Good morning, everyone, and thank you for joining me at the second day of the Needham Healthcare Conference. My name is Gil Blum, and I'm a Senior Analyst here at Needham & Company. It is my pleasure to have with me today Sean McClain, the CEO, and Zach Jonasson, CFO of Absci. Actually, I'm going to start with something that's not part of my regular schedule, something I've asked every company today. Any exposure to tariffs, your thoughts? How does this influence the company, if at all?
Yeah, from the review that we've done to date, so far, none of the tariffs that are in place are going to affect Absci and what we are doing. I think that, even given the market dynamics, we feel very strong with the upcoming catalysts we have, as well as where our runway's at to be able to execute on those catalysts and be able to be in a great position when the markets ultimately do turn. Right now, it's heads-down execution and doing everything we can to get to those catalysts as quickly as possible and drive to those value inflection points.
Another kind of general question as it relates to changes in the FDA and your conversations with them. Has there been any influence there or are things as usual?
Yeah, so one of the benefits of our current strategy right now is that both ABS- 101 and 201 are going to be run in Australia. With that, we see that actually being a benefit given the current changes that you're seeing with the FDA. By the time we run our phase II in the U.S., hopefully the FDA is all sorted out. I think the current strategy we have in place and running in Australia, I think, is quite a benefit given the current dynamic in the FDA right now.
With that, Sean, I think we can dive into the company and Absci, its main technology, and what do you view as the value proposition here?
Yeah, absolutely. We are a generative design company that's focused on using generative AI to design antibodies to difficult-to-drug targets, being able to really unlock new biology, being able to ultimately use design to test hypotheses you couldn't have tested before. The current R&D process for drug discovery is really this trial-and-error process where you're searching for a needle in the haystack. Where we want to be is being able to actually design the needle instead of searching for it with all the attributes that you want. We've been able to show to date that we can go after these really challenging targets, like our collaboration with Caltech going after the Caldera region in the HIV virus, a region that has never been drugged before.
We've been able to show in partnership with Almirall that we can target and block an ion channel, which to date, there has been no approved antibodies to ion channels. In our partnership with AstraZeneca going after a membrane-bound target that they had struggled to drug. All of this is allowing us to create novel and differentiated biology, again, through that design. We're not focused just on better and faster. We're focused on new novel biology. I think you're going to start to see that in our pipeline that we're building out. I think just a really exciting time where we sit in being able to merge AI and biology together. We have partnerships with both large pharma, smaller biotechs, but we also are developing out our own internal pipeline as well that's focused in I&I and oncology.
We have an anti-TL1A antibody that's going to be entering the clinic any day now with a readout in the second half of this year. Early next year, we'll have ABS- 201, our anti-prolactin receptor antibody for androgenic alopecia and endometriosis that will have a POC or an interim POC readout next year. We're excited to be pushing those programs forward.
There are quite a few of these tech bio AI plays out there. They are kind of broadly divided by target modality. It is either a small molecule or biologic. I am really interested as to how you view your key differentiation from other platforms out there.
Yeah, so it's really kind of four different areas. It's the data, it's the compute, the kind of multilingual talent that we have, and then the actual models that we're building. This is all centered around our lab-in-the-loop process where we're able to rapidly generate protein-protein interaction data, essentially antibody functionality data, use that to train our models. We are able to then go and use that same technology to actually validate how accurate these models are. We can rapidly iterate on the model designs and architectures to figure out what's going to create the most generalizable model for us. Those four components plus that wet lab in the loop and that ability to not just rely on in silico metrics, but to actually get experimental validation and being able to use the wet lab data for training is critically important.
When you look at these LLMs that are out there, these chatbots, they're trained on the whole internet. Biology, we have a small fraction of that that we can train on. We have publicly available data, structured data that we do use. We have our own internal data that we're generating. One of the new areas that we're going into that I'm really excited about is the synthetic data generation through MD simulation. Using MD simulation to get additional data that you can train on so we can have this multimodal approach. Those are kind of how we see our differentiation. Additionally, I think folks sometimes think that you can create these large foundational models that can be applied to every single modality.
What we found is that you have to really fine-tune these models for the particular tasks that you're trying to solve, whether it's for antibody engineering or enzymes or small molecules. The niche or the area that we've focused in on is antibody design and everything that is all the data we're putting in is around antibodies. We believe that you're going to have very specific models for each different type of use case versus one kind of generalizable model at the end of the day.
In your view, and considering that computational powering has been increasing and the models have gotten better, do you think that there's a point at which companies that don't even produce any wet lab data whatsoever could be effectively competing in this space?
