Good afternoon, everyone, and thank you for joining us at Needham & Company's 24th Annual Healthcare Conference. My name is Ethan Markowski, and I'm a member of the Biotech Research Team here at Needham. It is my pleasure to have Geoff McDonough, President and CEO of Generation Bio, with me today. As a reminder, any viewers who are watching through our conference portal are able to submit questions via the Ask a Question feature below the video feed. Before we dive into our conversation, I'll let Geoff provide a quick overview of the company. Geoff, with that, the floor is yours. Great, Ethan. Thank you for this very kind invitation. It's a pleasure to spend some time together, and I'm really looking forward to the discussion and Q&A. As you said, I thought it might be useful just to frame what we're up to and maybe front-run some of the content so that we have a scaffolding to anchor our conversation in. Generation Bio is really working to change what's possible for people living with T-cell-driven, predominantly T-cell-driven autoimmune disease. What that is about is the idea of a first-in-class targeted siRNA set of therapeutics that address the role that T-cells specifically play in driving autoimmune disease. It's the combination of our unique, selective, reducible LNP delivery system for T-cells and the very well-understood, well-validated characteristics of siRNA, predominantly the ability to hit undruggable intracellular T-cell targets. The undruggability here derives from the combination with highly selective delivery. Of course, there are targets in T-cells that can be included, but they also are expressed in other non-T-cell targets, which tend to drive a lot of the off-target side effects for many classes of drugs, as we'll come on to. We are very focused on bringing our first IND forward in the back half of 2026. This idea of bringing a well-validated payload into new tissues is one that has many nice analogs. Alnylam, of course, started the siRNA revolution, let's say, about 15 years ago, and about 10 years later, started to migrate from the liver to the eye and to the CNS. Companies have built the antibodies required to get conjugated siRNA into skeletal muscle, and now we're doing the same with our LNP delivery system for T-cells and other immune cell types. I'll come back in a few slides to why an LNP is required to make this work in immune cells and in T-cells in particular. Our LNP system is unlike any other LNP system that we're aware of in both its approach and its characteristics. We are not unique in using a small-format biological ligand to drive engagement with a target receptor on a target cell population. What's really special and unusual about our system is the ability to dial out the nonspecific biodistribution, which is a hallmark of all other LNP systems, that nonspecific biodistribution or clearance by the liver and the spleen. We refer to that ability to avoid the majority clearance by the liver and the spleen as our core stealth LNP. Those properties are driven by a proprietary ionizable lipid and a proprietary anchored polymer that really enable the key features of the system. In addition to avoiding uptake by the liver and the spleen, we also drive very efficient endosomal escape. It is a modular chassis that allows us to conjugate novel ligands to access new cell types and tissues. It is fully reducible, and it works with all nucleic acid cargoes. To give a flavor for the behavior of the system, I am showing data in the left panel here in a humanized mouse. Here you can see that we are reaching upwards of 90% of T-cells that express the receptor that we are aiming for or targeting, and we get equal access here to CD8s and CD4s. Although the data is not shown, we are also sweeping in NK cells. Really broad access to the key effector T-cell types that you want to get to to bring down the level of inflammation. Moving into non-human primates in the middle panel, what we're showing here is the selectivity of the system for on-target. You can see for the parental LNP in the left part of the middle panel, when we have no ligand attached, we get no access to T-cells. On the right, you can see with the ligand in place, we get very nice and broad access to T-cells. That is just making the point that all of our selectivity is driven by the activity of the ligand, not by non-selective uptake in the target cells. On the right, you see an example panel of non-target uptake or, let's say, lack of off-target uptake. These are monocytes. You can see that we have a really de minimis level of uptake in monocytes. This is representative of what you see in B cells, dendritic cells, and other non-immune parenchymal cells. It is an obvious idea. It is very attractive to think about knocking down targets in T-cells to address autoimmune and autoinflammatory disease. It is an idea that has been pursued for many, many years. For whatever reason, a conjugate approach for knocking down targets in T-cells and other immune cells has not been successful thus far. It is probably some integration between the receptor uptake kinetics and the intrinsic characteristics of endosomes in immune cells, which tend to be high integrity and do not allow for leakage of siRNA into the cytoplasm. This is an unexplored field of opportunity to bring siRNA to T-cells. Our hypothesis was that the fact that our T-cell delivery system is uniquely selective for T-cells and is intrinsically designed to break open the endosome could perhaps overcome this longstanding challenge in si delivery to T-cells. Indeed, that's what we see in a variety of settings. I'm starting on the left side of this panel showing in vitro knockdown at