I'm Derik De Bruin, the Senior Life Sciences and Diagnostics Tools Analyst here at Bank of America, and welcome to our 2023 healthcare conference, coming to you live from the Encore Hotel here in fabulous Las Vegas. Our next, our next presentation, our next discussion is with Pacific Biosciences, with Christian Henry, Chief Executive Officer. Christian, welcome.
Thank you.
Thanks for being here.
Appreciate the opportunity.
You just reported Q1 results last week. Did you wanna do some remarks, or should we just jump right into the Q&A?
Well, I think we can jump right in. You know, we did have a great quarter, and Revio, our new flagship sequencer, is off to a fantastic start.
Yeah.
Probably most exciting is how well it's performing in the field. Yeah, we can jump right in.
All right. I mean, since sequencing is very much on the top of everybody's minds these days.
Mm-hmm.
I want to talk some big picture questions before we sort of go into the nitty-gritty on your platforms. Look, you've been around this field for a long time. For the first time in many years, customers have a lot of new platform choices, including both on the long and short read side, options from PacBio. We've just also seen the price of whole genome sequencing drop from $600 to $200. You know, historically, when we've seen these price drops in sequencing, customers have just run more samples. The question I'm wondering is like, are we potentially facing a capacity overload? I mean, is there enough samples to sort of fill these machines right now?
Yeah. I think if you look out historically, there has always been enough samples to fill the capacity of the sequencers. I don't think today is any different. If you think about where we actually are in our journey to understand the human genome, let alone the rest of the genomes on the planet, you know, we only 5%, 6% of the genome is medically actionable today. There's still whole swaths of the genome that we just don't understand.
There is no shortage of desire, no shortage of funding, and you put those together, you know, I do think that although you may see in any particular 12 week period or 13 week period in a quarter, maybe there's machinations, but on balance, there's no question that we're moving to continued more sequencing, and there's really no end in sight.
Got it. I think historically that when we've seen price drops, you know, people just run more samples. I guess when you're talking to labs right now, and this sort of segues into talking about
Mm.
your products, are people just gonna run more of the same old samples, or are they gonna start looking at newer technologies like PacBio to basically decide to maybe get some more incremental information as opposed to just running the same old experiment again?
Yeah.
How do they spend that savings?
I do think it's very clear, particularly with the launch of Revio, that there's a huge hunger in the market to understand and develop and get a more comprehensive view of the genome. That's what long-read sequencing technologies do over the historical incumbent technologies, is that they give you, with single molecule resolution, they give you much more of the picture. They allow you to look at the dark regions of the genome that are unreachable by short-read technologies. They allow you to look at methylation status and epigenetic markers, which are being found to be much more important than we ever knew in regulating, you know, genes, turning them off and on. They look at structural variation.
you know, having a more comprehensive view of the genome gives, you know, gives researchers hope that they can uncover more of the mystery of the genome. What you're seeing is that you're seeing projects now start to move portions of their samples over to long-read sequencing. Over time, you know, perhaps in certain areas, the whole project gets done with long-read sequencing and, you know, hopefully on PacBio instruments. That's the paradigm that's happening right now in the market. Better, a better visibility and good economics is enabling a whole new wave of science that we're gonna see over the next several years.
You do think that that's ultimately just a cost question versus a,
Well, I think it's
science question?
I think it's actually, more a cost question. The science, it's pretty much undisputed that people wanna see more of the genome.
Yeah.
That's undisputed, I believe. It's really do you have enough throughput, and do you have the right economics? It's both cost and throughput to be able to do the scale of experiment required to develop that insight. That's one of the reasons why we developed the Revio system.
Yeah.
which is our new flagship sequencer. It gives you scale that's unprecedented. It's 15x more capable than anything that we've ever had before. It's at economics. In fact, I saw just a preprint that was just out earlier this week, where they show at 12x coverage using PacBio long reads, you get variant performance consistent with 30x short-read technology. What does that mean? At list price, genomes today, that you can buy today, using our technology are under $400 a genome.
Hmm.
You know, there's all this talk about the $200 genome. The reality is that the $200 genome's not here yet. Right? It's purported, and it will come, but the actual in the field, it's not there yet. On top of that, the only way you get to $200 is if you're a very high throughput lab
Right.
and you have the batching. Not that I'm doing competitive comparisons right now
Of course not.
It's all part of it.