Those that do not produce wet lab data, you're saying?
Yep.
Yeah. I believe that you have to produce wet lab data. You have to be able to generate your own data if you are going to compete in this space. You have to have this wet lab in the loop. I think those that are solely computational, that are solely only leveraging what is publicly available, the models alone will not get you the generalizability and the outcomes that you're looking for. I think it's critically important to have the wet lab, to have technology that generates the data, no doubt about it. We are seeing that, and it's critically important.
If you had to name some other company in this space, who would you consider the closest competitor?
Yeah, I can.
Oh, go ahead, Sean.
No, go for it, Zach.
No, I was just going to say, on the small molecule side, we're definitely following what Recursion and Isomorphic are doing. There's less visibility on the programs at Isomorphic, a lot more at Recursion. We do appreciate Recursion is trying to build their own internal data, getting back to your prior question. We hope they're successful. There's so much opportunity in this space. There's so much room for multiple winners. On the large molecule side, I do believe that we are well in the head, in the lead in that space, particularly with all of the recent work where we've shown an ability to de novo design antibodies against these really challenging targets that were undruggable previously.
The HIV Caldera, which Sean mentioned, the ion channel, the work we did with AZ, I think from my standpoint, it's just been really exciting to see the evolution of our models. We do keep an eye out for other players in the antibody area, including companies like BigHat and Generate. I think we're solely focused at Absci on improving our models right now at antibody design, particularly now focusing on these really challenging targets where we think the biggest value proposition lies.
Zach, maybe staying with you for a sec, walk us through the Absci business model.
Yeah, look, it boils down to creating great assets, either first in class, and as I mentioned, increasingly focusing now on these undruggable targets where the biology is known, but traditional antibody discovery technologies can't address. I think that's where you're going to see our early pipeline heavily focused as we start talking more and more about that over the course of the next year. We do this in the business in two different ways. We design antibodies for our own internal programs, which are wholly owned. Those provide the most upside to us and the greatest ROI for Absci and our investors. In that case, we're looking to develop those internally to a proof point where we can structure a transaction with a large pharma company that brings in a significant amount of proceeds and upfronts, downstream payments, royalties.
On the other side of the business, we're, again, generating novel first-in-class or best-in-class molecules for partners. In that avenue, we're either working on targets that a partner like a Merck or an Almirall bring to us. Those are structured in more traditional business arrangements where we're upfront payments to do in our research funding as well as downstream milestones and royalties. Those bring some non-dilutive capital in, which is nice. There's less upside because the payments are all structured around a target that the partner's bringing versus us bringing an asset to transact on from our internal programs. More recently, we've done more of a hybrid where we're doing co-development partnerships. There, we're basically offsetting a cost structure on developing a program that's owned 50/50 with us and the partner. We're de-risking those on a capital basis.
If you look at what we're trying to do there, that's a way for us to go after additional first-in-class novel targets, but de-risk them on a capital basis. We have layered that into the business model as well. You can think of it back to the high level. We're using the platform to generate assets for our own internal programs, which we're going to transact on later. We're using it to develop and design assets in partnerships with pharma and other biotech companies.
Would you guys ever consider taking something to the commercial stage? I know we've talked about this in the past.
For the right opportunity, yes. I think that that right opportunity in our mind is ABS-201 for androgenic alopecia, which is common baldness for female and males. We see this as a very large unmet medical need. It's actually a relatively quick and simple clinical trial process through approval. I think that finding the right commercialization partner and looking to do that ourselves, I think, is a unique opportunity. That is definitely a program that we would consider doing that. Again, we kind of look at everything on a case-by-case basis. That's the one we have conviction in and something that we are definitely considering.
There's been some contention around the meaning of the term de novo in the context of AI-powered drug discovery. I'd like to understand your position here. Also, do you even think de novo, quote-unquote, is even necessary here? There's a lot of known biologic problems to solve for.
Yeah, absolutely. I mean, if you look at, let's just look at the facts. Two and a half years ago, we came out with the first manuscript on antibody de novo design, where we were able to show that we could design the HCDR3 to HER2. This is the first time anybody had designed the most variable region of an antibody completely from a de novo perspective. Since then, we've been able to not only design the HCDR3, but now we're to the point where we can go after hard challenging targets where there is no known binder. No one's been able to drug a particular target or a particular epitope. We've been able to use our de novo design models to actually, from scratch, design an antibody that hits that particular target or epitope.
