Great. Thanks for coming today. Before we get started, I just have a disclosure statement to read. For important disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. Great, thank you all for coming. Great to have you here, and Sean and Zach, appreciate you making the trip to New York. Yeah, just maybe we'll just, you know, have a couple questions that we can use to get the conversation started. Maybe to kick things off, can you just walk us through Absci's strategic journey over the past few years? I think this year particularly has been a pivotal moment for the company, and also say a few words about how you're thinking about the business today. Yeah, absolutely. So first off, we didn't start off as a AI drug creation company. We started off by developing a platform technology that allowed us to scale data, and in particular, protein-protein interactions. How antibodies interact with targets of interest. So what epitope does it bind to? What affinity? And that was right around the time transformers were taking off in 2018 with Google. And it was this idea of, if you could take this data with these generative AI models, you could really go from this paradigm in drug discovery, where you're searching for a needle in a haystack, to actually creating the needle, in our case, an antibody. And by solving some of these design problems in drug discovery, what you're able to do is actually start to unlock novel targets with known biology. So, look at GPCRs or ion channels that have been difficult to drug in the past, but you have known biology. You can now actually start to target those and create some really interesting opportunities. And so that's where, you know, we started to apply the technology, and strategically, we were only focused on partnerships. You know, we had partnerships with AstraZeneca, with Merck, NVIDIA, and it was ultimately to leverage our platform, and we'd get upfront milestones and royalties. And now we've shifted into developing out our own proprietary pipeline. We've realized that in order to recognize more value and value creation, building out our own pipeline allows us to do that. So we're now strategically looking at taking assets all the way to potentially Phase 2 proof of concept, and then out-licensing them to pharma. Mm-hmm. Zach, you know, can dive in a bit more on kind of the strategic roadmap from there. Yeah, sure. I mean, last year, sort of end of the year, we announced our first programs, and but those were a while in the making. We'd put the infrastructure in place, with the hiring of Andreas Busch, former head of R&D at Shire, and the team that he brought in. So we had the discovery team in place, and we'd worked a number of years in big pharma partnerships, really cutting our teeth. And so, I think the end of last year was a seminal moment for us, and you're gonna see more and more asset creation from Absci going forward. And as we go through and look at how we allocate resources, you'll see we're extracting efficiencies from the platform. We see that year on year, and we're taking those savings, and we're redeploying them into asset creation. And also, too, you know, how we see AI is it's really a technology that's unlocking value in the assets. But the assets themselves are the actual value. And so again, AI is a tool, and we're using that tool to create differentiated assets and a really differentiated pipeline, and that's ultimately where the value lies. Yeah. Great. And I think seeing kind of what's been happening, you know, over the past year or so, you know, just in the space in general, what do you think is the biggest challenge for companies that are trying to do what you do, which is using AI-enabled drug discovery? What's from both a scientific standpoint, but also from a compute and pure AI and ML standpoint? Zach- Yeah, I can take that, and before I joined Absci, I led two venture funds focused in this space, and I'd say what I see in that landscape is, you know, it's really a data problem, right? You have to have the ability to create enough training data to train these models. The compute is sort of, you need to marry that with the training data, but in our space, the training data is the real bottleneck. So you need to have that capability to generate training data at scale, and you need to have the capability to go do the validation of what the models are designing at scale. And at Absci, the way we think about data is not one-dimensional, right? We think about data as in terms of quality. We think about data in terms of scalability, so the quantity, and then we think about data in terms of usability, and I think the platform we have, it sort of harkens back to the origin of the company, delivers on all three of those dimensions, so we create training data that looks at the functional component of the antibody, how it interacts with a given epitope on a target antigen. We can do that at scale, and that's what's really enabled us to have success in designing against difficult targets and designing differentiated therapeutic assets. Yeah, it's, you know, as Zach pointed out, it's really this lab in the loop. It's being able to, you know, rapidly generate data for training and then being able to use that same platform for validation, and being able to do that in a very rapid time. You know, we can do it in a six-week time period, and I think that that's allowed us also to recruit some of the best AI talent that's out there because they can rapidly iterate on the model designs and architectures and do this in a way that you haven't really seen in biology before. And so that active learning loop is super important. That's an important metric for us. We continue to look at how can we decrease that, that time period, 'cause the faster we can go on that, the more generalizable these models can