All right. Good morning and warm welcome, everyone. We're super excited to be here in New York City, and, super grateful that, we have you here in person. We obviously also wanna welcome the, participants that are here virtually today. We will all think and hope that we will have a very exciting, fruitful day today. Okay, our usual standard disclaimer, and the agenda today. First, I will give a little bit of an introduction, and then Carl will do an update on, our commercial execution. At around noon, or actually at noon, Ida will start hosting, a few fireside chats. Here, we try to pull together, true thought leaders that represents our industry in the best possible way. We have Claudia, who is an epidemiologist doing a lot of pop health studies, Anders Mälarstig from Pfizer, Charlotte Teunissen, a thought leader in neuroscience, Lori Turner representing Azenta, and then the grandfather or father of proteomics, Mathias Uhlén. Then Ida will blow you away, and sharing Olink Insight, something we are extremely proud of. And then I'll do a little bit of wrap-up, and then we have a Q&A session to end the day. What we're hoping that you will take away from today is, as an organization, you're never more valuable than the return on investment or the value that you create for your customers. I think what we have been extremely successful with and unique with our disruptive technology is actually to truly unlock an extremely complicated science proteomics. That unlocking value creation and the return on investment we provide to our customers is actually what's driving the strong attraction that we see across biopharma and academia. On top of that, how we deploy this disruptive technology, I'm super proud of our organization that we, I think, truly believe, understand how the customer works, their focus, and that we are extremely innovative on that technology and continue to innovate with an amazing pool of talent at Olink. We are basically just starting the first inning in proteomics as you will learn today. But we also believe that we have built a very strong foundation as a company to capture what we think a very, very exciting opportunity we have in front of us. I'll kick off my brief introduction. I wanted to start by, of course, celebrating a true milestone for us. I mean, how rapidly we've come to 1,000 peer-reviewed publications. This is of course an extreme validation to the underlying technology, but also its utility to science. What's also interesting, I think, the resemblance of this growth curve to our revenue generation and pointing in a positive way. As I prepared this introduction, I thought about what will be most useful to you. I thought about what type of questions that you are asking me. It's a lot about like proteomics as a complex science and trying to understand how customers uses it, and what value and return on their investment they perceive and receive. I took four publications that truly represents the most common use cases we see. The first is around proteogenomics, combining genomic and proteomic data. The second one is around predicting disease outcome and drug response. Then we look at drug differentiation through mode of action. Then last, early and more precise diagnostics. Let's get going. The first one is around combining protein and genomics data, proteogenomics. Here, we had really high hopes of that additional layer of scientific information on top of the genomic data. I'm actually kind of blown away of the type of feedback that we receive from our customers when they get these large data sets. We hear comments as p-values through the roof or exponentially more valuable data. We really make the genomic data that we have gathered over the past decades really come to life and become much more actionable. I selected a paper here, which actually has become sort of like a standard for how to drive or lead this science and the data analysis. It's done by the SCALLOP Consortium, which is the largest proteogenomics consortium in the world, that exclusively uses proteins from Olink. The real challenge that they're trying to unveil here is to differentiate between causal or confounders. If we take a look at the bottom left of the slide. Of course, you have the genomic information that you then tie to protein data, and you try to make a direct link to causal outcome. Not confounders. We can take a look at a concrete example. Here, in this study, they had gene expression of the IL-6 receptor gene. The IL-6, IL-6 receptor protein was expressed, and they were able to directly link this with causal outcome of RA. Of course, in RA, you will have a systemic inflammation, it's pretty significant such, and CRP will be very, you know, elevated as well. You could potentially think that that's a good drug target, but that's actually a confounder and will not directly affect the disease or its outcome. That's the type of approach that they do. In this paper, they ran 31,000 samples roughly with Olink Target panel, and they found 451 associations. If we go into the details and start at the upper left, the blue circle here represents already existing drugs where they could validate these targets for cis-pQTLs. They should work, and apparently they work, they are approved. The lower green one are novel causal drug targets, so true drug development opportunities. The larger box there represents trans-pQTLs or if we can think about them in a bit simpler way, label expansion opportunities. I thought I'd just try to make this very simple. If we pretend or put ourselves in the shoes of being of head of R&D for example, biopharma company. On the graph on the right or the table on the right, we'll get a list of protein targets. Here I took three. That has identified cis-pQTL indications and then trans-pQTL opportunities as label expansions. If you were that head of R&D and had this dataset, would you choose to go after IL-6 receptor here that had three cis indications or would you go after TNFR2 or TM that had one cis, but a number of label expansion opportunities? I don't know. Obviously, with this type of dataset, we will be able to in a much more precise and accurate and thoughtful manner look at the market opportunities for drug development, the potential risks and so forth, and make much more educated decisions. This is what the future will look like. In the past, we tried to have a mouse mimic a human disease and then try to treat that and then translate it into humans. Here we're now starting drug development on humans with human multi-omic high-quality data. That's the first one. No, sorry. Of course, we should add here as well. This was still in its very large study, but compared to what we're doing now, they found here, as we said, 451 associations. The UK Biobank consortium, putting out this first publication on bioRxiv for the 54,000 samples across 1,500 proteins. They found well over 10,000 associations, 85% that novel in both common and rare disease. Imagine when we do the rest of the UK Biobank or the tens of millions of population health studies that are out there, how that will provide insights to we will do a much more effective, targeted, and carefully planned drug development. Now, the