We've been able to show that we can do this with the HIV Caldera region. This is the work that we shared at our R&D Day December of last year. This was in collaboration with Caltech, Dr. Steve Mayo and Pamela Bjorkman. We're the ones that discovered this highly conserved region within the HIV virus, this Caldera region. It's this very deep crevice, hence the name Caldera. No one to date had been able to drug this region. We were able to use the de novo model to actually drug that region for the first time ever. We've been able to show in partnership with Almirall being able to design an antibody that can block an ion channel. As I said previously, no one's actually been able to get an approved antibody that's blocking an ion channel.
We were able to do that with our de novo design model. Again, no known binders. We've been able to also do that in our partnership with AstraZeneca. We have shown that we can go from two and a half years ago of just one CDR to now designing antibodies where there is no known binder to actually hard targets. That is what we set out to do in the first place was ultimately, let's build a technology that can address these targets that we have struggled with. Let's figure out how we can use AI to actually create differentiated biology. The few, or there has been discussion on what does de novo mean and debates on the definition. We feel, and we've shown that with the latest results, we can de novo design it regardless of what definition you use.
More importantly, we're actually able to use these technologies to create biology that is ultimately able to address massive unmet medical needs. That's ultimately what matters at the end of the day, being able to create these differentiated assets that can address this unmet medical need for patients. That can be ultimately differentiated and monetized effectively for shareholders. We're extremely excited about where we stand. I mean, I honestly couldn't be more proud of the team on what they've built and how quickly they've built it since the first release of our manuscript two and a half years ago.
On a separate topic, you guys have guided or discussed potential for additional collaboration deals. Any additional details you can provide at this point? I'm assuming these are ongoing discussions.
Yeah, they're ongoing. The guidance in this case refers to a platform deal. We expect to assign at least one platform deal. We believe this will be a significant multi-target deal. Separately, we have discussions around ABS-101 that continue and have good engagement there with pharma. We will be deciding based on how those discussions go, when and if we transact on that asset this year. I would say we have discussions in both areas. The guidance you refer to, though, is around a platform deal.
Okay. I do want to move forward to some of the clinical programs. I think both are very interesting. I do want to start with a TL1A and an IBD, very crowded space. Maybe you can explain some of the reasoning around moving forward and how you intend to differentiate ABS-101 here.
Yeah, absolutely. First off, we think patient convenience is going to be critically important. We have a half-life extension technology where we believe that we're going to be able to dose once quarterly compared to the once monthly of the clinical competitors. One of the other areas that we have seen in the clinical trials is the increased ADA percentage, in particular, the Roivant study, which showed significantly elevated ADA rates. Now, only 10% of those were neutralizing antibodies, but that was a short trial. We do believe over time that the neutralizing antibody percent is going to continue to increase. By developing an antibody that could have lower immunogenicity, we see as being a differentiated feature that would be important for a chronic disease like IBD.
What we've found is that when we ran T-cell activation studies, we saw very similar T-cell response across all the different competitor molecules, including ours, which led us to believe that this was B-cell mediated or complex-driven. It was based on the epitope. If you looked at the Merck ADA response, it was much lower than the Roivant, again, which led us to believe that this was complex-driven. We used the de novo model design to design an antibody that was kind of around the Merck epitope to be able to decrease the overall immunogenicity that was B-cell driven. We showed some data at R&D Day, some in vitro data that supported this hypothesis. We're going to obviously see this now play out in our phase I interim readout.
The last piece of data that we have that we think is really encouraging compared to the competitor molecules is the target engagement data. We showed at JPMorgan that the NHP data where we did show increased target engagement at the same dose. We are hoping to see a very similar response for our anti-TL1A antibody in our phase I interim readout. From that regard, we do believe that from a next-generation perspective, we have a best-in-class, very compelling profile. If we look at where the industry is headed, especially with what you are seeing J&J and AbbVie do with combo, we do think that that combo-based approach is the future. That is what is going to really help address this heterogeneity that we are seeing in the patient population. We are in, there are three large pharmas that currently do not have a TL1A antibody.
That is J&J, Takeda, and Lilly. We think that this next-gen antibody that we have here, I think will be really nice in use for combo-based therapy. Additionally, one of the other areas that we are competing is kind of with a new bispecific we have created. Again, kind of going on this combo-based approach, going after multiple pathways, we have developed a bispecific with one arm being TL1A, another with an undisclosed target. This target is a known target and a known pathway for IBD. It to date has not been drugged. We have figured out a way to drug this particular mechanism. We see this being a very differentiated first-in-class bispecific program as well, which has definitely gotten a lot of interest from pharma.
