Good morning, everyone, and thanks for being with us here today in the office in Boston. I really appreciate it. Everyone else in the webcast as well for our GE portfolio updates at Praxis. As you can see here at the beginning, the way we tend to think about everything at Praxis, there are four more. We will be there for more today as well. It is a great show here in a not-so-sunny day in Boston, and hopefully we are going to make things a little bit brighter and sunnier during the conversation. Steve and I are going to be doing most of the prepared remark conversation, but we have other members of the team here with us as well: Tim Kelly, our CFO, and Brian Spar, our head of the GE program, and many other—bless you—many other members of the Praxis team for us doing gals afterwards. The way we thought about this event is really mostly conversation about what we are doing, what the unmet need is, and the commercial opportunity across the portfolio. We are going to be structured that way. We decided to deserve plenty of space at the end for conversation and Q&A and discussions and how we see all of those programs. We are going to be making certain forward-looking statements. I strongly advise and recommend that you read the SEC filings, including our Form 10-Q that was posted this morning on the Edgar system. Before we jump into the contents, we thought it was very important to remind all of us here and elsewhere why we are doing this, right? It is not for us. It's for people that are living with those conditions and that need good treatments, better treatments, safe treatments, and that would impact themselves and their families and their loved ones and society at large. At Praxis, our mission is to do that multiple times and to keep doing that again and again until someone does better than us. That's a great thing. We're going to do again and again better as well. The focus is fundamentally on making CNS drug development and hopefully very soon as well commercial opportunities here. As I mentioned, Steve and I are going to be leading this conversation. I'm going to go a little bit about the opportunity in GE, what are GEs in general, I'll cover some of the commercial opportunities for our molecules. We are going to jump into the relutrigine program, particularly what we are seeing on the long-term extension within BOLD. There is a little bit of new data there and updates. Steve is going to talk a little bit about our EMERALD program as well and the rationale. There are updates on the current stage of development for Elsunersen, or what used to be called Praxis- 222, and SCN2A as well. What is probably the newest part of today is really our first foray into autism spectrum disorders with the PRAX-100 program. You might have heard about that before, but now we are at the points that are going to be giving a little bit more in terms of the plans in general for us. Probably it is obvious why neurology, right? Neurology is an incredibly large therapeutic area. Maybe one of the biggest uniqueness of neurology is this white space that exists. The white space, differently from other therapeutic areas, is mostly on areas that are fairly large. When you look into multiple indications, and of course, this by no means is exhaustive in terms of all the indication neurology, it's rare to see any therapeutic area where some of the largest indications are still the largest unmet needs. It makes it for an incredibly attractive market in general. It makes it even more exciting for us as a company because that's by chance and by design where we focus on. You can see here the focus, of course, as you know, in epilepsy, in autism spectrum disorders. Of course, it's still in essential tremor with the studies going on. Then adjacent to that, as we hinted last year, applications of some of those technologies in pain, in Parkinson's disease, and so on. Not only the current areas of focus, but the potential future area of focus are quite interesting from a commercial standpoint. We're going to go back to the entire portfolio at the end. You're fairly familiar with all of that with our platforms, Cerebrum, small molecule platform, and with Solidus, our anti-sense oligonucleotides platform. Today, we're going to be focusing on relutrigine primarily and on the ASO platform as well. Maybe before we jump much further ahead, there's been a lot of discussion. We're incredibly happy with the fact that GEs in general are like a topic of multiple conversations and a lot of interest in the industry, a lot more academic work, drug development, FDA, you name it, work here. There still seems to be more of an intent to divide than to unify GEs when some of those conversations happen. We thought it would be a good idea to bring them back to what they are, right? It's actually pretty simple. It's in the name. They have an epileptic condition with developments. Delays happen at infancy. That is really the hallmarks, right? Infancy, maybe in a liberal way, can go a little bit more sometimes into the teenage years, but really not beyond that. That is the universe. When you look into this universe, it's manifested by incredibly severe. I know all of you here in the room cover incredible number of diseases that are very devastating, pediatric particularly. When you look very closely to these, these are amongst the most devastated conditions anyone can present. Incredibly diverse, life-limiting, not only an impact from a comorbidity perspective, but reducing life quite significantly. Intractable seizures in the majority of the patients, not all, but in the majority of them. Maybe the saddest part so far that hopefully we're going to change, and you're all going to agree with us by the end of this conversation today, is that more and more effective and safe drugs are being developed. That is a huge change. Until not that many years ago, there was absolutely nothing. Now there is a small, timid, but growing number of treatments designed for GE. From our approach perspective, we're looking into two major ways to address the issue here. One is a very focused genetic-driven, like a one-to-one relationship with the ASO platform. There is a genetic defect, and we're addressing that with one of the genes, one of the ASOs to the genes, as Steve is going to discuss later today. Another one is really looking into the manifestation of the etiology that can be genetic or non-genetic, as you all know, can be familiar or non-familiar in most cases. It's incredibly exciting because you can really tackle both sides or that together or separately, as you're going to discuss in a minute as well. No matter how we cut it, the number of patients with GE today in the United States never end up being lower than 200,000. I think that's quite fascinating because, again, there is not a lot of other conditions with this relatively large unmet need and with just a very small part of them with current treatments in general. Quite importantly as well, because the majority of the conditions are either environmental or de novo mutations in certain genes, the overall reduction in rates that we see in other conditions like this does not apply here. When you look into very conservative estimates on the next 10 years as planning period, we're looking for very, very significant growth, like 50% growth on the patient population in general. When you compound that with the fact these patients should be living longer with better care, better treatments, should be identified earlier and therefore have the chance of intervention earlier as well, it becomes even more interesting in general. In a few minutes, Steve is going to discuss our EMERALD program as one of the key objectives today that we discussed where we landed there. One of the most exciting things for us is that EMERALD can really address all this like pie, right? It's the entire population. Now, patients are not seizing, of course, they're not going to be eligible, they're not necessarily the targets, but when you think about a large proportion of the overall markets is where relutrigine is going to be focusing on. To give like a ballpark of what a commercial opportunity would be, if you use what we would call modest assumptions here, right? If you look into not necessarily high price points and even relatively small, as I'm sure you're going to back calculate market penetration rates here, we're talking about a very, very significant market opportunity that can be derived from not only relutrigine, but others as well. I think the more the merrier as people develop other therapies to address this market as well. What might sound counterintuitive at first is when you look into Elsunersen, for example, right? It's a sliver of, it's a relatively small number, not a small for rare, but relatively small number here, about 5,000 patients eligible globally and by global in countries with well-established rare disease mechanisms to get access to patients. And using analogs that are, I would say, I'm sure you'd agree, fairly reasonable for this kind of size and condition, maybe even on the low end. The global opportunity here is about $1 billion and the U.S. only opportunity is about half of that. Quite interesting, feasible from a commercial perspective as well. The newest part to introduce today from a commercial opportunity is really PRAX-100. We've been working for a while on the understanding that a very unique gene, right? SCN2A calls for an NaV1.2 solvent channel and has this very interesting manifestation that when you have a lot of expression, gain of function, patients have uncontrollable seizures at birth. We know a lot what happens there. We are addressing that through relutrigine and Elsunersen. On the opposite direction, there's not enough, the haploinsufficient phenotype. There is a very profound and one of the largest, if not the largest, cause of genetic autism. Not all autism is genetic, as we all know, but there is a fairly large part that is. Using conservative numbers, once again, we're talking about at least 20,000 patients eligible and zero drugs in development. We know increasing the expression of a gene is significantly harder than reducing the expression using those technologies. I think you're going to be quite impressed with the work we're able to do here. This is the next frontier on this portfolio as well. We're going to remain quite focused on delivering on all this opportunity in general. In the short term, if I bring us back to what's coming on the next 12 months or so, for relutrigine, we are fast and furious about involved cohort two. We had some quite, in my humble view, phenomenal results last year. I think we're going to be impressed with the updates we're giving today as well. We right away switch to cohort two here. It's enrolling incredibly well. As you can see in our press release today, in our disclosures, we are guiding for no later than the first half of next year for the results. We are starting, as we speak, the EMERALD program. I'm going to discuss a little bit the design there and how we landed. That would allow for, if all goes well and if it's all approved, for an expansion of this market from the 2A and 8A into this broad 200,000, give or take, total number. When you move to Elsunersen here, the EMBRAVE program, as Steve is going to describe as well, has a number of parts. We learn different things from the very first one. We're learning other things from part A that is enrolling right now. We are starting as well imminently EMBRAVE 3, that is the registration program for the US. Lots going on. With that in mind and with this what I would call quite incredible opportunity in front of us, I'm going to hand over to Steve to talk a little bit about the details on some of those problems. Thanks, Marcio. What I'll do is talk a bit about the science and the development for relutrigine and then Elsunersen and then touch on the further pipeline for the Solidus or ASO program and highlighting the PRAX-100. Start from a sort of a fairly theoretical point of view. When you're faced with the challenge of how do you make a drug that can give you as much efficacy as you can possibly extract without bumping into tolerability? That's always the refrain you hear with current anti-epileptic drugs is that, well, we're always limited by toxicity and we're leaving a lot of efficacy on the table. Now, we're a precision medicine company and precision takes many different forms, whether it's genetic, whether it's precisely how you understand the neurophysiology and how you think about delivering medicines. In this example, what I'm trying to show is that in the blue line, there is what's healthy neuron. This is how you understand a neuron. You stimulate it, which is the bottom axis, and it gives a response. So input, output, stimulus, response, however you want to look at it. Generally, you follow that curve for the healthy neuron. In epilepsy, what happens? That neuron, for the same stimulation that normally underlies normal function that we need to operate every day, pushes into that epileptic range. That is something that all drugs are trying to stop from happening because a neuron has