next one, which actually might represent one of the most common use cases we see, so about prediction. First, either predicting disease outcome or predict drug response. In this particular case, we're doing it through a surrogate marker or efficacy. I'll explain that to you. They studied end-stage renal disease, and it's really scary, right, to see how prevalent that is here in the U.S. with 30% of type 1 diabetes patients and 10%-40% of type 2 that actually develop end-stage renal failure or kidney disease. Our challenge is that we have no clue which patients that will accelerate into such a severe outcome. Can we find tools where we could identify those patients and hence treat them much more aggressively much earlier to predict that? I'll come to the drug development perspective of that because that presents huge challenges here as well. What they did here, the Joslin Diabetes Center in Boston has a truly unique biobank with type one and type two diabetes samples that are banked longitudinally, and subsets of patients develop ESRD. Selected subsets of those from that biobank, and then they ran that cohort and everything, of course, here, in collaboration with Eli Lilly, it should be said as well. Let's take a look at the results. On the y-axis, you have the cumulative incidence of end-stage renal disease. On the y-axis or x-axis, you have time. They have identified signatures of proteins where, as you can see with these Kaplan-Meier curves, that the high up you are on the curve, the higher the risk is that you will develop ESRD, right? If you think about this is from baseline. If I have a waiting room with 100 patients in front of me, I do this test, and I see that these groups will very likely develop very severe outcome, and I should hence treat them much more aggressively as early as I can. That's very common. We see that across basically all disease areas to try to predict that outcome and find the really worse outcomes early. If we take then the biopharma setting. If we're a drug company again, and we would develop a drug to treat the ESRD, our phase 3 will likely be like 5-10 years of follow-up to show a clinical endpoint. How could I use these results to minimize my risk? Actually, if we then use that signature and see if we can compress these Kaplan-Meier curves or that they remain low, that will be a very strong indicator that we have efficacy of this drug. For example, in phase two, to reduce that risk versus sort of going blind into extremely expensive and long phase three, and if we can even potentially have FDA approve that surrogate marker for efficacy, we could use that as an endpoint and get approval. Tremendous opportunity for biopharma lies in here as well. This team now, they're doing Olink Focus development and are looking to take this into prospective clinical decision-making for everything that I just talked about. Great publication and use case that we see very often. The next one is around drug differentiation by explaining mode of action with very strong scientific evidence. I'll explain what all that means as well, hopefully clearly. Here is a true success drug story that we are excited to be able to work with. Jardiance from Boehringer Ingelheim and Lilly, it was approved in 2014 as an SGLT2 for type 2 diabetes. They expanded their label to also include cardiovascular disease. Jardiance inhibits reabsorption of glucose into the kidney, hence lowering blood sugar, right? It also has, you know, improves cardiovascular outcomes in many different conditions, but we don't know exactly or they don't know how. What this team did was to take a third label expansion going into a very large phase 3 clinical program with 10,000 patients. Starting in 2017, they were done in 2021, FDA actually pre-approved the drug for use in chronic heart failure as well. They could not explain the mode of action. They ran Olink Explore. This gets technical, I'll keep it at a high, simple level. Basically, what Jardiance does is to help cell to function optimally, getting rid of cell debris, improve cell repair, and cell renewal, improve mitochondrial or energy functions in the cell. It would also reduce negative impact from cell processes, basically stress, inflammation, and fibrosis. What does drug differentiation mean? As you know, the drug market is a very competitive market, a lot of companies trying to sell into the same indication. I have a very scientific customer as a physician. If I can, on strong scientific evidence, very clearly explain the mode of action, how these drugs actually work, that's a very strong sales argument, and really, what we say, differentiates this drug towards others in the same space. They were super excited here and now expand to run the 3K product. I think here also, from an investor community, if you think about it from the biopharma's perspective, that they have a very quick ROI here. We started off with proteogenomics. It's a long path to a new drug, right? Here they take existing approved drugs, they work with us, they identify, for example, new label expansion opportunities, which are much faster to get to, or they build the scientific evidence to be much more competitive in the market. The return on investment here is, like, very quick. The next is around earlier and more precise diagnostics, and this example is in ovarian cancer. This is a project that I've actually been working with, or been involved with personally all the way since we started the company. Ovarian cancer is a very horrible disease, right? With a five-year survival, only 30%-50%. The biggest challenge here is, you know, to detect early stages early, because if we do, we have very high survival rates. That we really need to do. The problem here is really around specificity or false positive rates because today we would bring in the patient and actually have to do open surgery for final diagnosis. There's really a strong need to develop, you know, biomarkers here. What this group has done, we have built our library of antibodies from 500 to 700 to 1200. They sort of continuously worked with us and added more and more proteins to try to improve the area under the curve. If we take a look at the bottom left of this slide, the perfect area under the curve is 1.0, right? As you can see here across all stages one to four, they have an area under the curve of 0.98 in the discovery cohort, and then they did a validation replication cohort, and they had an area under the curve of 1.0. It's like wow, right? But as I said, the more important one was this the early stages. If we just look at the analysis for stage 1 and 2. First, the dotted blue line you have here, that is, the gold standard that we use today. CA-125 is how we diagnose or detect ovarian cancer. That's its area under the curve. The second line you see is, actually an Olink Focus panel that they developed from the Olink Target panels that they had run earlier. That was that area under the curve. It was a little bit better, but, as you see when they added more proteins in Olink Explore, they got tremendous results. They will now also go to Explore 3K to see if they can get even better, and then they will prospectively validate it in a very unique UK longitudinal cohort. They see this, and what they are driving for is that it's actually quite low-hanging fruit. Today, we have cervical cancer screening program. You take blood tests, and you can just add on one more. Whole infrastructure is already there to add another test to that screening program and hopefully help a lot of patients to much better outcomes. 