are aware, an approved immuno-oncological agent. What you can see on the right-hand side in our graph is a measure of immune cell activation. In this case, it's interferon gamma, which is a pro-inflammatory cytokine known to have important anti-tumor effects. When we compare the changes in interferon gamma from the isotype control to 23ME-00610, we see increases in almost every donor and tumor type tested. Importantly, when we compare that to the approved immunotherapy anti-PD-1, in almost every case, we see a significantly greater increase in immune stimulation with 23ME-00610, and in some instances, in patients that have minimal or no response to anti-PD-1 therapy. It's this sort of data that really compelled us that 23ME-00610 has unique properties that could potentially differentiate it from existing immuno-oncology agents that are in the clinic already. Next slide, please. With these data, we filed an IND late last year and have started a phase I clinical trial. That phase I study has dosed its first patients, which we announced within the last few weeks, and it is moving through the dose escalation phases now. It's an open-label, non-randomized, multi-center trial for patients with very advanced locally or metastatic solid tumors. It starts off with an accelerated titration to get us into the therapeutic range quickly and then moves forward with a more standard 3 + 3 design to the recommended phase II dose. Once we've achieved the recommended phase II dose, we will expand into multiple different indications to test very distinct hypotheses that we've generated with our genetic and biology data at 23andMe. We'll be giving more details about the types of indications that we'll expand into nearer the time at a later point in the conduct of the trial. The primary objectives are safety for dose escalation and efficacy for the expansion cohorts. In addition to traditional endpoints of efficacy in cancer trials, we're also looking at immune activation, pharmacodynamic biomarkers, and also markers that are relevant to the action of 23andMe-00610 on the CD200R1 pathway to ensure that we're engaging it properly and seeing the effects that we anticipate from our preclinical work. Next slide, please. Now, as Joe mentioned earlier, choosing the right patients is critically important to a successful and efficient clinical development program for many drugs, and in particular for 23andMe-00610. Later on, Jennifer will tell you a little bit about how we're going to leverage genetics to try and inform which patients' immune systems are likely to be most sensitive to immune activation with checkpoint inhibitors such as 23andMe-00610. In addition to that, parallel to the work that we've done to develop the therapeutic antibody, we've also been developing agents that allow us to interrogate the activity of the CD200R1 pathway in human tumors so we can best select for those that are exploiting the CD200R1 pathway to evade the human immune system. We're showing just a snapshot of that data here today with immunohistochemistry for both the CD200R1 receptor on the left-hand side and the ligand on the right-hand side, showing brown immunoreactivity for the immune cells that express the receptor in the human stroma and also for both tumor and intervening stromal cells that express the ligands on their membrane on the right-hand side. We anticipate using a variety of approaches to try and identify which patients are most likely to benefit from these therapies. Next slide. Choosing the right target is also critically important, and as the genetics has shown you already, and as you perhaps have appreciated from the biology and the ligand signaling pathway for CD200R1, there's actually the opportunity to intervene at multiple nodes in the CD200/CD200R1 pathway. We've specifically chosen to drug CD200R1 because it's expressed on immune cells, and we're confident that we'll be able to inhibit the pathway fully by targeting that receptor. Some of you may be aware that the ligand CD200 has been previously drugged by Alexion Pharmaceuticals with their agent samalizumab, also known as ALXN 6000, in hematological malignancies, several of which express CD200 on the tumor cells themselves. In this particular study, phase I study, unfortunately, Alexion was not able to saturate cell surface CD200 with their samalizumab antibody. As a consequence, they weren't fully able to shut down signaling through CD200R1, owing largely to the abundance of CD200 ligand that's expressed throughout many different cell types in the body, including tumor cells. By contrast, CD200R1 expression is much more restricted just to immune cell subsets, and our PK/PD modeling, pharmacokinetic/pharmacodynamic modeling, indicates that we're much more likely to be successful in suppressing the pathway with the approach that we've taken. In addition, 23ME-00610 is a much more potent antibody than samalizumab or ALXN 6000, with greater than 100-fold higher affinity for its target, its ability to displace preformed complexes of ligand and receptor owing to that high affinity. Also in many biology assays, where we have run the two antibodies head-to-head, we've been able to show superiority of 23ME-00610 over anti-CD200 targeting approaches. We're confident that with these insights, and the diligent work by our biology team, that we're well-positioned to be able to drug CD200R1 pathway, as efficiently as possible and bring that benefit to patients. Next slide, please. Just to summarize, Jennifer's told you about the genetics, and we believe that both human and viral genetics have pointed us to CD200R1 as the right target. We've told you about the properties of 23ME-00610 and its ability to bind CD200R1 with high affinity, and we believe that that's the right drug. We've told you also about some of the insights that we have into that pathway and our ability to pick the right patients, the right populations in which to test this and ensure that we're able to demonstrate proof of concept early in the development program. We've started on that journey already with the announcement of a first subject dosed for the phase I study of 23ME-00610 in the past few weeks. With that, I'll turn you back to Jennifer Low, who will tell you about our CD96 program. Thank you, Adrian. Now I'd like to turn to the CD96 program, which is the first immuno-oncology antibody targeting CD96 to enter the clinic. This is a program that has been conducted in collaboration with GSK, and for which we announced this morning that we will be taking a royalty option on this program. Once again, next slide. Let's start with the immuno-oncology signature. Again, the CTLA-4 signature shows the blue bars for immune-related diseases and the red bars indicating opposite directionality for the oncology-related diseases. Meaning for this gene, there is an association, and it is an opposite direction for autoimmune diseases and for cancer. Here we see the CD226 pathway, and I'll explain why the CD226 pathway is important in a moment. CD96 is a part of the CD226 pathway, and as you can see, also shares a very similar IO genetic signature. Next slide. There are a lot of components in the CD226 pathway. You can see CD226 in dark blue here in the bottom middle of this picture on the immune cell. CD226 activates NK and T cells. This is a good thing because these cells can recognize and kill the cancer. The well-known PD-1 pathway also interacts with the CD226 pathway and may be one mechanism of how PD-1 inhibitors work. TIGIT and CD96 also suppress CD226, so inhibiting them may also lead to increased CD226 activity. Combining inhibition of PD-1 with CD96 or with TIGIT may