All right. Welcome, everyone. Let's go ahead and get started. This is the Fireside Chat with Exscientia and Recursion. My name is Vikram Prahlad. I'm one of the biotech analysts with the Morgan Stanley Research team. Really quick disclosure, for important disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/researchdisclosures. With that, happy to have with me, Dave Hallett from Exscientia, Mike Secora from Recursion. Thanks for joining us. Appreciate it. Dave and Mike, the best place to start maybe might just be the recently announced business combination between the two companies. Would just love to unpack the rationale for that, what you see as the obvious areas of overlap between the technologies, and then we can go from there. Sure. Maybe I'll start, and then I'll- Sure. I'll pass over to Mike. The way I think about unpacking that combination is that at a very high level, you have Recursion, who very mission-driven organization, who are applying technology, generation of significant amounts of proprietary data and applying that to decode biology. And that, what that really means is to try and kind of glean unique insights and to identify sort of root causes of the biology that leads to disease. On the flip side, Exscientia, again, at a very high level, has been very much focused on the generation of small molecules against targets. The kind of idea being here that, if you look at one of the major reasons why small molecules fail, and therefore why it is so expensive to do small molecule drug discovery and development, is that they were poorly designed, and therefore issues were baked into those compounds kind of on the day they were designed. And so we've focused very much on how can we use kind of AI and technology and generative methods to improve the kind of probability of success. And so by bringing the two together, the two platforms together, you create a kind of an end-to-end kind of capability, where you've got the kind of unique insights around, say, around novel first-in-class targets, if you look at kind of Recursion's pipeline, and then couple that with the ability to then derive truly kind of best-in-class compounds from that. That's a very compelling combination. On top of that, so you've got the pipeline. So our two pipelines reflect kind of our approach so far. So Recursion kind of mainly kind of first-in-class, kind of novel biology, kind of quite a lot of rare diseases in the first version of their pipeline. We're very much around best-in-class, larger target space in oncology. And so the two actually stick together very neatly. There's no overlap there. There's orthogonal kind of risk and reward. So that helps to de-risk the kind of two pipelines. And then the partnerships. The partnerships is a very compelling proposition. We, Exscientia, have our significant combination with Sanofi, kind of, in the oncology and I&I space that's been going a few years now. We have a collaboration with Merck KGaA. Recursion, on the other hand, have two significant collaborations, one with Roche Genentech, which I'll let Michael talk about, and also one with Bayer. Last, but by no means least, just the people. We have a huge kind of talent base of people, so both within North America and in Europe. Post-close, I'm really looking forward to kind of the sort of power that those people can bring to bear in modern drug discovery. Yeah. Great. Mike? I very much agree with everything David said there. I think it really, for me, it really begins with the native respect and admiration that the two companies have had for each other for really over a decade. I think both companies, founded 2012, 2013 timeframe, were consequences of an acceptance of the unique problem that defines drug discovery and development, that it takes, you know, over 10 years to bring a new drug to market, takes over $2 billion, 10% likelihood of success. And for that genesis, I think it's given rise to companies like Exscientia, companies like Recursion, who have adopted novel approaches, expertise. I think in the case of Exscientia, it has been how to pursue precision molecule design, chemical synthesis, to generate best-in-class molecules. In the case of Recursion, it has manifested in a desire to explore biology, chemistry more broadly, and start to find first-in-class opportunities. And so you see this kind of philosophical combination of these two companies, where both, again, are the consequence of, I think, a lot of things converging, things like AI, things like ways to control biology, like CRISPR, things to control chemistry, like chemical synthesis and all the different kind of digital chemistry methods that have come in the last decade or so. And I think it's that cultural appreciation. It is, I think, the success that both companies have had, I think it is the mutual respect that we've been able to kind of garner with each other over the last, you know, few years. And, you know, Vikram, I think that for all those reasons, if you look across, you know, the, the orthogonal pipeline, best-in-class, first-in-class opportunities, approximately