Good afternoon, everyone. Welcome to the Jefferies Global Healthcare Conference. My name is Tim Odutola, member of the Jefferies Healthcare Investment Banking team. It is my great pleasure today to introduce Sean McClain, CEO of Absci. Thank you.
Great, thank you. Yeah, I'm Sean McClain, founder and CEO of Absci. We are a generative AI drug creation company. What we're ultimately looking to do is use generative design to ultimately tackle the hardest problems in biology. We have a proven platform with a purpose-built team. We have an integrated data flywheel where we're able to have a wet lab in the loop, being able to rapidly go from data in the wet lab to training our models to validating this. This all happens within a six-week time period. This has enabled us to have leading AI models, getting the best AI talent to ultimately come to Absci and build breakthrough models that have allowed us to ultimately tackle hard-to-drug epitopes, allowing us to create first-in-class, best-in-class assets.
We have, in a very short amount of time, built a world-class pipeline that's focused on both I and I, as well as oncology. Our first lead asset, ABS 101, is focused in IBD. This asset recently just initiated our phase I trial, and we'll have an interim phase I readout on that the second half of this year. Our second asset, ABS 201, is anti-prolactin receptor antibody. The indication we're pursuing on this is androgenic alopecia, so common baldness in both male and female. Additionally, we have an early-stage pipeline, ABS 301 and ABS 501, which are both oncology assets. We started off as a data-first company. The technology was founded on a technology that was able to scale protein-protein interactions, how antibodies interact with targets of interest. Looking at the affinity and where the antibody binds, looking at the functionality of antibodies.
We were able to scale this from traditionally screening thousands of antibodies to now screening millions. We have been able to create what we call a wet lab in the loop, where in a six-week time period, we can go from data in the wet lab, this protein-protein interaction data, to training our models, to then using that same technology to ultimately validate the models to see how accurate the antibodies the model is creating. This wet lab in the loop that we have created has really helped us recruit the best AI talent because they are able to rapidly iterate on the model designs and architectures and have ultimately led to the breakthrough de novo models that we have today. We are utilizing these models to create novel and differentiated therapeutics. Ultimately, we want to use this technology to tackle the hardest problems that exist within this industry.
We want to go after these hard-to-drug or undruggable targets, such as ion channels and GPCRs. In a recent partnership with Astellas, we were able to show that we could take an ion channel that had been known for a long time, but the industry had been struggling to drug it. We were able to show that we could actually use our de novo model to actually drug this particular ion channel and block it. Additionally, we're wanting to leverage these models to have precise engineering and enable smart features, being able to have, let's say, an antibody bind to the tumor microenvironment versus healthy tissues. We can engineer in agonism versus antagonism. This is how we're leveraging our platform to create novel and differentiated therapeutics. I'll talk about two case studies on our two different AI platforms that we have.
We have our de novo antibody design model, where we're designing antibodies from scratch. And then we have our AI lead optimization model, where we have an antibody, but we want to improve the features of this antibody. First, on the de novo design, I talked about drugging hard targets. Another example of this is recent data we released at our R&D day. This was in partnership with the Gates Foundation and Caltech. Researchers, Dr. Steve Mayo and Pamela Bjorkman, were able to discover the Coldera region of the HIV virus. This is a highly conserved region within the HIV virus. The idea is if you could drug this, you could create a neutralizing antibody across all the different clades or strains of HIV. The difficulty drugging this Coldera region is that it's a very, very deep crevice, hence the name Coldera.
No traditional technology had been able to drug this. We were able to use our de novo AI model to design the CDRs that could bind deep within that crevice. We were able to successfully show that we could design an antibody to the Coldera region where other technologies were unsuccessful. Moving on to the AI lead optimization model, we recently came out with data showing that we could use our lead optimization model to be able to hone in an antibody to bind in a pH-dependent manner. We wanted an antibody that could bind in the tumor microenvironment, an acidic condition, and not bind to healthy tissues at neutral pH. Indeed, we were able to show that we could have this pH-dependent binding that we were able to achieve through our lead optimization model.
I like to think of this industry as a team sport. It takes a lot of different players to come together to ultimately get a drug approved. We have partnerships on the AI drug creation side, where we're helping partners such as Almirall, Merck, or MSK design drugs to difficult targets and co-develop these assets with large pharma. They're able to bring their domain expertise on the biology. We're able to bring our AI de novo capabilities, and we're able to discover potential first-in-class, best-in-class assets. Additionally, we have partnerships on the data and compute side with partners such as AMD and Oracle. We have a multilingual team here at Absci that really has led to the success that we've had at Absci. We're in this really exciting time in biology and medicine, where we're seeing biology merge with tech.
