Pacific Biosciences of California, Inc. (PACB) Earnings Call Transcript & Summary

Welcome to the last fireside here on day 2 of the 2023 UBS conference. I'm John Sourbeer, one of the UBS Life Science's analyst. Joining me on stage is Liza Garcia, my co-coverage. Last but not least, we're really happy to be hosting PacBio here, fireside chat. Joining today is COO, Mark Van Oene; and Senior Director of Investor Relations, Todd Friedman. Thanks both for joining us here.

Great. Thanks for having us John and thanks for hosting the stuck around.

Maybe we started Revio, lots of talks in demand here in the market, but the long replatform has launched us here. So year-to-date placements have been strong, beat consensus last 2 quarters. Can you just talk about where you're seeing customer demand for these placements? What are new to PacBio, new to long-read type of customers start there?

Yes. Revio has been off to a great start. One of the better introductions of a product into the market that I've been involved with. And so we talked earlier on last year about how we prebooked those 76 units. And that was really important just to give us the signal that there was demand for HiFi and for long reads at scale. And that was our issue with Sequel. It just couldn't sequence enough samples. And so the market responded really quickly to us, and we got those bookings, and it's continued. We're -- from November through now, we're ahead of our original plan on the bookings. And so what that did was it gave us the visibility into how we were going to grow and ship and just scale the manufacturing. And so in terms of that mix, most of those early shipments, easy sale, go and convert your current customers. And so the majority of those early shipments were all -- and sales were all to the existing PacBio customers. But since then, we're really encouraged now with this pipeline shift of these new customers to PacBio. And so as we work through this year, we're up to over 40% of our orders now coming from brand-new customers to us, which was the thesis, but that's ahead of our plan. We talked before about trying to target at least 30% new to PacBio, and we're ahead of that already, and we see that pipeline building. And that's what we need to do. We only have a few hundred customers with Sequel II and prior PacBio platforms. And we have to grow the installed base, and we're seeing that demand and that interest now with Revio.

One of the top investor questions I've received about is a little bit on Christian's comments on order intake in the quarter being less than orders shipped. How should we think about placements over the long term, the guidance this year in the second half had a sequential increase for quarter-to-quarter in the second half. But is this increase sustainable? And how should we just think about placements?

Yes. Think about placements. Everyone asked us today. One of the first 2 questions was around the orders and the backlog. We think of things annually, and so we're well on track for annually. And I would say that the orders were -- they came earlier. And so the cadence of those orders was different than we expect, but we're well on track to beat our annual plan for it on our orders. And we pay our sales team on annual orders, not quarterly. So the timing of that is a little bit different. But the shipments is where you should focus on this. And so we grew our shipments from Q1 to Q2, low 30 shipments in the first quarter, 45 shipments in the second quarter. That will grow sequentially in the next 2 quarters. We've built the manufacturing and scaled the manufacturing, which is one of the only risks to grow was to get to be able to manufacture 40, 50, 60 units a quarter, and so we scaled the manufacturing to support the demand now, which is encouraging. And so look for that quarterly growth this year. And then in 2024, we'll ship more revenues than we did in 2023. And so we feel really good about the revenue business and the fundamentals of that. So look at the shipments. And yes, we'll make sure that we figure out how to talk about backlog and bookings, but focus on those shipments there. I mean focus on those new customers. We have to maintain a high rate of new customer acquisition to continue to grow and diversify the use of long reads. The other thing is to look at is just consumable use and adoption. And we had a great first start. We did $6 million in consumables through Revio last quarter. Great first start. And as that starts to balance out this mix between existing customers that can crank away right away and the new customers that we have to onboard a little bit more slowly and hopefully, in a few quarters, we'll be able to narrow down that range of utilization and expectations around pull-through.

And on those new to PacBio by and those new to long-read customers, any just additional granularity on what customer type, what are the use cases you're seeing there?

