All right. I'm just taking this moment to welcome you all back to the 40th Annual TD Cowen Healthcare Conference for probably the 800th time today. We're glad to have you nonetheless. I'm Brendan Smith, Senior Biotech and Life Science Tools and Diagnostics Analyst, and I am joined today by the CEO of Absci, that's Sean McClain, and the CFO and CBO of Absci, Zach Jonasson. Thank you, guys, for joining us. Yeah, thanks for having us. All right. Again, we want to keep it as interactive as possible, so we will be checking our phones. If you guys have any questions or anything, you want to email us, it's brendan.smith@tdsecurity.com. Maybe, Sean, could you just take a quick minute and walk us through Absci's AI-powered Integrated Drug Creation platform, really how it's evolved over time, and where you see the important points of differentiation versus other players in the space? Yeah, absolutely. It's getting close to 13 years ago that we founded the company. We founded the company on really a data-first principle. We ended up figuring out how to scale protein-protein interactions, how antibodies interact with a target of interest, essentially looking at their drug functionality. We were able to scale that from tens of thousands of data points to millions. That was right around the time deep learning was taking off at Google. They had the transformer that came out in 2018. It was this idea if you could take these transformers along with this data, you could really go from this paradigm where you're searching for a needle in the haystack to being able to create the needle, in our case, an antibody, and really being able to solve hard biology problems through solving the design of antibody-based drugs, being able to test hypotheses that you haven't been able to test before. I think it's obviously, and we're seeing it, it's really important to obviously decrease overall cost and time. We really wanted to utilize AI to tackle the hardest challenges in biology that still exist. We believe that that's what's going to ultimately increase overall probability of success. That's what's going to allow you to have differentiated assets where you can go after known biology that, again, has just been difficult to drug. I think a really great example of this that has gotten a lot of interest from our partners and investors is the work that we've done on HIV. We had a partnership with the Bill & Melinda Gates Foundation along with Caltech. We are working with Doctors Steve Mayo and Pamela Bjorkman at Caltech. They had discovered the caldera region of the HIV virus. This is a region of the HIV virus that is highly conserved. The issue with it is that it's this very deep crevice, hence the name the caldera region. No one to date has figured out how to ultimately drug this region because of how deep it is buried within inside the HIV virus. We were able to use our de novo design models to design an antibody that could actually drug and target this particular region of interest. These are the types of problems we're looking to solve with our de novo model. We've successfully applied it to other difficult-to-drug targets, like our partnership with Almirall going after an ion channel that hadn't been drugged for the past 30 years, being able to work with AstraZeneca on a transmembrane target that they've struggled to drug for ADC purposes. These are kind of all the examples of how we're utilizing this to really create differentiated assets. That's really our strategy, as well as building out our own internal portfolios, looking at known biology that's just been difficult to drug to ultimately create these differentiated assets. Yeah, I think that's a great introduction to a lot of different aspects that we want to discuss today with you guys. I guess maybe first, just on the platform there, AI is everywhere you look right now. Every channel you turn on, every program you see is AI. Even within healthcare, right? I guess first, I want to ask, what are people misunderstanding, particularly about your platform, or what's the biggest disconnect between what you guys are actually able to do, how you're leveraging it, versus what maybe some people who are less in the weeds on what it actually means, and how you're actually delivering on what the capabilities could be? Yeah, absolutely. I think one of the things that a lot of times is overlooked is the data advantage that we have. Ultimately, and you're seeing this with LLMs, like the accuracy of the models and how well they work are really dependent on the data you're feeding into these models. We have three sets of data that we're feeding in. The first is sequence-based data. This is the affinity-based data that we get, as well as the epitope data that we get experimentally in our wet lab. There is additionally structure-based data that we're getting from the public domain. The last piece that we're really excited about is this synthetic data. Actually using molecular dynamics to generate synthetic data for training. All three of these pillars are creating this multimodal model that is ultimately allowing us to, again, de novo design these antibodies from scratch to challenging targets. Again, each one of these components that we have is helping us increase the overall generalizability of the model, as well as the overall accuracy. Obviously, we are innovating very rapidly on the modeling side. I think we have an amazing AI team that's helping push the frontier there. Again, the gains that we've seen is really driven by the data that's being fed into the model. I don't know, Zach, if you have any. Yeah, I mean, just building on that point, I think one of the things that's maybe not always appreciated is the pace of improvement in the models. It's all driven by the ingredients that Sean mentioned. If you look at what we were doing three years ago, we could de novo design a CDR. Today, we're doing all the CDRs in a novel framework against a target that you can't address with a traditional approach. It's really exciting to see how fast the models are improving. We're really excited to deploy that capability to generate these differentiated