lead to an increase in antitumor activity, and in fact, combining PD-1 with TIGIT has already shown promising phase II results. Next slide. Preclinical data supports combining CD96 with PD-1 and TIGIT inhibitors. Here we show published data on the left in a mouse model that supports this hypothesis. As you can see in the bottom line in pink, which has CD96 added to PD-1 and TIGIT, there is the most suppressive activity in this tumor model. Over the last few years, GSK has acquired or partnered to gain control of four of the assets known to act on the CD226 pathway, including PD-1, CD96, PVRIG, and TIGIT. Next slide. Here's a schematic of the phase I study design that GSK has previously shown. This study is a phase I first-in-human study in patients with solid tumors, and GSK'608 is what we are calling the anti-CD96 antibody. There is a standard 3+3 dose escalation for the single agent GSK'608, and then there is a dose escalation for the GSK'608+ dostarlimab, the CD96 PD-1 combination arm. This study started in July 2020, and GSK is expected to show data from this study later in 2022. Next slide. To conclude, the 23andMe immuno-oncology signature has highlighted the importance of the CD226 pathway, which includes CD96 and TIGIT. Combining the inhibitors of CD96, TIGIT, and PD-1 to activate the CD226 pathway may be more efficacious than inhibiting single components. However, in order to demonstrate the contribution of components, this will require more complex clinical trials. GSK has the relevant agents to target the CD226 axis. The phase I clinical trial is ongoing and data is expected later in 2022. This morning, we announced that we will be taking a royalty option on this program, and we remain optimistic about the success of this program. Next slide. So far, we've talked about using genetics to find targets. Now I'd like to talk to you a little bit about how we are using clinical genetics to improve the success of our programs in clinical development. Next slide. What if we could use genetics to predict immune function and immune response to immuno-oncology agents? Next. At 23andMe, we believe that an individual's genetics can tell them and us a lot about their biology, and that includes how they may respond to therapeutics. Genetics for an immuno-oncology drug may predict which immune-related adverse events they may have and whether they may have efficacy to an immune-stimulating drug. In our 23andMe clinical trials, we are having our patients spit in a tube just as they do for our consumer product, and then they are genotyped. We will be running genetic classifiers to do exploratory outcome research so that we can evaluate how genetics may be useful in predicting side effects and efficacy. Because we have a time-tested product with FDA-authorized reports, we will also be giving our patients the option to receive information about clinically informative variants just as we do for our customers. Next slide. In fact, there have already been some preliminary data evaluating whether genetic data can be useful in predicting immune-related adverse events and efficacy for the PD-L1 inhibitor atezolizumab. On the left are two or three phase III clinical trials with atezolizumab that were negative, the top one in breast cancer and the bottom in bladder cancer. However, in these two papers shown here, scientists looked at the polygenic scores for hypothyroidism and psoriasis, two immune-mediated diseases that also appear similar to the side effects that one may get with atezolizumab. What they found were that these polygenic scores calculated from the subjects in their clinical trial not only predicted which patients were likely to have side effects, but also whether they were more likely to have progression-free or overall survival benefit. Because of the size of our database, we may be able to generate more accurate and impactful polygenic scores, and this could improve our clinical trial outcomes. Next slide. In summary, 23andMe is incorporating clinical genotyping into our clinical trials because this could enable more efficient clinical development and improve the probability of success for all of our programs. Developing drugs in genetically defined populations may differentiate future medicines based on better outcomes and improved benefit/risk profiles. We will be providing the right drugs to the right patients. Now I'm going to turn the mic over to Adam to provide the summary. Thank you. Thank you, Jennifer. I think that concludes our section on therapeutics, so I just wanted to give a summary of the things that we've discussed today. I think we've highlighted that 23andMe has generated an impressive research platform that is covering multiple therapeutic areas in a number of indications with high unmet medical need. The platform has continued to be very productive, and we now have more than 40 programs that have been generated from data within our database as part of our collaboration with GSK. Reflecting that success, GSK has extended their exclusive target discovery period for the collaboration with 23andMe for an additional fifth year, as we announced today. A recently announced target, CD200R1, is advancing into the clinic. Adrian and Jennifer have given you oversight of how that program is progressing. 23andMe has chosen to take a royalty option on the immuno-oncology antibody collaboration program targeting CD96 into later stages of development. That was an announcement that we made this morning. Finally, managing our therapeutic portfolio investments based on scientific data will help us optimize our investments and mitigate the risk of our future potential returns. With that, thank you very much. With that, I'd like to hand over to Paul Johnson, who will now be talking about the genetics-based primary care. Thank you, Adam. My name is Paul, and I was the Founder and CEO of Lemonaid, and now I'm really excited to be leading the consumer organization at 23andMe. Next slide, please. I think everyone would agree that there is a lot of room for improvement when it comes to healthcare. In fact, we know from an article published in the New England Journal of Medicine a few years back that healthcare is only a small factor when it comes to preventing premature death. As you can see from this chart on the right, healthcare alone is only able to reduce early deaths by about 10%. On the other hand, when you combine genetics and individual behavior, you make up 70% of the factors that determine premature death. We believe that by linking genetics with primary care, we can help drive changes in behavior and treatment that can significantly improve people's overall health and well-being, all provided in a consumer-centric setting. Next slide, please. By combining 23andMe and Lemonaid, we're able to deliver something that's never successfully been done before, genetics-based primary care at scale. Together, we have all of the components required to deliver this. We're helping patients through telehealth solutions in every state in the U.S. and the U.K. We're ordering and interpreting, providing diagnostic tests. We're providing actionable wellness reports, and we have a nationally licensed and operating pharmacy. Combined with that, we have medical records that are longitudinal, so we can help our patients for more and more of their healthcare needs over time. Next slide, please. What is genetics-based healthcare? We have a really unique and impactful opportunity to create truly personalized and preventative healthcare that's founded on individual genetics. This is really the realization of the mission that Anne talked about at the beginning to help people