Certara, Inc. (CERT) Earnings Call Transcript & Summary

Good morning, and thank you for joining us. My name is Mike Ryskin, and I'm on the life science tools and diagnostics team here at Bank of America. We're excited to be hosting our first Tech Solutions for Drug Discovery Conference. By way of intro, we really see this tech-enabled drug discovery topic as a growing new subsector in public markets, and it's a very exciting opportunity at the intersection of tools technologies, software technologies and computing and biotechnology. The underlying problem that these companies are all trying to address: less efficient R&D at pharma and biotech, lower ROI, falling success rates, higher cost of drugs development. They've all been well characterized and are well understood. But despite years of efforts, we haven't seen significant progress until now. Using genomics, proteomics and other omics technologies to generate vast amounts of data about underlying human biology and then leveraging AI, machine learning and software algorithms to process these massive data sets in an unbiased way, scientists are starting to design new drugs that have a higher chance of success or come to market in a shorter amount of time, either through preclinical or clinical algorithms. It's a sector that's seen tremendous growth in recent years and significant buy-in from pharma and biotech customers and partners as the evidence base in support of these technologies continues to grow. At the same time, we still see a market that's largely underpenetrated with significant future growth opportunity, but there's also no shortage of challenges and potential roadblocks. So given the heavy interest we've seen from investors and all the activity in capital markets with recent IPOs in the field, we thought a 1-day conference focusing specifically on this topic would be a great opportunity. So with that being said, we're ready to kick off. And for our first session, we're excited to be joined by Certara. Joining us is CEO, William Feehery; CFO, Andrew Schemick. With me is also Derik De Bruin, the senior analyst on our life science tools and diagnostics team. Gentlemen, thank you for joining us.

Thanks, Michael.

I guess just to get it kicked off a little bit, for those of us that are a little bit less familiar with the story, can you please provide a high-level overview of Certara's biosimulation offerings? What is biosimulation? What's the value add to pharma partners or customers that are using the technology?

Great. Thanks. It's great to be here this morning, and thanks for the question. So I think it's a great intro to what we do. So at a high level, biosimulation is really just a computer model. It's a complex computer model of what happens when a dose of a drug is introduced to a human body. And what we have is a pretty complex model. It encompasses lots of different organs of the body. And we're trying to model what happens to the concentration over time in places in the body where you want the drug or alternatively, maybe where you don't want the drug. So we're -- what we have here is a very useful predictive model for a drug developer, and we've developed that over the course of about 20 years of iterating, and each year, it's gotten more complicated. So it's useful -- if all it was, was a model of one body, it would be a really useful tool, but we've taken it a lot further. So what we've done is we've developed inputs to the model that match lots of different subpopulations of humans that a drug might encounter in a clinical trial or on the wider market. So things like age, weight and sex are obvious factors that affect drug metabolism and drug dosing, but so are factors like genetic markers or ancestry or pre-existing conditions like diabetes or kidney failure or liver disease. So using these subpopulations, what we do is we conduct -- we basically run the trial over and over again for maybe thousands of times and slightly different person characteristics each time. And so you can conduct what we call a virtual clinical trial. So you're modeling what happens to lots of different people with different characteristics when this drug gets out there in a trial or on the market. So with this, you can do a lot of different -- you can answer a lot of different questions during different phases of drug development. So for example, in a Phase I trial, we're going to want to know the first-in-human dose. And of course, you're going to have some lab data and ideally, some animal trial data. But biosimulation can give you a lot more refined estimate on which to base the trial. And then as the drug moves into clinical trials, biosimulation can be used to determine things like exclusion criteria, so who should never get this drug, or to design the trial more efficiently. So we're taking advantage of all the science we know so we're designing the trial to measure some of the things we don't know about that drug. In later stages, biosimulation is used in drug-drug interaction studies, and it's also used in pediatric translational studies in which we use data from adult clinical trials to predict the dose for our children or babies. And one of the big reasons for the success here is that effectively, you can use biosimulation to bypass or reduce certain clinical trials altogether. And we've demonstrated this for more than 80 novel drugs that have been approved thus far by the FDA. So in terms of where does this go, there's significant value across virtually every therapeutic area. We work with all of the top 35 leading biopharma companies, and we have thousands of biotechs as well. Because our work tends to match what pharma is investing in, so oncology is probably the largest portion of our work. We've got a lot of rare diseases projects. And we work with both small molecules, which is kind of our history, and then over the more recent years, we've gotten involved in some of the interesting questions around the complex biologics. I think a lot of -- and I'll just end here, but a lot of people ask us, "Well, that's great. Did you work on anything related to COVID?" And yes, like a lot of companies, it was very important for us to switch and to help out there. So we work on lots of different therapies and vaccines. We've worked -- I think we've done about 30 programs to date from novel therapies to repurposed drugs to vaccines. And it's been very exciting, and I think it's been good to be part of the work and the conversation that's been going on there.

