JEFFERIES GLOBAL HEALTHCARE CONFERENCE IN LONDON - TWIST BIOSCIENCE CORP
All right, great. Thanks, everyone. I'm Matt Stanton on the Life Science Tools Diagnostics team here. Jeffrey's happy to have Twist with us at the conference this year, joining us from the company. We have Emily and Adam. Thanks both of you for being here this afternoon. You know, Emily, I thought it'd be helpful to kick off, for those that are newer to the Twist story, maybe just talk a little bit about who Twist is, some of the differentiation in your technology, writing DNA on a silicone chip, why that matters, and then also maybe You talked to the evolution of the company, started gene synthesis, you've added a number of markets, NGS, antibody discovery, protein diagnostics, and how that evolution, you know, expanded the TAM over time as well. Yeah. Thank you for the invitation. And it's our first time, Jeffrey, so hopefully we get invited again. So we started Twist in 2013. The idea was to miniaturize the chemistry of writing DNA. I'm a DNA chemist. Our CTO is a DNAchemist. Our CO is a RNAchemist, so we are very much in DNA chemistry. And the chemistry for writing DNA is well-known. Four letters, A, C, G, T. And everybody uses a 96 web plate. It's a piece of plastic the size of an Apple phone duo. And people traditionally make 96 pieces of DNA or oligos on that format. We miniaturized the chemistry, and now we, in the same footprint, we can make one million pieces of DNA, one million oligos. So it gave us massive throughput, much lower cost. I'll spare you the details on fluid dynamics, but it also gave us better quality, enabled us over the years to do the direct synthesis of longer pieces of RNA than the competition. So that's the technology base, platform for the company. And very quickly, actually from the beginning, we realized that we will have a foundry business model. The fixed cost would be substantial, but the viable cost, because we miniaturize everything, would be very small. And so we built a strategy where we will bring volumes on the chip. For a foundry model, volumes is your friend. Again, the viable costs are very small. So to get to bring volume, we developed a new product introduction engine where we will develop newer and newer flavors of DNA to be able to serve more applications. And so we, in 2021, for instance, we were serving a SAM serviceable market of $2 billion. And right now, in 26, we are serving a sum of $7 billion, and we have a line of sight to get to $13 billion. dollars from that that expansion into flavors of DNA we also expanded expanded the different applications that that we will support and the two big ones are the diagnostic field where we enable the library prep and the enrichment prior to sequencing in areas mostly in cancer diagnostic, as well as a smaller participation in rare disease. So that's one big application. Another big application for us is in therapeutics or drug discovery. Drug discovery is a big market that was very, very fragmented. Very few companies cracked ever the $50 million revenue per year. And if you ask five drug discoverers how to discover a drug, you will get six answers. Everybody wanted to do things differently. And so our idea was to enable all the workflows and do it in a way that was based in class. And that has been a very, very successful strategy. And lately, a new way of doing discovery has emerged, and that is AI-enabled drug discovery. And there, in AI discovery, our customers use the engineering principle of design, build, test, learn. The learn and the design is a dry lab effort, and then the build and testing, that's a wet lab effort. And so our customers do the compute for learning and design on their servers, and then they outsource to us the building and the testing. That has been a great source of growth for us. I'm sure we'll come back to that, but we had $21 million of revenue in 2026, which ended in September, the September months for us. And then, sorry, 2025, $21 billion, and 2026 we grew 175% to $59 million. So great, great growth. and we've guided $100 million of AI discovery revenue in 2027. And then overall for the company, last year we got to $459 million and we guided for next year $550 to $560 million. So we have been on the march to deliver quarter after quarter after quarter, year after year, revenue growth. And the last quarter was a very important milestone for Twist, where for the first time we got to address the EBITDA break-even. So, again, the idea of Twist is a technology platform with silicon chip, building a new product and introduction engine on top of it, layered with application support in key areas of life science and then delivering financial performance that we think is best in class compared to our peers. Great, thanks. Maybe Adam, just on the 27 guide, I think you talked about low 20s growth on a 20% comp, kind of rarefied air. I think a lot of focus is on just potential upside drivers given in 26, I think, you guys ended up beating the initial guide by eight percent right so obviously ai is in focus and we'll get to that but any other kind of upside drivers you talk about you know you have a healthy new product cadence um you know maybe academic government comes back