A lot of times, when you bring two different cultures together, it can cause a little bit of chaos. We have been able to figure out how to recruit the best talent on both sides and be able to work together. You have to have scientists that understand how to develop models, but then you also have to have folks on your team that actually know how to run and design clinical trials. We have a team across the board that can go from the drug creation to successfully running these models. This is both on our leadership team as well as the board and SAB. As I mentioned, we are building an industry-leading pipeline. Our lead asset, TL1A , just entered the clinic for IBD. We have our anti-prolactin receptor antibody for androgenic alopecia that is going to be entering the clinic early next year.
We additionally have two oncology assets that we're working on as well, as well as a slew of other early-stage pipeline assets that we have yet to disclose. Diving in a bit into our potential best-in-class TL1A antibody, we used our generative design capabilities to develop a next-generation TL1A antibody that we believe has best-in-class potential. First, we were able to use the platform to increase overall potency. We were able to show recently in NHP data that we were able to achieve greater than 3x target engagement versus leading TL1A antibodies at an equivalent dose. We see the ability to have sub-Q formulation as convenient for patients to be able to enable at-home self-injections.
Additionally, we've seen great durability and safety in our 13-week GOP talk study, and we're able to confirm the once-quarterly dosing, which we believe that we'll be able to see translate into our phase I study. Additionally, we are expanding the TL1A franchise. We've been able to leverage our AI model to go after a new novel target. This is a known target with known biology, but has been difficult to drug. We are using that antibody along with the TL1A antibody to make a bispecific. These are two pathways that we see as synergistic. We are advancing this bispecific as well as the monotherapy to a lead. You'll be hearing more about that here in the future. As I mentioned, we recently just started our phase I trial.
At the end of this year or the 2nd half, we'll have a phase I interim readout. Be expecting that end of Q3, beginning of Q4. Now going on to ABS 201. This is a really exciting asset in the category of hair regrowth. There hasn't been much innovation within the hair regrowth market. You have minoxidil, finasteride, which have limited efficacy and not great durability. What we sought out to do was to see great hair regrowth and strong durability. That's what we believe ABS 201 delivers on. Additionally, there's a massive unmet medical need here. 80 million patients in the US alone suffer from androgenic alopecia. Let's dive into the mechanism. If you look at the follicle life cycle, it starts off with the anagen phase. This is the active growth phase of the follicle.
Ultimately, the follicle goes into the catagen state, where you end up having apoptosis and regression. What ends up driving the follicle to the catagen state is actually prolactinemia, so the buildup of prolactin. For those that have androgenic alopecia, they have increased prolactin levels on the scalp, keeping the follicle in the catagen state. We believe by blocking the prolactin receptor, you're going to be able to shunt the follicle back into the active growth state, into the anagen state, and be able to ultimately have that follicle resume growth. One of the areas that we're excited about is that we're actually seeing strong durability. The genetics of the anagen state shows that you can stay within that anagen state, depending on your genetics, from anywhere from 2-6 years.
We believe six months of treatment could lead to durability of anywhere from 2-6 years. That is what we actually see in an NHP study that was done. How does this translate in terms of animal models? There was a study that was recently done in stumptail macaques, where an anti-prolactin receptor antibody was examined in stumptail macaques. This is a really great model because it is a naturally occurring model. These stumptail macaques naturally go bald. What you are seeing here is the tops of the heads of these bald monkeys. You have both the male and the female. You can kind of see the bald heads at the beginning of the study and the gray hair. Treatment was for 28 weeks. You start to see hair growth. Then treatment stopped.
What was really remarkable was post-treatment, you continue to see hair growth and durability. One of the things that we also got really excited about from this study was not only did you see the hair regrowth, but you actually saw the hair follicle go from a gray color to their naturally colored black hair. We actually believe that pigmentation could be additional upside in this particular mechanism, being able to go from your gray hair to your naturally colored hair. This is a naturally occurring model. That is what gives us really strong confidence in this overall mechanism. Additionally, we did a study comparing ABS 201 to minoxidil. This is a shaving study. What we did was we shaved these mice. These mice are naturally in the catagen state.
When you shave them, it takes a while for them to enter back into that anagen active growth state. You can see when you compare ABS 301 versus minoxidil and the control, the hair regrowth is far superior. We're achieving efficacy that's far superior than what you see in both the control as well as minoxidil. Additionally, we released NHP data, where we were able to show we could have extended half-life. We believe that likely the dosing interval will be six months. You'll have two to three doses over that six-month period based on the half-life we're seeing in NHP. We did see really great bioavailability in our sub-Q dosing of greater than 90%. Additionally, we are seeing really great manufacturability and developability. We think that we'll be able to achieve 150 mg/mL sub-Q concentrations for this particular antibody.