Great. That was a great overview. I want to ask a couple of follow-ups right on that point you touched on. You mentioned oncology. What about other therapeutic areas of drug modalities? Is biosimulation more amenable to any part of the field than others? Is it more suitable for small molecule versus large molecule? Or given you talked about PK/PD and some of those properties, is it the sense -- is it the case that biosimulation may work better on, for example, oral drugs or on drugs targeting particular disease state, particular part of the body sort of in terms of technology, in terms of the entire ecosystem of potential drugs?

Right, right. Great. It's a good question, but it's a little bit of a complex situation. So biosimulation fundamentally can work anywhere we have some good fundamental understanding of what's -- of the biology that's going on in the human body. And so that's why I say over -- every year, our models grow more complex as we're basically tapping into a greater body of scientific knowledge and we're also increasing the reach of our models to additional targets or different additional systems of the human body. We started out working on small molecules. So that was originally the kind of the use of this, doing PKPD. And ours is really PBPK. It's -- physiologically based pharmacokinetics is really what this is. But the -- a couple of kind of like interesting examples of where we've gone maybe kind of helps. So in addition to all that we've done in oncology and kind of the 80 drugs we worked on, but recently, we've been really excited to develop a virtual bioequivalence model. I'm just going to give you a couple of examples to give you some sense as to where this -- what we can do with it. Virtual bioequivalence, what companies -- what are -- what some companies wanted to do is to prove that their generic drug was virtually bioequivalent, ideally without doing a lot of expensive trials. So this particular drugs were skin -- topical ointments. And so those drugs are -- those types of trials are like large and expensive. We -- our client received the first approval using virtual bioequivalence, having done basically no additional clinical trials. And the FDA wrote a manuscript about that and published it. So it's -- one of the things that's going on, I think, in biosimulation in general has been we are fortunate to have a lot of regulators who are very supportive of the idea, and that's really kind of -- one of the keys here. A couple of other options have been -- we've done a lot of work in avoiding drug-drug interaction trials. We've worked on some quite large cancer drugs, obviously. And there's -- like I said, there's 80 of them. There's been, I think, over 250 label claims where you can see directly where the evidence of biosimulation was used in those drug development. So really, it's -- our aim is to extend it everywhere that pharma companies are interested in exploring with our R&D organizations.

And can you give us any examples or any guideposts in terms of how to think about the cost savings, the time savings, anything that comes to mind in terms of recent programs where a pharma company was using your software and was able to eliminate a clinical trial or move progression faster by -- of the clinical trial process by 3 months, 6 months? Sort of what's the economic benefit we're talking about here just to help us scale it?

