Seer, Inc. (SEER) Earnings Call Transcript & Summary

Great. Thank you, operator. Good morning, everyone. I'm Derik de Bruin, the Senior Life Sciences and Diagnostics Tools Analyst from Bank of America. Welcome to our 2021 Virtual Viva Las Vegas Healthcare conference. We appreciate you joining us today. With me from BofA is my colleague, Mike Ryskin and our next company up is Seer. With us from Seer, we have Founder and CEO, Omid Farokhzad; we have David Horn, CFO; we have Omead Ostadan, President and Chief Operating Officer. Gentlemen, good morning. Well, thank you for being here.

Good morning, Derik.

Great. So Omid, given that you're a relatively new company, I don't know if you want to make some opening remarks or like just a nickel sketch on the background of the company, just to like set the stage because we have a whole set of questions to go through. But just because you may not be as familiar to the broader audience, maybe a little bit of background.

Sure. So I'm Omid Farokhzad. I serve as our CEO, at Seer. We found Seer in 2017 with 2 other colleagues, Philip Ma and Bob Langer. Bob is a professor at MIT and Philip was CEO at McKinsey and then Biogen and then starts here with me. The premise of the company was that we had a technology platform that allows us to get access to deep unbiased proteomics and to do it at scale and efficient, in essence, to be able to, for the first time, impedance match our ability to access deep genomic content in an unbiased way, with deep proteomic content in an unbiased way. And so you can go from genotype to phenotype, tie your risk factor to your functional status and so it was quite exciting. And before Seer -- I was a professor at Harvard Medical School, where I was for about 20 years, and I left to start Seer. Now I'll fast forward. In late 2020, we brought our first product to the clinic -- sorry, to the hands of the customer. That is the Proteograph Product Suite, which comprises our consumables that includes core nanoparticle technology. And automation instrument that's an OEM that allows us to automate the assay. The assay is a 7 hour assay, 30 minutes of hands on time, 6.5 hours of automated time to run 16 samples. And then a software suite that lets a customer go from biological data to biological insight. And essentially, really make it possible for most labs, not just the key core labs that does for proteomics, but most labs to access proteomic content and complement that with the other studies that we do. We took the company public in December of 2017. We have a commercialization strategy, which is a 3-phase strategy, one that's being used in commercializing other disruptive platforms in the past. And my colleague on this call, Omead Ostadan, is an expert in that area. In fact, I say he's the genius that actually invented that commercialization strategy, having launched 11 sequencers at Illumina prior to joining Seer as our President and Chief Operating Officer. We completed the first phase, which was the collaboration phase, where we installed our product in 4 different labs. And announced recently that we have now entered the second phase of our commercialization, which is a limited release. That phase will continue for the balance of the year before we switch to the product release, and we can get into all those details on this call. Thanks, Derik.

Great. Thanks. So I've been around the life sciences sector for 20-plus years now. And when I sort of think about the evolution of the space, you had the genome project and then everybody said, that's the cloud, then we're going to do the human protium project. And there's massive investment, and there were some companies that were out there that bought [ 50 Q ]tops, and there was the top top wars to sort of like do this. And then those technologies never delivered because they basically couldn't do what your company can do today. So you went through the situation where everybody got excited about proteomics, it's fallen off. Now it's sort of making a renaissance. With that in mind, you were the first company, you were the first -- the new generation proteomics companies that went public. And there -- since then, there's been a number of companies that have come after you. And I think one of the main questions we're getting from investors right now is you've got Seer, you've got QSI, you've got OLink. And I think investors are really struggling about how do all these puzzle pieces fit together? Like how does this go on? So I think if you could help us understand the proteomics landscape and where Seer fits in relative to everything like that, I think that would be hugely beneficial for a lot of people on the call.

