Good morning and warm welcome to Olink Proteomics inaugural Investor Day. We are very excited to be here and hope that today will be both fruitful and fun. Thank you all for attending. I'm not gonna read the disclaimer, but I need to share it. The only thing I want to highlight is that the KOLs are not paid to participate and do so from free will to help educate Wall Street on proteomics, which I think is quite gracious of them. Starting off with a brief walkthrough of today's agenda, where I will kick things off with a short introduction. Following that, I will host fireside chat with Rob Gerszten. Ida Grundberg, our Chief Scientific Officer, will host fireside chats with Arnav Mehta and Emma Guttman. Then I will speak to Ferhan Qureshi and Chris Whelan. Carl Raimond, our Chief Commercial Officer, will provide you with a commercial update. We will wrap up today with a Q&A. If you have any questions, please feel free to share them in the chat. Let's dive into things. We're starting with a quick recap of who we are and what we do. We are a market-leading proteomics company founded back in 2016. We're rapidly growing with roughly 70% CAGR since inception. We see a strong desire to add proteins to the mix of other omics. We have the world's best technology that has overcome most, if not all, of the complex challenges in proteomics. We use NGS and qPCR for various use cases, and we will dive into this a bit later today. Proteomics and Olink is happening today. Major biopharma, academic institutions, and thought leaders use our platform globally. They do so across all stages of research, basic research, clinical development, and now also in clinical decision-making in the U.S. The market is very large and expected to be larger than genomics. Proteomics will be run longitudinally, meaning that you will test patients over time since proteins are dynamic over time. Our fastest-growing segment is highplex, driven by our NGS platform Explore, but also extremely exciting opportunities in mid and lowplex, which Carl will talk to you about later today. We continue to deliver across all points in the strategic plan that we communicated. We expect to deliver on the revenue guidance we communicated early this year with yet again roughly 70% CAGR and revenues ranging from $90 million-$92 million. Explore is most important growth driver and now representing more than 55% of our business, basically in just a year since launch. Extremely impressive. Carl will also share more details around the successful and exciting Signature launch and commercial wrap-up, and we continue to be very successful in rapidly growing our library of assays and expanding our R&D capabilities to make the most of our disruptive PEA technology. Today is a day where we'll mostly focus on the science and the customer's perspective on proteomics and Olink. We want to bring you under the skin of proteomics. Hence, I wanted to tee it up with a quick glance of the whys and hows our customers use our products. This slide we use in basically all customer meetings, a great door opener where we check off what we call the needs assessment, i.e. understanding the challenges our customers are seeking solutions for, and proteomics is a great strategy. Starting at 12 o'clock, mouse to human models in drug development are yesterday's news. Human biology is a better starting point than mice to develop drugs for humans. Here, one combines genomics with proteomics to identify novel and causal drug targets. This is actually one of the major reasons for the UK Biobank proteomics project, which we will talk more about later. A better understanding of biology, pathophysiology, and how drugs interact with that biology is one of the most prominent use cases as well. Prediction of disease and disease outcome or prediction of drug response is also one of the major use cases we see. A great example is one we shared in last week's earnings call, where researchers used Olink to predict mortality risks in cardiovascular disease to identify patients with highest risk and treat them much more aggressively to avoid very serious outcomes. Using the fruits from, for example, prediction of drug response, we can move that into clinical development through patient stratification. Here provides an opportunity to enroll a higher rate of potential responders to a new drug. That will increase the power in a study, which means that the N or the number of patients will go down. Cost, risk, and time to market will go down. This is how modern drugs are developed. What we also try to do is to identify surrogate markers for both safety and efficacy to, as early as possible in clinical development, assess these very important factors to potentially change a dose or take patients off a certain treatment, et cetera. Last but not least, early diagnostics, where we really see a big need to find disease earlier or actually, where we really try to be predicting. You'd rather be finding disease progression much earlier to be able to prevent that. Okay. Personally, I don't think that I've been in any customer meeting where the research question don't fall into one or actually several of these use cases. Another dimension is but equally important is the scalability and actionability. Scientists are not using these experiments for fun and really want to use it to drive their projects or improve patient outcomes. Starting off with casting a very broad net of a combination of known and novel biomarkers in discovery, followed by verification of findings in a secondary cohort. This part of science is mostly done on retrospective sample collections. Next step in development is prospective validation of clinical utility before moving into clinical decision-making. No other technology or platform other than Olink offers this complete, scalable, and actionable solution. We also believe that modern research is done in a very transparent way versus black box thinking and share all the details of our performance in an open access format on our webpage on every assay. To study these critical aspects of health and disease and how drugs interact with real-time human biology with a proteomic strategy, you need to cover all key performance criteria across all assays. Specificity, sensitivity, dynamic range, precision, and scalability. Here, we really pride ourselves as the only platform in the highplex space that performs an extremely robust analytical validation of basically every assay. We strongly believe that this forms the foundation of great science, highest data quality, and trust why customers choose to work with our platform versus others. To understand the details of every assay, you need to develop a so-called standard curve for every protein target. You run the assay in the presence of the antigen across certain predefined dilution steps. This is where you define the key properties of every assay. Here you have in the bottom of the slide, the predefined dilution steps. In protein research, you will always have some background noise, and that's why it's very important to set the limit of detection or lower limits of quantification. Then you also define the upper limit of quantification where you still have robust response from the assay. It's very. The difference between the upper and lower limits of quantification defines the dynamic range of the assay. This is also very important that it covers the dynamic range from potentially very low numbers in a healthy individual versus extremely high concentrations in a severely ill patient. A great example is troponin, where most of us have very, very low concentrations, but when you have a heart attack, you will have very high levels of troponin, and the assay for troponin needs to cover this very broad dynamic range. On the y-axis, you have the unit NPX. So it's normalized protein expression, which is the unit we at Olink use for the readout of our assays. How you use the standard curve is also to identify the concentration of a certain protein in a sample. At a certain point on the NPX curve here, you can follow out to the dynamic range and see where you're at here in lower picogram levels per milliliter. What we are providing through this is actually protein research at a massive scale, proteomics. Without knowing that it's the correct protein you're looking at, so specificity, or without a robust validation of each assay performance at this level, I would claim that it's impossible to understand real-time human proteomics and challenge every provider to provide the same robust validation to support the best possible science in proteomics and to do the same. This forms the foundation of solid research. This is where we excel, and in my mind, one of the most important points as to why we win and customers choose to work with our platform. To cater to all of the use cases I discussed initially, and also the dimension of scalability and actionability, we have very carefully designed our product portfolio. Starting with Explore, using NGS as a readout, where you want to cast a very broad net of a combination of known and novel biomarkers in extreme high throughput and in a very cost-efficient manner. In Target, you follow up with more targeted research questions, often closer to the clinical or later stages of research, when you, for example, know which pathways are involved with a certain disease or which pathways your drug is interacting with. Last but not least is Focus, where we custom develop each product and protein signature for a specific customer or use case. No other company or technology support research across all of these dimensions and truly makes Olink unique, and something that we know is very important to our customers. With that in mind, we have designed today's investor day to present the opportunity to talk with thought leaders across all aspects of our product portfolio, with an emphasis on Explore. Firstly, we will talk to Professor Gerszten, followed by Arnav Mehta for Explore. Emma Guttman-Yassky will join us, more focused on Target, and then Ferhan Qureshi from Octave Bioscience, on Focus. To wrap things up, we will talk to Chris Whelan, who is heading up the Pharma Proteomics Project, executing on the UK Biobank project. Now it's time to start these very exciting fireside chats, and by that I would like to also introduce Olink's Chief Scientific Officer, Ida Grundberg, who will be supporting these discussions with me. Very welcome, Ida. Great to have you here. How are you? I'm great, Jon. Thanks for having me. I couldn't miss this. I'm looking forward to these exciting talks. I should say, I'm also very proud to have this prominent list of influential scientists join us today and to hear from them their view and their experience working with us. I can't wait. I think we should get started. Let's kick things off. So let me first introduce Professor Robert Gerszten. Rob is the Chief of Cardiovascular Medicine at Beth Israel Deaconess Medical Center in Boston. He's the Herman Dana Professor of Medicine at Harvard Medical School, and also a Senior Associate member of the Broad Institute. Hey, good morning, Rob. Great to have you here. Warm welcome. Rob, your career is truly impressive with lots of high impact publications and or you're considered a very strong and important thought leader in cardiovascular biomarker research. Can you please briefly share with us the goal of your research and what you hope to accomplish? Perhaps you know share a few use cases that you use. Sure. Thank you very much, and it's a pleasure to be here. By way of background, I'm a physician, a scientist, and for better or for worse, I'm also now an administrator as well. I did my medical school training at Johns Hopkins, went out to do research at University of California, San Francisco, and then came back to complete my cardiology training at the MGH, where I joined the faculty in the mid to late 1990s. I still remain very active clinically, serving as an attending physician in our coronary care unit. I really believe that a lot of our work has as its springboard our clinical interactions or how we interpret our work is really influenced by the clinical questions. How did I get into Olink and these sort of proteomic profiling technology? You know, over time in my research career, I have to say, I became tired of studying other people's molecules, if you will, you know, things that other people discovered and thought, you know, maybe if we could begin to understand how these new blood profiling technologies for metabolites and proteins work, you know, we might be able to identify new disease pathways and new disease biomarkers as well. We developed this interactive program where we apply these types of technologies to work in collaboration with the Broad Institute, the Framingham Heart Study, the Jackson Heart Study, the Diabetes Prevention Program. As maybe we'll talk about a little bit sort of we're a core for the TOPMed program, which is the personalized genomics program at the National Heart, Lung, and Blood Institute. Okay. What we really do, John and others, is we kinda make observations in large human populations, and then the goal is to go back to test for functional associations in model systems. The work kinda has two goals, right? One is we make these observations, do we have new disease biomarkers? And then number two, have we identified any new pathways? I think you asked, you know, can I give one kind of example from our work. About a decade ago, in a paper published in Nature Medicine, we identified a profile of amino acids, particularly, if you will, greasy amino acids, branched chain and aromatic amino acids whose levels are really elevated over a decade before you develop diabetes. Even if you adjust for all our standard clinical risk factors, what somebody weighs, what their blood glucose is, what their hemoglobin A1C is, for example, these are better predictors. We have these new disease markers. Okay, that's one area of translation. Can we integrate these or the subject of today's discussion is protein biomarkers for better disease prediction and, you know, better precision medicine. At the same time, we're very interested in the genes that regulate these sort of circulating factors, right? Let's say you see something in the blood. It could be that you went to McDonald's, you ate that, or maybe in this audience, people don't go to McDonald's, but, you know, you ate something and your blood sugar went up, and it was really because of dietary interactions, or it could be genes. As it turns out, we and others have now found genes that regulate these circulating factors, and they in turn are associated with diabetes. We've done this sort of triangulation experiment, if you will, or what's known as a Mendelian randomization experiment, to identify not just a disease marker, but genes that are contributing to the disease itself. We're not saying there aren't many ways of getting to diabetes, but these circulating factors that the genes that modulate the levels are independently associated with atherosclerosis or cardiometabolic disease. They are in a pathway that we're quite interested in. That's, like, kind of a long-winded introduction in the general type of work that we're interested in doing. No, that's great. Thank you very much, Rob. So, if I understand correctly, I mean, you mentioned genomics here as well, but I think your work is mostly focused on proteomics and metabolomics versus, say, genomics or transcriptomics. Is that correct or if so- Yes. Could you sort of elaborate on that? That's a fair statement. For example, we're interested in the molecular signals by which exercise, for example, confers its beneficial effects. If you're going to profile somebody before and after exercise at the genetic level, well, their genes aren't going to change. I mean, this is a sophisticated group. Genes make messenger RNA. There's all kinds of modifications, and then they make proteins, which in turn make those that are enzymes that make metabolites. You know, the proteins and the metabolites are most proximate, if you will, to any given phenotype. If you're interested in an acute exercise perturbation and how it might be signaling throughout the body in a hormone-like way, you know, how, maybe how does muscle transmit signals to other organs, to the fat or vice versa, well, that has to be done at a it has to be interrogated at a proteomic or a metabolomic level. We are particularly interested because they're right next to any disease entity. In addition, though, we certainly are interested in the genetics because they tell us a little bit more sometimes about the pathways. We're far more