Yes. So I mentioned the virtual bioequivalence, so that's the extreme version. We don't see -- honestly, it's unusual for a drug to be approved with no additional clinical trials. But there are many examples of drugs where people have avoided drug-drug interaction trials, for example. It's pretty common nowadays. And those are going to take at least a few months and a fair amount of -- in some cases, a fair amount of money. So we're also -- we've also done a lot of interesting work on rare diseases. So in those cases, the use of the simulation is a little bit different, right? We may not have a whole lot of patients to really even put in a clinical trial. So every data point that we get out, we really want to make sure is basically adding to the case for -- hopefully, for the approval of that drug. And I think there's quite a number of them in which we've helped -- it's a little bit -- it's always a little bit hard to measure versus what someone could have done, right? But I think in a lot of the rare diseases, we've really helped make the maximum use of the data that they have to prove that those drugs really can help that population.

Got it. And I also want to touch on the interplay between software and services. There's 2 major segments to your business. I think for the most part, we've sort of been talking about software. But could you discuss the service offering, how that's connected, how that's distinct, sort of what the difference is between customer base and sort of how each one gets applied?

Right. Yes, it's a great question. So look, Certara has several different pieces, several different software products that are targeted to slightly different parts of this. And what we started out doing in our history was we were basically a software company to large pharmaceutical companies. So we have -- back in our history, we had, I don't know, 20 of them got together and were very interested in this type of modeling. And we supply software to them on a typical basis, right, on a per seat per year kind of basis, right? But what we found as we wanted to get bigger and we wanted to be more influential is that this is pretty complex software. It's not just something that -- let's just say I couldn't download this and go design a drug, right? I'm not one of the drug experts. So this is designed for expert use. And so what we found is that there was a big demand for -- in a lot of companies for us to have a set of drug development experts ourselves that could use our software and do those projects for them. And so that -- when we talk about tech-enabled services, about 70% of it is that. Those are -- I think a lot of investors think about services and they say, "Well, that's going to be kind of low margin, not sticky, not repeatable." But that's not the way to think about us really. I think about this as we have this technology and we have different delivery mechanisms for the technology depending on the segment of the market that we're after, right? So if you're a big company and you have big internal biosimulation groups, great. We have software you can license. If you're a biotech and you have 100 people and you don't want to have investment in those groups, that's great. We've got recognized drug developers, lots of them, internally who can do your project for you with our software and give you direct access to the same kind of technology. So if you look at us right now, we're about 30% directly software, about 70% are tech-enabled services. The only thing I'd tell people about that is it's probably a little bit -- at least conceptually, when we do services work, we kind of -- there's an implied value for the software in there. So maybe we're understating it a little bit if you think about at least conceptually.

Okay. And how is that mix going to evolve in the future in terms of the relative growth between the areas? And as far as you're concerned, is there a part of the business that you want to push more than the other? Or are you agnostic to sort of how your customer sees it?

Yes. I got a lot of questions on that earlier this year. So after we went public, our services grew faster than our software for the first half of the year. That wasn't necessarily a problem because our EBITDA actually grew. So services were not really -- obviously, weren't -- didn't appear to be diluting the EBITDA margin. But what's going on here is that -- as I mentioned, our past was sort of a software company to big pharma companies, and we have a lot of those clients. We've been making a big push to get into biotech, which is causing the services side to grow slightly faster. I think over time, that will probably even out. We're a little bit under -- let's just say that we're relatively underpenetrated in biotech. So I kind of think of it as we're catching up a bit there. But both segments are quite active, and there's -- we're tiny compared with the R&D spend and the potential value that we can deliver here. So I think we'll catch up. We did actually -- I won't -- maybe we'll get onto this, but we're -- we did recently announce an acquisition of a software company, so that will make our -- like our software percentage actually grow as we close that deal. But I think -- I'd like to think about this as there's great work you can do. There's great margins on both sides. And so we want to be able to meet the customers' need, which -- whatever -- however they want to access the technology.

Right, right. And you touched on a point earlier in terms of the regulator acceptance and sort of your interactions with the FDA and other agencies. We just saw a press release come across minutes ago about Certara receiving an FDA grant to further advance bioequivalence assessments. So could you talk about FDA, European regulators, China, Canada? Sort of what's your relationship there? How does that factor into adoption of the platform and sort of recognition of, "Hey, this is a real value add by your customers?"