4 use cases, which are very, very common, and you see the sort of value that Olink and proteins bring to the table. Okay, I'm soon gonna wrap up my introduction, but I also thought about actually it was one of you investors that in a call with me said, like, "Hey, Jon, you know, I love proteomics. I love your technology, and I love Olink, but it's like make me wanna dream about Olink. Where, where is this gonna take us? I thought about that and tried to summarize that in a couple of slides. When we thought about the purpose, why do we exist? It's to reveal the truth of human disease protein by protein. In global diagnosis today, we use something called ICD or the International Classification of Diseases. Here I just took a few examples. Actually, you will hear another example that will fit beautifully into this from Charlotte Teunissen with Ida today. She's done exactly what I'm laying out here actually but in Alzheimer's disease. If we take multiple sclerosis as an example. If you go into ICD, and you google multiple sclerosis, you will get one classifier, MS, full stop. It's a known fact today that multiple sclerosis is a very heterogeneous patient population, so with subgroups or more homogeneous populations that should be treated differently between each other. How we think this will evolve is that when we go to, for example, the next revision of ICD, that we will, for example, in MS maybe have defined MS version 1 through 12. This will, you know, be much more precise diagnosis for those patients, much clearer therapeutic strategies, and much better outcomes for those patients. When it gets really exciting is when we look at it from the pathway and the protein perspective, and here it opens up a tremendous opportunity for biopharma. As Ida will also do later today with Olink Insight, here are a few images from Olink Insight. When I took those, I took here four examples of different MS subgroups, and I looked into the proteins and the pathways that were activated in them, and you see how different they are. Here is the roadmap for biopharma how to develop drugs and target those. On top of that, they can use this stratification in the clinical trial to stratify in my subgroup of patients that will now be very, very likely to respond. I'm gonna have a high power. The end of the study will go down, time, risk, cost, time to market will go down. This is how I think 21st century healthcare will look like and evolve. The good news is that this will, of course, have a tremendous impact on patients. We will get much more precisely diagnosed. We will receive much better and more targeted therapies and have much better outcomes. It will affect, of course, whole healthcare and society at large. The bad news is that this will take a long time. Which in a way is good news if you sell protein agents to make this happen or a reality. Hopefully, this helped you a bit to understand in practice how our customers operate, what kind of experiments they do and what results they get. This is actually how I and we think that the world will evolve. Proteins will be the most important part, and we believe that Olink will play a very significant role on this journey. With that, I'm gonna hand over to Carl, who's gonna give you an update on our commercial execution. Great. Thank you, Jon. Welcome, Carl. Thank you. All right. Okay. It is the best time ever in history to be a mouse. There are few diseases we can't cure in a mouse. I think we've all seen the headlines over a long period of time, right? Oh, you know, we can cure cancer in a mouse. We can cure this disease. We can cure that disease. A lot of it, as we all know, has not translated to human health. You know, in a lot of our sort of top-selling drugs out there could have an efficacy of 1 in 10, right? Imagine that. Patient 1, congratulations. Patients 2 through 10, I'm sorry. You know, you've spent your money. You've potentially had side effects, and you're gonna have to go find another therapy. You know, this speaks to there's some hole here, right? There's clearly some gap. The genomics revolution, right, took hold about a little over two decades ago. I was involved in that industry. Very exciting. I sequenced the human genome. New technologies came out in the world of genomics, and it's been fruitful. It's, I think it's transformed biology. Again, there's a huge gap. We knew this back then. It's not that nobody wanted proteomics. It wasn't possible to look at it at the scale we can today. You know, genomics, not to oversimplify, it's a, you know, it's a complicated technology. Right, you're looking at four bases A, C, T, G, and you're looking at sequences of them. The complexity of proteins is some order of magnitude beyond that, right? That's why these technologies sort of didn't exist back then. You know, DNA, incredibly powerful all over the what, you know, could happen stuff. You know, RNA, I think the hope was, well, that's fine. We can look at RNA. If you all remember sort of, you know, our first biology classes. You know, RNA translates into protein. Well, it's not, of course, that simple. Turns out RNA is not a terribly good predictor of what proteins are actually going to wind up being translated. And so this gap exists, right? I think Jon was alluding to this and the power of proteomics. There has been, and we've known this, a gap in our fundamental understanding of biology for, you know, quite some time. Of course, this is where Olink comes in, right? In the aid of sort of modern proteomics. I'll talk about four themes. I think you all know a bit about our business historically, but then, of course, looking forward, you know, what's driving Olink's business? Well, I'll talk a bit about scientific momentum, you know, touching on what I just spoke to, but I'll put that more in numbers. The Olink portfolio, so why Olink in this space? Technology adoption. We've spoken about our strategy and our business about exporting our technology to the larger scientific community. Then finally, world-class execution of the Olink business. Scientific momentum, you know, sometimes you don't need it too sophisticated to spot a trend. If you look at our numbers just over the past few years, you know, you can see the growth rate. You can see something is happening here. In fact, if you stretch this all the way back to 2016, and I think Jon has a slide on this later, you'll see the trend has been going on for some time. This is a strong indicator that something, you know, is happening in the marketplace. Importantly, it's quite broad, right? I'd like to distinguish that because I've been in this industry my whole life. It's not a niche application. It's not an academic application. It's not a small translational clinical type of application. We're seeing adoption across all segments of our business