to fire and that underlies all the paroxysms and the seizure manifestation. An ideal drug would push that back down onto the normal curve. What does not happen, though, is most drugs today, or all drugs that do this, push to stop you getting into that epileptic zone, but also impair function at the lower side of that curve. That is where all the normal things that we do have to happen. The cognition, the movement, speaking, everything happens down in that zone. If you are impairing that, you are going to have tolerability concerns. The ideal drug would just take that red curve and plop it straight back onto the healthy neuron curve. If you're trying to do that, how do you think, and what are the levers that we've got for developing a drug against the sodium channel to enable that? If we look at relutrigine on the left compared to mainstays of anti-seizure medicine treatment, carbamazepine and lamotrigine, I'll just walk you through. There are four properties there shown with different symbols. The diamond is that property called tonic current, which is the standard operating mode that a sodium channel supports an action potential in a neuron that then supports normal bodily function. Totally critical to let that happen, leave that alone. If you start to interfere with the function of that, you start to get tolerability concerns. The other currents, persistent current, voltage-dependent current, use-dependent, shown in the black, the purple, and the square symbols, these are all the currents that sodium channels can support to make a neuron fire into high frequencies. That's what happens in epilepsy; those currents are what support that epileptic firing. The challenge for a drug is to, when a neuron starts to do that, selectively impair those so they don't get out of control, but leave the tonic current alone. That slide for relutrigine on the left shows the pharmacological separation we've been able to achieve between the tonic current and that triumvirate of persistent, voltage, and use-dependent current. No other drug on the market is able to separate those three from tonic to this extent. This is what this is the profile that you would need to support the previous graph that I showed you, where you block activity at the high level, you spare it at the low level. It has exactly the right pharmacological profile. This is the work of a lot of computational neurobiology, a lot of experimentation that we have done, and the field has done over the years, bringing that together. It really is the next step. We've probably all heard about use-dependent, state-dependent channel blockers. That's from 20 years ago, but this really is the next evolution of that to bring it to this level. What happens? What are the advantages if you achieve that with a mechanism? One really important advantage, and Marcio alluded to this, is that network excitability, hyperexcitability being at the core of all epilepsies, whether they're a DEE or anything. That is because everything has to funnel in the end through a sodium channel mechanism in order for a neuron to fire an action potential. It is unequivocal. You cannot fire an action potential without a sodium channel. Everyone, that is a fundamental property of neurophysiology. What that lets us take advantage of, irrespective of the upstream etiology, you now have a choke point where all that etiology has to flow through in order to manifest as hyperexcitability. It affects the neuron, the channel affects the neuron, and that affects the network. Just a quick clarification question. You mentioned a, I believe the word you used was triumvirate of sodium channel. It was tonic, it was excitability. The ones on the bottom left, the pathological excitability currents, persistent, voltage, and use-dependent. If you can draw a circle around that, that triumvirate separate from the tonic, exactly. Just to clarify, did you mention that PRAX-100 was also a sodium channel drug? We're going to get there. We'll get there. Opposite direction. There's some nuances to that that we can explain when we get there. As I was mentioning, irrespective of the upstream etiology or the upstream seizure driver, however you want to look at it, you pass through this choke point of sodium channels. If you can modulate the sodium channels, you're going to modulate the output of the neuron, you're going to modulate the emergence of pathological firing. Again, relutrigine is perfectly positioned to serve that role as a modulator of that choke point. That gives us the opportunity to say, we should be broad now. Because we've gone from a computational understanding to a pharmacological and biophysical profile for a drug that we think is going to meet that computational and theoretical understanding. Now this is where we first test it. What happens when you put relutrigine on animal models that are harboring different etiologies? On the left, it's the obvious stuff. It's a sodium channel etiology. This is where you would expect it to work in its best state because the sodium channels themselves are mutated and are not just driving the seizures, but they're the source of the pathology. That is why you will see our first study in bold, it asked that question of patients that have got gain of function mutations in 2A and 8A, what happens when you give them relutrigine? We saw the remarkable data come out from that last year. We will give you an update from the extension today showing you how that has panned out. The next step was, what happens in a potassium channel? This is no longer the same mechanism. It is not a broken sodium channel. It is a broken potassium channel that is now making that neuron fire higher. The sodium channel is participating in this process because that is what it does. There is an incredible synergy between activity in the neuron and what a sodium channel does. Remarkably, and consistent with everything we have been saying, you give relutrigine, you completely abolish the seizures in a potassium channel etiology. That was the real first acid test that is this going to work with different mechanisms. If you look at where else we have taken this, and this is really a little bit of something that is a bit of a disruption, I would say. You probably heard a million times from all the KOLs you've spoken to and things that you've read, you don't give a sodium channel blocker to a person with Dravet because the disorder is already due to an insufficiency of sodium channel function in the interneurons. The interneurons are the brakes, the excitatory neurons are the accelerator in the brain. Not enough braking. That is why it's a, why would you do something that impairs the braking further? There are cases where you do that with some of the sodium channel molecules that are on the market, and you can see exacerbation. This slide shows that in a zebrafish model, if you give relutrigine, you see essentially a complete reduction in seizure activity, all the way to zero. At a very, very potent 0.04 micromolar EC50. As I did mention in the previous slide, relutrigine is orders of magnitude more potent against sodium channels than anything else on the market. That's great because all the off-target stuff is also going to be limited because of that. Very, very potent. Now if you compare it to the molecules that are approved for Dravet, fenfluramine, most recently, it's as efficacious, but more potent than fenfluramine. You compare it to bexicaserin, which is the upcoming molecule that's now with Lundbeck, hugely more potent. In this model, it couldn't even achieve full efficacy. It really, really compares well with molecules that are out there for Dravet. This is a bit of a surprise. We think, again, it's another manifestation of the biophysical properties of relutrigine. Because an interneuron that's not firing hard enough is not going to be subject to interaction by the relutrigine because it's not the sort of state of activation of a sodium channel that's going to favor binding and inhibition. We think it's another remarkable tick against what we expect for the profile to look like from an in vivo perspective. Didn't the results here in the previous slide quite mind-blowing? A lot of KOLs that we spoke with at AES said that sodium channel inhibitors are contraindicated for Dravet because it's a disease of haploinsufficiency. You don't want to further reduce the activity of sodium channel, but you're showing efficacy here. Again, mind-blowing. In this model, what happens if you apply a standard regular sodium channel inhibitor? Does it get worse or better? We can show some of that data. Look, in many cases, you see exacerbation in Dravet models with standard. Which drives the contraindication. That has been shown in the literature. That's something that, the model doesn't just show benefit. You can see exacerbation in some of these models. We're happy to try and dig some data up to talk about that, John. Does that particular sodium channel in the interneurons not have a persistent current? It doesn't have a persistent current. The normal persistent current is driven by pathologies that cause gain of function. It's a property you see, a normal physiological property you see in the dendritic tree of normal sodium channels also have that, but not 1A. It doesn't have a big persistent current. That's why, and also then it's not available because it's not that active in those neurons. It's not available to interact with this drug as well. This was perhaps the most remarkable thing. When you look against 10 different disease models, and some of this is from the literature, and some of this is work that we've commissioned, and what you can see, look at Dravet first. There was the fish data that we showed you, SCN1 lab fish, the data we just showed you. A tool molecule, very similar pharmacology to Dravet, was also applied in the mouse model of Dravet. That independent validation of that efficacy, unexpected efficacy in Dravet, again, telling you about this biophysical profile, how it works. 2A, 8A I showed you, but a whole range of potassium channel etiologies, full efficacy. A hyperpolarization activated cation channel, HCN, also efficacy in this model. Because that whole idea of the choke point being the sodium channel, you have to manifest through that in order to have a seizure. We are modulating that behavior. None of these things can get through that choke point when relutrigine is around. It might be interesting to say, right, because HCN1, the very last model here, we cannot find any other drug that has this property. It's not like we know there are approved drugs for Dravet. We know that other mechanisms work for Dravet. We know there are ways to modulate potassium channels, right, and certain potassium channels. There is no way to modulate that pathology. We used to have a table that has multiple ways. That is the one that is red, meaning pro-convulsive in any other molecule. It really reinforces no proof that things are going to work across every possible model, but it definitely maybe is one of the most compelling cases we have to really understand how this drug can benefit patients globally. What's the profile of the relutrigine modulation that prevents, I guess, this is the interneuron activation that's the theory. Interneuron inhibition. Inhibition, yep. Further inhibition in Dravet. I'm off a red eye, forgive me. It's good. What aspect of that modulation allows for that when the other compounds don't have? Yeah, so most compounds just bind the sodium channel and block it. They don't care what state it's in. They don't care that it's in a resting state or an active state. They interact with it. Because they do that, essentially, that sodium channel is out of action all the time. If you can stop that from happening, it only interacts with the sodium channel when it's in that pathological state. Yeah, when it's supporting hyperexcitability, right. There's no other drug that you believe modulates only the hyperexcitable state? There's some that will modulate a little bit of persistent current, but none of them that take all of those three. By taking all three, you're blocking all of the drivers of pathology. Yeah. They're anatomically not on the same place in the brain, right? I think that's on the neurons. I think that's important as well. Maybe a little disservice that we call them like 1A, 2A, 3. It feels like they're anatomically on the same place, but they're not, right? Maybe you can help. 2A is only in excitatory neurons, and 1A is only in inhibitory neurons. That is why Dravet is a disease of interneurons, and 2A is a disease of pyramidal neurons. They do anatomically separate and drive different behaviors in those. Our Dravet relutrigine will interact with all of them under the right circumstances, but it is the circumstances, it is the manner in which you interact that matters so much and that differentiates us. Maybe what was sort of the pharmacology with which you were able to achieve that? Maybe give us some color on that. You showed us a lot of different models, and I understand some of it is literature, some of it is work that you have done. Is the sort of impact the same across the different models that you are seeing, the effect and kind of the safety? The safety profile is the safety profile. When you give the same concentration of the drug to the different models, there is not a different safety profile that emerges from that. The inhibition is universal across all of these, that you can have either stop or have a massive effect on the seizures in these animals. The pharmacology, when you do discovery on sodium channels, you can separate those tonic current, the use dependent, the persistent, the voltage dependent from a protocol perspective when you are manipulating the channels in a cell. You can isolate them, and then you can look at the specific pharmacology against each of them individually and in a separate fashion. You can say, okay, we profile this drug. We see ones that hit these three and spare that. It is very easy to do that. There are multiple ways of achieving that with patch clamp electrophysiology in the assay. Is it fair to say the binding site is what's unique? 