interested in the more distal end of the paradigm of human biology. Cool. Great. You mentioned a couple of diseases here, atherosclerosis, a difficult word for a Swede, and diabetes. How do you think, you know, proteomics and you also talk about this sort of, you know, really trying to help patients out. From your perspective on the proteomics work that you do today, how do you see it have an impact in a more clinical setting? If so, what in from your perspective, the greatest impact? Yeah. We've described, we and now many other groups have described a number of different markers that presage the onset of diabetes and these or heart disease in large population-based studies. That's an important first step for sure. Really for these to have clinical impact, I think that they have to really modulate our decision-making process, i.e., is there an individual who is going to benefit more from drug A than drug B? Again, I realize this is a sophisticated audience, but that's really when we're going to adopt these and when I think we're going to make the biggest impact. Now, at the same time, I also believe that by understanding groups of proteins that are modified in any disease state, it's going to get us to primordial pathways that might serve as the targets for drugs as well. Again, adding value on the biomarker side, but in the question of interaction with therapy and then in terms of identifying novel pathways. Well, what I will drive home in terms of atherosclerosis is, you know, we have lots of drugs that work quite well, for sure, our statins and many other drugs, and our blood pressure controlling drugs such as ACE inhibitors, et cetera. That said, even if you account for all of those, there's a huge amount of what we call residual disease, and that's the disease that we're going after, right? The non-cholesterol disease that we'd like to prevent. Great. For listeners here, we briefly talked a little bit about the use cases. You might have heard. There are novel pathways to understanding biology, prediction of sort of severity and even patient stratification. We checked a few off just here. We're gonna turn into something that I think many of the audience are quite interested and excited about, Rob. At your lab in Harvard, we know you're running both the SomaLogic and the Olink platform for proteomic analysis. But we should also say that you have a broad experience with MassSpec, and I believe there is a big interest here to hear how you compare and contrast between these three different proteomics technologies. Right. First off, Jon, maybe before we get started, I can say that I receive no resources from anyone at a Mass Spec or an immunoaffinity company at all. You know, we work in collaboration for sure. I just want that very clear to the audience. Those who know me well will agree that I will occasionally accept a little bit of sushi or a beer, but that's where the limit is. Let's start now by, like, parsing this discussion that I think about, I don't know, on a daily basis. Why don't we start dividing the Mass Spec versus the immunoaffinity platform? Mass Spec is really, you know, unparalleled in its ability to give you molecular characterization of some species, for sure. I mean, it's really remarkable. That said, I am a firm believer that we are still probably two major quantum leaps away from being able to use Mass Spec for the types of questions that I'm interested in, understanding at a population base who's gonna develop disease or a group of individuals that I wanna segment in terms of who's gonna benefit from a drug therapy. The throughput is simply not there. I know that there are companies that are coming out that propose to have really solid, better throughput systems. I've tried them with collaborators, and I'm in one experiment right now. The types of technologies that, say, Olink offers or SomaLogic offers are, to my eye, still probably one and a half orders of magnitude faster for sure. That's my 10,000-foot view of affinity versus Mass Spec. I just don't think you can do what we wanna do with mass spec. I'm not saying mass spec isn't a great tool for many other questions or, and it's also a great tool to verify the specificity of a lot of these reagents. Now, let's talk about Olink and SomaLogic. So we use both platforms, for sure, and we think that both platforms, as I said previously, are a quantum leap over what's available by mass spec. Most recently, we put in a large contract to the National Institutes of Health, and we had to kinda put our chips down on one or the other. Overall, we decided to put our chips down, and we got this large contract, we put it on the Olink platform. For the reason being that we believe that with 2 reagents to every single target, that you're enhancing specificity of what you're doing for sure. That's really what gives us the gut reaction to do that. We think that you guys, this company is a little bit more, you know, into the mode of democratization of these tools. You've been like that, right? I mean, you don't have an arm that's using it. You may have, you know, looking at some IP around your technologies and whatnot, but really your goal is to kinda democratize these tools. That's why we leaned in this large study towards Olink, okay? That's why particularly since we're working with a lot of population epidemiologists who don't have the tools to then further work on downstream validation of these signals. Great. That's my 10,000-foot view. Oh, great. That's super exciting. Investors like, you know, even further clarity, perhaps. Can you share with us, like, how you think about how many samples you will run on the Olink platform versus the SomaLogic platform, say, in the next couple of years? Yeah. We are, I think we're a bit of an anomaly out there in terms of, you know, luck that we've had in terms of getting large contracts from the NIH and grants as well. Our own group is probably gonna be running now on the order of over 30,000 samples over the next few years on the Olink platform. We're running fewer on SomaLogic. Now, partially that relates to the fact that SomaLogic really spent several years kind of changing their business model, and that wasn't kinda compatible with the type of science that we wanna do. We want free science where we can, you know, generate data, and we have to, by mandates of the NIH, share it almost immediately. We wanted to perform all the assays in our own laboratory, and so that we, you know, have control over the technologies, you know. That's really the driving force behind, you know, why we're running so many samples right now on the Olink platform. Great. Obviously, we are excited to hear that, Rob. So you earlier mentioned the grant or the contract that you were provided with from TOPMed or NIH and the TOPMed, and in particular around the MESA cohort, which I think you're gonna run roughly 20,000 samples. Yeah. This is obviously a very, very significant project. I think it would be quite interesting for the audience to hear a little bit more about that contract. Yeah. the study itself. Yeah. The MESA study is the Multi-Ethnic Study of Atherosclerosis, and it's really a study that reflects the heterogeneity of individuals in the United States. It's really a wonderful study. It turns out that there are about 5,000 or so individuals in each that are overall in the study or a little bit more than that. What's neat about this study is that it's at multiple time points, so that we'll have serial samplings in a trajectory analysis. There are other studies that are starting to do that, but this will certainly be one of them for sure. This will be the largest, I believe, Olink study that has serial samples. You get an idea of not just can we predict at one time point who's gonna develop diabetes or heart disease or other phenotypes, but also can we look at the trajectory of the changes of these proteins over time. It turns out this is a particularly well-phenotyped group of individuals, such that we're not just gonna have, you know, other disease biomarkers, but there'll be imaging studies assessing the heart calcification or in the vessels, et cetera. It's a really richly phenotyped sample across multiple time points that we'll be studying. That's the largest constituent, if you will, of this contract that we just received. There are other cohorts as well that are gonna be studied. No, this is super exciting. I think I saw a reference of already now, like 2000 publications on this MESA cohort. It's obviously a very, very well-studied population or cohort. I know that, because we worked on this already last year with a pilot study and there was, you know, the NHLBI did one with SomaLogic for this particular MESA cohort, and the reviewers, you know, obviously favored Olink here. Maybe you already talked about it, or do you know sort of what their thoughts were and why they leaned the Olink way? You know, I had nothing to do with the review. That's for sure. We stated our case that we thought that the specificity was better, that with the 3,000-plex, we're starting to get pretty close to where the total number of assays that are in the SomaLogic, but that the individuals that were participating in this collaborative group would probably want to be more sure of potentially fewer things than not so sure of a larger screen where there's less specificity. You know, those are hunches. That's, you know, that's what we set forth, and that's what was funded. Look, again, I also think that Olink's commitment to really kinda get the data out there, to get the technologies out there, to let people do the assays, you know, that is that resonates with a lot of scientists for sure. Great. We're very proud to represent that for sure. As we chatted a bit, we talked a bit about the UK Biobank project, which is pretty significant as well. Sure. Can you sort of share with the audience your thoughts around that and also the transparency in the data and so forth, and what that will mean for Olink in particular? Oh, yeah. I mean, I think the UK Biobank teaming up with your group to generate proteomics data is really an extraordinary movement that's positive for the entire field. I think the UK Biobank has really been the gold standard for integration of genetics and phenotypic information, and now part of that phenotypic information will be proteomic information. I'll say this, there are some limitations to it. It's gonna be a huge project that you guys are doing, but I, you know, I'll give you my angle as to where we're gonna fill it in, my own group. Number one is, to the best of my knowledge, the work that's ongoing in the UK Biobank is dedicated to analyses in individuals of Northern European descent. I certainly don't wanna offend my Swedish colleagues, but the world is increasingly heterogeneous for sure. I think it's important that we look across individuals of diverse ancestries, and I know that you guys agree with that as well. In fact, if you look at genetic risk scores in individuals, if these scores are tuned in individuals of Northern European ancestry, then they work very poorly in other populations. That's one thing. I think that looking at heterogeneous populations is gonna be important, okay? I think that there's ancestry-specific questions that can inform genetic discovery as well that many are familiar with. Finally, in the MESA study and in other studies, there's gonna be multiple follow-ups in these individuals. I You know, we're just beginning to wrap our arms around this knowledge of, okay, this is what your profile looks like now, but how does that change again? What's that trajectory over time? Is that trajectory immutable? Does it predict disease? I mean, there's a tremendous amount of work that needs to be done. Olink, the Olink UKBB project, spectacular. Lots of other things to do. Our own angle on this in my own laboratory is, you know, to really also focus on perturbational experiments in humans. Yeah. You know, we've been lucky. We've gotten grants to do large studies like the Framingham Heart Study, which is sort of the grandfather, if you will, of cardiovascular and maybe all of the epidemiological studies. We like to go back also and do perturbational studies. For example, exercise. We also are the core, and we'll develop Olink data on about 3,000, or a little bit less than 3,000 individuals who undergo an exercise intervention as part of the MoTrPAC study. MoTrPAC stands for Molecular Transducers of Physical Activity. It's another $180 million NIH study. Mm-hmm. We're certainly not getting that much, but we're getting a reasonable amount to profile those individuals. We have many exercise cohorts in our own laboratory. We also do many other studies, you know, before and after, a feeding study, you know. Mm-hmm. A mixed meal tolerance test. We love to look because we think that those kind of stressors will illuminate pathways that are relevant to human disease biology. Then we want to integrate the information from those provocative studies with the large clinical studies as well. Mm-hmm. I think that there are many studies that are, you know, possible in the future. You know, we're not at a loss as are many other groups for thinking about ideas where these types of technologies, like Olink and other proteomic technologies, will be highly applicable. UK Biobank will be an anchor because it's so large for some of the findings, right? UK Biobank is large, but they can't do all the phenotyping that's available in all these other studies. That is great. The word spectacular, I think, is a great way to round things off. Rob, we're super appreciative. There's lots of more questions here, both from us and the audience. We'll have to have a new Investor Day and spend more time with you. Unfortunately, we have to round things off and move on here. Yet again, thank you so much for participating today, and we'll talk soon again for sure. All righty. Thank you very much. Okay, great. Ida, please. Great. It's time to meet Dr. Arnav Mehta. It's still early days in his career, but this rising star has already accomplished so much with a PhD from Caltech and MD from Harvard. Arnav is currently doing a postdoc at the Broad Institute of MIT and Harvard at Eric Lander and Nir Hacohen's group, and a clinical fellow in Hematology Oncology at the Dana-Farber Cancer Institute. Let's welcome Arnav. Hi, Arnav. Hi. Thanks. Nice to see you, Ida. It's great to have you here. I'm gonna start then. As I understand it, you as a physician scientist, you are interested in biology, cancer biology, immunology, single cell genomics and mathematics to better understand real-time biology and how it applies to discoveries of new cancer therapies. Can you, in your own words, describe or give a background to your research and what you hope to accomplish? Yeah, absolutely. We're very much interested in how, you know, the immune system behaves within tumors, and in particular, trying to understand what the mechanisms behind immune response are, and especially immune resistance. To do that, you know, we have quite an infrastructure, and building a GI immunology program at MGH. But looking in Nir's lab, we look across a multitude of different tumors, whereby we can serial sample blood samples, so you know, peripheral blood lymphocytes as well as plasma, as well as collect serial tissue biopsies across a whole range of clinical trials, so that we can look before and on treatment, while patients are receiving different immunotherapy combinations. Our goal here is to combine a range of different approaches from single-cell RNA sequencing and single-nucleus RNA sequencing, where we've spent a lot of time optimizing protocols, to spatial transcriptomics and spatial proteomics on the tumor biopsies themselves, and then leverage peripheral studies so to look at plasma proteomics, and also markers on peripheral blood lymphocytes to understand the mechanisms behind immune immunotherapy resistance and response in a whole host of different tumors. Great. That was a great segue into my second question. You already mentioned that you apply different type of omics and also different sample types, plasma, tissues and others. Where do you see the value of the emerging field of proteomics and maybe in particular plasma proteomics, why that is important, and also in perspective to the other omics that you mentioned? Yeah, absolutely. It's a great question. I view it as twofold. You know, there's a whole arm of our interest that's interested in biomarker discovery. There, you know, the criteria by which we go by is that we want to be able to easily assay accessible tissue. You know, patients who