all across the world. There's a really, you know, fundamentally a very strong foundation for, you know, the growth of proteomics and that sort of mega trend in the life sciences space that you really see a birthing of this sort of mega trend of the use of modern proteomics biological research. Then if you sort of get underneath those numbers a little bit, and I think this is important, there's a very strong foundation. Again, we're not talking flash in the pan type technologies that many of us have seen over time. We had a press release, I think, just quite recently that we crossed the 1,000 publication threshold. Any company can wave hands, provide marketing metrics. This is irrefutable, right? The value that a company delivers to its customers that is translated in a very meaningful and tangible way. We're very proud of this. There's companies that have been around longer than us, yeah, who have not been able to deliver this kind of value back. We think this is a really strong indicator, and I think it's an excellent one to look at that again, sort of blows past any sort of marketing messages or anything else, right? This is real, right? This is the rubber hitting the road. Then customer account acquisition, I think we all know this from our personal lives. You know, if you have a good experience with a technology or a product, you're gonna tell your peers, you're gonna speak about it, you're gonna have success with it. We've seen this over time. We have a loyal and growing customer base that continues to grow and expand, and we're very proud of this as well. You can also see that, you know, we've continued to grow our customer base consistently over time. Again, it also speaks to this foundation that this isn't sort of a niche application in one corner of the market, right? We continue to expand further and further. The company you keep in the customer base, I think is also something that's quite important. We're proud of the customers that we're working with and the work that they're doing. We are working with all 20 of the top 20 global biopharma companies. We are working with 19 of the top 20 global academic institutions, and hopefully very soon that will be 20 out of 20. We're working with a significant number now of global CROs by revenue, which is also a sort of an exciting evolution of our business as we move forward as well. This is really important. These are early innings. You know, again, if you imagine genomics way back at the very beginning, you know, we're still at the early stages of what proteomics will be, and is estimated to be a market for various estimations from, you know, analysts and other analysis that, you know, should be, could be, you know, in far in excess of what the genomics market looks like today. Just a little piece of that to give you an indication. You know of our work with the UK Biobank, right? We were chosen by the pharma consortium to, you know, to look at over 58,000 samples for the UK Biobank. You know, we had a press release with FinnGen, which is another big population health study. We are working with others that we've not spoken about publicly. Tip of the iceberg is an understatement. We estimate there's over 42 million samples out there that have been collected or are in collection just in the population biologies. Forget about everything that we're talking about in pharma, which represents over half of our business today, and broader applications. Just in population biology, there's over 42 million samples. Just think about that market opportunity in and of itself. Again, we've sort of scratched the surface today. Next, why Olink? You know, why has Olink seen this success? I'd like to talk a little bit about our portfolio, which is completely differentiated in the marketplace. We uniquely have the ability to scale from the thousands of proteins down to the very few. When we launched Explore on the NGS platform, we created, you know, a whole new category in the proteomic space. We believe this was, of course, very transformational, as you know, as you've seen in our numbers. This represents a significant amount of our business today. This is the sort of parallel to thinking about high throughput genomics, that's this giant gap that's been out there for a long time. We have the mid-plex portfolio, the 96-plex products that has been a cornerstone of our business, you know, going all the way back to the early days of Olink, our Target 48 that sits in there. Our Focus product, which is a bespoke sort of product development for individual customers for their projects, like the Octave product is actually based on our Focus. But I'd like to call your attention to, and sorry, and of course, the Signature Q100, which is the instrument we launched to support all of our qPCR-based assays. Yeah, two really important product launches took place, one just about a month ago, and Ida Grundberg will be speaking to that later, and one that we just press released this morning that we've opened the order books on Olink Flex. We already had an extraordinarily differentiated portfolio. We have just now extended the distance substantially between Olink and the rest of the marketplace. I'll sort of dive a little bit deeper into these two since they're quite new. Olink Flex opens up a whole new category of business for Olink. We have been competing, as I talked about, the high-plex and mid-plex. This sort of smaller plex custom is a large market, an existing market, and one that is filled with older technologies and is ripe for disruption. We're really excited to bring the PEA assay into that realm, which represents, again, a tremendous opportunity both commercially, scientifically for our customers, and again, it further stretches out the ability of Olink to serve, the entire breadth of the market with the same, again, the same fundamental underlying PEA technology, the same super robust assay that provides the highest accuracy and specificity in the market now in this, lower custom plex space. It's a mix-and-match product of up to 21 proteins that you can select out of a library of over 200. These were also screened for compatibility with each other. This is quite important. This is a pain point in the marketplace with other platforms. So we've pre-screened, so we know that these assays will work well with each other, and it's a true plex. It's not like here's a five plex and a five plex and a five plex or something. It's a single plex. It's also uniquely available in both absolute quantification or our NPX values. For clinical translation type applications, this absolute quantification, so you can sort of read your protein in picograms per milliliters, for instance, is, you know, it's quite important. I think it's important for that market segment which is used to this, and they really appreciate and want, you know, that kind of capability. Finally, our customers can access and browse content in the context of pathways and biology via Olink Insight, which is a great segue into Olink Insight, which is something we launched about a month ago. Again, I'm not gonna steal Ida's thunder 'cause she's gonna go into this in much more detail. The point I would like to make is