Nautilus Biotechnology, Inc. (NAUT) Earnings Call Transcript & Summary

Hey, everyone. Good morning. My name is Tejas Savant. I know lot of you in the room. I cover Life Science Tools and Diagnostics sector here at Morgan Stanley. Before we get started, I just want to read the disclosure statement. Please see the Morgan Stanley research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, do reach out to your sales rep. It's my pleasure this morning to host Nautilus and speaking on behalf of the company, we have Sujal Patel, CEO. Thanks, Sujal, for joining us today. Maybe just to set the stage, for folks you're in on the webcast, not as familiar with Nautilus, could you just give a brief history of how the company came into being. And where do you see Nautilus fitting in the evolving Proteomics landscape?

Great. Well, just let me start by thanking you and Morgan Stanley for the invite to the conference. Happy to be here again. Let's maybe start that question just, just at the very beginning of what is Proteomics, right? Proteomics is a study of proteins. Proteins make up all of the functional pieces of your cell, they do all the work, and I like the genome your DNA doesn't really change from the day you are born the day you die proteins are dynamic. They change every day in response to disease, in response to what your body needs. And so measuring proteins is really critical to understanding what's going on with biology. And most of our FDA-approved drugs target proteins, most molecular diagnostic target proteins. And the problem that we have in the world today is that the technologies to measure proteins in sample are incomplete, they're hard to use, and they really don't give you a precise signal. And that's in stark contract to genomics, right? Over the last couple of decades, we as a scientific community have conquered genomics. Genomics, they can take a sample from any organism, they drop it on genomics [ sequence ] every time it works in the day, I've got the full genome. And that's just far from what we had in the proteomics world. And so many different companies have tried to attack this problem in different ways, trying to make incremental improvements. And Nautilus, which is founded in 2016, really is the only company that's taking a completely new approach to trying to measure proteins example with the goal of building an instrument that's able to measure 95% of the proteins in a sample from any organism and do it in a push button simple, easy method from sample to answer to bring proteomics in parity to where genomics is.

Got it. And talk to us about how it's differentiated from some of the peptide sequencing efforts out there, Quantum SI and Encodia, et cetera. And then you've got the targeted platforms like in Olink and SomaLogic as well. And then, of course, the [ Seer ] as well, which is sort of positioning themselves as complementary to mass spec. So in that ecosystem, where does your approach kind of fall?

Yes. Well, so, you kind of have outlined 3 different camps of approaches to tackle the proteome with us being the unique one, peptide sequencing being a new one, traditional assays being the standard way, right? And each of them has different pros and cons, right? Their traditional assay Olink, SomaLogic as well as a number of products that are inside of larger organizations are really focused on building antibodies or affinity reagents that are specific to different types of proteins. And then putting them in an assay so you can measure those proteins and example. Those approaches are good, but they're not super sensitive. And you have to build an antibody for every single different target. And it's really infeasible to go and build one for all of the 20,000 gene coded proteins and do that in that effective way. And even if you could do that, you would then not be able to see all of the different forms of these proteins, the different modifications. And those modifications and isoforms have a profound impact on the distribution of these molecules and cell, their function, their degradation. And so if you want to understand biology, you have to understand that complete picture. So that's one camp. Another camp you mentioned where these companies that are focused on peptide sequencing. There are a number of new initiatives out in the world that are trying to, like we sequence DNA trying to sequence protein and for sequencing protein is impossible. So what they do is they break these proteins into little pieces and try to sequence very short fragments. And it is a very complex hard thing to do, and it's a hard thing to do at scale. And scale matters in proteins because just in a single cell, there's a million protein molecules on an average. If you want to understand what's going on in the traditional pharma sample, for example, you would need to see billions of molecules and analyze them in a precise way. And so those types of approaches are good for answering niche types of questions, but really don't scale to being able to look at a whole proteomics. And so what our -- the unique thing what our approach is that we're really focused on bringing the scale and the sensitivity to the problem that matches the customer samples and the needs inside of biopharma, the goal for us is to be able to analyze sensitively any of these samples that have 100 to 1,000 cells or blood serum so that we can help our customers improve the efficacy and speed of drug development, we can help to build more precise diagnostics which can help them with basic science research and so forth.

Got it. So on the last earnings call, Sujal, you talked about balancing time to market versus minimum product specs. What specs do you see potentially receiving some ground on version one of the product? And you talked about even on those sort of note down specs, it could still be a game-changing sort of introduction to the field. So talk a little bit about what are those game-changing features of the platform that you think cannot be met with other sort of efforts out there?