assets. Sean, you touched on some of the partnerships that you signed a little bit earlier. I want to kind of double-click on one of these. You have a new partnership with AMD in particular. Maybe just give us a sense of what is the driving rationale behind that, and what do you see as that unlocking for Absci over the near term as that partnership kind of gets up and running? Yeah, absolutely. I think it really comes down to the memory on the AMD hardware. I think we're seeing that in a lot of different applications, including the recent reasoning models that these are very memory-intensive workflows. We're seeing the same thing on the protein design front. One of the areas that we got excited about with these AMD chips that have the higher memory capacity is the ultimate training resolution we're able to get in. What do I mean by this? Essentially, when you train models on your protein on lower memory chips, you are only able to model a certain portion of the protein. You essentially have to crop the protein for training. If you're able to have a higher memory chip, you can actually give the model the full context, the full protein structure. More context going into the model means more information. The more information you have, the higher accuracy your models are going to ultimately be. Training resolution is really important for the particular application that we have. The other area that memory comes into play is being able to increase the batch processing that we can do, which ultimately gives us just better efficiencies across the board, whether in training or inference. We believe that this is ultimately going to decrease our overall compute costs over time as we continue to scale. These were some of the applications that we saw as being beneficial. I will note that, I mean, we still have our partnership with NVIDIA. I think that we're not tied to one particular hardware provider. I think that gives us, I think, really great flexibility to utilize each of these companies' chips in the most effective way possible for Absci. Yeah, I think it's a really helpful context when you're thinking about how you're building up the pipeline, right? To that point, look, I've been having multiple conversations today already about just I think we are obviously in the early days of AI integration into all of this, and to the point that some people are still figuring out what is the right or the best business model of applying it, right? Yeah. Some people want to have an entirely wholly owned drug discovery, drug development pipeline on their own. Some people want to just license out the software. Some mix of both. How has Absci decided on effectively kind of being a sort of engine for the broader sector? I guess, why does that make sense given the capabilities of the platform itself? Yeah, do you want to tackle? Sure. I would just point out we've evolved our business model so that it's in lockstep with our capabilities, right? As we brought in a true drug discovery team with Andreas Busch, who used to be Head of Discovery at Bayer before that, we built out that team. That allowed us to have the confidence to develop our own assets, select the right targets. As the AI platform has improved, we're going after more and more difficult targets, more and more differentiated targets. In general, the approach is to use the platform to partner with pharma to work on their targets, and then to use it to develop our own assets internally, which we will then, in most cases, partner at a later stage of development at a proof point. I think we've definitely learned the lesson, too, that I fundamentally believe a SaaS model is never going to work in pharma. Pharma is just never going to pay the value that's needed to make a scalable SaaS company within biotech. I mean, we didn't try a SaaS model. We tried a partnership model. Even there, you leave a lot of value on the table. Actually being able to create your own assets and being able to partner that, that's where we've seen where pharma is ultimately going to step up and pay is being able to create the differentiated assets. Again, I fundamentally believe that the value is in the assets that you create. The AI is a tool. That's where we're seeing the most value generation is building out our own internal pipeline and partnering those assets. I think that that's where we find the scalable model and ultimately where the value lies. That is a great segue right into some of the value cases for some of your individual assets. Maybe let's start with 201. This is a drug that is androgenic alopecia. You guys announced this at your R&D day in December. A lot of excitement for the asset and, frankly, the opportunity since then. Maybe just help us level set. Why does this mechanism make sense here? What is the actual addressable opportunity here? What are people not quite grasping about what you are able to do with this drug? Yeah, so this is our flagship asset that we're extremely excited about. It's going after androgenic alopecia, just common baldness that affects 80+ million Americans. The opportunity here is very large because there's a big unmet medical need. Patients are looking for solutions that can actually have durable hair regrowth. That's exactly what you see with this anti-prolactin receptor antibody. We're targeting the prolactin receptor. What this does is it basically shunts the follicle that's in what's called the catagen phase, where you have hair growth and where you have apoptosis and regression, which is essentially the start of your baldness. You're able to basically shunt it back into the anagen phase, which is the active growth phase where it stays there for anywhere from three to six years and actually get the hair regrowth. We have seen both in the mice studies as well as NHP, where when you dose the mice as well as NHP with an anti-prolactin receptor antibody, you get the hair regrowth, which is really exciting to see. We presented some of this data both at JP Morgan as well as our R&D day. The other area that we are really excited about is not only the hair regrowth, but the repigmentation as well. Both in ex vivo models