nanomolar potency against beta-2 microglobulin, which is a very highly expressed housekeeping protein in T-cells. You can see that, and that's a delivery in vitro with our ctLNP system. In the middle panel in a humanized mouse, you can see across three different siRNA chemistries for the same tool siRNA compound directed against beta-2 microglobulin at 0.1 mg/kg. It's a very, very low dose. You see really total knockdown in the chemistries represented on the middle and right side of that slide. On the far right panel, you can see the persistence of those same chemistries also at 0.1 mg/kg, also a single dose that we're getting two to three weeks of really nice knockdown in this early model, again, with an unoptimized siRNA. It's important to say that this is a mixed model. Here you have both active and naive or inactive T-cells. The degree to which you're seeing the half-life, you know, or the recovery occur here, it's really driven by the dilution because in this model, you see expansion of T-cells in those T-cells where we're not dividing the knockdown as total throughout the period. This combination of very T-cell selective delivery and then an intracellular mechanism of action with very nice predictable longer-term pharmacology has us thinking about what's possible in addressing unmet need in the INI field by hitting T-cells selectively. As a primer, of course, this is the broad trajectory through which T-cells become activated. They start naive. They encounter their antigen, whether self or infectious. They activate and proliferate. Downstream from that, they differentiate into four different effector cells, at least four effector cell types. You have on the top CD8-positive cytotoxic lymphocytes. These are primarily the drivers of self-cell destruction. Then you have three predominant kinds of CD4 effector cells referred to here as Th1, Th2, Th17. They all are active more or less in a variety of different indications. They have this very important fifth subtype, which are T regulatory cells, which generally have a self-balancing effect in inhibiting the activation cycle of the effector subsets. Now we have good classes of drugs that activate or that function on the left side of this figure to inhibit activation and proliferation. They tend to have a relatively broad spectrum immunosuppressive effect. You can see a representative group of JAK inhibitors here on the right side of that square. You can see some relatively older calcineurins and glucocorticoids and other drugs, all of which have really good efficacy in broad-spectrum immunosuppression, but have very significant and sometimes very long-lasting side effect profiles that really limit the kinds of indications where they're used and within indications, the degree to which they have uptake and penetration. If you look into these effector pathways on the middle of the slide, you can see that in these pathways, the second-generation INI drugs, predominantly monoclonals and bispecifics, do a very nice job of knocking down the activation in any one of those pathways. Interestingly, many of them have an efficacy ceiling. You have substantial proportions of patients who either respond transiently or do not respond at all. There is kind of a level beyond which addressing the pathology downstream at this level does not seem to be able to move the needle. With our approach, the thinking that we have is around the role of selective delivery of siRNA to access upstream mechanisms that are critical for multiple arms of autoreactive T-cell pathology. In other words, inhibiting activation and proliferation and at the same time broadly knocking down the activity of these effector subsets. The mechanism, the ability to go after intracellular genetic targets can open up the idea of accessing pathways that are critical for effector cells, but not required for regulatory cells. In the T-cell context, the idea of restoring balance is something that we're really excited about and looking to explore in the combination of our delivery and novel targets. We intend to name our lead target and candidate indications around mid-year this year. As a kind of primer ahead of that, you can see a list of what we consider to be primarily T-cell-driven diseases. All of these have some of the characteristics of what I mentioned before, either standard of care that has a therapeutic ceiling or one that has significant safety limitations that limits overall access and penetration and/or have significant gaps in the patient experience. These are the kinds of things that we're taking into account in prioritizing the indications according to each of the targets that we're going to lay out. Just a final slide to say that the T-cell delivery system is just one aspect of what is really a broad and broadly applicable platform technology. As I mentioned earlier, the core stealth LNP and the conjugation really remain fixed. As you cycle through additional ligands, it opens up the opportunity to go to new cell types. HSC is one of those cell types where we've done a lot of work up to now, but there's a long list of others that would be amenable to this approach. I think maybe I'll stop there and maybe use this as a jumping-off point to get into the discussion. Thanks for letting me show a few slides, Ethan. Yeah. No, thanks, Geoff. Thanks for the overview. I think it's very helpful to sort of frame the conversation. Maybe before going deeper on Generation Bio itself, I do want to spend a minute or two on the current market conditions we're seeing. I think they're hard to ignore. We will focus the conversation on Generation, but maybe could you just start by addressing how the team at GBio is viewing