access, understand, and truly benefit from the human genome. As an organization, we're executing on four fundamental pillars to deliver a genetics-based healthcare organization to our approximately 12 million consumers. First, identification to drive targeted prevention, monitoring, and management of personalized healthcare predisposition insights. We're able to provide really detailed and personalized interpretation and solutions to patients based on their genetics. Second, providing detailed insights on carrier status with guidance and advice on understanding potential risks. Third, tools and support to help you feel your best, from personalized wellness coaching to lifestyle guidance, advice, and partnerships. Fourth, pharmacogenetics. We're prescribing the best medication that will work for you as an individual based on your genetics with the lowest risk of side effects. Next slide, please. All four of these pillars and the capabilities that Lemonaid and 23andMe bring together have one thing in common, the ability to provide personalized interpretation and healthcare at scale. Healthcare today is too often just about the average. We're looking to change that and build healthcare that's truly about you as an individual. In order to do that, we're becoming a consumer-centric, holistic wellness and healthcare brand. Much of how well you feel today is driven by lifestyle choices you make, your environment and your genetics, and 23andMe is becoming the first brand to combine all of these elements so that people can live healthier, happier, longer lives. By doing this, we're truly driving strong strategic differentiation in a way that no other organization today is able to. Now I'm going to turn over to Jeff Benton, who's gonna talk about the power of polygenic risk scores for personalized healthcare. Jeff. Thanks, Paul. I'm excited to talk a little bit about, dig into the details about what we're doing on the consumer side of 23andMe and leveraging the science and the data that we have. Next slide, please. We've already seen the flywheel a couple times this morning from Anne and Joe, and I'm excited to be able to drill in a little bit more on the novel consumer products of this flywheel and how we leverage the phenotypic data and genetic data that we've collected over the last decade plus, and how we're actually using that data to build exciting data-driven products that bring more value to our customers, that incentivize them to provide more data because they know they will receive more value and thus get this flywheel spinning and moving. Next slide please. Paul already talked about the four pillars of our health service and this gives a little bit more detail into the offerings that we have from health predispositions all the way through to pharmacogenetics. Today, I'm gonna be talking a little bit more about aspects of the health predispositions and wellness portions of our product and how we leverage the genetic and phenotypic data to build new products and the next generation of products as we move beyond these. Next slide. Adam did a great job earlier today digging into the details of genome-wide association studies and how we can leverage the insights from these associations that we find to find new drug targets and then to go on and build the amazing, incredible therapeutic portfolio that we have today. The beauty of genome-wide association studies is we can leverage these key insights for many other aspects of our product and service, especially on the consumer side. Many of the health reports on the health predisposition and wellness reports are actually driven by the insights that we've gleaned from these genome-wide association studies. We can take these same insights, hits that we see from these studies and use them to build advanced models. Next slide, please. At a very high level, we use the GWAS as the foundation for our predictions using cutting-edge machine learning technology and techniques. We can do feature selection and modeling of incorporating those genetic features, sometimes upwards of tens of thousands of these single nucleotide polymorphisms, incorporate them together to build personalized risk estimates that we can then give to our customers. Let's go a little deeper on what a polygenic risk score actually is, and the value that it can provide. By doing so, I'm gonna use an example of body mass index to show what this looks like in our customers. Next slide. In this graph on the X-axis down below, we've split the 23andMe customer base into 20th percentile buckets, so 5% each from one to 20. The people on the left-hand side in bucket one have the lowest predicted BMI based on this polygenic risk score. Those in the highest bucket, the 95th percentile, bucket 20 there, have the highest predicted BMI. What you can see on the Y-axis here is what we've plotted is the mean actual body mass index of the people within these buckets. You can see if you compare the lowest predicted bucket one to bucket 20, you can see nearly a fourfold difference in BMI from those people. In these buckets, this is the actual body mass index of the people that fall into these buckets based on their genetics alone. You have, you know, close to a 16-pound range difference between individuals just simply based on their genetics and the power of their polygenic risk score. That shows that this is real information and real predictions, but what does this look like for a customer? On the next slide, you can see an example of what this genetic weight risk report looks like for our customers today. In this case, the example report here, "John, your genes predispose you to weigh about 8% more than average." We've taken this polygenic risk score and the interpretation thereof and put it into text and information that a customer can understand. Importantly, you'll see kind of below the top line result, we also mentioned that this is a predisposition, that this is, you know, there are clearly lifestyle and environmental factors that play a large role and that genetics is not destiny. We're gonna talk a little bit more about how we're going to incorporate those lifestyle and environmental factors shortly. On the next slide, though, we'll give you a sense of over the last two fiscal years, just the broad range of things that we have been able to release based on polygenic risk scores. Adam mentioned scale in his portion of the presentation. Scale is what allows us to have this breadth of phenotypic coverage for our customers, everything from coronary heart disease and other heart-related outcomes to things like allergies and eczema, acne and nearsightedness. This, you know, broad range of phenotypes allows us to hit topics that our customers are interested in and concerned about and give them reports. As you can see, we have a robust pipeline that we have the ability to continue to release these on a regular cadence and plan to do so for our customers going forward. Next slide please. As Paul alluded to and Anne alluded to earlier this morning, clearly genetics is not the only part of the prediction puzzle or what goes into our risk for disease. Clearly, lifestyle and environment play a really key and important role in this puzzle that we are trying to put together to help people understand what they can do to be healthy at a hundred. The future of the product and where we're headed is to incorporate genetics, lifestyle and environment, use machine learning, and to get to a truly personalized disease risk assessment. This will allow us to provide personalized interventions, help people understand their risk over time, and figure out how they can change this risk with