Generate Biomedicines, Inc. (GENB) September 2026 Conference Transcript

Good morning. I'm Sean Lummon, head of SMIDCAP, biotech equity researcher here at Morgan Stanley and welcome to our global healthcare conference. For important disclosures before we begin, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com forward slash research disclosures and if you have any questions, please reach out to your Morgan Stanley sales representative. For this session we have Generate Biomedicines and we welcome the CEO Mike Nellie and President and CFO Jason Silvers. So welcome to the both of you. Thanks, Simon. Thank you. Yeah, maybe just for some introductory questions and we're doing this for all our companies.

We have three and you know I guess sort of how is the rise of China origin innovation changing your competitive position in R&D versus your business development playbook if at Yeah, so I think China, Sean, has been, I think, nothing short of miraculous in terms of the ability to brute force their way to exceptional answers across a broad range of therapeutic modalities and therapeutic targets. Certainly for us, it's forced us to be very thoughtful about how far we push the innovative frontier because I do think if you're not pushing pretty far out, China can rapidly copy, and I don't think it'll stop at just copying. I think they'll continue to push their own innovative frontier, but what we've tried to do is actually think about where we can use our machine learning-based platform to drive answers to more complex therapeutic questions. And so, you know, our TSLP program was largely an affinity maturation program where we basically took the affinity of a molecule like TESPIRE and improved it by about 20-fold. China can get to those sort of answers now. And so, you know, for us, we've started to ask, you know, what complex biological functions are more difficult to brute force your way to answer? So stuff like pH-dependent binding, you know, ultra-high selectivity, these sort of things that can combine and answer more and more complex biological questions has been, you know, kind of an evolution of our own discovery process.

Sure, thank you. I guess one of the most exciting things to me about Generate is that you are an AI-driven company. We have this question across all our coverage, so it's almost a redundant question for you guys. But is there anything you can share that you think about the adoption of AI across the biopharmaceutical industry?

Well, I think in some ways the lead times in drug discovery are masking state-of-the-art of AI in drug discovery. For us, I mentioned our anti-TSLP molecule. That was discovered five years ago, and that was on a model that predates the original release of ChatGPT. And so the state of the art of these generative approaches have progressed on a comparable basis to what we've seen with the large language models. And so, you know, where originally, you know, we were solving kind of, you know, simple questions around binding with these sort of approaches, now we're able to really take on very, very complex biological functions. And the answers we're getting and the speed at which we're getting those answers really is distinguished from anything we've ever seen in drug discovery. And so clearly, you know, I think it's important, you know, to kind of note that AI is not a panacea for all drug discovery and development. There are certain parts of the value chain that are probably more prone to disruption. And, you know, right now we believe molecular generation is at the forefront of that. We think there's going to be some really interesting use cases in clinical development in terms of taking the waste out of the system. But the complexity of biology is so great that AI is not going to solve biology in the near term. But if you can be discerning and figure out which parts of the value chain can be disrupted, the tools can be extraordinarily powerful.

Sure. Thanks, Mike. And last question before we really get into the meta generate. But, you know, is there any policy variable that you'd highlight in terms of the outlook for either yourself or the industry that you think would be impactful? FDA, Medicare negotiation, MFM might be a little way away for you guys, tariffs, global pricing, anything?

Well, I think, you know, clearly, you know, the U.S. industry has benefited from the gold standard regulatory body for the last, you know, 50 years. And, you know, I think the uncertainty within the FDA has not helped anyone in the industry. You know, when you make these sort of long lead time bets, you want to make sure that you have the confidence in the regulator that when you make those bets, if you're successful, they'll ultimately yield, you know, medicines for patients. And so I think the uncertainty at the FDA has been one in particular. I do think the MFN question is something that is going to be deflationary for the whole industry, right? I mean, the reality is the U.S. economics have covered or carried a lot of the profitability of the industry. And, you know, with MFN, I think in the first instance it hasn't been too harmful. But, you know, having spent six years leading large parts of Merck's business in Europe, usually when these sort of policies are in place, it's the start of a slippery slope that, you know, ultimately the government will see this as a cost savings lever and will continue to turn that crank. It won't be a one-time opportunity. And so I do think having the government negotiate prices in the U.S. is a huge swing, given the free market dynamics that have kind of underpinned the industry to date.

Sure, thank you. I've got a few questions here on the strategic framing with your business in post-IPO. Should investors think about your business as a pipeline with a platform or a platform with a pipeline?

