Welcome to the Growth Conference. I'm Kyle Mixon. I cover lifestyle and solution diagnostics for Canaccord. Please join me for this fireside chat with Seer. Seer offers the Proteograph Product Suite for deep, unbiased proteomic analysis. With us from the company, we have Omid Farokhzad, CEO. Thanks so much for joining us today. Appreciate it. Seer has been public since 2020. We were talking about it, almost five years. Maybe just provide an overview of the company and its strategy and product that it offers.
Yeah, and Kyle, first of all, thank you for having us here. We really appreciate it. We enjoy your conference a lot. Seer’s mission is to enable access to the proteomic content and at depth, scale, speed, cost that was previously unimaginable. Now, the proteome is very different than the genome in terms of its complexity and the level of information it provides. We all have 20,000 genes, and then as those genes make RNA and then the RNA becomes proteins, layers of complexity go up considerably. So 20,000 genes make about 200,000 transcripts or RNA. Those 200,000 transcripts make a million plus, multiple millions of proteins that, importantly, those variants of proteins can have completely different functions. That body of information in terms of how the proteome operates, it hasn't been possible to access it, and we started Seer with the idea to do that.
We went public, actually, as Kyle said, in December of 2020. At the beginning of 2022, we did a broad commercial release of our product, which is called the Proteograph Product Suite. It's a combination of an instrument, consumable kits, and a software for analysis. Since the launch of the Proteograph Product Suite, we've now done two new assay updates. The first version of the product that we launched could run 16 samples in a span of eight hours, and then those 16 samples would go into a mass spec. The subsequent product that we launched, Proteograph XT, you could now do 40 samples in about eight hours, and then that would go into the mass spec, but it actually reduced the amount of time on the mass spec by about 150%.
Most recently, at the ASMS conference, June of this year, we launched our high throughput product, the Proteograph ONE, with a new instrument, the SP200 Instrument. Now you're at the point where you can do large scale, population scale proteomics, something completely unimaginable just three years ago. We also announced, since the beginning of the year, now two that are publicly announced, others that are forthcoming, large scale proteomic. In Q1, we announced a 10,000 sample study with a corporate customer. In Q2 at the ASMS, we announced a 20,000 sample study with Korea University. 20,000 samples on a mass spec to do deep proteomics at the depth of about 10,000 proteins is just something that was previously impossible to even think about. Before Seer, the largest deep proteomic study done in a mass spec was 48 samples.
The deepest proteomic study done in a mass spec was right here from Broad Institute. That was 4,500 proteins. Now we have multiple customers that are doing thousands, and now we're getting into the realm of tens of thousands. I had predicted last September at the Morgan Stanley Conference that 2025 will be the year where we'll see population scale proteomics on a mass spec become a reality, and we've seen that. I will predict now on Kyle's conference that I think the first 100,000 sample study on mass spec will become a reality in 2026. That is really the velocity by which access is being enabled today.
All right, interesting. You're basically thinking like a U.K., B iobank size, or maybe even bigger than that, actually, type of a sample project would be performance using mass spec. That's something that we haven't really seen, like population mass spec in the past, right? To this point, what's been the bottleneck and the challenges that prevented that from being a reality? Why is the Proteograph Product Suite enabling this and just streamlining, unlocking what mass spec can really offer the world?
Yeah. Kyle, I mean, I think there are a lot of similarities between the genomic space and the proteomic space and some of the barriers that got broken down that enabled access to the genome and now enabling access to the proteome. There are some similarities, but there are also some differences. I would say similarities are just like the genome and access to the genome enabled the creation of entirely new end markets that didn't exist before that. Companies like Grail, that we just finished, or Natera or Guardant exist only because access to the genome became a reality. My prediction is that as we get access to the proteome, that entire end markets, new end markets may be created, existing end markets will significantly expand. That is not the same as interrogating a set of known proteins. That is not what I mean by access.
The analogy to that would be like the microarray or Affymetrix, for example, that we interrogate the same thing. That is not where Grail would not have existed if it was because of the Affy technology. Grail existed because of content generation in the genome or Guardant or Natera, for that matter. That is the content that's going to become available. Okay, so that's similarity, between the genome and the proteome that exist. There's a difference, though. In the case of the genome, improvements in the detector is what allowed access to the genome to become possible. Became faster and faster, costs dropped, and that's where the innovation came.