Yes. Thanks. We did just put out a press release that we had an FDA grant for virtual bioequivalence. There were -- there's 2 things that I think are relevant about that. So I talked a little bit about these complex generics. We were working on topical medicines that they published about. We've also worked more recently on a pediatric topical for acne. And so what that enabled the customer to do was to bypass the clinical trial on children, right? So they already have the adult clinical data. There was no need to go do another one. We knew -- we were able to show what the differences are in dose and activity on children. I think in general, what we've seen over time has been -- and I mentioned this a little bit before, but the regulators have been favorable to biosimulation in general for a long time. And I think the reason for that is a couple of folds. So one is, obviously, this is -- this enables a drug developer to take into account even more that's known about the science, which the FDA wants to -- or the regulators should want to approve. The other thing is that it gives the regulator -- because they use the software themselves, it gives them some insight into what's going on in the clinical trial data. So you can ask questions like, "Hey, what -- in this data that is submitted, what's actually surprising, right?" And if it's surprising then, it hasn't been adequately proven and you can ask questions about that. And just in general, the FDA doesn't -- isn't trying to get drug companies to do more and more expensive clinical trials, right? They want the industry to be successful. So the FDA has been kind of the lead of the drug agencies here. The EMA and the PMDA and increasingly, the Chinese regulatory agency have been also fostering it. We actually supply software to, I think, 17 of the global regulatory agencies, but the FDA is really the ones on the cutting edge here. And their -- they put out basically guidances over time that have expanded the use cases for biosimulation, which they would either accept it or they would maybe encourage someone to say, "Come on and be the first one to try something there." So that's not really targeted at Certara. It's targeted to biosimulation in general. But I think as one of the leading companies here, that's certainly something that benefits us.

Got it. And is that something that you've seen drive uptake among customers? Is that an argument -- is that a selling proposition you can push along, especially to maybe some of the more biotechs and the smaller customers that may not be as familiar with it that say, "Hey, look, here's the FDA. They're all onboard. They're one of our biggest customers. So clearly, they value biosimulation. You should too?" So is that sort of like a channel for you to drive further adoption?

It is, it is. But because -- obviously, if the FDA is using the software, it kind of makes sense to get an idea of what they might see. But the bigger push to use biosimulation is really around the potential for either cost savings or time savings or ideally changing the probability of that drug getting out, right? So where a lot of companies come in for first use of biosimulation has been in drug-drug interaction studies because that's been a use for some years now. There's quite a number of examples. In a lot of cases, you're not going to feel like you're the first one going in like this. And who wants to do the trial if you can do simulation? But even that's kind of a small version, I think, of some of the value that we've added to some of our customers. There are -- as you all know, there are thousands of drug companies out there. There's like, what, 25, 30 drugs that get approved by the FDA in a year. So there's a lot of drugs that don't make it, and there's a lot of money that gets wasted on drugs that don't make it, right? So even if you've helped the company figure that out earlier, you added a lot of value. So we've had these -- a lot of -- most of our big customers have been with us for years. And I think that kind of gives some indication that they're finding value in multiple ways than this.

Got it. Is that -- I mean what would you say are your biggest levers for driving further adoption? And any sort of sense of where we are and what inning we're in, in terms of market adoption? I mean is this the top of the first? Is this the players are running out into the field?

Yes. I think we're still early stages in terms of where the wider field of biosimulation will be in 10 or 20 years. I think in the -- I think quite frankly, in the early stages, Certara must have been little bit lonely, right? This was -- we were a bit the disruptor here, and so it took a long time to get this industry. And the regulator is comfortable that the models are reflective enough of reality that you could use them to make real decisions, right? So that's happening. But we haven't pushed it everywhere. We're launching new products to extend our reach all the time. The science is also getting -- the biologies -- the amount of money that's going in here and the learning in both academia and in drug companies is expanding. So there's a lot of really interesting things going on. So we've -- we're taking advantage of this. We're launching kind of more products here that are specifically targeted as we were able to expand the reach of biosimulation. So I can just give you a couple of examples. Maybe that helps a little bit.