be. Mm-hmm. And can maybe we can spend a little bit more time on what allows you to be so efficient with the six-week iterative cycle. Can you talk a little bit about kind of your approach, specialization, and perhaps even talk about moat, which prevents others from replicating that? Yeah, it goes back to the original technology we had developed, which was producing antibodies in E. coli. So normally, you produce antibodies in mammalian cells, and you can scale that to maybe producing thousands of antibodies in a given week. But by producing antibodies in E. coli, you can do what's called a pooled approach. So, like, you can take your engineered E. coli in your test tube, where you have billions of cells. You can take a million-member DNA library that encodes, you know, a million different antibodies. You can transform that into your E. coli, and now in that population, you have every E. coli making a different antibody. So you've, you know, scaled from thousands now to, you know, millions to hundreds of millions of unique antibodies being produced in that single test tube. And then, what we've done is figured out how to then interrogate every E. coli that's producing a different antibody and look at what the binding affinity is to a given target of interest. And so we can then get the binding affinity data, and then that's ultimately what we use for training these AI models. And that same technology, again, can be used for the validation. We can actually validate up to three million unique AI-generated designs in a given week. But it's that technology that's really allowed us to drive down to this six-week iteration that we have, and again, I think has led to the success that we've had to date. How important is that in your kind of conversations with, you know, potential strategic partners? And I guess said in another way, what are some of the capabilities that really draw partners to Absci? Yeah, absolutely. I think one of the differentiating factors is our de novo design model where we can actually design antibodies from scratch. And so what do I mean by this? Essentially, what we can do is we can take a structure of a target, we can feed that into the model, and then we can specify the epitope we want the antibody to bind to, and the model then is able to design the CDRs that can then bind to that particular epitope of interest completely from scratch. And we've applied this technology to our own internal pipeline. ABS-101 is a great example of that, our TL1A asset. But we've recently now just applied it to a partner program, AstraZeneca. So yesterday, we made an announcement that we met a key milestone in our partnership, and that key milestone was actually developing some antibodies towards a hard to drug target, a transmembrane protein, an oncology target, where we used the model to design antibodies from scratch that could bind to epitopes of interest that AstraZeneca was interested in. And this was a really key milestone for us to achieve, and I think it's a really great validation of this de novo design platform that we have, and being able to achieve epitope specificity. 'Cause if you look at all the technologies that exist out there, whether it's AI or other biological technologies, none of these are able to achieve this epitope specificity. What we're seeing is that you can now start to unlock both known biology as well as knowable biology by going after, again, some of these hard-to-drug targets, like ion channels and GPCRs. And so, again, now we've seen both validation on from a partnership standpoint as well as, you know, using it in our own internal pipeline as well. Mm-hmm. Now, maybe, going back to a point that you made earlier in terms of, you know, the company's strategy and direction shifting a little bit and, more focus on... And we'll come back to the TL1A in a second. What was the biggest learning for you over the past year as you're kind of focusing more on your internal development? Absolutely. I can hit on this, and then, you know, I'd love to hear Zach's perspective on it as well. But the biggest learning that I've had is that you can make the greatest AI model to design an antibody against where you can de novo design an antibody against a, you know, any target that you want, but if you get the target wrong, the asset is worth nothing. And so you have to be able to have amazing drug hunters, people that really understand the target biology, that can take the technology that we are working on and apply it to relevant disease biology that's ultimately, you know, gonna create value. 'Cause the value lies in these assets, and it's getting the targets right. It's understanding that biology, and it's really having a team that's multilingual, that can understand both the AI as well as understand the disease biology. And so I think that's what we've really learned is that both are extremely necessary, and that's why we brought in Andreas Busch. I mean, a large pharma executive, he's had over, you know, 10 drugs approved under his leadership, and, you know, now he's able to kind of take this really powerful AI tool and apply it to targets that he's been interested in over the past decade that he hasn't been able to drug, and we can now kind of unlock new kind of novel biology. And I think you're gonna see, you know, you're seeing that in, you know, ABS-101, but we're also excited to, you know, disclose at R&D Day, ABS-201, which is an exciting derm target, as well as ABS-301. But, you know, that's, I think, a key learning that I've definitely seen. I don't know, Zach. No, I would completely echo that. You know, what we've done at Absci is build in those competencies so that we can bring these drug candidates forward through IND-enabling and into the clinic. So we've put in clinical capability. We've put in translational biology. We've