ten, readouts, you know, coming in eighteen months as we've disclosed. You look at the partnerships, I think some outstanding partnerships there with Roche Genentech, with Bayer, with Sanofi, with Merck KGaA, so on and so forth. I think. And we're seeing already delivering on our promises to our partners. You know, we had our first neuro map, about a month ago, where Roche, Genentech, paying us $30 million for that. You know, Exscientia team doing extraordinary work with their partners, advancing, you know, multiple programs across their partnerships. If I look across the platform, you know, I see multiple virtuous cycles that are now being integrated together, ways to explore biology, chemistry at mass, ways to design chemistry and produce it in mass, ways to invoke patient-centric data, and thereby, having all of those modules connected together, one has that full stack solution. Right, because I think, I believe, and I think a lot of us here believe, that it is only when you have all those modules connected, integrated in a holistic way, do you truly have that profound impact on the drug discovery design development process that is characteristic of, you know, $2 billion to spend, 10 years of development time to get to market, and 10% likelihood. And so, I think for all these different, you know, complementary reasons, pipeline, partnerships, platform, people, there's just also such great appreciation around what we've been able to develop thus far, and I think we're very much looking forward to what we can develop going forward. Great. And I guess with any combination, right, there's always a little bit of a time period that you need to integrate platforms, to integrate people, to integrate cultures. Kind of looking forward, how long do you think it'll take to be able to operationally fully integrate the two companies and have that kind of day one, where you're both operating as one, and then you can start ideating and generating kind of as one unified platform? I actually don't think it will take that long, and the reason I say that is that my personal experience of kind of mergers and acquisitions of this type is that the thing that always gets in the way of ultimately a successful kind of bringing together and a combination is cultural misalignment. Mm-hmm. And here, if you go to our respective websites and look at the kind of value statements that are on there, they're, they're almost identical. It's like we could have written them together, but they were written separately many years ago. Having kind of met Michael and Chris and Najat and the team during the diligence process and met them on the investor circuit, and now we're starting to kind of meet a bigger swathe of people within Recursion. I think that's reflected, like we're very aligned, we're very curious organizations. We are, we are both trying to fundamentally change how modern drugs are both created and developed. And so there's really good cultural alignment there. We're using this process at the moment between now and anticipated close. So, doing... Starting to do kind of the planning exercises that you're allowed to do. So we're getting to know each other in more detail. But I'm more than optimistic that kind of very rapidly after we actually close this combination, that you'll start to... the world will start to see and our partners will start to see the kind of benefits that Michael and I have alluded to in terms of the kind of benefits of the overall combination. Yeah, I completely agree with Dave. I mean, it begins with, do you have cultural alignment or not, and that is the foundation upon which you can build. I think both companies have been mission driven. Both companies have had extraordinary cultures, and based on that foundation, you can build something, I think, really important and impactful. I think with respect to, you know, right now we're in the midst of planning, and before we can actually close and then do the process of integration, which we've talked about giving guidance that we believe that's the early 2025. But I would also just call out that Recursion, this is, this will be the fourth time that Recursion has done M&A activity. You know, Vium, which gave rise to our In Vivo omics platform, the acquisitions of Cyclica and Valence, both for on the digital chemistry front as well as generative AI front. All those acquisitions, all of those M&A activities, all were predicated, did you have a foundation around cultural alignment? And I see the same here with respect to Recursion and Exscientia. And I think having that as substrate makes this, it puts this, I believe, on the best footing to carry out that planning, and then once we close, integrate, and start working together once that's done. Great. Great, that's helpful. Operationally speaking, I know you've mentioned, with the announcement of the merger, that you expect roughly $100 million in synergies per year moving forward. Where could those come from? I think they can come from a variety of sources. I think, there's the obvious ones, is that. As you probably, everyone