Hello, and welcome to the Canaccord Genuity and MedTech Diagnostics Digital Health and Services Forum. My name is Alex Vukasin. I'm a member of the Life Sciences Health and Diagnostics team here at Canaccord Genuity. We are very pleased to have CEO and Chair Omid Farokhzad with us today. Omid, pleasure to have you here.
Pleasure.
Again, Seer. Really quickly on Seer. Seer offers the Proteograph Product Suite, which enables deep, unbiased proteomic analysis at scale in hours and can be incorporated by nearly any lab. I think a good place to start would be a brief update on Seer's platform and recent performance. Particularly, for those less familiar, could you just walk through the Proteograph Product Suite and how it's evolved with the launch of the Proteograph ONE and the SP200?
Awesome. Thank you for that. At Seer, our goal has always been to decode the biology that is embedded in the proteome, and by doing that, to improve human life and the quality of human life. With that vision, we launched our first product in 2021. Since then, we've iterated three times on that product. We just launched our most recent product in June of 2025 at ASMS. With it, we launched the Proteograph ONE Assay and a new instrument called the SP200. What did that enable? The combination of the Proteograph ONE and the SP200 lets customers access the proteome deep, unbiased, at speed, and robustness that was previously not possible. Importantly, at a cost point on a per-sample basis that is substantially lower than, let's say, the first iteration of the product that we launched in 2021.
Throughput increased, Assay time decreased, robustness stayed the same, and then there was a reduction in cost. What is that enabling? It is enabling the customers to do larger and larger studies. When we started Seer, shipped our first instrument, at that time point, the largest study that looked at the proteome in plasma that was ever published was 48 samples. The deepest study that looked at the plasma proteome was about 5,000 proteins. That, by the way, was from Broad Institute. Fast forward to the current year, we now have multiple customers that are now running studies in the 10,000s or tens of thousands of samples. We announced first half of the year that Korea University started a 20,000-sample study. Our colleagues at Discovery Life Sciences had a 10,000-sample study. The biobank in Singapore precised a 10,000-sample study.
While it hasn't been publicly announced because of all the government shutdown issues, I did allude to it on our Q3 earnings, so I'll share it with you, which is NIH recently funded a $50 million multi-center grant. It's a multi-omic study looking at 50,000 samples over a course of that grant, looking at a multi-omic data set to create really a benchmark of human multiome. In that context, they were looking at proteome, genome, metabolome, lipidome. Because it's a multi-center study, they wanted the data to be cross-referenced across each center. They wanted everyone to use the same platform for that large grant. They boiled the ocean. Everyone has their favorite platform, if you would, in terms of the proteome. They ran on the same samples the different platforms.
Ultimately, Seer was selected as the sole proteomic platform for that $50 million NIH grant. I was really thrilled to see that. More importantly, customer validation has been really accelerating. I mentioned that we launched the first product in 2021. In 2023, we actually saw our first customer paper get published. Since then, as of the end of Q3, 66 papers have been published. Importantly, 13 of those were just published in Q3 alone in that quarter. 13 papers got published. More than half of those 66 papers have been published since the beginning of the year. The velocity of customer publication is also increasing. The biological insight that is coming is fantastic.
I think the ultimate test is that we had spun out a company called PrognomiQ from Seer in 2020 that was focused on liquid biopsy, multi-omic, to which core to it was deep, unbiased proteomic. About a week ago, PrognomiQ had their press release that they launched their LDT, a liquid biopsy test for early detection of lung cancer. They have the best-in-class data, better than their peers at Delphi and Guardant and others. It is really a proteomic-based platform. Those biomarkers could only be discovered because they did deep study at a very, very large scale, identifying biomarkers that you would never be able to go look for. It has been a fantastic year of progress for us in 2025. I look forward to the rest of the year.
That was great. Let's take a step back and let's look at those things you just talked about. You mentioned the Proteograph ONE, that doubles weekly sample throughput while improving runtime and precision. Can you just briefly discuss what some of the early customer feedback has been on the workflow? Are there any specific sample types or study designs where the improvements are proving the most impactful right now?
