Hello, welcome, and thank you. Well, good afternoon. Welcome to our next session.
Oh, and we can share-
Here at the Jefferies Healthcare Conference. We have a number of members of the Schrödinger management team up here with us. We have the CEO, Ramy Farid, on the far left. We have the CFO, Geoff Porges, here on the left-hand side, as on the end of the table, and then in the middle, we have Karen Akinsanya, who's the president of R&D, Therapeutics, and of course, very important because a lot of internal pipeline drugs are moving forward. I know that Ramy would love to have the opportunity to give some slides.
Okay. Yep.
That would be great, and then we'll have a lot of questions and a lot of topics I'd love to hit on in our 24 minutes together.
Yeah, and we'll go as quickly as we can-
Sure
... so there's enough time-
Sounds good
for questions. So just launch right in?
Great.
A good place to start is what our vision is for the future of drug discovery, and in general, molecular design. It's pretty simple. The first step is to enumerate as much of chemical space as possible. It turns out, chemical space is almost infinite. So that's actually kind of a big deal, essentially designing every possible molecule that you can make. In the case of drug discovery, it's drug-like molecules. In the case of materials, it's molecules that have general structure that a particular material might have. And then to develop, that's the thing in the middle, to develop algorithms for accurately predicting the properties of those molecules. And the reason this is incredibly important is, in every molecular discovery project, you have to design into a single molecule, a whole bunch of properties that are fighting each other. They're anti-correlated.
So to design a drug is extraordinarily hard because as you design a molecule that's potent, it tends not to be soluble, and as you make a molecule soluble, it tends not to be permeable, and as you fix that, then you screw up potency, and you get the idea. So this is an incredibly hard multi-parameter optimization problem. But if we can enumerate much of chemical space, if we can accurately predict on a computer before you go doing the really expensive thing of actually making a molecule and testing it on a computer, then you can do this multi-parameter optimization problem very rapidly, very efficiently, and identify molecules that actually have the properties that you're targeting. That's the vision. And what we've been working on for quite a number of decades is that well, actually, two things.
One is the algorithm in the middle to actually predict properties of molecules very accurately, and also enumerating chemical space. That turns out to be a challenge as well. And what we have found, and we talked a lot about this, is that there are two, at a very high level, completely orthogonal, completely different approaches. One is to do something that you all know about. It's called AI. We used to call it machine learning. We used to call it deep learning. We used to call it QSAR. It's all the same thing, machine learning, knowledge-based methods. That turns out to be pretty powerful, but by itself, it turns out not to be actually that useful. We've also been developing first principles methods. This does not rely on training. It doesn't rely on knowing anything.
It relies on understanding the underlying physics that governs the properties of these molecules. It turns out by combining these two methods, the really accurate way of predicting properties of molecules using physics and using that to build training sets for machine learning, is a really powerful method for developing that algorithm that we talked about in the first slide. We built this platform. It's taken many decades. It's super complicated, but we have shown now, and I'm gonna show you in a second, many ways in which this platform is highly validated and having a really big impact on the industry. What are we doing with this platform? We license the software to pharma companies, biotech companies, material science companies, academic institution, government labs all over the world in both life sciences and material science.
Because it's physics-based, it's completely agnostic, not only to the modality of a drug, it works on small molecules, peptides, antibodies, but also on a very wide range of material science applications. So that's the software business, what we call the software business, and we'll talk about that probably, right, Mike, a little bit?
Yes.
It's an important part of the business. We also, for a while, have been using that software to advance programs, discovery programs in collaboration with other companies, many of which we've co-founded. I'll show you a slide on that, and that's both in drug design and in materials. And that business, you can see there, we have quite a number of collaborations. That's been really quite successful. Quite a number... I'll show you in a second. We'll look at the names, you'll recognize them, but this has been very successful. Let me leave it there, and we'll look at the details. And then more recently, and we're getting more and more excited about this, and Mike, you just mentioned it, we have a proprietary pipeline of drug discovery programs.