the decisions that they're making in their lifestyle and in their health decisions. I'm gonna go into a specific example to show how this might work and show the promise of where we could be heading with this. Next slide. This, I'm gonna use an example of type 2 diabetes and the data that we have within 23andMe to show what this can look like practically. In this graph you can see we've graphed incidence rate over time. This is one year incidence rate. The power of the 23andMe database is that we have a health survey that we ask on a yearly basis, and people answer that on a yearly basis and give us longitudinal data. That allows us to find and identify people who have gotten diabetes in the last year since we talked to them last. What you can see here is the total population there in blue. This is the incidence rate over time, and you can see the age of onset here. In the red line there, those are the people in the 95th percentile of the polygenic risk score. Imagine bucket 20 when we looked at the BMI graph. Individuals that are in that 95th percentile for type 2 diabetes risk based on genetics alone, they're at nearly a threefold increase in risk for developing type 2 diabetes throughout their life. If you contrast that with people in the lowest percentile of risk, the lowest 5th percentile, their risk is substantially reduced. This shows the kind of power practically of what this can mean for customers in the context of type 2 diabetes. This is kind of what we've been doing as a company on the consumer side for over a decade. Now what does this look like if we incorporate lifestyle factors and environment? Clearly, I think people understand type 2 diabetes as being also driven by lifestyle and having a key component to the risk for type 2 diabetes and the progression of that disease. Next slide. In this slide, what we've done is we've added over 20 lifestyle factors to our prediction model to the genetic risk itself. What we've done is we've taken this, the red line of the kind of highest risk profile people, and we've split it out and stratified it based on their lifestyle factors. You can see in this purple line that goes up, these are folks that have the highest genetic risk and unhealthy lifestyle. Again, their risk is going up nearly twofold just based on these lifestyle factors alone. The exciting part about this and the empowering part for our customers and the part that we're excited about as a company and as a consumer group is to look at the power of a healthy lifestyle and decisions that can be made that actually bend the curve, flatten this curve down, such that even people with high genetic risk that was putting them at, you know, three times the risk of acquiring type 2 diabetes, developing type 2 diabetes based on the lifestyle decisions that they can make. This is everything from the kinds of food you eat, movement, you know, exercise, standing versus sitting, how you deal with stress. All of these components are in this model. We can bend that curve down to essentially population incidence and which is a really empowering message for our customers and something that we're gonna be incorporating into our reports as we go ahead. Next slide. What this finally allows us to do is to actually get at personalized healthcare at scale. Right now, the practice of medicine is all about reacting to the diseases and conditions that they come. With 23andMe+, with 23andMe, we are working towards getting to a truly individualized form of healthcare. With the acquisition of Lemonaid, we're very excited to be able to move beyond simply giving individuals reports about their risk, but incorporating lifestyle interventions, helping them take action, and ultimately connecting the dots all the way through to healthcare. Davis Liu will be telling us a little bit more about that, so I'll turn the time over to Davis. Great. Thanks, Jeff. Good morning, everyone. My name is Dr. Davis Liu. I'm the Chief Clinical Officer of Lemonaid Health. I am a board-certified family physician. Before Lemonaid, I was actually a primary care doctor at Kaiser Permanente in Northern California for 15 years and also served on the chronic disease management team over at Kaiser Sacramento Roseville for three years. As the only doctor in my family, I'm super excited about what we're talking about today. The possibility of actually taking what Anne talked about, Paul and now Jeff, about taking the vision of genetics into reality, how to make sure that everyone has the insights, genetic insights offered by 23andMe, and understand what behavior changes and interventions they can do to live a long and healthy life. Next slide. A bit about Lemonaid Health. Lemonaid Health is a fully integrated healthcare offering that's available in all 50 states and D.C. We currently offer over 20 services online. We have a dedicated patient support team, medical team, as well as pharmacy team, and we have a common leadership structure that it also includes a technology stack that is both proprietary, customizable and flexible. This allows us to ensure that every patient has a seamless experience at every touchpoint. It also allows us to ensure that they get the most up-to-date guidance and information, every patient, every time. This is what Lemonaid actually brings to the 23andMe family, and that's why I'm super excited to talk about how this allows us to offer a genetics-enabled primary care at scale. Next slide. Imagine the Lemonaid platform that's frankly supercharged with the robust, scientific insights that 23andMe offers, specifically making the art of medicine into the science of medicine for the first time. Imagine your doctor knows you at a really deep level now, really at the genetic level, and talk about specific risk factors, lifestyle factors, and also, goals that you have for wellness to understand what you need to know in a way that no one's actually been able, done before. You can do this through any sort of modality you'd like to, whether it's messaging your doctor, a phone call, or a video. Now imagine what Jeff talked about and that now your doctor actually knows your polygenic risk score. Let's talk about type 2 diabetes and what that really means for you, the individual. This is actually not something that can be done in traditional healthcare today. Now imagine your primary care doctor actually knows what tools and resources now to implement at scale, what lifestyle changes you need to do, what coaching might benefit you, and be proactive and start testing for the first time in a way that's much more proactive than the traditional healthcare system today and monitor you, more actively today and actually intervene even sooner today. Because we know the healthcare system in America, it usually takes 17 years for something to be published to actually take place in common practice. Today, now, we have the opportunity to do that sooner than later and ensure that every person has a long, productive, and healthy life. Next slide. Now you can imagine it's the first time healthcare can actually focus on the material factors that actually make an impact on an individual's wellbeing. It's not just the medication. It's not just the therapies. It's actually your genetics. It's actually what choices you make, and it's actually being informed about those choices you're making and how that makes a meaningful impact on your own personal life. Now for the first time we have a long-term relationship with a clinician who actually understands this level of scientific insights, and over time, these insights get more advanced, more thoughtful, and more interesting. Of course, your doctor's able to loop