You know, I think we think of it as a platform with a pipeline. I mean, the reality is that the first program, our anti-TSLP antibody, is an extraordinary molecule. It'll do a lot of good for patients. But at the same time, it's the first manifestation of what we think is a transformative way of making drugs. We think ultimately every molecule in the future will be generated, generated, not necessarily, you know, leveraging random discovery processes like immunization campaigns and screening campaigns. And so for us, you know, we're at the starting point of this kind of evolution of how drugs are made. And, you know, we think, you know, clearly it's important for us to constantly show, you know, a platform in and of itself is worthless. The platform has to create extraordinary products, and so we think if and when GenerateSuccessful, we'll have leveraged this platform for a number of different products that will make a big difference in the world rather than be a single asset or kind of therapeutic area company.

Sure, and with your lead program, GB0895, how should investors think about that? Would you credit the platform? I mean, it's moved pretty rapidly through trials and showing some great data, so would Do you credit the platform, or do you think it's more that you've picked a de-risked biological pathway?

Well, I think it's actually this combination, right? I think if you think about 0895, what the technology allowed us to do, and the observation we had from the outset of the company was if you have a good prior generative model, you can actually change the CDRs in more profound ways than any other technology we've ever seen to date, right? So historically, if you'd use error-prone PCR or you'd use computational techniques like Rosetta, you could only change about 10% of the binding region before you'd find no functional variance in a library. What we saw with our technology and with a good prior model is that you could change up to 70% of the binding region. And that allowed you to search the functional space in a much more efficient way. So I think the molecule and the molecule design was platform-derived. At the same time, where you direct it was, I think, a strategic choice by the company. We saw that Tespire was kind of an emerging medicine with a lot of multi-indication potential. We thought there was an existing liability with both the half-life and the affinity in the initial molecule that could be addressed with our technology. And then I think what we also did, though, and I think critical to the success, has been really clever drug development. And so I think, you know, this is going back to the statement, you know, AI is not a panacea. It requires really savvy drug developers to come up with innovative clinical pathways that allowed us to kind of go from phase one to phase three.

Sure, sure. And still on 0895, so clearly a very potent molecule compared to the benchmark and the long dosing interval for those that might be less familiar, you know, six months. You know, how should investors think about that? Is it a play on convenience, or do you really hope to show better efficacy?

Yeah, so efficacy is probably not the direction that we'll take this, although in preclinical studies we did show kind of a five-time improvement in potency over Testfire. But in reality, at our 300-milligram dose, which is a dose we're in Phase III with, we're 99.9% saturating the target, which is very similar to what Testfire is saturating the target at, at 210 milligrams, which is their approved dose. And so ultimately, there probably is not an efficacy play in the clinical trials. So this is really a convenient play. Now, every six months versus every one month, we believe, is material for patients and for physicians who are seeing their patients every six months anyways. Now, ultimately, the other interesting piece we're shown over time, you may see an efficacy benefit, not necessarily in the clinical trial, is Klaxo had shown about a year ago that there's only about a 20% adherence rate to biologics for patients. And so if patients are not adherent to their medicines, they're going to unfortunately resort back to some of the symptoms that they had in exacerbations. And so with a very convenient play every six months, if the patients are much more adherent in the real-world setting, you might actually see a much better efficacy benefit over time.

I think Jason's point there is really important, Sean, because, you know, what you see in the sort of domains is, you know, massive cost to the health system of noncompliance. And the fact that this drug will line up with, you know, a severe asthma patient's, you know, normal visits gives you a strong underlying driver of maximal adherence that could ultimately lead not only to an efficacy benefit but also a cost savings to the health system.

Yeah, I think it's a super important point. and just wonder what your view is on how well investors understand that. So while you're saturating the target and maybe you can't really improve upon biomarkers and outcomes in that sense, but certainly improved efficacy does derive. Improved compliance does derive better efficacy and then ultimately better efficacy, less exacerbations, less hospitalizations, less cost to the system. And do you think that's really grasped by investors?

I think it's partially grasped. I think, you know, the nice thing is there's a lot of analogs in both kind of the immunology markets, so markets like psoriasis. We've watched them mature over the last, you know, 20, 30 years where, you know, first-generation therapies were more shorter-acting, longer, you know, second-generation, third-generation therapies were longer-acting, and you saw these sort of compliance benefits. I can go back to, like, domains like osteoporosis. You know, I worked on Fosamax earlier in my career, right, which was a daily and then a weekly oral pill. Amgen came along with Denosimab, a six-monthly biologic, and you saw exactly the same dynamic where, you know, in the clinical trials you could not show a difference in hip fracture rates, but in the real world you saw a significant difference in hip fracture rates. And so it's those sort of dynamics that ultimately both, I think, investors need to understand, but also importantly payers, you know, physicians as well as kind of the patient at the end of the day.