You see there a number of active, wholly owned programs, where now, and you'll see in a second, Karen will talk about these; these programs have actually made it all the way from target selection through optimization into IND-enabling studies and even now in the clinic. And we're very excited about those programs. The important thing, and there's a little subtle feature in this slide, you see these cards are overlapping a little bit. That's by design. These are highly synergistic businesses. We're learning so much from our customers, and that gets fed into the platform. We're validating the platform from these collaborations and proprietary programs. That's helping to drive demand from pharma companies, so there are lots of really interesting synergies between these different businesses. So I've been alluding to this. Here are the collaboration programs.
We have a number, you know, some in Phase 3, even some approved, a number of programs in Phase 2, a number in Phase 1. We think... I don't know if this is statistically meaningful or not. You can decide that, but this is a pretty good track record.
...And we think this is a reflection of the impact that the platform is having on these discovery programs. A number of the companies on this slide, we've co-founded. Obviously, some of them, you know which ones we didn't co-found. They're pharma companies there we've been collaborating with for a while. And maybe Karen can speak. Can you see this slide, actually? Karen, this is. Yeah, maybe you can speak to it.
Yeah. So as Ramy alluded to, we've been building a proprietary pipeline of drugs. We have two now in the clinic, our MALT1 inhibitor. This is a target in the BTK pathway, that's now validated in the clinic, in third-party hands. We think we have a best-in-class molecule, and that is now in a Phase I dose escalation. Our next program is for AML. That's our CDC7 inhibitor, also in the clinic and progressing well. And our third IND has just been cleared. That's a Myt1/Wee1 inhibitor that is beginning trials very, very soon. Behind that, we have a number of other oncology programs, a SOS1 inhibitor, PRMT5-MTA, as well as an EGFR brain-specific inhibitor, and also now emerging our immunology pipeline and some undisclosed targets.
The first batch are sort of best-in-class plays, and our first-in-class targets are coming along in the pipeline. Next slide. As I just said, for these three clinical stage assets, we've done a healthy volunteer Phase I study, which is complete. We showed data on that last year. We also have the global dose escalation study in relapsed refractory B-cell malignancies ongoing, and we expect to have data through the end of 2024 into the end of 2025 as enrollment continues. Then CDC7, same sort of timeline with respect to data. As I just mentioned, our Wee1/Myt1, we are initiating dosing very soon. Geoff?
Okay. Take us through the numbers, Karen?
Yeah.
This is just the Q1 highlights. This is more or less old news now. We're most of the way through Q2, but for the first quarter of this year, software revenue came in at $33.5 million, compared to $32 million last year. We continue to think that we've got strong prospects for growth for the rest of the year. Drug discovery revenue was $3.2 million. There were two large milestones that contributed in the first quarter of last year that were non-recurring this year. Software gross margin remains very high, in the high 70s%. We think it'll be sustained in the sort of 80% range going forward.
Operating expenses increased principally due to an increase in R&D spend, and there's a lot of noise in the other income because last year we reported the benefit from the Nimbus distribution of $147 million in cash that came to us. Our cash position at the end of the quarter was $436 million. Just quickly on the full-year guidance, we've guided to software revenue growth of 6%-13%. That's down from last year, we were at 17.5%, but that's because of the large contribution of two multi-year deals in the fourth quarter of last year. Those are, since they're multi-year, they don't recur this year, so we're growing through that kind of headwind in the fourth quarter.
Drug discovery revenue, we guide to $30 million-$35 million. Same thing, we had $25 million in the first quarter of last year. That's non-recurring. We expect the software gross margin to stay high, and we're bringing down the operating expense growth. We expect the cash use in operating activities to be somewhat above the cash use last year. In terms of the priorities that we've been communicating to investors, we are focused on maintaining the commercial momentum in the software business, particularly focused on drug discovery. We're aiming to expand the scope of application of the computational technology platform, and we can talk about that in the Q&A. Advancing the lead programs to proof of concept and progressing the second wave of programs that Karen talked about.