as well as in vivo studies in NHP, we were able to see the repigmentation. Basically going from your gray hair to your naturally colored hair. We see this as, again, a very large market opportunity. I think very, very conservatively. We have estimated it to be at a minimum $14 billion a year. That assumes an 11% conversion rate, which is what you see with Botox. We assume a $1,500 treatment cost as well with that. We see that potentially being much, much larger. I don't know, Zach, did I miss anything there? No, I just say we think that's the low end, right? We think that there'll be significant more demand for this, and the conversion rate should be much higher. We think the pricing probably will be much higher, too, because it'll correspond with the durability of the effect. If on average a patient's seeing three to four years of durable regrowth of hair, we think that's significant economic value. Yeah, I guess that this kind of ties right into the development path for this drug too, right? I think we've talked about this a couple of times now. It's likely being pretty underappreciated. You speak to the demand, right? Not just the market size, but the actual demand, the kind of lack of clear and safe alternatives out there. I guess, excuse me, sorry. When you look forward to development pathway, first of all, are you planning to develop this internally yourself? Yes. What is it about the market that would potentially make this a faster time to market and potentially more appealing to you guys to develop? Yeah, yes, we're absolutely planning on developing this ourselves. We have no intention of partnering this at the current moment. We want to take this up through approval. The great thing from a clinical trial perspective is even in your phase I, we can get efficacy in that and a POC. We plan to power up our MAD study to be able to get a POC at the end of that. By being able to get your POC in phase I, I think that really accelerates time to near-term value inflection. The plan would be to likely do a combined phase II, phase III with, I think, approval in, call it 2029. I think you're able to get to some pretty meaningful value inflection points in a very short time period. I don't know, Zach, if you want to talk about the cost associated with that compared to other indications. Yeah, I mean, one of the really nice things about this program is the trials are very manageable. The endpoints are objective. You're looking at hairs per square centimeter. You're looking at density. You're looking at color. All of that's an objective measurement. What we've heard from the KOLs is that as we get into the phase II and beyond setting, there'll be waitlists of patients to get into these trials. Recruitment won't be an issue. As you're recruiting, you're basically going to be building market awareness, too, for the launch. Everything lines up very well for trial execution. The pivotal trials will probably likely need two of those in the order of 200-300 patients. Again, the measurements are very straightforward. I think this is a very clean path. It's something that we feel comfortable we could execute. It seems like a pretty obvious option for you guys to pursue, right? What has the market missed here? Why has somebody else not stepped into the space? I mean, for one, I don't think pharma was looking at spaces like this. In fact, the mechanism here was actually uncovered at Bayer originally, and they didn't know what to do with it. It wasn't in cardiovascular. It wasn't in one of their defined indications. It was also discovered serendipitously, right? They were looking at endometriosis at the time. I don't think anybody's looked at it from the angle of hair growth, this target. I think the timing is perfect, given where the market is and the understanding of aesthetics and also the provider networks. If you look at the number of medispas and derm practices in the U.S., there are more of those than there are Starbucks. The go-to-market and the provider capability to network into patients exists today. That didn't exist in the same way five years ago or 10 years ago. I don't know if you want to add. Oh, no, I was just going to say it was actually our Chief Innovation Officer, Andreas Busch, it was him and his team that ultimately discovered this mechanism. When he came on to Absci, he was like, we need to pursue this particular target. I think, again, with kind of what you're just seeing with GLP-1s and how kind of everyone is kind of focused on wellness and aesthetics, the medispas, I think that and I also think you're seeing a lot of pricing pressure caused from Medicare and Medicaid. Having this be out of pocket, I think, is starting to become really attractive from a large pharma perspective. I think, again, I think there's a lot of things in terms of timing that we're hitting this just right. I think it's an exciting opportunity. Yeah, and just on timing there, now give us a sense of 201. Where are we in terms of entering the clinic? When might we see initial clinical data? Yeah, we are targeting to be in the clinic early next year. We are looking to accelerate things as quickly as possible. We are potentially looking to have a POC readout next year in the MAD study. Fingers crossed. That is, I would say, kind of rough ballpark of what we're shooting for there. All right, great. I did want to shift gears a little bit over here. Can I just one other comment on this? We just announced this program at the end of the year. We just started having discussions with analysts like you about how do you model this and also having discussions with investors who are just starting to do the homework. I think one of the things that's exciting for us is we've got this great KOL network around this, but it's not yet really priced into our stock. I think part of our job is to make sure we articulate this story and bring the KOLs out to talk about the big patient need. We're very excited about the potential for this program. I'm putting you on the spot because I know you're going to. No, that's great. It's part of our job, too. At what point do we start? Yeah, look, I also wanted