what a differentiator this is. I challenge you to find another company in our space that has this kind of support for their products as well and their customers. Not only on the front end, like I talked about, the ability to go and browse and choose a panel with the new Flex product, but also if you're just choosing one of our standard panels or you're looking at Explore, you can browse the content in the context of biology. Then importantly, on the back end, once you've run an experiment, you know, this is the last mile, which is perhaps the most important, is making sense of your data. We, you know, we've created this platform to assist our customers in that ability. Not only do you get data, but now you're able to make sense of that in biological context. You're able to, you know, tie in, Ida will highlight some of these features to look at sort of publication, how do my results relate to what's out in the public domain, and to be able to think about your next experiment and sort of rinse and repeat and come back to choosing content again. Again, a very differentiated portfolio. That's why Olink in the product space. Then, in terms of our strategy, we've talked, you know, of course, about exporting our technology and the importance of that. We somewhat uniquely started off as a service business, incubating our products, primarily, within Olink. Over the past few years, as we've talked about, we've been sort of more aggressively exporting our technology as a key strategy, both in terms of its, you know, its great margins on our kit business, but also in terms of the amplifying effect of exporting your technology, right? You ask all of these other networks that when you sell your technology, of course, especially if it's a core facility or something like this or a CRO, you're just keep expanding the network of customers that you're exposed to. In the high-plex space, you saw on our very last quarter, and we had a lot of questions on this, yes, we've seen an acceleration of externalizations around Explore. We focused on quality, and I wanted to, you know, repeat that statement. This is not a vanity metric that we're simply just trying to drive numbers up to say, "Hey, look how many." It's important that we're supporting those customers, that they're successful with the platform, that the end users of that data, you know, are happy and satisfied with what they have. We've scratched the surface of this opportunity. There are thousands of sequencers out there, you know, that are potential outlets for the technology. We're talking about 40 external sites. There's, you know, the potential of thousands still in front of us. We've taken an NGS-agnostic approach. We've talked earlier in the year about some collaborations we have with some of the next-generation sequencing companies. We wanna be wherever our customers want to be from a platform point of view. Of course, we have a very good relationship with Illumina, who, you know, launched some new technologies of their own, which is exciting. You know, we believe this all spells, you know, great news for our customers and for proteomics as it potentially reduces the sequencing piece of the cost of experiments in the future. Great opportunity for us there. It's really a similar picture in the mid-plex and low-plex space. We launched the Q100 just about a year ago. We began shipping it. You know, we have an existing base of BioMark labs out there, and then suddenly just over this past year, you can see we placed 63 units as of the end of Q3. Now with the launch of Flex, we expect a really nice tailwind behind the Signature business, so there's even more reason to adopt our mid-plex platforms. Again, we estimate there's at least 4,500 of these labs out there, probably in excess of that, and they continue to grow. Again, if you look at 103 in the context of thousands, you know, that speaks to the opportunity that still lies ahead of us. Then finally, world-class execution. Importantly, you know, so how are we doing it? I'd say we're doing it in an aggressive but responsible way. You know, we continue to invest in commercial execution, all aspects of it across all regions. We're seeing growth across the board. You know, of course, we're doing what any responsible company should be doing. We're growing aggressively, but then we're leveraging the value of those employees and that capability, so we don't need to keep doubling and tripling those commercial investments. We can get plenty out of what we're doing and yet, and still clearly we're gonna continue to invest to expand those markets. As quickly as we can. One of the key investments we've been making over time, and it's worth pointing out, going back to the comment on making sure that our customers are successful with our platform, is investing in our support organization, right? To ensure that our customers are successful with the platform, and that again, they create happy end users at the end of the day. Finally, as you see, we are sort of expanding the portfolios and bringing into more application areas. We're looking at how we're now able to, now that we've sort of grown and created a really strong commercial base, how we think about expanding into different parts of the market and really focusing on that, be that the clinical translation space, population biology, pharma applications, et cetera. I think a great reflection of sort of how, you know, we're executing commercially can be seen again in these numbers on customer loyalty. I think this is a really important metric to look at in terms of not just getting more customers and then seeing attrition of those customers, but gaining customers and keeping them over time. We have a loyal, customer base that tends to repeat over time. Then as I showed earlier with the customer numbers, continue to sort of grow on that base over time. We're very proud of what this means. This reflects how well, how much value we're bringing back to our customers and how we're supporting, their success. I could go on. There's, you know, plenty of metrics to speak about, but, you know, I'd like to leave you with a few things. I think our commercial execution, financial execution is excellent, and we see 59% growth, right, most recent quarter, that we've shared. Our strategic execution is on point with what we've said. You know, posted good numbers around our internalization efforts and the kit mix of our business. Finally, from an innovation point of view, right? I just shared two, you know, significant product launches that just happened over the past month. You know, I believe they are the things that innovation is the lifeblood of our company, and that we'll keep ahead in the years ahead. Finally, the whole team is strong and confident, and I think exceptionally optimistic about the future of not business, but, you know, the impact that I believe we can make to science and humanity overall. With that, I'm going to end, and we are going to take a short break. We will come back at noon, and then, Ida, Chief Scientific Officer, will start some of the interviews with some of our KOLs today. Thank you. Ladies and gentlemen, we'll start in two minutes. Great. We're back. I have