you have some capacity ads coming online how do we think about kind of guidance philosophy on tough comps and then some of the potential upside drivers moving through the year no matt thanks for the question great to be here everyone um when we look at the business um we we are always looking for new areas and opportunities of growth and we see that the diversity as being one of our strengths and so for us when we look at 2027 the guidance is based on the same principles we've guided to in the past and that is understanding the demand that's in front of us that we have a high degree of confidence in and so we're as equally excited about the opportunity for upside in areas like ngs and mrd expansion we saw over 20 growth in q4 in our academics we know that that that funding has been under pressure but we we believe we're taking meaningful market share given our unique value proposition in that space and of course would be reluctant without saying ai drug discovery is a huge potential opportunity an area for upside but also an area where we're continuously investing into and part of what we're doing in 2027 is adding capacity to ensure we can continue that accelerated rate of expansion okay and maybe just real quick on academic government it used to be a big focus now it never comes up but it's still you know low double digit percent of your your mix like you said you've seen really nice growth despite kind of a choppy end market. If we do see academic government markets come back a bit more meaningfully in the US, is there further upside? How should we think about kind of those share gains you talked about and then just a potential broader recovery on the horizon here? Yeah, no, I think that one of the most exciting parts about Twist is in any market, we find a way to grow. And as part of that is the value proposition of high quality, high speed, best in class pricing too. And so you have all three together. It's a winning proposition in a strong economy, but it's an even stronger winning proposition in a tough market. And when you're in a rough market, you're not going to be able to do anything. When you're a tough marketplace, as a customer, they're looking for more value, and they're considering switching suppliers. And so that's our opportunity to take share, which we believe we've done. Again, if the market rebounds, we haven't built that into our guidance, but if that happens, that is an area of opportunity for continued upside as we grow with the customers we've now acquired. Okay. And so maybe shifting gears to AI drug discovery, you talked about orders accelerating, I think, throughout the year in 26. That kind of is informing this triple-digit growth you're talking about for 27. Maybe just unpack a little bit more how orders evolved over the course of 26, where you saw kind of strength and what's kind of underpinning that acceleration you guys talked about throughout the years. Yeah, and so for us just to maybe set the stage of how we are looking at our customers, we've been in AI discovery since the first half of calendar 2025. And so now we have multiple cohorts. When it started, we had just a few customers. One in particular was a top five pharma, and also was a native AI discovery company. and they kind of set the the stage for the phases that are being reproduced again and again as we add new customers and there's mostly three phases the first phase is is a pilot very small dollar orders you know ten thousand dollars hundred thousand dollars um um where this and those sequences, targets, and a list of assays that they won't run. If it's a cell-free workflow, we ship the data in 10 days. If it is a cell based workflow, it is in 20-ish days. Then after that pilot, then they come back in a second phase where they will do model building. And so there, the numbers of sequences that they want data for just increases tremendously. It's a very lumpy time where now they want thousands, tens of thousands of sequences delivered in 10 or 20 days. And then when we ship the data, we have to wait because they need quite a bit of time to digest that data. And then the end of phase three, which is for us, is production of drug discovery. So at that point, they're done with model building, and then they go into the turning of the crank where they declare a target with them. We do a drug discovery campaign where there's discovery optimization cycles. And then that ends for that target, and it moves on to another target. And there the revenue is a lot less lumpy. It's a lot more predictable. It's run rate, it's recurring. And overall, the revenue that we get from that phase three is bigger than what we get in phase two. And so we've seen that for now that early cohort, and we see that pattern being repeated cohort after cohort, which means that now it's a compounding plan where we have a new customer that starts a pilot, for instance, in summer. It's only a few tens of thousands of dollars, but we can predict where the demand is going to be a year from now based on our experience. And so then we can build the demand that's going to need to be built and that's what's going be needed for us to match our capacity. And so that's where we were able to see that in 2025 