Welcome to the final day of our Leerink Global Healthcare Conference. I'm Puneet Souda. I cover life science tools and diagnostics here, and it's my pleasure to be hosting the Twist Bioscience team, CEO Emily Leproust, and Adam Laponis, the CFO, joining us. Thanks for being here.
Thank you for this year.
Yeah, great.
Lot to talk about, you know, on the NGS and bio protein expression services, the AI capabilities that are ramping up. Maybe to start with maybe a high level question. I mean, you've been able to put up strong growth. You've delivered 20% growth for fiscal year 2025, 17% last quarter, you know, and versus the, you look at the rest of the sector and even the peer DNA, oligos, probes, and kits and life science tools companies, they had a tough year last year. I mean, you know, how much of this is, you know, share taking versus new product introductions versus new account growth? Maybe take us through at a high level and then maybe just on the NGS ramp and/or the SynBio and other product additions that you've had.
Walk us through that and maybe give a sense of how much of this is just, you know, you gaining more share versus expansion.
No, thank you for the question. It's the key question. It's a little bit of all of the above. I think if we step back to the strategy, where the silicon chip and our economics is such that the variable cost is very small. Any net new products, any net new that we can do is and it massively beneficial to the financials. We've been very focused on revenue ramp, and that has been done through new product introduction. We are an NPI machine, and we can see it in the SAM. It was $2 billion in 2020, now it's $7 billion, you know, five years later.
That's the launching new products. In addition, we also are deploying what we call commercial excellence to outcompete, outperform our peers.
Mm-hmm.
therefore we're taking shares. Once we land into an account, we radiate into more labs. It's not an accident, it's a result of us being highly differentiated on the product side, on the speed, on the quality, on the cost, but differentiated in how we reach the customer with our pre-sale, post-sale support, our e-commerce, our APIs, our B2Bs. We are firing on all cylinders to make sure that at the end of the day, we delight the customers. If we go maybe even deeper on the new product production, a big driver for us last year was the trend of AI drug discovery.
Mm-hmm.
A year ago and a couple months, we had not sold any data.
Mm-hmm
For companies that are doing AI-driven drug discovery. Now, it's a very fast-growing product for us.
Mm-hmm.
That shows kind of the speed at which the Twist team can identify a need, respond to a customer need.
Mm-hmm
We have scale. You can come to us with thousands of sequences. We have speed. We can deliver that data to you in, you know, 15-20 days, which is best in class. We have the cost structure. Those companies that have a fixed budget per campaign, we're able to delight them. Last but not least is with the quality. It's a given that in drug discovery, if you don't have the quality, if you can't be trusted, you're not in business. Back to your question, it's really a combination of the two.
Yeah. That's great. You guys have been good at, you know, capitalizing on what is an emerging trend or an emerging growth market in the MRD, NGS, and now AI, and we'll get more into the AI. Adam, you know, at a high level, you know, before I get into each segment, I want to just see how you're balancing the need for, you know, growth investments while staying focused on both gross margin and adjusted EBITDA breakeven by fourth quarter. I mean, I guess, speaking of commercial leverage, what's the leverage in the P&L?
Thanks, Puneet. If you look back over time and you look at the last, I'd say about 13 quarters
Mm-hmm
We've seen over three years about a 24% CAGR in revenue growth.
Mm-hmm.
During that time we've also seen about 20 points of gross margin expansion from the low 30s to now over 50%. The majority of that expansion is what Emily just described, that low variable cost, where about 75%-80% of the revenue growth drops to the gross margin line. What we've also seen during that time is significant discipline in how we manage our operating expense investments. For about the three years it was relatively stable. We did take a step up in Q1, but that discipline still continues. The discipline now is, let's make sure we're titrating the expenses, not just to make sure we hit adjusted EBITDA breakeven in Q4, but also make sure we do it at the maximum velocity for revenue growth.
