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

Hi. All right. So here on our next panel on NGS technology and how it applies to both research and clinical. We're excited to have Christian Henry, President and CEO of PacBio, who has Zoomed in. And then Sean George from Invitae is also here in person. So hi, guys.

Hi.

So maybe to start things off, you have a perspective on both sides, and PacBio and Invitae has -- have talked about the partnership that they have there. But what is the basis for the partnership between Invitae and PacBio? And maybe starting with Christian, could you provide your thoughts there on the partnership?

Yes. I can. First of all, thanks for this invitation. Sorry, I couldn't be there live. Looks nice. But the -- Sean -- when I joined the company, the first -- one of the first calls I made is I called Sean. And I've known Sean from the past, and I think he is one of the visionaries in the field thinking about the power of genomics in clinical applications. And the companies had been having some discussions about long-reads and whole genome sequencing in the clinic, but the truth is it hadn't really gone anywhere. And so I called Sean up and introduced myself and said hi. And the truth, we share the same vision of long-read sequencing being a critical component of really improving our understanding of the genome in a clinical application. And so I think the conversation took maybe all of 5 minutes to say, "Hey, we should put something together and do something really big." And the rest is kind of history. And so it was a pretty straightforward discussion. We shared kind of the same vision and passion for clinical genomics and where we're going and where the tool -- where are the picks and shovels. And he is the digger, for lack of a better word.

And just I forgot to mention earlier, if you would like to ask a question in the audience, there's cards on the table and we have the iPad up here. But then I guess, Sean, your thoughts on the same questions from Invitae's perspective.

Yes. No -- I mean, as Christian pointed out, we had been talking with PacBio for a while. And as soon as I heard Christian was taking the job, I knew -- again, we had worked together briefly in the past, also having been at Illumina during that period of time when sequencing cost per base rapidly declined, and they resulted in explosion of application space. I kind of knew, okay, we might -- this might go a lot better. So when Christian called, it was pretty obvious. And he's right, it took all of 5 minutes, like, hey, we both think this, why wouldn't it be? And then we just got going on it. And I think a lot of it is kind of -- I actually think it's kind of -- it's analogous to, I think what's going on in our space in general, right, that you can have great technologies with niche applications. Or you can ask what could the utility of this be at a certain price point. And again, this is where I think Christian understood this from a direct experience. And I also think we're a unique -- not solely unique, but there are only so many players with the appetite for the kind of scale and volume that we're looking at. And so there was also the basis of a partnership there that we could kind of codevelop the same together, have a benefit of our application business, the diggers -- we could dig faster. And also it benefit the tool business as a result, the kind of the ancillary learnings and kind of scale therein. So it's kind of a clear win-win. On top of it all, we -- in particular, there's certain disease areas where having the PacBio HiFi sequence, we've demonstrated together actually prior collaborations with PacBio that in some of these cases, your diagnostic yield increases by 40%, 50%, 60%, which for a family with a kid with a problem and everybody is pulling their hair out trying to figure out what's going on, that's a very real -- that is a real value in -- for our customers out there. So it was kind of obvious just to try and push that forward.

Got it. And Sean, I guess, maybe adding on that, what do you think the first product of this partnership might look like? And any upcoming milestones or time lines that you could discuss?

Yes. We're doing -- pursuant to our last conversations, milestones and time lines, we keep it out of the public for a reason. There's some technology risk there. With that said, I think Christian and the team are very clear eyed as to what those are. There's implementation questions, like what format and what kind of scale and what cost do we go for right out of the gate versus what we know this can lead to over time. And our joint steering committee is frankly just working through those on a monthly basis. It's kind of a decision, do you go for something sooner that will get you part of the way there, do you do 2 steps out at 3 bites of the apple, what have you. The time lines, we're certain it's going to get to where we ultimately get. It could be a couple of years, 2, 3 years, maybe longer. I think the -- I mean christian can speak better than I can. But from what I can tell, the technology is advancing to the point where you could see some early returns if we decide to peel off kind of an off-ramp along the way, but we haven't really decided that yet.

Yes. I mean I think that -- I think Sean is right. The core technology that we're going to deploy is deep in development already, and it's a core -- the core technology revolves around the sensor and how many individual fragments of DNA can you look at simultaneously on one chip and then how do you deploy that in different configurations to drive whatever scale you're looking for. And I think that the -- what's so great about this partnership is that we can work together to define some of those milestones and some of those product configurations. And we can leverage Sean's expertise [indiscernible]. Turns out they handle a lot of samples every single quarter, and that workflow is going to be so critical because our belief is that this can -- this tool can potentially replace the majority of exomes that are done, for example, and that's a lot. And so you really want to have high capacity, high capability workflow from INDEL and not just in the core sequencer. And that's so appealing and exciting to me. And then I'm sure Sean will probably comment that what we're going to learn as we start to push hundreds of thousands of sequences through the long-read paradigm is going to be really, really exciting and transformative.

