Schrödinger, Inc. (SDGR) Earnings Call Transcript & Summary

Hello, and welcome once again to the 19th Annual Morgan Stanley Healthcare Conference. My name is David Lebowitz. And before I get going, I'm going to read the requisite disclosures. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. And with that, I'm happy to have with me today of Schrodinger, I have from the company CEO and President, Ramy Farid; Chief Biomedical Scientist and Head of Discovery, Karen Akinsanya; and CFO, Joel Lebowitz, and that has no relation to me, I might add. And with that, I guess if you could start out by giving a top line discussion or explanation for the company itself? It's certainly a unique corporation, and I'd be happy to hear your thoughts on how it came about.

Absolutely. Yes, the company has been around for quite a while, over 30 years. And over that period of time, we have developed a computational platform based on physics that is -- its aim is to compute molecular properties as accurately as possible. We started selling that platform to pharma companies, biotech companies, more recently materials companies, academic institutions, government labs worldwide. And that business has been generating a significant amount of revenue, and it's been growing very nicely. About 11 years ago or so, we started to make some really significant breakthroughs in the technology. And in situations like that, when you get to a point where the technology is transformative and at the same time, disruptive, it becomes challenging to convince large companies to adopt the technology on a really enterprise-wide level. And so we recognized back then that in order to really provide the kind of validation that was required to really transform the industry, we needed to do drug discovery ourselves. We started by forming collaborations with companies like Nimbus and Morphic that we help to cofound, and we've been working on those, probably we still are working on those using the technology, learning from those collaborations and continuing to improve the technology. We eventually got to a point where we were able to run our own internal programs because we've built up that expertise. And we're realizing really extraordinary synergies between these seemingly disparate businesses, where we're interacting with our software customers with every pharma company that's using our software learning from that, those learnings go into the into the software, and we apply it to our internal programs and our collaborative programs. And we even see synergies between life science and material science part of the business. So hopefully, that gives a good overview. I don't want to take up the whole half an hour with...

I'm sure you can go on and on.

Yes, yes, yes. I hope that was helpful.

So the company has a physics-based software, runs a bunch of algorithms and there's a lot of modules to it. What are the most popular modules right now? And what other types of modules do you look to introduce in the future?

Yes. We do -- it's true. We have quite a few pieces of technology that combine into a complete platform for doing everything from target validation to hit identification to lead optimization on the life science side and on material science side, design of all types of molecules from polymers to organic light emitting diodes to many other types of materials. So I think saying that -- it used to be the relevant question, sort of which tool is more popular when we were sort of viewed as a tools provider. But now it's really a complete platform and complete workflows that, again, like I said, are aimed at solutions like hit identification or lead optimization. So I would say, really, it's the entire platform, the ability to identify a molecule as a starting point for lead optimization and then to go through the entire lead optimization workflow, which is really a multiparameter optimization problem. Now where the technology is going is very exciting. Where we're seeing some really exciting advances is these physics-based methods require as input a protein structure. And as is well known, the number of structures available, experimental structures from X-ray crystallography or cryo-EM of the human genome is a pretty small fraction. So the need to produce high-quality structures computationally or develop computational methods for refining low-resolution experimental structures is very high. And if we do that, it would expand the domain of applicability and the scope, the number of targets we can work on in a really significant way. So that's -- in a lot of ways, that's where the technology is going. Now that's quite complicated. I said a lot of things, there's -- again, it's not one tool. It's a complete workflow for producing high-quality structures. The other area we're focusing on is there are quite a number of properties that we can compute with very high accuracy. But there are a number that we can't. And so we are working on expanding the number of properties that we can compute with high accuracy. And the more we can compute, the more you can do, the more of the multiparameter optimization you can do on a computer saving that and not having to synthesize those molecules and test them experimentally, which is obviously very costly and slow.

Since you went public, certainly the amount of people paying attention to the space has grown substantially. And there are a lot of new companies and other competitors out there doing artificial intelligence, some, I guess, starting also physics-based approaches, how do you see yourself fitting in with this wealth of other companies that are coming on?

Yes. We should make clear first that although the platform is -- our platform is based on physics, it leverages, in a very important way, machine learning. That's a very, very important piece of the technology because these physics-based methods are computationally expensive. And although we can process and analyze quite a large number of molecules, in order to be able to analyze hundreds of billions of molecules, it's necessary to integrate the accuracy of the physics-based methods with the throughput of machine learning methods in a very clever way to which we've talked about. So I just want to make that clear. We see some companies that are focused solely on machine learning. We think that's very limiting that the types of training sets that machine learning models need are too large to be able to produce those using just experimental data. So you need a physics-based method. The few very, very small companies that are claiming to be developing the physics-based methods are very far behind. We're aware of that technology. We continue to invest very aggressively. So we're ahead now with regard to domain of applicability and accuracy. And we think by continuing to invest in the platform, we can maintain that lead and continue. I think this is a very important point to leverage these 2 methods, machine learning and physics-based in a unique way.