Yes. Maybe that's a great question. Let me back up just for context and then we'll kind of get to that, right? So the products that we're building, much like the first time this next-generation genomic sequencer was built. It is a hard complex task and involves lots of components. It involves different pieces of science that have evolved over the course of the last 6, 7 years that we've been building these. And there's a huge amount of effort bringing those things together -- just like Illumina did or Solexa did, which was the company Illumina bought to get into NGS, there are some compromises that you would make at V1 to go and get to market sooner. And from our standpoint, we were going to make some of the same sort of very pragmatic compromises. And for me, the important thing is to be able to talk to a customer about the fact that you buy the instrument today, and just consumable updates are going to get you to the full specifications that we had originally disclosed because that means the investment they make in that instrument, that roughly $1 million instrument, that instrument isn't something you have to rebuy afterwards. And so you asked like what are the sort of compromises that you would make? I think that the first thing I'd like to just kind of head on is the compromises that we would make are really informed by conversations with customers and looking at the competitive landscape. So to give you an example, we need roughly 300 different affinity reagents in our system to be able to achieve 95% of coverage of the proteome, if the coverage was 50%, 60%, that's still well beyond any other method that exists. Of course, we would still launch. Dynamic range is another, right? We have a chip in our system that has 10 billion spots for proteins. That's 5 orders of magnitude in excess of where the competition is. If we need to make compromise there, have an order of magnitude or even an order of magnitude, that's not a problem. We're already so far above where the competition is, and we can resolve those types of things with consumable updates in the future. And so those are the types of compromises that we'll make initially. And we've pressure tested that with a lot of our early customers. And I think that pragmatic is certainly how I describe those.

Got it. On the point on the multi-affinity binder, Sujal. How confident are you that you can get to 50% to 60% by the time of the V1 launch?

So I think that I will hedge my answer in that regard. We said that on the last conference call, we said we continue to target roughly the middle of next year to have our commercial launch with platform. A big piece of that depends on us having enough multi-affinity probes that operate within specification on our platform to be able to get to 50%, 60%. And I think that we are extremely confident in the development path and we're confident in that rough time line, but there's still a lot for it to be done.

Got it. And then in terms of the sensitivity and ease of use, I think you call them sort of the nonnegotiables on even the V1, what is the minimum level of sensitivity you think will be required to drive meaningful adoption?

Yes. So that is a simple one. We are a single molecule counter. There is nothing else our play. We can't have any sensitivity other than single molecule. And in terms of single protein molecules we are the only platform that is even conceived of that can analyze a single protein molecule. Everyone else who uses those words, breaks their proteins into peptides and then those get analyzed. So for discovery platform, single molecule sensitivity is what we do. That is the only mode that we operate in. And it is one of the key things that customers look at for our platform. To give you an example of what's the practical use of that, right? If you are developing a drug for a particular disease, the first thing you would do is look at healthy cells, look at cells that are affected by disease, look at the cell surface to look for differences. The differences between healthy and sick cells are very, very minute, sometimes even single molecules that are changed. And so for a pharma company, being able to push that threshold down and see things that are more and more rare, those could be the biomarkers that are the best potential drug targets for the next generation of compounds that they're going to go and put through the FDA and get out into the clinic. That is an incredible value proposition for customers. So that absolutely is nonnegotiable in our platform. Ease of use is the other. We set out, when we founded this company in 2016, we set out to go and build a platform that could be used by any biologist. Any biologist who wants a proteome can get a proteome from any sample from any organism. And that is a nonnegotiable goal for us. Mass spectrometry-based Proteomics workflows are extremely complicated, even companies like [ Seer ], which you mentioned earlier, which make it incrementally easier to prepare samples for the mass spec, they still don't fundamentally change that mass spec based proteomics is extremely complicated. We are bringing an easy-to-use platform to market so that any biologist can get the types of answers that are going to translate into massive discoveries.

Got it. You talked about not requiring an instrument upgrade as you go from V1 to V2. It sounds like you're further along in terms of instrument development at the moment. Could you just give us a sense of what remains to be done at this point versus middle of next year when you launch?