Now, the detector in the proteomic space to do unbiased proteomic or discovery proteomics is the mass spec, and the challenge is, while the mass spec is the gold standard, and the mass specs have gotten so much better, and they continue to get better, my prediction is that that velocity of improvement in mass spec is very consistent for years to come. The mass spec could never scale, and that's where the Proteograph comes in. The complexity of the proteome isn't just about protein structure. It's also about protein quantity and the dynamic range. The challenges of the proteome required workflows that fed into the mass spec. Those workflows were not scalable. Dynamic range was broad, so you would have to immunodeplete abundant proteins. You would have to do serial fractionation. All of those is why the deepest study ever published before Seer from Broad was only 16 samples.
Seer took that bottleneck away and made it possible to do no upstream workflow and take a sample in an automated way, get it to the mass spec. That time is shrinking. For example, in our first two iterations of our workflow, the workflow was eight hours. In the one that we just released in June, you run 80 samples in 4.5 hours. The interesting point is that as we made it possible to access the proteome, the next bottleneck that we actually created for the scientific community was the data. The data analysis became a massive bottleneck, and we made a very large investment. In fact, there was a period of time where about a third of my organization was software and data people. We then innovated in terms of our data analysis suite.
The Proteograph Analysis Suite made it possible to do large-scale data analysis, large content. My prediction, by the way, is that we'll continue to innovate in this space. As Kyle said, we are now very close to approaching a complete impedance match of genome, meaning as you do a 500,000, 600,000 UKBB genomic study, you could do a 600,000 proteomic study in an unbiased way. The discovery power of that unbiased proteomic is not even comparable to what happens when you look at a targeted approach looking at proteins.
Yeah, just to clarify, you're basically saying like you could buy a bank, like, you know, they're kind of investing in the targeted side right now with the Olink, doing Olink with all the samples. They would eventually be interested in doing like an unbiased view of all those samples as well.
Yeah, Kyle, and by the way, I think targeted platforms, I listened to Alamar this morning here, Alamar, Olink, Soma, Quanterix, these are all great products, and they've had their time in various different periods over the last two decades. Targeted approaches will be absolutely important as they are today in genomics. They'll be important in proteomics, but content discovery, where we today are at the very, very tip of the iceberg in terms of understanding proteomic content, discovering that massive content is where biological insight, new biomarkers, new drug targets, new diagnostics are going to emerge from. In the land of the blind, the one-eyed man is king. When you could not scale unbiased proteomics to do a UKBB study, you were forced to do targeted approaches. Untargeted approaches at scale are just becoming a reality today.
By the way, keep in mind, there's exactly 52 papers published on Seer, only 52 papers. The fact that on the back of 52 papers, customers are now starting tens of thousands of sample studies tells you the pent-up appetite to do this because UKBB studies came after 500 papers were published on Olink. I do think that a lot of these organizations, these biobanks, will begin to interrogate the proteome in an unbiased way. As you begin to understand those content, you can put those content on targeted panels for interrogation, test development, and others. This is not to say that untargeted approaches are not suitable for test development. They can also expand all the way from discovery to clinical application. In fact, at Seer, we've done ISO certification, getting our platform ready for clinical utility. You don't want to force the system.
If the biomarkers you discover can fit in a targeted panel, and that is the most cost-effective way of doing it, then that's exactly what you should do. In fact, we spun out a company called PrognomiQ from Seer in August of 2020 before we went public. The CEO of that company is in this room in the back, Philip Ma. PrognomiQ's mission was to look at unbiased proteomics at large scale, also look at other omics as well, and to be able to develop a clinical test. They ran at the time the largest, but no more, because now there's other customers doing larger studies than them. At the time, the largest deep, unbiased proteomics study, they developed biomarkers for early detection of lung. They're on the cusp of launching their LDT in about a month or so from now.
That LDT exists only because Seer existed, just like Natera exists only because Illumina existed. That LDT for early detection of lung has the best performance across all the publicly available data for early detection of lung.
All right, great. Really interesting, especially about PrognomiQ. Just going back to the research side, do you think there's been a lot of mass spec updates over the years? ASMS has been really big announcements every couple of years. I think 2025 was a good one as well with new products and stuff. How much is the work and the new product launches, the releases that companies like Thermo, and SCIEX and Bruker are doing, how much does that kind of help you and enable Proteograph to really have that kind of use case and be truly appreciated?