100%. Certain binding sites become available. That's either a genetically determined mutation that causes that binding site to be more prevalent, or the temporary state of the channel has that. Exactly. The mutation might make the channel more active, therefore more exposing that binding. It's not necessarily at the binding site where the mutation is, but it makes that site more prevalent because the channel's doing more and it's making the whole neuron and the voltage changes. These channels are exquisitely sensitive to membrane voltage. That's what causes them to change shape, to open, to let sodium in, but also to then expose novel binding sites that drugs like relutrigine can interact with. You can favor that binding site over the binding sites that are available when the channel's just sitting there minding its own business. Is it allosteric in that it's changing the conformation of the channel back to normal, or is it just binding and then able to bind and inhibit? Binding and inhibiting. I think it's your question, right? The binding dynamics are quite important as well. There is a K-on and off ratio that we believe to be ideal. Are we at ideal? We don't know, right? By definition, we'll have to test all possible combinations. What we know right now is that binding very fast is important. Staying there for a while, but not too long, is important as well. A few years back, we showed the profile, the binding profile of this molecule. People normally do not pay a lot of attention to binding profiles. That is incredibly important as well. When that site is exposed, binding fast and inhibiting is maybe the reason why other drugs miss the mark and are not able to help here because it is incredibly fast that it binds to the sites. Is the K-off rate important or normal here? Say it again, sorry. Is the K-off, which is that you un? It is important, yeah. Yeah. That ratio is very important. That gives you affinity. Does it just come off once the current is gone, or is the off independent? The off is independent of the on, is dependent on the amount of drug and the binding site. The off is independent of that, and it just falls off. The current itself, like once the current is gone, the confirmation changes. Does the drug come off, or does the drug? No, it comes off because the residing time on that binding site is governed by a statistical process, and it just drops off. Okay, so it's independent of the binding. That's called, yeah, called a Markov process, how it drops off. Any comment at all on how relutrigine differs structurally and from a binding profile from matrigine? We have not, but I will, Laura. When we first designed, so relutrigine came first, right? They are from the same series. You're going to be able to find their structures right now. They're both published, so we might as well bite the bullet. They're incredibly similar. And to the point that I think most untrained eyes would say these molecules are going to do something very, very, very similar. And they do from certain properties, right? Maybe the oversimplification is those channels are incredibly sensitive to small changes on the, what I'm going to call the electric fields around the molecule, right? I think medicinal chemists would call that something a little bit different. Just that one nitrogen change between those two molecules significantly changed that interaction and the metabolism of these drugs. It's completely unexpected. The beauty of that, not to expand way beyond the scope of today's conversation, is that there are many more of these molecules that we have patented. You can imagine certain molecules that would be having a better either primary or secondary pharmacology for other neurological conditions that can use that property as well. Neurons are not made to fire and create seizures, right? We have all of them for other functions as well. Modulating this fundamental process with molecules is definitely incredibly important. We know that sodium channel blockers are used for mood stabilization, for pain, for other indications that are not necessarily today's discussion, but you can imagine how even understanding those small subtleties between the interactions can be important for future discussions and indications as well. Maybe a related question. How would you modulate, or how are they different? Is it the binding site that's different, or is it sort of the K-on K-off ratio? The clinical pharmacology is different as well. The primary pharmacology is relatively similar, meaning the way by which they inhibit different occurrence, but not the amount of drug, the persistence of the drug. One of the concepts we discuss quite often is when you look into the half-life, for example, of each one of those molecules, they're incredibly different. There are molecules you want shorter half-life. There are molecules you want longer half-life. Once we discussed the profile for relutrigine, it was important to be that long, to be as long as we possibly could, because there are tons of intercurrent events that happen with patients with DEEs. We want to make sure that when they take the drug, it doesn't matter for the effect of the drug. Like we know, for example, there are cyclic vomiting in a lot of these patients, right? So independent of when they are taking, it goes through a lot of vomiting one day, several days without, a lot again, no one knows exactly why it happens. On some of the GE patients, actually very common manifestation. They get a lot of infections, so they might need to actually treat the infection before. Imagine if every time that happens, the drug concentration goes to zero. Now the seizures go all the way to azu. It's terrible for these patients. It might be one of the issues why there are uncontrollable seizures with other medications. We wanted to make sure that small changes don't affect fundamentally the ability of the drug to control the channel. We were able to do that with relutrigine. It's less important, too, important to Vormatrigine. I think when you're looking to these drugs, right, we spend a lot of time on the mechanism and whether or not they work, but then real people like us end up using those drugs, and we have real lives with real complications and intercurrent events and so on. We try to match those drugs on the clinical pharmacology, not necessarily the primary pharmacology. The K-on-off is slightly different. I'm more than happy to post this afterwards as well for those molecules as well. I believe Yas has a question as well. Who is on Zoom? Or maybe not. She's coming in. We didn't design an interruptive, but it works. Not now. Go ahead, Yas. Oh, so sorry. I wish I was there with you in person. Thank you for taking the questions. On this really amazing pre-clinical table that you guys put together, I assume, I guess, have you also looked in the predictability of these broad DEE models, the translation to human data? Could one say that the translation of one pre-clinical model is stronger than another? I guess you probably will talk later about it, but I assume you dose in the pre-clinical model sort of uniformly. I guess the question becomes, how do we think about dosing and figure out the optimal as you contemplate? In the pre-clinicals, I guess, did you use the same doses or different doses based on your modeling, and therefore it could warrant different doses as you open it up into broader indications? The fundamental question on the overall dose at the end, right, for all patients, I think we have to be incredibly honest that it's going to be an open question for a while, right? We are taking the approach that we have a dose that is effective or we expect to be effective for all those patients, knowing full well that some of them might need more and some of them might need less. When you look into the predictability, to your first question, of the exposures on those models and correcting for protein binding and properties that humans have from the animals, they are all within the therapeutic dose that we have been giving to kids and adults on the EMERALD study. There is none of them that require a lot less or a lot more. Once again, we are all different. Some of us have full response at full dose, and some of us do not have for any approved drug. It is not a question as much for the clinical development, what things should be standards, but a lot more for after-market. I can tell you without giving much details that we receive a fair bit of emergency requests for this drug because of, and some of them are on certain conditions that would not expect effects. Some of the things we're saying here are a little bit rhetorical because we actually know it already has effects in some of those conditions in humans, and the doses are very similar to the ones we are using right now. I'll stop there and I'll not entertain questions on that specific topic. Thanks. Thank you. Doug? Marcio, thanks for taking the questions. Similar to Yas, sorry I can't be there. I'm just curious, in terms of this process of identifying unique binding sites and binding sites that are sort of available only when you want to interact with them, in terms of your process for identifying those, is that something sort of proprietary to the company, or is it just sort of like an orthogonal method that in sort of knowing how and when to sort of put together somewhat sort of available information and work that's been done broadly in the community? I'll let Steve comment on most of this, but maybe because he's not going to do himself, because he's a very humble man. The most complex interaction of neuron system in the world was developed at Steve's lab when he was a professor at the University of Melbourne. We leverage a lot of that work to understand maybe a fallacy of a lot of those models is we clamp one of the cells, as you know, and you look into the behavior on that given system. Our brain is not like that. There is not a single neuron or a single part of a neuron firing. There is a system. What we do not know completely, if that is the full secret sauce, is definitively a lot of why we are here today. Steve, go for it. I'll just talk a little bit of it. It's a way of bringing a computer and a neuron together, a computer and a cell together with a channel, like a sodium channel in it. You can produce a hybrid thing that in real time, you can investigate the way drugs and the behavior of the system works. It's called dynamic clamp, and it really lets us get incredible insights into how not just the neuron affects the sodium channel car, but how it would affect the behavior of an actual cell, of a real neuron. So how the drug affects an actual hybrid neuron. That really compresses the time it takes to understand mechanisms, to have predictions around what might happen to a real neuron. Using pretty much the same method, you just throw a computer, a real-time computer in the loop, and now you've got a virtual neuron to play with. It It normally would require a mouse model and slow data. That method has been very helpful in exposing some of the fundamental properties of the system. But I guess as a follow-up, Steve, does that help you identify the specific binding sites, or is it just simply the sort of neuron functionality and the firing that you're looking at? Yeah, so without going into too much detail, Doug, you don't necessarily have to know the binding site. You need to understand what state you're moving the channel and the neuron into, and then look at the sensitivity of that state to the drug. Because that's what you really want. If you might know the binding site, it doesn't prove that you're going to be able to have the interaction you need. This is the downstream step. Okay, we want channels that are being activated too much to be inhibited. So we can directly look at that functional consequence without under. We can go back and deconvolve and say, you know what part of the channel is probably going to be exposed here, and then start to look at where it's binding. That's not necessarily knowledge you need in order to first characterize and profile the pharmacological