the great honor to chair the next session, where we gonna get to talk directly with top thought leaders in our space of proteomics. First off is Professor Claudia Langenberg. Claudia is the director of the Precision Healthcare University Research Institute at Queen Mary University of London, UK. She's also professor of computational medicine at the Berlin Institute of Health at Charité, Germany, and visiting senior fellow at the University of Cambridge, UK. Claudia is a public health physician by training. Her research is focused on the genetic basis of metabolic control, and her team studies its effects on health through integration of molecular with clinical data in large-scale patient and population-based studies. Welcome, Claudia. It is great to have you here. Thank you so much, Ida. I can't see myself yet, but I see. I guess yes. That's better. We can see you. Yeah. Thanks so much for the kind invitation. It's a real shame that I can't be there in person, but a real pleasure nevertheless. You've already introduced me kindly, maybe can we go straight to the next slide? Yes. Perfect. I hear a slight echo, but I hope that's not the case for you all in the audience. I just thought, before I start, it's useful to give you a bit of background on kind of, you know, why proteins, why proteomics, and why we are excited about this. It's really the little inset graph that matters here. You know, what it shows you is, over time, how the number of people that we've included in our genetic studies have led to an increase in the genetic discovery or the strong correlation between how many people we include and how many robust findings we get out of our genetic association studies. Really even though this timeline now stops a few years ago, there's really no sign of slowing down at the moment. In 2007 is when I entered the field, when it really this graph starts. I like to always point out in 2013 is when I went on maternity leave, there's a little dip. 2014 is when I came back and it really accelerates. Obviously, that's just a joke, but it really underpins the massive success that we've had in identifying these robust gene disease or gene phenotype associations. The problem is that this huge success has not really been coupled with a deeper understanding of the mechanisms and the causal genes that underlie these kind of hundreds and thousands of associations that we've detected. Really that is the point of why we integrate what we call the omics. From the genome to the transcriptome proteome metabolome to what you could call the phenome or the diseasome, really a large quantity of different disease representations across almost all clinical specialties that we can now integrate based on bespoke studies or electronic health records. It's really this for the first time helps us to identify the mechanisms that link the genome to the phenome or diseasome, and in my view, presents the most comprehensive way to systematically screen for causal genes and variants, identify new potential drug targets, but also look at the very widespread effects of existing drug targets, which can lead to repurposing opportunities or anticipate adverse events that could never be really looked at at scale in trials that look at primary or secondary events. Next slide, please. I just wanna exercise this. I don't have much time, just illustrate this with one recent example, and actually both examples I show you, they're not really published. This is already a preprint, so if you wanna have more detail. But really the study for our understanding is relatively small. It's only it's under 2,000 individuals. It's just over 1,000 actually. But the number of proteins that we screen is very large, so almost 3,000 protein targets based on the Olink Explore and Olink Target 96 Expansion panels or assays that are now available. It really shows you the degree of how you can integrate different types of data and information to make sense and not just explore the underlying mechanisms, but create robust links to disease and to actually not just one disease, but many diseases. We systematically look and identify genetic variants that regulate the levels of these proteins in our blood. This is, you know, blood is our preferred matrix. It's easily accessible, can be done at scale. Then we use this information on, you know, the genes or the protein encoding genes and overlay from a whole variety of sources, information on the genetic determinants of a whole variety of diseases, and we integrate all of this together. We can also use this information and look at in databases of sequenced individuals where the rare variants kind of converge with the knowledge that we obtain in these, what I would call proteogenomic approaches. Finally, of the first graph I showed you of the many, genome-wide association studies that often use the nearest gene or, you know, a gene, postulated, hypothesized to be the underlying causal gene for the kind of observed statistical association. We can then go and take a step back and say, from the protein encoding gene and the robust links that we identify to diseases, are these hypotheses likely to be correct or can we refine the list of causal genes or even pinpoint with great certainty the causal gene for any of these previously published reported associations that overlap with our studies? Next slide please. This is kind of a fancier view of what I've just described to you. What we've done is kind of, you could call this like a proteogenomic map where we really take, you know, in this case over 200 robustly identified variants that determine protein levels. It's the cis-related variation directly from the protein encoding gene to any of the health outcomes or diseases across any specialty, as you see the color coding here, identified in this one study, just, you know, based on just over 1,000 individuals, but integrating data from a whole variety of other databases and robustly linking these variants we've identified to over 500 different health outcomes. What does this look like? Maybe to make it a little bit more tangible is the plot on the right. If you now zoom into one of these examples, so you go straight in and this is the genomic region encoding that sits around the variant that we identify. What you can see is kind of with the color coding at the very bottom is just the link that we identify from the variant to the protein level. Now you see in blue. The opposite direction is where we see an inverse association with the risk of type 2 diabetes, i.e., the variant associated with higher levels of this protein in blood protect from the risk of diabetes or the development of diabetes. Likely, as you can see in all the other integrated plots in the middle through a reduced amount of fat in different body compartments and reduced level of BMI. Now, that's very interesting. Nobody has really focused so much on this gastrin-releasing peptide. There is kind of, I don't wanna call it anecdotal, but kind of older evidence of an effect on satiety and weight loss in mice and also some kind of smaller human studies. There's also in the literature evidence of an additive effect with GLP-1, which is already used for the treatment of