But I mean, you know, basically, there was some arbitrary designation that 30x was the right definition. It's like, is that. Moving away right now is the accuracy because, I mean, look, there's always these questions of accuracy and all the blah, blah, all this type of stuff. That means there's a reason why 30x was sort of done. Is that antiquated?
I do think that that's becoming more antiquated. I was actually in the room when we decided, you know, 30x was the right coverage and in collaboration with lots of different people. Back in the day, David Bentley was one of the leaders of that at the time, and that was a happy medium. The reality of biology is that it's complex.
Yeah.
In some areas, you wanna do 100x or 60x. In other areas, you wanna do 2x or 1/2x , and it really becomes the experimental design. I do think the community understands that 30x has been kind of just this arbitrary benchmark that we try to match coverage, which equates to economics to a certain level of performance. Now that long-read technologies are really becoming much more mainstream and much more capable, that paradigm is changing.
Let's segue and talk about Revio. Can you talk about the initial demand you saw? I think you reached, I think you had, what, 76 orders in Q4. I think you shipped 32 in Q1. What's your backlog? How should we sort of think about, is demand greater than expected? Just sort of your initial thoughts on the launch.
Yeah. I think that, I mean, it was quite remarkable for us to start the launch with 76 orders in the fourth quarter. That gave us a huge jump-start on, you know, when we knew we were going to start shipping in March. It gave us a very strong confidence that we, first of all, we had the right product. Then it was just, could we execute? In Q1, we executed. We delivered more than I expected us to deliver. The field performance of the machines has been really quite strong. You know, every new instrument launch has its nuances and challenges. We really have seen that on balance, the hard failure rates, the data that people are getting, is all above our, generally above our specification and working quite well.
All of that's gone well, which has driven us to have strong order book and a strong forecast going into the rest of the year. We've decided not to give quarter-to-quarter orders. You know, I'm not gonna give any new information today on that, but what I can say is that our orders in Q1 were more than we shipped in Q1. We built net backlog during the quarter, which sets us up really well for Q2. In Q2, we will scale our manufacturing to its steady state, and matter of fact, we're pretty much almost there now. Over the course of the quarter, we're scaling up. What we will do is, as a result, we expect to ship more systems in Q2 than in Q1, as everyone would expect.
We will evaluate, whether or not we need to scale manufacturing even further, you know, down the road. The good news is that we have a combination of contract manufacturing and, a manufacturing floor, and so we're able to scale with reasonable time horizons, so it's not gonna take us that long.
Got it.
Hopefully that helps.
Yeah. In terms of the orders, how many of these are people that have already had some sort of experience with PacBio? They either had a Sequel or a RS, or how many of these are, like, new to PacBio customers?
Well, if you look at Q1, you know, 2/3 of the orders were from people with experience, and about a third was new to PacBio. In fact, our largest order in Q2 was a brand-new customer. They bought multiple Revio systems, and what they're going to do is they're actually gonna replace their whole exome sequencing with whole genome sequencing using Revio. This is in a clinical
Mm.
a clinical research setting. That's actually quite remarkable.
Yeah.
It shows you where we're going.
Yeah. Well, that's funny 'cause when further down in my question list is, like, what needs to happen for long-read whole human sequencing to make a significant adoption to the clinic? That was sort of like one of it. Clearly that is. What's unique about this customer?
Well, this customer is a very large, you know, clinical diagnostics customer. Their this particular order came out of, out of Europe, that's actually a little bit remarkable.
Yeah.
You would think normally that would start in the United States and then feather its way out through the globe. This is actually, some folks that, you know, are in Europe.
Yeah.
For competitive reasons, we're not gonna actually give the customer name today, but it's actually quite exciting.
Mm.
It actually hits the fundamental belief. If you remember back in 2021, we did a project with Invitae.
Yeah.
that, you know, didn't bear as much fruit for different reasons. That was the vision. The vision was how do you convert, you know, the human exome market in a clinical setting to genomes? As you do that, you get so much more value out of the genome that everyone does well, patients do better, the clinical diagnostic labs do better, and obviously, we are the engine that provides all of the technology for that. That's really super congruent with our mission, right? We're trying to.
Yeah.
trying to leverage the power, the promise of genomics to better human health.
How do you think about the potential for the installed base, right? I mean, are you expecting your legacy platforms, those buyer customers to all upgrade change.
Yes.
Right? I mean.
Yeah.