Yes. Derik, by the way, I'll make 1 high-level remark, which is most disruptive products that come, they come on the back of a lot of other ones that have failed, right? I mean, just look at gene therapy and all the enthusiasm that went into it and then it was dead in the water and then now it's back and boy, is it back with a vengeance. And so proteomic is no exception. I think we were missing the technology and the technology caught up, right? So if I look at the proteomics space, at a high level, I would put it into -- these companies into 3 buckets. Those that approach proteomic in a targeted way. Those that say, I don't like a mass spec, and so I'm developing a new detector. And then those that say, I'm going to approach the proteomics in an untargeted or unbiased way. So let's just take a step back. Remember that we go from our 20,000 genes where every one of our cells in our body has those same genes, we don't go from 20,000 genes to 20,000 proteins. We go from 20,000 genes, and we end up at 1 million-plus proteins. And that's because biology is complex at every step when you go from the DNA to an RNA and an RNA to protein. And then after the protein is made -- modifications that happen. Biology is complex and proteomics is far more complex than genomics in terms of content. So if you approach proteomics in a targeted way, let's say, you're one that has a panel of 1,500 proteins in your panel or 5,000, et cetera. If you get a panel of 5,000, you're missing some 995,000 protein variants that you may not be able to distinguish from each other. If you approach a targeted-proteomic approach, keep in mind that an average human protein is 470 amino acid long. An average targeting ligand that this targeted approach is used binds to an epitope that is 5 to 8 amino acid long. So when you bind to a small epitope of a protein, that protein could change anywhere else, either through post transitional modification or through amino acid changes that happens because of genetic variance for that happens at the level of the transcription, et cetera, you will miss that. Now so a targeted approach is like looking for your keys under the lamppost. You find exactly what you're looking for, but you don't find new content, right? In order to find new content, you would need to have ligand for all of those variances, and it's just practically not possible. Now who does targeted approaches? Well, you know them Quanterix, Olink, SomaLogic, and each of them have their own unique way of doing it, and they each have their own strengths in doing it. Then there is a group of companies that have emerged that say, we think the mass spec is yesterday's technology, and we want to develop the next-generation technology. And so let me make one point very clear. I actually think mass spec is a great detector. The problem isn't with the mass spec, the problem is with the workflow that sits upstream to the mass spec because not only are proteins complex in terms of their structure, proteins are also complex in terms of their quantities. And so the mass spec requires a very complicated workflow upstream for it to go deep into the fractionation depletion, et cetera. That takes a long time to get through a detector. That the ones that are trying to replace the mass spec, they are the Nautilus, , Encodia, QSI. And in some cases, they've guided the street to a product that may come at the end of 2023 or maybe some of them sooner, and some of them haven't guided at all. And they're in the research stage. From our perspective at Seer, we're detector agnostic. In fact, I want to be Switzerland and I want to service all of them. Today, we've got a base of 15,000 mass specs. That's an awesome base of potential customer where the Proteograph can sit upstream to their detector. But in the world where these other detectors get commercialized and get commercial traction, the very end of the Proteograph workflow then could be changed and the Proteograph will then sit upstream to their detector. But for us to invest R&D dollars to do that, I first need to see the detector come get commercial traction. So it actually merits and investment in R&D dollar for us to do that. And we would hopefully do that should that happen. The third bucket, which is the entity of the companies that are approaching the proteome in an untargeted unbiased way to access the proteome in a deep way, there is little exactly and of one company that does it, and that is Seer. Now if I take a step back and look at what happened in the genomic world, we were doing sequencing all along and I put plenty of my gels to run my S-35 label sequencing gel, as we had there too. So we share that pain. Now when NGS came, targeted-genomic approaches like a PCR did not go away. In fact, as NGS allowed us to go ahead and sequence a lot, now 1 million genomes have been sequenced, 10 million exomes have been sequenced, 695 million genetic variants have been identified because of that. The need for PCR actually increased. And so the proteomic TAM, which is $50 plus million, a lion share of that TAM is going to be in content generation, exactly the way NGS brought much of that TAM of the genomic space. But as this content gets generated in an unbiased way, actually, the utility for targeted approaches will increase. Now that's in the foreseeable years to come. I actually think the same movie will play again as it did in the genomic and the [proteomic] space, where if you look at Illumina today sits at, whatever it is, $70 billion market cap, and the ecosystems that it created 10x, Garden Health, Grail others, maybe $30 billion or $40 billion around it. Fast forward another decade, that $30 billion, $40 billion is going to be $200 billion or $300 billion and maybe Illumina would be $100 billion. The application is going to grow. My suspicion is that Seer will grab a massive amount of that TAM, create end markets that don't even exist today, expand end markets that exist today. And then the next thing that would happen is a lot of these applications will grow. And by the way, that includes targeted approaches. I hope that helps.

Yes. A great overview. And I think Mike wanted to jump in.

I was going to jump in real quick. Omid, on that point of sort of the targeted versus untargeted and kind of really splitting up in 2 different markets, not just proteomics, but looking into the sort of the subsegments. Of that mass spec installed base, I mean, how many do you think are -- how many are interested, how many are amenable for an untargeted approach? Because I would think that a large amount of the labs out there are going to be more focused, are going to be more -- whether they're translational or what, they may be more looking for a more focused targeted approach. So how big is the ecosystem out there, sort of who are the labs that are going to be doing untargeted? And how much of that sort of groundwork do you need in that market?