Yes. I mean it's extremely diverse. We've sold now to over 100 different customers Revio system. And so a big increase there again in customers. But it's across all the different academic and government and centralized labs and pharma. So we're seeing it across all of our different market segments, plant and animal. So it's very similar to what we would expect in the diversity of long reads, but I will say the majority of them are coming from human genetics. And so that's where the real growth driver is. Historically, long reads have been used in plant and animal. Complex, difficult to sequence genomes and resolve where they really just had to spend the extra money for long reads. But now the human genetics world is catching on to -- at this price point and at the scale, the utility of long reads for human genetics. And so that's where we're seeing our biggest growth right now and the biggest opportunity for Revio going forward. Todd, anything you want to add to that?

I guess within human genetics, any additional color there, where you're seeing demand breakouts like population sequencing, rare and undiagnosed disease, RNA-seq, can you give some additional color?

Yes. Yes and yes. So population -- so that's a very different market, right? It's kind of a made-up word, let's not kid ourselves, but it's really just trying to baseline understand genetics and usually with some health care spin, but it's really a research study. And so those require a different amount of information. And so traditionally, those have been a long-read market. But all of us last year with some of their surplus budget and they were really sample limited. Their recruitment was going slower and they had to spend their money. And so rather than just doing more short-read genomes or more short-read applications, they started to weave in some long read of some single cell and some RNA work. And so the All of Us program has been a great new way for us to learn about the impact of long reads in PopGen. And then since then, there's 1,000 of those genomes available publicly now, and they'll do a 10,000 sample study by the end of this year. So that continues to grow for us. But since then, MRBU (sic) [ MBRU ] in Dubai has adopted HiFi for their PopGen study and Sampled here in the U.S. is taking on some of the Veterans Affairs project. They've had this big MVP, Million Veterans Program, going on for years. And they're now going to start spending on long reads. So we're seeing now that the surplus of money from these programs is moving outside of just short reads and into long reads and RNA and single cell. So it's exciting, and that's going to be a big thesis. Now outside of PopGen, to me, it's all around how do you drive medical genetics. It's going to be research right now in understanding the disease, a lot of neurological, but this whole genome sequencing and where we're doing a lot of proof studies right now was with children's hospitals, rare undiagnosed disease. And the reason there is to show the value of the information you can get and the solve rates. What's the solve rate using long reads versus exomes or short reads. And a lot of these programs have always been done with clinical exomes, and there wasn't a big enough advantage to go and spend the difference between an exome and the short read genome. And now we're starting to see people like Bioscientia in Europe, say, look, we're going to go from clinical exomes to HiFi genomes and make that leap right to us. Here in the U.S. GeneDex probably has the majority of market share for clinical exomes, they're now scaling with us to go move all of that business into HiFi genomes. And it's the same thesis that Invitae had with us years ago. That's why we were developing a high-throughput long-read sequencers. They wanted to go all their clinical exomes to HiFi genomes. So we're starting to see more of these big service providers take this on at scale, and it's going to be a big part of our business.

And I think we're about 2 quarters in now with customers having a revenue on hand. Just any early feedback, learnings, areas of focus, updates coming.

Yes. I mean, no really high-tech instrument is going to go without some hiccups. We've had a few software bugs that we fixed. We've been really proud of this on the failure rates. We had an internal target that we weren't going to launch it unless we were under a certain hard failure rate. And so we knew that we were fine with. We're having run thousands of cells internally. So -- but you still want to make sure that plays out in the field. And so our hard failure rates are way below where we set our current goals. So that's been encouraging. There's little learnings along the way, but it's been quite robust and quite exciting for us. The one thing that's throwing people off is sort of the variability they get right now. On our old Sequel platform, we had a mechanism to prevent them from overloading a cell. And if you overload it, you'll lose it. And so people are underloading as a result. So they're not always getting their 90 gigs because they don't want to overload so they're underloading. And so we're going to build that feature back in and later this year release that so that people running 150 KB libraries or greater are always going to get that 90 gig.

And manufacturing capacity, you touched on it briefly, but just maybe dig in there a little additional color. Just any color around that, the ramp in the second half, what do you think you're going to be burning through that backlog, maybe in demand, how should we think about this?

Yes. I mean I feel good with how we set up manufacturing. We do a lot of the Revio manufacturing through contract manufacturers that have that burst capacity. The biggest wildcard is some of these PopGen programs. They have to go and spend their money in 18 months, we're going to have to be able to support that really quickly. And so we have a good sense of the forecast for the baseline business reviews. And then any of these big things, we don't forecast them. They'd be incremental. And that's where we would just use our contract manufacturer to scale it. But I'll let you talk about it.