in the audience is aware, there's a significant cost that comes with a NASDAQ listing. We take two of those and take them into one. Yeah. As we announced, pending close, our board will join Recursion's board, but so you take two larger boards, kind of, and, and shrink that, and the same with the exec. But I think a lot of it will come from operational synergies. Is that kind of what I learned from, from the diligence exercise is that, Recursion kind of spend a significant amount of money on, on CROs. So do we at Exscientia. That's kind of one of my, my... CSO, but in that hat, it's one of my biggest kind of line items. So there's a real opportunity here to kind of like, okay, how do we, how do we internalize that? How do we kind of leverage the kind of capacities that the two organizations have? Obviously, a significant kind of strength and depth on the Recursion side, particularly around digital biology. I think we've spoken before about with the automation studio that we've created in Oxford, starting to re-onshore the synthetic chemistry aspects that we've been outsourcing for so long. Mm-hmm. So I think a lot of it will come from operational synergies like that. Understood. Yeah. No, completely agree with Dave. I think, I think just to some, you know, high-level headline things, I mean, on Recursion side, do we have CROs working in, in some of the chemistry, medicinal chem, synthetic, you know, chemical synthesis? Sure. Is that something we can internalize with Exscientia? Yes, that would, that would be an easy, easy synergy. And I think on the Exscientia side, as they perhaps have had CROs, doing some work on in, you know, in vitro biology, in vivo biology, those are things that we can also internalize. I would also frame that if you think about the early stage of the pipeline, now you start to centralize early, development work around one, you know, one set of objectives. That also, you know, helps to kind of frame what you can kind of do going forward. And also, as we think about where could there be natural synergies between each other across the execution, the delivery for any of the partnership objectives, I think this all starts to drive, you know, these costs, you know, drive these cost synergies that we can see going forward. Got it. Got it. Understood. Going back to the topic of partnerships. You both mentioned the partnerships with biopharma companies you both have respectively at Exscientia Recursion. Through the combination, do both sets of partnerships just transfer over as they are right now? Or do you think the structure, the form, the nature of those could change, given that it's a combined entity now that the partner is dealing with? Operationally, the structures will continue as they are. That both Recursion and Exscientia are busy companies that busy pipelines, lots of partnerships. And so, part of the integration planning we're doing at the moment is to make sure that we don't disrupt that. The kind of partnerships, as Michael pointed out, are kind of critical to the organizations. They, both companies have always had kind of a partnership kind of structure as part of their business model. Kind of, I think we said in the kind of, in terms of some of the near-term details, the collective partnerships have the capability over the next twenty-four months of delivering north of $200 million in milestones. Mm-hmm. So, we're not going to do anything to kind of jeopardize that. In fact- Yeah ... The whole point of leveraging the combination, like how can post-close, how can Exscientia help the Roche Genentech and the Bayer collaboration post-close? How could access to capability at Recursion help us and offer our Merck deal, and likewise with the kind of, with our pipelines as well. Yeah, I think for both companies, I think we very much appreciate our partners. We appreciate working with them, and we want to deliver well for them. I think with respect to the contracts, I think that we have... You know, these are well scoped out as to what are the, you know, areas of focus, how are we going to deliver for these partners? And I think both companies have already been doing so, for all of our partners. You know, that being said, you know, we are certainly open to having a conversation to see if maybe some of these new capabilities help to evolve or expand the collaboration, and if so, you know, happy to kind of have that conversation around what a combined entity can also maybe do for a partner going forward. Understood. Understood. Okay. Just taking a step back then from the combination, maybe a bit of a foundational question for each of you. Dave, we'll start with you, maybe. There's a lot of talk in the investor community about trying to identify kind of proof points to kind of understand how a platform is working to develop the next generation version of drug development. When you look at the Exscientia platform, what you've been able to do to date, I guess, what are the one or two key proof points that give you confidence that things are headed in the right direction? Then, Mike, I'm going to ask