Yes. That is pretty clear. The closer you look and the more research program, the collaborator is chasing that down and then you kind of even think all the way to like the T2T Consortium, right? You just start seeing a lot more. And it's a virtuous cycle. That's the other aspect of this. Regardless of kind of the immediate use of a long-read HiFi genome, the virtuous cycle of more discovery, more rapidly increases utility, drives the utility towards a broad medicine in the first place.

And then maybe one more on the partnership, and then we can maybe dive in a little bit on some of the technology on long-read versus short-read. But maybe looking out, say, 3 years, how do you see this partnership evolving? And what are the future products that you could see being developed from this partnership?

Yes. I mean I'd love to say like, we're going to evolve into the Wintel duopoly that we saw in the '70s and '80s, and we're going to be $2 billion -- $2 trillion companies. I can't -- it would be crazy to argue it's going to be that, but that's the idea. Like that's the general idea is, like, hey, we push the application space for in this particular application. Christian, I mean, you can speak for your part. There's a lot of applications for the PacBio platform. I'm hoping to pull it as fast as possible in the medical genetics as we can, but...

Well, I think that's what's so powerful for us. As Sean is pulling us along in his space, and we're pushing on the technology front to drive even more applications for long-read because what we saw in the first wave of NGS, and I was part of that, is that in the early days, there weren't any applications. And it took the development of the technology and driving the installed base. And the customers really define and develop the applications, many of which are the precursors to what Sean uses today and what the other testing companies and customers are using today. And the same phenomenon is going to happen in long-read, and we're seeing it now actually all the time. We're seeing fundamental papers coming out that are showing use cases of the technology that, quite frankly, weren't possible before HiFi. And when you couple that with where we're going on, evolving the platforms to drive cost down and throughput up, you're going to see a pretty dramatic increase in the number of applications that -- where long-reads will be ideally suited. And that's one thing that's really compelling. And I'm sure Sean and I will cook up a whole bunch of different ways to use the instruments as the technology gets into their lab in a very real sense.

Christian, maybe even diving into the long- versus short-read market a little bit more, any thoughts on how the short-reads play out? And how important is accuracy in the market?

Well, I think from my perspective, I think accuracy is one area that hasn't really been challenged in over a decade. And if you look at the short-read accuracy today, it's good. And lots of companies and folks work on algorithms to improve the accuracy in SiLiCO after the sequencing is done. But I think it's really clear that if you can be highly accurate at the raw read level, then you just have that much more power. It's less informatics. It's better. It's fundamentally differentiated. It gives customers fundamental flexibility on whether -- how they use that accuracy. Do they use that accuracy to lower cost and drive cost down? Or do they use that accuracy just to gain deeper insight and look deeper into the particular sample? And I think what we're -- what we've found is that with our proposed acquisition of Omniome that, that accuracy is transformative. And we'll see, and I'd be curious to ask Sean the question because he's going to be the user of the technology.

Yes, yes. No. Yes. No. Thanks, Christian. The trade-off is exactly as he laid out. There are applications for which that increased accuracy can translate directly to better sensitivity and specificity, which you use either to -- like you can think of like an early screening application, where it is not quite there to screen the entire population with the current sense and spec and the false positive rate. So you can trade that accuracy into even more exquisite utility of the application itself. I also think of another area that we're developing and we're doing research and developing is in prenatal diagnostics. Not -- there's pre-natal screening, like we talked about for the big trisomies and the microdeletions and whatnot. But then there's ultrasound -- abnormal ultrasound follow-up, which we do a lot of that testing today. And then there's also, in the case -- an unfortunate case, where nobody got the Carrier Screen done in time, and you've got 2 positives. And now you want to know -- a noninvasive way to do that is another example of where that accuracy will translate to your capabilities to even get the test right in the first place. And then there'll be other applications where you'll trade that accuracy and for costs when it's not such an exquisitely important sense and spec issue, but where you're going for massive volumes or you're going for kind of quick turnaround. Whatever you're looking for, whether it is a cost or a time trade-off, like Christian pointed out, it gives you the flexibility. But that -- and that's the way that shows up in the application space. And you kind of go across that spectrum one or the other, depending on the application. But it's always great to be able to tap into if you need it. That's for sure.