come to see us in clinic are getting blood draws every visit, and getting a little bit of extra blood is minimal hassle to them, right? When you can essentially liquid biopsy the, you know, patients and collect a range of samples, it enables high resolution biomarker studies. The other important part of that is that you can do the sampling at very high frequencies and in very narrow time windows. That enables a very close investigation of the dynamics of, you know, patient responses. The other aspect of this is really about studying, you know, underlying biology and mechanisms. That's where, you know, you can get multiple views on what's happening. You know, a lot of cancer immunology is focused around looking at the tissue itself, which is fundamentally important and critical. It gives you information about what's happening within the tumor microenvironment and how therapies are affecting the tumor microenvironment. What it doesn't give you is a very high level, you know, or a broad view of how the host is responding to the tumor globally and how the, you know, and how the tumor is affecting the host in sort of a global stage. That I do believe there's a lot of information lost in looking just at the tumor itself, which is why even in the context of biological discovery, I think looking at the peripheral compartment is particularly important. So when you look at the approaches we have to look in the periphery, you have circulating tumor DNA, which is often used clinically but only gives a very tumor-intrinsic view of what's happening. You can look at peripheral blood lymphocytes. I mean, we do a lot of TCR sequencing, but again, that gives you a very immune-intrinsic view. A really nice, you know, advantage of doing plasma proteomics is that you actually get a sense of how the tumor and the immune cells are interacting, right? You get a holistic picture of the multicellular interactions that are happening in the body. Again, with the downside that you have, you know, a problem of deconvolution of those signals, which is why combining proteomics with other modalities is also very beneficial. Great. Great answer. Arnav, we have collaborated with you and your respective teams on two very important projects, and the data has been shared with the investor community. We have the melanoma project, therapy response to melanoma and COVID-19. If we start with the melanoma project, maybe as a background, can you describe how these patients are treated today and also how the immune system plays in. Absolutely. Yeah. You know, tumor immunology is fundamentally important to how we think about melanoma. Both in the localized setting, where the disease is potentially curable, we treat patients with immunotherapy, adjuvant immunotherapy, after they've had surgery. Then in the metastatic setting, melanoma's really led the way in establishing immunotherapy as a paradigm. First line treatment for any patient with melanoma, whether or not they have a targetable BRAF mutation, at least, you know, in the way we practice, is always to treat with immunotherapy. We tend to give single agent anti-PD-1 therapy for most all patients, but in rare cases, such as if the patient has brain tumors, we treat with dual checkpoint blockade. That's to say, you know, understanding, you know, why and how patients respond and which patients develop resistance is an essential part of, you know, taking the melanoma field forward, as well. Great. As well. Jon was just nodding here, following up to where he started with the therapy response and responders, non-responders. That was great. In short, can you describe the goal of the project that we collaborated on and also what you see as the greatest accomplishment so far? One of the limitations I brought up, you know, with doing biopsies and getting biopsies, especially in these cancer patients, is that it's often hard to get serial biopsies. In an effort to set out to discover biomarkers that are associated with response and non-response over time, in addition to being able to sort of biopsy a much larger cohort, we, you know, established this relationship with Olink to do plasma proteomics in a cohort of over 200 patients. Here, in our cohort, we had plasma samples available. We analyzed plasma samples both at baseline, then at 6 weeks and then 6 months on anti-PD-1 therapy. Really, you know, two fundamental goals. The first being, can we use early time points to predict which patients are responding? Secondly, with the secondary goal of understanding if we could predict which patients would develop toxicity, because that is a fundamental limitation of how, you know, we can treat patients with immunotherapy, and oftentimes patients have to stop treatment for that reason. Secondly, to understand how we can use this plasma proteomic data in the context of all of the other tissue data that we have to deconvolve the signals in the plasma. In essence, then, understand how future efforts of using plasma proteomics can inform our biological understanding of how the disease is responding or resisting. I'll highlight a couple of things that, you know, I think have worked very well for us. The first is that, you know, we saw very clear signals between responders and non-responders, and we identified a co-expressed group of proteins or an entire module of proteins that is enriched in a non-responder subgroup. This is important because, you know, the scale of the number of proteins that we were able to assay enabled this, which, you know, historically using much more limited assays like Luminex, you know, folks have not been able to do. Secondly, we were able, because we had paired single-cell data from tumors from a lot of these patients, to take that set of proteins and then deconvolve which cell types were likely contributing to the proteins in circulation. We dissected this module into components that were coming most likely from tumor cells versus most likely from myeloid cells. We're able to use this co-regulated module to identify a very unique subset of myeloid cells that seem to be driving immune resistance in a set of immunotherapy non-responders. In essence, the plasma proteomics took us not just in terms of biomarkers and prediction, which our non-response module does a really good job of predicting non-responders, but also took us to fundamental biological discovery, where now we have cell types that seem to be expressing this non-response module that we can then investigate as a mechanism for non-response. Great. Well, it was great last sentence there, that you both could use the data for prediction of therapy response as well as giving more insights into the tumor biology. That was great. I know we are now expanding this study with the Explore 3000 project. What do you think? What are your hopes with this expansion? Yeah. Short. Yeah, absolutely. I think, you know, in terms of our predictive capability, we're actually, you know. If you look traditionally in melanoma, most of the predictive models leverage tissue information like Tumor Mutational Burden, PD-L1 expression, a number of other genetic and immune-related signatures. But using plasma proteomics, we're actually able to do better than that, which is honestly, I was surprised to see how well we were able to do. Our cohort was well designed with training and validation cohorts. With the larger panel, I think we can yet do better. I think that's really important, because I think right now, even with the Explore 1536 panel, the set of proteins is biased, right? Based on the things that have been of general interest. I think a more unbiased look at a range of proteins, you know, outside of the immunology, neurological or cardiac space might actually further inform our predictive models because it's clear that, you know, there are multiple cell types contributing to the plasma proteome. I mean, I think we would also be able to better define signatures that are more tumor specific versus myeloid specific versus other cell types. I think we can do much better at doing that. I'm particularly excited for that. Secondly, I think we can leverage the full spectrum of 3,000 proteins to create much better signatures of organ toxicity. Mm-hmm. which I think is going to be very helpful in refining our models for immune toxicity with checkpoint blockade. I think that's been a real limitation in the field as well. One, because you know, immune-related toxicities tend to be very heterogeneous. As such, I think a larger spectrum of proteins will enable our predictive abilities. Great. Very exciting. Looking forward to that data. Last question on the melanoma project. How do you see, maybe especially you as a clinician, that these results potentially can impact how drugs are developed, administered and treated in the future? Yeah, absolutely. I think the most important thing is that, you know, looking at peripheral biomarkers and proteomics enables patient stratification in a way that we can't otherwise do with tissue. It enables dynamic patient stratification, and that you can biopsy at serial time points. You can envision, you know, designing clinical trials that are focused around the specific subsets of patients that are more likely to respond versus those that have expressed this in non-responder module, for example, that we found in our first dataset. In that way, you know, you can focus or tailor your therapies to particular subsets of patients. Secondly, I think there's you know huge potential for biological discovery outside of just melanoma. So that's something to think about. I think doing plasma proteomics is much more accessible than a lot of other tissue approaches. As we think about other clinical trial strategies, you know, at least in our labs, we're incorporating plasma proteomics into pretty much all of our clinical trials. Great to hear. If you then go to the COVID-19 project, can you give a background to what the goals for the study were and also maybe highlight the most important findings so far? Yeah, absolutely. You know, the COVID-19 project was obviously a very fun collaboration with Olink. Olink came to us and was very eager to do this and share data, so we had a lot of fun with this. This was early on in the pandemic. We realized we had a very good opportunity to contribute to the understanding of the disease biology. We had infrastructure at MGH to collect serial samples from patients that come to the ED, and this infrastructure was set up based on previous studies looking at sepsis. We assembled a cohort of over 300 COVID patients and collected over 800 samples. You know, we performed a range of studies, but plasma proteomics was the fastest. I mean, it's a good testament to how useful it can be. We were the fastest aspect to perform. The single cell studies that we performed are still sort of being analyzed and being processed. Our goal here was twofold. One, we wanted to understand again what predictors of severity would look like and understand how the immunology of the disease changes over time. For that, we had serial time points from the date patients arrived to the ED to three days after and seven days after. Overall, we found the data was extremely fruitful. We were able to sort of define distinct signatures of patients that had more severe disease and that passed away versus patient signatures of patients that recovered from, and we found there was a specific subset of proteins that seemed to rise during the acute inflammatory phase of the disease, but then seemed to resolve in those patients that did better. Importantly, by comparing to tissue expression data, we were able to find a module of proteins that underlay the interaction of myeloid cells with T cells, with epithelial cells that seemed to drive disease pathogenesis as well. That module, you know, that mechanism has since been validated in multiple tissue studies as well. Great. On this cohort was part of the COVID-19 Technology Access Framework, where all data was shared publicly, and we are amazed by more than 700 research sites around the world has downloaded the Olink data, 1 million the Olink data points. Why do you think it attract so many groups, this data set? Yeah. Yeah, absolutely. I think for a number of reasons. One, because of the serial sampling is very influential. Sort of the range of proteins that were sampled. I think overall, the quality of the data set and the quality of the proteomic data that we were able to obtain. It enables a lot of different groups to then look at the data and validate their findings or use it as a discovery pipeline for future findings. I think for multiple reasons, the data has been proven to be useful, and we've had a number of people reach out to collaborate in analyzing the data. Great to hear. I also know that on this cohort, you have both access to Olink data, of course, but also SomaLogic data. I know the investor community is really interested in hearing your view on this, how you compare the two datasets in terms of performance- Yeah. also how you apply the two datasets. Yeah. It's a great question 'cause I honestly, you know, till then, I didn't know what to expect, and I hadn't had the same, both Olink and SomaLogic from the same cohort. What became very clear and apparent right up front was, you know, the Olink data set had a much larger dynamic range. For findings related to individual proteins, we really focused on the Olink data set. We couldn't really rely on individual proteins in the SomaLogic data set. Most of all the findings from our paper about, you know, differential proteins and modules that we discovered were from the Olink data. That's because of the sensitivity and the dynamic range of the data. Where the SomaLogic data was helpful was that because it had a much broader range of proteins, we sought out to define tissue signatures of tissue destruction to understand, you know, the pathology over time of how, you know, for example, cardiac proteins look like or lung proteins look like. In that sense, you know, we wanted to look at intracellular proteins and define those signatures based on intracellular proteins because they would reflect a leakage of those proteins during the destructive phase of the disease. We were then able to re-leverage the SomaLogic data just because of the larger set of proteins. Because we couldn't look at individual proteins, we had to use them as signature scores. We had to combine, you know, dozens of proteins in order to faithfully get signals that we thought were believable. Mm. Um, or... I understand, especially in an infectious disease as COVID, that in particular sensitivity and dynamic range is extremely important, that you can cover those very low abundant one in an early stage and then when it raises. I understand that that was very important for you then. Yeah, absolutely. Both in COVID and in our cancer data, it's fundamentally important, which is why we, you know, we stuck to doing COVID, so Olink for both our COVID study and all of our cancer immunology work. Great. Thank you, Arnav. We could continue talking with you for hours, but we have to wrap up. It was great having you here, and thank you for sharing your view, and we are looking forward to see how the continuation of these two projects will become. Thanks again, Arnav, for taking the time. Great. Thank you for having me. Wow. Great. Very exciting, interesting discussions. Sure. For sure. We thought we'll allow everyone a short technical break. We'll take 5 minutes, and then we will be right back here with more exciting science. Thanks. All right. Welcome back. We're gonna continue with these exciting fireside chats. I just wanna remind everyone to keep sending in your questions for the Q&A session we will have towards the end. I'll hand right back over to you, Ida, to lead the charge with. Great. Thank you, Jon Heimer. Now it's time to meet Dr. Emma Guttman-Yassky. Emma is the System Chair of the Department of Dermatology and the Waldman Professor of Dermatology and Immunology at the Icahn School of Medicine at Mount Sinai, New York. Emma is also the Director of the Center of Excellence in Eczema and the Laboratory of Inflammatory Skin Diseases. Very welcome, Emma. Wonderful as always to see you. The same. Emma, as you know, we are very impressed by your career, which has received recognition for excellence by various academies, and I know that you are considered one of the senior thought leaders in the field of inflammatory skin disease, with groundbreaking research done on atopic dermatitis. Can you give an overview of your research and also in layman terms, maybe describe what is atopic dermatitis and why biomarker research