and biophysical profile you want. If I can have just one follow-up, when you think about other indications, perhaps even those outside of epilepsy, would this also be a situation where you'd be looking at somewhat differentiated binding sites and sort of neuron firing than what you're seeing right now in epilepsy models? Nice try, Doug. We'll be there soon. Fair enough. I got to try. Oh, yes. I'm never going to punish you for trying. Maybe a very quick related question. What you've patented is the clamping method, not the binding site, or both? No. The actual methods, you can go ahead and buy supercomputers and build the system, and good luck to you. It's published, actually. Steve published a lot of that for other years, not the final system. The ability to actually interpret this, it's so computationally intensive that until recently, it was actually very difficult to completely integrate it. Now it's a little bit easier. There's much more computation. But knowing what to do with it. For a while, this shall pass as any understanding do, and we're going to figure out, others are going to figure out. Right now, the integration of how it works, how to de-risk, how to de-risk the tox later on, it's something that we have a significant amount of internal knowledge just by confidentiality and internal knowledge about how we designed some of those drugs. There's a lot of luck here as well. I mean, you know me, I'm not going to just say, oh, it's just like a bunch of people that can come up with anything. That is a lot of luck. The original idea for some of these molecules did not come through this. Came from observing the environment, what happens, and then, oh, can we design something? And we didn't know. Here we are. May I have a quick question? Of course. What is the chance to rationally design state-specific activation inhibitors, inhibitors of hyperexcitable? We would argue this is rational because we are precisely isolating that process and looking at its pharmacology. If you're saying rational from the perspective of a protein perspective, with some of the new AI tools, I think that's in the future, might hold that because we can model the binding of this drug, dock it into a model to molecular dynamic simulations and ask it. I think that's in the future. Right now, I think that's things that we're looking at for not deploying it right now because we've got this ability to sort of not have to worry about the protein interactions because we can look at the functional interactions first. Is there any structure of a hyperexcitable sodium channel out there to be also docking? There are structures of sodium channels on which you can impose a mutation and then say, what would the conformational change be, and then use that. Obviously, this is a lot of our thinking. We're obviously on top of all that literature on how to deploy. We'll deploy at the right time. We can make in silico predictions of the change, if that's the question. Yeah. Brain penetrance? Through other methods, but yes. All right. I love the questions, but. All right. Leave it in. This is just a snapshot of where we're at with relutrigine. We talked about how well-positioned the in vitro, the biophysical, the in vivo properties are. Now we're going to start talking about what happens when you start to put this in people. It's a bit of a highlight, and I'll talk in a little more detail on them. First thing is we've got robust seizure reduction in our first trial, one of the EMERALD trials that ran out last year, and an incredible increase in seizure-free days. That'll be clear. We're looking at the safety in our healthy volunteer studies. What was really remarkable with this, and again, another sign that this profile we've got is actually meaningful, is that we couldn't find a maximum tolerated dose. That's something that you can find with a sodium channel molecule quite easily. We could dose these healthy volunteers up, no MTD, at fold levels higher than we would think is necessary for therapeutic coverage. DEEs were all mild to moderate, so it was a really good profile. Now, if we look at the trial design, it was really, I think a lot of the trials we're going to look at, probably the best word to describe them without being too proud of what we've done is just elegant. They're very simple, but they're highly informative, and they're very, very efficient in pulling the signal out from these patients. This was another design, obviously randomized to the placebo arm or the drug arm. The placebo arm was a very novel design where each of them is four-week periods. Anywhere within one of those four-week periods, you could have received a placebo period. It was blinded to the investigators. It was complete double-blind on that when you got it, which we think contributed to the very favorable placebo rate that we saw here because no one had any idea. If you were on placebo for three or four months, you might guess it. If you've gone, well, I could be on it, I may not be on it, you removed a big element of the human bias that influences, we think, some of these things. Patients were able to go on to an open-label extension. This is perhaps the most remarkable result. You probably recall from last year where we talked about the top-line data from the EMBOLD trial where, and these are the periods, baseline at zero each month. At the end of the fourth period, the fourth month there, that was what we reported, about a 47% reduction in seizures. This is now looking at the open-label extension, and there we are assessing every three months. What you can see is the continual reduction in seizures. Absolutely remarkable. At 11 months now, we are seeing around a 90% reduction in seizures. Bearing in mind that these patients at baseline were already on the most number of drugs that the treating physician dared put them on. They had gone through all the trials and put them on this. We put this drug, 70% of these patients were on sodium channel drugs. We put this on top of that and didn't see any new serious AE events emerge. Again, is that another sign that the profile of this drug, only acting when you need it to, it's not inhibiting and contributing to the pathology that you'd expect if you put on a phenytoin or something? Do you know how many were on how many drugs they were on at baseline? Yes. There were an average of three. There was one patient without any drug at baseline, incredibly unusual. Everyone else was in more. We consider that one quite important as well. While it's an N of one, I think it's important because they had tried every drug available. They could not tolerate any. She's the only receiving a drug today and responding really well. I think that it goes back to every piece of evidence helps us understanding the profile of this drug. It gave us enough confidence to design the next trial with even more liberties for these patients. Because one of the key, I'll argue, pushback we got from particularly families is the restriction of having one sodium channel blocker at baseline. 70% of them were in one, as you can see here on the right-hand side. We did not allow for a stacking of multiples. I think it pays to be safe at the first time we were in pediatric patients, right? Most of these patients are not mobile. They have several seizures per day, as you might recall. We tried an average of three, but they were only on. They were on three. The average is like, for most patients, Brian and I was just reviewing some cases actually this morning that came in on emergency requests, and they tried, I think, 10 drugs. The majority of these patients, to the point that they got to this trial, they tried everything. The other points that we got asked questions before, it's not on the slide, but I think it's worth mentioning, is were they really at therapeutic levels? We measure therapeutic levels of every ASM during baseline. And no, they were not. They were at supratherapeutic levels. They were actually higher at almost toxic therapeutic levels. If something were to happen in this study, we can only talk about this study or reporting, I think there was a high probability that if our hypothesis was wrong, we would have seen it. Quite the opposite, right? I think what you're seeing here is an incredible seizure reduction that deepens, as you stated, the drug, as it makes sense. Why? Why should it deepen over such a long time period? Why shouldn't you see the maximum benefit within a month? From a neuronal perspective, you've all heard about plasticity, homeostatic plasticity. The systems change, and there's the immediate effect. Then having a lower firing rate has a longer-term effect, and as the system completely resets its function. You see that in many other states. There is a strong acute effect, but continual exposure starts to actually really do a new set point. That does not happen because you first put a drug on, the neuron wants to go back to where it was, because that's what homeostatic plasticity does. I'm firing at this high rate, even though you pull me down, I'm going to change other things. Eventually, everything comes down to a new set point. We're seeing, I think, a really good manifestation of that in this data. Yeah. This might help as well. On the next slide here, we showed another version of this before where we showed the baseline and the controlled periods. When you look by months of exposure, what we are seeing is not only the seizures are reducing, but there is spacing. You might still have one seizure, but now instead of every day or every three days maximum, it now takes two months for one seizure to reemerge, or three months, or four months, or 10 months, whatever it is. It might be now, we don't fully understand as humans, scientists, why those seizures show up. I think we all should be incredibly humble as well that we don't know the key for any of this, right? We can observe a phenomenon as we're observing. This phenomenon is quite positive. We don't completely understand why it resolved. There is an increase in concentration as well of the drug. There are, for many patients, a reduction on their background therapy. It might be, it's not an unreasonable hypothesis, not one that we aim to study here. I can't tell you definitively that what was causing some of this was actually those non-specific drugs as well. They might reduce to a level, but they might create other issues in the brain. We just don't know, but it might be a synergistic combination here about reducing a little bit of toxic drug and increasing the concentration of one that is not. When you say increasing exposure, do you mean in the CSF in the brain or the serum or both? Yeah, both. All the patients started at half a milligram per kilogram per day in this study. There was an opportunity just once during this study to go up to one. In the open label, after month four, there was an opportunity to go up to one again if you haven't. What we can tell you is that at this data, the majority of them got to those exposures. Patients have different clearance, right? We all have different clearance to different drugs. Some of them can get to one milligram exposures by staying at half. That's why we always like to talk about exposure versus dose because exposures have a lot of variability around them. We do see a very strong correlation between exposures around one milligram or target one milligram and resolution of seizures and reduction of seizures. It is a multi-system, like multi-factor system. Isolating one parameter is not going to be the solutio n here. All of this might be contributing. We've nerded out on technical reasons. If you look at the clinical phenomenology, one of the earliest things that people learn in clinical epileptology is seizures beget seizures. It is a really old saying in the field that you start having seizures, you start having more seizures. You look at the wording for the definition of a DEE, it is described where the seizures themselves contribute to the pathology. That does not happen overnight. That's a slow feedback that makes the seizures worse and worse. Now, when you're unwinding that, it's going to be unwound on a similar timeframe. That's the phenomenology. I think Marcio has very elegantly told us about some of the underlying reasons why that might happen. Absolutely. That's, I think, part of the response to your question. One other aspect just to add in too about relutrigine's profile is there's no titration required. It's an oral pill or it can also be given through the G-tube, J-tube once daily. Very, very easy, relatively, to some other drugs out there to give in the dose. I think the second part too, and it'll lead into the next slide also, it's not just the patients who see the benefit of these extended period without seizures. There's an entire caregiver ecosystem as well who's supporting these patients and the children who also see the impact of fewer seizures, less severe seizures also. Yeah. Doug, I think you have a question. Yes, thank you, Marcio. Steve, I think you sort of described this reset of sort of what is needed for sort of like the firing for an epilepsy. I think in the