the potential impact of tariffs as well as the recent leadership changes at the FDA? Yeah. I mean, you know, I don't think any of us know how long the tariff storm will last. It's not going to go away quickly, but in principle, you know, it's something that can be negotiated into either a new equilibrium or into a place that we've lived in previously. I don't think that's true for the FDA or the NIH. I think these are two institutions that are populated by a very deep bench of people who are, on the one hand, building the next generation of insights and scientists who in part can collaborate with and/or join industry in the case of the NIH. In the case of the FDA, of course, this is the gateway for us to build a development program. Having competent and ready interlocutors and discussion partners and reference points for building our programs is just critical for the industry. I think it's hard to see how the FDA can rebuild. Now, is the FDA broken today? No. We just fervently hope that nothing further emerges there in terms of challenges for the agency. I think as a development-stage company, obviously, the tariffs are less impactful. I think the health and readiness of the FDA to engage with companies around their programs, that's obviously really vital and much more near term. Okay. With that out of the way, we can go deeper into the company itself. You did give a nice overview, but maybe just starting with sort of a broader topic, which is at its core, Generation Bio's ctLNP platform is focused on selectively delivering nucleic acids to tissues and cells beyond the liver. How important do you think that pursuit in general is for unlocking genetic medicine? I think it's really vital. In fact, I don't think it's overstating to say it's probably the single most important gating facilitating technology to allow genetic medicines to achieve their full potential. I mean, it's kind of too simple even to say it, but that means it's probably worth saying. If you want to create a gain, a function change, or knock something down, or ultimately edit in a given disease, it's important to do that in the cells that matter, the ones that are impacted by that disease. Especially when you're knocking down a target or editing a target, it's vitally important that you don't make those changes in cells that are vital for other function or cells that are not impacted by the disease. The reality is that for most of our technologies, and I'll just talk about two, for AAV and for traditional LNPs, they are pretty non-selective delivery systems. They share in common what we consider kind of a tropic approach. When you say that AAV9 has an ability to get into the CNS, for example, it doesn't mean it's selective for the CNS. It means that it has been tuned to include the CNS, and you're still delivering hundreds or thousands of variants for every cell in the human body in order to include the CNS. Now, LNPs are similarly non-selective at their base because 99% or so of every dose ends up in the liver and the spleen. That can be pretty selective if you want to treat the liver and the spleen, but it's not selective if you want to treat other tissues beyond the liver and the spleen. I think the combination of the needs of payloads to be selective and the fact that most delivery systems are fundamentally not selective, this defines post-hepatic, post-splenic delivery as a really key gatekeeper. I'll make one other point as it relates specifically to LNPs. If you're losing 99% of your dose to the liver and the spleen, that's a real headwind for potency. For many of the next-generation payloads, the potency of the payload is also at the margin. If you combine those two features, it's another selectivity aside, but just as a broad feature of getting to the right doses for patients, that's a headwind for sure. Yeah. No, I agree. There are quite a few companies, mostly on the private side, trying to develop novel ways to detarget the liver, adding lipids, adding ligands. You did touch on it, but do you think that GBio's main differentiation is that ability to sort of stealth its LNPs? Maybe the way I would say it is we've taken a fundamentals-based approach to dialing out that non-selective clearance by the liver and the spleen. We have been very forthcoming in publishing quantitatively how we have done that. At the moment, between the liver and the spleen, we are delivering less than 0.1% of every dose. We are extending a basal half-life of minutes for all other LNP systems to about eight hours in primates. I think we have done something to kind of flip the ratio on its head from 99% non-specific delivery to less than 0.1% non-selective delivery. To be clear, I think, as I said earlier, the idea of attaching a ligand to an LNP to relatively increase its ability to get into a target tissue is a good idea, and it works. Without stealth, it does not work selectively. That is probably okay for certain payloads. That is not to criticize. It is just for payloads where you are knocking something down or editing, selectivity becomes really important. That is why we've really optimized for not only getting on target delivery, but for removing off-target delivery. Right. You have demonstrated the ability, in preclinical studies, to significantly increase the half-life of your LNPs in blood. Maybe two questions on safety that we get occasionally. First of all, is there any potential safety concerns with this increased persistence? On a similar topic, is there any increased risk of redosing LNPs, which I know is an option companies are going for? Absolutely. They are a bit related, aren't they? Toxicity and redosing and safety. First to say, let's just