put in all those elements to ensure that we can be successful with the asset development. Mm-hmm. Okay, thank you. I think one of the themes that's really resonated over this year really is how you, Absci, has been able to demystify what, you know, people think of as the black box that is AI drug discovery, and giving people something tangible to look at and also compare against other, you know, known assets going after the same target. Now, how, when you talk to investors, how do you help further, you know, articulate your ability to demystify your platform in terms of what you can do going after a certain target using TL1A, your TL1A experience as a case study? Yeah, absolutely. As you mentioned, I think, how we applied the de novo model to create a differentiated TL1A, I think, really, again, helped demystify the platform and really show how unlocking the design of biologics can create differentiated assets. And now, that's one example. I think now we've just come out with a second example with AstraZeneca, showing how we can, you know, take a transmembrane protein in oncology and design an antibody from scratch. Again, this is another great case study of how we've been able to unlock new novel biology, not only in our own pipeline, but with a partner. And then additionally, I think, you know, come R&D Day, you're gonna see how we've, you know, taken this and applied it to two other assets, which we're excited to unveil. And I think, you know, this is not just a TL1A story, this is a platform story, and I think you're gonna continue to see that with the pipeline that's, you know, unveiled at R&D Day. And I think you're gonna continue to see these wins, you know, be put up on the board over time. Yeah, I mean, I would add, like, one of the things we're having a lot more discussions about with partners or potential partners is sort of even going beyond the selection of epitope. We're now, like, I think, pushing the frontiers around understanding how the interaction with an epitope can actually drive the potency, and in some ways, sometimes modify the method or the mechanism of action. And so we've got our models now where we actually were able to show how we deployed that in TL1A as well, because there, we've specified an epitope for a good reason. We selected an immunoprivileged epitope, but then we used that interface model to really sample all these potential interfaces between the antibody and that epitope to uncover an adjacent epitope that actually delivers much higher potency, but which we also believe will maintain that immunoprivileged aspect, and then we've done some more recent work, which we have not unveiled yet, but really illustrating how landscaping at a global level to find the right epitope for biology, we've done that actually in our AZ partnership, and then at the local level, we've done some recent work where we've shown by defining a new interface for a very well-known epitope, we've been able to overcome resistance mechanisms, so I think we're really pushing the frontier around leveraging this epitope capability, and I think that's been really interesting for a lot of our partners to dig into. Mm-hmm. Yeah. I think, when we look at some of the others that are doing AI-based drug discovery, I think one of the challenges is that, you know, they have all these great insights that are coming from their AI and ML platforms, but ultimately, they are unable to make their own drugs. So I think that's a theme that's resonated well with, you know, the investor and some of your, the partner community, which is your ability to actually develop drugs, that are best in class. Now, coming back to, the IO- Can I just comment? Yeah. Ultimately, the only way to show the proof of the tool is to build the asset. Yeah. I think we recognized that early. So I mean, because there's a lot of in silico models that produce designs, but they're never... If you don't validate them in the wet lab and then actually validate them in humans- Right ... you don't have the validation. Yeah. Right. Yeah. Yeah, and I also think, too, developing your own assets, you're able to get a lot more value as well. Because if you look at pharma, pharma's only willing to pay a certain amount early stage, which is, you know, low millions. But if, you know, you can take it to a Phase 1, Phase 2, I mean, you're now talking of, you know, hundreds of millions to, you know, potentially billions of upfront payments and milestones that you could never get earlier on. Mm-hmm. And so we're able to, you know, we believe in the platform, we believe in the differentiation that we can create, and so that's why we've really shifted that strategy from, you know, just partnering to developing our own assets that we partner later in order to extract that value. But we also recognize as well, pharma's great at late-stage clinical development, they're great at commercialization, they're great at manufacturing. And so let them do what they're great at, and, you know, help them, you know, create, you know, pipeline assets. And the other great thing is, we're not competing with them. We're never gonna compete with large pharma, because, you know, large pharma knows that they can buy an asset from us at any point if they're interested. ... Got it. I think on that thread, where do you see the AI drug discovery field going in the next couple of years? I think ultimately, those that have to rely on an in-licensing strategy, there's obviously a cap and a ceiling to where they can go with their platform. Where do you see-- what's your vision for the next couple of years for Absci, but more- Yeah generally, thematically speaking, the field? Yeah, absolutely. You know, I think we're right now, you know, tackling, you know, easier problems with data with AI problems that where we have data. You know, right