Got it. Let me jump into what Emily was talking about with the AI and drug discovery. You know, from my view, it's you know at the end of the day it's protein expression services then you're running assays with that and producing data. There's you know AI-driven dry lab to wet lab service work. Obviously, you know, it's a theme in the market. I guess I wanted to understand from your perspective, number one, your differentiation in the market when we look at the RUO discovery work, which is largely for pharma companies, but academic labs and other service providers also provide services of protein expression.
You know, your advantage seems like it's on the DNA writing side, so you can write faster and then go to the next door lab, so maybe save more time, and maybe that is the differentiation. Maybe just walk us through the factors that are differentiation, because the protein expression market has been around for a long time. Obviously AI is boosting it. Maybe just give us a sense of the differentiation and, you know, what is meaningfully material here that you have seen and how sustainable is that?
That's a good question. You're correct that expressing a protein, an antibody and characterizing that protein, you know, it's easy. Everybody can do it in 15 days.
Mm-hmm.
Where our differentiation comes from is the scale. Where our customers, they don't want one, they want thousands of proteins, and they don't want to wait 1000 times 15 days. They want it all in 15 days. That's where the technology that we have comes from. Making the DNA at that scale very easy for us. We have capacity for three million genes a year, so you can see, I mean, that's per day. Expressing the protein at that scale, even though we don't do it on the silicon chip, but we leverage the same automation capabilities that we've built. At this point, I will call them almost extreme automation.
Keeping track of everything with the paperwork in a way that is as little human decision as possible. That software layer that we've built over the last 13 years that all enables us to get to the production of protein in 10 days for HEK, in 13 days for CHO.
Mm-hmm
Thousands of sequences. Then after that, the protein gets split and goes to all the different tests in parallel. Do people want affinity testing? Do they want functional testing? Do they want epitope binding? Do they want developability? Do they want expressibility and so on? They all get split to the machines. Because those machines are also automated, it's not a PhD scientist working nine-five, sitting at the mass spec or the HPLC, and the HPLC doesn't run overnight. It's a machine that runs the mass spec or the HPLCs, and it runs 24/7. We have more capacity. We can get done faster. The data comes back at the end and is shipped to the customer.
All of that, again, 15-20 days, for thousands of sequences. Back to what's the differentiation.
Mm-hmm.
It's a little bit, I know it's gonna speak to people, but it's a little bit like making spaghetti for four or making spaghetti for a thousand people, right?
Mm-hmm.
It's a different skill.
Mm-hmm.
This is the, you know, the AI drug score is almost perfect for the Twist moment.
Mm-hmm
'Cause we have the infrastructure, the hardware, the software, the extreme automation infrastructure to be able to almost on a dime realize there's this market, put it in place, deploy it in a fab.
Yeah
Be able to grab the market.
Maybe just two brief questions on this topic. You know, when you have competitors in the market that have been both public and private companies and service providers that have been providing the service in terms of expression of proteins, and you could say Adam or Charles River, Aldevron, some of the other service providers and whatnot, and sometimes they're projecting that they can do a better job with the proteins because of their expertise. How much does that matter in this more scaled environment? Or is it more about numbers of expressions or sequences versus more on the actual protein design and protein engineering?
The protein design or protein engineering of the sequence.
Mm-hmm
is done by the customer, right? We don't do that at all. The customers come to us with, "Those are the sequences.
Mm-hmm.
That's the target.
Mm-hmm.
Those are the assays we want to run." That is the decision of the customer.
Mm-hmm.
You're absolutely correct that expressing protein is a skill, and for those customers, quality is paramount. I mentioned it last. You know, I talked about scale and speed and cost. I mentioned it last, but almost because it's a given. If you don't have quality, you know, when you work with pharma companies, if there is no trust.
Mm-hmm
You're not in business. Trust is absolutely the foundation.
Mm-hmm.
We've been able to CHO that the quality of our protein expression, the quality of our data, is very stable top line. There's even actually some tests where you get quality that is better than once you automate. There's a test in particular where it's so well known that the test is a little bit flaky that you have to do it three times. The customer expects three data points. We used to do it manually, and like everybody else, we were providing three data points.