Yeah. It's been a good set of feedback from the customers. Last year was a tough year for us. All of the year, we shipped 11 instruments. The first nine months of this year, we've tripled that, almost tripled that. We're just shy of 30 instruments the first nine months of the year. Importantly, 2/3 of that, roughly 30 instruments, are the SP200 and the Proteograph ONE, even though the formal launch of it happened in June at ASMS. The feedback has been great. What that has allowed the customers to do is to really do the large-scale studies. In other words, it's not a coincidence that this is the year that these population-scale studies are being announced, because Proteograph ONE really makes that possible to do tens of thousands of samples on a mass-spec-based proteomic, deep, unbiased proteomic.
By the way, the same platform today, really, with kind of modest CapEx commitment, will allow you to do hundreds of thousands of samples to do on a mass spec. Also, that's large scale. If you're interested in just biomarker discovery, we're seeing customers adopt it for neurodegenerative diseases. There's a tremendous amount of interest in neuro and also oncology. We have customers that are using it now for biomanufacturing, looking at wholesale protein in the biomanufacturing process. Importantly, we also have customers that are looking at model organisms, because these are mass-spec-based proteomics and not ligand-based. Species differences in the protein doesn't make a difference. We have customers that are doing mouse studies and monkey studies, dog studies. There's a large animal health company that's actually a good customer of ours.
They have multiple instruments that do this kind of work.
You were selected as the proteomics platform for a major NIH-funded multi-omic study. Just curious, can you share any details on how that selection occurred and what the program will entail exactly? What does it signal about broader institutional confidence in deep, unbiased proteomics?
Yeah. Look, I think the scientific community, appropriately so, is skeptical. They should be, because that's how science is done in a rigorous way. They want evidence. What the NIH did in the context of choosing the proteomic platform was, on the same set of samples, they ran competing platforms. Of course, every platform has its strengths and weaknesses. It's not that one platform dominates in every possible dimension of strength. If you're interested in discovery and you want depth and you want robustness, because remember, these are multi-year studies, what that means is that they're going to cross multiple manufacturing lots of products. They may be done at different sites under different instruments. Robustness matters a lot. Our process is ISO-certified. Our manufacturing is extremely robust. Seer is a fifth company that I started.
The other four were all therapeutic companies, all nanoparticle-based therapeutic companies. I'm used to building particles that are suitable for clinical application and in-human injection. The robustness by which Seer operates in terms of its scalability and manufacturing and quality is really differentiated. Those were very attractive. The NIH wanted to know if they start this process, that we're going to support them with the same product through the duration of the grant, or if we cross over, that we'll do crossover studies for them. Again, a really collaborative relationship that was well received by them. I'm very proud that we were selected.
Your Insights Grant Program has been gaining traction with applications more than doubling in 2025 every year. What have been the most compelling projects thus far? Importantly, how do you view this initiative strategically, whether it be customer activation or for scientific validation?
Yeah, this grant program is really interesting, because initially, when it was suggested by our marketing team, I was a bit skeptical about it personally. It turned out to be actually brilliant on their part to do it, because it does two things. In the process of the grant applications, you see a lot of what the customers want. In other words, it is the customer's wish list, if you would, in terms of the kind of study that they're doing. It's also a source of lead generation, if you would. Ultimately, because we have a high-margin consumable business, to support this study for us is actually relatively small dollars, because our business model is a razor-razor blade model.
The consumable, when we run this in our own labs as a service, is a relatively small expense for us, but a huge value add to the customer. It has been fascinating to see the kinds of work that the customers are interested in. Our selection process is quite rigorous. We engage our scientific advisory board. We guide them in terms of the therapeutic areas that we are interested in. The SAB selects them. Once the customers do those studies, it takes months for those studies to get done, data to get accumulated, et cetera. They present it at conferences. Seeing the kind of work that gets presented is awesome. For example, at the most recent American study of human genetics, there was work presented in idiopathic pulmonary fibrosis.