And I guess maybe, Christian, following up, I guess, where does HiFi provide a differentiated offering, where you see you competing versus the short-read market?

Yes. Well, this is actually one of the most fundamental reasons why we're so excited about the short-read opportunity because it's completely additive to our long-read opportunity because the long-read technology in HiFi, in particular, has really powerful use cases in whole genome applications and in some RNA applications where you're looking at specific isoforms. And In the short-read application, the short-read technology will really help when you're looking for needles in a haystack and you're looking for small fragments of DNA in a deep sort of way where the number of counts matters a lot. And so we see the technologies as completely complementary, both with the fundamental accuracy hallmark because accuracy is the fundamental element that makes the technologies fly easier and can create differentiation in both HiFi and the SBB technology that Omniome have, have really strong accuracy. And we will focus both in different parts of the market. And what this will do is it will allow us to be much more flexible in how we go to our customers. Rather than asking the customer, can you fit this technology into your workflow, we can ask the question, what questions are you trying to solve. And we have a portfolio of technology. It's actually one of my core strategies for building PacBio is how do we become a company that offers solutions rather than technologies. I think it goes way beyond the technology and really starting to ask the solution, what questions are you trying to answer so that we can offer products that actually solve that problem.

Excellent. Sean, you've spoken earlier a little bit on the mix between long- and short-read. But I guess, maybe specific to ArcherDx, how do you see that future impact there? And what's more important there, the long- or the short-reads?

Yes. I think it's a great example. Depending on the application, for a lot of the cell-free applications like liquid therapy selection or MRD testing or we call it Personal (sic) [ Personalized ] Cancer Monitoring, noninvasive prenatal testing, all of those -- those fragments are short, they're short. So short-read works great. And that's where if you have better accuracy, then you can kind of ramp down that sensitivity curve or the cost curve that we were talking about earlier. But then when you're -- on the other hand, when you're doing hereditary testing, you're doing the pediatric testing, you want the most comprehensive, most complete picture possible. And what I really appreciate about Christian's approach to this is he said right. And look, I used to work at Affymetrix, right? The number of tool providers that take the hammer and hit everything with that hammer, I mean, it's almost the entire industry. We were already thinking kind of -- if you think about our bread and butter sequencing or everybody born getting a genome sequenced, you probably do a mix of a long- and short-read approach. And the reason is because there's inherent systematic biases in both. And if you can run them both to their max efficiency, you can basically kind of cancel out the systematic bias of the other, depending on what percentage of which you're looking at, and you can maximize the best genome at the best cost. And I think it's a great example where if you think of it from a solutions approach as these guys are, it kind of makes obvious sense. There are going to be some applications where you want to do all long-read, some you want to do all short-read. Many, I believe, where you want to do a mix of both. And I think that was just -- so that was -- actually, we -- obviously, we read about that, and it was just kind of -- that was another really quick phone call. It was like, "Hey, we're buying this company. You want to add it?" I said, "Yes. We're already thinking about it. Great." That's just rolling in. And I think also, it's good for the -- I also think kind of personally, it's good for the space to have. PacBio has been around a while. The -- so it's not -- this is a company that can engineer, can build boxes, can produce, can ship, can service. And that's also another real benefit is there's 5 or 6 new next-next-gen short-read technologies. In these guys' hands, you've got the form factor, you've got the engineering, you've got the manufacturing. Like it's all -- it's going to accelerate it for everybody, which is really -- it's really exciting. But it kind of -- again, depending on the application, you'll want one or the other or a mix of both. And I think it does take away the question, like which is better, long or short. What are you trying to -- what data you're trying to get, and the answer will flow from there.

Great. And so for Sean, we actually have a question here from the audience. Considering the partnership with PacBio, what were the considerations around how, what the response might be with your current NGS provider? And how does this factor into this...

You can say Illumina. Like everybody knows it's Illumina. Christian used to work there.

Right.

The I guess how did this factor into the decision, if at all? And what did you do to mitigate the risk to the core business?