literature, they sort of refer to this kindling effect. I'm curious, it sounds like you would characterize this as maybe like a de-kindling effect that you're seeing with relutrigine. I'm curious, is this manifested? Does it change in terms of sort of the confirmation of the neurons? When you de-kindle, reset, whatever the terminology ends up being for this phenomenon, you do change things. You change what things express in a neuron. You change the amount of potassium channels and other things that express. That homeostatic response that keeps neurons firing at one rate is because of that. If you change it, you're going to reduce the things that express. That's how you change the set point, Doug. It's not necessarily how it could be structural parameters, the amount of dendrites and the size of the field. Other things can change. This is a whole area of investigation. We won't nerd out on the science. We do know how this can happen. I guess as a quick follow-up, though, that as you change the set point, would that imply to get further gains that perhaps you might need a different drug with a different binding site and availability? As most disease, there are no silver bullets. I think it would serve us really well when you are at about 90% or whatever, 90% seizure reduction to be very arrogant right now and to say, "That's it. That's the be-all for these patients," right? I think one must remain fairly humble about this. There might be—and we know it's the only standard of care—how much is contributing each of those drugs. I think the important point here is those patients are completely uncontrolled. Let me remind you, this was the most severe ever DEE study conducted. There were 50 seizures per month at baseline. Most Dravet, and it's not a jingle on the Dravet, it's a very important condition. It's five times less than that, three times less than that. We already started at an incredibly high bar. Seeing this is incredibly encouraging. Is that going to be like that for every condition? I think to your question before, likely not, right? We're likely going to see variability throughout. I think at this point, it should be encouraging. Maybe you can move a little bit. I know Candace has a question because we're trying to get to the end. We're probably not going to be able to stop at 11:30. I love the dynamic. We're going to stick around. I appreciate it. FDA Commissioner has supported expediting drugs for rare diseases publicly. That's granting conditional approval based on an RCT, a little bit vague, but post-marketing as well. Do you have any plans to explore conditional approval for relutrigine SCN2A and 8A based on the Cohort 1 data, which appear quite strong? The base of dosing for Cohort 2 is fairly strong. We believe there might be a point in time in the near future that makes sense for us to ask that question more formally to the agency and try to understand the overall available data at that point. I think right now, head down, executing, gets the 80 patients enrolled on that study and submit as quickly as we possibly can. I'm sure it did not get lost because I repeat that all the time. We're talking about no later than the first half of next year, which means we are really working hard for that, not to be as late as the first half. If there is an opportunity for a subpart H submission under the current federal codes, we're absolutely going to explore here. I think right now it's a little bit premature with 15 patients or 16 patients of data. Marcio. Yeah. Can you help us understand the patient disposition once you finish the 16 weeks? How many opted to be participating in open-label expansion? Yeah. Thirteen of those entered the long-term. One patient was, I'm going to call, it's not really lifestyle, but I can't find a better word right now. They decided not to transition for other reasons. One patient, unfortunately, had a significant infection and could not, and is actually no longer with us. And we have in this data set, twelve patients are being represented. From month five onward, the number of patients are consistent. Twelve. Yet the seizure rates reduced. So it's not just because some severe patients dropped out that you got better. Oh, no, no. This is not a survival effect, if that's the question, which is incredibly common. You might be thinking about another AES poster. Look at how everyone knows what I'm talking about. Sorry, it's supposed to keep this precaution, right? I'm going to go there. We all know us. It's Friday. Yes, it's a good point. No, no, this is not driven by survival effects, statistically speaking. All right. It's just as though the final, if you're looking at other functional domain, this is really, I found this really exciting because if you are going to do disease modification, because we know stopping some of the seizures in these DEE patients is very important, you're going to see improvements in other functional areas. These are some of the ones that are very important to the parents and caregivers, obviously. Disruptive behavior, communication, alertness, getting the kids' gaze is really important. You see 60%-70% great concordance between the clinician and the caregiver here, which gives us a little more credibility, I think, in believing this data. I think just one final thing here, the seizure severity. Even if you're going down 90%, each of the individual seizures that are now remaining are much less severe than they were before. That looks binary when you just look at the counts. When you start to look at the gray levels and the sort of analog side of this, the seizures have gone down, which you probably would exp ect. Did you do CGIS then? We did as well, but the comparison here was on the improvements that was based on both of them. What I can tell you about, it's not on the data set, but it's a very, very high concordance on severity as well. Now let's just turn to Cohort 2. We had a really good baseline start within bold. What's the next step for this project? It is to do a pivotal study, obviously. This is the design, very, very similar design. What we have done is we can start at a higher dose. It was important in the other trial. The reason we're confident in this, not one dose reduced in the previous trial. That's one thing that you might do. You can start at 1 mg. Same innovative design on the placebo. We think that'll be great for controlling placebo rate, a good balance. We know the drug's effective. It allows us to have some good ethical perspectives as well. We want to make sure that these kids are treated. That is important. We do care a lot about patients at Praxis. This trial is currently underway. It's going to be measured in the same way as we did the other one. We think it's going to make an obvious way that we can get move-to-move registration as efficiently as possible. The duration is good. We know that we can see effects in this time period. We're poised now for registration. You need to have a certain mix of 2A versus 8A in the study. We do not. Yeah, we do not. We're attempting to have as many patients with each one. I think there's a very good understanding from the scientific community and the regulators that that is not necessarily a requirement. When thinking about if, for example, the placebo is the second four weeks, how many weeks of those is the drug still on board? Yeah. Should we just assume it's going to be the first four weeks so that they're actually placebo? You're free to assume whatever. We're going to stick to the design and the disclosure. I like how you said about the offering. Yeah. Maybe I'll answer your question about the carryover effects, right? Of course, we've had five days half-life. I would be speaking both sides of my mouth if I say there's no carryover effect. It must be, right? What would that do is maybe the question that we have to—and what that would do is just reduce the ability to compare drug and placebo negatively to Praxis. If you believe that that's what happened, you should be even more impressed about the results that we had. That is an obvious placement for this. I think we talked about this before. There's an obvious way to look into this data. I encourage you to think about what would be the most obvious here. Because it worked incredibly well, we thought it would be advised to actually change the design. I think we hear again and again from all of you and from our shareholders, try to keep things as stable as possible, especially when you are winning, when things are going well. That is what we attempted to do at this study. The second part, and I can't reinforce that enough, we all worked with a lot of incredibly severe conditions before. These patients are not in great shape. Keeping them on several months of placebo was not an option. It's not that it's like, "Oh, we thought about this because we want a clever design or something like this." Yes, that was part of it. The most important is they decompensate. They die. They have a series of infections that leads to aspiration pneumonia, pseudo-death. That was just not okay, right? There is an ethical component of this. When you consider all of that, what is the most efficient way? If you think a little bit, you're going to figure out which is the periods. The beauty is it didn't. When you actually ask, "Can you guess which period you were in?" people cannot guess which period they were in because there is no expectation that you could always be in. We know very, very little about the psychology of placebo effects, number one. Two, I think it's oxymoronic to think that a kid that is seizing several times a day, a parent would be just like, "I'm not going to count." Compliance to diary was through the roof. They feel every single day. These people are incredibly compliant. Human recall is—we don't remember anything, right? That's what the literature on recall shows. So variance on severity of these seizures such that if there is a placebo, could it be exerted on a parent's judgment of what is a seizure and what is not? Yeah. We're only showing countable motor seizures, right? So the patient described the seizure. All epilepsy studies do the same way. You describe the seizures, then there's a shorthand definition, right? Meaning a simpler way. They put that on the diary. What we see in the diary is your description of the seizure. What's on the back end is the classification of the seizure by the PI and in most studies by the Epilepsy Study Consortium, including this study. So the patients never know if they're counting a motor or no motor. Oh, okay. Maybe that helps with the question. They're just counting what they are seeing. You might see a lot of absent seizures. For example, most absents are no motor, as you all know. That wouldn't be counted because he wants to see reduction in general seizures. He always wants to do that. Now, 2A and 8A patients, virtually all seizures are motor. They're very severe. In general, in epilepsy, you have different types of seizures that are count. I have a hard time with placebo rates, as you can see in epilepsy, because seizures are real manifestations of—and they're very severe, and they lead to a lot of other consequences. The final really comes back to everything we started saying at the beginning. Because of the properties of this drug, it's incredibly well-positioned to serve the broader DEE market. That is what EMERALD is all about. This is—we have great biological rationale that we went over. We have clinical proof of concept in the 2A and 8A population of phenomenal reductions in seizures, phenomenal increases in seizure-free days, improvements on a whole bunch of other functional domains. Everything is really well-positioned. From a patient and a commercial perspective, a huge, large unmet need where this drug could be used and deployed for a really broad range of these DEEs. There is a market of 200,000, no matter how we count it, as Marcio said, in the U.S. alone. Before I move there, I think she has a question as well. Thanks, Marcio. I guess, given that you saw in cohort one, percentage of patients already are titrating to their lower background therapy. Did you pick that up at the 16 weeks? I'm just trying to think about if you could, in cohort two, potentially also show that because that's going to be really important for physicians as well, not only seizure reduction, but sort of down titration. If you could just remind us of that and how you're thinking about this. Yeah, we do. It's fairly individual. Yeah. After a number of months on drug, the physicians can assess if they have a tolerability issue only. They cannot simply reduce the dose arbitrarily because that could create other issues. There's a trigger. Let's say, "Oh, a few of these patients are a little bit more somnolent," for example, and then they reduce the background meds. It tends to resolve. It tends to be a one-to-one relationship. You reduce seizures, don't increase, and the side effects decrease. We are capturing all of this. It's going to be all on the safe tables or reporting to the FDA, all of this, as you can imagine. Because it does not happen for all patients every single day, it is