Sure, yes.

So we've launched what's called a lot of work in quantitative systems pharmacology, right? So here, we're modeling the interaction of the drug with its targets. So we did 3 of those this year. One of them -- well, they're really not just this year, but we talked about them this year. Some of them we've launched a couple of years back. But we've launched one which is getting some attention in immunogenicity. So all, basically, complex biologics are going to elicit an immunogenic reaction. All of them will. But you can design the molecule to perhaps minimize that, and so that's an interesting thing we can do. We have a product in immuno-oncology looking at the issue that there's basically -- the treatment is combination therapy, so they're going to be specific by patients. So doing some modeling there can really add some value. And we've unsurprisingly moved our technology over the last, what was this, 1.5 years into a vaccines simulator. So it's not that we didn't work on vaccines, but it's obviously gotten a lot more interesting. And we published some really interesting work about the time between doses and what the dose should be in different populations and things like that. So those are some interesting ones. We've also done -- another one I'll just highlight, which were -- we've talked about and we're launching now, we call it Secondary Intelligence. But if you can model how a drug interacts with its intended target, you ought to be able to model how a drug interacts with all of its unintended targets, so we call those secondary targets. And so this is reusing of the base of the same technology to go look at safety effects and specifically targeting the toxicology area in pharmaceuticals. So that's kind of early stages. We just pushed that out the door. But I think we're excited about it, and it's kind of an interesting example of, hey, we're still finding all the places that we can take it right now in this market.

Yes. That's interesting. I appreciate those examples. And if we look at sort of alternatives to Certara, what's the competitive landscape like? Who are the companies working on the space? And is this something that some of your pharma customers could develop internally? As the field grows, could they sort of bring some of these capabilities in-house over time?

Right. Well, I think that the wider idea of biosimulation, which is that drug development should use modeling and data and analytics in addition to kind of the gold standard of clinical trials, is a theme that you're seeing across the market. And I think you've got other people today coming in to talk about that. But it's a really big theme, and there's different places that you can bring it in. So there's -- right now, you have a lot of companies that I see them as more complementary than directly competitive because this is a big market and you can use -- you can think about the use of modeling it in a lot of different ways as you get into this huge drug development field. The -- a lot of our biggest -- well, a lot of our biggest clients have large internal groups that are doing this, and in fact, they probably effectively are the biggest users internally. So what you see is that they'll be using our software, then they'll be doing their internal work to customize it or make it so that it's proprietary to them. On some of our services, we'll do that as well so that it's not becoming part of the overall software. We're working just on their drug. But I think that if I really had to look at the biggest part of the market that's not us, I would look into -- like into some of our big customers right now. And so -- and then the other parts, I would say -- there's really interesting stuff going on in biosimulation. I know you're talking to companies later today that are taking it into discovery or taking it into modeling of molecules. There's just different concepts that you can do. We're not -- we certainly don't have a monopoly on this whole field. And so...

Yes, yes. And as you try to drive higher adoption among your customers, especially among customers that have used biosimulation in the past on 1 or 2 programs, why do you think that adoption isn't happening any faster? What are the challenges you're running into when you're having these conversations with pharma and biotech?