Sure. So yes, I mean, like obviously, you've got -- at the time, it was the dominant provider of sequencer. And by the way, it has been and continues to be a great partner for us. Like we've alpha and beta tested a lot of Illumina kit over the years, and we've helped them develop application. We've helped them -- like all the stuff that Christian and I are talking about right now, we've done that with Illumina in past years. And Illumina has, over time, maybe more in earlier years than lately, but they have been driving that cost curve down, enabling our work. We have a great relationship with them. We have -- I mean we have a team of Illumina service engineers that pretty much might as well be employees. They're on 24/7 on-site. Yes. The consideration is -- it's a little touchy when you think about, okay, great, we're going to go and directly invest in and enable a competitor. But what I would say is with that good relationship, we've been very upfront and very straightforward for years. This is what we're trying to accomplish. This is what we're looking to do. Here's what we're willing to invest to drive X sensitivity, specificity, this, that, the other. And I feel really good about it. I feel really good about it. We've had a very upfront discussion. And obviously, yes, it creates a little bit of tension. I'll be honest, I think that they've got bigger things to worry about than what we do with PacBio, so that's also helped a little bit. It's not been kind of front and center existential. But yes, it's -- as you might imagine, it was a touchy thing for such a long-term value partner to kind of introduce a new player in the mix. But it's also -- I mean it's good for the space. It's good for everybody. And so we pushed through it. For mitigation, I'll be honest -- again, I don't think it really required a lot of mitigation. We know those guys well. We know they're not going to cut -- it doesn't do them any good to cut off one of their major customers or do anything like that. We did ink some longer-term supply agreements just in case, but it never came to that, right? The response has been like, what can we do better, what can we move forward. And we'd say, "Well, now it's clear where this is all going. Let's proceed." So yes. I appreciate the question, but it's -- I don't think it's as tensed as kind of as you might imagine, just a bunch of companies trying to push the space forward.

And then, I guess, maybe, Christian, while we're discussing long- versus short-reads and Illumina, can you maybe provide us a little bit of PacBio's rationale on the pending acquisition of Omniome and entering the short-read market?

Yes. Sure. So the rationale is actually pretty simple. It starts with the notion that we want to offer solutions rather than just technologies. And we've already talked a lot about that. But the long-read opportunity is still our core asset and our core strategic focus. But the short-read opportunity does a couple of things for us. First, it allows us to get into markets that this long-read technology couldn't really penetrate very well or at least not in the near term. It allows us to have something to talk about with a significant number of customers that aren't really in the plans of adopting long-read technology anytime soon and gives us a cross-selling opportunity such that we can start to manage the transition from short-read to long-read. And we can talk to customers about -- when we're talking about the Omniome short-read technology to solve their current solutions, we can also introduce them the power of the long-read technology and what we're doing. And so we think by making this acquisition, in fact, we will accelerate the revenue ramp of our long-read business in addition to getting into an entirely new market. And then finally, of course, a very modest penetration into the short-read space yields very significant amounts of revenue, at least to PacBio. And so you could -- very -- you can do the math. Anyone can do the math. But potentially, hundreds of millions of dollars of incremental revenue that could come into PacBio and accelerate our ramp as we get our -- we continue to grow the long read part of our business, but then we get this additional revenue, which will be a decent gross margin to create the scale we need. Because I do think, at the end of the day, these life science business, and Sean would probably say exactly the same thing, scale matters. It matters a lot because it helps you drive cost, it helps you with innovation, it helps you with building, competing against those that are significantly larger. I mean, let's face it. PacBio is a drop in the ocean still relative to some of these other players. And my job is to make us a really big drop or at least a bucket as we kind of -- as we go from year-to-year. So those are some of the key rationale. There's incredible synergy, both -- from -- one thing I think people may not appreciate is how much technical synergy there is. In fact, the PacBio team has lots of experience dealing with individual molecules of DNA. Turns out that experience helps you when you're thinking about clusters of lots of molecules. The single molecule problem is a lot harder to solve generally. And so -- and the Omniome team has some really high-quality research folks, R&D people that will help us continue to scale as we're trying to grow. We're trying to support Sean in our partnership. And more R&D is better at this point.

Yes. And to add on that, I actually think it's a parallel to our entire space. I mean I think kind of we are -- if you kind of like use the tech industry as a kind of analogy, we're emerging out of the Fairchild-ren kind of era where we're selling transistors and resistors and analog. We're selling it all to each other. And I think the real applications are starting to emerge, and they're starting to get -- they're growing now on a ramp that's going to start picking up here. And the scale -- I mean I've used it in the right way. I mean I've been a part of it my entire career, but it's kind of a cottage industry that we're involved in here and doesn't affect a whole lot of people. It's moving to where this is going to be, just mainstream stuff, mainstream technology and scale matters. I mean it really is important. And it's going to be a fixture, I think, here in the next many years.