you the same question right after. Sure, it's a really good question. I think I'll answer that in two ways. I think one is about kind of specifics of kind of delivery, and then I'll get into some detail, maybe one or two examples. Sure. So I think the kind of proof points for me are the small molecules that either we or we've helped our partners kind of take into development. That is, they were designed using generative AI, not kind of artisanal, kind of medicinal chemistry methods. And so across a kind of a broad swathe of therapy, as you think about the three molecules that we helped to design and discover for Sumitomo Pharma in the psychiatry space. You think about the PKC theta project that w as in-licensed by Bristol-Myers Squibb, advancing kind of through phase one. Think about the CDK7 asset that's now that we recently acquired full ownership of and are moving through phase one. Think about MALT1 and LSD1 that are literally on the cusp of going to the clinic. Is that I think we've proven on more than one occasion that so it's not a fluke, it's like it's reproducible, that we've been able to deliver high quality development stage assets with a significant kind of cost saving to get there. So 60% over industry benchmarks. We've been able to do it by actually performing kind of 80%-90% fewer experiments to get to those proof points. The molecules are good enough that they've actually got through kind of IND-enabling studies. So we're designing safe molecules as well from that perspective. I think in terms of some specific stories, I think PKC theta is a good example of where technology has played a role here. So this is a target space that is well known. Mm-hmm. If you look into this patent space, probably nearly twenty companies tried to get a subtype selective PKC theta inhibitor for inflammation use, and pretty much all of them failed, but we've been able to design a molecule, as I said, the BMS kind of been licensed, is progressing through our IND studies, and then one more example before I'll pass to Michael, is probably our MALT1 story. Where we use technology and the experts we have, we think really carefully about what is it... Ultimately, kind of we're thinking about the patient. How's the patient going to see this drug? How's the patient going to take this drug, so from MALT1, the story's always been the kind of clinical path here will be a combination, likely with a BTK inhibitor in CLL. You're giving two drugs, you're giving your own drug, the MALT1 inhibitor, you're giving the BTK partner. I think it's well known that a significant or the majority of BTK inhibitors that are actually approved, or even the ones in development, all have a drug-induced liver injury flag. We've been very cautious and very cognizant of that, thinking, "Okay, we can't add to that kind of liver burden." That's the story behind using technology to design on target potency selectivity, while being aware of where some of our competitors are in the MALT1 space, about avoiding exacerbating kind of potentially fatal liver injury. I think there are two specific data points where we've actually gone out ahead of time. MALT1, we actually, we told the world that we think J&J, for example, would have an issue with hyperbilirubinemia in the clinic, and they subsequently kind of told the world they did. So I think they're the kind of things that we've done today. I think clearly where we're going next is, okay, so we can design molecules that are safe enough to be tested in humans. Can you see efficacy? And that's kind of where the CDK7 combination trial is going. Obviously, it's kind of why we're pushing LSD1 and MALT1 into the clinic. Got it. Got it. And Michael, I'll ask you the same question- Yeah. And this might be a good opportunity for you to touch on the data release from yesterday on CCM. Sure, absolutely. You know, I'll touch on that as well as maybe talk on a few other points more broadly. But I think proof points are across what are the leading indicators, what are the statistics that you're able to see? You know, what's the pipeline showing? What's the partnership starting to show, and what's the platform starting to show? On the statistics side, one thing that we started to do since our IPO is we give out statistics around the time and cost to get to some development stage compared to the industry average. We often found ourselves to be half the time and half the cost. And also credit to the Exscientia team, who I've always enjoyed the shift the curve concept. It's precisely the same idea. It's... I would actually encourage more and more life science companies to actually frame the statistics of how they're operating, because this is a leading indicator. Are you having impact or are you not? I've always enjoyed that graphic. On the pipeline side, you know, I think that I look at them a lot of the novelty that Recursion is able to frame, where we're finding, in many cases, a target that is not known or well known in the corpus of scientific literature, or some kind of biological insight, and I think that characterizes all the