is so important? Sorry for the long question, but I'm sure you get it. Yeah. No, of course. You know, I'm a little bit unusual maybe in the field because I did two residencies. I did one residency in Israel in dermatology, and I did a PhD in Israel as well. After that, I came to the United States to do a fellowship, thinking that I'll go back to Israel, which never happened. I stayed, repeated another residency, and after that I joined Sinai. From the beginning, I was very interested in atopic dermatitis, a disease that at the time there was this debate whether it's a barrier disease or an inflammatory disease. This debate really prevented therapeutic development because pharma companies didn't know what they should go after. We published the first molecular maps for the disease showing that immune abnormalities have a major role in the disease and likely are contributing to the barrier abnormalities. We followed a study showing the systemic inflammation in atopic dermatitis in patients that have a significant disease, showing that to treat patients with a moderate to severe disease, you cannot just treat with topicals. You have to give a systemic medication. I think what's maybe unusual or unique maybe about me is that I'm really leveraging both the clinic and the lab. You know, many people do one or the other, but I'm doing both. I'm doing, let's say 10 hours maybe of clinic a week, and the rest I'm doing lab work, and now I also became chair, but that allows me to do the bench to bedside approach that I think has worked very well for me because I take ideas from the clinic to the lab and then vice versa. Great. Yeah, you're unique for sure. What biomarkers are used in atopic dermatitis today or just maybe yesterday, but let's say today, and why are they lacking? Atopic dermatitis, first of all, to those of you that do not know, is a very itchy disease, and I'm sure that everybody has some family member that itches and has red patches. It's actually becoming now the most prevalent disease we have, inflammatory disease, not only in dermatology but beyond. We are talking about 7% of the adults in the United States. About a third of these will have moderate to severe disease. You can imagine it's a sizable chunk of population. Biomarkers are really important in atopic dermatitis, both at baseline, they were very instrumental in order to understand the disease phenotype at baseline, and very importantly also with treatments in order to identify which treatments work and how do they work. Some of the biomarkers that are also in the Olink panels are Type II biomarkers such as CCL17 and other Type II biomarkers that we utilize every day. Great. I think you were one of the first users in the U.S., so we're very happy to work with you for five years or more. When you first started using Olink, did you compare our PEA technology with other trusted technology? If so, how did they compare and how did you do it? Yeah, that's a great question. I did compare. I was doing, for example, IL-13 in a Singulex and in Olink. You know, Singulex is very sensitive but requires large quantities of serum. One thing that we didn't discuss, atopic dermatitis has a very large portion also of children that are affected. For us, it's really important to utilize techniques that utilize a really small quantity of serum. We compared IL-13 and IL-17, for example, and we found very similar data with very high correlation between Olink panels and a Singulex. Great. Great to hear. You already touched upon it a little bit, but which attributes of our platform have been the most value for you to drive your research forward? First of all, the fact that with the same quantity, that is a very small quantity, I can do not only a single biomarker, I can do really discovery. Many times we are going for, like, if I want to do a pre and post or even a baseline for some diseases, I do not know what I will find. I like the idea of non-directed biomarkers that we see what we find as compared to normals, and we don't do only a single biomarker because particularly if that single biomarker utilizes 115 microliters, and even more sometimes, there is a limited number of biomarkers you can do. I like the idea that with very tiny quantities you can do so many biomarkers. Great. In your many, many projects that you have run, I know you mostly use our targeted panels. We call them Target with qPCR detection. How have you decided on which panel to run? How do you select them? We looked in and published quite extensively in profiling using RNA-Seq, and we saw which panels are, you know, they have gene expression that is concordant with those panels. We decided on several panels that are inflammatory, cardiovascular, and neuroinflammation. The good news, recently, you guys also have the Explore panel that we've done now successfully as well in one study. Soon we will do a huge study with multiple inflammatory skin diseases. We are very excited about that because it allows again to explore more biomarkers. Great. That's obviously, of course, great to hear. Maybe I had that question later, but I'm gonna raise it up here since you brought Explore up. As you said, you wanna expand, and you have a very ambitious project in the pipeline where you want to develop sort of a Skin Disease Atlas. Can you say a few words about that and what you hope to accomplish with it? Yeah. We want to accomplish a Skin Disease Atlas in both skin and blood. One thing that I didn't say, we utilize the Olink platform successfully, not only in blood like many others do, but as dermatologists, we have the luxury of also sampling the skin. It's accessible. We managed to get a good amount of protein, and we found very nice biomarkers using the Olink proteomic platform in the skin. That can also be achieved in studies where you can only take a tiny piece of skin. It doesn't necessitate in other studies, we do 4 millimeters. You can only do a 1 millimeter biopsy and achieve that, or you can even do it in tape strips. Great. Yeah. I know you have a long list of high-impact publications using both plasma, serum, skin sections, but you also mentioned tape strips. Why is tape strips to collect patient tape strips so important in your research area and for your patients? Absolutely. Yeah, tape strips are important because, you know, we are dealing sometimes with infants and children and in these you cannot do biopsies. I think we are seeking more minimally invasive techniques, you know, not fully invasive, but, you know, blood and tape strips are possible to collect. You can collect only a few tape strips and achieve a good success with Olink proteomics in tape strips, as we have seen in these populations. Great. Maybe just so everyone understand what tape strips are, that you collect just skin cells from regular scotch tape, or how does it- Similar. Yeah. Yeah. Think about the Band-Aid that you apply. Yeah ... multiple times. Here with your technique, you don't even need to apply it multiple times, only a few times. That's really amazing that you collect enough of information or material. That's fascinating to hear. How do you hope that these results will improve patient outcomes in the future, Emma? Yeah, no, 100%. First of all, I heard the one that spoke before me, and I agree. It allows to sample more patients because biopsies are good to have, but you cannot have biopsies from all the cohort. Many times, patients will not agree. You know, the more minimally invasive technique you use, you'll be able to sample more of your population, allowing more of biomarker discovery and enriching your population. It's particularly important in a longitudinal studies where you want to follow up children or a clinical trials of large populations. Great. I have a couple of questions around clinical studies. Do you include Olink measurements already today in clinical studies? And if so, it would be great to hear an example of a study. Yes, I do. I do include Olink measurements in studies to inform me of what are the molecules that we are shutting off with treatment. For example, if we want to investigate a treatment that targets the type two pathway, the Olink has very good biomarkers in that pathway, and not only to see if you target specifically the type two, but you want to see if other pathways are also inhibited to some extent, and you can compare to see the ratio, for example, TH2, TH1. I heard with great interest, for example, the COVID study. You know, these are all things that are relevant now as well because we want to know how patients also cope with COVID responses and so on. Great. Yeah. How do you see that clinical study designs have evolved over the years, maybe patient selections, or maybe you already now are implementing some of the new designs or strategies? Yeah, no, absolutely. You know, in the forward thinking is that maybe at some point we'll have a personalized medicine approach for inflammatory skin diseases that will put the patients on specific treatments based on biomarkers. I don't think we are so far from that era. It's probably more relevant to diseases that are more heterogeneous, like atopic dermatitis, maybe less so in psoriasis. I do see it coming based on age, ethnicity and so on. Great. You are the top thought leader in your space, and I know you collaborate with many or maybe all biopharma companies in inflammatory skin disease. Can you tell us how these collaborations work and how you impact the study of choice of platform and biomarker development? Yeah, no, of course. I do, you know. I'm working at Mount Sinai, so I always say when you work with everybody, you don't have a conflict, because all the grants and everything comes to Mount Sinai. I'm really excited that I'm part of this really rapid therapeutic development. I collaborate with many companies in which some of the work is done in my lab, some of the biomarker work and the thinking behind the biomarker work, which biomarkers to implement. I do think that biomarkers, both invasive at some stages and non-invasive, have played a huge role in the therapeutic development and in understanding very early if it's a go or no-go, particularly smaller companies that, you know, don't have money for larger trials. Emma, you are very successful in your collaborations with biopharma, so I have to ask, what's your Olink pitch? How do you, what parts do you highlight maybe so you understand my curiosity? I think you know me by now. I am very convinced because, you know, when I see multiple studies that show, for me a biomarker is not a good biomarker unless it correlates with the severity of the disease and with the treatment response. When I see across studies that I've done with Olink, several things. First of all, that it correlates very well with the severity of the patient and also correlations with RNA-Seq and PCR, so we see protein-gene correlation. I'm a believer, and particularly that you need tiny quantities, I think it's important to implement, and you can translate it later on to children and, you know, other applications. Great. Jon, do you have a question? Yes. Hi, Emma. There's actually an audience here with a few questions as well. If it's okay, I'll chime in with a couple of those. How do you envision the frequency of utilization of proteomics assays in real practice compared to oncology and genomics today, where a single therapy management assay is used and maybe annually a recurrence monitoring assay might be utilized, say, 3, 4 times per year? How do you think about that from your clinical perspective and a proteomics perspective? Yeah, that's a very excellent question. You know, it depends what we want to achieve. Maybe in the future we'll have like a panel of biomarkers that we can utilize. Or maybe like in Japan, that we'll single out one biomarker, to utilize and to identify patients, for a particular drug. I think both are good and, you know, it's probably also a difference in cost, and the approval through the, HMOs. I can see how in the future this will be potentially done. Great. Credit should be where credit is due. So the great question was from Puneet Souda at Leerink. Here's another one from the audience. Have you seen any change in overall funding for proteomics technologies like now compared to a few years earlier or whatnot? Absolutely. You know, like anything in life, people are more enthusiastic to fund a project based on proteomics after they've seen data. I'm a great believer in new technologies and I like to adapt new technologies and try them. I think now people accept it. They understand that this is a good way to really study and understand systemic immunity and inflammation and different immune pathways that are upregulated in patients. I see it utilized more and more in the systemic space. Also, I see that it may penetrate in the biopsy or tape strip arena as well. Great. Emma, what are the next steps for you to implement your findings into the clinic? I think, you know, it depends on different drugs that need to show a promise in different subsets. Based on these subsets, I think we can think about a single biomarker or maybe a panel of biomarkers that will be able to identify responders that can go into the drug, either in trials or in the clinic. I can see that in indications like alopecia areata, in atopic dermatitis and more, and I want to also implement additional proteomic studies in other inflammatory skin diseases that need a new treatment. Great. We just have a few more questions or maybe just one last one. You already mentioned that you apply both transcriptomics and proteomics in your research. Can you discuss a bit on how you use the different data sets and maybe also how they complement each other? Then I'm gonna let you run. I know your time is precious. Thank you. Yeah, you know, it's important to see gene-protein correlation. I think both have a tremendous value. I do like to do both, and then I like to correlate them as well. It depends also what's the quantity of skin that I have. If I have a tiny quantity and I cannot do at all transcriptomics, then I would actually even favor proteomics because again, I think the quantity is a huge plus of this technology. Great last words, Emma. Thank you. I know you have to run, but a sincere thank you from all of Olink to you and looking forward to next time hearing about the Skin Disease Atlas. Thanks, Emma. Thank you. Bye. Bye. Nice to talk to you. Bye-bye. Great. All right. Good discussions, Ida. Sure. Thanks very much for hosting those. We will move on onto the schedule and with us we will have Ferhan Qureshi from Octave Bioscience. Good morning, Ferhan. Are you here? Yes, I'm here. Fantastic. Great to see you. Good to see you as well, John. Yeah. You're an accomplished leader in the precision medicine and proteomics space, and one of the most experienced and insightful customers we have had, we've interacted with from a protein biomarker research and development perspective. We as an organization have certainly learned a lot from you, me personally as well, and which we are obviously very grateful for. We thought it would be super interesting for everyone to, if you could, briefly summarize your career and the accomplishments that you've done. Sure. I'd be happy to. I was very fortunate to have had a fantastic example and some great inspiration right at home. My father had a very successful career in the pharmaceutical industry, and he encouraged my fascination with the natural world and steered me towards studying biology and then a career in life sciences. I actually started my career as an intern at Alza, which was a company that he was working for at the time, and I had the opportunity to spend a summer in an analytical chemistry lab supporting novel drug delivery systems. That internship helped me earn a stint at a company called Santa Cruz Biotechnology, and I started working there while I was still in college. Learned a lot at Santa Cruz Biotech, antibody production, kit and reagent manufacturer. Eventually, it led to a leadership position supervising the quality control department. We were doing routine ELISAs, Western blots, all in support of a very vast and expanding catalog. Soon after Santa Cruz Biotechnology, I took a position at Genentech. This was in the early 2000s, so that was a great place to get experience, a really exciting time. Herceptin, Avastin, Rituxan, they were all evolving or had been launched as products. Eventually, I was running a GLP lab operation, developing, automating, deploying pharmacokinetic, pharmacodynamic, immunogenicity assays for oncology, autoimmune, and all of this was done in a regulated environment. I had