How difficult is it for these companies to be able to catch up?

Yes. So it's a very good question. I think if we stood still and didn't do anything, I think in 5 years, that sort of timescale, they may catch up to where we are now. I think it's still quite challenging. This has been, as I said, the company has been working on this for quite a long time. There are some very complex physics and math and so on that's required to do this. But I think if we did stand still, it would take that sort of time. I think if we continue to invest at the level that we have, I think we can maintain the competitive advantage and the lead that we have now.

So let's start talking about the details on the software business. Right now, 80% of revenues come from the software business. Software business itself breaks down into -- I know investors tend to focus on biopharma because, frankly, it's easier to focus on biopharma. It's not all that's there. Tell us about the actual breakdown of these revenues at this point in time and how you think on the biopharma side and on the material science side, things might have evolved going forward?

Yes. At the moment, we still have not disclosed the breakdown of life science versus materials. That's something that we're going to do in the future because the materials science business is becoming a larger portion of our software sales. We're investing in that division very heavily. We see really significant opportunities. We've been realizing those opportunities. So as I said, it's becoming a larger portion of our business, and it will be appropriate in the future to disclose the breakdown. Joel, did you want to add anything to that? Is that...

No, I think that covers it.

Good.

So there's been a lot of focus lately after you gave guidance earlier this year on pharmaceutical buying patterns and how they might differ from traditional buying patterns in the software space. Could you elaborate on what that actually means and how we should look at the company's revenues going forward?

Yes. The very large majority of our contracts with pharma companies and biotech companies, which, by the way, go back, in some cases, decades, are annual renewable subscription-based licenses. Most of them are annual. We have some multiyear deals. And what we have seen, again, because a lot of these deals go back quite a while, that the initiation of the contracts were coincident with their own sort of budgets, annual budgets, which is usually in Q4 and Q1. So we do see the sort of seasonality that maybe doesn't exist in other companies. And that has to do again with just the way the contracts work and the fact that they're annual. So I think that's what you're sort of getting at. Does that answer the question?

It does, to start. First of all, I would ask is, number one, do they always re-up once a year, do some add during the year? Does that occur? And given that there's an element of trial and error and they're learning how to use the software and have to expand over these large clinical platforms, how -- what -- do they buy heavy one year and buy lighter the next year? What drives that?

Yes, yes. Yes, it's a good question. First of all, the renewal rate, as we reported recently, I think this was in 2020, was 99%. Joel, is that right? Yes. So it's very, very high. It's extremely unusual for a customer once they've acquired -- licensed the software to not renew. And in the cases where it doesn't happen, that's usually the company either getting acquired by another company or essentially ending their preclinical work and shifting their focus to clinic. So they just don't have research efforts anymore. But again, that's obviously pretty rare, given the renewal rate of 99%. So now with regard to renewals in the middle of the -- or not renewals, but upping up in the middle of the year, that's unusual. Usually, these conversations about the needs for the next year occur a quarter before the renewal date, something like that. We obviously have very good visibility into that. We know when all those dates are. We begin those conversations in a serious way. Obviously, we're interacting -- and this is very important. We're interacting with them through the whole year, understanding what they're doing, we're understanding what their needs are. We're giving them webinars and training and seminars, all sorts of things to tell them about the new technology that's coming, the improvements, the kinds of work that we're doing, right? So they see the kind of impact that larger scale use of the software is having on our programs and our collaborative programs. And so we see, obviously, companies for the most part, increasing their usage. They're still way below where they should be, given the level that we're using it internally. It is -- it can be lumpy. We do see some companies doing some very large investment, and then it takes time to see the impact of that software to learn how to use it. And so you may see in the following year, not exactly the same rate of increase. So that's the origin or the source of some of the lumpiness in the business that some had observed.

Is it typical that the clients that re-upped last year and increased are the ones most likely to continue with that type of trajectory going forward? Essentially, they show an embrace of the technology 1 year. Does that have a positive feedback?

Absolutely. Absolutely. There's no question about that. That usually is the result of not only the company recognizing the impact of the software, but also the nature of the relationships that we have. So at those companies where the amount -- the spend starts to get into larger and larger amounts, that requires essentially sign off from, for example, the Head of Research or the Head of Chemistry, and obviously, having those kinds of relationships and the attention of executives and companies can obviously lead to larger growth in the future.

That makes sense.

And that growth can come in different forms, not just in the software but of course, in other types of partnerships like you saw with BMS, for example.