now we're tackling de novo design of antibodies. That design problem, being again, able to now go after these undruggable targets. But our models right now aren't predicting the biology. That's ultimately where we want to go in the future, is like, how can we start generating data to be able to start to predict the epitope that's gonna give us the biological response? How do we start predicting the target for a particular indication that we wanna go after? At least to the best of our knowledge, that data doesn't exist right now. I think over time, we are building it up, others are building it up, and you're gonna, you know, start to go from, you know, predicting antibody design to predicting, you know, epitopes that give you the biology, to actually then starting to predict the biology, itself. But that's a ways out, and again, it goes back to, you have to have the data to create these, you know, generalizable foundation models, and that's what we've focused in on. I think we're already seeing the impact by just solving this design problem. I think the future is extremely bright. And again, I think that AI is a tool, and how to... You know, it's figuring out how to use those tools to create these, you know, very differentiated assets and have insights that others don't have. Zach, if you have anything else to add? I just think you're gonna see, maybe this is more of a midterm, you're gonna see a separation of companies that are really doing AI, like we are, where it's integrated throughout the company, the data capabilities there, from the ones that are using it more as a marketing term. I think the proof is within the assets, and that's what we're focused on, is bringing those assets forward to show the proof points on how you use that AI tool to develop a different, you know, differentiated assets, and so I think you're gonna see a separation from the noise, and I firmly believe we're in the leadership position for antibodies, so our objective is to keep generalizing our models and extend that lead. Yeah. Great. Now, last month, you released results from your non-human primate studies for your lead asset, ABS-101. Can you share a little bit more on some of the key findings from those studies, and also how ABS-101 is best positioned to be a potential best-in-class asset? Yeah, absolutely. So we're able to show in the NHP studies, when we, you know, compare our molecule to the clinical competitors, that we can have an extended half-life of two to three X. And, you know, this is likely going to allow us to do dosing potentially up to once quarterly. And, you know, right now, the competitor molecules are doing dosing once monthly. So we see this as an extended half-life as a key differentiation. We're also able to target both the monomer and the trimer, which, you know, that could lead to potentially better efficacy in the clinic. We'll wait and see. And then also, going back to the de novo design and being able to hit the specific epitope of interest, we were able to hit an epitope that's similar to the Merck epitope, which we believe will allow us to have a lower immunogenicity response or a lower ADA response in the clinic. If you look at, you know, the Merck antibody versus the Roivant molecule, the Roivant molecule had a much higher ADA response than the Merck, and we believe that that's due to where the epitope is binding, so it's complex-driven or B-cell mediated. So we see that as another advantage. And then also, we're able to do subcu. We're able to show that we could formulate up to 200 mg per mL, which will allow us to do subcu dosing. Zach, did I miss anything else? No, I think we're well set up for differentiation. Yeah. Mm-hmm. Got it. Thank you. And you know, over the years, you've, you know, had a number of partnerships and collaborations, including the recent Sloan Kettering collaboration to develop novel therapeutics using generative AI. Can you talk a little bit about maybe that particular collaboration with MSK? And also, how does your platform provide value to partners, and how are you thinking about the various partnerships, whether that is of the early stage partnerships with someone like MSK or, you know, a partner, a large partner like AZ? Yeah, I mean, I can... Well, I'll comment on the MSK partnership and- Mm-hmm ... talk about the broader strategy. I think with MSK, we're really excited about that partnership. You know, on the face of it, they're covering half of the development cost. So for us, we can develop some novel assets at half the cost. That's great, and that means we can do twice as many of those for the same fixed pool of capital. But I think what's really more exciting and why we did this partnership, and why it's a six-program partnership, is there's a tremendous amount of synergy there, right? MSK is world leader in cancer research. They're gonna bring the target biology, and they're gonna nominate or they're gonna bring forward novel targets that we can jointly decide on pursuing, but they bring that biology. And then, on the other hand, once we've built in our side, once we design the asset or design the antibody against the target, MSK can then run and conduct the Phase 1 clinical trial. And so there's a tremendous amount of synergy in working with them to make things go efficiently and to leverage the capacities and competencies of both groups. And then ultimately, we're aligned at, you know, post Phase 1 proof of concept, we'll jointly market those for potential transactions with pharma. So I think it's a well-aligned partnership that leverages multiple synergies. Yeah, and then, you know, on kind of partnerships with large pharma, you know, we're gonna continue to pursue those, both on from a platform standpoint, but also from, you know, out licensing our assets. And, you know, from a platform