programs that Recursion has brought forth in its internal pipeline. Perhaps, you know, the most recent program I would highlight is RBM39, where you're finding a novel target. In the context of CDK12 biology, you're designing a novel chemical scaffold. You are using patient-centric data to think about how to target patients, stratify patients, develop biomarkers, and so you're getting at that novelty, and I think that is across all of the programs that Recursion has advanced, where we've had a number of phase one successes. Just yesterday, we had phase two data come out for our CCM program, where the primary endpoint around safety tolerability met, finding a very safe molecule in the context of, you know, effectively, in, you know, compared to placebo, no adverse effects. And then also in terms of seeing multiple encouraging trends on the exploratory efficacy measures, particularly with respect to MRI-based objective measures around reductions of lesion volume and a reduction in hemosiderin ring. That's the leak that's coming out around the ring, and coming out around the lesion. And those lesions, you know, many experts are seeing that as the drivers of this disease, knowing that CCM, cerebral cavernous malformation, disease of the vasculature, disease of the endothelium. And very much, you know, seeing this data, we're very much looking forward to having a meeting with the FDA as soon as practical, to talk about what the next steps for development can be. And I think we also saw a time-dependent effect, meaning 12 months better than six months. So you're seeing the longer a person is on drug, you're seeing increased effect. You know, I think we have strong confidence we'll continue to advance this program, and we've had a lot of KOL support and a lot of patient advocacy support. So I think that's an important proof point, and I think also an important proof point, going back to the work that Chris and Dean had done, you know, back at the University of Utah during Chris's MD PhD work, and that, that kind of, that playing out, all that being a product of the Recursion OS from an earlier vintage. But you know, you see those also, you know, clinical successes where we had earlier data with C. diff, phase one data being positive and looking forward to other readouts we have coming as well. That's on the pipeline side of things, and I think more progress to be made. On the partnership side, you know, I look at how we've been able to serve both Roche Genentech and Bayer. As I talked before, we had a $30 million payment for how we've been constructing these neural maps with Roche Genentech. I think that's, that was an extraordinary feat to do, genome scale knockout in a neuronal cell type, iPSC-derived neuronal cell type, for which we have been operating as one of the world's largest producers of iPSC-derived neurons. And then also, you know, Roche Genentech optioned their first program last October. Mm-hmm ...a novel target in GI oncology, novel chemistry there. Those continue to progress. And also with respect to Bayer, how we've by the end of Q3, here, we're gonna have all 25 data packages to them, start additional program selection. And we also talked about how Bayer is gonna be our first beta user of LOWE, our sort of software engine for doing drug discovery. Not just kind of connecting different modules that can be called, but also be a collaborative environment to be able to do research together. And then lastly, I'll just say on the platform, when I look at the OS, the platform Recursion has constructed, when I look at the OS that the Exscientia team has constructed, I see automation. I see an ability to collect standardized, systematized data in a methodical way. I believe that both of these approaches, right, biological chemical exploration, precision chemical design, having been an investor in this space, scientist by background, I see this as being increasingly the way drug discovery will be. That's my belief. Because if you want to explore all of this space, a lot of space to explore, it just... You need to be so fanatical around the design choices that you're making. Can you design? Can you make? Can you test? Can you learn? Can you continue to do that over and over and over again? Mm-hmm. Because that is active learning, biology, chemistry, complex, but we must use the tools of our day to continue to make sense of this all, explore it, advance it. Got it. Got it. That's helpful. We're digging into the data a little bit more. For the CCM readout, there were also PROs that were evaluated. You mentioned in the release that at the twelve-point time mark, excuse me, you haven't seen a benefit yet. What should people make of that signal? What's the best way to interpret PROs versus the lesion benefits you mentioned? Mm-hmm. How do the two endpoints map against each other, and how important could they be for a future pivotal program? Absolutely. Great question. I think as we talked before, Recursion is the first. This is the first industry-sponsored phase two. So we are truly in pioneering space, you know, bringing the