the opportunity to participate in FDA inspections and also get some exposure to some novel assay techniques beyond routine ELISAs and Western blots that I had worked on before. I then went to a company called Aviir. That was a startup, in vitro diagnostic company that was focused on cardiovascular disease, and that was really my first exposure to several of what are now considered the conventional multiplex proteomic platforms. From there, I went to Crescendo Bioscience, which was really an incredible journey. I started early on, research and development, all the way through successful validation and launch of a multivariate proteomic assay for rheumatoid arthritis disease activity. At Crescendo, I got really heavy exposure to many more multiplex platforms, Luminex, Meso Scale Discovery, Quanterix, Simoa, others as well. We were scaling up at the time, as the product evolved, so lots of exposure to automation, batch to batch, manufacturing. That was actually my first intro to Olink. We did a project with Olink when I was at Crescendo. After nine years at Crescendo, I joined Bill Hagstrom, who's also was the Founder and CEO of Crescendo, who then founded Octave Bioscience, and that's where I am today. At Octave, I'm leading a cross-functional team, R&D, clinical lab, CAP-accredited, and a team of data scientists, and we're developing multiplex immunoassays for clinical use in neurodegenerative diseases. We looked at multiple immunoassay platforms involved in our product development, and we selected Olink as the platform for our first biomarker product, which is a custom blood-based immunoassay panel to manage multiple sclerosis disease activity. Wow. That is certainly a very comprehensive career. Let's continue and dive into a bit of the detail of the work that you've done in Octave. To me, it's been a true pleasure. We've collaborated or worked together for five years now and Yeah. I've been following your fantastic progression. If we go back and when you start from discovery to cast this very broad net of a combination of novel and unknown biomarkers and using sensitive, specific, precise, and accurate protein biomarkers to drive your science forward. You're also, I know, you know, really appreciative of scalability. Based on everything we've talked about, how could you sort of compare and contrast Olink to all the platforms and technologies that you've used in your very extensive career? I think it would be quite interesting for everyone to sort of get your expert insights into that perspective. Sure. Well, I mean, I think one of the advantages to Olink that resulted in us selecting them for our custom assay development product. I mean, there's many great platforms out there, and all of them have a lot of advantages. But one of the things that really attracted us to Olink was the wide net, the number of analytes that we could target. You know, a wide net has many advantages, particularly when you're dealing with a complex and heterogeneous disease like multiple sclerosis. There's many pathways, many cell types, different mechanisms that are involved with MS disease pathology. Olink's panels and their menu, which is rapidly expanding, enabled us to do what we refer to as a deep scan of the proteome and really hone in on a signature. We used a dynamic and iterative process starting with, you know, over 1,400 biomarkers and down selecting those to a focused panel of 21 analytes. With Olink's commercial panels, we felt the content was very thoughtfully organized, with pathway-specific panels, disease-specific panels, inflammation panels, immune response, neurology, et cetera. That really allowed us to successfully target that focus set of markers for our products. I also think it's very important when you're doing biomarker discovery that you're focused on both correlation and causation. That's critical. And both were very central to the process at Octave. We used many techniques to explore causation for our biomarkers as well. Protein, protein interaction mapping, gene set enrichment, literature curation. Spatial expression analysis and even some other techniques to corroborate the proteins that we associated with the clinical and radiographic endpoints in our studies. Regarding the Olink platform, sensitivity and specificity were key aspects for us to consider. We were measuring the signal in the periphery, so in blood, serum, plasma. Olink had sensitivity because of the PEA technology, qPCR readout that was equivalent or better than many of the conventional and even some of the more sophisticated immunoassay techniques. We looked very thoroughly at analytical performance. As I said, our team has a lot of experience across multiple platforms. Accuracy and precision are also of the utmost importance to ensure that we could reproduce our results plate to plate, batch to batch, lot to lot, and then over time. There's been a lot of advancement in the field over the last 5-10 years. There's lots of options, many more platforms than before, and, you know, more content, higher sensitivity assays. Olink had many advantages that we've leveraged for our program thus far. I think one important one, especially when you're in discovery, is the very low volume requirement that's critical. You know, we had the great fortune at Octave of working with some fantastic collaborators who had access to deeply phenotyped cohorts, and those samples are so precious. The ability to do what I mentioned earlier, the deep scan of the proteome using a very low volume of sample also was critical early on in our development process. Wow. No, it's super exciting to hear. Personally, I mean, being really evaluated by such a competent authority as yourself and your team, and to pass that test is obviously great. One of the aspects, you know, we try to educate investors more and more about proteomics, but what we haven't talked too much about is how we use relative quantification in sort of the higher plex, where it's really a case control experiment that is being done. Sure. As you hone in towards clinical decision making, we introduce calibrators, and run those and actually do everything in absolute quantification. Can you talk a little bit about relative versus absolute and how important that is, for the research that you do? Sure. I think if you include methods and appropriate techniques to bridge your batches, relative quantitation can be very effective for R&D purposes. I think Olink offers great advice on how to do that effectively, using pooled controls, stratification, and randomizing your plate maps, how you treat longitudinal samples. I think as you move forward with your project, especially if translational is your goal, incorporating absolute quantification is something that, you know, would be highly recommended, and something that we did effectively on the Olink platform. I mean, Olink is now doing this with panels as large as 48-plex, even on, you know, commercial panels. The focus panel that we developed included 21 proteins, and each one of those had absolute quantification, so the result is reported in picograms per mil, nanograms per mil, et cetera. It's something that I think people are used to seeing when they see results from a proteomic assay. They're looking for absolute quantification. But if you understand how relative quantitation can be handled in an R&D setting, I think it can be appropriate in those early stages as well. Cool. Great. How do you think Olink compares to this relative versus quantitative quantification versus all the other technologies that you've used in your past? You know, I think it compares very well. I mean, it's really all about standard curves and making sure that you develop them appropriately so they bracket the intended range of your sample distribution. We put a lot of effort with Olink into this process by running cohorts patients that were in different disease states, characterizing limits of detection, limits of quantitation. Olink's approach to do this is relatively novel with a larger than, I would say, typical range, up to 16 or more standard curve points that get incorporated at the lot level. As you mentioned, several calibrators run on each plate that are used to normalize the result back to that, what we refer to as that gold standard curve. Using that methodology, we've observed similar performance in terms of accuracy, precision, sensitivity on Olink versus several other multiplex platform. One thing I'd add about it too is this; this relates to the smaller volume requirement. We think about how that impacts samples, but smaller volume requirement can also impact things that are very important, like critical reagents. This includes protein stocks and even more importantly, the antibodies that you use to develop and run the assays. With Olink's approach, you can source and store amounts that would support a very large number of kits compared, I would say, to other platforms. That's really important when you're thinking about lot-to-lot matching, batch comparison as well. A small amount of reagent can also support an extremely large number of kits. You know, one of the most important things is that it relates to the team that's developing the assay. We're very fortunate at Octave. We have a great team, a very experienced team, R&D, clinical lab, and there's a fantastic team at Olink as well. The collaboration that we had between both groups to execute this project and develop 21 assays with absolute quantitation was fantastic. Yes, we were equally excited. Actually, you're really pushing boundaries at Octave here and, you know, pushed us also to the next level. You have completed the most comprehensive first analytical validation, followed by a clinical validation for utility before developing your final product and moving towards the LDT. I think it would be fantastic if you can share with us a little bit on that comprehensive analytical validation. Perhaps think about it as you also talk to laymen here. What does it mean? Why do you do it? What are the factors that are important, et cetera? Yeah. Well, I would say both are essential, especially when you're talking about bringing a test to the clinic. We've just this year completed both a very extensive analytical and clinical validation. In fact, we just reported last month at the ACTRIMS conference, the results of our clinical validation for this custom assay panel. The final algorithm that we use in our test, it's an 18-plex assay with proteins that represent four key pathways involved with multiple sclerosis. We started, as I think I mentioned this earlier, with a candidate pool of over 1,400 proteins. We measured it across several assay platforms, not just Olink, and then selected these 21 proteins onto our custom panel, which were then analytically and clinically validated. It's an important process to go through. With Olink, we developed a very thorough development characterization and validation plan that I mentioned was executed collaboratively over several months. We manufactured 2 lots of reagents, and in that process we varied critical reagents for our probes, for our standards, and for other reagents used in the test. We also added some new assays onto the panel, and in that process we made sure that they correlated extremely well to their counterparts that were run on other platforms. Olink also gave us this opportunity to blend together proteins that are found in both high and low abundance in the blood. They have a very wide dynamic range on the Olink platform. There was also some really novel techniques that were used to do what's referred to as dynamic range optimization, so you can blend proteins together, high and low abundance. Olink did a very thorough characterization, and we established pre-specified performance criteria that was evaluated thoroughly at Olink before the assay was transferred to us. We then analytically validated the assays in our lab. Even before starting the custom assay project, we had a lot of experience with running the platform. Our lab got certified by Olink back in 2018, and that gave us the experience that we needed to perform this extensive analytical validation. It included all the typical fit for purpose elements that people who are familiar with assay development would know about accuracy, precision, robustness, sensitivity, specificity, stability of reagents and samples. In addition to that, we also did some tailored experiments that we designed that looked at diurnal variability. We looked at the impact of disease-modifying therapies on interference in the assay. We reported these results earlier this year, and that completion of the analytical validation led us to do that extensive clinical validation study that we just reported. In that study, we took 600 samples from four different sites. This included both prospective and retrospective cohorts. We demonstrated that the multivariate approach, multiple proteins used together in an algorithm, significantly associated with several disease activity endpoints, and importantly, outperformed any individual biomarker as well. The output of our assay is a disease activity score as well as four disease pathway scores. Those are scaled from 1 to 10, with thresholds corresponding to the level of disease activity, so low, moderate, or high. After we completed that clinical validation, we went back and did another level of analytical validation, which was looking at the score level that I just described. We'll be reporting those results in a forthcoming manuscript as well as at the ACTRIMS conference in February. It was analytical validation of the biomarkers, clinical validation of our algorithm, and then another round of analytical validation of the output of our algorithm. I think through this process, we've demonstrated that Olink's platform is capable of achieving the highest level of analytical stringency. Anyone who's interested in more details, I'd be happy to talk to them about it, or they could look at some of our presentations and posters and even the webinar that we did with Olink earlier this year. We went through much more details on analytical and clinical validation. Great, Ferhan Qureshi. It's obviously super impressive to hear you represent the work that you've done, and frankly speaking, also from an Olink perspective, to pass your extremely rigorous validation. I spoke a bit about the importance of actually the standard curves, the upper/lower limits of quantification and so forth earlier today, and that we pride ourselves in the validation that we do for the Explore product in, like, a 3,000-multiplex setting, but also passing the test for PEA in this very, you know, in a clinical setting. I mean, it was the first one for us, and so it was a big threshold for us to pass as well, and very impressed with all the work from both parties here. One of the things that has been asked of us as well. I wanted to pass the question to you. How are the CVs on Olink platform perform, and are they to your satisfaction? Yeah, absolutely. You know, that's something that is critical. CVs relate to the precision intra-assay, inter-assay precision, and even beyond that, inter-lot. I mentioned that we manufacture 2 lots of our reagent kit, and the CVs have performed quite well. I think a lot of this has to do with the reagents that you choose for your assay. You can have a fantastic platform, as I think Olink is a fantastic platform, and there are several others as well. But an assay is really only as good as the reagents in some cases, and especially the antibody pairs. With any platform, some assays might perform better than others, and I think a lot of that is dependent on the antibodies that are selected. With our custom panel that I described earlier, the 21 proteins, only one of them, one out of those 21 was dropped from consideration for inclusion in our algorithms based on higher than desired CVs. I think that's quite impressive. In all the custom assay projects I've worked on, multiple different platforms, there's always been analytes that don't perform as well as others, and that's why we characterize and validate assays. Wow, cool. Thank you so much, Ferhan. This was super insightful. Really appreciate you taking the time. Now great, good luck with your launch of LDT, and really hope it will make a serious improvement of how patients in the MS space will get treated. Thank you. Thank you. Thanks, Jon. With that, been a lot of science here, so we thought we'll take another short technical break. We'll give you 5 minutes and then continue on to Chris Whelan, representing the UK Biobank Pharma Proteomics Project, before Carl Raimond will give more or provide a commercial update. 