Yes. I think that when you kind of look at all the platform pieces, right, we have to have an instrument. And yes, the instrument is -- it's on a time line that would certainly meet that middle of next year because it is just engineering, and we are operating on the final generation of instrument now in our facilities. The other thing -- the next thing which you have to have is, you have to have the single molecule array, right, something you and I have talked about quite a bit. Our platform depends on taking molecules from a sample and spatially separating them on a chip, and that chip is a combination of a chip with sort of semiconductor fabrication processes, chemical functionalization and then an assembly that's built into a flow cell. That system and our single molecule deposition are also in good shape, good enough along that we're willing to put the pencil down and say, yes, that's good enough for inputs good enough for launches. Evolve those processes for scale, let's get the production manufacturing scale, what's set on, right? The next 2 sections are the actual multi-affinity probes, cycling them through our sample, one after the other and being able to detect those binding, and that's the area where we continue to spend a lot of our time and energy in building those multi-affinity probes and figuring out what the right mix of them will be. And then the last piece, which is kind of tied over that is really the bioinformatics and the data science associated with turning that data into identification set of proteins. And so that's where the predominant effort is today is on those pieces as we go and try to get to that set of initial multi-affinity reagents that we'll use to launch the platform.

Got it. And on that point, Sujal, the repeat sort of buying [ watch cycles ] that you need to do, how many of those cycles do you need to do? And then is there any concern that, that might sort of change the signal or sort of add noise?

Yes, the question you bring up is a question that many life sciences instruments have to deal with, right, in particular in NGS, you have to deal with this. So in our assay, we flow one affinity reagent in, we go as we do our analysis. We wipe that slightly, send the next one in over and over again. We need 300 affinity reagents to get to about 95% coverage of the proteome, which is our target. We do 2 at once and 2 [ colors ], so that's 150 cycles. And across any number of cycles, there's always a degradation in signal. So you start a little bit high and you have a little bit of degradation. But we've done enough experimentation through -- we've talked about in our earnings calls that every quarter, we're pushing that higher and higher. We're up at the 100 mark now and are comfortable with the degradation of signal and we're at. So I think that's not going to be a concern that's going to stop.

Got it. Makes sense. And then one of the features of the platform you've highlighted in the past is just the ability to characterize proteoforms in PTMs. Is this important in your mind to have on V1?

So I think that you should think about proteoforms and different isoforms as kind of a short- to medium-term type of focus for us, right? I think that at V1, we may have very simple proteoform detection. For customers that want to do more there will be improvements in consumable kits in the future. But as well, we have a number of customers who have very specific proteoform questions who come to us and say, "Hey, help us to go do that." Genentech did that, for example. Around the different forms of the phosphorylation patterns and the Tau protein. And we're happy to work with customers and customer engagements on those types of things. This question of proteoforms and isoforms is really a question of different biological insight, right? We think that over the course of the 5 to 10 years after we launched this platform, we will move down a continuum of enhancements that will give us more biological insight. First, understanding in deeper depth what the protein modifications and splice forms are. The next thing after that is understanding what cell did that, will come from after that spatial is definitely something our customers are interested in. And so for me, I think that there is a road map for us that gives us the ability to make enhancements over time that gives more and more biological information to the customer.

Got it. Do you think it's fair to expect some sense of lockdown specs and form launch time lines by year-end?

What I will tell you is, I'm not going to commit to a time line on it, but I think that if you back up from launch, right, and you said, okay, wherever our launch is, you back up 3, 4 months, we'll probably have some physical betas out with customers so that we make sure that we're testing the physical hardware with our customers. You back up another 3 months from that, and we'll start some early access samples with customers. That's a model where the customer ships sample to us. We run them on our equipment in our facility. We give them the data. We use that not just for papers and publications marketing value, but as well to drive the pipeline and the preorder pipeline for the instrument. And then if you back up from that, there will be a point where we talk about real data coming off of our platform. We will likely have a paper poster that shows a significant progress. And, that will be about the point that we will then lockdown final pricing and talk about it with Wall Street.

Got it. Makes sense. And then in terms of just the areas where you see the most challenges today. Outside of the multi-affinity probes, is really that -- the main focus at this point with everything else more a function of time and you're being sort of relatively confident that it gets there for that mid-'24 launch?

Yes. I mean I think when you say like what's the big fair challenge? I think even the multi-affinity probes, we have antibody development experts who run that organization. It is one of the largest organizations in our company. We have built an at-scale antibody development shop that probably rivals most companies are not doing antibody discovery. And there's a lot of iteration that goes on there, but there's nothing there where I say, "Oh, that's a huge challenge at this point." We've solved the challenges and being able to build affinity reagents that targets very short amino acid sequences. We've figured out how to have -- those probes be broad enough to get coverage of a molecule so that we can get enough information out of them. And so that's time and innovation, but that's not really a challenge. I think the biggest challenge is just bringing all of these components together and having them operate together in a reliable manner every time when the customer pushes go. And then just kind of getting it out and push it to the launch. So for me, it's time not necessarily challenges.