to spend some time on TL1A, right? Because this is a heated space, lots of competition there, too. Lots of excitement for the target as well. Maybe give us a sense, where are we with ABS-101 today? When could we kind of get next data and how you see potential points of differentiation here versus some of the later stage players that are already in the clinic? Yeah, I think if you look at what we've been able to generate preclinically, I think we've been able to show when compared to the clinical competitors that we do have better biodistribution. We've been able to show better half-life for patient convenience and then also a better ADA profile. We think that all of these contribute to a potential best-in-class TL1A asset. If you look at the market right now, I mean, we're firm believers that combo-based therapy in this area is really the future. You have J&J that's kind of leading the charge on this. They have the DUET trial that's going to be reading out soon. I think you've seen AbbVie start to go into the combo-based approach. Everyone, I think, is seeing that TL1A is going to be an important pillar of this combo-based strategy. I would say those that are focused in INI and IBD in particular, there's probably three-plus large pharmas that don't have a TL1A asset. I think that we're really set up, I think, really nice in terms of our profile, I think, to fill the need for likely one of these three large pharmas that don't have a TL1A antibody. Additionally, I think what's also gotten some of these pharma excited not only about our TL1A antibody, but is also an early stage TL1A bispecific that we have as well. Again, kind of our belief in these combo-based approaches, we're approaching it from a bispecific standpoint and kind of pulling in the other arm to this. It came from where it's coming from our de novo model. This target's been, I think, a really interesting target in the INI space. It's a known target. One of the issues with this molecule and why it's never really advanced or why no drugs have advanced for this particular target is that you get agonism, but at high doses, you get antagonist or you get agonism, which you don't want. We believe that this is a really interesting target. It's not a target that everyone's talking about. It's not an IL-23 or an alpha 4 beta 7. We think that this is a very differentiated target. I think it's a really exciting bispecific. I think that plus just the profile of the TL1A antibody by itself, I think sets us up to be really differentiated and execute on a nice partnering transaction. I guess just remind us, when do you think we could see for the not the bispecific, obviously, but the 1st- gen, let's just call it that? How are you thinking about what do you need to see in that data versus what's already kind of been presented out there from other TL1As to really kind of get them to the finish line? Yeah, I can take that. I think, first off, we're going to see data in the second half of this year. We're going to get an interim readout. We'll be able to show target engagement, and we'll be able to show the PK profile, so the half-life. If that data looks like what we've shown in the NHP, I think we've accomplished something significant. We're also hoping to get at least an early signal on a lower ADA profile as well. All right, great. I know there's still a few other things left to talk about, and we only have a few minutes left. I did want to shift gears to the rest of the R&D day. Also discussed, you had this 501, ABS-501. Can you just give us a little bit of a sense how does 501 compare to trastuzumab and how you did design the molecule with that kind of in mind in the context of your existing pipeline? Yeah, so this was actually early work that we had done on our de novo model where we designed the HCDR3 of trastuzumab to bind to HER2. We got some really interesting results that ended up coming out of this model when we went and tested these antibodies that came out of the model in some in vivo settings where you saw trastuzumab resistance. We were able to actually, with these molecules, overcome this trastuzumab resistance. We did this in collaboration with Dr. Dennis Slamon out of UCLA, who's a renowned expert on HER2. He was the one that did some of the early work on Herceptin. He was really shocked that we were able to, with these antibodies that came out of our de novo model, be able to overcome this trastuzumab resistance cell lines. We think that there's some really, I think, exciting opportunities to be pursued in some trastuzumab resistant patients as well as potentially in HER2 resistant populations as well. Okay, and I know just in the last couple of minutes here, I did want to just give you a chance. With everything now, we've covered a lot of ground, right, with the platform, with a few different assets here. Just kind of let's sum up the calendar for the next 12 to 18 months. In chronological order, what are kind of the most important inflection points for Absci now moving forward and things that, in particular, people should be paying attention to as longer-term drivers of value? Sure. Yeah, I mean, look, we're giving guidance. We're going to sign a significant partnering deal with a large pharma on the platform. That's something we're actively working on today. This year, we'll have an interim readout on ABS-101 in the second half. We'll be initiating the 201 study either likely early next year with the potential to get some proof of concept by the end of the year. Because, as I mentioned, as Sean mentioned, we're going to power that study to move from the SAD into a MAD in a way where we could get an interim look that I think is going to give us a POC readout. I think that's very exciting. In addition to that, there's additional upside around a potential transaction around some of our assets, including ABS-101, our TL1A asset, and potentially around other of our earlier stage programs. All right, I think with that, we're just about at time. I want to thank everybody for joining us today. It's always a pleasure to see you guys. Thanks for coming out. Enjoy the rest of the day and the week. Awesome.