Yes. So really, the way this field has been going, it's been kind of a steady tailwind. I mean pharma is a big conservative market, and by the way, it should be. But it's kind of like we're converting -- we've had a nice growth, but it's not like all of a sudden everybody wakes up and does something different, right? So a couple of things that have happened. One is what I said before. There's -- every year, year in, year out, our process, we make our models more complex or more accurate, right, or we extend them, let's just say. And we have to go through a cycle with the regulators and with the pharma companies to show that these models, in fact, reflect reality enough to make decisions on. So that takes time, right? We make advances. We have to go through a cycle and then someone has to decide, "Okay, this is good enough. I've got a drug, and I'm going to go take it to -- I'll be the first one to take it to a regulator." And so that part takes time, but we've had 20 years of kind of building up some tailwinds, right? So we can see how that's going to go for quite a while in the future. I think the second piece has been, for a lot of our history, we were really not very large so we haven't pushed our technology into all the places that we can. We tended to go where we had a big customer who got really excited and pulled us in that direction, right? So that means that there's still areas where we need to kind of fill in the white spaces on the technology, and we're going to -- that's kind of the strategy we have as we keep launching new versions and new products here. And then I think the third piece of it is the regulators. So like I said, there are some parts of this that have become pretty recognized where people have done this for a while and so you feel pretty confident when you go in, it would be good. And then there's other parts where you say, "Okay. I'm the first." And so those -- if you're -- we always have to work with a company to feel comfortable that when they are making -- because if you're really using biosimulation, you're making decisions on the course of your drug development well before you get to a drug application. So we have to work with our customers on the science and on the regulatory strategy so that they're comfortable when -- because they're first, that everything will be fine. And so that always takes a little bit longer than one where we are second, right?

Yes, yes. And you've touched on a couple of times on sort of the evolving nature of the business, new capabilities coming on. How do we think about new product launches going forward? And by product I mean new capabilities, new software solutions. Sort of what's the cadence? What -- are there any particular areas you're focused on where you sort of see an opportunity where there's not a lot of solutions available right now?

Well, we have some competitors that are showing up today, so I don't want to get into that, right? But I would say, in general, pharma is moving really quickly, right? We just saw these mRNA vaccines get launched, which were -- 5 years ago, everybody would have never thought that would have happened on this time line. And that's just kind of one example of some of the technologies coming out there. You've got cell and gene therapies which are really interesting opportunities for biosimulation, for example. There's a lot -- I mean there's endless numbers of -- I don't mean to say that in the wrong way, but there's a lot of rare diseases out there with really interesting technologies. Biologics are getting -- the understanding behind complex biologics is getting much greater all the time. So you've got CAR-T therapies for cancer. So I mean all of those are really -- are kind of where you want to bring biosimulation, right, because those are all the things that everybody is wrestling with. We're trying to -- everybody is -- the industry collectively and all the academics are all trying to understand the technology and produce something useful. And so as we think about where we want to extend it, those are kind of front and center. And so we're doing a lot of work in them.

And historically, Certara has certainly been very active in M&A and using capital deployment to sort of bolster those capabilities. How -- sort of how deep is the opportunity set there? How many more technologies, how many more platforms that are out there that are smaller that would provide value for you? And again, any particular area you're focusing on, more on the service side, more on the technology side? Sort of what's the go-forward plan on capital deployment?

Yes. So as you know, we've done, I think -- in our history as a company, I think we've done 14 acquisitions. They've all been very successful, and they've kind of -- they've been targeted at collecting the right technologies that we needed for this. Our most recent one was last month, we're still closing it, where we announced Pinnacle 21. So you'd say, well, 14 acquisitions, that's a lot obviously. But the way I'm thinking about this is that there's a tremendous amount of organic investment opportunity here. So that's really what I think about first, right? And then when we look at things like Pinnacle 21 or other acquisition opportunities, a lot of times, we're looking for either groups of drug developers, maybe small groups of drug developers that can join Certara. We've had a lot of success with that. They get a bigger -- there's a bigger field to play in here, and we have lots of interesting work going on in modeling to bring them in. Or what we did with Pinnacle 21, the larger ones tend to be in the software area. There are some interesting things out there. We can be opportunistic when they become available. And they're strategic. But we're in a nice spot, right? We have plenty of interesting opportunities to invest internally. And then there's -- because it's a growing kind of -- or as I said, relatively early-stage market, there are certainly interesting ideas that people have out there that could become part of us. So we'll kind of -- we kind of keep our eye out. We move opportunistically if we see something there.