standpoint, I think these large pharma partnerships really provide kind of three advantages. You know, one, they bring in, you know, upfront non-dilutive capital. It provides validation of the platform. And then third, it allows us to go into therapeutic areas that we are not experts in. You know, right now, you know, most of our pipeline is focused on cytokine biology, and in particular, focused in on I&I as well as oncology. But we're not planning on going into, you know, in, you know, areas like neuro. And so partnering up with a large pharma that has real domain expertise in neuro, you know, makes a lot of sense 'cause they can provide all that expertise. We can, you know, design, you know, antibodies towards those, you know, interesting targets and really start to create this diversified portfolio that we couldn't achieve on our own. And so we see that as being able to, again, diversify the assets that we couldn't do on our own. Okay, great. Thank you. Now, as we look forward into the future, what key milestones should we be looking forward to in the upcoming quarters with regard to your internal pipeline and- Yeah ... also future partnerships and collaborations? What can you share with us today? Yeah, absolutely. I think we're headed into a very catalyst-rich six to 12 months. So we just announced the NHP data for ABS-101, where we were able to show, you know, the extended half-life that we were looking for. That asset will enter the clinic early next year, and the key catalyst on that will be a Phase 1 interim readout, the second half of 2025. Then regarding ABS-201, that is a potential best-in-class derm target. We will be announcing that target, as well as the preclinical data at our R&D Day on December twelfth. And we'll have some KOLs as well that will come speak on that target as well. We see this as a really exciting opportunity, as it's a large market to go after. Then additionally, either at R&D Day or at J.P. Morgan, we'll be announcing the in vivo efficacy data on ABS-301, which is a novel IO target that came from our Reverse Immunology platform. So we're excited to be unveiling that, you know, those preclinical data packages at the end of this year. Then additionally, we do plan to announce more partnerships through the end of this year. I think we gave guidance on four new partnerships. We've announced one already and plan to announce another three partners by the end of this year. So I think again, from now through, you know, the end of next year, I think both on our own internal pipeline as well as with partnerships, we have some exciting catalysts upcoming. Great. How in terms of your you know, cash position, cash runway, and also talking about your strategic vision, where you know you require capital investments, can you talk a little bit around that? Yeah, so you know, we haven't changed any of our guidance. We have runway into the first half of 2027, and that allows us to prosecute the Phase I clinical development of ABS-101 and bring some of our other assets forward, so I think we're well positioned to execute on our goals and right now, we haven't changed any of our guidance with respect to our gross spend this year. We're still estimating approximately $80 million in gross spend. Obviously, on a net basis, we run well under that. As an example, for Q2, our net spend was roughly $16 million, so we're not running at a full $20 million per quarter, so I think we're well positioned, and the other thing I'd come back to, which I mentioned earlier, is we're seeing efficiency gains from the AI platform. And so what we're doing with those efficiency gains is really reallocating some of that resourcing capital into the asset development, and I think that's a really wise decision based on the return on investment there. Yeah. Now, maybe a wrap-up question: What is one thing that you wish people knew about more about when it comes to Absci, maybe for each of you? ... Yeah, absolutely. Yeah, I, we did a raise earlier this year. I think a lot of the focus was on TL1A. TL1A is a really exciting target. I think we have a differentiated asset. We have some exciting catalysts coming up, but we are not just a TL1A story. We do have a platform that can, you know, create other TL1A-like assets, and I think we're really excited about our R&D Day that's coming up on December twelfth to really talk about ABS-201 and ABS-301, and really show how this isn't a one-trick pony, but we can, you know, take this and apply it to other exciting assets, and we're continuing to develop these. I also think too, we're starting to really show the breadth of the technology, not only internally, but externally, and I think the AZ partnership is a really strong validation of that and meeting that milestone and, you know, having AZ elect to move forward. So, again, it's... TL1A is exciting. We're gonna continue to pursue that, but this is a platform story, and I think that that's the number one thing that I would like to get across to investors. Zach, I'll- I'll just emphasize that TL1A is just the beginning. Yeah. Great. Any questions from the audience? Well- Do you have any insight into additional partnerships or is that with the biopharma partners, and you don't have much visibility? Yeah. We try to make announcements on a regular basis as we achieve these milestones, but we don't. We're not giving guidance on when the next milestones will be hit, but we try to disclose when they are achieved, just like we did with AstraZeneca. And so I think you're gonna continue to see announcements like these in the future as the partnerships progress. Great. Well, Sean, thank you for coming to New York, and happy belated birthday. Ah, thank you, Rock. Really appreciate it. Glad we could celebrate together, and thank you, everyone, for coming today. Yeah. Awesome. Thank you. Yep.