Maybe a question that's a little challenging to answer, but how far do you think you're going to be going forward with ABS-101 before it's partnered?
Yeah, absolutely. We believe that after the phase I interim readout is a likely point in time to partner this. I think the other opportune time would be after phase I- B, where you are able to show some efficacy with patients. I think after the safety readout showing potentially lower immunogenicity, showing the target engagement and the half-life profile, I think will be an opportune time to be able to partner that out.
Just to remind us the timeline there.
Yeah, so we will have the phase I interim readout the second half of this year.
Okay. I want to shift to alopecia, also a very interesting program, maybe one that's going to stay around a little longer. Lots of excitement around the asset. Just walk us briefly through the development history.
Yeah. Yeah, I can walk you through the development and then hand it over to Zach to talk about the commercial opportunity. This is a really exciting program that actually our Chief Innovation Officer had discovered. He was actually doing research on endometriosis and had a very serendipitous find, very similar to Viagra, where when they shaved the mice, the mice that were shaved for this endometriosis study, those that had the anti-prolactin receptor antibody grew back their fur faster than those that did not, which showed that this prolactin receptor is involved in the hair growth. What we found is that you essentially have prolactinemia that ultimately builds up over time, which basically takes the follicle from this active growth phase in the anagen state to this catagen state where you get regression and apoptosis. The increased prolactin levels essentially keep it in this catagen state.
By blocking the prolactin receptor, you can actually shunt the follicle back into the anagen state, this active growth state, and actually get very great durability for two to six years. Essentially, when you're back into this anagen state, the follicle naturally stays in this growth state for two to six years. We have built out a really exciting program going after the prolactin receptor. We have done studies that have shown in mice superior efficacy to minoxidil. When we have done these shaving studies, we have been able to show that the prolactin receptor antibody that we have is able to regrow hair much faster than minoxidil. There are also studies that were shown in non-human primates, the stump-tailed macaques that go naturally bald. When you take these stump-tailed macaques and you dose them with an anti-prolactin antibody, within six months, they have full hair regrowth.
After treatment, which was incredible, they were monitored for the next four plus years. They saw no hair regression after treatment had stopped, which really shows this really strong durability. We see this as a real game changer in the hair regrowth market for androgenic alopecia. The other interesting piece is that we're seeing that the follicle, when it goes back into the anagen state, actually repigments the hair. You essentially go from your gray-colored hair to your naturally colored hair. We see that as additional upside. When we talk to dermatologist KOLs, everyone is talking about how important hair is to their patients and that there hasn't been good treatments out there that can really have this strong efficacy as well as durability. They, I think, see this as a really exciting new mechanism to go after.
There is a big unmet medical need. Zach, I think it'd be great to maybe talk a little bit about the market and how we view it and kind of the commercial opportunity.
Yeah, absolutely. I mean, the market is immense, right? There are over 80 million androgenic alopecia patients in the U.S. This is common pattern baldness for both men and women. As Sean mentioned, the treatment options today are very limited. You have minoxidil, finasteride. Both of those show very poor efficacy, very variable efficacy across patients. In the case of women, you do not even have access to finasteride. It is counterindicated. They also have side effect profiles, and they are a daily treatment. It is a regimen where you have to stay on that therapy lifelong. What we are developing here is a brand new category of therapy, which is what is really exciting that can address this 80-plus million patient market in the U.S. alone. It is a durable treatment. As Sean mentioned, this would be a pulsed therapy. You would have maybe two to three injections over six months of duration.
Maybe once every two months is our expectation. Could expect to see those follicles that are dormant or in the catagen or telogen phase move into the anagen phase. Experience stimulation back into the anagen phase, which could last two, four, five years for a human. A durable therapy that could actually regrow hair. If you put aside the pigmentation, which is a whole market expansion, we think that this is, on a conservative basis, at least a $14 billion market in the U.S. alone just for the hair regrowth. Pigmentation could expand that dramatically. The other point I would make that's really exciting to think about is not only do we have a straightforward clinical trial development path, which Sean mentioned, where we could get to a proof of concept in phase I.
Once we get through the clinical development, there's a very active, diversified provider network that's ready and primed to go. If you look at the dermatologist office, the number of dermatologist offices, plastic surgeon offices, and med spas in this country, there are more of those by a large margin than there are Starbucks. They're all looking for therapy, particularly for hair. This product profile is well-timed for the market. It's a new category. We think, again, that this is a very significant opportunity north of $14 billion on a total TAM basis.