slightly harder to make in trials this small as an endpoint. I think as we dose more and more patients, that's going to become a quite important feature, right? At the end of the day, and we say this with all cautious, this drug should become monotherapy. The overall promise of this drug and relutrigine, I know it's not relutrigine day today, but is that we would be able to get as close to monotherapy as possible in some of those patients. Not the time to explore. There is no regulatory framework in the United States to do a study like that. You probably know you cannot run a monotherapy study in the United States in epilepsy patients. Rightly so, by the way. It is certainly something we're going to continue to explore in the future. Again, this is just another example of a very simple, elegant, and efficient trial to how to answer this question. It's straightforward. It's one-to-one randomization. You either go to the drug arm for 16 weeks or you go to the full placebo arm for 16 weeks, around 160 patients. Really, really important, these patients will be phenotypically defined, the DEE patients. That's how DEEs were done until 10 years ago when we didn't know anything about the genetics. Everything was phenotypically defined. That's how LGS trials are still run to this day because we have no idea of the number of etiologies that give rise to LGS. A couple of them are genetic, but the majority of them, we don't really know. That is very important because this is how most physicians come and see their patients that have got DEEs. That is because of the way we set the drug up, that ability for it to be downstream of the etiology, it is really a natural consequence of it. Within this trial, the drug, as Tim said, there is no titration, once-a-day dosing, pretty much unheard of for a lot of these types of medicines. I'm sure you know about a lot of the drugs in development that have two or three times-a-day dosing requirement. All of these things matter for the conduct of the trial, for the attractiveness of the trial, and then for the long-term benefit of the patients and the families in dealing with this. I think this is really well set up to answer this question in this broad population. Do you see your ability to enroll a broad DEE population when despite all the great data you've shown so far, many of the perceptions do not give sodium channel blockers to certain types of DEEs? Yeah. I think it's particularly the 1A, right? Not all Dravet patients have mutations on 1A gene, as you all know, but let's call Dravet 1A, something similar there. It's probably going to be less. Rightly so. There's a lot of stuff that works, right? There's a lot of patients on Dravet that have actually very good control. If you go back to that pie that we showed before, we talked a lot about Dravet for all the years because there are drugs approved. They're a very small proportion of the patients. We might not end up having a lot of that in the study or maybe any. We just do not know. We need to start. The premise of 200, let's call it just to round it up, patients with DEEs that would be available for this trial, if the early indication on all the conversations we have been having with sites can be used, there is going to be no problem whatsoever actually getting this off the ground very quickly because the majority of the patients are not controlled on that. There was a lot of conversations with the agency, as you can imagine, about, "Let's make sure we do not put all 2As and 8As in this study." Maybe that is more of the concern. I would not necessarily call it concern. It is more for us to be careful on allowing for a lot of 2A and 8A. We do the overall approval before randomization anyways, medical monitors and company. In the words of even people that were in the meeting with us, kind of, "If you know it works, right? And if you know it works, we shouldn't be adding a lot of those." We're not seeing that as a problem whatsoever here, something to watch. The perception changes a lot when data is available. We've been doing, I believe, quite a good job, and Steve particularly, and our medical team meeting with all the key opinion leaders in the U.S. We have this once or twice per week evening meetings where there's a discussion. I think they have the same initial bias, prejudice, and very quickly question that, "Well, when you know anatomical is not in the same place, why are we using the same thing?" These are very smart people, very well-intentioned. They want to do good for the patients as well. Is the FDA going to require you to have a certain representation from each DEE type? Or are you going to try to make sure you have a certain representation? No. The premise is really, if you seize, if those seizures are motor, right, quite important, if those seizures are not in a cluster, then go back to what Steve showed on, as Tim calls it, the tree of life, right? There should be no reason not to try this drug. At the end, there is a disproportional amount, of course, going to need to run subgroup analysis, make sure they're consistent, and things like that. A priori, there is no restriction on the mix. There are over 1,000 DEE genes now. It says, "No, we don't do that." It would be futile. LGS is the ultimate proof that that is not necessary, right? Maybe that's the discussion we don't have publicly. LGS is not a condition. LGS is a basket, if we want to call basket because people call things baskets of things that manifest similarly. A patient with a focal epilepsy does not have a single etiology. It's just manifest in locus, in a focus in the brain. Sometimes you stay there, sometimes generalize. It's the same general principle, right? What are you treating, the manifestation or the overall etiology? The target label here would be ages two and older, DEE, no other qualifiers. Yeah, with presence of seizures, right? Because there are some of them that don't have seizure as the primary driver. Why the incorporation of the full placebo arm in this study versus the prior? Yeah. Levels of comfort are, I think, particularly from us and the agency, right, in terms of, "Okay, this is a little bit more that should be a little bit less severe population," based on what we discussed before. Keeping this longer is going to allow for better understanding of this diverse population. Some of them might take a little longer to act. Some of them might take a little bit shorter to act. I'm going to call this the bread and butter. It's normally 12 weeks, as you know. There's a question here as well, why 16, right? We know two things so far about this drug. Many things, but two of them are important. One, the more time, the more effects. That is why we are seeing this. The other is the ability to go to 1.5 mg is relatively late, at day 35. With about 20 days to get to steady state, you want to make sure that if those patients went to 1.5, they reach steady state. It just required a slightly longer study here. Patients would be allowed to be on their background therapy, but then only one sodium. No, not anymore. In the previous study, we were a mess. In both cohort two, there is no restriction for one. They can be on two. They are on two as of today. The patient has been moved from one to two in both cases. Correct. Yeah. It is global. I mean, that is going to be a critical part of getting to the patients. We have developed a portal that you are going to see for a lot of these trials. It is important for patient engagement, alignment. This will really unify the way we look at these trials and the way that we understand the cohorts that we are dealing with. Those are the committed geographies, so geographies where we either have applications or we have commitment from physicians and are working on applications. This is not theoretical. This is the practical footprint at this point. Just a clarification for EMERALD, will you be excluding SCN2 A? We are not going to be excluding, but we do not expect that to be a significant number of patients to be included. That was not a requirement to exclude, just to watch that they do not become 80%. All studies will have SCN2 A. What if you see some different efficacy? What does that mean? It's hypothetical, but yeah. Size, I think it's incredibly unlikely that we would see no efficacy. But it's different. Yeah. And we know there is variability every time you repeat a study anyway. What you're looking here for is a large significance effect. We don't believe that that is a large risk. How do you arrive at 1.5 mg? At one, we are already seeing the results we are seeing. The trigger is seizure reduction. We're not going to tell you how much, but the trigger to go to 1.5 mg is a level of seizure reduction. We're gauging in terms of, let's say, one etiology requires more than what is one. It's almost an insurance policy, right? We have that ability to go there. Are we going to require to go there? No. At 1.5 mg, it is very high, the exposure. You can try to simulate that with a drug that has the properties of relutrigine. I think that is proper that we stay at the maximum. Did FDA say that there was a ceiling of any particular? We did agree on a ceiling of 1.5 mg for this indication. A ceiling of 1.5 mg, but of any proportion of the subtype? No. Okay. It was a very interesting debate. I'll call. Yeah. Whether or not proportions would be necessary and whether or not you see you've been avoiding to use the term baskets, right? There is a reason why we're avoiding because a basket is something that you know what you're putting in, and you try to keep proportional. This is not a basket study. This is a phenotypically defined study. A basket would require proportions, but somebody who moved apartments recently, I fundamentally disagree with you. It is in a basket. God only knows what's in a basket after you've moved. I would say that it is really random. It is not random. You are very deliberately putting in. No, this is not a basket. Meaning patients come in, they manifest a given way, and they get randomized. A basket would be, to underline, correct. Correct. Inclusion criteria match as Steve just described, that can be randomized. A basket would be one patient with mutation A matching mutation A on placebo. That would be an infeasible, unreasonable, scientifically idiotic trial to run with 1,000 potential. You can even, from a combinatory perspective, just that trial couldn't be run. Do you want to ensure a minimum representation of a non-DS, LGS, SCN2A and DEE? We don't. If you have a drug that can work across the boards because mechanistically, as Steve showed, it's like this is a fundamental mechanism to reduce seizure. Why would you want that? I'm not saying not have those patients. I'm just saying limit them because you want enough representation from the others. Obviously, we're going to monitor this. So we're not going to let any phenotype be overwhelming on the study, but a priori, there is no reason to believe that there would be an issue. Did you ever consider taking an approach like Longboard had done where they had carved out certain groups such as LGS separately from a broader DEE population, just sort of as maybe a little bit of a hedge as to getting to a larger segment? Should the overall trial have any issues? Yeah, absolutely. I think it's incredibly important to consider what everyone else has been done. Then we go back to the science as well, right? Serotonergic mechanisms are not downstream, as Steve showed. They've been shown. They've been on the market for a long time. Cannot ignore the evidence in the markets. Fenfluramine does not work in most of those indications. We love narratives, but I love data more. The data shows that it doesn't work. Other serotonergic mechanisms are unlikely. If I believe that was unlikely, if we all believe it was unlikely, we would have designed that study because that is a hedging. If you don't believe, then you have to follow the science. The science is incredibly clear that it's not a requirement. We say this, I don't know. Sodium channels, there is no they're all going to get to EMAtS. There is almost no other mechanism that can guarantee that you'll stop a seizure. You will stop a seizure if you block enough sodium channels, guaranteed. Because it is the only way to generate an action potential is with a sodium channel. Whether you're modifying the output of an interneuron, which is what a serotonergic system does, there's a limit to the effect you can do through that particular mix. It limits. We don't have that with the sodium channel. That brings a lot of conviction that this is going to work broadly. Should never disclose the genotypes. We shouldn't. T hat's great advice. Thank you. We all still happen to like Acadia for everyone. Not always wise. That's what you're saying. All right. We'll switch to the Solidus platform now, which is our antisense oligo platform. It's been in the company for some time. There are four molecules, one late stage, three preclinical. It's a very well-defined platform here. We've got a computational start to help us select ASOs. We've got a very well-trodden way we think about development as