Once we automated it, we realized that the CV of the three data points were so tight that now actually you only need one data point, because now you can trust it because the reliability of loading the instrument is so good. Quality is stable state. It's hard to get there, but we're able to provide the top-of-the-line quality again at scale, great cost, great speed.
Got it. Just briefly on, you know, how I'm sure you had plenty of time to think about this market. How would you quantify this market, like how large it is? Again, the new piece here is the dry lab that is feeding, or the more, model work that is feeding into, you know, into the wet lab. Maybe on sort of how large a market this is in your view, and then it does seem like it's a market that's lifting all the boats. Is that fair to say in this because of the nature of, you know, discovery and the research process?
Well, hopefully it's not lifting all the boats. Hopefully it's lifting only our boats.
Maybe the speed boats.
Yeah. Maybe only the Twist's boat gets lifted ideally, right? We want to be number one in a world where there is no number two.
Mm-hmm.
Definitely we're building capacity to make sure that we have the choice and you don't need to consider anywhere else. In terms of the market size, I don't have a number today to provide, but we're definitely quantifying it, and at the right time, once it's validated, we want to share it with the world. What we see is there's net new dollars coming in, right? There are some of the Magnificent Seven. There are new startups that are AI drug discovery startups, and they do not have a wet lab by design in the foundation of their company. That's net new that are coming into the market.
We see in the top 20 pharmas, there's a spectrum. There's pharmas that don't seem to do any AI. They would be late adopters. As expected, there's also a number of early adopters of AI in the top 20 pharma. What we can see is, there's some teams, even in the early adopters that remain traditional R&D discovery. There's some new teams, we are still trying to tease out if it's net new dollars that those AI teams are spending or if it's a real internal reallocation.
Mm-hmm
From traditional to AI and there is a blend. We're still characterizing that. At the same time, for us, the strategy in therapeutics for Twist has been build the menu so that we can meet the customer where they are.
Mm-hmm.
Early on our strategy at Twist was to be strong in drug discovery as we are strong in diagnostics. Those were two key pillars that we wanted to reach. When we looked at the market of drug discovery.
Mm-hmm
What we saw is that there was very few companies that were able to break the $50 million revenue barrier. You know, our view was because the market was very fragmented, and you had companies that focused on a niche. You had a bunch of them, but they could never break. At Twist we said our strategy was, let's add everything. We have in vivo, we have in vitro, now we have AI, even in vivo we can do single cell, we can do the regular hybridoma analysis. In vitro we can do phage display, yeast display, whatever type of displays you want.
Mm-hmm.
We have lots of libraries. We're still providing the fragments, and genes, and preps, and IgG, and HEK and CHO. Again, we'll meet you where you are, enable anything. I think that has been a very successful strategy.
Mm-hmm.
Last year we had $111 million of revenue in therapeutics drug discovery. We not only broke the $50 million-dollar barrier-
Mm-hmm
We are now double that. We grew 21% and 25% last year. The strategy is working. To complement that strategy, there was one piece where we were not doing well, or I'll say we were doing equally bad as everybody else, which was in bispecific.
Mm-hmm.
Bispecifics are notoriously difficult to express purely.
Mm-hmm.
It's hard enough to make an antibody where you have two plasmid, but with bispecific, you need four plasmid. The ratio is important and very difficult in any format to express pure bispecific. With the licensing of the B-Body technology from Invenra, now we have a bispecific format that almost auto-purifies. It comes out pure, you know, without effort. Therefore, we think that we'll be the first company to be able to provide bispecific in high throughput.
Mm-hmm.
Again, that's us looking at the menu. There's one part of the menu that's not differentiated, not strong, and now we have in bispecific a solution that we believe is strongest. That will continue to fuel the growth. It's our strategy is working. We'll keep adding new product, the bispecific is one example, and we'll keep deploying commercial vigilance to grow fast.