we had $21 million of revenue. It was $59 million in 2026. And we're already guiding at least $100 million in 1027. And we still do see the potential for triple-digit percent growth in 1026. It's just that six weeks into the quarter, we can only predict for sure $100 million. And the difference is, frankly, not a demand question. It's a capacity question. We have a number of projects in flight to add capacity. And as we add capacity and we can build that capacity, we anticipate that we'll be able to realize an even bigger performance in air hydro discovery beyond the $100 million in the guide. Okay. And just maybe one last one just in terms of the customer base. It sounds pretty dynamic depending on where that kind of cohort is. but is there anything to kind of talk about in terms of, you know, legacy pharma, some of these tech bio companies, frontier labs? Is the message kind of all of them are contributing to that $100 million, or is there any way to kind to think about the relative, you know, demand across those three? Which one's the strongest today? I think it's a pretty broad-based, where, you know, all AI-native discovery companies, you know, have needs. there's only 7 magnificent 7 so you can't grow beyond that and then there's only a few very big AI companies in the trillions of dollars of valuation so that's a limited market even though their demand is not limited then in terms of pharma we're seeing a few pharma that are really deep into the AI discovery for instance the example is You know, like Lili, they are building a tune lab that have a fantastic strategy and a very vibrant work around AI. So that's an example of a top 20 pharma that's all in. But there's also a number of pharma datar not yet, I would say, converted to AI. So there's a lot of growth opportunity of people that have not quite started. and even in the companies that have started there's another dimension and that is the mention of the complexity of the modality a lot of the AI work today is done on what I would call more simple therapeutic protein so single domain protein such as VHH and it's great again the business is booming, but over time, we know that those customers that have started with single domain, they're going to want to graduate to the next level of complexity and go to multi-domain antibodies, such as IgG, and even over time go to multi-specific IgG. It's pretty clear that to go from single domain to multi domain to multi-specific, you're going going to need significantly more data points and so uh even for today so basically what i'm trying to say is we see increasing demand in one area one is expanding ai usage into all of pharma and two even in the one that are currently using ai there's going to be increased demand to go up the modality value chain and so you know again it's a it's a question for us of making sure that we can add capacity in a deep discipline way as quickly as we can to to capture that that that demand okay and maybe that's a good segue to the capacity you guys talked about 150 million in capex the next two years 75 this year and 27. um maybe just talk about where that's capacity is going is it mostly on the on the protein side and then also Anything to say, Adam, in terms of phasing of that? I think you guys had always kind of earmarked some space at the site in Oregon for something like this, so you kind of have the roof. Just how do we think about the timeline and cadence of that capacity rolling online, just given how significant it is? It's a significant step up in our CapEx spending and investment in 2027, and it really is driven primarily by the acceleration of the demand in AI drug discovery. And if you look at the business, though, and the way we're doing manufacturing, it starts all on the riders, as Emily talked about. And we've mentioned that, you know, we've got 10 riders online today, and we've put in orders to go up to 16 in the coming quarters a year. And those take a long – there's a long lead time with that. And so that's one area of investment. But if you go down through the labs, we talk about the automation we've done in GeneLab 1. We talked about that in Investor Day. We talked with the automation. We're doing in Gene Lab 2 and the capacity expansion we're making in proteins. But it's also that last step of characterization and adding more equipment where, you know, it may have been one or two pieces of equipment running a characterization assay, and now we need five or ten pieces of the same equipment. So it's really investing there. That's the first step of it. But it'S also about doing it in a disciplined way. And I think it's making sure you've got the capacity on hand to meet the needs of the customer, but you're not putting it in too far in advance as well. So it'S finding that balancing act. And when you're seeing the growth rates we are in AI, keeping up with that demand is non-trivial, to say the least. Okay. And maybe just to put a finer point on this, Emily, I think it would be helpful on the protein side. I think there may be some confusion. You guys are doing this today. It's more of a scale-up, right? So how do we think about kind of tech risk that is or isn't associated with kind of ramping this capacity