those in sooner. It allows you to live a long, productive, healthy life. This is just the beginning. I think it's really the first time that we can offer patients the ability to have genetics-based primary care at scale. With that, I'm gonna turn it over now to Anne. Next slide. Great. All right. Well, thank you to everybody, and I hope you got a taste of some of the things going on at 23andMe besides just, you know, other than just hearing me and Kenneth and Steve speaking. You know, the joy I have at this company is being surrounded by truly phenomenal individuals who are really pursuing this mission of helping people, you know, access, understand, and benefit from the human genome. Where I see the company going is, you know, leveraging everything that we just heard here today. We will absolutely continue to be the leader in direct access to your genetic information and really helping people benefit from that information. When I think about, you know, where we put our priorities, it's on, you know, people like Jeff Benton and his team and pioneering what is that next generation type of report. Obviously on the Lemonaid team, we're thinking about, well, how now are we going to integrate all that into true genetics-based primary care services. How can our customers really leverage all of this information to ultimately have that goal of living to be healthy at 100, using this information, being able to potentially change their lifestyle or get more, you know, more proactive care from the medical community to really have a different outcome with their life. Obviously I'm also very excited about the therapeutics team and everything that is happening there. It has been just an absolute, amazing journey to see the collaboration with GSK and how productive that has been. Like we mentioned before, we have a pipeline of over 40 clinical and research stage programs. Those continue to move forward. Very excited to see how those continue to progress. Last, obviously one of the highlights from last year was our IPO and our SPAC. The consequence of that is that we have a strong balance sheet, and we constantly are thinking about how we are using that cash to either support activities in the consumer side or on the therapeutic side. By having a strong balance sheet, we are in a great position to be able to continue to think about how we wanna grow both of those areas. With that, thank you again for joining, and we are very excited to answer questions. I will hand over to Wade to start, getting Q&A. Thank you, Anne. Our first question is along the lines of how our relationships with healthcare systems are involved, particularly public healthcare systems. The person asking this question wants to know, are we able to increase the data we acquire by having relationships with healthcare systems like he mentions New York City Health and Hospitals. He thinks that- Mm-hmm ... access to a public health system might help us gather more data and understand more data. Mm-hmm. I can jump in and then and Paul, you can see if there's anything else, or the research team, or Kenneth, if you wanna jump in and answer some. We do not have any of these partnerships right now. We occasionally have very specific research programs. If it's a specific phenotype and we wanna collect data on, you know, deeper data on a specific area, we can do programs. We don't have anything broad. Part of what 23andMe is very different in is the re-contactability, and a lot of these data sets that come from public health facilities are de-identified and not re-contactable, so it's a static data set. It's interesting, and it's potentially valuable for us for, you know, validating a finding that we have, but it is not something that we've actively pursued as much because they lack the re-contactability that is important to us. Thank you. We have Tiago Fauth from Credit Suisse on the line, and he wants to join and ask some questions, so we're gonna bring him on. Great. Great. Thank you so much for hosting the event. Appreciate it. Actually, I have a couple of more technical questions on 610 specifically. I guess the first question would be how to think about this target and its utility, 'cause there are some differences in expression patterns for CD200 receptor and ligand relative to other checkpoint inhibitors across different cell subtypes. Is there any way to compare this to other checkpoint inhibitors in terms of how broadly you could utilize this molecule? And to a related question, you alluded to that to a certain degree. Can you actually prospectively try to enrich your studies for cancer types that are more likely to benefit from that depending on tumor microenvironment or whatever sort of biomarkers you can get? Let me stop. I have a couple more, but let me stop there. Kenneth, you wanna jump in or point to someone else? I think Kenneth's having a- I think Kenneth's having technical difficulties. He can't hear us. We're gonna have Jennifer jump in. Great. Jennifer, are you there? I'm here. Can you hear me yet? Yes. Yes. Hi. With regard to 23ME-00610, we're really excited about the target for the reasons that we explained. The question was around whether we could compare it to other checkpoint inhibitors and what indications that we might be going after. We are going to be evaluating additional indications in the expansion cohorts. We aren't talking much about that right now, but we will be talking about that more later this year. In terms of how we select patients for this therapy right now, the trial is a solid tumor all-comer clinical trial, but some of the data that Adrian showed you showed the high expression in certain tumor cell types, the ligand, and one of the pictures we showed you was ovarian cancer. I think that indication plus others might be very interesting. As you probably know, existing checkpoint inhibitors haven't been addressing certain of these disease areas as well as the ones that there's broader approval for. We are hopeful that 23Me-00 610 will be active in disease areas that other checkpoint inhibitors have not shown as much activity in. Got it. Perfect. Perhaps so one follow-up on 23Me-00610 and one more broader on the therapeutics engine. For 23Me-00610, like given the mechanism of action, is there a way to try to handicap expectations for monotherapy efficacy by any chance? I know it's kind of hard to do based on preclinical animal models alone, but any thoughts on expectations for monotherapy activity early on? I don't think that there's much that we can say at this point. Obviously, you know, we've shown you what we're excited about for the preclinical activity, and as you point out, it is difficult to predict based on that. You know, this is something that, you know, obviously we've done a lot of biology on and we're very enthusiastic about. I don't know how else to answer your question, right? Got it. it. No, that's absolutely fair. Perhaps just a bigger picture one, just thinking about the target discovery productivity of the therapeutics engine. Of course, the increasing size of the database has contributed to a lot of the recent developments and helped you guys identify new targets and programs. Currently, if you start from today, like what are some of the limiting factors or bottlenecks and what can actually. How important is the consumer engine growth to kinda keep feeding the therapeutics engine? Thanks a lot. Should we get Adam to answer that one? Or Jennifer, you wanna answer it? I can answer that the growth in our ability to find targets has remained linear, and a lot of our data is also driven by the answers that our customers provide to us when they participate in