Mike, let me -- I want to answer one bit to that question, and I'm going to hand it to Omead Ostadan. The bit of our answer is that the broad mass spec installed base is about 50,000. The 15,000 is the one that does proteomic. In that 15,000, our folks that do targeted, often clinical, and marketed, right? But remember, not only is that a big market for us, and Omead will get into that in a second. But also, there is an entire segment, folks that are doing genomics, transcriptomics, other omics today that just did not have access to proteomics. Not only do they not have access to mass spec, but they just have no ability to get to those content at the scale and level that they need that actually Seer enables, right? In fact, we have examples of that, not only in the collaboration phase of our customer but examples of that coming into the limited release phase of our commercialization, let me hand it to Omead, who can break that mass spec segment for you, but also the customer base that is not just exactly your mass spec proteomic customer. Omead?

Yes. So Mike, obviously, I haven't talked to all 15,000 owners of mass specs, but up to the [ ones ] I have, talk to. So there is a little bit of a selection bias here, so I want to do is I want to acknowledge that. But the level of interest has been exceptionally high in being able to do unbiased in deep plasma proteomics. And it really comes down to what Omid said earlier. I think the intent, the clarity or division around the science has been there, it's been technology that's been gating people's ability to do this, right? So in a lot of ways, this -- in some ways, it's actually been easier of a conversation than conversations I had on the genomic side, trying to sort of get people's heads around why you should do genomes at large scale. To some extent because I think mass spec has been around for a long period of time. I mean, people have tried to do unbiased and deep plasma proteomics, clearly, there've been examples of it, all limited in terms of scale. And now that we can enable that application, that scale is very attractive because, again, it goes back to what Omid said earlier, it's estimated that there's roughly in the order of 1 million different protein variants across the human population, of which we know arguably a tiny fraction of it. And so much of the value, if you think about what drives publications, what drives content that can perhaps be helpful in biomarker discovery. Or what can drive biological insights that can be helpful in therapeutics development or monitoring of clinical trial populations. It is all in those variants, I think, not all, but a very substantial amount of it is in that undiscovered uncharted part of the proteome. And so the desire is there. And so that's why the conversations have been actually more about help me understand how well your technology works and how does it scale as opposed to, do I really want to do unbiased and deep plasma proteomics. So I actually think the pool, certainly based on what we've seen from those who are in proteomics and have mass specs, is strong and is only going to get stronger. And to the point that Omid raised, if there is substantial part of market opportunity for us, we believe is in multiomics and particularly among translational and commercial companies who've scaled upon genomics, who've cataloged all of these variants. And really now, 15 years ago, the bottleneck was generating the genomics data. Now I would argue the bottleneck is characterizing the genomics data. And this gives them an extraordinarily powerful and unique tool to go about the motion of characterizing those variants and really separating the weeds from the chaff. And importantly, also really building into the proteome to figure out aspects of biology that might only be elucidated through the use of the proteome. And so there is also a strong pull and a part of it as well, and that will only mean expansion or the opportunity for not only us, but mass specs as a detector type.

Okay. That's helpful. Thank you.

And so since you basically mentioned it, how should we think about the pace of rollout of presentations and publications of data for the balance of '21 and into '22. How should we think about milestones? I mean, unfortunately, we are financial analysts, and we have to think beyond the science. So how should we benchmark you guys for success?

Yes. So let me take a crack at that, and Omid and David, please chime in. So -- and I think we've talked about this in previously, in fact, I think it may have come up in our earnings call earlier in the week. And so here's what you can expect. Clearly expect to see an increasing number of peer-reviewed publications using or describing Seer's technology. You will see more of it from us. Now we're a little bit ahead of the curve because obviously, we've had the technology, and we were working, but you can expect to see it coming from users of our technology as they are scaling up. And they are scaling up at a good clip. Now keep in mind, and you know this, the lead time for peer-reviewed publications is typically measured in months, not in weeks. So how do we bridge that gap? And the way we're looking to bridge that gap is provide opportunities for users of our technology to be able to talk about their experiences, either mediated through us, an example of it is actually a video recording that Dr. Flory of OHSU recently did that's posted and available on our website. And additional examples of it is through conference presentations. In fact, with the upcoming HUPO conference, we're going to have one of our -- Dr. Flory's colleagues talk about their experience and some of the data. And so what you'll see progressively through the year, Derik, is as we add more customers and as those customers scale more of their technology that they will begin to talk about it. Both in the context of webinars and interviews mediated by us as well as poster presentations and abstracts and podium presentations at conferences, and all of this will all -- we believe, will culminate with a stream of peer-reviewed publications that are going to likely come out, I would suspect, the earliest is probably the tail end of '21, but likely going into '22. Omid and David, do you guys want to add or elaborate on any of these?