We're looking at the proteome of that at the level of protein isoforms, so not at the level of a total protein. Because if you look at it in a targeted way, you miss the isoform differences between protein or the variance of proteins. That the different isoforms of the protein could predict, for example, a patient who would be a rapidly progressing patient versus a stable patient. This has been a fantastic mechanism for us, number one, to connect to the customer, number two, to hear what the customer wants, and ultimately, number three, a source of lead for us, because if they don't get selected as part of the grants, we still go back to them and see if we can find a way to work with them. It has been a great effort.
Just one more on the work that's being done or could potentially be done with the Proteograph. Dr. Gloria Shankman's work on the IPF identified isoform-specific biomarkers missed by affinity-based methods. How do you see peptide-level resolution and isoform discovery shaping the clinical value of proteomics over time? Could this be a potential runway into more direct diagnostic or biomarker-driven applications?
I mean, look, I think the answer to all of those things is yes, meaning biology is complex deliberately. All of us have about 20,000 genes. But as our genes make messenger RNA, they make about 200,000 different messenger RNA from the same 20,000 genes. As those messenger RNA, a subset of them, get made into proteins, some of them don't, but many of them do, those variants of protein that are originating from the same genes further get modified through post-translational modification. We end up going from 20,000 genes to 200,000 transcripts to a million-plus different protein variants. The thing is, evolution is deliberate. Those variants of proteins have evolutionarily emerged because they have important biological function. To ignore that, you're ignoring a massive amount of biology.
To capture that, you are honing in on differences between health and disease that was otherwise not possible. When you look at proteins at the level of a total protein structure, you miss a lot of those variants. Why? Because an average human protein is 472 amino acids long. An average targeted ModE that binds to a protein binds to an epitope that is only five to eight amino acids long. That means when that targeted ModE binds to an epitope, it ignores a huge majority of that protein in terms of variation that may exist, either post-translational modification or domain differences or domain deletions or domain addition or amino acid substitution. All of that biology gets missed. These variants of proteins can have vastly different functions. One variant of a protein may activate a target. One may shut off a target.
To not study protein at the amino acid resolution or the peptide resolution leaves an enormous amount of biology untapped. Honestly, it's like if you're in the genomic world, if you stick to doing genomics on a microarray platform instead of doing a next-generation sequencing and looking at the totality of the sequences that may exist in the exomes, by the way, and also non-exomal sequences. Absolutely, the case is true. It doesn't mean that targeted approaches are bad. Not at all. These are very complementary platforms. I make the joke that I usually say, if you sit at your dinner table, you've got your fork, you've got your spoon, you've got your knife. If you pick up your fork to eat your soup, it fails miserably because that's the wrong utensil for that application. That fork has a very specific purpose if used correctly.
Targeted approaches are excellent if you know what you're looking for. No better validation of that than our own spin-out PrognomiQ that leveraged our platform to do large-scale studies for biomarker discovery, develop the best-in-class lung cancer test. Once they knew what they were looking for, they put their targeted approach. They put their LDT on a targeted approach for clinical use. That's OK, because that's exactly the right purpose of that. By the way, if I just look at the proteomic space, what has been very exciting to see is how rapidly eLemur in the neuro space has gained traction, $25 million of revenue last year going up to $70 million this year. They did the opposite of what the other targeted approaches do.
While the other ones were kind of busy in this arms race of trying to add ligands to kind of increase their panel size, they went the total opposite direction. They went 120 panel. That's it. But they looked at variants of protein that mattered, like the p217+ tau. That is what the neuro folks cared about. I think targeted approaches are excellent if you know what you're looking for. Today, in the proteomic space, we know exactly almost nothing in terms of the complexity of the proteome. When you know almost nothing, you need to embrace the total knowledge that an untargeted and unbiased approach brings. You can only do that if you're hypothesis-free and willing to look at the totality of information that comes your way. Seer absolutely uniquely delivers that.
Some great colors. Shifting to the commercial aspects of the business, your partnership with Thermo yielded its first co-sale in 3Q. Just looking at one, what does a joint Thermo-Seer customer look like? What can this relationship yield in the long term for Seer?