And, Sean, I think what's probably not well understood by investors or appreciated to this point with GB0895 is the fact that we moved from Phase I to Phase III and did not do a Phase II efficacy trial with exacerbation. And some of the points that Mike made earlier kind of justified the move that we made. But in reality, given we are following the same pathway as tezapelumab, we hit the same epitope as tezapelumab, our biomarker data is as good, if not better, than tezapelumab. We have full saturation of the target at our 300-milligram dose, and this is not a unique pathway Glaxo took it with their IL-5 and, frankly, is now taking it with COPD as an indication, moving right to Phase III without Phase II data. It's very highly probabilistic in our view that this will be a strong, efficacious drug. And so I think investors have yet to fully grasp the fact that the probabilities of the Phase III success are probably very, very high. Yeah, I'd agree with that.

It's just so it brings me to, I'll just throw this question out. It wasn't on my list, but it's just given what you just said, Jason, like what keeps you awake at night on the trial? Like what could go wrong? Whenever you put a drug into humans, there's a lot that can go wrong.

The reality of it is, you know, the things that keep me awake are, you know, we've seen you need to have very balanced enrollment geographically. Standard of care of disease like asthma varies by geography. And so, you know, and we've seen a number of different trials that, you know, in Eastern Europe, you're not showing a difference between, you know, placebo and active arms in recent trials. And so making sure you have the right caps, the right adjudication procedures for entry criteria, you know, the details really matter on these sort of trials, Sean. So, you know, making sure that we do that very well, making sure that it's stratified appropriately by EOS level. You know, one of the benefits that anti-TSLP medicines have is that it works for all comers in asthma, but you want to have the right proportions in the EOS less than 150, the 150 to 300, the 300 and greater. And so how you actively enroll this trial to make sure that you have the right balance ultimately determines the outcomes in these sort of trials. And, you know, I think that only becomes more and more complicated as you go into other diseases like COPD, where the heterogeneity of the disease is even greater.

And I think the other point, Sean, is this is a very competitive space. And so speed matters. Getting to patients faster is really important. So the speed in which we execute on enrollment in the trial, this is a 52-week endpoint, so getting to enrollment faster will get us faster to market. And there's a lot of other companies out there which are behind us on the long-acting molecules. Now, we think we have the best long-acting molecule of a 98-day half-life, which is meaningfully better than any of the other long-acting next-generation TSLP, anti-TSLPs. We do have that binding affinity of 20-fold improved on affinity than tezapelumab, so 100-femtimolar binding. So we feel also that in addition to the long half-life, we'll be grabbing the cytokine tighter for a longer period of time, which will enable us to have the benefit for patients out to six months, which is not necessarily clear for some of the others. And so speed is critical to us as well.

Well, it flows nicely into the next question. So it's area one and two. So what can you tell us about the rate of recruitment and your confidence on completing and getting to top line? I think guidance is first half of 28.

Yeah, so the studies are off and running. You know, as of Friday, you know, we had regulatory approval in 39 of the 42 countries in which we're recruiting. You know, we're seeing uptake across all regions. You know, the team has done an extraordinary job standing up to replicate 786 patient trials. And so we feel like, you know, we're on track with all of our prior guidance in terms of having, you know, enrollment completed by the end of 2027, data toward the end of 2028 given the one-year endpoint. points, and all of those details that I had just mentioned a moment ago, Sean, have been kind of front and center, right? So we're making sure we're recruiting broadly geographically, but also making sure that we're getting the right types of patients in the trial. One of the key entry criteria that we looked very carefully at was having patients have two or more exacerbations, documented exacerbations, in the year prior to the start of the trial. And ultimately, you know, we believe that sort of an entry criteria will ultimately show meaningful therapeutic effect.

Sure. Thank you. And what level of exacerbation reduction do you think you might have to show to say you've got a competitive drug? I think TESI showed a 70% reduction over 52 weeks. So how should investors think about that as a benchmark?