Got it. And then in terms of, you touched upon customer conversations a little bit earlier in terms of informing which proteins you would go after on V1. Can you share some color on that? What do customers really want? And could you perhaps like narrow the set of proteins to [ ones come in ] blood or something like that in terms of how you choose them?

Yes. So much like the mass spec, we don't get a choice of what proteins are going to come out when we say, "Hey, we do 50%, 60%." And for the customer, that's totally okay. Like when we talk to the customer, our customers are already using multiple techniques to analyze samples and particularly the customers in biopharma, these are customers where -- this information is absolutely critical to their business in building drugs. And so getting the widest possible -- widest possible net to get biological insight using the mass spec, use Olink, go and use another assay. That's a normal part of their process. And so if we're able to give them information, new molecules that they haven't seen with one of the other techniques for them that's really the important thing. And the key thing for customers that are using mass spec is that the mass spec always picks the proteins that are most abundant, which are almost always the least interesting proteins in variable, and our 50% to 60% will be a will be a distribution across the dynamic spectrum that will have some abundant proteins, but it will have just as many of those rare proteins as well. And so for the customers, they're really excited to see that, and the customers are really excited to see data at a single molecule level. With those 2 things, there's a huge amount of interest even at half of proteome.

Got it. Makes sense. Now obviously, you're planning to do early access before the full launch. What are your key aims during this phase? And will customers have an instrument in their lab at this point in time? Or will it be, you guys processing their samples?

Yes. So the predominant model for early access will be customer sends us samples, we analyze them and send them back the data. The goals are to work with customers early that are going to publish, and that are going to want to talk about what they're doing, which helps us build the base of papers and publications that enable us to get the next set of customers. The second goal for us is to start building interest in buying instruments and hopefully finding preorders for instruments. So those are the goals of early access. In early access, we've talked about it as kind of activities leading up to launch, but they don't stop after launch. We will continue to run samples as a service for proof of concepts with customers for burst capacity. And for times when a customer may have submitted a grant proposal, but they need 6 months, but they want to run some samples. But we'll still have that capability for our customer. Right before the launch, we will, as we talked about earlier, ship a few physical betas as well, just to make sure that everything we're seeing in our facilities is the case with our customers as well. You and I have talked about before, we are spread across 4 facilities on the West Coast, San Diego, San Carlos and the Bay Area and Seattle. And so we've already got a bit of geographic distribution and have that kind of -- we already know how our instruments operate in different places. So that's helping us to some degree, making sure that the instrument is not, doesn't have any particular issues with where the customers are falling. But we will put a few out there.

I see, got it. And then in terms of who the key target customers would be, any color you can share there, Sujal? And how do you see the mix sort of trending between core labs, academic labs and pharma as we get closer to the launch?

Yes. So the way I would expect that market is, I wouldn't look at it as like kind of let's look at it academically first, right? Look we take the top level of the Proteomics market. About half of it's biopharma, about 30% academic and nonprofit research, 20% is supplied market, so environmental, agriculture and so forth. For us, I think that the initial business is going to split 50-50 between biopharma, which is drug discovery and diagnostics, and 50% in academic and nonprofit research. That is pretty typical for what we're seeing for other new entrants in the [ essential ] space as well. Now you mentioned some other types of customers like CROs. I view those as partners and on-ramps for the customers that are the ultimate end customer in those 2 segments, biopharma and academic research. So customer doesn't want to buy an instrument, they use a CRO for some of those services. CRO may have our instrument. But still the ultimate customer is in one of those 2 areas.

Got it. Fair enough. In the past, you've talked about -- this is sort of the potential for tunability in the platform where the customer could do higher plex, I guess, in exchange for a narrow or dynamic range. How are you thinking about balancing price per sample when considering dynamic range in multiplexing?

Yes. So the platform itself in the mode of operation where we think mostly the customer will operate it is where they've got about 1 billion of our spots available for molecules. So that means you could run 12 samples with about 1 billion molecules per spot, you could run 8 samples, you could run 4 samples. There will be some specialized, and those customers will pay a few thousand dollars a sample. There will be some specialized high dynamic range applications where the customer wants to use all of our lanes for one sample to get a very high fidelity type of data set often. That will be quite costly. It will be something along the lines of tens of thousands of dollars. And we've talked to customers who have specific applications to do that. But my guess is that 98% of our volume is going to be on that. Few thousand dollars per sample running as many as 12 samples per day.