Yeah. First of all, Thermo has been a fantastic partner for us, our teams. My CFO, David, has a biweekly call with his counterpart at Thermo. Our technical team have monthly calls together. The relationship is fantastic. That was the basis for deepening it and doing this co-sale/co-market agreement. What that means is, as a Thermo sales rep is out there selling a mass spec, and they have the best there is, the Orbitrap of Astral and now the Astral Zoom, they can offer to that same customer also the Proteograph. Now, Astral is a fantastic instrument. Astral, if you took plasma and digested it, put it down an LC column, and then into the Astral, you would get about 750 proteins. On the Zoom, which is their latest instrument, you get about 850 or so proteins.
If you take the same exact plasma, run it on a Proteograph, then put it on the same LC column on the same instrument, you now get 6,000, 7,000, 8,000, 9,000, depending on the sample proteins. The depth that you now achieve with the combination of the Proteograph and the Thermo is about 8x-10x of what the Thermo alone delivers. That is because the Proteograph and the Thermo instrument each solve different problems. The Proteograph allows interrogation across the wide dynamic range. A mass spec instrument has a dynamic range of about four to five orders of magnitude. Human biological samples like plasma can have orders of magnitude of about 10 orders of magnitude in terms of complexity. No matter how sensitive that instrument gets, it will always grab the top four to five orders of magnitude and leaves the rest of it alone.
The better that mass spec gets, the better Seer can serve it. In fact, if you look at the ability for the Seer platform to continue to augment the iteration of the mass specs that have come over the last five years since Seer launched its instruments, the magnitude of improvement that we deliver for the mass specs has actually continued to grow, not shrink, meaning as the mass specs are getting better, we can deliver even more value to them. The Thermo reps offer that to the customers, and then they sell both. In fact, in Q3, Alex, the Thermo folks placed two instruments, which is great. In our own revenue model, we kind of put very small emphasis in terms of their contribution overall to our sales. It's nice to see that the customers are adopting both at the same time.
I think that has room to grow over the coming quarters.
Other aspects of your commercial strategy include the stack as well as the SIP programs. Can you walk us through how these programs contribute to long-term platform adoption and what the typical customer journey looks like from initial engagement to scaled deployment? What are you doing to ensure customers transition successfully from the pilot stage to recurring or expanded use as well?
Yeah. The stack and the SIP program have been fantastic, fantastic for adoption of our platform. When we launched Seer, there was exactly zero customer publications. Our reps were showing up saying, can you write a check for us for the instrument? Can you write a check for us for consumables? The customer would say, listen, I need a proof of principle study or something. We did not want to be in the services business. We started these things called Centers of Excellence or COEs. The problem is even the COEs had a hard time selling it because there was no evidence for the customers to adopt. We started a Seer Technology Access Center. We said we do not want to be in the services business, but the stack allows the customer to send samples, we process, and we give them data.
Once the customer gets comfortable with that data set and what it means, the customer could then choose. I mean, there are customers that actually wish to work on a services platform. If that's the case, we would then refer them to one of our COEs. There are customers that want to have an instrument in-house. The stack is an excellent mechanism to show a customer what the value is that they could get, and then the customer can adopt it. In fact, if I look at Q3, now it's a little bit biased because that happened to be a pretty significant quarter in that way. In Q3, 2/3 of the instrument placements were former stack customers. I think if I look back more, it's closer to maybe half or a little bit below half. Stack has been great. That is the stack.
What is SIP? SIP was Strategic Instrument Placement program. SIP essentially is like a loaner program, which we did not have initially. About 40% of our customers are academic and government customers. These folks usually do not have a CapEx budget to write you a check. The list price of the instrument is $175,000. What the SIP did was allow the customer to borrow the instrument as a loaner with a clear contract where it would end in the expectation that they would buy it, but we would not commit it to them. In order to accept a loaner, they would have to buy a reasonable amount of consumable kits so that we are reassured that the customer is going to use the instrument and get familiarity with it.