Additionally, when you look at Hope HMI-115, which is another prolactin receptor antibody, what you see with that antibody is that they are likely going to need 6-12 doses to achieve the efficacy that's needed. Due to the low formulation that they have, it's likely going to be two injections per dose. You could have actually 12-24 injections, which ultimately is not commercially viable. Based on this data, we actually believe that we'll be first to the market for an anti-prolactin receptor antibody. Yeah. Just to recap, we see ABS 201 representing a significant untapped new market opportunity in hair regrowth. The treatments that are there are currently not durable. The efficacy is hit or miss depending on the patient. There's significant upside.
We see this being at least a $14 billion-plus a year market. Based on the durability that we're seeing and the efficacy so far in the animal models, we see additional upside to that $14 billion. That does not take into account the potential in the recolorization as well. Yeah. Just to recap, we have an exciting next 12 months in terms of catalysts. We have ABS 101 readout at the end of this year for our phase I. We're looking to enter the clinic end of this year, beginning of next year for ABS 201, with a POC interim readout on that the 2nd half of next year. Additionally, we have guided to a new large pharma partnership this year, which should bring in significant upfront capital, which would help extend overall runway. We're on track to get that executed this year.
It is an exciting 12 months, very catalyst-rich, and really being driven by the AI or the generative AI design platform that we have built. Thanks.
Thank you, Sean. We'll now open it up to general Q&A if any questions.
Question about the model. Is that developed from scratch in-house? Maybe some detail of the type of model you use?
Yeah, absolutely. It is a diffusion-based model that has been developed in-house. It is our own proprietary architecture that we are using. In addition to the proprietary architecture, we do have proprietary data that we are utilizing. We have our data that we are generating in-house that is sequence function-based. We have structural-based data that we are leveraging that is open sourced. The last piece of data that we are utilizing is synthetic data that we are getting from MD simulation. All three of those pieces are really key to being able to increase the generalizability of our models.
Would you kindly elaborate on the study design of the phase I, like where you're conducting it and what you expect to get out of it, that you have an interim plan?
Yeah. Currently for 101 and 201, they're both being run in Australia. For ABS 101, we are looking to achieve similar target engagement that we saw in the NHP study, confirming low immunogenicity. We'll be looking at ADA. Last is showing the extended half-life, which we believe we'll be able to achieve once quarterly dosing.
Thank you. I'm not an expert, but how much product can you generate with your platform in terms of dosing, in terms of availability of the product? To generate an antibody, I'm not sure how do you translate from an algorithm into a viable antibody that's injectable into human beings. Sorry.
Yeah. Yeah, absolutely. That is where kind of this wet lab in the loop comes into play. You have to be able—it's not about just finding a binder. You actually need to figure out, does it have the functionality you want? Does it have the developability? Can you formulate it? Can you get the right titers? We have a whole screening cascade and technology where we're able to first identify, does this bind to the epitope that we want? Then we go into, does this give us the function? We have in vitro screening. Then we go into in vivo modeling. That is all done in-house at Absci, except for the in vivo portion of the work.
I will say, working on these hard-to-drug targets like ion channels and GPCRs, those screening cascades and technologies we've had to actually develop ourselves because those are actually extremely difficult to set up. That is a core piece of the technology. It is not just about generating the antibodies to these hard targets. It is actually then being able to effectively screen those hits as well. We have proprietary technology we've developed to be able to do that.
Can you talk a little bit, the two questions, about the history of 201 and this target? And anything besides sort of animal models that might give you confidence in the target? The second question is around TL1A. It seems like it's getting to be a crowded space. Do you think that there are still potential partners out there?
Yeah, absolutely. With TL1A, we do not—yes, it is crowded, but we think three large pharma out there still need a TL1A in their pipeline. I think that there is also room for improvement in terms of efficacy and patient convenience as well. In addition, the combo-based approach we really believe in. The bispecific we are developing along with that monotherapy, we actually believe is an additional advantage for the TL1A asset that we have, and I think is gaining a lot of interest from pharma. In terms of 201, that originated from our Chief Innovation Officer, Andreas Busch, when he was looking at this target for endometriosis at Bayer. He worked on this early on at Bayer, and it was a serendipitous discovery.
When they looked at the mouse study that they had performed, the mice that had the anti-prolactin receptor antibody grew back fur faster than the control arm, which led them to dive in and find out that the prolactin receptor was involved in the hair regrowth cycle. Bayer did not want to pursue that asset, so they out-licensed it to Hope Medicine, which I referenced in the presentation. When Andreas Busch ended up coming on to the team, that was the first asset or the first target he wanted us to work on. I'm sure glad he pushed us in that direction.
Is the idea of the interim in the phase I because of potential partnering? What's next? Is it a 1B or 2? Will you bring it back to the U.S. for U.S. study sites?
Yeah, that's a great question. We do see the 1A as a potential place to partner the asset to large pharma. We are fully prepared to pursue the 1B 2A. We have plans to pursue that. We think that that could be another opportune time to partner that asset. With regards to 201, we actually plan on taking that all the way through approval ourselves. We see that as a very large market opportunity.
All right. Great. Thanks.