Yeah. Mass specs are getting better and better. Let's put that in context. In 2020, when we filed our S-1 before we went public, at the time, the Bruker instrument was a really, really good instrument. You could get to 300 or 400 proteins on that instrument. Subsequently, the Astral came and that became really the best instrument. With the Astral, you can get to about 750 proteins in plasma. That arms race in proteomics and the mass spec will continue. Over that five-year span that I just mentioned, going from about 300, 350 to about 750, there were serial improvements that went from 350 to 450 to 550 then to 750. The value proposition of Seer, which is if you pair Seer's platform Proteograph with that mass spec, has remained the same. If you take the Astral, the Astral gives you about 750 proteins in plasma.
If you put the Proteograph in front of an Astral, now you get to about 7,000 proteins. The Astral next generation just got announced at the ASMS. Astral can now do 900 proteins. Put Seer in front of it, you're now close to 10,000 proteins. If we go back five years ago when, let's say, the Orbitrap 480 was a great mass spec, on its own it did maybe 300, 400 proteins. With the Proteograph, it would get to 2,500 proteins. The answer is, Kyle, as the technologies are improving, the baseline is improving, but the full improvement in terms of what the Proteograph adds to that has remained the same. They have not narrowed that gap. I don't expect that gap to narrow. The question is, why is that? Because they solve the fundamentally different problem that Seer solves.
What Seer does is it quantitatively compresses the dynamic range and allows for the proteome, which is complex, to go as a bite-sized unit into the mass spec. As the mass specs are getting more and more sensitive, they'll pick up more of that content. That solution is a Seer unique solution. By the way, since then, there's been some copycats that have come. None of them perform nearly as well as Seer does. I think if you live in an island by yourself, maybe you were wrong. The fact that there's a bunch of folks trying to copy what Seer does, I think is actually an additional validation of what Seer offers. We have 250+ patents in this space. I do think, Kyle, the mass spec improvements are an absolute important part of this.
Okay, yeah, good point. In terms of being cell-based and, you know, your products have been available for three years. Where are they, where have they been placed right now? Where are they being used? If you could just talk about what happened over the past few years with your stack program when you had more of a service going on, how did that kind of help drive either ultimately placements or equalization, things like that?
Yeah, so thank you, Kyle. I'm going to try to make the answer brief because it's really complicated, but I think it goes to the heart of the matter, which is when we launched Seer ., we said we are not a service business. We want to be a tools provider to our customer. We don't want to compete with our customer, which is why we spun out PrognomiQ as a diagnostic company. We didn't want to be in their business. We wanted to provide tools and be a trusted partner. The thing is, we were inventing a market that didn't exist. What do I mean by that? Mass spec market for decades has been a CapEx market. You save your money, you clunk down $500,000 or $1 million to buy an instrument, but then you don't have a consumable component that goes with that.
Seer came and said, here's an instrument. Now, every time you run a sample, you have to spend hundreds of dollars per sample. That customer was not used to that. The genomic customer is, because of, you know, decades of Illumina or 15 years of Illumina, but the proteomic guy was not. Now we're a new instrument with exactly no publication from any customer and saying, clunk down a bunch of money and do this, and we're not going to, you know, we're not in a service business. That was a mistake, I think, in hindsight in terms of commercialization strategy, because the customer doesn't know what they're getting yet. You're asking them to write a CapEx check and then to write a consumable.
When we did two things, one was the stack, as Kyle mentioned, the Seer Technology Access Center, and the second one was our Strategic Instrument Placement, SIP, p rogram, which was our loaner program. They were very important. What the stack did was the stack said, okay, you can send your samples to Seer. W e will run the sample for you and give you data, essentially like a service business. The SIP program says, okay, you don't know if this is going to be a good thing for you, we'll loan it to you, but you have to buy consumable upfront to take the loaner, and then the loaner would then convert after, you know, X number of months into a sale.
The stack ended up being a really excellent asset for us because there was a significant number of stack conversions into instrument placements, and stack lowered the barrier to entry to get the Proteograph done. Stack was also a good source of customer getting data and publishing papers. It's been tremendously helpful to us. It's been a challenging year. Last year was very hard on us, the headwind is still there, but the tailwind is getting stronger for us. For example, in the first quarter, Q1 of this year, we placed as many instruments as we did all of last year, and Q2 was about the same. In Q2, a significant number of those were largely SIPs, but we're now seeing customers adopting the technology and doing at-scale studies.