well. I think that's important. The platform encompasses that. For Elsunersen, the ASO is a Gapmer. The Gapmers are a very specific instantiation of antisense oligos. They've got a chemical modification in the middle that's different to the chemical modification on the wings. That sort of ASO, when it binds an RNA, it targets it for destruction. It's like a tag for destruction. It attracts the attention of this molecule called RNase H. That destroys the RNA. You can, with a disease like SCN2A gain-of-function, the therapeutic hypothesis is, let's just reduce the number of channels. That is going to be beneficial because you've got too many overactive channels. We've got a whole, and we'll show you a whole slew of preclinical work that validated that concept. See here, this is a mouse harboring a human mutation. These mice are really sick. They're dead by day 30 if you do nothing on the left. What you can see when we inject at birth with two different concentrations of an SCN2A Gapmer ASO to reduce the amount of channel, either by 50% in blue or by 80% in orange, you get an extension of survival. That is remarkable. Now, if you give a second injection to compensate for developmental changes and sizes of the animals, that second, you can see an even more pronounced increase in survival. These animals, the second dose at a 50% reduction extended out to 120 days before you started to see any, and that's because of the pharmacokinetics of the ASO wearing off and then the disease remanifesting. The final test was, what happens if you intervene a little bit later in the natural history of the disease? Because in a human, you can't always do it at birth. We did it as late as we possibly could before the animal started to die. You can see at the point where we injected with a control ASO, they all just died within a few days after that. You can see in the 50% and 80% for that later stage injection, incredible. Really, that delay was just the time it took for the ASO to kick in and start to work. The timelines for the human are so different that this is a non-issue for people because the disease doesn't develop as quickly. The ASO kinetics is much better suited to the disease kinetics, if you want, in people. I'm going to move through here. Right. This is the overall program. It's like a roadmap, but we've got so it can be a little bit confusing because EMBRAVE, part one, that we, I think it was the fall of 2023, we first released that for patients, open label, and saw the first signs of remarkable seizure reduction that was consistent. I mean, this is really great proof of concept that this ASO is doing what we wanted to do, and it's having the clinical effect. EMBRAVE, part A is an ongoing dose escalation study. It can be between 8-16 participants is an option. Three-to-one randomization. This is a proper control study. We've got a three-to-one for the treated arm and the sham arm. We're going to get a very good look to see what happens. In EMBRAVE 3 is a really, again, I think a simple, elegant design that lets us get a registrational cohort in a controlled fashion with a bunch of patients. There are two additional cohorts that let us wind back the intervention time to birth. That's a really important thing. The label, we're shooting for a label that's going to say from birth onwards, you now can take this ASO for this disorder. We've got a very powerful way of understanding the genotype and how to include patients in this study. We know that they're validated as gain of function. In the other studies, this just gives you a taste of what that EMBRAVE 1 from 2023 readout looked like. You can see 30% reductions, etc. In part A, this is the design of the trial that's ongoing right now with the randomization. We expect it to finish this by mid-year. These programs are moving ahead at light speed. This is going to be very important for supporting registration with the EMBRAVE 3, which is the registrational study. Finishing enrollment or finishing? Finishing enrollment. Yeah. This is a registrational trial. Very simple design, one-to-one randomization between sham, Elsunersen for 24 weeks, and set to start shortly, very shortly. We're in really good shape with this. This is the cohort one. You see it's the control part. If you look at the next part. Yeah. Just maybe, Steve, sorry for interrupting you. One of the important things here as well is this is the shortest disease-modifying study being conducted with an ASO for a CNS condition. It was an incredibly productive discussion with the FDA on what is expected based on what we know from this drug, what you've seen, the overall biology of the channel, and the number of doses we've been giving to show a disease-modifying effect. We are pretty happy that we landed at 24 weeks because I'm going to bring back to patients are incredibly severe. Keeping them on sham for one year, 18 months, it is when necessary, it's necessary, but in this case, it was not necessary. We were incredibly happy with the fact that we're limiting the overall exposure here as well. I would say even happier because that is even more unique on what Steve is going to show now. Yeah, the way to reach back to zero is actually really, I don't think you can get more streamlined than this. Two cohorts, uncontrolled. There is no placebo arm, NF5 in each group. You can start first between the one to two-year group, have a look at some of those, and then start to introduce it to the younger group. We can fairly rapidly move in. There can be a bit of overlap potentially here between these cohorts as well. We can very quickly get to the point of a zero, which we think this is the disease that manifests early. The earlier options are absolutely very critical for these patients. We're doing what we can to be able to include them and give them that option. Can you remind us about the four patients where you gave us the data? Eventually, are they still on treatment? Where are they now? Yeah, it's a good story. Maybe you want that sometimes you got a little bit sad to report. At that point in time, we could only dose those four patients. You wait for a review by the FDA on the next step. Remember, it was a seamless design. We had four patients review. How many doses did they get? They got four doses. There was a natural stop, a planned stop on the dose at the end for a decision on the next cohorts. Unfortunately, we saw the natural history playing full-blown after we stopped. Two patients died after stopping the drug several months after. They end up getting more seizures, as we would expect, that they had regained. They got into the overall complication. Were they the most severe or the oldest? They were similarly severe. It's very idiosyncratic. Sometimes happens, right? Infections, ICU, status. I don't have complete confirmation, but I think they helped a lot with these conversations with the agency. It's unfortunate what happened with them. Because they're not on drug and because it was several months later, there could not be any association whatsoever other than that the drug was keeping them alive. The other patients are on drug and doing well. The other two. The other two patients. They were allowed to stay on drug? They were quite quickly when certain things are certain decisions are made for administrative purpose, and they end up being resulting to patients dying. Those bodies making the administrative decisions tends to reverse, tends to revert their decision quite quickly. They are on drug today. They've been treated for roughly 18 months. 18 months. 18 months. With the highest mortality in most of the DEEs are two A patients. Yes. The natural history is terrible. Six times higher chance of dying by childhood than drug at two A patients has. If those patients, let's say, have two commercial programs here, if they have inadequate control on one of these genetic, could they still go on? I think there's natural complementarity there. The ASO tackles the genetic root cause. This is now brain is different because it's developmentally everything about it. There's going to be residual excitability due to other things that have happened. Relutrigine is very well positioned to address that. It's not going to be only for this case. Relutrigine could be a drug you might add on for a whole host of other genetically defined therapies. We think that combination of root cause, residual hyperexcitability is a really important match. We are definitely thinking deeply about what this synergy could look like. Can you give us some sort of sense of, I think you guys reported the percent reduction in seizures, but the residual seizures, how many are they? Put it in context with the SCN2A for Relutrigine. Yeah. They're definitely way less severe. I think that's something we see as well here. Right? Can you imagine that it's likely different motivations to participate on intrathecal trials than on oral trials? Right? And these patients tend to be fairly uncontrolled as well, but relatively anatomically stable. Right? There's a lot of anatomical change that happens because like wheelchairs and other things here. What we see in the short run, we didn't talk about what's happening now with 18 months and so on. It is, I wouldn't call modest, but about 50% reduction is not modest on that. There is somewhat of that plateau there for a while. Reason why we have this slide here as well. I think the combination because they are on other drugs that are not necessarily adequate to control. The combination of both of them might be, and it's likely to be the way to really round it up and get the best treatment for those patients. I believe Yas has another question as well. I don't know if she's still on the line. Yeah. Yeah. Thank you so much. I hope you can hear me. I guess, team, as you're thinking about sort of the complementarity of both a small molecule as well as an ASO that you're developing, is there an opportunity, I guess, if patients obviously with genetic mutations are going to be identified earlier, like as we think about modeling both of these drugs that have $4 billion total opportunity, how do we think which population is easier to recruit, get to pivotal data, think about it? Obviously, relutrigine has a broader population. Maybe help us understand sort of the developmental timelines and first-to-market and uptake between these two. We've been working on the recruitment for EMERALD 2 for about seven months or so now. I'm not going to give you exact numbers, but I'll give you order of magnitudes that are more patients interested than slots on the trial right now. The question is more eligibility. It's not necessarily whether or not we're going to finish this study. It's going, again, as we said, incredibly well. We don't have a single site open for EMBRAVE 3 yet. They are in the process. Every single patient needed for that study is already identified. Not making any promise here. That's all of them are going to randomize. You know what the sites are? They're just not open. The site directors have their records. Patients, their willingness to participate, their seizure counts, their concomitant med. It's just unheard of. They're not even open. It's a small study as well. Together, there are 120 patients. It's not a small study. When you look into other GE studies, it's taking two to three years to recruit similar size. Right? Again, this is not a promise, not a guidance that we're giving. It's just the expression of interest is clear here. I can't help myself but to think, and Joe's going to remember this really well because we worked together back then on different situations, on similar questions that are being asked about type 1 SMA back in the day. A lot of you drew us parts there as well. Why do you need three different drugs, three different mechanisms for if you have a gene therapy, an ASO, oral splicing modulator for that? Fast forward today, and you all know that market's better than I do at this point in time. I think what we learned is that biology is tricky. The majority of them probably needs a little bit of them, or there are specific situations. Each one of them are multi-billion-dollar drugs. Not a straight line to get to the same results. That is one good comparator because pediatric, severe, high mortality, similar mechanisms as well. The only difference is like a gene therapy will never, and I do not say never often, going to be developed for this because precise control is important for these channels. We cannot deliver at this day and age like vector-based gene therapies with precise control. I think it is just another sign that we are going global. We have got these sites lined up. Portal is the same. We are taking advantage of that interface with the participants. I think, again, we are just poised to really move as fast as we possibly can on this. What I wanted to just in closing for my part is just give you a really brief introduction to the pipeline for the next ASOs coming forward in the company. There are three of them: PRAX-80, PRAX-90, PRAX-100. One of them is for PCDH19, where we've practiced, again, I think, is sort of leading the way in a novel way of thinking