Excellent. I do want to focus a bit on the differentiation from the original bio molecule, which has been developed by a Chinese company. If you can kind of put that into context.
Yeah, absolutely.
Go ahead, Sean.
You got it.
Yeah, I think there's a couple differentiations that we see here that actually, I think, are really important for the aesthetics market. First is the overall half-life. From what we're seeing from the Hope molecule, you have to dose every two weeks. Based on the PK data that we're seeing so far, you're likely going to be dosing every or you're going to have two to three doses over a six-month time period, which we see as really important from a patient convenience standpoint. The other piece is the formulation. I think that we're seeing like 60 mg per ml, something that's really low, which means that you'd have a high volume for injection, which, again, from a patient convenience standpoint, we don't see being minimal in the aesthetics market. The last piece is the patent life.
The patent on this is going to be expiring soon. From kind of all of those aspects, we think that we definitely have a superior molecule compared to the Hope. Zach, did I miss anything else to add?
No. And look, I mean, just put it into context on the convenience. If you need six months of exposure, which we believe you do based on the NHP data that we've looked at, then that's 12 doses from the Hope. You got to start thinking about COGS and other issues then versus two to three for what we're developing. It is a major benefit not only to the bottom line, but also for patient convenience. I also, just to emphasize what Sean pointed out, all of these factors roll together. We think we're in good shape to be first of the market in the U.S.
Just reminding the audience, because we only have a few minutes left, if they want to ask a question, they can use the question box. I do want to spend a minute on other programs. You guys have quite a lot going on, very diverse portfolio, including HER2 lead. How's the company thinking of moving other assets forward given limited resources?
Yeah, absolutely. First off, we're laser-focused on ABS-201. We're planning on bringing in some of these timelines, which we talked about on our earnings call. Now we're on track to be in the clinic early next year with an interim POC readout second half of next year. With that in mind, we have various opportunities to monetize our assets like 101 that brings in non-dilutive capital to continue to push that program or to push 201 forward. We are going to be bringing in, as Zach mentioned, a large pharma deal, which will bring in non-dilutive capital. The ABS- 301 and 501 are being worked on with collaborators. ABS- 501, the anti-HER2 program, is being worked on in collaboration with Dr. Dennis Slamon on running these in vivo efficacy studies. That is minimal effort and cost on our part.
We see a partnership like that being able to, one, help validate the technology, being able to keep the spend low, and setting it up to be monetized before it hits the clinic. We see minimal cost too going into ABS- 301 for the validation. We do see strong interest in that. We did just show some target validation data, which is extremely compelling. Again, we see this being a large pharma asset. We have earlier stage programs that are in the I&I space that we're continuing to push forward. In terms of just capital utilization and just the overall market environment, we are very cognizant of getting the capital and the resources funneled into 201 to get those readouts.
I think because we have a diversified portfolio, we have a chance to monetize other assets, to bring in non-dilutive capital to really put the focus in the assets that we want to develop ourselves. We are kind of laser-focused and making sure that resource allocation is being prioritized to bring in the biggest near-term capital or to ultimately focus on the catalysts that are going to bring the most shareholder value.
Maybe looking a bit into the future, not next year, maybe five years from now, and given how target-agnostic the company is, do you consider eventually specializing more or you're going to continue taking this kind of one program at a time approach?
We are definitely going to continue to look at targets in biology that we think we can be highly differentiated, where our AI platform can ultimately create kind of this novel new biology, and where there is big unmet medical needs and large market opportunities. We are going to kind of take it case by case. I think one of the trends that we're seeing emerge is areas that are focused in what I'll call age-related diseases, women's health, aesthetics. I think that there's kind of some general trends that are emerging from some of the targets we're going after. We are going to continue to stay agnostic and utilize the platform to ultimately create the differentiated assets that can have, I would say, big upside scenarios.
Gil, I would just add, this is one of the areas where we're highly differentiated from, I'd say, some of our peers, is that we have that capability in-house, right? We have a great SAB, but we have Andreas Busch and the team under him. They have over 10 approved drugs. So when it comes to really thinking through targets and indications, we have that capability in-house at an A-plus level, right?
Zach, you get the last word. Given market conditions, cash runway, cash position?
Yeah, I think we've maintained our guidance. I think we've got good runway into the first half of 2027. As Sean mentioned, we're prioritizing 201. There's additional upside in that forecast if we're to do a transaction around ABS-101 or if we do a very large partnering deal around the platform. That would be additional upside to that forecast.
Okay. All right, guys. We're at time. Again, I want to thank you for joining us today.
Yeah, thank you so much, Gil.