Yes. And so I mean, in terms of, yes, backlog, we want to get to a place where ultimately exiting the year, we're able to deliver a Revio to a customer who places the order in the quarter. So at the start of the year, entering with 76 units in backlog and growing that in Q1 that we want to get to a point where customer places an order of that quarter we're able to deliver it, that they're not waiting 4 to 5 months to get their Revio. So that's how we think about building manufacturing and getting to that type of constant rate.

Yes, the caveat being, I don't want to make 100 systems this quarter just to bleed backlog and then have to wind this bell back down again. And so we're trying to work through that backlog surgically so that we keep customers all happy. But if you order 3, I might just give you 1 or 2 for now and another customer that 1/3 and hold your other backlog. So we're trying to spread them around to everybody and not overbuild the manufacturing capacity. But yes, I'm with Todd, the 2 to 4 week backlog would be great going forward if we can get there.

And I know you're not providing pull-through guidance on the revenue. But when you look at your internal expectations, how are things tracking there? Are customers finding the samples to sequence?

I mean, it's really early. So we -- Q2 was the first full quarter of having Revio in the field and it's only 32 systems to start the quarter. So it's definitely too soon to call a trend, but we had about $6 million in consumables for the quarter, which was a little bit ahead of our expectations. But in terms of utilization, what we've typically said is that the sequencer typically follows this cadence of 25% to 40% of max utilization. And what we saw in Q2 was Revio was right in that range, so above where we expected it. So customers not only buying SmartCells to get going on sequencing, but they're turning them through their Revio. We saw some standing orders in the second quarter which is a good sign. That means the customer is placing an order for the next year for a certain amount of SmartCells to be delivered at a certain cadence. A good sign is this that customers expect samples to continue to come in and is ready to commit to take those shipments over the next year. So it's too early to kind of call a pull-through number, but like we've been saying 25% utilization to 40% utilization is typically where a sequencer falls. And right now, that's what we're seeing. But I imagine sometime next year, we'll have a better sense dollar-wise what Revio has in terms of pull-through.

And those systems are connected for our tech support team to remotely service and support and troubleshoot. But as a piece of that, we can monitor runs. And so we'll have enough data to say, are they running on average 2 or 3 or 4 of the cells at once? How regular are they running? And so we can start to say for that information, and we know exactly what's sitting on their shelf versus what's been run based on our shipment data. So we'll have some pretty good data to triangulate once that installed base is a little bit bigger.

And when you think about that -- speak of the current users in the Sequel IIe got into the Revio, how do you envision that upgrade cycle going? Is it 3 to 2 today? Like [ 2 to 1 ], like 1 to 1, how does that track?

Yes. So it's a multiyear cycle, right? So you have to always think of it in that context. And so over the multiple years, to me, it's 1:1 or sometimes it's even greater than 1. But in the initial start of that, if you've got 5 Sequel IIes, you might only buy 2 or 3 Revios and then as you fill up the capacity start to expand. And so most people now are a little bit more cautious about expanding as they need capacity versus just pulling it all in right away. But over time, we do expect that to be a 1-to-1 transition, but again, over multiple years.

And I guess, maybe just broader long-read sequencing market. So seeing strong demand, seeing new long-read customers come in, I guess, where do you think the market plays out over the long run? I say if it goes from 10% of samples today, could it go to 20% to 30%? How should we think about the trajectory just on long reads?

Well, let's hope so. Now it's -- long reads have been desirable, but they were just too expensive and the throughput wasn't there. And what we're seeing now with Revio is the throughputs there to do meaningful science and customers like the Broad are talking about how they can convert 30%, 40% of their work to long reads to take on these programs. And it's because you have the information they're going to gather and how many more Caucasian short-read genomes you need to sequence to look at a disease, right? And so I think short reads will move into diverse populations and start to understand those. And a lot of the studies that have been done on arrays and on short reads in Western Europe and North America will start to shift to long reads to look at the other parts of the genome that they have and to include the epigenetics. And so that epigenetic component is going to be a big driver of why we think 10%, 20%, 30% of some of these markets will shift from short to long reads for big genomic studies. And it's also just different applications. I'm still really big on the single cell long read play and looking at the whole transcriptome, just like we want to look at the whole genome, we want to look at the whole transcriptome and all those isoforms and what those are contributing to disease and how those then correlate to the proteome. And right now, the RNA protein correlations aren't great because people aren't looking at the hundreds of thousands of isoforms that exist. They're looking at the 20,000 genes. And so that whole transcriptome field is going to shift for us in a meaningful way as well.