In fact, that was complete. We have nothing to add.

So what's interesting, I commend you on your financial model that you've built in a very conservative ramp this model. And one of the things I've sort of been thinking about the market, I mean, when you think about like when next-gen sequencing was introduced, right? I mean you had years of ] and the automated sequencing and flat gels. And so when NGS hit, everybody knew what to do, and particularly in the HiSeq kit, it was like wham, bam, we're off, it's gone. So the proteomics market is different, right? I mean, to your point, people understand the value of proteomics, but it's not ubiquitous, it's not there. So how do we think about the revenue curve? I mean, so -- and the question -- it goes at this point, you locations, what's the inflection point? I mean, because I think a slow steady sort of like ramping growth makes sense. But in theory, there's a potential for it to go like that as well, if it takes off. So that's what I'm trying to understand is like think about how to model the business, thinking about what the inflection points or that revenue inflection -- the revenue growth in it.

Look, Derik. I'll comment, and then I'm going to have Omead and actually, David as well comment. I would say, for your models, please keep these models as conservative as possible. The comment I would make is that again, many thanks to our colleagues in the genomic world. They have paved the road for us that we're now driving on. And so I think that this slope of adoption would be a bit more robust than we saw in the genomics. I do think though that what's going to drive a lot of that is going to be the initial wave of publications that's going to come. And as Omead highlighted, publications lag by, a, the time that it takes to generate the data; and b, the time that it takes to go through the publication process. And I don't expect any papers to be published from our collaborators or customers in 2021. I do... [Technical Difficulty]

So perhaps, if you don't mind, I'll just have a -- I think what he was going to say, he does expect him to hit in 2022, right, in terms of the publications. And so let me maybe perhaps pick up where we Omid dropped off. There actually -- maybe it's because I lived through it, Derik, I have different memory of genomics because you make my job at Illumina seemed a little easy, make it a harder as far the people the technology. Because what we had is that, look, there was a couple of things that were tailwinds that were really helpful, right? It was -- you had the push for the next wave of genome sequencing, you had just -- this really pent-up demand where people wanted to chart the next wave. Ironically, it took a long time before genome sequencing became a key application on it. So with some of these other applications like ChIP-Seq and RNA-Seq that really sort of catapult the sequencing and then followed within IPT and all the rest of it. So where we are right now with our thing is I want to say, what's interesting is that the level of interest in this technology and in doing unbiased and deep plasma proteomics is very high. We anticipated this coming into the year. Now we chose this commercialization model because I firmly believe that for a disruptive technology, the right way to do it is to go to a measured number of customers, who can really operate at scale because the full value proposition of this technology can be displayed at scale. And you want to go to those customers who have the wherewithal, the vision and the incentive to do it at scale. And in doing that, they not only exemplify the technology, but they create a blueprint that other people can follow. And this is one of the things we saw very successfully at Illumina with the introduction of the HiSeq X. Limited number of customers, they really showed the utility of genome sequencing at scale and that you could do it at scale. And the next sets of customers who adopted the HiSeq X were able to ramp their consumables at a much faster rate than in a preceding sequencer that had been introduced. And a lot of it was because they had a blueprint to follow, and it was very, very clear. And so the interest here, again, not surprising to us, is quite high. And we're seeing it in our ongoing conversations. And so my expectation, you said what's the ramp, it is a little bit too early for me to look deeply in that crystal ball and tell you, except to say that I completely share Omid's sentiment that the groundwork is there, the biology and the need for this biology is there and the fact that you have all of these genomics and translational companies having advanced is actually super helpful, which is why we're seeing more of a pull from commercial than we're seeing from academia. And that's the flip side of what it was with genomics. That's why I do think, you're going to see -- and those commercial entities, quite frankly, are less sort of like dependent on publications, right, which is why our view is that going into 2022, we're going to have the level of interest and demand needed to continue to ramp adoption of this technology in a very successful way. So I am optimistic as I have been over the course of the last 12 months since I've been at this company. And if anything, my outlook and optimism for where we are, the opportunity and the uniqueness of our technology is extended.