type 2 diabetes for dual agonism, but this is from rats. Altogether taking this human genetic evidence together with the published literature points towards that it might be worthwhile assessing GLP as a target for the prevention of type 2 diabetes and weight loss or influence on satiety. It's just one of many examples. Then I wanna end with a notion. This is kind of mechanistic insight, just trying to understand, you know, links between genes and disease. Next slide, please. Next slide, please. Thank you. Another important use of the technology in my mind is for disease prediction and indeed not just prediction of onset of disease, but prediction of complications. I yeah would refer to this as prognosis. Just on Thursday, Julius, the student who's led this work, had a publication in Nature Medicine on how a 3-protein predictor can improve the prediction or identification of individuals with isolated glucose intolerance, and this is, you know, a severe co-condition, a post-challenge hypoglycemia. It's not detected by existing tests like fasting glucose or HbA1c. If it's isolated, it's extremely common globally, specifically in certain ethnic groups that are at high risk of diabetes. Very important to find these individuals which otherwise go undetected and present with kind of microvascular complications of their high post-challenge glucose. You know, sadly, most of us spend the majority of their day not fasted, and hence a huge amount of time our vessels are exposed to this. That's important. Because of the success, that's all published, and you can read up on this, because of the success, we wanted to expand this and now say, if this works for this condition, which is kind of pre-diabetes, can we do this for a variety of different diseases? We took 23 different diseases and premature mortality as an outcome, 24 outcomes, really as a very small scale preliminary proof of concept study, and these are the results. I've just picked out the nine outcomes where this worked very well. What you can see in each of these plots is just how well do five proteins, the top five proteins, you know, not selected by us, selected by the machine through a machine learning algorithm where we have an independent holdout set and then the separate discovery set, how well do these five proteins only do in predicting the onset of these different diseases? That's what you can see here. They do really well. You know, not for all diseases, but for the majority of diseases. Importantly, in the second row, we compare this to how well do you do with existing risk factors such as age, sex, BMI, and smoking. It's important. It seems so simple, but it's important to reiterate because many genetic prediction studies never assess themselves against this benchmark. Now, finally, the important question which the little bracket shows you is how much better do we do with this existing cheap patient-derived information that we have in the health record? How much better do we do if we add the five proteins to that? It's often really hard to do better than this. Actually, as you can see here, the proportion of people, this is a Net Reclassification Index, so it's the proportion of people which are assigned to a more appropriate category, either being at risk of developing the disease or not being at risk, is substantial for some of these outcomes. I just wanna leave you with, you know, a bit of a word of caution, but also excitement. If you take type 2 diabetes, the huge proportion that you see here, that clearly is an exaggeration because we haven't assessed ourselves because purely in this study it was impossible against the gold standard which will be including let's say glycemia, glucose, HbA1c. If we did this, of course the improvement wouldn't be as big by the proteins that's one example. It depends which predictors are available. You can look at prostate cancer, and there it's PSA is already a marker that's used in the kind of follow-up for prostate cancer. It's not really a screening test certainly in the UK, but that's a proof of concept because the 5 protein model includes this existing predictor. Really not knowing anything about the data or the biology, the machine learning algorithm detects what we already know clinically and use to manage. You pick out two other examples that you really, premature mortality or lung cancer, you know, outcomes that are severe and really you do not want to get. You see the substantial proportion of people reclassified. That's, I think this is only a small pilot study. It's a start. Clearly, there are larger efforts underway that will substantiate that and also allow us to look at the prognosis of different diseases. I find this extremely exciting and look forward to kind of using this, you know, in the future across a whole multitude of other diseases. Final word for polygenic risk scores is the third line included here. This, in this example, the polygenic risk scores really do rather poorly for all of the diseases I show you. That's not the case for all, so there are selected examples where they do better, such as breast cancer. It's not included here, but nevertheless, in compared to the dynamic proteomic profile that you can assess that also reflects, of course, the existence of health status on the proteome, the comparison is really almost on a different scale for the majority of 24 outcomes that we've looked at. I'm gonna end here. I could talk about this forever, but I know we want to have time for questions. Thank you very much, and I look forward to discussing this. Great. Thank you, Claudia. That gave us a really good introduction and overview of how you see both proteomics and proteogenomic contributing to your research. You actually answered many of the questions that I had prepared, so that was good. That saved us some time. But then I prepared some questions, and then I'm gonna make sure to leave some questions for you and then encourage you to take the opportunity to ask directly to Claudia. I'll go first. In your view, historically, what have been the main key technical challenges for doing this type of large scale, high quality proteomic studies that you need for your research studies? Also which key developments and improvements would you say have played the most important roles to address these challenges? Yeah, I mean, it's a good question. I mean, I'm a molecular epidemiologist. I think it's fair to say, you know, we love large numbers, and what the genetic studies have shown us how much we do need large numbers. Certainly, some of that is not quite as true for proteomics, but I would say the ability to have a high throughput method, so, you know, being able to run a large number of samples, that has been a stumbling block previously. Together or coupled with that, you know, certainly for historical studies where samples are stored, and they're extremely precious, you need a technology that's mature enough to have the specificity to justify using the samples for this in this context. High throughput, you know, broad capture, proteomics with a specificity. Also importantly, I mean from my perspective, that is sample sparing because as I said, these are precious samples. Many