No, that's great. It seems that you are spending more investments and resources in the scaling of antibody discovery as well as early stage AI discovery. Just maybe at a high level question on the biopharma therapeutics antibody drug discovery side. I mean, you were involved in that area early on. It was a smaller piece of your revenue, and it is a long game to get to a phase I, which is usually the benchmark. I mean, then you sold some of those assets, I believe 50% to Astellas, and now you're, you know, doubling down in bispecifics. Maybe just, you know, thinking about antibody therapeutics, drug discovery piece, you know, how important is that to you know, longer term?
Obviously, you're providing AI services, which is a little bit earlier than that. I'm just trying to understand how much you want to be involved here because you know, NGS and SynBio and other areas are still very important, and you are a product company. How should we think about the mix versus product
That's a great question. For us, you know, if we wake up in the morning, we don't want to be number two, right? If we go into an application, we want to be number one. In liquid biopsy, we are number one in MRD. For high sensitivity MRD, we believe we enable the best tumor-informed MRD. In therapeutics, our strategy had been we want to become number one. This is an area where the match between the strengths of our technology and what the customers needs is as near perfect as can be.
Over the last year, through our NPI introduction, we have been chipping away at everything that drug discovery scientists at the bench, head of biologics
Mm-hmm
You know, may need. The strategy is paying off. We'll continue to push both our products and our services. It's true that sometimes we're able to get milestones and royalties.
Mm-hmm
For monospecifics. I was just speaking about monospecific. We were pushing really hard milestones and royalties in the past.
Mm-hmm.
We've changed a little bit our strategy. We think that first of all, it's really hard to get milestones and royalties for monospecific. We were just delaying the
Mm-hmm
Delaying the contract signing, the negotiations. What we are actually more interested in is revenue.
Mm-hmm.
We've changed the approach where we are trying to get fee for service and sign a contract as quickly as possible, and do a great job, repeat it, repeat with the same customer, find new customers. Over the last few quarters, there's been less emphasis on getting milestones and royalties.
Yeah. It's more service. Yeah.
More services. Things will change with bispecific-
Mm-hmm
Because now we believe that the Invenra B-Body technology has unique advantages in terms of speed of discovery, the ability to mix and match multiple antibodies very quickly. We think that we will be able to get more milestones and royalties that are less back-loaded and more front-loaded. We will be pushing maximum royalties for bispecific.
Got it. Maybe, Adam, in terms of, on the gross margin size now with this, data offering that you have, maybe just walk us through, you know, as this mix shift happens, you have more services versus product, how should we think about the gross margin? Obviously, you know, gross margin improvement has been, part of the story. You know, you have steadily reached, you know, last quarter you did 52% gross margin and, you know, there's an expectation that that should improve, but maybe just walk us through that.
No, happy to. The way we think about it, and I hit on this earlier, is, you know, any incremental revenue growth on either side of the business, whether it be on the DNA synthesis and protein solution side or be on the NGS side.
Mm-hmm
We're looking to target about 75%-80% of that revenue growth dropping to the gross margin line. With the services element of AI-enabled drug discovery and protein characterization, the first order we had was really coming out of that combination of, hey, we have the team that does some of the services work out of Boston.
Mm-hmm.
They had the equipment, they had the knowledge how to do this at lower throughput. The first order, it was a bench scientist.
Mm-hmm.
It was at lower throughput, and the margins probably were not at that point in that first order at that target level. That's consistent with how we do a lot of our new product introductions, get it to the market, delight the customer, and then as it scales and you have a proof of concept, then let's now automate heavily.
Mm-hmm.
What we've been able to do over the last year, and we're still doing it today, is we've moved a number of those assays into our Wilsonville facility.
Mm-hmm.
We've automated them with a high throughput, so it's now a robot and an operator running 24/7 versus a scientist in a lab. We see that margin profile on both sides of the business.
Mm-hmm.
We're excited to see that continue to ramp.
Got it. Maybe just on the OpEx side, I mean, you highlighted growth in OpEx to support the e-commerce rollout and the sales force growth. Maybe, you know, just tell us about, you know, changes in store there, and the motivator for that change and sort of, when do you expect to reach, you know, full productivity for some of these new sales reps?