research and help with understanding the biological basis of disease. We are continuing to have growth that we expect to continue at this time. I can maybe add on to that. I think our strategy is thinking about, as Kenneth alluded to, our power, need, and speed pillars of our strategy. We're certainly focused on the power aspect, which is driven by the growth within the consumer database. I think we also have to remember that the need and speed thereof identifying where are the opportunities for us to really prosecute targets quickly, and where's the unmet medical need. Going forwards, we'll have to be thinking about how those pillars are kind of balanced against each other. Perfect. Thank you so much. Okay. I think we have Daniel Grosslight from Citi is on and would like to ask a few questions. Daniel, go ahead. Hi. Thanks, Wade, and thanks everyone for this very detailed presentation. It's incredibly helpful for us. I just wanted to go back quickly on the CD96 program and the shift to a royalty structure rather than the 50/50 cost profit split. Was that at the behest of GSK? How does this really change or influence your capital deployment strategy in the near term? Do you anticipate slowing down R&D a bit or in the near term, or will you be reallocating some of the capital you would have spent developing CD96 in conjunction with GSK on other therapeutics, particularly wholly owned therapeutics? Shall I take that one? That'd be great. Yeah, for sure. You know, there's very much an ongoing, you know, like most companies that are in the business of developing therapeutics, right? We review those opportunities all the time when we come to decision nodes about significant capital decisions. This one was no different in that we, you know, looked at the sheer size of the expected clinical program there. You know, we modeled out what we thought, you know, the future of that program would look like from a long-term, you know, risk-adjusted cash flow point of view. You know, our cost to capital is relatively high versus big pharmaceutical companies similar to GSK. You know, we will look at that independently given our cost to capital and decide whether we think another set of things to invest in are better returns for the company than that is. That's really the basis of how we looked at this. Yes, capital redeployment for sure into more things that are at earlier stages. Okay. It sounds like this was 23andMe's call to shift to the royalty. Yeah, absolutely. Based on our specific circumstances. Yeah. Okay. On the targets of your two most advanced programs, CD200R1 and CD96, can you talk a little bit about the competitive dynamic here? Are there any other companies targeting those specific areas? Obviously these are targets that have been around and known about for a while. Can you just talk a little bit about who else might be developing therapeutics that target those areas? Kenneth or Jennifer, you wanna take that? For CD96, the CD226 pathway, more broadly is very interesting, and a number of other companies are now developing inhibitors against one component, another component of that pathway, TIGIT. To our knowledge, we are the first and only CD96 inhibitor in the clinic, and as the data I showed you before, we do think that adding CD96 to the combination will be even more interesting based on preclinical data. We're very supportive of GSK's evaluating additional combinations. With regard to CD200R1, we are not aware that anybody else has targeted this in immuno-oncology, that they are inhibiting the blocking of CD200R1- CD200, and so we believe we are the first for that as well. As Adrian stated, there has been a prior agent that targeted CD200, but there were technical reasons why that inhibition was probably inadequate. We believe that targeting CD200R1 will be the better strategy. Got it. Okay. Helpful. Once a target is selected, can you go over in a little bit more detail how your database actually impacts drug design? If the database can be utilized to assess how druggable a target might be and how best to approach that target. Jennifer, you wanna take that again? I'll start off a little bit on that, but I think Adam or Joe would be great people to answer that. We do use the combination of genotyping and phenotyping information to evaluate how to drug a target and what modalities and what other things to think about in terms of side effects and indications that might be interesting to target. I'm gonna go back to Adam to answer more thoroughly. Yeah. I will likewise let Joe jump in. I think one thing that's important to realize is that genetics does also give you indication about the directionality. Specifically, the genetic information can tell you about whether you need an agonist or an antagonist. Certainly it does give some information about the specifics of at what type of drug you want to develop. As I alluded to in my presentation, I think it really can offer situations where you can see indication expansions or risks that you want to discharge early on. With that, I'll probably just hand over to Joe. Yeah, just expanding on that a little bit. I think one thing we have to consider is the, you know, what kind of target is it? How is it druggable? Is it an extracellular protein that could be targeted with an antibody? Is it an intracellular enzyme that can be targeted with a small molecule inhibitor, or is it something else? When we think about that, it's really important to be able to look at the indication space that is implicated by the genetic hit. Because obviously, in certain indications, we can tolerate different routes of administration or frequency of dosing more than in other indications. For example, in a disease like asthma, we're not really going to be tolerating, you know, weekly IV infusions within the broader population, but in a more severe or acute indication, perhaps we would. We really have to think about the genetics are implicating a particular target. What is the biology of that target, and how does that influence our thinking, not just about what sort of modality are we going to use, but what sorts of pharmacology is going to be consistent with the treatment in a particular indication. Got it. Helpful. Just a couple more from me. I don't want to monopolize your time here. Do appreciate it. Getting specific on 23Me-00610. You mentioned that oncology or ovarian cancer specifically may be well suited for this. What in the data right now at this point would suggest that this is an amenable population? Joe, you wanna take that again, or Adam? Maybe Jennifer would be best suited to answer this one. Joe. Sure. Can you hear me now? Mm-hmm. Yeah. Ovarian cancer is not a particularly well treated disease at this point. Especially for more advanced disease, there is a lot of potential for new agents in this space. Immuno-oncology therapies in particular have not been particularly effective in this space. Having an immuno-oncology agent that could address this disease population would be really interesting and helpful for those patients. Okay. Last one, because I don't want the consumer folks to be left out here. Paul at Lemonaid, I'd love to get your thoughts on the evolution of the Lemonaid platform now that you're combined with 23andMe. Can you give us a sense of what percent of Lemonaid's revenue has historically been driven by preventative health versus just those seeking an easy access point to prescription drugs? Because it seems like much of the logic behind the acquisition is the ability to develop more personalized preventative care plans based on genetics. I'm trying to get a sense of how the Lemonaid platform will have to evolve