And we're seeing customers are benefiting from a price decline in short-read sequencing. In the past, they might have just put those savings back in to do more samples. What we're seeing more of these days is spending less on short-read sequencing and that money that I have now, I'm going to use for a different omics. I'm going to look in -- deploy that into long reads. I'm going to deploy that into maybe single cell or spatial. So you see long reads and other omics being kind of beneficiaries from some of these savings that customers are realizing.

Great. Well, maybe let's do that shift in from long read here to short read to talk about some of the other products in your portfolio instruments there. So at 2Q earnings, you announced the Apton Biosciences acquisition and the move to a high-throughput sequence. Maybe just can you talk a little bit of how this could develop and accelerate that SBB chemistry to high throughput? And just any color on valuation there too as well?

Sure. So first off, we started shipping the Onso sequencer just in the last couple of weeks here. So that's our mid-throughput short-read sequencer. And part of that Omniome acquisition we did a little over 2 years ago now -- around 2 years ago now. When we acquired Omniome, we had made the decision it was important to go after the oncology market, and you weren't going to beat Illumina on price ever. But if we could find something more accurate, we could differentiate ourselves and hopefully use that sensitivity to win a foothold in oncology. And we looked at all the different players. And so when we bought Omniome, we would also have been in discussions with Apton because they had a high throughput sequencer, but it was just an SBS sequencer, and we just didn't think that was going to be competitive. But we were intrigued by the idea of in the future, if we can really get 90% Q40 data in the market for short reads, what Apton can do to help accelerate that into a higher throughput short read platform. So we stayed really close with John and his team as they made a little bit of progress on their system. And as we got near the end of the Onso program, we had that confidence to go and make another investment into the short-read market because we've demonstrated now in the multiple collaborations have shown that this accuracy is going to matter for oncology. It's not just that you can detect things with less reads than you can. So there's an economic advantage to our chemistry over SBS. But there's also things that we see that you can just never see with SBS. And so it's going to be an economic thing plus oh God, like we're going to miss it. And so in a lot of these applications, mid throughput is going to get people an entry point, but they're going to want higher throughput sequencers. And so we went back and we actually had Apton do a little collaboration with us. So we sent some of our team over and used their clustering, their flow cells, their optics and their whole sequencing system, but our SBB chemistry. And within 6 weeks, we were able to sequence a human genome at pretty high quality and derisk the idea of how this can accelerate the program. And so -- to me now, they've got really high resolution imaging and optics that are hardened that they've been using for years. They've got multiple alpha systems that they're running day to day. And so rather than us having to go and build out that optical train and an optical system and all -- and then reinvent their clustering, they've got great in solution clustering as well. We can take that and to start integrating with SBB. And so that's the acceleration and the derisking that we're talking doing. And -- it's always been a part of our 2026 model. And so when you look back at our Investor Day, we had some high-throughput short-read revenue in that model. Thus, if anything, it derisks that and it may accelerate some of that revenue.

And, I guess, maybe just run through or any additional color on just value addition, maybe the milestone payments, maybe how this is different than Omniome?

Yes, it's hard. It's really hard to value in time. Todd spins the models for me. It's really hard to value a stand-alone sequencing company right now. That's sort of that far from having a product on market. And so we've booked it as many comparables as we could, what we're emerging short read sequencers that are either public or private or how they've done with their fundraising on their last rounds. We looked at some other comparables in the space, as you can imagine. We knew that they'd raised just under $60 million. So we knew kind of what their investors would take for a flat route. And the reason why we got to this $85 million was ASL stock, probably not worth $85 today given their performance. But it's all stock. And we thought that it was an appropriate valuation given what our synergy models were telling us. Now the milestone is very different. So our Omnia milestone should get triggered later this month, right? So once we start shipping our reagents to TGen this month, we'll trigger that Omniome milestone. But it was on first customer ship. But with Apton, we said, look, because of this concern over what's the right valuation, the milestone $25 million can be stock or cash, depending on what we want to do at the time. It's based on achieving $50 million of revenue with their product. And so if we get to $50 million of revenue with that high throughput product, we'll have no problem paying that milestone, more than support it. And so we structured it very differently and intentionally because of some of the questions we were having around what was the right value.