SIP has been fantastic because as the SIPs are getting to maturity, the customers will then kind of adopt and buy them. If I look at our placements, let's say this first half of this year, SIP and sales placement are about 50/50. Half SIP placement, half just purchases. Remember, a SIP is just a delayed purchase, if you would. The number of times that a SIP comes back is actually much less frequent than the number of times that a SIP gets purchased. The SIP program, it's been in existence for just a couple of years. We're essentially one year into SIPs coming to maturity. The SIPs that we're placing now will see those sales a year down the road.
SIP has been a great program because then the customers can write grants and they can adopt the system when they have the CapEx budget.
I think we have time for just one or two more. You ended Q3 with about $251 million in cash. You reaffirmed break-even runway. Can you speak about capital allocation priorities for 2026? Also, recognizing the multi-pronged strategy you've implemented in order to expand adoption levels in a negative CapEx environment, could you potentially leverage your sizable cash position to acquire some less capital-intensive technology?
Yeah.
To support the Proteograph?
Look, I think we see two things. One is I like to think we are an extremely diligent steward of the capital that's been entrusted in us at Seer by our investors. Despite our strong balance sheet, we've gone through four rifts in the last 18 months. That's because my revenue expectation was much sharper than what it has been. I mean, we've guided the street $17 million-$18 million this year, up from $14 million and change last year. At the midpoint of the guide, it's about a 24% growth. 24% growth is not my definition of a growth company. In other words, I think when the flywheel kicks in, when the biological insight becomes obvious to the customer, the growth and the velocity will pick up to what I consider is the normal expectation of what I would have in terms of our revenue growth.
Until then, we're going to be super diligent about burn. Now, that said, we are investing in innovation because I think the edge that we have over our competitors is the velocity and the pace by which we innovate. We invented this space. Prior to Seer, the notion of protein enrichment for mass spec to do with scalability did not exist. That is a field that Seer invented. It's a field that Seer has 250+ patents in. By the way, imitation is probably one of the best ways to flatter someone. The fact that there's now copycat products that are coming is great. IP aside, those are inferior products. Why? Because as they are coming, the next iteration of our products is already in development and is already coming.
What's beautiful is our customers are publishing papers showing that those inferior products perform significantly below par relative to Seer. Also, the manufacturing and robustness and all of that. These are areas that we're focused on. Now, we need to get to a company of scale in terms of our revenue. To me, that number is somewhere around $100 million. We have plenty of cash on our balance sheet, and I have so much incredible trust and belief in our own platform to know that we are more than enough cash on our balance sheet to get to that point of profitability. If we can accelerate that by bringing that timeline closer, investing in some great technologies that can help us shift that slope and get there faster in terms of becoming cash flow positive, we will do that.
Now, the beauty of the balance sheet that we have now is that we get calls from bankers all the time to look at a variety of things. We have looked at so many things in the last 18 months. None have triggered us to pull the trigger. We are continuing to look to see if there are areas that we could leverage in terms of our ability, our customer base, our knowledge synergistically with another to be serving our customer better. We are always looking.
That's great comment . One last one to end on, I think, would be helpful. Any upcoming clinical collaborations, new application areas, or product development milestones that we should keep in our radar for 2026 and beyond?
I think just to do two things. One is keep an eye on the scalability of the project because I think they are going to be increasing. I had predicted at the last Morgan Stanley conference that 2025 will be the year where pop studies, population scale mass spec studies, will happen. They did. I had predicted that they would happen by mid-year, but it actually happened the first half of the year. It was, in fact, in your last conference that I had predicted that 2026 will be the year where the first 100,000 sample study on a mass spec will get done. I stick to that commitment, which I think is going to be, I think if I look at the dialogues that are happening, the conversations are now approaching tens of thousands of samples.
By the way, people are now viewing tens of thousands of samples as, quote, "pilot project" to doing much larger studies. The fact that the conversations have taken that tone is fantastically rewarding. What needs to happen next is the biological insight from these findings need to come. That is then what pushes the revenue to emerge.
It's a great place to end. Thank you very much for your time.
Thank you, Alex.