5 minutes, everyone. Hope to see you soon. Thanks. Great. Welcome back, everyone. Now we'll have the great pleasure to talk with Chris Whelan, who's a PhD and Associate Director and Head of Translational Genetics at Biogen Research and Early Development. He's also the Chair of the UK Biobank Pharma Proteomics Project. Hi, Chris. Great to see you again. Super excited that you are here, and thanks very much for joining us today. You certainly are a rising star with a very impressive resume despite your young age, and now also the chairman of the Pharma Proteomics Project, heading up the UK Biobank Pharma Proteomics Project. Could you please give the audience a brief introduction of the very important work you do at Biogen? Sure. Thanks, Jon Heimer, for having me. Yes, I am an associate director at Biogen. I lead a small team of geneticists called Translational Genetics, and we focus on combining genetic data with imaging data and fluid biomarker data to identify new drug targets and to better understand the pathophysiology of neurological diseases like Alzheimer's and Parkinson's disease. Great. Thanks very much for that. Over the past year, you've spearheaded the collaboration with the UK Biobank and the 13 biopharma companies and Olink. Can you please give an overview and background and the purpose of the project, please? Absolutely. The Pharma Proteomics Project, or PPP for short, it's a, as you said, a consortium of 13 pharmaceutical companies, and we are exclusively funding the generation of Olink data from 62,000 plasma samples in the UK Biobank. UKB probably represents one of the largest epidemiological studies ever undertaken, and we felt this would be an amazing opportunity to conduct proteomics at the population scale. Fantastic. What is the goal with the project from these 13 biopharma companies' perspective? Sure. I mean, I'll start off with my personal goal. My dream has always to be to conduct proteomics at a at a similar scale to genomics, so in other words, at the population scale. Genetics is now routinely employed as a healthcare tool and also as a research tool. It's applied across hundreds of thousands of samples in big GWAS studies. This is kind of commonplace nowadays. My dream was always to bring proteomics to a similar level to genomics and get it to the point where we're applying it at the same scale. With PPP, we set out to apply proteomics at a similar scale, and we believe that if, once we get this project off the ground, that it will help us identify new drug targets. It could help us identify new biomarkers for patient stratification. It could also help us uncover new insights into disease biology, of course, and ultimately help us make better medicines. Cool. Great. We will certainly want to join you in that ride to bring proteomics to sort of where genomics is today. What is the current status of the UK Biobank project? The consortium received the first batches of data in mid-October. We're currently going over those data, conducting QC, and readying them for downstream analyses. We're awaiting data on another 41,000 samples, and I believe that the next batch of data will be delivered to the consortium very shortly. Overall, the project is about one-third of its way through. It will continue on into the summer of 2022, and we have 9 months exclusivity over those data as a pharma consortium before they're released to the general public or the approved UK Biobank research community. Yes, exactly. That really highlights an importance there that you will have exclusivity to the data for 9 months. What do you expect the 13 biopharma companies, how will you analyze and what will you do with the data during those 9 months? Yeah, it's a great question. The data undergo rigorous QC, you know, a couple different steps. Obviously, Olink themselves conduct very high standard of QC on the data before they're released to UK Biobank. UK Biobank then conducts their own QC. They make sure that the data satisfy a number of pre-specified metrics around intraplate, coefficients of variation or rates of quantification, for example. Then once the data meet UK Biobank's standards, they're then released to the consortium. At that point, it's fair game in terms of how the data are used. This is a pre-competitive collaboration, but at the same time, each company will have its own use case. For the most part, the data are going to be analyzed independently across the 13 different companies. There is one exception. We have all agreed in principle to work together on a pQTL GWAS, or in other words, a genetic study of SNP influences on protein concentration. We're currently in the midst of finalizing protocol to collaborate across all 13 companies for that study. Oh, that's super exciting. I know that we've talked to many investors about the SCALLOP Consortium and exactly this sort of proteogenomics approach to identify novel and causal drug targets. Do you think that SCALLOP will be a part of your data analysis? At this point, we've kept it strictly within the consortium, but we hope that once the data are officially released, that we will have a number of academic collaborators, SCALLOP, of course, being chief among them. SCALLOP was a big influence on me in putting together this project. Okay, cool. When you kicked the project off, which proteomic platforms did the consortia consider for the project and why did you end up and move forward with Olink? I originally proposed this idea pre-pandemic, I think it was the winter of 2019, to another pharmaceutical consortium led by Regeneron, and that consortium was focused around exome sequencing of the UK Biobank. Now, at the time, we considered two main proteomics technologies to expand upon that project, Olink and SomaLogic, and we had good reason to believe that these were the only two multiplex technologies at the time that were capable of scaling to half a million people. We spoke with both vendors extensively. We laid out our terms very directly, and those terms included things like making sure the cost of the project would be competitive, that the vendors made no ownership claims over the data. That would be very important for UK Biobank. That the vendors agreed to extra stringent QC processes, and of course, that the vendors could actually kick off the study no later than the winter of 2020. Olink won out on all of those points, cost, reliability, speediness to data execution. They satisfied all of our requirements. That's obviously great to hear from my position. In your opinion, how is the Olink technology differentiated to other proteomics technologies? I would probably use one word, specificity. I have confidence when I run the Olink assay that I'm measuring the protein I'm supposed to be measuring. We've conducted Olink alongside gold standard assays, you know, targeted assays developed to detect a single protein with remarkable specificity. We've almost universally found strong correlations in the data. Olink holds up against these gold standard targeted assays. That's not necessarily the case with every other proteomic technology. Some will claim that getting more proteins is paramount. They might argue that having more coverage is better than having a more specific measurement because you want to capture as many proteins as possible. I would firmly and passionately disagree with that stance. There's little point in casting a wider net if that net has bigger holes. Other technologies can certainly increase the numbers that they capture, but I've yet to see how they can beat Olink on specificity. If we want proteomics to truly be a transformative technology, we need to cast the best made net, not necessarily the larger net with potentially bigger holes. In my mind, Olink is the best made net. If we wait another two or three years, I think it will be the best made net as well as the largest net. Wow. Thank you. Thank you, Chris. Those are nice words for sure. Does this feedback that you provide as well resonate with the other pharma companies different sort of views or feedback on the Olink technology? Yeah, that's a slightly difficult question to answer. I don't want to speak on behalf of the other companies, especially since we're all parsing through the data as I speak. I think what I can say, based on the last three weeks that we've had the data, is that the data look robust. The consortium across the 13 different partners has a number of world-class statisticians and geneticists and biomarker experts, and they're all inspecting these data with a fine-tooth comb. I don't wanna put words in their mouths, but it's probably safe to say that the data are holding up against their very high standards. What I can comment on, more directly, is that there's been universal praise for how Olink have conducted themselves throughout this collaboration. They've held themselves to extremely high standards. They've made their top research scientists available and accessible, you know, every day of the week. They've answered all of our questions, no matter how silly or obtuse, and they've been entirely transparent about, you know, slight project delays or logistic issues. Officially, Olink is the service provider on this project, but they've acted almost like a partner, like a true collaborator, which is fantastic. I'm sure that the other consortium partners would agree with me on that point. That's also great feedback, Chris. Thanks very much. I'll definitely forward that to the internal team. The consortia recently also expanded the project to cover our next 1500 proteins, which we're just launching now to get to 3000. How important is that to you? I think expanding from 1500 proteins to just under 3000 proteins is a no-brainer. Actually, as one person put it in the consortium, we're literally going to double the number of proteins that we'll be measuring in UK Biobank, and that effectively doubles the number of potential biomarkers and drug targets. I've seen the list of new proteins to be included in the 3K product, and it looks great. There's a lot of exciting markers in there, like ApoE, which is obviously a really important protein in Alzheimer's disease. There's STXBP1, which is a super important synaptic protein involved in epilepsy. There's a lot to parse through there. Just going from 1500 to 3000, it expands our discovery space, and it effectively doubles our chances at finding new targets and new biomarkers and ultimately making better medicines. Wow, that is great. There are apparently questions pouring in here. I'm gonna hand over to Ida, and she'll share some of the questions from the audience as well. Great. The first question from the audience is actually great follow-up to what you just said, talking about the expansion to 3,000. The question from Puneet Souda at SVB Leerink is how important is the total number of proteins that can be detected in the single run? Currently, PEA at 3,000 and on its way to deliver 4,500. What is the upper limit that we need? And maybe also, I know it's been discussed in the consortium, maybe you can also say what you think is needed. I love that question. It's scientifically fascinating, and I'm not sure that we have an answer to the, you know, plateau effect in blood. It's been rumored that maybe it plateaus somewhere between 5,000 and 7,000, but of course, I haven't seen a lot of peer-reviewed literature yet that really digs into it. I think that the number of proteins is obviously paramount. It increases our discovery space, as I said. At this point, 3,000 is a fantastic start, especially considering we have high confidence that those 3,000 measurements are very specific. I think going to 4,500 next year and potentially to 6,000 the year following, we're going to get very close to detecting what I believe is the full proteome in plasma. Now, I could be mistaken, but, you know, based on what I've heard, based on conversations with other scientists, we may plateau somewhere between 6,000 and 7,000. We're gonna get very close to that number within the next couple of years. Great. Yeah. Obviously, it would be extremely interesting to see that. Another question that came in, if you think 50,000 samples in the UK Biobank consortium will that be sufficient to gain meaningful insights, population scale, genomic scale at much larger? How do you see that? I think it's several times larger than anything we've seen before, certainly on a biomarker discovery level. You know, a couple years ago, I was conducting Alzheimer's disease biomarker discovery in a Swedish cohort of 1,500 subjects, and that was considered relatively large for a fluid biomarker study. We're going from 1,500 now to what would be 54,000 participants or 62,000 samples in total. It's already a huge leap for biomarker discovery. I think for genetics and for pQTL discovery, it's also a pretty big leap. I think that Claudia Langenberg just came out with a paper today which was in, I believe, over 10,000. I need to double-check that number. We're starting to get into the tens of thousands. Once again, this study, as a single study of a single cohort, I believe will still be the largest of its kind. We do view this as a pilot. We view it as the first step towards potentially doing the entirety of the UK Biobank and then moving beyond UKB to other cohorts of non-European ancestry. This is just the start in our eyes. Yeah. I think you can be proud and happy to have a pilot of 50,000 samples. Of course, it's great that you have 500,000 samples waiting. Another question, how important is concordance across platforms over time? Maybe I'll add some to that question, both concordance across platforms, but maybe also within a platform. You wanna apply this over time longitudinally. How does the scientific community take into account potential differences in results? It's of paramount importance, and we are looking into that. Within the platform itself, we already have data. There's a couple of proteins on the Olink panels, a couple of duplicate proteins like TNF. Just as a quick quality check, we've already looked at TNF, and I think there's an interleukin in there that's measured across all 4 panels. Safe to say the correlations across the panels is extremely high. It's I think on average, it's coming out at 0.96 or something like that. That's a good first proof of principle. Of course, in terms of over time on how these protein measurements hold up over time, we'll also look into that. We've baked that into this pilot project, so we are integrating a longitudinal component to this. That's why it's 54,000 samples, but it's 62,000. Sorry, 54,000 participants, but 62,000 samples. There are going to be longitudinal samples for some of these participants, and we'll specifically look to see how the Olink measurements change over time in healthy people and people with certain diseases. Then when it comes to correlations with other platforms, it is something that's also of high interest to us, and we are actively discussing ways that we can actually tackle that question head-on. I'm hoping we can speak more to that over the next couple of months. Great. That sounds promising. Another question. There are many questions for you, Chris. Do you see beyond biomarker discovery, the Olink platform transferable into routine clinical use in the future? I would love to see that. I think that that's the ultimate goal. I think the, you know, that is the dream, to actually make discoveries that turn into diagnostics. That is my hope. I think that the Olink platform has shown, based on my experiences, that it is high quality enough to achieve that goal. It'll probably take a few years to actually reach that, but the foundation is very strong. Great. I actually sneak in a question from myself as well. In the beginning, you said that you wanna take proteomics to the same level as genetics. What do you think you need to get there, or what do we need to get there? I've spoken about this before at other presentations, and ultimately, you need somebody to take the first giant leap financially. I think in the case of genomics, you had grants from the Wellcome Trust and many other government institutes kick-starting genomics and helping it follow Moore's Law, where initially it was prohibitively expensive and it costs you know, ungodly amounts of money to sequence the first human genome. Now it's routine. You know, it doesn't cost very much at all to do a single human genome. That was spurred along by initial, very large investments from government institutes. I guess, you know, I got a little impatient and I said, "Well, let's see whether pharma can make the first big leap this time