Yes. And really the value to around accelerating the development. So if you could get this high throughput product out quicker in the middle of -- you have a lot of customers going through a high throughput transition right now to a new product. So if you're able to launch a product in the middle of the cycle versus the tail end of the cycle, it gives you a much better -- a bigger opportunity to get into more customers and getting into those customers more quickly, we'll accelerate that consumable revenue faster into high throughput and more high-throughput consumable revenue leads to better gross margins. So ultimately, that acceleration is how we got comfortable to with the valuation. Also looking at OpEx. This -- we're committed to 5% OpEx growth through 2026. And this was a pretty lean team that developed this product. And so we're bringing in forward about 14 full-time employees. And these are employees and move over roads so we would have hired either later this year or early next year. So it doesn't really add any incremental OpEx. In fact, our guidance, the other week, we're calling for lower OpEx growth than originally anticipated in 2023. So looking at kind of the valuation in the model, how quickly can we accelerate this program and really derisk it by getting this already in place office systems and doing it without spending that much incremental OpEx.

Yes, it's probably less OpEx for me next year if they don't have to go and build all those sequences to go and start integrating on. I guess, I think I counted Omniome to Onso was 22 months. Is that the proxy you should think about here? Or any time lines there?

Well, I mean, high throughput sequencers are a lot more difficult to make robustly than the throughput. It's -- that's why Illumina hasn't had a lot of competition on the high end, but there's a lot of emerging throughput players. So it's probably a little longer than what it was with Omniome, but it's not far from that. It's in that time frame. And again, it's within our guidance around our 2026 revenue model.

And last one here on this, I guess, just any thoughts on pricing in the market, where you compete with some of the existing high-throughput launches on there today when you enhance maybe SMB, increased accuracy, how that should fit in there?

Yes, that's -- we haven't talked about when we've obviously modeled some different price scenarios, but we haven't decided where we're going to price it. A key factor for that is going to be the performance of Onso over the next couple of years because as multiple collaborators have shown us now, if they only use 1/3 of the sequencing, price per gig is not the way to think about it, right? It really is -- what's the price is going to cost them to get that answer. And so if it's 1/3 of the sequencing, our $3 per gigabase becomes a $1 per gigabase comparable. And so we feel -- we're going to learn a lot of these next 2 years about the performance and sensitivity of Onso that will help us inform that pricing. But for sure, we think we can price this competitively, not with where the market is today, but where we think it will be when we launch it. And so we know enough about the space to think about what we have to compete with in that time frame. And so we're very comfortable. Apton built everything around low cost. If you look at their old investor day, can we do a $10 genome. So all the decisions they made was off-the-shelf components, really, really simple flow cell design relatively inexpensive components and lasers, the reagent volumes are pretty small. So everything they built was how can they make it so they can do a $10 genome and makes the margin. And so they made some really good trade-offs. So they are going to let us go and build a product that ultimately lets us price it wherever we need to. And then the extra value will likely come from the accuracy of the SBB.

And shifting to Onso, I think it's been about 2 weeks since you announced the first shipment out the door. Any just additional updates there, color customer received anything up and running yet?

Yes. No. So we shipped it on the day of our earnings. We installed at the following Monday. By the end of that week, it was up and being validated. It's passed all this validation runs now. So all of the validation runs exceeded our 400 million clusters per lane and greater than 90% in Q4. So it's hitting all of our metrics already. This week, we're training the customer. So TGen is excited. So they're running our libraries and our samples this week. And then next week, we'll likely ship them a bunch of consumables and trigger the milestone and turn it over for them to run their own samples. So it's going exceedingly well. We expect that typically 3 to 5 days get it in there and install and validate it and start training. So going -- and we're likely to ship a few more this quarter. We're sort of limiting the release here. We just want to make sure this is a really good customer experience. But overall, really, really happy so far.