What do you think is sort of like the new incremental? Basically, it's like how do we sort of think about the potential for the TAM on the boxes?
Yeah. Well, you know, if you go back and start with the history.
Mm.
Basically PacBio's been able to double their install base with every generation of technology. If you just use that as a starting point framework, we have a little bit more than 500 Sequel IIs in the field, II and IIe in the field, so that would give you math of 1,000 instruments if we followed that paradigm. That's certainly not out of the realm of possibility. We see where the customers are, and we see the scale of the other sequencing platforms. One of the good things about being a fast follower or maybe slow follower, but a follower, there is that we know, we see who in the world's doing sequencing, and there's a lot of sequencing to be done.
I do think, you know, 1,000+ instruments is not an unreasonable thing.
Yeah
to be thinking of. You know.
Yeah, yeah, of course, of course.
We can debate about the time horizon, but that's what I think, could happen.
What do we think about the potential for consumable pull-throughs? I mean, everyone's always, you know, your box costs X, and it can pull through Y on this. It's like, how do we think about the consumable stream? Obviously, you know, you're launching a bunch of new machines right now, so people are not gonna be fully up to speed.
Right.
Right. There's gonna be some bleed down in terms of the consumables, just the way the math works on people ramping up.
That's right. Well, you know, on our earnings call, you know, we got that question about 18 different ways. What we've said very clearly is that we don't know the answer yet. But the capability of Revio is such that it can. At its maximum, and you think list price and maximum throughput, it's roughly 1,300 genomes a year or $1.3 million a genome, give or take. Or $1.3 million a system per year. Typically, you know, based on my experience, instruments run between, say, 30% and 35% utilization over the course when you look at the whole install base. The early install base is likely to be use their instruments more than the late install base because you're going to the end customer.
What actually happens is you end up with a distribution, a bimodal distribution.
Yeah
effectively of customers that use it maybe at 60%, 70%, 80%, and customers that use it at 10%, 15%.
Yeah.
You'll see that. I don't expect this to be any different. What that means for us is that on balance, you know, you probably are in the 30%-35%, 40% range.
Right.
You know, for us, that's massive because it has so much more capability. It changes the dynamic of our entire P&L.
Yeah.
Consumables are generally higher margin than the instruments, and therefore we get, as that product mix shifts, we get a lot more gross margin, which is a good thing.
What's the list price of the Revio?
The Revio is $7.99.
Yeah
$7.99.
Okay. Well, Okay, that makes sense. I mean, we've always used like 1/3 to 1/2 of the cost of the box.
Yeah
as the reagent pull-through. That's how we sort of have always modeled
That's right.
when we do things like this, depending on where you are in the curve with it. I have to ask the obligatory question on synthetic long reads, you know, obviously, there's been a new product launch out of Illumina with the Infinity. Has that made any sort of impact? I mean, have you seen any sort of like people saying it's like, "No, no. This is, this is fine. We don't really need the bigger long read information." Seen any impact on the market?
I don't think. I mean, to be honest, we've seen zero impact. In fact, we've sold Revios to all of the beta customers already that are beta, that have been beta testing their synthetic long read technology. You know, the reality is that it has some serious challenges. It's still short-read sequencing.
Yeah.
It still can't access the dark regions of the genome very well. There's a lot of problems with synthetic long reads relative to, you know, Revio. I don't think it's slowed us down at all. If anything, it continues to amplify the desire of the market to have long read technology because of what it can bring to the table. No, I don't think it's slowed us down one bit.
Just staying on competition, I mean, obvious, you know, your competitor in the UK, Oxford Nanopore, has made, you know, significant progress on that platform.
Mm.
It feels like you're targeting a little bit different niches in the market right now with this. You know, what is sort of your thought on that technology and the competitive dynamic between the two labs? I'm just sort of just, you know.
Yeah.
How does that play out? 'Cause it's one of these ones where it's the Nanopore technology has always sounded good on paper, but it's always a little bit different in market experience.
That. Yeah. I think that still holds to be true, is that, what you see on Twitter isn't exactly what's happening in the field. You know, they have done well. They've built a nice business. They reach a different customer than we do.
Yeah
at a fundamental level. We are going after, you know, the higher end, high throughput customers who need the reliability, who need to know that it works every time, who need the level of automation that we provide, whereas they seem to be going after more customers where it doesn't need to be as reliable because they.