versus where it has been historically. Thanks for-- Okay. Thanks for letting me jump in. We really believe the future of healthcare is around holistic healthcare solutions, and we talked a little bit about that in the presentation and that's always been our future direction at Lemonaid, and we are now able to accelerate that as part of the 23andMe family. We really think a lot about how do we combine genetics, wellness, actionable insight- Insight. ...all together and really provide good solutions to our patients. I think data becomes a large, a big important component of that as well. The more the patients tell us, the more that we're able to help patients, particularly in the areas of preventative care in a way that's truly personalized. That's definitely our strategic focus. That's what we are- Focus. ...building together, when we're combining 23andMe and Lemonaid, and we're excited to make that happen over the coming weeks, months, and years ahead. Maybe just pass to Anne in case you have any other thoughts. Yeah. Yeah, I think a couple things. One, it hasn't. You know, primary care online is really just starting to take off. It's, you know, I think when I think about pre-pandemic world- World. ... most of it was focused more on the reactionary, I need something right now, what can I do? Part of what I loved about Lemonaid, which I think is a real differentiator versus everyone else, is it's direct to consumer, and it's very hard to find another company that actually understands what it's like working directly with people. Second is Dr. Davis Liu comes from Kaiser Permanente- Kaiser Permanente. ...which I know quite well, and is really focused on, you know, on prevention and the holistic you, and how can you actually best manage an individual to stay out of the hospital and avoid the physician as much as possible. The combination of those two, of having direct access as well as with the foundation that Dr. Liu has really put together, is what made me feel confident that we have a platform that really can be adjusted and fit to add in genetics and really help people have a preventative-focused primary care service. That makes sense. Thank you, everyone. Great. Okay. We have a few more questions from the audience. First, next question is, the payment, the $50 million payment from GSK for the one year extension on the collaboration. The question is that reflective of the value of our data, and do you expect to be able to ramp up monetization of our database after the exclusive agreement expires? Want me to take that? Take that. That'd be great, Steve. Yeah. I mean, first of all, for background, remember, you know, this collaboration was signed back in the middle of 2018. The size of our database at that point was significantly smaller. This was kind of a look ahead several years as we were pricing this. But, you know, the second thing to understand is that that was quite a comprehensive multi-part deal. There was an equity component. It's highly unusual in these kinds of collaborations with really small biotech and big pharma to have all programs initially be 50/50 sharing as opposed to being just a straight up royalty and milestone deal. The flow of value within the deal, you know, really had multiple points, and I wouldn't draw any conclusions about the value of the data platform as we go through time strictly related to this $50 million. It's just part of a very comprehensive deal, and it's been really productive to have those partners. All right, the next question is about Lemonaid Health. Is Lemonaid Health pursuing relationships with primary care providers, insurers, or networks? Is Lemonaid Health limited in the data that they can gather in the primary care setting? Paul, you wanna take that? Yes. At the moment, Lemonaid Health is direct to consumer cash pay service connecting directly with patients, as Anne mentioned earlier. We think a lot about what are the partnerships and what are the tools that we can put in place so that we can help patients for more and more of their healthcare needs over time. That might be connections into fitness or wearable data so that we can truly understand more about the patient's behavior, so that we can start to help guide that patient to make better clinical decisions. Or that might be partnerships with others in the healthcare system and, we're excited to explore that over time. Okay. Next question is about data collection. What is your plan to scale up the collection of data? One would think that the company with the broadest, richest database would be the company that other drug companies would be interested in partnering w ith to find drug targets. I can take that, and then Adam, you can feel free to jump in a little bit. We think about additional data all the time, and we think about pulling in medical records. There's obviously all this wearable data you can pull in now, and some of it is gonna be very valuable and will absolutely be part of what Jeff Benton talked about in terms of the future of our health reports. A lot of it is also not nearly as valuable as self-report. I think it's one thing that 23andMe has really pioneered, is how we can have a way of collecting a lot of information that is incredibly high quality from huge numbers of people that is scalable- Scalable. ...that is self-reported information. Every time we think about an additional type of data offering, it's the bar is self-report. Adam, I don't know if you wanna talk about other additional types of data. Yeah, I'd love to. I really do wanna emphasize that point about self-report. As I talked about in my section, scale really is key- Yes, key. ... for the type of analyses that we do, and self-report is really one of the few data paradigms where you can really collect at the scale required. The other aspect is that you need your data to be structured in a way that it is actually useful. Often when you are collecting things from health records in such a way that those data are really difficult to work with and because they are generally unstructured, they're not collected for the purpose of integrating with genetics. You know, often we look at the data that's available out there, but the bar to utilizing that data might be extremely high. As Anne sort of alluded to, self-report is actually remarkably good for the type of analyses that we want to do. We do constantly explore where are the other opportunities. The next question is about the IND that was recently filed for 23Me-00610. The question is, what's the end game for this study, and will 23andMe end up manufacturing the drug? Jennifer, you wanna take that? I'd be happy to take that. For 23Me- 00610, that is the drug that we have manufactured for the purposes of the clinical trial. That was the drug that was filed with the IND in order to initiate the clinical trial and as we announced a couple weeks ago, the first dose has already been given. It is the intent of 23andMe Therapeutics to develop and commercialize drugs in the future, and so we will be really excited to tell you about the progress we're making towards that. Thank you. The next question is as follows: Understanding which patients would be best served by treatments may become an essential factor in designing treatment strategies. Is there an opportunity to do research and identify which alternative existing therapies for diseases are most appropriate for a specific patient based on their genetic profile and to provide that information to the clinician? Lemonaid seems to imply they are doing this, but there is research coming from. Where is the research coming from, and are the drug manufacturers willing to participate