And, I guess, just any additional comment or color on the backlog there, customer type, new to PacBio customers, domestic versus international mix need, can you provide us on that?

Yes, it's going well. So we talk a lot about our ability to enter short reads globally because of the scale of the commercial team and the distributor network we have. And so we've gotten more than half our orders from Asia. And so there's just like early Revio, they want that new technology. So more than half of orders are from Asia. We've already got orders in this system from over 8 different countries. And so there's -- you're starting to see this is not going to be just a U.S.-focused opportunity for us. We really are able to globalize this from day 1. And so I'm really excited by the diversity. Now a lot of these are in and around oncology and oncology markets, and we expect that. These are customers that want to explore the sensitivity of Onso that are really intrigued right now. But I'm encouraged. I don't know if there's other color you want to provide, but I'm really encouraged by the early order book.

Yes. No, across applications from liquid biopsy, gene editing research. So several different applications, service providers, academic labs. So pretty satisfied with the diversity of customers that have ordered so far.

And actually, TGen is an interesting one because they were collaborating with Omniome when we acquired them. And so they've had early insight into the value of the accuracy. They weren't a beta customer because they were just happy with us running their samples, but they would've always been sold on this. But when they bought Onso, they've never been a long-read customer of ours. And so there, they bought a bundle. And so now TGen became one of our early Onso orders, but also a new customer to PacBio with Revio because of that collaboration and that joint sales. So we're starting to see that ability for us to satisfy a couple of different market applications for them to really play out.

And I'm going to ask one more here, and I'll let jump in and ask a couple of questions. But you did touch a little bit about, I guess, from the Apton acquisition, but that's sequencing less, and I think you've mentioned at AGBT it was roughly 50% -- 57% reduction in all-in costs, given the accuracy in the sequencing. [indiscernible] now that you have shipped units, you've had these customer discussions, how is that resonating? Does that come up in the conversations? Any just additional feedback now since then?

Yes. I mean I don't know, 57% is the right number. And so I think we'll learn more about that. It's probably in that range. But there's a couple of different ways they're going to save money. What people are seeing is they don't need to do like molecular gymnastics to do error correction. And so the workflows and the library prep and the deconvolution of those is much simpler and much easier. And so there's going to be savings there. And then it really is just about less -- having to sequence less to get to your variants. And that's probably in the order of a 1/3 of the sequencing and may be even greater than that because as Chris showed, it's not just a Q40 sequencer. He's now publicly put out bunch of his sequencing runs at Q50 plus. And so that's why wait and see. If it stays at 90% Q40, it's probably a 3x reduction in sequencing. If it plays out the way I think demand in more of these applications turn into be a Q50 type quality, then it will reduce it even further.

Great. Great. I'll keep it quick. But I guess, rounding it out on products go to the original Sequel, so how should we think about kind of the pull-through back half of 2023 and maybe longer term and how to think about that instrument?

For Sequel?

Sequel -- Sequel II, sorry, Sequel II.

Yes. l [indiscernible] the details. But at a high level, there will always be some customers that will stay on Sequel, a small number because they just don't need 15x more data, right? And so there will be some Sequel that will persist over time. But if you think about what we've done, we've given the person, the ability to get a 25 M chip versus an 8 M chip. So 3x the data on a chip cheaper, right? And so if you've got access to both, why would you pay more for 1/3 of the data? And so I think that's why we're seeing this really quick overall transition from Sequel to the Revio, but Todd.

Yes. I mean in the first -- so second quarter, it was the first full quarter of Revio already, 44% of the consumables were related to Revio. So we expect that to -- Revio to take over the majority of consumables here in the second half of the year easily and the ramp down continue. We talked about the first few -- several orders we got were from existing customers. So most of our largest customers that were production, Sequel II customers have already gotten their first Revio and are going to be the ones that cut over first to Revio. So we'll likely see the Sequel II pull-through come down pretty quick and customers start to decommission those. But yes, so I expect the second half Revio to take the majority of consumables and that trend continue into 2024. But having some tail of Sequel II consumables for those customers that opt to stay on the platform.