about delivering a therapy. It's a disease of mosaic expression. Our idea is to completely ablate because we know that males that have none are clinically normal. So that's much worse to have a mutant and a wild type expressed. PRAX-90, SCN1A, haplo-insufficient. You need more of it. What I'll focus a little more time on is PRAX-100. That is a haplo-insufficient autism, one of the most common monogenetic causes of autism. Not enough 2A. The channel that when its gain of function gives epilepsy, when it's truly haplo-insufficient, 50% less than normal, autism manifests. There are about 20,000 of these patients. It's the analog to Dravet. Dravet is the same thing in 1A. This is 2A, the excitotrogene in the excitotrinuron. When you do not have enough of that, you do not have enough activity now, and that gives rise to autism. The therapeutic hypothesis is, can you double the amount that you have got? This is just giving you a taste for what we have seen with four of Praxis development candidates measured in a humanized animal that contains mouse that contains the full human gene. We are giving it the human ASO that is designed to increase the amount of protein. You can see all of them are pushing at more or roughly 200%, which is a doubling, which is a restoration of the normal amount of 2A. We think we could not be more excited by the fact that, and this is hard, as Marcio said, being able to increase the amount of a protein is difficult. We are not disclosing the full mechanism here that we have deployed. We obviously know about retained introns. We know about everything. We think we are just very excited by these first signs of direct increase in protein. It is the first time we have described this publicly. Just giving you a feel of where this platform could go. I think, Marcio, I was a bit floored by the size of the potential market here as well. Very excited by what we can do for these kids with 2A haplo-insufficient autism. We'll hear more about this soon. This is not getting rid of the poison exon strategy. It is not. It's undisclosed. We looked at everything, and we'll reveal more later. There are other ways to increase the amount of protein that do not involve that. There are other things that you can do, bind RNAs and different RNAs and in different spots that make you end up with more protein. Another complementary way to answer a question, we are certain that retained intron strategies are important in certain situations. They are highly dependent on the rates of the events in the population. When you understand the population, the rate of events, people should do a little bit more advanced molecular biology, not the 101. If the event is not there, there is no effect. The event is not 100%. No event is 100% expressed. Retained intron events sometimes naturally disappear throughout life. We like when they stay. You could assume maybe one here is, maybe one that's not here is. We had hundreds of those. We like when they stay. When they do not stay, they are a big problem because it means they are going to work for some patients, but not for others. We like the idea to start with working for a lot, not a few of the patients, because you can always walk back from there, but not walk forward from not working into people. Maybe I will leave there until next time when we show the many more things we have in this program that get us so excited. All right. I know Doug has a question, and then we can get more questions or wrap it up. Doug? Yeah. Thank you, Marcio. Just going back to EMBRAVE 3 with the additional cohorts to get the age lower, I'm just curious, how quickly do you think those would be completed? Would those be sort of, would you seek labeling through an sNDA process, or is there any way that you could perhaps even get those completed and included at the time of launch? Thank you. It is fairly fast, and we're incredibly grateful to the Division of Neurology to work with us on this from the FDA. There's a requirement for a few doses for the first patients on the first cohort, but not for the complete time, not for all patients, three patients, four months. And then we can move to the second cohort, and then we can move to the third cohort, as one would expect, right, to be. In a sense, we are the rate limiting. Those four patients today, it means that before the end of the year, we're already dosing the next cohorts. There are integrations of the different studies here. As you can imagine, they might allow for that to go faster. Stay tuned. I would say right now, without guiding too strongly, probably the cohorts one and two would be the initial application. It might take a little bit longer for cohort one because it's birth, right? Which means that we need to know the patient was just born, available, willing, able to consent, stable enough. Those studies tend to take a little longer, those cohorts think, in general. It might be the case. It might be that we get lucky and we are accelerating that. I think all the way to one, we should be pretty certain. All the way to zero, supplemental NDA is probably the way to go. Okay. Great. Thank you very much. That's helpful. Of course. Sorry, you cannot start cohort two until you dose, at least give one dose to all of the 40 in cohort two. No, no, three patients. Only three. Correct. I have a question for Tim. With so many things going on, can you remind us what your cash position is and your runway and what is baked into that? We disclosed this morning that we had $472 million at the end of Q1. I'd continue to confirm that our runway is into 2028, and it funds all of the trials through their readouts that we talked about today. We've got the cash to execute all of this. Sorry. We got to test with the press releases. No, it's totally cool. Marcio, just to clarify, for the EMERALD study, are you doing site activation in the middle of 2025? Is that when it's going to start? Y es. Our goal is to actually get patients enrolled at that point in time. It's happening basically right now. Okay. No pressure on Brian who you're going to activate. Yas has a question for Tim. Yes. I guess I'll let Tim take a start in, and Marcio you can add on. I mean, obviously, with this really rich pipeline, how do you think about which assets could warrant to be partnered at an earlier stage of development versus taken all the way to the end? Or maybe help us understand how do you think about factors that determine partnering versus bringing to the finish line and potentially even commercializing on your own? That's one. The second one is you guys have always done a great job using great resources for getting interested patients. You did that for your essential tremor study. You're currently doing that for your focal epilepsy studies. I'm assuming that same resource is being utilized for the entire pipeline. If you could just talk about that aspect, because I'm sure as enrollment continues, we're going to learn more, but maybe both would be helpful for us. Sure. I'll kick off on the idea about partnering these assets. We're very excited about them. I would expect there's other people who would be excited about them as well. The best thing that we can be doing for patients, and I think for shareholders as well, is focusing on the execution of advancing these trials to further de-risk them and to validate the mechanism and what the applicability is. I think when we look at which ones we might partner in different ways, there's different things to think about there. For example, on Vormatrigine with more of a focal epilepsy, much broader market, maybe a partner who has broader commercial reach might make sense. Although neurologists are a fairly targeted market and something that we feel that we could do. Again, if there's a partner that comes with a good offer, and I think that part is critical. We don't want to be wasting our time on deals that don't make any sense. That might make from a commercial use perspective or use of resources and realizing the full potential of the molecule make sense for some of the smaller market areas like what we talked about today. The rare diseases can be done with a fairly focused field force and commercial effort. They also have commensurably higher pricing as well that we talked about. Even globally, I think there is an opportunity for us to look at commercializing that as well. There are companies set up with infrastructure outside the U.S. that might make sense to partner with also from a bringing-to-market perspective and being sure we realize the full value of what these molecules can do. All of those conversations we are best prepared for if we continue to advance these trials. We are not waiting for somebody to come. As you'll get from looking at our pipeline slide, all four of these assets that are in late stage right now, if we were to include Vormatrigine and essential tremor as well, are all in parallel enrolling quite quickly. We are not taking our foot off the gas with any of those, which I think leads to your second question. Let me just address that quickly, and I'll leave it to Marcio to round out as well. We have done some very thoughtful, I think, innovative ways to recruit and identify patients for essential tremor. I think what we disclosed today is now it's over 200,000 people who have been interested in the essential three study for Vormatrigine, reutilizing the EMPOWER registrational study to find patients. We will apply similar type tools as well for both the DEE programs that we talked about today. I think the other aspect that we will lean into more on the developmental epilepsy side is the patient organizations. They're very well organized. They have a good sense of who in their networks are good candidates for these trials as well. That gives us, if you will, a third leg of the stool on ways to identify and find these patients sooner. There's a whole breadth of tools that we'll be using. We very much feel that it's important to invest on the recruiting side of the trial because the faster we can get the right patients into the study, the less resources we need in the long run, but also more importantly, the faster we bring a new therapy to these patients as well. I'll pause there and open it up to Marcio too. Maybe on the recruitment, we get a question a lot, right? It's like we get patients faster. The quality of the patients is great. The quality of the study is great. It's like, how do you guys do this? Maybe I'll take a little bit of a detour for a second, right? Back into when everyone out of manufacturing the world sucked at producing cars, people would ask, how does Toyota do this, right? You might recall back then, you'd say, come and look into the plants. I will show you everything. We don't have any problem if you actually take a look under the hood. We take the same general approach with competitors, with colleagues. You cannot imitate, and you cannot copy passion, love for delivering on this, competency, and lack of silos. Go ahead and try. That is the way I would say, right? I would leave people behind, but it looked like Taylor, who leads our recruitment, and Megan, and Kelley Dalby, and many more people here. They are obsessed on getting this done. It is impossible to copy obsession. No one stops until they understand what is necessary. I could give you financial reasons, system reasons, the fact that we actually look for the best possible technology, that we lever technology. Yeah, that is all important. At the very end, at the very core, there is something no one else in Cambridge, in Boston, or anywhere else can copy. These people love it, and they want to deliver to these patients. I am incredibly grateful to everyone here at Praxis that does this. That is the long answer. Thank you. All right. I think that was it. I do have a closing slide, but I'm not sure we even need to go through this. We went through absolutely most of it. Maybe one aspect that we haven't touched before, I believe is quite important here, is all these molecules are pretty new, right? They are all discovered very recently. One of the questions that is quite important in drug development, as you know, currently, the past was important. I think currently is even more important is how long is the intellectual property runway? For how long can we actually see the benefits of this while we are working on the next thing? We're not going to give you specific terms for each one of them, but they all start with a four zero, or they all end with a four zero here. I think most of us, because we're all very young in this room, are still going to be doing this and still going to be loving this. A lot of people are not even going to be with us that develop drugs right now when those patents expire. We have a lot of grounds to bring to life to patients the promise that we have here today. I'm insanely grateful for all of you not only attending, asking great questions, engaging, but staying 40 minutes longer. It shows the importance of these events and shows the passion that we all share about helping these patients as well. Thank you very much. We shall meet again and have a wonderful weekend, everyone. Thank you.