Mm-hmm.
they're operating at a different scale. It's lower throughput and et cetera. When you look at the technologies, you know, their technology is improving, but so is ours. If you look at some recent preprints that have just come out in the last week or so, it still shows there's a very significant gap in accuracy. There's a very significant gap in variant calling ability. You know, there's a lot of a lot of advantages of using HiFi technology.
Yeah.
over the Nanopore technology. The other thing I would say at a fundamental level is the level of industrialization that we have and full product that we have. In fact, we just won a deal just, it's happening kind of right now, where we're replacing Oxford Nanopore in a perceived high throughput laboratory, because they see the, You know, Revio is really the first platform that PacBio has had to fundamentally have the throughput, have the economics
Yeah, yeah.
have the complete product. You know, people that were trying to use ONT on, in a, in a high throughput setting, I mean, it's pretty obvious when you look at the benefits. Different, different markets, still long-read technology. They're making improvements, but we continue to make improvements, and I think we're in this acceleration mode right now.
I'm gonna finish off with this, then I wanna talk about the short-read, the Onso technology. Going back ancient history, there was always this complaint about raw read accuracy.
Mm.
Right? Is that still an issue with customers or are the HiFi sort of, like, taking care of it and like this? I mean, I'm just sort of questioning, it's like, where has that debate gone, and is it even is this more academic than sort of anything else right now?
I do think it's a bit more academic.
Yeah.
As raw read accuracy increases, the total.
The costs go down.
The costs go down.
Yeah.
The way we've overcome raw read accuracy is with HiFi, and the way we've improved HiFi is. For example, on Revio, you need fewer passes to achieve HiFi than you did with Sequel IIe because we have Google DeepConsensus built on board. By having DeepConsensus on board, we're able to use fewer passes to get to HiFi, which allows us to shorten the run times.
Yeah.
We shorten the run times from 30 hours down to 24 hours. Now every day, Revio can crank 4, you know, 4 SMRT Cells at a time, 100 million, look at 100 million molecules simultaneously, and it changes the whole equation. We still focus on raw read accuracy, but in terms of continuing to improve it, but it's not, it's not so fundamental because HiFi actually overcomes, you know, overcomes the problems with that.
Got it. Going into accuracy, I think one of the selling points you're making on Onso is that you believe that technology has a higher, you're targeting Q50
Yeah.
With that, right?
That's exactly right.
Right. Could you talk about this, like, how you sort of see that initially? 'Cause I mean, there are a lot of options now in sort of like that benchtop mid-throughput range.
There are.
How does that whole dynamic go out, particularly when you've got Illumina talking about wanting to put them on their, I forget what they're calling it, whatever they call it or they're currently calling Chemistry X on their, on the NextSeq and sort of like that. How do you sort of see the whole thing playing out?
Well, you know, first of all, with Chemistry X, you know, for some reason, they didn't get the accuracy that they were purporting that they were gonna get. But that's a different topic. If you look at Onso, you know, we thought about short reads will have a very significant place in the market in the long run, particularly in needle in a haystack applications. Well, to do well in needle in a haystack applications, accuracy is essential. You have to know that your sequencer is calling the bases accurately. If your sequencer can do that, then you can reduce the coverage, right? Because you know you've got the right answer, or you can look more deeply.
When we acquired Omniome because of this SBB technology, we believed we could create a highly differentiated product that could create a, you know, beachheads in the market. It turns out now that we'll start shipping Onso late this quarter, that Onso in fact does that and does it swimmingly. You see, you see routinely on Twitter people talking about our beta users talking about getting Q50 accuracy from the first read.
Mm.
from the first base to the last base in the sequencing run.
Is on I'm, yeah, I've heard.
No technology's ever done that before, ever.
Yeah.
That changes the way you think about experimental design. It changes how you what you can look for. We just sold a bundled deal. We just got a PO for a bundled deal the other day. It was because of the data. What this customer was able to do with the, They're doing cancer research. What they were able to see with Onso, they did a head-to-head comparison. They couldn't see the same things with Illumina technology. That, you know, sold the deal. Then there, away you go. I think having that differentiation really allows you to sell a solution.