By the way Derik, my apologies for technical issues, on my computer, but I'm back on. Sorry about that.

No problem. No problem. We got it. So let's talk a little bit -- I mean, I think what's interesting is at Illumina, you obviously control the box and everything that went into it and the users here. But here, you've got to look at -- you've got to worry about Orbitrap and timsTOF and triple tops as sort of like the machines, and the standardized approaches and protocols are doing that, it's like how are you thinking about those relationships you have with Bruker Thermo, Danaher and the standardization to make sure that if somebody's doing a timsTOF that day, and they do the same thing on an Orbitrap, it's reproducible, and they get the sort of same results. So that there's not this question about like, well, it's not -- yes, well, it's like a liquid biopsy questions like, well, here's liquid biopsy assay, not [indiscernible] liquid biopsy assay and . So how are we looking towards like reproducibility and uniformity between the different instrument types.

Omead, do you want to take a crack at that? Or do you want me to answer that?

Sure, why don't I head off and then -- start it off, and you can only improve on it, I suspect. So here's how I sort of think about it, one of the biggest areas where lack of standardization and reproducibility exist -- just certainly in terms of unbiased and deep plasma proteomics, is the variable upfront workflow of depletion and fracination. Not only is it lengthy and cumbersome, but there is some very specific black art in it. So our workflow had full stop, takes out a substantial part of that irreproducibility add an overall paradigm of the experiment because we're standardizing the front end of it. And that's critically important because you're now reducing and eliminating a lot of the sources of noise that come into the experiment. Now on the back end, but you're absolutely right. There is a range of mass spec technologies that are available. But also sort of if you think about it from a sequencing perspective, that's not dissimilar to what's happening in sequencing. Now sequencing -- Illumina has a substantial share of the sequencing space out there, but you've got companies like PacBio, you've got companies like O&T, and you've got BGI and other emerging companies. And you know that if you take the same genome and use sequencing at an Illumina platform and then on a PacBio platform and O&T platform, you're going to get a venn diagram that's overlapping, but each of them are going to have their own unique spaces. So the fact that you have, if you will, detection specific signatures is not uncommon, and it's certainly not slowed down the adoption of sequencing or its propagation, and I expect it will play out the same way here. The thing that's more important in some way is that, look, here, the question in a lot of ways, especially early on, is discovery of the unknown. And that, as you increase the sample size. If you increase sample size, that is going to cut away a lot of the noise that comes out, if you will, from range of mass specs. And which is why my view is that it's really not an issue, right? And these deeper relationships only help us standardize platform specific workflows that then tighten the output of data from a timsTOF across all labs that have timsTOFs, for example, or across all labs that have Orbit. So Omid, if you want to add to that, go ahead.

That was amazingly perfect. The only thing to add i Derik, on top of that, the Proteograph could work with an instrument that comes out anomalous or anyone else for that matter. It truly is, it solves the problem upstream, which is a, cumbersome; and b, variable, and it makes it robust and it makes it scalable.

We're at time, but I want to squeeze in one more question. You -- and mind you, if you get to this point, it would be a great point to have. But obviously, Illumina was the arms dealer for this and -- but it got to the point where they needed content, right? I mean that's clear with the Grail transaction and going after some of these other things. And there is -- I agree with you. There's going to be value created around the ecosystem that you're building. 100%. I mean you've done PrognomIQ. I guess, could you sort of like talk about why you're not worried that, that's going to be in competition with your customers. And I mean, the feedback that we certainly have had from people post the Grail move at Illumina has not been kind. So how do you make sure you don't butt heads with your customers as you sort of go down that content development?

Derik, the very reason why we spun off PrognomIQ is to clearly signal to our customer that we would never butt head with them. That we would be there supporting them, we would help them develop content, develop methods to help them achieve their objectives and their goals. PrognomIQ, we retained a 19% ownership, but PrognomIQ, gets a very -- an identical deal as any other customer would at the volume that PrognomIQ is signing up for. So in other words, if Grail gives us the same economic to us as PrognomIQ is in terms of volume and commitment, then Grail will get a very same pricing that PrognomIQ would get, right? In fact, if Seer is exactly right about its business, every one of those liquid biotech companies will have a further Proteograph in their labs.

Got it. With that, gentlemen, thank you. Good luck. Enjoy the rest of the conference. Investors, thank you for support, and have a great day and be safe and see you next year in Vegas, we hope.

Thanks so much, Derik and Mike.

Thank you.

Bye.