Yeah, so I think if you think about the anti-TSLP space, and you mentioned the TESI data that was in the EOS greater than 300 cohort, right? So they showed a blended rate across all comers of 56%, 70 in the high EOS cohort, 40 in the low EOS cohort. I think that's kind of the, you know, TESI has sent the benchmark as the incumbent, right? And so I think, you know, somewhere in the 50% reduction across all comers and then pushing up, you know, I think you've seen with the IL-5s and with, you know, products like Dupixent and the high EOS cohort somewhere in that, you know, 50%, 60%, 70% range as well. And so, you know, I think that's kind of the mark that you're going to have to hit to be competitive. And so, you know, I think obviously doing that with also a great safety profile will be key to the success of the medicine.

And given the FDA feedback for our trial, and the reason the size of our trial is 786 subjects in two different studies, is to be powered to capture the below 300 EOS endpoint. And so we're 94% powered to capture the less than 300 EOS endpoint in terms of reduction. and that means we're basically 99% powered across all companies.

Can you remind us or tell us what the evidence you have around pharmacodynamic evidence that the T-slip suppression is maintained through the 26 weeks?

Yeah, I mean, the data we just showed at the European Respiratory Society actually shows that we're sustaining those biomarker reductions out now up to about a year. And so, again, Jason mentioned earlier the 98-day half-life. You're seeing, you know, 50% reductions across a number of the key biomarkers like EOS, IL-5, IL-13. You're seeing substantial reductions in pheno as well. And what's really been encouraging is, you know, at that 300 milligram dose, you're not seeing kind of a waning effect toward the end of the period. And so we feel, you know, very, very confident that the medicine is very active for, you know, beyond six months, actually.

I guess COPD is the largest swing factor in our valuation, and the market will read the ERS poster as a proxy for the multi-billion dollar opportunity of Biomacrum PK data align. Is that a fair basis, or does the real answer need a phase 2B?

No, I think, listen, I think, you know, as we've seen, you know, GSK just took a very aggressive approach going straight to phase 3 with no COPD data that we're aware of, right? So I think, you know, TESI showed a really interesting signal in their Phase II study. If you think about the currently approved biologics for COPD, you have Dupixin and Nucala. They are only approved in the greater than 300 EOS cohort in COPD, which is about 28% of the market. the data that TESI showed would take that down to about the 150 EOS cohort where you would add another about 40-45% of the market. So, you know, obviously the landscape in COPD is changing with the recent AstraZeneca data for the IL-33 where they were the first medicine to show a benefit in the less than 150 EOS cohort. And so, but, you know, we still think there's a huge opportunity. The COPD market's only 2% to 3% penetrated from a biologics perspective. As you know, in these sort of immunology conditions, you usually see when they're mature markets, you have a 40% to 60% biologics penetration. There's a lot of room to run. COPD is, depending upon which statistic you look, one of the top five burdens of disease globally. So the cost of the health system is extraordinary. The cost of patients' lives is extraordinary. and we think if you had somewhere, you know, the TESI data in the greater than 150s was about a 37% exacerbation reduction that's a really meaningful medicine for patients and so we think the opportunity as you rightly point out is profound and we think, you know, GB0895 is very well situated in that and the last thing I'd just say is the six monthly dosing may actually be more valuable in COPD given the frailty of the COPD population You know, COPD patients oftentimes are carrying multiple comorbidities and having a long-acting therapy that almost provides background protection for that population, you know, we get a lot of feedback from doctors and patients that, you know, would be very, very meaningful for their lives.

I normally ask these questions towards the end and want to give some time to, you know, some pretty exciting oncology programs, the MMA E1 in particular. But, you know, given Solaria 1 and Solaria 2, the data you've shown at DRS and COPD, I think you had $457 million of cash on balance sheet at the end of Q2. How do you think about the balance sheet and funding?

So we are funded through the first part of 2028, and that will carry us through the Phase I trials in oncology. It will carry us through the enrollment in the asthma trial, as well as a lot of the initiatives that we're doing from the platform side in the next generation of molecules that we're moving toward the clinic. There's a number of different paths that we're simultaneously pursuing from a capital perspective. Doing partnerships like our Amgen and Novartis partnerships is part of our core strategy, so we will continue to do partnerships like those. We've been very successful in terms of getting toward the end of resolution on most, if not all, of those targets and certainly will be over the next several months. that opens up a lot of capacity for us to do more partnerships. And so those types of partnerships, as well as achieving milestones on the Amgen and Novartis partnerships, will bring in non-equity dilutive capital. We'll certainly explore equity capital markets, as well as product-specific financing, potential current partnerships around some of our products over time. And therefore, we believe over the next 12 to 18 months or so, a lot of these pieces will carry us through not just asthma data but beyond. Sure. Thank you.