Great such a quick launch. But I guess I know this is probably a little bit less applicable for you. If you could just remind us kind of biopharma customer exposure kind of thinking at a global level for the company? And then kind of what you're seeing from biopharma customers just given how topical that conversation has been?

Yes. So biopharma and pharma as a percent of revenue is in the single digits. So it's not a huge part of our revenue base. But that is one area where that was a customer base where we expected them to continue to use Sequel II, specifically around AAV applications. As Mark mentioned, some customers just don't need to pull through of Revio. And AAV gene vector sequencing is one of those applications. And so that is an area where we saw some biopharma customers push out. They were planning to sequence on Sequel II or even purchase a Sequel II this year, delay that purchase due to the biopharma funding atmosphere.

Okay. Great.

Actually, maybe just one -- maybe dig in here a couple of applications in the last 10 minutes, but I just liquid biopsy, just any thoughts on the demand applications there, short read, long read, how that's coming through on your different platforms?

Yes. So liquid biopsy is one of those specifically why we invested in the short-read technologies. If you think about what that test is trying to do, you're taking a blood sample and you're trying to look for DNA from a tumor or DNA from a baby, you're trying to look for something other than the normal DNA. But in your blood, that gets chopped up into a little short fragments. And so they're all about 200 base pairs plus or minus 20 basis pairs long. So it's a little short pieces of stuff floating around that you don't need the sequence that can sequence 20,000 basis for. And so specifically, that market, I think, will always stay on short reads. But because you're looking for very minor amounts of this tumor DNA in this sea of normal DNA, you need that sensitivity. And so liquid biopsy to me is our play into short reads with our sensitivity and our differentiation. And that's why I think you need both mid and high throughput to fully capture the different market segments within the biopsy. And the big one there that excites me is MRD. Like -- there's a lot of talk about early detection. The market that's nearer term is MRD, this residual monitoring of the disease. And it's multiple tests per year, you're staying with that patient. A lot of it is tumor informed and so is specific to that patient. And if we can use our technology to get into MRD, that's the beachhead that we're looking for. And we're directing our organization specifically at that. Let's not try to be a short-read sequencer for everybody. Let's play where we've got a strength of accuracy. The rest of the other markets, the germline genetics, the transcriptomics, all the different microbial things, those are much better applications for long reads. It's just people haven't had affordable accurate long reads to do it. And so they've been using short reads to attack a problem that's better suited for long reads.

And one piece that doesn't get talked about a lot, I think, is agricultural and applied markets. Just can you remind us what percentage of revenue that is there? And just how that's been trending where you've you seeing on customers?

Yes. So that's historical PacBio, that's been kind of the bread and butter for long reads because you're looking at complex genomes sometimes significantly larger than the human genome with several different -- multiple times of chromosome. So it's extremely complex genomes. So that's always been in the wheelhouse for long reads. But what's great is with the Revio launch, those same customers are benefiting from the increased throughput. So they want to scale their agricultural samples on Revio just as much as customers want to -- human genomics customers want to scale human genomes on Revio. So that's still -- as a percent of total revenue of Revio, it's second next to human genomics. So still a lot of demand from [ AGG ], from biodiversity projects like the Darwin Tree of Life program. They're using Revio to really kind of ramp up the amount of reference genomes. They want to build -- so it's still a focus for us and still -- I think it's a strong market.

Yes, it's interesting. I had a -- I forget about the dramatic increase in throughput with Revio. We had a post stock presented at our company meeting from the East Coast, and she was so excited because what took her a year on Sequel she just did in 3.5 weeks. And then just sort of put it in perspective for me, it's like, oh, you're right, we did drop a lot more capacity on the market. And plant and animals is one of those where the whole movement to a pangenome reference is really important. And so it's not just one and done anymore. If they're going to do a big study into some fungus or some pine tree, they're going to want to create this big pangenome reference. And so long reads have always been used for that, but it's just taken years to get it done. And so we're seeing a lot of interest in researchers, not even just the Corteva's of the world, but researches go and start to create these pangenome references with us as well.

And I think at the Investor Day, you talked about developing the high throughput short read. You've seen the Apton acquisition. Also, there was discussions on the desktop in long reads. Is that still in the plan? Just any thoughts on how that could play out?