start with single dose. Our system is designed with all biodegradable components. On a single dose basis, the half-life itself does not have any impact on tolerability. We do not see liver function abnormalities or tolerability limits. We're obviously still exploring the outer envelopes of dosing in non-human primates. To date, at doses up to 1 mg per kg, we see no tolerability concerns related to the LNP. Just to point out that in the data that I showed you with siRNA cargo, we're very effective at doses as low as 0.1 mg/kg. We think we're in a range where we have a lot of room to play. The other kind of aspect of safety is related, as you pointed out, to redosing. Maybe that meshes a bit with the longer half-life in the sense that longer half-life and redosing, and certainly the two together, would raise the question, do you get accelerated blood clearance? Do you bring the effectiveness of the LNP down by essentially inducing antibodies against your LNP, either because it's lasting longer or because you're redosing? This is a well-known phenomenon with polyethylene glycol as an example in LNPs where you can see the efficacy of the LNP decline if you give weekly doses over time. It has been something that we've been really engineering our system to avoid. We have a novel ionizable, as I mentioned, and a novel anchored polymer. That combination with weekly dosing over time shows no accelerated blood clearance. We do not see the development of antibodies against the system with multiple doses in humanized mice. We will explore that also in an HP, but we think we've got a system that has engineered out some of the known considerations, whether it's acute tolerability in the liver or accelerated blood clearance with repeat dosing. Okay. I do want to move now into a little bit more about your specific application, your lead application, which is for T-cell-mediated autoimmune disorders. Obviously, people are familiar that there are a lot of companies, especially cell therapy companies now, going after B-cell-mediated autoimmune disorders. Does this competitive aspect have any impact on why you decided to focus specifically on T-cell? Yeah. I mean, I think we're always interested in doing something that can create a major step change in outcomes. I think we are just as impressed as the field is with the kind of early patient series data out of Germany showing this durability for B-cell resetting. There is a lot to like about the idea of using T-cells to bring down B-cell populations, whether in vivo or ex vivo. Our view is that we would not necessarily add much incrementally to that field. Whereas when we started to look at whether there are any existing technologies that can get at intracellular targets selectively in T-cells, we ended up drawing a blank. We really could not name a technology that was able to do that today. Therefore, this idea of what selectively knocking down individual pathways in T-cells could do became really appealing to us. I should say at this point that we have drugs that kill all your T-cells. This approach is non-depleting. We are not reducing the number of lymphocytes or creating a lymphopenic state. In fact, when we transduce T-cells, we do not change their state unless the payload does that. If they are naive, they stay naive. If they are activated, they stay activated. We like that idea of adding a passenger to the T-cells in whatever they're doing and not changing their state so that the changes that do occur are really only driven by our payload, not by our delivery system. Great. Speaking of payload, you guys have worked with multiple cargos now, mRNA, DNA, ultimately landing on siRNA. I know you touched on a presentation, but what characteristics of siRNA make it attractive to use for your lead application here? Yeah. Let's work backwards to forward. In the rear, like the very rear seat, we've really found that siRNA behaves beautifully in formulating with our system. It's small. It's very stable. It creates beautiful particles. I mean, only a parent can say that about their LNPs, but they're beautiful particles with siRNA. All the other stuff we like is the stuff that everyone also likes about siRNA. There's a very clear pathway to designing and optimizing and bringing an siRNA candidate to the right level of potency and stability. It displays very predictable, sometimes referred to as clamped pharmacology, meaning that you can bring a level of stable messenger RNA degradation to bear in a cell that stays at that rate even when you have loading of siRNA in excess of what's required to achieve that level, which then gives you pretty predictable longer-term residence times and longer dose intervals based on optimizing around those characteristics. It is reversible, which is very appealing when you're making modifications to a really critical compartment of the immune system. One thing we really like about the specific combination of siRNA with our delivery system is that the receptor we're targeting to get our LNP to T-cells is not expressed on somewhere between 5% and 10% of T-cells at any given time. I think preserving a pool of normally reactive T-cells is quite important as a safety floor. We like this combination of the delivery and the specific si payload. Sort of a related question, but is there anything unique about the siRNAs you're using themselves, or is the differentiation mainly on the delivery side here? We are using very, very straightforward si designs. As I showed in the data, we're exploring a variety of chemical modifications. I should say those are kind of in the category of relatively vanilla chemical modifications. They're not at the level of sophistication