And in terms of milestones and catalysts, Karen alluded to this already, we're guiding that we'll be initiating the first patient being treated with 3515, that's the Wee1/Myt1 inhibitor, in Q3, and then we're looking to clinical data from both the CDC7 and the MALT1 program late this year or next year.
Very good.
Okay.
Very good. That was an excellent summary. Let me maybe break that up into some parts. So, first on the software business, which I think we agree is a significant majority of the valuation and your significant revenue contribution, that there's been various debates about the rate of growth and also the guidance. And so, and the good news is that you've actually come in at the high end of software guidance every year since you've gone public. And then in 2024, you gave guidance in January or February, and the rate of growth was 6%-13%, which is down from the above 15% growth that you've done historically. And Jeff says that that's because of a large contribution in the fourth quarter, which creates a difficult comp year-over-year because the number was so big in the fourth quarter.
So my question is, are there data points that would suggest, hey, yes, don't look at the year-over-year, the rates are accelerating, the adoption is accelerating, and if I backed out that large contract, it is accelerating?
Yeah.
Can you talk to that?
Yeah, that's right.
Because people saw the number, it's lower.
Yeah, yeah.
I hear you on the large-
Yeah
... year-over-year comp-
Yeah
But that doesn't make me feel great.
Just to be clear, that Q4 deal that Mike is referring to-
With Lilly.
Exactly-
Lilly
... is what's called an on-prem deal, which requires us to recognize a majority of a pretty significant portion of the revenue in the quarter that it closed. That's why it's a difficult comp, right?
Significant, so like if I took the number that you reported-
Yes
... on the fourth quarter, which was, how much was that? 30%.
$70 million.
Seven? Okay.
Yeah.
A big chunk of that was a Lilly deal?
... Right, and a big chunk of the Lilly deal had to be recognized in the quarter that it closed. It isn't, it wasn't a so-called hosted deal, where if it was, the revenue would've been recognized ratably over the whole thing, so it's got this lumpiness. So that's-
Was Lilly an existing customer?
Oh, yeah, yeah. Yeah.
Okay, so they went from a current customer to a bigger customer-
That's right
... using a multi-year deal?
Yeah, and again, because the deal is not hosted, we're required to recognize most of the revenue in the quarter that it closed.
Okay.
So now, to get at the hard... Just to—
So if you back that out, my question is-
Yeah, yeah
... not going into all those accounting things-
Right
... tell me-
Right
... how I should look at the growth rate-
Right. Yeah
... on a more normalized pattern?
That's right.
Is it accelerating, or what?
Yeah, yeah. We're definitely seeing signs of it accelerating in the following way: the demand for computation and our, I mean, our platform, I should say, but obviously, that's computation, is really increasing. There are more and more pharma companies coming to us and saying, "We're done messing around. We've been messing around for decades. It's time to start doing drug discovery the right way." This is clearly the right way. We've developed the best platform. We've validated it over and over and over again. In many of these companies, by the way, we have drug discovery collaborations with them. They're seeing the impact of an actual live project. And I think they're starting to get to that point. Now, why aren't they just adopting it on, you know, immediately?
Well, there's a lot that has to change at a pharma company for them to be able to make use of all of this technology. They know they need it, but who's gonna run it? And what kind of processes are going to be changed inside the pharma company to do drug discovery, and really, let's be honest, it's a completely different way. We're not gonna put so many chemists on the project. We're not gonna make as many compounds. We're going to actually synthesize compounds that a human looking at it, their intuition doesn't tell them it should work. That's weird.
Mm.
That's a weird thing. That's how we've been doing drug discovery this whole time. It's based on intuition from so-called medicinal chemists that have some kind of magical ability to look at a molecule and see physics. And that's supposed to sound silly, by the way, what I was trying to say.