Okay. And I wanted to ask a little bit more on Pinnacle 21, especially since you brought it up, a little bit more of a recent deal. Could you give us a little bit more color on exactly the services they provide? Data standardization, sort of how does that fit within the portfolio and sort of why the rationale to move in that direction?

Yes. That's great. So Pinnacle 21 is not a biosimulation play, but it does extend the biosimulation because, as you kind of referred, the issue in -- one of the big issues in clinical development and in biosimulation is data standardization. So we -- our drug company clients have data coming in from different CROs, different sites within CROs, from labs, from animal data, and there's a lot of work that has to get done to put all this data into some kind of format where you can process it and compare it and analyze it and then eventually turn it into something that you could submit to a regulator. So the regulators are ahead of everybody on this a little bit. The FDA, a number -- a couple of years ago, decided that clinical trial data needs to be submitted to the FDA in the CDISC format. So it's the standard format, and this is set by a standards body. And Pinnacle 21 is the software that's used by the FDA and by a lot of the industry to determine whether your data meets that standard. So we're interested because, number one, Pinnacle 21 is a great company. Once you do -- once you determine if your data meets the standard, you can start doing what we call data fitness, and you can start asking lots of questions about data and what's in there. But we need this for biosimulation as well because every -- all of our software eventually creates data that needs to go to the FDA. And as the software gets more complex, the data we're generating were kind of contributing to the overall problem here. So we're excited. Pinnacle 21, in the short run, it's a great company that's growing well and is really, I think, created a solid need for themselves in the pharma area. But in the medium term, there's a ton of value by combining them with our software products. And in the long run, we see that there is -- I don't know exactly how long, hopefully not that long, but there's clearly a huge need in preclinical and clinical around data standardization within pharma. And we think that they have hopefully some opportunities to play in that trend.

Okay. And along those lines, I want to touch on just the broader segments of regulatory science, market access, like you said, the other parts of the business beyond biosimulation. Is that an area where you're running -- you're going a little bit more head to head versus the traditional CROs, contract research organizations, because they do have some regulatory submission services. Sort of how is Certara differentiated there? Is it really comparable? Or are you really sort of focused on biosimulation as you are just...

Right. So about 70% of our business is truly -- is basically biosimulation, about 20% is in regulatory and let's say the last maybe 10% is in market access. So the regulatory is really key to our business because, as I was kind of indicating earlier, clients are very reluctant to work -- to really make decisions on biosimulation unless you get some good regulatory advice. So we need to be credible in regulatory, which means that we gain a lot of that by participating in that. And then the second piece is ideally and in practice, when clients have worked on really interesting biosimulation projects with us over the years, then it's kind of a natural to continue with us for regulatory work. So that portion of our business is pretty efficient. We have a lot of tech enablement of it. So we have, I think, a very good product and a good price point there. But as far as like competing directly head to head with CROs, although I agree that there's similar offerings out there, what we're targeting on doing is just kind of maintaining continuity with the biosimulation and just kind of helping those projects that we worked on continue their acceleration through the FDA. The market access addition to Certara is much more recent. And the reason why we did that is because we kind of see this as what we think that -- we're not a CRO, right? The whole point of Certara is that we believe that companies should take into account everything they could know about the science, the data, the regulatory -- similar drugs that have gone through regulatory approval, for example, and what happens to the drug after it's on -- just taking account all of that as you're thinking about your trial strategy and your drug approval strategy. So it's not -- drugs obviously don't stop when the FDA approves them, right? You have a wider market around whether the payers will accept them and how they compare with the existing standard of care. So we added that as an opportunity we thought to bring that data in kind of earlier on as our companies are making decisions. And so that -- I would say that's still -- I'd say that the thesis there is definitely holding out, but it's still kind of a work in progress obviously. It's still just a relatively small part of us.

It's Derik De Bruin. A couple of questions here. We've covered Charles River since 2003, and a lot of -- that entire time, there's always been some conversation about replacing animals and preclinical studies and doing this, which never really seems to happen, although animal usage has gone down. Where are we sort of in that preclinical talk space of reducing animal usage? Is that something that's still being actively worked on by yourselves and others?