that is required for conjugate approaches. I mean, just as a reminder, the siRNA in our case is contained within the envelope of an LNP. It is not exposed to systemic circulation. It's really only released into the cytoplasm after the LNP has been brought into the endosome. Much of the stabilization and so forth that the more advanced chemistries are there to help create, they're not necessary for our mechanism of delivery. The differentiation is really all on the delivery side. We obviously want to make really good, high-quality siRNA molecules, but we're not doing anything particularly novel there. Because it's on a similar topic, and it's kind of more of a minute question, we have seen it before. I guess, how is IP around siRNA being handled then if they're not potentially differentiated? Are there any issues there, or what are your thoughts? From the point of view of standalone IP for siRNA, I kind of think of it in the traditional buckets, right? Do we have freedom to operate? I think the answer is yes. Are we going to establish standalone novel siRNA? Maybe there'll be some limited opportunities on a per-target basis where those targets have not been accessible before by siRNA. I think the most powerful IP will be the composition of matter around each product, which is the unique combination of our unique and proprietary LNP and the siRNA as a product. That is the most important fundamental IP as it relates to the SI side. Great. Maybe one last one on the cargo itself. From your preclinical work, have you had any sense yet of what the anticipated pharmacokinetics might look like in terms of should we be expecting suppression potentially for weeks or months? Great question. The path to date has shown that the half-life tends to get longer as you move through the species ladder. Obviously, we'll have a view on that as the year goes on. Our estimate based on what we've seen so far is that our dosing will land somewhere between once monthly or once every other month dosing. Some of that may depend on the disease. It also may depend on the ultimate dosing regimen that we land on. In other words, will it be stable dosing, or will there be a loading dose and then a maintenance dose? Our expectation is that once monthly dosing certainly will be feasible. If we end up seeing that same kind of extension in higher species, maybe it goes out to as long as every two months. I would not expect it to go much longer than every two months, but we'll see how the data develops. Great. I do want to take a minute or two to at least remind investors about your ongoing collaboration with Moderna. Maybe if you could just provide some of the details around the deal, any recent progress, and if you're expecting any more updates from the collaboration in the near future. Sure. We are just completing our second year of that collaboration really right about now. As I mentioned earlier, I think the subject of the collaboration is this creation of a set of T-cell selective LNPs. We brought the background IP into that collaboration, and the work together was in identifying and optimizing these selective ligands. Moderna has the rights under the research term to pursue that delivery system with mRNA as the cargo, not surprisingly. They have not announced their development intentions there, and so I can't comment on that. I can just say that our experience of them as collaborators has been terrific. We've really had a very seamless, very constructive collaboration. We are just bringing different payloads to the same delivery system that we've been able to successfully put together over the last couple of years. To answer your last part of the question, Ethan, no, I don't think there's going to be any sort of business-relevant shifts in the collaboration in the next year or so. We'll see how the program develops in Moderna's hands and update accordingly as we go forward. Okay. Yeah. I know, obviously, working with partners, only so much you can disclose at this time, probably. In terms of other collaborations, you have your one with Moderna, are you considering exploring additional partnerships similar or in different therapeutic areas? How are you thinking about potential business development opportunities? Yeah. I mean, I think the T-cell landscape is really broad. I mean, we've talked today about what we'd like to do in the INI field. There's obviously a large application space in oncology. There is a development space, as we were saying earlier, that's beyond T-cells, whether it's B-cells or dendritic cells or HSCs or arthroid progenitors, etc. I think both of those dimensions are open potentially for an additional partnership here. I think later this year, as the data matures around this approach, we'll have a view of where those opportunities may come up. We're obviously in conversations with folks as they stay tuned on our turn towards INI and T-cell and immune cell delivery. We are certainly open to doing another partnership. It would likely be a little different in terms of the structure and goals from the Moderna agreement, but we're pretty open-minded about how it develops. Great. I do want to take this time, as we are getting a little closer to time, to remind everyone listening and watching that you are able to submit questions via the ask the question feature below the video feed. It looks like we do have one, so I do want to address this while I see it. Question for you, Geoff, is the immune evasion does not solve for CRIM negativity potentially. Is this the reason they are using siRNA knockout as an initial target instead of expressing a missing protein? I think I heard you say in the question, it doesn't solve for CRIM