No, it's not just in the plan. I've moved people off Revio onto that and pushing hard. The whole idea for the desktop version is to leverage the technologies that are in Revio, but lower throughput, lower CapEx, more decentralized, probably would become our first clinical box if we ever take something through the CE-IVD route or the U.S. FDA route. For sure, we're going to be taking it through the China NMPA with Berry Genomics, and that was the collaboration with Berry was around using a benchtop device to go and just diversify the testing for beta-thalassemia. And so that's well underway. It's not going to be -- it's not one of these really long multiyear programs because we're leveraging 25 M chip and core technologies you already have. So yes, we're pushing hard. And that's an important part. You get Revio here for a year, 1.5 years and then prove even more value of HiFi and then let people go and get into that market with a little bit less cap expense.

Maybe just hitting here in the last couple of minutes, a couple of high-level long-read questions. So you've seen the short-read market. You've seen talks of $100 genome, [ sub-$100 ] genome. Revio has definitely brought down that price of the long-read market. But where do you see the dynamics playing out in here from a pricing standpoint?

Yes. So it depends on the application where that matters, right? So we always use the price per genome as -- it's just -- to me, it's just a measuring stick of the pricing of the technology. It doesn't mean that short reads in my mind 5 years from now are used for genome sequencing. I'm not going to spend any time building a whole genome workflow for Onso or for short reads. That's where I would invest in making sure that people can use long read data and do genomes with long reads that include the epigenome. And I think that's going to become a more and more important part of genome sequencing. So I always think that the long reads will have a premium for whole genome sequencing, a premium price because of the differentiated value over short reads. So where we have to pay attention to that $100 genome is that $1 per gigabase or that $2 per gigabase that Illumina and BGI and [ Element ] whoever else is going to drive towards, that's important for us to think about in the context of liquid biopsy or some of the different microbial or infectious disease. And so it's a benchmark for where we have to think about the ability to price short leads, but not necessarily whole genomes. And so I have to think -- you have to differentiate it that way. But I think there's going to be a premium value for long read genomes over short reads forever. And I think there's going to be a premium value that we can get in the short-read market with a higher accuracy because of the extra sensitivity.

Maybe just [ we have time ] touch on just academic demand. Any updates we're seeing there? We're seeing the NIH budgets, but how is that playing through to your customers?

Yes. I mean -- so just like we've been a little bit protected in China because of the new product launches, we're also protected a little bit with the academic budgets, especially NIH because even if that budget is flat or slightly down, there's the shift within the spend. And so what we're trying to capture now is the spend shift within that NIH budget, not looking for them to increase budget so that we can grow. And it is the shift that we're seeing. And as Todd alluded to, people don't have to spend as much on their short reads, so they're adding extra omics into their studies. And so that's a shift of the spend of NIH. Anything we're doing for All of Us is the shift of the All of Us dollars to us. It's not new money that All of Us was granted. And so it's the shift within that, that we're paying attention to more so than just the overall magnitude of the budget.

So, I guess, last question here on to bring a high level philosophical. So maybe where do you see we are, what inning today in long read? And then if you get the desktop, you start to see clinical penetration like where you say we would be in 5 years from now?

No. God, whatever 1% market share, so we're probably not even into an inning yet. We're still warming up. It's a huge opportunity for us. You talk about these rare disease cases and children's hospital networks. They want to help these families with this diagnostic odyssey. And we think we can help them and we can help them faster and for cheaper than what they're doing and actually solve more of these cases. So I think there's a real opportunity for us in that clinical aspect. I looked at the GeneDxs and the [indiscernible] and some of these larger providers to really accelerate our win in shifting clinical exomes to clinical HiFi genomes. So I'm going to be leaning hard on our partners to help us drive that marketplace. But there's -- the payers are getting more accustomed to paying for genetics and genomics in a lot of these cases. And so I feel like a lot of work has been done over the years to get the payers heads wrapped around the way we're doing this testing, and we're going to be prime for growth, but it's so early. And that's why it was important for us to get Revio when we did because these markets are earlier in their potential. And I think we're poised with the right technology to take them on.

Right, well, I think with that, we're at the top here. Thanks for the time. Mark and Todd, thank you very much for joining us today, and thank you for those for listening in.

Thanks, John. Thanks Liza.

Thank you.