first phase two. Hence, the reason why with the FDA, we looked at about a dozen or so efficacy endpoints. Some of these more on the objective side, like MRI imaging, where you look at, you know, lesion volume, hemosiderin ring, so on and so forth. Some of these being a bit more on the subjective side, like patient-reported outcomes, PROs, physician-reported outcomes, so on and so forth. I think what's important here is if we look at the effect we're having on lesion volume reduction and hemosiderin ring, six months better at twelve months. That timescale, if you look at the literature around, say, hematoma and half-life, starts to kinda tie out with what that half-life might be. So you're seeing, at least from my perspective, those timescales starting to kind of match up. That makes sense. And again, the fact you're having a bigger effect at 12 suggests that the longer you're on therapy, the more effect you can look to have. Overall, I think that this is a slowly progressing disease. We see it manifest itself slowly over time. And the fact that, you know, PROs in general are a inherently noisy measure, I think that as we have patients... Again, we have the vast majority of patients continuing on the long-term extension study. I think we'll continue to accrue data on that longitudinal, on those longitudinal effects. I think that, you know, and I'm very much looking forward to see some additional data around the functional measures. But I do think that a lot of experts see the lesions, again, primary driver of this disease, and it would seem as if our drug, REC-994, already have an effect on those lesions, which many folks believe to be, again, primary driver of the disease. Got it. Got it. Now, Recursion's platform has different data, data layers, many different inputs, many different insights you can derive from the platform. And you've also iterated on the platform over time. And you've mentioned that CCM, 994 kinda came from the first version of Recursion's platform, so to speak. So from that perspective, what do you think the data readout helps de-risk for the platform, and what do you think are the kind of the pending open questions, especially when people think about the future data readouts coming from the platform over the next year or so? Sure, sure. Absolutely right, I mean, it did come from the earliest version of the Recursion operating system. I think what's important is that when Chris and Dean were looking at CCM as a potential indication, you know, they were using a lot of gold standard approaches that were, I think, that defined a lot of the industry at the time or thinking on the space. And they were coming into contact with a lot of failure around what was preconceived notions of what should be success. That ultimately gave rise to taking a step back and thinking: How could we think about this disease in a target-agnostic approach? You know, Anne Carpenter at the Broad was doing good work or early work on phenomics and how one could actually, you know, apply high-resolution microscope images with computer vision to extract out morphological features, i.e., giving an approach that could truly be functional, not target and be target agnostic. So with that, when they took a step back, they started to apply, you know, an earlier version of a phenomic approach, taking a disease model, looking at how to perturb it with perhaps many thousands of compounds, for which they then saw a functional rescue. I'd say, Vikram, over that timeframe, over that decade plus timeframe, if you look at a lot of large pharmas, if you look at a lot of biotechnology companies now, a lot of them have brought in a phenomics approach to complement how they're doing the drug discovery development. I think Recursion, one of the first to really adopt how to apply this target-agnostic approach, how to apply a phenomics approach coupled with computer vision to extract out these morphological features and see how a cell looks, how that could be characterized healthy, disease, and could you perturb it back to health? I think that validates that early thesis that a target-agnostic approach could bring in something that was not expected. I think that that gave rise to REC-994. I think that gives rise to what we see happening in the clinic, that you are having an effect on the, you know, what is characteristic of these lesions, that are driven by ultimately, loss-of-function mutation of, CCM one, two, or three genes. Got it. Got it. Great. Now shifting over to Exscientia's platform, and your pipeline. So 617, there's a data readout expected from the ELUCIDATE study by year-end. Dave, just characterize for us what we could learn, what you hope that ends up showing, what would be a win for you, and what should people think about when they're looking to interpret that data readout? Absolutely. I think a brief reminder, I think, kind of just CDK7 is a really important kind of cellular enzyme in both healthy and a disease state. So the story behind CDK7 has always been, we have a lot of data, there's a lot of data from the literature as well that says that this is a really good target, you know, and a target potential in a lot of solid tumors. But you're walking a fine line between safety and efficacy, so therapeutic index. So one of the things that, as part we're looking at and what does good look like at the end of this, this monotherapy dose escalation, is that have we been able to deliver a molecule that can substantially inhibit the target? Ideally kind of showing that at the level of the tumor, but also maybe using blood as a surrogate. So that we know that when we then go into any subsequent efficacy study, we know we're testing the mechanism. 