Moving on to GB4362, so the MMAE neutralizing program, maybe just give investors a snapshot of what that is because I don't think it might be broadly appreciated, particularly for people new to the story. And what clinical evidence do you need to show toxicity can be decoupled from efficacy? Yeah, it's a great question.

4362 is a program that is very unique, right? I mean, we talk a lot about medicines and, you know, you started the conversation with China. We're not aware of anyone else pursuing a concept like this. And so the beautiful thing about 4362 is the dose-limiting toxicity for MMA-based ADCs has been peripheral neuropathy. And this has kind of been seen in pretty extraordinary rates in the clinical trial setting. So in Fortimab, Vidotin, Padsev, Pfizer's extraordinary molecule for urethelial cancer shows about a two-thirds rate of peripheral neuropathy in that trial, which leads to about 20% of patients discontinuing. It leads to down-dosing, and it leads to dose holidays. Ultimately, what we're trying to do with this molecule is selectively bind the cleave payload without interrupting the intact ADC. And we have a molecule that can distinguish the cleave payload from the intact ADC because it binds the cleavage site, and that cleavage site is only exposed once the payload is circulating systemically. And so we think this has a huge opportunity to not only address this kind of core dose-limiting toxicity, but also extend the utility of all MMA-based ADCs. And so if you, you know, Pfizer had some really interesting data at ASCO in June, which showed a direct correlation toward survival on drug and overall survival. So if you're able to keep people on these ADCs, you're seeing extraordinary survival rates. And that's ultimately what we hope to be able to show, is that by reducing the neuropathy, you're reducing that dose-limiting toxicity. We think the endpoint that the FDA will look for is a reduction in neuropathy. And what you'll try and show is a non-inferiority from a survival perspective. Because you rightly point out, like, you know, what you don't want to do is interrupt the bystander effect. And so we're, you know, partially through the design of the molecule, partially through the work we've done in terms of dosing in a line to the second dose of MMA, we think there's a way to actually thread that needle nicely.

And what we've done in, so preclinically, what we've shown is we can reduce free MMAE by up to 80% in non-human primates and mice without impacting tumor killing, but with getting substantial reduction in skin toxicities and neutropenia, which are other toxicities that occur in additional peripheral neuropathy. Our phase one trial is designed to find the dose of our antibody that reduces free MMAE by 50%. The feedback from the FDA, which this has gotten fast-track designation and is in first-line therapy, you can go up to 80%, but we think 50% reduction is the right. It gives us enough room to not impact the bystander effect, but to reduce the peripheral neuropathy and other toxicities meaningfully enough that patients can benefit from continuing on the drug longer. Sure.

Thank you, Jason. And what milestones should investors look for in terms of data release around that program? cohort of subjects in the dose escalation portion of the phase one.

We believe, and again, the dose escalation portion is to find the dose that reduces free MMAE by 50%, and so we believe we'll have that dose toward the beginning of 2027. Once we have that, we're going to do a dose expansion study or cohort expansion where we'll take somewhere between 40 and 60 subjects, let's say, who already have grade one peripheral neuropathy on PAD-CEV and Keytruda that are urothelial cancer patients. They'll receive our drug, the MMA neutralizer, commensurate with the cycles of getting PAD-CEV and Keytruda to see if we could reduce the progression of grade one peripheral neuropathy going to grade two. Because once you hit grade two peripheral neuropathy, it's irreversible. And so that, over the course of 2027, we believe will have the data, which is the safety and efficacy data we're looking for to be able to potentially move directly to registrational trials, as Mike mentioned, where safety would be the end point. Sure.

And what relationship might you have with PAD-SEV and how do you fill out or unfill further MMA-containing partners?

Yeah, so there is no formal relationship. The beautiful thing, this is just simply layered on top of that standard of care. As we look forward, though, I think there's two different pieces, Sean. You point out, like, ultimately the way you will develop this, we've done what Pat said, because that is the leading MMA based ADC. You'll do a basket trial, likely, with all MMA based ADCs to kind of show that this works across the whole continuum. And then, you know, I think in the future there's an opportunity to potentially co-formulate, you know, which would be, you know, I think really interesting as well, where you could co-formulate 4362 with the MMA-based ADCs.

Is there anything specific about PADSEV that might mean that you're, compared to other MMA-containing ADCs, that might make it more robust in combination with 4362 about just soaking up the free MMAE, or you think it's broadly applicable, your confidence level around that?