multifold, what you're doing today, how long you've been doing that, you know, just to kind of level set us on that as you look to kind to lean in and scale it up? Yeah, so from a technical risk, it's very minimal. The vast majority of the CapEx that we are deploying right now is instrument that we already have operating now, and so it's just copy-pasting of it. There is some software integration that has to happen at the same time, and software is always the rate-limiting step, But we're doing really well in leveraging software to build a bigger and bigger menu. So I would say that the technical risk is minimal. It's a question of execution. There's a big cake to eat. So we have to be organized in cake eating. But yeah, overall, I think I would say quite low risk. Okay, maybe to zoom out a little bit. I think there's been this view of the emergence or resurgence of cell-free synthesis relative to the gold standard cell-based synthesis. Maybe just talk about, in terms of the market, why or how people are leaning into cell- free, where that plays, and then also what you're doing at Twist today, maybe in the future. And is cell-free versus cell-based an important attribute relative to pricing, turnaround time, some of the things we think about today, just given how rapidly evolving that kind of ecosystem is for AIDD? Yeah, I mean, in AI discovery, one thing that's very clear is the sense of urgency that that field has is of the chart. We have a huge sense of agency at Twist, and we've met our people, right? It's just fantastic. and the main benefit with cell-free is that it's just weeks faster than a cell-based assay. There is something you have to pay for it. There is you don't get as much material and so you cannot do the full binding and developability data set and so, you have pick which binding and which developability. so you don't have as rich data and and the the quality is not just quite as perfect but it's way faster from a from a production point of view it's about sort of the the same value similar value but it so that's simpler to to make from from an operator's point of view, so it's definitely margins are creative. We have been doing early access in cell-free for a while, just to make sure that when we launch, we have something that's super solid, and two weeks ago, we launched to the entire world. There is a twist on our cell- free. Most of our competitors' cell-frees only serve single domain, and I think we're probably the only supplier at scale of cell-free for single and multiple domain protein and so we definitely see as a growth opportunity for twist then I know the opportunity that we are we are trying to enable is right now the vast majority of AI discovery is done in an arrayed manner so by arrayed discovery I mean one sequence, one tube. And so if you want to look at a thousand sequences, you're going to get a thousand tubes. And then the price per data point is the $150 to $600 per data points. We see an opportunity to drag the market onto pooled discovery mostly because the way we make the DNA, we have, I think, a pretty unique way and we're the only one, frankly, to make pooled DNA at the quality, the speed, the price that we do. And there's a potential to reduce the price for data points significantly. That will not reduce the market. If people have a million dollars, they'll spend a million dollar, but they get more data points. And so we're trying to drive the market towards full discovery just because it's going to give us a better differentiation in the marketplace. Okay. Maybe just the last minute, and this deserves more than a minute, just maybe hit real quick on MRD. I think you have a number of programs you're supporting that are on the cusp of kind of commercial launch here. How do we think about the MRD opportunity both in 27 and over the midterm as well as some of these other tests start to come to market relative to Natera with their gold standards in Naterа? Yeah, so we're very excited with MRD, it's the next leg to our NGS franchise. We're already doing really well in early screening, in multi-cancer early detection, in therapy selection. In MRD, I think Natara has done an amazing job at creating a market for testing if there is a recurrence of the cancer. And where we see broadly the market going is towards high-sensitivity MRD. If you can tell sooner that the cancer is coming back, there's a huge medical benefit, and we see that's where the market is going. So that's the good news. The other good news for Altwiz is that to do high sensitivity, you need to massively increase the number of probes that you're using, from 16 to hundreds, actually thousands of probES. And so there's not a lot of... I think we have a pretty unique platform to be able to make tens of thousands of probes for millions of potential people. And so we have a supply advantage to beable to support that. And MRD was a couple points of revenue last year total for Twist. It's a very fast revenue growth, and we think it's going to be a huge driver of growth in 2027 and in years to come for Twist+. Super. We'll have to stop there. Thank you both. Appreciate it. Thank you, Matt. Thank you.
- Company
- TWIST BIOSCIENCE CORP
- Keywords
- TWIST BIOSCIENCE CORP