around. Great. One final question that is actually not about the consortium, but more on you with the Biogen hat, you actually have experience with also developing a custom Focus panel. Can you speak a little bit about that journey and how you used it at Biogen? Yeah, absolutely. I think my first use case for Olink was with an Alzheimer's disease cohort. The reason that we wanted to apply Olink in Alzheimer's disease is that inflammatory pathways and innate immunity have been linked to Alzheimer's disease for many years now, mainly through human genetics. We saw studies that uncovered TREM2 and CD33 and other genes that are innate immunity genes expressed in microglia. One missing piece of that was that we weren't really seeing that reflected in real-time in patient tissue, in blood or in CSF. We took a cohort of around 1,500 people from Sweden, and about a quarter of them had some form of Alzheimer's disease, and then we had some healthy elderly, we had Parkinson's patients. We conducted Olink on their spinal fluid samples as well as their blood samples. We were able to find that there are several inflammatory proteins that are significantly upregulated in living humans with Alzheimer's disease in real-time. There are markers like YKL-40, MMP-10, Chitinase 1, and they all appear to play an active role in disease pathophysiology. This was very useful. It showed us that inflammation isn't just involved in the development of the disease, it's also very much an ongoing part of the disease. As you mentioned, Ida, we followed up on that by developing a custom panel from using some of the proteins that were most highly up or down-regulated from that study and made a custom panel that captures those proteins, and we're hoping to be able to use that for further studies downstream, potentially in some of our Alzheimer's disease clinical trials. Perfect. Great. Chris, thank you. Yeah. Thank you very much, Chris. Ida stole all my questions here in the audience, which is fantastic. That was the hope of today, so great that you could answer all those questions. Now we're gonna switch gears a bit, and I just wanted to round off, Chris, by thanking you so much for participating today and sharing the time with us. Really much appreciated. Thank you so much. With that, I've had the pleasure here to work alongside Ida, and now I'm gonna have the great pleasure to introduce another dear colleague of mine, Carl Raimond, who is heading up our commercial organization and will share a bit of his work and what his team does. I'll hand over to you, Carl, mate. Great. Oh, sorry, Carl. Thank you, Jon. All right, and thank you, everybody, for joining us today. First I'd like to take a couple of minutes and just share a little bit of my background. I've been in the life science tool space my entire career. I actually started in research, quickly discovered my love for the business side of life sciences. I took my first foray with Bio-Rad back in the day. I then moved to Affymetrix for roughly nine years, where I was able to participate in the early days of the genomics revolution to be part of that and to witness it all in real-time. Then I moved on to Agilent Technologies, where I was vice president and general manager of the Americas for the life sciences business. Among my responsibilities included Mass Spec-based proteomics. I moved to PerkinElmer as a global commercial VP GM of what was then the discovery and analytical solutions business. Of course, here I am. I had the opportunity to work with the great folks at Olink, and this was not an opportunity I could pass up for several reasons. One, to leverage my experience and capability in commercial management. Also, You know, the experiences I had in the early days of genomics really struck me as incredibly analogous to where we are today in proteomics, as the proteomics revolution, I think, stands in front of us right now. Finally, what I really saw was a breakthrough technology. I mean, I think something transformational for our industry, and I think the very thing that's going to usher in the age of proteomics and the promise of multi-omics that's been in front of us for some time. Lastly, you know, a well-established business. It wasn't merely an idea on paper. This was a company that had been built sort of customer by customer, and sort of, you know, publication by publication since 2016 by Jon and team. That was really exciting to me. Putting all those things together, you know, I'd like to share sort of where we are today with the Olink business. We're very excited. We have 700+ customers. I have to say, a number of those data points are single accounts which we have many customers. In fact, I'd say we're far north of 700. We're spanning over 44 countries around the world. We have business in the Americas, in Europe, in Asia Pacific. We're covering all the major market segments. We're working with biopharm customers, academic, and government institutions, and as you saw in clinical translation as well. You can see a number of logos here that we're working with. This is just a sampling. We are working with the top 20 pharmas, global pharmaceutical companies in the world, and we are working with the majority of the top academic institutions around the world as well. We have a large and established commercial base. And then in the center here, you can also see something that we've spoken about, but it speaks to our externalization strategy. I think you would be hard pressed to find success stories over history in the life sciences tool space where a company has been successful purely as a service, but externalization of the technology offers many advantages, which I'll speak to shortly. You know, I think again, sort of the age of modern proteomics is here today, and again, the established business that we have here at Olink is quite exciting. We shared some of this data last week, so this shouldn't be new to all the investors out there. We are at our plan year to date through Q3. We're very proud of that result, with 91% year-over-year growth. Then in the third quarter, of course, we executed well on our plan as well to generate 82% year-over-year growth, and we reiterated our 2021 guidance at that time. Now, not all revenue is created equally. So I wanted to share this slide because I think it really demonstrates the quality of the revenue that we're generating here at Olink. You can see vintages of customers running from inception in 2016 through full year 2020. What that tells you is we build a loyal base of customers. You know, it's a very sticky technology, and I think because clearly it's demonstrating value to our customers. You can see they tend to repeat. We tend to land and expand within our accounts as well. Finally, I want to pay special attention to this. We continue to add new customers over time. That is a key focus of our business, is continuing to expand new users. I can tell you from early data through Q3 of pre-orders that we received an interest in the Signature Q100 product launch and the Explore 3072, that we are not only exciting and accessing our existing base of customers, but we're also turning on a whole new set of potential customers that represent quite a bit of the interest that we've been able to generate in Olink. We're very excited about that and the story that carries forward in two ways, and I wanna spend a moment on this. One, the Signature Q100 instrument really democratizes access to the Olink technology. It's a very affordable, small footprint platform that's easily adoptable versus the prior generation, which was the Biomark HD instrument from Fluidigm. We have a modern, sleek instrument that really improves access of the technology to customers all over the world. We've already received several pre-orders in Q3 and you know, we're looking forward to the future there. We also are launching the Explore 3072 this quarter. This is something that is igniting the customer base as well. As you've sort of heard from a few of our customers, this product now accesses the full Reactome and several levels of detail underneath that. You know, the depth of content is fantastic, and we believe this really represents a tipping point in the market. In doing so, we've reduced the cost per data point, which is also something that I think is really exciting to the customer base in terms of the accessibility of proteomics. Now, when I speak about new customers and that expansion with Explore 3072, I can't help but spend a little bit of time here, of course, on talking about this product. We mentioned some of this data last week that we now have 21 externalizations as of the end of Q3. A very important point to make here is that not all customers are equal. We've spoken about accessing a potential base of over 5,000 Illumina instruments, and some of these customers have access to multiple systems within their labs. When we think about capacity, it's not merely the 21. Those are the number of external customers that actually represents access to a larger number of total systems. You know, we're well on our way to achieving our 21 goals that we've spoken of. Really there's something important here about the quality of these externalizations, and you've probably seen a couple of press releases from us recently, talking about Fulgent and Somagen, two world-class CROs. What happens when you externalize is you create value and access to your end users through giving them capabilities that trying to own all of that purely as a service business is difficult to deliver. Some of these externalizations, for instance, can deliver genomic data plus proteomic data, perhaps other omic data, and to be able to do that analysis. Some are expert in bioinformatics. Some customers want access to highly regulated environments, and they have access to CLIA and CAP labs, for that kind of, that sort of need. Then the inverse of that is when we externalize, we also gain access to our customer's user base. It's sort of a hub and spoke model, as we expand out and we acquire new externalizations that turns on the user base and the relationships that our cores have with their community. It's really a virtuous cycle we have when we externalize and the technology continues to mushroom. Now when I talk about Explore, I wanna be careful not to eclipse Target. I don't have a specific slide on that, but I also want to emphasize the importance of the Target and Focus product lines that you heard about from some of our KOLs today. Because I think this is a very important part of the story here at Olink, and I think you heard a little bit of this in some of the talks. You know, Olink Explore 3072 is very exciting and a lot of discoveries, but then as you think downstream, many of our customers want to carry some of those discoveries through to perhaps clinical trials or to clinical application. Leveraging the same fundamental technology in our Olink Explore 3072, which is based on an NGS technology, down to using our Olink Signature Q100 is tremendous that they can use the same assay and get the same assay performance as they scale from large down to small. I think it's a very unique story here at Olink and our ability to execute the broad vision down to, you know, a very applied set of focused biomarkers. It's early innings in proteomics, I think is an important point to make. We are aggressively investing in our commercial organization and expansion. From 2019 to 2020, you can see we expanded the organization tremendously. Then, through Q3 2021, we're already at 141 commercial employees and we still have a quarter ahead of us. We're continuing to expand and execute on our hiring plans in commercial through the fourth quarter. I can tell you, we've already begun recruiting for our 2022 hires because we will not miss a beat on commercial execution and the opportunity that lays in front of us. I mentioned this earlier, but I think one of the important things that's really driving this and that we want to be positioned to take advantage of is the falling cost per data point that's represented by the launch of the Olink Explore 3072. As I noted earlier as well, I think this is a real tipping point. We're hearing this from our customer. The inquiries in terms of new customer interest have been tremendous, and we're very excited about how that speaks to the future of proteomics and what Olink has been able to deliver to our customers. Finally, I wanna comment. We are recruiting across all regions. You know, we've seen success, and we talked a little bit about this during Q3. We've seen success across our Americas region, our European region and our Asia Pacific, and we're going to continue to execute across all regions. Because we have, you know, an already sort of maturing and growing commercial organization versus some in our immediate peer set, I believe this creates a tremendous advantage for us. It's easy to say you're gonna grow and, you know, and hire a commercial organization. As somebody who's spent their career doing this, executing on that, getting all of the parts working together, and creating all of the organizational capability that's required for strong commercial execution is not for the faint of heart. I'm proud of the commercial organization that we have today and the execution we have and we'll continue to have in the future. As we're making more of those investments, we're now moving on to a new level of sophistication in terms of building out more capabilities in commercial, in a number of specialty areas, and the ability to focus on a lot of key opportunities like population biology, cohorts, specific focus in pharma and specific applications. Lots of great data points I shared there, but I think this is a really important slide because this is an external data point. I think few things will speak as well as what's really happening with the technology and the science. As you can see here, we've already crested 750 publications that integrates the Olink PEA technology. That's a 50% growth from the end of last year. Again, we have months ahead of us before we end this year. Then you can see there's a great diversity in the number of disease areas that we're playing in. I think one of the takeaways you can grab from that is that there's lots of opportunities. We've had great success in areas like cardiovascular, but areas like oncology, metabolic disease and neurology continue to spell great opportunities for Olink and great opportunity for our customers to make incredible discovery. All of it is tremendous validation of the value that our customers are gaining. Just like you saw on the previous slide in that stickiness, I think you see it sort of, you know, represented in some ways here in terms of the utilization and expansion of utilization of the Olink technology. To wrap things up, you know, I wanna make a couple of points, and I shared a lot of this data already, but it's worth emphasizing. We have far north of 700 existing customers. We are working with the top 20 pharmaceutical customers around the world. We're growing our number of externalizations quite robustly. We are on plan through our third quarter for our annual plan. The number of data points we're generating is becoming somewhat mind-boggling and exciting in terms of what that means for discovery. Again, the peer-reviewed publication's a great validation, and there's so much opportunity for externalization and expansion that still remains in our business. You know, we feel quite excited about our capabilities, about our footprint, about the commercial execution that we've demonstrated and we continue to build upon in the coming quarters and years. Finally, we're a leading technology with strong commercial footprint, a solid foundation. Clearly we're investing aggressively for growth and not just in our commercial organization. Across Olink, we're rapidly expanding, and I think we're well-positioned to continue being the leader in our space. Thank you. All right. Great. Thank you very much, Carl. Now we should all gather up here for our Q&A session. We've collected questions from the audience during the day. Before diving into that, take a moment to introduce Oskar as well. Most of you have met Oskar several times, of course, but great to have you next to us here as well. Great to be here. Thanks, Jon. Good. Okay. We have a few questions from the audience. We can start with one from Tejas Savant at Morgan Stanley. Maybe direct this one to you, Carl. Are you seeing the UK Biobank work start discussions in the community for adding more proteomics arms to PopSeq projects such as the Million Veteran Program or the USPMI? Yes. Yeah, I think the