'Cause if you look at the kind of the competitive landscape, and there's a handful of companies out there, it's still very difficult for us to kind of understand, but are any of those competitors really inhibiting the target hard enough for long enough? And so, we've presented this before. We have our preclinical data, both in kind of from rodent models but also using primary tissue, suggest that the kind of line you want to walk is that you probably want to inhibit this target at sort of 70%-80% levels for about 8-10 hours, but then you don't want to be on the target anymore. And so the way that will then reflect in the kind of it reflected in how we design the molecule. So it's deliberately a reversible inhibitor of that enzyme because there's quite a few covalent out there. It has been deliberately designed to have a modest half-life so that it's not hanging around for too long. And it's also deliberately designed, so it can be very well behaved in terms of how uniformly between people it's actually absorbed. So I think a good readout for towards the end of this year would be: How well did we do in terms of the design characteristics? What's the pharmacokinetics like? How well is the compound tolerated as we kind of dose escalate? What are the kind of indicating pharmacodynamic biomarkers? So we're looking at both tumor-based biomarkers from paired biopsies, as well as kind of sort of broad transcriptional readouts, both in blood and tumor. So I think that will be the indicator. Have we found a compound that we can dose where there's a good separation between the undoubted mechanism-based side effects you will see with a CDK7, i.e., particularly kind of GI side effects that others have seen? And do we have confidence as we take it forward into that ELUCIDATE breast cancer study- Mm-hmm. But also, as we start to think about, can you dose a CDK7 inhibitor with a checkpoint inhibitor, for example? Have we got a molecule that we know going into kind of combination efficacy studies, that we're gonna, we know fundamentally we're testing the mechanism? And so that data we're looking to kind of present towards the end of this year as we come towards the end of the ongoing monotherapy dose escalation. Understood. That's helpful. We have time for one final question. We started the discussion with kind of a philosophical conversation about the business combination, so maybe we should just go back to that and end there as well. Just looking out five years, ten years, at the combined company, what should people look forward to, as kind of the value add of the two platforms coming together? Should it be more focused on the identification of new targets? Should it be the ability to drive better safety, better efficacy with known MOAs? Should it be novel chemistry? What do you hope? If you had to pick one of those kind of three pillars of AI-driven drug development, where would you hope to see the most benefit from the two platforms coming together? I think the long-term vision here is about from the combination is that to be both first and best. There's little point being first in class only to put kind of inferior kind of chemical matter together with that, because all you do is end up telling the rest of the world what a great piece of biology it is. And so I think I look forward to kind of building a pipeline kind of with the Recursion team post-close that delivers that. Delivers novel targets, but with a kind of definitive kind of small molecule chemical matter that goes with that. I love the question. I think that we talked before about just the problem of drug discovery: time, cost, translatability. And I think that when you are starting to align these multiple cycles of learning iteration across biological, chemical exploration, around chemical design and synthesis, it's all of these things coming together that I think can have an effect on all of what novel insights can you find? How do you drug it? And what does that mean for, you know, first-in-class opportunities, best-in-class opportunities? And what does that mean for working with, you know, new and potentially, you know, existing and potentially new partners? And I think, very much looking forward to, you know, working with Dave and the team and going forward. Great. Great place to end. Thank you, Dave. Thank you. Thank you, Mike. Appreciate your time. Thank you. Thanks for being here. Thank you. Thanks, everyone. Thank you. We'll close out. All right, thank you.