Yeah.
rather than sell a technology. Yes, it's a crowded market, and certain... you know, what you're gonna see over the next few years is different sequencers will do things better than other sequencers, and therefore, most labs will run in a mixed lab, a mixed technology environment, and then pick their technology for what they're
Yeah.
trying to accomplish. Onso. You know, really enables that with this unprecedented accuracy. What you'll see from us is continued push to show more data, more use cases, more applications as we start to get the product more into the market, it will become, you know, it will become more and more clear all the time how that accuracy changes everything.
Yeah. I mean, I've got the sense that the sequencing market is gonna evolve a lot like the mass spec market, where people have
Mm.
various different machines there.
Mm-hmm.
-to go with it. Let's talk about being able to call modified bases epigenetics like that. I mean, how big of a driver is that to your overall sales pitch?
I think that it's really important that you get, you know, epigenetics with every run on Revio. It's just embedded.
Yeah.
It's just another form of data for you to analyze. You don't have to, you don't have to pre-treat the sample or lose it, and you get it in the same actual assay. The experimental differences from run to run, all that stuff is taken away.
Yeah.
I do think it's really important because if you're trying to be complete, you know, you wanna see everything.
Yeah.
Today, you know, today we call methyl C. We do have capabilities to call other epigenetic markers that we will introduce in the. Most of it's software, right? It's, it's understanding and analyzing the traces that come off the sequencer and understanding what that means. We're able to, you know, improve the software over time and add these other markers so that you have the most complete view of the genome.
Yeah.
that you can have.
Stepping away from the science and talking about, I don't know, financial issues.
Oh, boy.
since that's my job. Can we talk about margins and so, like, margin progressions, and
Yeah.
You're, you know, you're sitting on a nice cash position, OpEx. It's sort of like all that, and it's like, how do we get the margins back up, you know, as these products come out? 'Cause obviously you're taking a hit right now and...
Yeah.
Like, how should we think about OpEx going forward and, you know?
Sure.
deep in you certainly are well capitalized
Yeah.
in the near term, so.
We'll talk about, We'll start with margins, and then we'll go to OpEx. On the gross margin side, you know, first of all, getting that product mix shift that I talked about earlier is really important because that will drive the mix in such a way, and that could add substantially to the gross margin line. The second thing that will happen is as we scale... First of all, 2023 is gonna be a transition year, so margins will be a bit-
Yeah.
will be depressed as a result of that as we transition from Sequel, the Sequel II paradigm into the Revio paradigm specifically, and also as we get Onso out of the gate. You know, margins at the beginning of the product life are a lot less than margins later on. 2023 will be a more of a challenging year on the gross margin front for us, and you'll see improvements in 2024, both from mix but also from actual production costs of both Revio and Onso. That will help us drive towards what we've said for 2026 is margins in the 55%-60% range is kind of our current thinking. Could we be above that? We could, but right now we're, you know... That's what we, that's what we highlighted as our initial targets.
Yeah.
At that place, we can live and build a business off of that, but we definitely wanna go strive to get above that over time as we launch more platforms and take cost out of Revio. For example, some of the compute cost over time, we think we can eliminate some of it, which will drive, you know. It's actually one of the biggest drivers of cost in Revio itself, is the compute that we have all on board. From there, if you go down to the operating expense line, we do have a very significant R&D spend right now, and what we've committed to in terms of our long-term models is no more than 5% operating expense growth.
I believe we can actually achieve that, and we will achieve that because what we'll be able to do is moderate our R&D expense and then trade R&D dollars for commercial dollars, for example, as we scale and grow. The R&D itself won't really suffer because what we've done over the last couple years is develop two major platforms simultaneously, and what you'll see from us is better phasing of the product launches such that you are not doubling your spend...
Yeah.
So to speak, for the prototype phase, the very expensive parts. But we had to do that in order to get to this point where we had the capabilities to, you know, kind of accelerate. As part of my growth plan was in the phase one, scale the organization for commercial, develop a multi-product portfolio, be working in parallel on research and development rather than just, rather than just in serial. That phase has ended, you know, with the launch of Revio, and now we're in what I call the acceleration phase, where we're moderating R&D, we're feathering out the projects a little better, and we're scaling our...
We'll continue to build our commercial organization as we see growth, which will drive us to positive cash flows and that durability, which is really my end game right now is getting to growth and durability, so that we have strong free cash flows into the future.
Got it. With that, we're out of time. Christian, thank you. Thank you, everybody, for listening. proxy season's coming up, so don't forget to vote. appreciate that. Thanks, everybody.
Good to see you. Thank you very much.
Thank you.