Got it. And you probably don't want to get into too many specifics. But in the past, you've talked about $1 million sort of price tag for the instrument, software and services against a sort of challenging macro backdrop. Have you had any sort of second thoughts about it? And with the price of mass specs going higher as well, could that give you some flex in terms of perhaps pricing it even higher?

That's funny. You say that I harass my team about prices get higher all the time. Now the mass Spec has pushed their high end up to $2 million. I think that we're still feeling pretty good around $1 million. But around $1 million for the deal size could be somewhere [ 1, 1, 1, 2 ] somewhere in that range, but I think we feel good about the price point of that range based on conversations with customers. We had a recent formal contract study that we've done. Not recent enough to pick up that $2 million actual price. But I still feel good about the price point around there.

Got it.

And in terms of macro, the macro has small effects up and down 10%, which, of course, if you're a 6%, 7% grower, it has a big impact. We're a company that is going into an established market. We have no revenue to start, those little variations don't affect us.

Fair enough. How are you thinking about pricing consumables? And so what's the thought process behind that?

Yes. This is an area where we have done an extensive amount of work with customers. And we haven't done the exact pricing and we haven't disclosed the exact pricing, but it's a few thousand dollars of sample. So call it, $2,000, $3,000 of sample. And that's all in. Many of these other solutions require you to have a sequencer and do some sequencing and have sequencing reagents, depletion column. For us, it's the entire account is $2,000 to $3,000 somewhere in that range, and it's sample to answer. There's nothing else needed.

Got it. Next sort of 2 or 3 quarters? Are there any sort of key publications that we should be watching out for?

Yes. So I mean I think I told you this before. I tell it to most investors who ask the same question. The key thing that you should look for is that at some point here in the not too distant future, hopefully, we will have a poster or a publication that shows real protein identification out of lysine. It doesn't matter if it's 100, 500, 1,000, 2,000, once we are there, all of the components have come together, and it's just more affinity reagents and more cycles. Once we're able to publish that type of data, I think it will be readily apparent to the scientific community that they approach works and that it's just a few more iterations to get it fully to where we aim to get it.

Got it. Makes sense. And then just quickly on the financial side of things. You've been pretty prudent with your spending. You've extended your cash run rate into in '26, can you just remind us of your cash position and also examples of how you're managing costs and improving efficiency without sort of not compromising on the launch process and efforts there?

Yes. I'll take them in last minute to give ourselves a pat on the back, right? I mean this is, I wouldn't say we are mostly prudent with our cash. I would say that we've done an exceptional job on the cash front, right? I come from the tech world, my lab company was -- that I was founder and CEO of went public in '06. We were profitable by 2010, 20% positive GAAP operating margin before we sold. So this is like in our DNA. We raised $450 million roughly since inception. We still have $287 million of that on the balance sheet. We only burned $48 million last year. In every corner of our company, we spend time making sure that, that spend is incredibly effective, making sure that it's targeted, making sure that we're scaling our processes and our people appropriately. Like for Anna, my CFO, who sits to my left, it is a day in and day out job and I think we do an exceptional job of it. I think that, that gives us, we think, the opportunity to have the capital that we need to finish product development to launch build a commercial team and get started with our scaling.

Got it. And, as you think about the upcoming launch, can you walk us through the freezing of that commercial channel build-out?

Yes. So in terms of the build-out, this is a $1 million instrument. It's got a significant pull-through associated with it. And so with that, it doesn't need to be a huge direct salesforce, right? I often get this question from investors like, what is that commercial build look like? How many sales reps do you need, even hire 20 on day 1. And what I answer is that the productivity of salesforce is based on what's the average deal size because transaction is a transaction takes a fixed amount of work, how big is the value proposition relative to the competition? And how easy it is to identify and reach those customers, right? We know who the 1,000 top customers are. We know how to reach them. We have a very wide value proposition, and we have a high ASP at roughly $1 million just to have initial deal, that's the recipe for being able to build an effective salesforce. And so you're not going to see us hire a whole lot of people to start. We're going to start small, start nimble. And start with direct salespeople and their support personnel, get those initial sales rolling and then learn before we go and continue to scale that.

And do the specs of the V1 once you lock them down, also play a role in the scale up?

Yes, to some degree. But what I would tell you is that the entire range of possibilities that we're contemplating still have a transformational value proposition relative to the competition. At least today, I'm not contemplating shipping anything that isn't well above their capabilities of anything else out there.

Got it. Well, that's a great place to leave it at. Sujal. So thank you so much for your time this morning.

Thank you, Tejas.

Yes, of course.