Well, no-
That's ridiculous
... doesn't it kind of imply-
Right
... that, you know, when you get into it, there's some legacy-
Yeah
... dynamics-
Exactly
... at the big companies-
You got it.
... run by medicinal chemists-
That's right
... who are traditional organic chemist guys-
Exactly
... on whiteboards and computers-
You got it. Yep
... doing stuff.
Yeah.
The software says, "Do this"-
Yeah
... and they go, "Why?
Then they go. Right, and, and that's the nature of physics-based methods.
Uh
... is that the predictions aren't intuitive. That's sort of the whole point. If they were, you wouldn't need any of this. It would've worked. We wouldn't have the problems we're having now of these huge, higher, high failure rates and, you know, programs taking forever to get to development candidates, with most of them not succeeding. So it's a process. And, but-
So, is there a metric?
Yeah.
You know, is it-
Yeah
... ACV numbers?
Mm-hmm, mm-hmm.
Is it there's more people doing over 1 million, or whatever it is?
Yeah.
What are the metrics I can look at and say-
Yeah, right, right.
... "It is getting better"?
Yeah.
'Cause people are seeing the revenue number not accelerate.
Yeah, exact-
There are a few metrics.
But again, that's an artificial, right, because-
I hear you.
Yeah, yeah.
I hear you.
So go ahead, yeah.
The first is the KPIs that we disclosed.
Okay.
Mike, we talk about the number of customers with ACV over $5 million a year and the number of customers over $1 million a year. The number over $5 million increased last year. It was 2 in 2022, it was 4 last year, and over $1 million, it went from 18 to 27. We think those are very strong indicators of the underlying dynamics in the market. That's probably the most important. Overall ACV is also a significant indicator, and what we've suggested is that ACV growth this year is going to be substantially higher than revenue growth. And that was because ACV growth last year was lower than revenue growth because in 2023, there was the benefit of some revenue from 2024 and 2025 that was reported in 2023. So ACV growth was lower.
ACV growth's going to accelerate, and if you take a sort of three-year look-
Mm-hmm
... at ACV growth, that gives you a sense of the underlying dynamics. The other thing that we should point out is that this hosted on-prem transition, it kind of annoying to talk about, but hosted is basically SaaS revenue. It's, it's quarterly recurring revenue, and our business is migrating slowly to more of a SaaS-based revenue model. Now, as you can imagine, that is a headwind compared to if you report it all up front. And so this year, in the first quarter, SaaS revenue was, I think, close to 20% or so, or hosted revenue of our revenue, and that is going to continue to increase. Slowly, our large customers are migrating over to hosted, i.e., ratable revenue, but that's not gonna occur quickly, but it's definitely a headwind to the reported revenue growth that we're looking at-
A temporary headwind-
Yeah
... right, until it normalizes out.
Yes, until it-
So I-
Right
... since I'm not a software accountant, the hosted-
You are now.
You're becoming one.
Most of the experts here are excellent in looking at-
Yeah
... drug pipelines and data.
Yeah, yeah. We'll get to those.
But in terms of the hosted and on-prem, that's you hosting it or where it's done separately? 'Cause that, you're telling me that's a big difference in how the-
Yeah, great, it's a very important question. So in the traditional sense, that's what it means. In a traditional software company, it means that the vendor is essentially running the software and-
Okay
... you know, like, the company has some kind of web interface, and-
Yeah, and you log into it.
Exactly.
Yeah.
That's not what it is. That's not what we're doing at all.
Okay.
The difference between hosted and on-prem for us is essentially nothing, actually. With regard to the compute-intensive work, it's the same. They run no matter whether it's hosted... I'm gonna get to the difference, but just let me make this point. Whether it's hosted or on-prem, the software, the meat of the software, the compute-intensive software is running on their own hardware or their own instance, cloud instances.
Okay.
The difference for us between hosted and on-prem is this tiny bit of code that controls the number of calculations they can run. We call it the license server.