negativity. Is that what you said? Yes. CRIM? Yeah. Yeah. Yeah. No, cross-reactive immune material. I got it. Let me try and get the question. I think if I understand the spirit of the question is, what about gain of function? What about payloads that can create a missing protein in lysosomal storage disease or other places where you're just missing either some or all of the material? CRIM negative just means you're missing all material, which means you have no tolerance to a gain of function payload, whether it's an enzyme replacement or messenger RNA or DNA. I think for us, the choice of siRNA is just a decision based on the idea that we can go into a space where others aren't present and try to do something that hasn't been possible in the INI space before now with a payload that's really ready and has so much of the development infrastructure mature around it, both in our hands and in others. The question is a really good one because it actually goes back to the founding of Generation Bio. I think our initial thesis was we don't have a good, long-lasting, reducible gain of function payload to address protein deficiency diseases. And much of our early development was focused on developing such a payload. We refer to it as immune quiet DNA or IQ DNA. To me, that is the preferred mechanism for getting long-term gain and function replacement, whether CRIM negative or in partial insufficiency states. It's just that that payload is not quite development ready yet. When it does become so, we're very interested in addressing your question. Interestingly, the answer in CRIM negative patients might depend on where you express the protein. If you express a protein from the liver, you may be able to induce tolerance differently from if you express that protein in other tissues. I might not have got all the nuance of the question, Ethan, but I think I got the main things. Yeah. No, I think you got the gist of it, but definitely a very interesting topic. Thank you for submitting that question as well. We do have a couple of minutes left, but I do want to give you a chance to highlight Generation Bio's manufacturing process. It's a bit unique feature of the company. You just give a brief summary of the process and the benefits you've observed so far after switching over to the enzymatic version. Oh, sure. Yeah. I think you're probably referring to DNA, IQ DNA. We do have a proprietary process for making that enzymatically, and it has a lot of benefits, but the main ones are that it's very quick and very pure and therefore very efficient. That is a trifecta that's really intriguing because a little bit in the spirit of the prior question, people who need protein replacement generally need it for life. Having an opportunity to create high-purity, efficiently produced material means you can access points on the cost of goods curve where you can really make those kinds of therapies available globally, which is one of the things we're really, really interested in. The flip side, of course, is that now that we're focusing on our proprietary LNP and siRNA, we don't need to own as much of the manufacturing capacity. We own all of the kind of process development around the LNP because the LNP's characteristics really drive its behavior as a drug. We want deep control and knowledge over that. On the flip side, there's a very nice and very mature supply chain to take that process over when we're ready to scale it up as we are now. siRNA has an even more mature and more capable external supply chain. We intend to really leverage both of those things as we go forward. Great. I know an important question that we've been pretty much asking every company throughout the conference this week because it's such an important topic. Can you just remind investors, company's current financial position? I know that's a strength of Generation Bio, cash runway, and what you hope to accomplish with that runway. Sure. Yeah. We ended last year with $185 million in cash and project our runway to go into the second half of 2027. Ideally, we'd like to get our first experience in humans under our belt in that runway. As we talked about earlier in the presentation, we haven't yet made a final choice on target and indication. This is really the goal that we're working towards is to get our first in-human experience well underway within the cash runway with the IND coming in the back half of 2026. Great. Maybe with the last minute, I'll give you the floor to talk about if there's anything at this point that we didn't cover or you feel is underappreciated about the company still. Yeah. I mean, this has been a great conversation. I think we've covered really most of the bases. I think the field broadly, whether it's in genetic medicine, which mostly has its feet in rare genetic disease, or in the broader sense now of seeing the first nucleic acid or genetic therapies coming into traditionally much larger biologics-dominated fields like INI, the role of delivery is really something that I think people are starting to appreciate is unlocking. I think our prior views of where genetic medicines could go were really defined by where delivery could take you. As delivery now starts to become more sophisticated and more precise, especially in the sense that it can be more selective, I think we may see precision genetic technologies like siRNA have inroads into tissues where they haven't played historically, and we hope to be part of that. Awesome. With that, I think we are at time. Thanks again, Geoff, for coming and attending the conference and looking forward to see what Generation Bio has to offer in the coming years. Thank you for the interview. Really enjoyed it, Ethan.