We'll certainly be getting into that again in just a few moments. So you guided towards fourth quarter to have this increase. What type of visibility -- how long ahead of that fourth quarter decision-making process do you actually have true insight? Because at the end of the day, meeting or beating guidance are falling short, you might not actually know until the order is done. How long ahead do you have an insight to what the order is going to look like vis-a-vis last year?

Yes. I mean I think that's a complicated question. It depends very much on the account. We have some visibility, obviously, into the nature of the discussions, which is why we provide -- we can provide guidance. And -- but the reason it's a range is because, as you said, what happens on the -- and this is a common [indiscernible], right? I mean, you don't know until it's closed. So there's visibility. There's a sense about what's going on, but you don't know until it's closed. So there's still some uncertainty.

Got it. Yes. So let's go from the software side to the drug discovery side. We'll start out with the internal programs. You have 5 that you've been highlighting, but there are 3 that are going into IND-enabling studies. Could you talk top level about those various programs?

Of course. Karen?

Yes, certainly. So as you've highlighted, we have 3 programs that have transitioned over the year, not just into late-stage discovery, but now also into IND-enabling studies. These 3 programs are in oncology. So we have MALT1 program for B-cell lymphoma, CDC7 inhibitor for solid tumors and AML and Wee1 inhibitor for also [ solid ] tumors. Each of these programs has benefited from the technology in terms of the target product profile. We've been designing what we believe to be best-in-class compounds in each case, slightly different challenges. But we're really pleased with how these profiles are stacking up against the target product profile. So as you pointed out, this is a transition year for these programs. The first is in IND-enabling studies, and that sets us up to be able to take our first program into the clinic in the first half of next year. And given how we started these programs, there will be a gap between each of them, but moving each of these forward into the clinic. And I'd say that we've really been pleased by the packages we've been developing. We've published at ASH last December. The MALT1 profile in vivo showing really nice data. We believe it's one of the most potent, if not the most potent, MALT1 inhibitor described to date. And then on Wee1, we haven't published yet. We haven't shared the data, but that's coming. So I think this year, we published at AACR on CDC7. Similarly, really these molecules have translated in silico to in vitro and now in vivo in ways that we're very pleased with and I think support us moving them forward.

So what made you choose these particular targets? Now there's a lot of targets out there. And clearly, one of your competitive advantages is the software itself, which helps to improve chemistry, potentially resolving challenges with these targets. What about each of these targets opened the door for you to proceed?

Yes. I mean I think it's very close to our evolving strategy. So when we started thinking about launching an internal portfolio, we really wanted to pick targets that met certain criteria. Number one, was there any human evidence that these mechanisms were going to be tractable from a therapeutic standpoint? Now obviously, you don't know that until you're in the clinic, but we look for certain hallmarks. In the case of MALT1, this is in the BTK pathway. We know that people are resistant to BTK inhibitors or that they relapse. And MALT1 is downstream of BTK and is involved in NF-kB signaling. So from a biology and a sort of clinical potential point of view, it certainly met that bar. It's also an allosteric site. We've had a history of doing well with designing small molecules for allosteric sites. In other cases, in the case of Wee1, there was evidence of human efficacy. But what we know about that molecule, and this is characteristic of several programs we've worked on, is that while there was initial evidence of efficacy or antitumor activity, the molecule itself, the most advanced, I would say, in the hands of others, has some challenges with regard to its properties. And that's where, as you point out, the technology allows us to try and solve for those issues, for example, selectivity or something like time-dependent inhibition of CYP3A4. These are challenges that we felt even with that data in hand showing antitumor activity are going to be potentially challenging for those molecules to realize the full potential of the mechanism. And so those are sort of reasons we decided to deploy the technology against these particular targets. Now as time goes on, the extent to which we lean on breakthrough biology and so on, in addition to our platform, we think that this is a really exciting space for us going forward.

Thank you. Now each of these candidates that are entering the trial soon, when is -- could we initially see data? And I guess relative to that data, I imagine you're considering partnerships. When could we see a move in that regards?

Yes. So I mean, obviously, it really depends on the clinical trial design. We'll be taking these molecules into safety, tolerability, dose escalation studies to begin with. Being oncology programs, we will be obviously recruiting patients and looking for signs of proof of biology, target engagement and ultimately, early signs of efficacy. So I would say that a year after the first in human, we'd be looking to report on the progress that we're making against target engagement markets, and then potentially there on looking at efficacy endpoints. So I would say it's 2023 before we're probably going to start seeing any patient responses or anything like that, 2023 and forward. That was the first half of your question. The second half?

Potential timing of partnerships?