Uh-huh.
It's just this tiny little thing, but because of the rules, if you have one line of code that's hosted, the whole deal is considered hosted.
Okay.
From the point of view of the customer, from the point of view of the cost associated with us delivering the software, from the point of view of every single thing other than how you recognize the revenue, the deals are the same.
Okay.
That's it.
Okay.
Sorry, is that? I know that's just really confusing, but-
It is. It's a transition-
Yeah, yeah
... from that-
Yeah
and that's impacting the revenue. So-
That's all it does, right?
Importantly, you have guidance of 6%-13%. We're now into June.
Mm-hmm.
For the first time, actually, last year, in the years I've been covering you, you actually did, for the first time ever, modify guidance in the middle of the year towards the higher end. That happened last year.
Yep.
Are you holding to some potential that you will look at guidance again at the midpoint of this year and change things?
We try and guide to the most likely range of outcomes, Mike, and we constantly evaluate what that range of outcomes looks like. We're certainly not going to be a company that changes guidance on a whim or with a lot of frequency, but we're certainly going to be taking a hard look at it through the balance of the year. And as we dial the business in, as you can see, our business mix is skewed towards renewals at the end of the year. That's because that's when large companies have their purchasing cycles. But as we get more visibility towards that end of the year, we'll try and narrow what the range is of expected outcomes.
What % of software revenue is in the fourth quarter?
Uh-
Forty-five percent?
It's close to 40%.
40, 40%.
40% is in one quarter?
Yeah. Yeah.
Right.
Yeah.
Okay. Let me hit on the second thing, and that will lead us into Karen as well. But the drug discovery line, which is collaborations, which you had a very nice slide on, is various mix of revenues coming in from-
Mm-hmm
...ongoing research collaborations, et cetera, et cetera. But that was a number that had declined, and people said, "Well, Ramy, if things are going well at your partners, aren't the milestones going up?" But it was a harder look at the probabilities of different things, but it was a decline year-over-year. And so can you speak to helping us gain confidence that partners are doing better if the milestone revenue has been declining, not increasing?
And then, of course, we'll keep in mind that, you know, we've been, you know, as we've said, there's been a shift in focus on the internal programs, which, of course, the revenue from those will be delayed. But sorry, I just wanted to-
Yeah, well, your own drugs are coming. We'll get to that in a second.
Yeah, exactly. So that-
But the-
That's an important part.
... but, Jeff, the revenue guidance for drug discovery-
Sorry
... declined year-over-year.
Yes.
People said, "Jeff, well, if the things are getting better, shouldn't they be going up?
Yeah, there are a couple of factors that contribute to the change in expected revenue for this year compared to last year. First is, there's no doubt that the payment from Bristol in the first quarter was a significant contributor. That's not going to occur this year, so that, you know, just a mathematical change.
That was first quarter of 2023?
First quarter of last year, yes.
Right. Got it.
Associated with that, of course, you know, BMS has made a host of decisions in their portfolio about prioritization not going ahead-
Mm
... all manner of aspects, and so some of that has come out of our expectations for our business.
Okay. Okay.
That's been a change. The last thing I would say is what we're trying to guide to is milestones and transitions that we have control over, not that are in third-party portfolios. We're not in the business of estimating when a global pharma company who's in license a program for one of our partners is making a decision to start Phase 2 .
Yeah.
But we will include when a program that we're in control of reaches a particular transition point, whether it be IND or something else. So that's what's included in our guidance. We expect that revenue to be back half, second half, heavily second half weighted. Unlike last year, where it was first half weighted, we expect the revenue contribution for drug discovery to be significantly second half weighted.
Will that grow sustainably year-over-year, including in 2025?
I have to say, we haven't sat down and modeled all of 2025 and 2026 yet because it depends upon the progress we make this year. If we're successful in hitting the milestones this year on the, in the time course that we expect, then we would anticipate there'd be milestones in the future.