So we do have versions of the simulator that model several of the animal types that are used in pharma, mouse, dog. I forget all the ones that we have. And that is used, and we do maintain them. Practically speaking, however, most of the money in drug development, as you know, is spent on human clinical trials in the clinical space. And so that tends to attract more of the attention. So we share in the desire to minimize that. But if you -- I think what you're seeing in the industry is that people want to do that, but it's somewhere on the priority list.

Yes. So another question coming from our coverage -- our broader coverage universe. Mike and I both have epigenetics backgrounds, and right -- we went through the genomics wave. Now we're going through a proteomics wave as that picks up, and there's epigenetics and metabolomics. I guess when you look at all these data sets, it's like where -- how do you see adding all these different data sets on? And what are you doing to sort of go out and build these? And I guess is there a particular sort of information data set that if you had your hands on you would say, "Oh, my God, this is perfect. We need this?"

Yes. So there's kind of a couple of answers to your question. So the -- what we're doing are mechanistic models. So we're interested in the underlying biology, turning those into basically mathematical equations and making predictions from that. In order to validate them, we need clinical trial data, clinical trials that passed, that failed because we needed to show that what we're predicting is indicative. But then the other question which I think is more interesting is -- I said this is kind of an early-stage field, right? So we were doing our version of biosimulation. There's obviously a lot of people looking at these huge data sets. That's not us right now. But over time, these 2 things are going to marry themselves. I mean I'm sure you'd like to model everything you can and then go look in all these big data sets and see what you can do with AI and machine learning. So I think as this field matures, you will see some -- I don't know, combinations of these things start to become quite useful. But we're -- it's still early stage. People are still feeling out what's actually practical when you develop drugs there.

Got it. I think we're coming towards the end of the allotted time for the session. I think we're going to end on sort of Derik's usual question of what do you think is most misunderstood by investors. Where do you still get questions? What are the misconceptions that people have about the field? Sort of what would be your closing remark on that?

I think that the biggest one has been that just a lot of people said they had no idea that you could do this, right? And I think that has been the thing that Certara has sort of been working against over the years, right? So the drug industry started out with a gold standard of placebo-controlled clinical trials. That always will be the gold standard. But the idea that you could actually predict clinical trials to really, in a lot of cases, a pretty accurate extent using modeling, I think, has been eye-opening not only to the people in the drug industry, but now, as we've gone out in public, to a lot of the investor base. And so I think that's helped us in some ways because it's intriguing. And obviously, we're not anywhere near peak in terms of what you can do with this. But there's lots of opportunities here. There's lots of science that's known and is actually quite useful as you go through the drug development process.

So I'm going to ask a follow-up. Even though I said that was the last question, I'm going to ask a follow-up then. Sort of you yourself said we're nowhere near peak. What is peak then, right? What is -- if you look forward 20, 30 years from now, where is biosimulation? And what percentage of clinical trial process is done the traditional way versus biosimulation to one degree or another, either in terms of size of the trials, the number of trials? Sort of just how good could this technology get in the future?

Yes. I mean ultimately, the goal here is that all -- everybody's drug development and particularly clinical strategy is going to be informed first by biosimulation, right? So let's go model what we think will happen and then conduct our trials to go prove that that's the case. And that kind of philosophy leads to good drug development and leads to good data, and it will ultimately lead to a more efficient industry. For that to happen, obviously, we've got -- we have some more adoption to do, and we have some more extension of the technology, right? So we've shown that this is certainly possible in a lot of areas. But we haven't hit everything yet, and so we're pretty excited to keep pushing.

Great, great. Thank you so much. On that note, I want to say thank you for joining us, William. Andrew, I know we didn't get many questions to you, but glad to have you here as well. Clients, thank you for listening in, and we'll talk soon. Thanks so much.

Thank you, Mike.