interest in population biology has been exciting. The number of conversations has been continually expanding. You know, you heard from Chris Whelan in the UK Biobank, you know, about the potential value there and the value of biology at scale. I think that's, again, that's very exciting, I think, for what that means for the market, what that means for proteomics. I think we're gonna continue to see some great stories come out on this. Nothing, you know, we can speak of publicly, but lots of fantastic conversations are happening and you know, we feel very excited about the population biology opportunity in the market. Yeah. It's definitely the time to add protein to that mix, clearly. All right. Tejas had, Tejas said another one here as well. I'm gonna direct that to you, Ida. So several KAs today spoke about casting a wide net upstream during the discovery stage and then levering those efforts to down select a handful of biomarkers for use among patients. That would imply narrower panels will see greater uptake in the clinical use case, which comes with much greater volumes. Does that mean for Olink's consumable stream down the road? Put another way, over what timeframe do you expect broader panels to gain traction for clinical use? Perhaps this is sort of a question for both of you, but you can take the science part. I can start. Yeah. For sure, I guess is the short answer. That's exactly what we see, that you cast this wide net and then you find a number of different profiles and signature that you can take further down the line. I think Arnav presented some good examples, both in the melanoma project and in the COVID project, that we had this broad screen of one cohort, and he identified a signature for prediction, for therapy response, therapy guidance, as well as toxicity. Then it potentially could lead to several smaller specific panels. That's exactly what we see across all our applications area. Yes, I think that's. We're just about to see the real tipping point of that from the results of the from the casting a wide net. Cool. I think something we've heard clearly from our customers is one of the value propositions of working with Olink, and I just spoke of this, so apologize for being redundant, but you know, is that ability to sort of scan the proteome and then to sort of drive down to more specific content. Again, the ability to do so without having to move technology or platform so that they can get the same analytical performance with their assays working with the same company. I think that's something that's clearly resonating with our customers and something that again we think is an exciting and differentiating part of the Olink story. For sure. Sorry, maybe I can just add one more thing to it as well. I think going back a few years, we also thought that maybe this would lead to less exploratory discoveries, but that doesn't seem to be the case. You still need to screen more cohorts, validate your studies, find new insights. We're still not far done with the discovery part, but at the same time adding more focused studies to it. Yeah, for sure. For all the work we do with our customers, but we also, I mean, really closely together then build our offerings. I mean, I think it's very clear that, you know, the common denominator, and it sort of somewhat surprised me in a way that, if you are at biopharma or within academia, obviously your opportunities to drive these efforts downstream varies a lot. But they all, you know, why they, you know, jump out of bed in the morning and how they direct the science is really to make a difference at the clinical level. Obviously, you know, we see a great market opportunity in the research tool space, but as these fruits are reaped later on, obviously in a clinical setting, it will be, you know, a repetitive use and as Steffi has pointed out here, obviously, you know, a higher sort of pull-through on a smaller number of proteins likely, but in larger quantities. Right. There was a question from Matthew Sykes at Goldman Sachs. Matt, hopefully we answered your question, which I think touched on your comment about the value proposition of high-plexed, too, you know, to the Target platform. I think we just answered that question. Please, feel free to add something to the chat if you'd like to get a little more resolution on that. Okay, great. Here's another one, and I'll pull that to you as well, Carl. How important is having a widely accessible readout modality such as NGS rather than a dedicated instrument specific to Olink, for example? How important will continued adoption of NGS be for the growth in proteomics technologies? Yeah, I think that was a key strategic decision, the ability to launch on an existing install base of instruments, you know, that could utilize the Olink technology. As I mentioned earlier, you know, there's roughly 5,000 Illumina placements out there that could be leveraged to run the Olink technology. You know, sometimes capital equipment can be expensive and prohibitive. You know, being able to sort of jump straight past that hurdle in NGS and leveraging the power of the technology that's been developed in that space, it has been incredible for us. It allowed us to really sort of speed ahead with the technology and make it very accessible to our customers. When we talk about driving it down, you know, the launch of the new signature instrument sort of makes this technology also, you know, quite accessible to translate as well. You know, launching an instrument that's an affordable price point that can be sort of a workhorse in many different labs is very exciting for us as well. That said, we also leverage the existing Biomark HD install base that was out there. We have a number of existing target users as well. That's been a key and really important part of our strategy and execution. Yeah. No, just adding one point to that, which I, as a very customer-oriented organization, I mean, our scientists are really thinking about the future as a multi-omics one. Here, I think it's, you know, providing a unique opportunity. You invested in genomics platforms to do a lot of, you know, DNA and RNA. Now you can add what we think the world's greatest protein technology at massive scale, at affordability, which we heard today, to complete that multi-omics perspective, which truly, hopefully, will really change how we look at disease, health, you know, progression from health into disease, and be much more effective in healthcare moving forward. I think that's sort of an important point to think about. Yeah. An increase, I think it improved the number of opportunities for these external sites as well. Exactly. Which they've responded to very well, just to your point, Jon. They can expand their offering and sort of combine the data. I think it adds more value to the people adopting the technology as well because they have more, again, as to your point, more utilization for that investment and that system that they have. Yep. Another question perhaps at the back end of that, so you can maybe just elaborate to educate. It was a question around the benefit for Dr. Gerszten here to use NGS as a platform. The question is if it's more cost efficient than qPCR. The qPCR platform is, I think it leverages well our Target and our Focus platforms, which was fundamentally developed for, and our future custom offerings. Looking at sort of large scale, you know, what we saw was the great opportunity with NGS to look at proteins at great scale. NGS-based technology, again, all the investments and development that's happened over time was a fantastic leverage point for our technology. You know, a lot of the value lies in the sort of the creation of the assay, and the technology that we have. You know, whether you're talking about NGS or you're talking about qPCR, you know, those are fundamentally sort of detectors for our technology. As long as we see opportunity there, you know, we're gonna, you know, continue to drive that and leverage the existing install base that's out there. Sure. Yep. Here is another one. I'm gonna direct that to you, Oskar. Sung Ji Nam added another question. We actually asked one of the questions to one of the KOLs. Here she asked about the seasonality of our business. What are the key drivers, and as the percent of recurring customers continues to grow, what are some of the efforts that are on the way to alleviate the significant seasonality in your business? Maybe you wanna comment too, Carl, but I'll start with you, Oskar. Yeah, sure. Great question. I think to sort of start things off, you know, we've lived with this seasonality since the formation of the company. I think every year, you know, everything at Olink is, you know, done with an eye towards the big Q4 ramp. You know, a lot of things we do in terms of planning and execution throughout the year is to lever out the very big Q4. Then I think sort of the key drivers of the seasonality, you know, it's sort of a lot externally driven by customers and their year-end spend and year-end money. We do think as sort of proteomics, as we've heard today, becoming a more and more strategic omic and moving up the value chain and in the strategic budgets. You know, we hope to phase out some of that seasonality over time and also enabled by the kit strategy with sort of placements out there that you know that will sort of run projects on a more recurring basis. Unless you wanna add anything to that, Carl. Yeah. I think those are great statements. We, you know, we launched the Explore platform as a service in June of 2020, and we began shipping kits, of course, fully commercially in 2021. We sort of announced that the Explore 3072 was going to launch in the fourth quarter, and then we also announced that the Signature was going to launch in the fourth quarter. I also think it was an important year for us in 2021 that customers are really now thinking, as Oskar said, about how they're sort of adopting and budgeting for proteomics in the future. I think like a lot of businesses, you know, we would hope to see, you know, the business smooth out, I think, further into the future. Right now, there's a sort of established seasonality that if you sort of look back in our business is a very clear signal and is still, sort of typical as our business as we've guided. Great. Thanks, gentlemen. Ida, here's one for you. What are the keys, the key areas of research, diseases or indications where you expect PEA and proteomics-based approaches to have the most impact and provide the most insights? I don't think there are any key research areas, and that's a very good problem to have. That's how it is that I think today we saw representation from cardiology, immunology, oncology, dermatology and neuroscience with Chris. That's what we see all the time, that you can apply it in basically endless opportunities. Of course, with the span of the library, we have a very good representation of the low abundant inflammation. We see tons of opportunities in the immunology and inflammation space. As I said, we don't have any key research areas, but across all fields. That's only from the disease or research area perspective. If you go to the applications area that you showed in the beginning, it's from early biology discoveries to late post market. It spans for all research areas, all applications area, academic, clinical research and biopharma. That's where we are, and that's terrific. Exciting. Another one for you. Great. How do you foresee the potential of expanding your portfolio into other areas, species, et cetera, to further expand the library of targets? That's a great follow-up question. That's beyond what we talked about today, is human biology or human studies, with mostly blood samples. Of course, we can do. Chris presented, cerebrospinal fluid. We can do other type of matrices as well. We haven't developed any products yet for other specific matrices, but that's one thing we could do. Emma wants to push for skin or tape strips. That's different matrices. We could also go into other types of targets. PTMs, post-translational modifications, has been a hot topic, so that we're looking into. That could definitely be another path for us. Then other species, of course. We have a mouse panel, which is great, but we could definitely go beyond that into other species. We are already running canine and pig and non-human primates. Mm-hmm. We could develop many more specific products for other species. Yep. Great. Cool. Thanks. Here's another couple for you, Carl. First off, how are your customers thinking of their work downstream over their Explore experiments? Again, not to be too redundant with the prior comments, but yeah, and Chris, I think, pointed this out as well. There's been a tremendous opportunity to combine proteomics data with, you know, there's this treasure trove of genomic data out there, and I think we've heard pretty clearly from customers that, you know, genomics has delivered a lot. But the potential to fully leverage those investments that have been made in genomics are tremendous by combining those with sort of the real-time biology of proteomics. You know, we're quite excited about what that means for our customer and for the industry. Going back to the comment, a lot of customers wanna carry their discoveries downstream, so be that in a clinical trial or for clinical translation. That's again where our sort of signature Q100 plays in their ability to carry that technology down to a more focused set of markers. We've talked about the development of a custom product line or expanding on those capabilities that we have into next year. I think those are some of the things. Now, there's a lot of things people can do, of course, with their discoveries, but I think those are a couple of big ones that we're hearing from our customer base is the ability to leverage it from multi-omics, and two, to be able to translate that into smaller, more focused panels for applied applications. Great. Next one. Obviously we're growing our organization quite rapidly. From your particular team's approach, do you see any growing pains in the commercial organization? I wouldn't say growing pains, but it's certainly challenging to, you know, to scale an organization like this. I've spent a fair amount of my career, building and executing high-performance commercial organizations. It takes a fair amount of effort, to be clear. However, no, I wouldn't say any particular pain points. You know, the exciting opportunity, I mentioned it earlier, is as the organization has really matured, I mean, you saw that scale from 2019 to 2020, and then through the course of 2021, we really have a commercial organization of scale now, which is allowing us to really move on the maturation curve toward, again, some more sophisticated moves in commercial execution. That's something we feel exciting about, but you know, it's always a fair amount of work and development to build a strong commercial organization. Which you're doing a fantastic job with. Thank you. Great to have you here. With that, it's sort of a great transition as well. I wanted to take an opportunity to introduce one of our brand-new colleagues. We're super excited and happy to have Jan Medina, so we have a full-time internal IR person, and it's sort of a great opportunity for him to introduce himself to you guys because you're probably gonna have a lot of interaction. Hey, Jan, great to have you here. Thanks, Jon. Yeah, like, you know, I said before, it's really great to come, you know, aboard this company and this culture at this time. It's really fantastic, and I think what we're hearing today, it's also very, very early, you know. That's really, for me, very, very attractive. You know, I think I'm looking forward to interacting and communicating and emailing, speaking and texting with members of the investment community. Yeah, we're here to start work. Great. Very much warm welcome, Jan. Okay, with that, we will round things off. I've had a fantastic morning here. I hope you all enjoyed it. We have really tried here to direct the conversation today, based on the many interactions we've had with you to really dive into the science and look at things from a customer perspective, which all, of course, is sort of the real driver of any business and ours for sure as well. With that, I really wanna thank everyone, my fantastic team for your great job in the day to day and in particular today, and all of you as well that participated, and we very much look forward to continued discussions. Hopefully, we will see you in person early January as well. Thanks very much. Have a great day, rest of your day.