My expectation is that in the near to mid-term, and by that I mean maybe the next three years or so, stack will be about 20%-25% of our revenue. I think longer term, it would probably drop to about maybe 10%-20% of our revenue, from 20%-25% of our revenue. At scale, my expectation is that, today, if I look at today, instrument is about, sorry, consumable is about 60%-65%, stack is about 20%-25%, and the instrument, because we do a lot of loaners initially, is about 10%-15% of revenue. That shift is going to shift more toward consumable, and then a little bit of stack will go toward the instrument.
My hope is that as the customers get comfortable with the value proposition of the Proteograph and what this unbiased proteomic means for them, they'll begin to utilize the service providers that are working with us, like Evotec or Discovery Life Sciences. These are two big ones of ours, but Pamom and others are now coming online at scale. If the customer wants service, we'd rather if they work with a CRO who owns our instrument, buys our kits, but we don't want to be in the service business long term.
Got it. It sounds like there was almost like a bolus of placements that converted from the stack program, for example. It was almost like an inflated quarter in Q1 and Q2. Is that fair?
No, no, Kyle, I don't think so. I actually think the velocity of instrument placement, I think what we're now seeing is probably appropriate for the year. In other words, I don't expect there to be a slowdown. I don't know. I don't expect it to be a pickup. I mean, the macro picture is challenging, especially for CapEx, but the tailwind is getting stronger. What's driving, what's going to drive this year, and we guided at $17 million- $18 million this year, and we reiterated that on our call last week, which I admit points to about a 25% increase to last year, but that isn't because the headwind is getting easier.
I think with the tariffs and the NIH budget and the MFN of the drug pricing, all of those things, I think it's hard for our customers, but we're in small revenue numbers right now, so it doesn't take a lot to push you up. The data that the customers are getting is really important to them, and it's helping us kind of move up despite the headwinds.
Yeah, so given the macro outloo,k for a U.S. academic and then for biopharma, I guess, how is the picture setting up for like next year, for example? Because you have nice momentum, almost like inflecting, and like you said, like a 100,000 sample study maybe done on an aspect and using the Proteograph as well.
Yeah, yeah, Kyle, I think the money to be made isn't on biobanks, meaning biobanks are excellent validation in terms of your value proposition. I think the bread and butter of making money is the customer that runs, let's say, high 100s, low 1,000 sample number. You got to get into a run rate business where you have a lot of these customers that are coming. Those customers are academic, government, and biopharma customers. If I look at 2024, and I think 2025 is shaping to be about the same, academic and government were roughly about 30% of our revenue last year. It'll be about in that range this year. Government was about 18%, academic about 12%. I think that will probably shift. I look at this year, 2025, we're probably in that same 30% range, but it's heavily academic, less government this year.
Biopharma is kind of consistent, but very CapEx constrained and very conservative. What we need is a run rate business where the guy with his grants, the academic guy, has the money to buy his samples. Academics are being very cautious. Government is very cautious in terms of their spend. Biopharma is great because once they see the value proposition, it's a nice run rate business. If I look at some of our very stable customers, they're the large pharmas who've been our customer, who buy on a regular basis, but Kyle, challenging.
Got it. As you think of one financial question that I want to possibly end with, a lot of good things going on finally with the business. It's been years of investment in this. Given now like you're sort of, again, like I feel like inflecting a bit, I feel like the future is pretty bright. What is your cash runway and burn outlook looking like at this point, given maybe you have to invest more to drive that growth and the ROI is going to be positive, so you're giving those investments now maybe. How are you thinking about that kind of path?
Look, we have about $260 million of cash equivalent investments and no debt. We run the business super tight. Burn to CRO will probably be low 40s, $40 million, $45 million, and probably on the low end of that. If there was no growth at all, then you know we have six, seven years of burn. I do expect growth. My expectation is that we don't need to raise capital in the future. In fact, if anything, I feel very bullish about where we are. I think that there is a massive disconnect between Seer's value proposition and our stock price today, that we are doing a stock buyback. In fact, we had an authorization of $25 million stock buyback. We bought back about $20 million of that. That's about 13% of the float. I'm buying back stock because I feel that I have enough cash.
I don't think we're ever going to need to raise more. To win in this space, you always have to be pushing the boundaries of what is possible in terms of innovation. We are constantly investing in innovation. Our latest product launch, which we did in June, reflects that. We make highly concentrated bets in innovation. I don't have a gazillion things going on to see what sticks. We make very significant investments in certain aspects of what we think is the next big thing. We are making very significant investments in what I consider to be the next big thing. Hopefully, in the not-too-distant future, we can talk about that. My expectation, Kyle, is that access to the proteome in an unbiased way is here too.