Right. So one of the things, as Ramy pointed out, is we have good relationships with the pharma companies. And so we've been pretty much in constant conversation with pharma companies about these mechanisms and the profiles that one would like to see. As you may know, each of these is a combination partner for either a marketed product or one that's in development, and BTK inhibitor is a great example. We think MALT1 inhibitors are going to combine with BTK inhibitors. We've had a number of productive conversations about what does good look like there, saying with these DNA damage repair mechanisms. In terms of timing, we think that we have a lot of flexibility. We can generate data that takes us to -- into Phase I, allows us to see how the mechanism is doing, decide to partner or indeed continue with those programs. So I think it's hard to predict right now when we'll see a partnership. But as you know, with BMS, when we started talking about the science behind some of our programs, we decided to partner those relatively early on in the process. And so I think the full spectrum from very new ideas that are amenable to our technology, all the way through to products that are in the clinic, there's a lot of partnering opportunity across that spectrum.

And with that, let's move over to the partnerships. You have a handful of partnerships, and they could produce substantial revenues over the years, although we don't necessarily have a great deal of visibility on those revenues. Could you talk us through, number one, about those partnerships? What's in the clinic? What's going to enter the clinic soon? And what are the potential catalyst structures for these partnerships?

Yes. Sure. I can do a little bit of that, and then I can hand it over to Karen and Joel for more detail. So there are -- at the moment, 6 of the programs that we have worked on are in the clinic. And of course, these programs all have unique structures. But in general, there is research funding associated with them. There is preclinical milestones. There are clinical milestones, which are generally larger than the previous 2 things I mentioned. In a number of cases, actually, there are royalties on sales. And of course, in many of the cases, and there are more than a handful by the way, but anyway, there are -- there's equity stakes, right, in a number of them. And we can -- and we've already realized the value from our equity stakes in a number of the companies that we've been involved in. So we're very pleased with the progress these companies are making. I don't think that's an accident. I think these are -- there's a lot of examples where these are mechanisms that pharma went after and was not successful and where the partnership and the application of the technology led to a high-quality molecule that went in the clinic. I think that's -- it's building up to be a pretty good track record. And the other thing I'll point out is, of course, as our technology becomes more validated and as our ability to produce development candidates that go into the clinic becomes more obvious, we've been able to negotiate better and better terms. So the programs that are still in discovery, some of them entering and have already entered IND-enabling studies are generally associated with higher milestones than ones we've seen in the past, which we -- obviously many have reported on. You can see that -- actually, that's something we have broken down, right, the revenue from the drug discovery in the software. So hopefully, that answered all the questions that you're asking. Yes.

Certainly. When looking forward, how should we view revenue growth and the franchise, considering we only know the drugs that, frankly, the partners talk about?

Yes. That's a challenge, as has been pointed out to us a number of times, and you just said it. Obviously, these are programs that are run by these other companies, and it's really up to them when they're going to disclose progress. And so I think I would go back to what I said before, right, the economics are getting better. So to the extent that we have more programs that have started more recently, that's an important statement. And I think the fact that these programs are advancing to later and later stages, and those are associated with higher milestones, and that the equity, that's when the value from the equity can start kicking in. Obviously, some way down the line, we would certainly expect, in some cases for the -- and hopefully, all the cases for the royalties on sales to be kicking in. So we see really very significant opportunity on that side of the business.

I don't -- I'd just add that if you look at the revenue profile that we've guided to you on drug discovery this year, it's growth versus prior year. But more importantly, if you look at it over multiple -- you think about it over a multiyear period, the opportunity for growth is really significant. And I just point to the BMS deal, in particular, one of our most recent deals, as kind of game-changing economics with $2.7 billion in potential milestones and $1.6 billion of which are in the start -- are in the research, development and regulatory phases, so not even commercial. So the opportunity and the probability of starting to achieve some really sizable milestones, at least with that deal, in particular, we think, is very attractive. And I think you'll start to see that over -- in the relatively near term over the next several years, you're going to see that trajectory continue.

And we're starting to get towards the end here. Should we expect to see more of the Bristol-type partnerships going forward?

Karen, do you want to address that?

Yes. I mean, I think that there's a lot of inbound interest, actually from both large companies and small companies who, if you think about Nimbus and Morphic, they were able to leverage this platform for their entire portfolio. And I think -- so both the large companies and small, there's a lot of discussion about targets, about particular design challenges. And so I think that, yes, there are very likely to be more collaborations for Schrodinger to proceed in the future.

Excellent. And the last question here is just on cash position and runway. You have done 2 financings, including the IPO. How do things look going forward?

Joel?

I can answer that. So we had $617 million in cash on the balance sheet at the end of the second quarter, and that's a reduction of $27 million in the first half of the year. So relative to the size of the cash balances, we feel really good about the resources that we have to execute on our strategy. Obviously, we're not commenting on any potential capital markets activity, but we feel really good about the resources we do have and our ability to execute on the strategy.

Excellent. And with that, thank you very much for sitting down with us via video today. Maybe next year, we'll be in person, and look forward to chatting again soon.

Thank you very much.

Thank you.

Great to see you.