Okay.
But it's conditional upon the success of the programs.
Okay.
Yeah, and not necessarily commenting about 2025, but you know that many of these programs, these collaboration programs, have significant milestones associated with them.
Mm-hmm.
You know, some of them are clinical milestones. There's even royalties on sales, so that's the future. Again, not making a comment on 2025, but you see where these are going, right?
Well, if they... I would love to see them hit.
Exactly.
Okay. All right.
That's right.
Well, uh-
And look, they're progressing pretty nicely. Look at the data, you know. Nimbus just presented data, Structure just presented some data, you know, and so on.
Yeah.
So, you know-
Yes
... to the extent that there are milestones associated-
Yes
... with those are in the future.
Is there a milestone related to any of the obesity Structure compound? We own stock. You own stock in Structure.
We own equity.
Yes.
That is the main form-
Mm
... of value that's being created.
There's no milestone or economics on the deal?
Not in the that program.
Okay.
That's right.
All right.
That's right.
So Karen, you get to finish off with us. Mm-hmm. So tell us about the MALT1 compound. J&J has had very promising response rates in DLBCL, but has been sort of paused because there's some tox and some other issues. But so far, you've reported in a bunch of healthy volunteers. Yep... and cancer patients, but at least in healthy volunteers, it's remarkably clean. Yes. Usually, I think if it's remarkably clean, that's shocking if someone else has just had tons of tox. If that's an on-target thing, maybe they have off-target tox. So can you tell us how confident you are and when you're going to have data on this MALT1? And if you show many, many, many responses, like 30% response rate, that would be remarkable... with no tox?
Right. So as you just pointed out, we have healthy volunteer data. We completed a healthy volunteer study in about 80 people, and the drug was very well-behaved. What we've obviously been looking out for this year is, are we seeing anything different in patients? As you know, we have a global trial ongoing in relapsed refractory B-cell malignancies.
Yep.
And to date, we are very pleased with the profile. We're not seeing anything that says it's completely different in patients. So very well-behaved so far, but obviously, we're still enrolling and aiming towards that recommended Phase II dose, where I think it would be fair to make an apples-to-apples comparison. What we do know, and what we published last year at ASH, actually, is that our compound is about 50-fold more potent than J&J's, and we know they had to dose up to extraordinarily high exposures, where, to your point, there's the potential to overlap into other mechanisms. And so-
So you're 54 more potent on MALT1.
Correct.
They had to dose way up to get potency on MALT1, so they're probably hitting all sorts of other stuff.
This is based on a whole blood assay that we did-
Okay
... head-to-head with our two molecules. Now, of course, the-
Yeah
... Janssen have moved a follow-on molecule into the clinic.
Oh!
And so we think they have a lot of conviction about the mechanism, as do we.
Is that, is that, recently disclosed in a J&J slide deck or something?
Recently disclosed, indeed. And so-
So they like MALT1, and they moved a second compound in?
Correct.
Oh.
AbbVie and others who have large BTK franchises are there. We're very pleased with the fact we have a global, ongoing trial, well-behaved drug, aiming towards that recommended Phase II dose, with the expectation to start combinations with BTK inhibitors or Venetoclax. Really pleased with the progress so far. We've talked about our CDC7 program. Similarly, enrollment's going really well there. That's a drug for AML, where we are through all of our single-patient cohorts. We're now in cohorts of three patients at each dose level, and so a lot of progress.
Hmm
... I think, on the clinical programs.
Think you could have data at ASH?
We don't think so. The cutoff is in August.
Hmm.
We think it's a little bit too early. We wanna obviously have more data in patients before we-
Okay
... jump the gun and talk about efficacy. We have nice safety PK information, which we're accumulating, but we'll probably wait for-
Okay
... a future meeting.
Very good. Thank you guys very much for the update. Look forward to progress this year.
Yeah.
Thank you.
Thanks a lot, Mike.
Good.