Yeah. I'm happy to. Sales team in particular, we've always invested ahead of the growth.
Mm-hmm.
It takes, you know, three-six months to fully ramp up a sales representative. What we see is we know we have certain territories that we've gotten oversized or even in some cases we had instances where the reps were associated with a single zip code.
Mm-hmm.
What we saw is by, hey, having a customer that may be in an academic setting in that same zip code as someone who's negotiating, you know, a multimillion-dollar AI drug discovery contract, it's difficult to split the mind share. What we've really done is within some of these larger territories, we focus some on the academic and we focus some on the bigger, I'll call institutional, sales. We've been able to do that in key markets like Boston and Northern California.
Mm-hmm.
In terms of the other side of the business, we did just launch our first NGS e-commerce site. It's just about a month ago now, I think.
Mm-hmm.
We are in the early innings of enabling e-commerce. If you think of the initial customers, some of the big diagnostics, you know, a $1 million order probably not going through e-commerce. The part of that strategy of we know that we are under-penetrated in some of the smaller academic institutions, and this allows a more seamless transaction over time in there, so excited about that opportunity as well.
Talking about NGS, I mean, there was the air pocket that you had talked about, but then you noted about 18% growth in 1Q, that large customer, who has since restarted purchases. But maybe how should we think about the customers, you know, sort of spending ramp here and any other considerations in the NGS end market, just given that, you know, overall competition, any other considerations that you have for growth?
Yeah. I think, you know, for NGS, we are very excited about where the growth is going.
Mm-hmm.
We talked about the one customer going through their commercial transition over the last couple of quarters. Very excited to have them on the other side of that and seeing that account continue to growth.
Mm-hmm.
We've talked about publicly the idea of we have a path back to 20% revenue growth in NGS as a whole by Q4.
Mm-hmm.
We're excited about the progression we're seeing week over week, month over month in the account. Relationships are really strong, and what we're seeing is customers don't leave Twist. What we're seeing is the inherent, and we had this in some of our previous presentations, some of the inherent ordering patterns are inconsistent week to week, particularly when folks are managing inventory levels and whatnot. We're here to meet them wherever they are.
Mm-hmm.
Some accounts order once a quarter, some accounts order twice a month, and we're here to meet those accounts wherever they are. Excited about that future growth and see, as we continue to ramp, we're seeing it also be more predictable as well.
Got it. On the MRD, I think you previously noted about 1%-2% growth contribution in FY 2026 for MRD. My question is, you know, on, are these mostly sort of production or are they, you know, clinical testings that are ramping up in order to spec that into a potential MRD product? Maybe the first question, how good is, you know, in terms of once you are specced in, maybe just tell us how sticky that is. The second part, maybe I should just ask that question as well, just given the time, we had a large trial failure on the screening side in the market. What is the implication of that to you guys? Obviously, there you're supplying products to that customer.
Yeah. First, we're here for all our customers. We're very diversified. We are not sensitive to one particular test. If you're talking about the GRAIL test, I still think it's aggressive. I do it once a year on my birthday. I think I rather find a tumor when it's tiny rather than it's big for the most deadly cancers. In terms of stickiness, for MRD tumor-informed, if that's the question.
Mm-hmm
It's very, very sticky. What we provide is thousands of probes at amazing economics. We used to provide them in five days, which is, you know, nobody else can do it. We're moving to MRD Express, we'll be able to provide the same thing in one day instead of five days. From a competitive positioning.
Mm-hmm
I don't think anybody can provide that many probes that quickly at the price point. We think that's once adopted. Right now, they're still in the early phase of R&D, validation, verification, clinical trials. Once it goes to production, the scale that's coming our way is gonna be great for us, great for partner, and really good for patients.
Okay. Last one. Investor Day coming up in May. Any quick thoughts? Preview.
If you're not there, you're gonna regret it.
That's a nice way to put it. All right. Okay.