in bearing the cost of this research? If I understand the question correctly, a lot of this is around pharmacogenetics. Yes. For instance, there is a wide database of information. The FDA has a group called CPIC. There's something called PharmGKB which is out of Stanford, which is a database that is kept of all of the drug-gene interactions. There's well-known examples like Plavix, which is, there are genetic variations that show one individual is potentially, you know, some individuals are likely to not respond and might want a different therapeutic choice. That is, again, one of the key reasons why we were so interested and excited about Lemonaid Health is because they have pharmacy, and they have a team of pharmacists. They send out their own medications. Pharmacogenetics is something that has been around for a long time, but it is not well integrated into care right now outside of a few small examples. This is absolutely something that we think about delivering to our customers is being able to take the FDA-cleared FDA pharmacogenetic reports that we have, give them to our customers, which they are already receiving, and now help them actually apply that to their prescriptions and help them understand in key areas like depression whether or not they are taking a medication that is ideally suited to them. Thank you. Anyone wanna add anything? I think that was probably okay. The next question is about polygenic risk scores. The question is there a plan to further monetize this information? For example, business-to-business model, or would insurance carriers cover the cost of a subscription as it would ultimately lower cost for healthcare treatment? We definitely think about alternative models for polygenic risk scores, and I think that there's a lot of applicability potentially in clinical trials for health systems. There's a lot of ways, but I would say it's still early days right now in how polygenic risk scores are being used. It is something that you should see and expect that 23andMe will pioneer with, as we have customers getting it and also as we have a substantial research arm that is continuously researching, studying, and publishing on it. The next question is about survey data. Since it's fairly easy for us to collect, the question is, why not collect more data from around, all around the world? That's a great question. You know, launching and we are in, Wade, you can correct me what number of countries we're in, but we are in quite a few countries, not necessarily with the health product, but with our ancestry product. Each country has its own set of rules and regulations around research. Depending upon those regulations, we are either collecting data and engaging customers in research or not. It is you know part of it is just about the complexity of research regulatory rules and making sure that we are compliant with all of those in each country. The one thing that is spectacular about the U.S. is the diversity. It's one thing that we absolutely have this incredible privilege of taking advantage of, is that, you know, we have diversity all throughout the country and done calls to action for specific communities. We have really been able to engage those communities and collect a large amount of data. We have a substantial program right now with the African American community, looking at how we can actually engage that community more and more and actually collect a lot of data on that specific community. The next question is a two-part question. First of all, thank you, Anne, and everyone for an interesting presentation. Very insightful and concise. First part, in the beginning, Anne, you mentioned 1,000 data points a day can be collected. What could a data point example be, and how could this be useful? For example, will it be defined as a phenotype or a precisely defined case control or other usefulness? The second part is in relation to science and research. I happen to be involved in the research, a study of similar phenotypes to a previous 23andMe research paper. I would love to obtain summary statistics, or is there any chance that we can get in touch with someone from 23andMe Research in this regard? First question about the 1,000 phenotypes. We collect well over that number. Just in the spirit of being accurate, I know it's well over 1,000, but I don't know exactly the number that we are daily collecting and that can be something as simple as somebody telling us that they, you know, answering a survey question about are you a morning person or a night person, or it could be actually a wearable data point or it could be, we have some prototypes actually pulling in more medical record information. It could be one of those types of data points, but it's majority of individuals answering self-reported phenotypic data about themselves. Second, in terms of research, we actually do have a fabulous research program where we do a call to action for the community where we ask if you have a question and you're looking to engage with us on research. We do a number of those collaborations. Wade, I don't know if you remember the email address for that, but I'm sure if you email something probably on our website somewhere where under research it will say something about when we are having a call to action for people to submit proposals to us. Yes. You can check the website and, if you want further information, just shoot us an email. The next question is about Lemonaid Health integration and, the question is what will we see next? What will consumers see with that integration? Paul, you wanna take that? The first thing that we are building is genetics-based primary care and the pillars that deliver on that. The first things that consumers will see, and we are very excited to deliver on, is the components that will form part of that. Watch this space. Okay. I think we have time for one more question. It's a two-part question. One is what's the total addressable market for 23Me-00610, and is there any proof of efficacy for 23Me-00610 in ovarian cancer? In other words, you know, what gives us confidence that that'll work in that indication? Jennifer, you wanna take that? Sure. I'll be happy to take that. Obviously we're incredibly excited that we have, you know, just initiated our first in-human phase I study and this data's gonna take a little time to collect and based on that, we'll be making additional plans on how to move forward. I don't think at this point we can give projections to any of those answers. As Paul says, watch this space. We are really looking forward to being able to share data at a medical conference, hopefully in the not too distant future. Thanks, Jennifer. That wraps it up for our Q&A session. Turn it over to Anne to wrap it up. Thank you. Well, thanks to everyone, all of our speakers for presenting. Thanks to everybody for joining and tuning in. Like I said, I'm incredibly enthused. I think we really are just at the beginning of how customers are going to access, understand and benefit from the human genome and we're really just starting to scratch the surface of the power of this data set and how that's gonna translate to everybody in terms of how it's integrated into their care and the opportunities for really being able to live healthier lives, as well as we're just starting to see the tip of what's coming on therapeutics. My hope is, we will absolutely be able to have a very different type of world over the next decade as we prove out both of those on the consumer side and the therapeutic side about what's potential with genetics. With that, thank you very much. We look forward to engaging with you all in the near future.