In the preclinic, it's very clear that it's broadly applicable. So we think that, again, it's partially tied to the linker and the payload. So a lot of the early Seattle Genetics ADCs all kind of used the same sort of core technology there.

And philosophically, at the beginning of the conversation, we're talking about platform, pipeline, pipeline, platform, that I think many people might look at Genera and think it's a respiratory company and not really an AI company and that the molecule could have come from the wet lab, could have come from AI. We don't care as long as it works. you get a payback. But the way I'm thinking about this, this is potentially a step in the validation of the platform to investors. So maybe put some meat around the bones on that.

Yeah, I mean, I think, you know, one of the things that we've spent a lot of time thinking about, Sean, is actually, I think, you know, it's easy sitting here today to think, oh, well, where is the AI? The real question we had to answer with GB0895 was, would computationally generated molecules be immunogenic right that was that was the unanswered question that nobody had ever tested because there had never been a truly computationally generated protein that entered the clinic what we've seen you know now across the five programs that have entered the clinic is that uh you know the these are very well-behaved molecules the ada rates have been very very low and so in some ways you know for us what you know from a platform perspective 895 de-risked that question across all of our programs. Now, as you rightly point out, where we're going is to these more complex biological tasks, like distinguishing this intact ADC from the circulating MMAE. And you'll see that, I think, across the next generation of programs. They're taking on more and more complex domains of biology that I think traditional tools have not allowed us to drug.

Sure, sure. Maybe an opportunity, you know, just given we've got limited time yet, I just sort of wanted to touch on your Novartis and Amgen partnerships and maybe give investors a flavour on those partnerships and, you know, what could we expect in milestones, et cetera, or announcements on candidates.

Yeah, so the way these are structured, each of Novartis and Amgen, Amgen back at the end of 2021, Novartis late in 2024, so almost a couple of years ago, We agreed on a select number of targets. We generate molecules that ultimately get to either lead candidate criteria or DC nomination based on some pre-specified criteria. And then once we get them to that level, they take time to verify it, and once they verify, they pay milestones, and then ultimately they have the choice to move them forward into the clinic, and they would pay milestones over time and potentially royalties. Milestones are in the mid to upper $300 million on each program, and the royalties are some mid-single digits up to low double digits on each of those programs. We've been extraordinarily successful in prosecuting across all of the targets. There are a total of nine targets between the two of them at this point. Some of them we will get to the criteria, but they will not be moved forward because, frankly, biology that they could not test for decades, let's say, or more, we finally were able to get them a molecule to test the biology, and the biology doesn't work. So I think we were very successful in getting there, but ultimately, if the biology doesn't work, then that comes with them forward. We have received the first milestone on the first program from Amgen already. We believe we'll hit multiple more programs on theirs, and on Novartis, we've been extraordinarily successful, even much faster, just given that we'd signed that deal a few years after Amgen, so had many more people at the company, a lot more capabilities. And so we believe we'll also hit multiple milestones. The timeline on verification is kind of the biggest kind of impediment to receiving those milestones, but now we believe over the next 12 to 18 months or so we'll receive multiple milestones in both of them.

So one of the really cool things, though, about the partnerships is Amgen and Novartis are probably two of the best protein engineering companies on the planet, right? You know, certainly in the top 10, right? You know, and so when they hand you tasks that they can't solve, you know, this goes again to this question of China. If the best in the world can't solve these sort of things, you're basically finding complex biology questions that if you solve them for them, we then have the right to apply that across other domains. So one of the tasks that both of them had given us was, can you stabilize a native hormone and drive a thousand-fold improvement in selectivity? Neither of them had solved that for almost a decade. And we gave Novartis 40 molecules that had over a 10,000-fold improvement in selectivity in four months. And so, you know, the power of these technologies once you get them up and running is really, really profound.

Well, gentlemen, we're right at time. Is there anything I didn't ask that I should have or any message you would like to leave investors with before we call a close?

I think, you know, I think we've covered a lot of ground here today, Sean. I think the most important thing is, you know, I think the point you made at the outset, you know, Generate is a company that with our first manifestations, We're showing, you know, clinical proof of concept for these AI design molecules. But we're really only scratching the surface. And I think, you know, as we continue to push the frontiers, what we're going to find is that there are a whole series of undruggable domains that we have never had the tools to prosecute that will unlock biology in really meaningful ways. And we're excited to be part of that journey for the industry. Well, thank you, Mike. Thank you, Jason. Appreciate your time.