of them have been sitting in liquid nitrogen for over 20 years. You want to use them when you know that you can generate data, particularly if it's a very, you know, new and expensive technology that would benefit for research for the years to come and not just. If you have the choice of the same volume for measuring three proteins versus 3,000, it's kind of, you know, it's a non-question. With these things haven't really come together previously at the same time, and I think for me then that has been coupled with the development of technological or computational advances that really now allow us to to identify the most informative protein signatures in an agnostic manner. Using machine learning approaches and really doing that. It's a very exciting time for these things to have come together. It hasn't really been possible previously. Great. Great answer. Thanks. In difference to our genome, so the proteome is very dynamic, and we know it changes over age, gender, lifestyle, health and so on. How do you think this difference will have an impact on future biobank and cohort initiative? I'm thinking more of this longitudinal monitoring versus single time point. How do you see that? Yeah, I think it'll become more important, certainly in comparison to the genome. That's very clear. I mean, I think the reason why I believe what we have seen where the proteins are very predictive or prognostic is because they are dynamic. As I said, they change as a function of our disease and underlying etiology, so that's a strength. I think the ability to test the effect of interventions on this dynamic change is also important because we have this intermediate measure where we know the protein is causal for a given disease using these genomic approaches I've outlined. We then have the ability to go back and test which interventions actually outside pharmacological interventions also affect the levels of these proteins in blood. That's very exciting. Kind of a whole variety of different study designs can come together. I think there are some challenges as well. We're interested in the biological changes, but of course, as you've just said, there are also technical influences, circadian rhythm. I get these questions a lot. You know, how can we see anything given how, you know, volatile or changeable dynamic the proteome is? But I think the data that we have seen speaks for us, and the ability to integrate it with genomics really has helped us to overcome some of these challenges. I should say the request to remeasure will be there. The need for repeated samples will be there. I think for proteomics, it's not that there's kind of a single test that you do once, and then you archive it in your electronic health record, and that's there forever. That's not the case. Of course, that comes at a cost to have to do the repeated measures, but that's also the strength of it because if it wasn't reflective of the changes that occur, it wouldn't be as strong a predictor or prognosticator, I think. Yeah. Exactly. You gave beautiful examples on proteomic risk scores versus the more traditional polygenic risk score. Maybe you can help us to understand how we should think about the different usage of them, maybe more moving forward. Yeah. I mean, again, people have very strong opinions about polygenic risk scores. I'm not one of those people. I think I'm open-minded. I think what's important is. I guess, yeah, I'm not a skeptic. I think I see the clear utility for specific outcomes, such as I've mentioned, breast cancer, and there are trials underway how you can more efficiently design screening around this. I think there is clear value, and it's a simple and cheap test that could, in theory, just be assessed once. Of course, we know that. As people sequence, that will also change the way we view this. However, it doesn't change. It's not dynamic. It is, as you saw for the examples I'd shown you, it's not a very strong predictor, and that's the bit that I feel is not as obvious. The understanding is not as clear. It doesn't predict as well and for vast majority has not been assessed compared to existing predictors. Where it improves over and above is often not assessed, and for the vast majority, it won't have any added value for most diseases, is what I predict. That doesn't mean that there aren't very good use cases where it should be done and would justify the small cost. Great. Yeah. Exactly. We're gonna see if there are any questions from the audience before we. Yes, Puneet. Hopefully you can hear me. Puneet Souda, SVB Securities. Thanks for the overview. I think the question from investor community point is, you utilized PEA technology, Olink Explore, and I mean capabilities to narrow down to some actionable disease markers. Sort of can you talk to us about sort of how frequently that you need to do this and how repeatedly you're gonna continue to utilize this technology in sort of other projects that are diabetes-related and metabolic-related disorders? Once you do get to these targeted markers, let's say a list of five markers, are you still, is it the time to go to ELISA or is it still something where you would prefer to utilize the PEA technology in a small form factor? Yeah. I think, I mean, certainly for, from my perspective, the predictive analyses and the derivation of the kind of, as you said, sparse models is at an early stage. Not much has been published across different diseases, so the next stage is not to go to an ELISA. The next stage is to validate in different, you know, validate this in different populations. Certainly the case. People haven't even done it, yeah, as I said, broadly across many diseases. We've done this across hundreds. We will need other large datasets to do this. I think there are many, many different approaches one could take to do this. This is possible, but it will first need to be done. Even once it's validated for a specific disease, the generalizability across different population subgroups is also important. All of these are, you know, important steps as you would have with any other discovery that need to be done. I think the next stage is not to straightaway go to that. Of course, I've in the interest of time and also because this study was relatively small, focused on sparse predictors. Of course, for some diseases, larger predictors are even more strongly predictive of the outcomes. That's not the case for all, but certainly we see examples for that. Does that answer the question? Yes. Great. Claudia, we are already out of time, but I wanna thank you for your taking the time today and join us here. It's a pleasure. Thank you very much. Thank you. Bye, Ida. Bye. Great. I think we're gonna go straight to the next KOL. I can start. We're gonna have the pleasure of talking with Dr. Anders Mälarstig. Anders is director in the Target Sciences unit at Pfizer Worldwide R&D, with responsibility to develop and execute strategies in target discovery and biomarkers for precision medicine. Anders is also a senior researcher at the Karolinska Institutet, with focus on integrating genetics and proteomics for target identification in immuno-oncology. He's the chair of the SCALLOP Consortium. Is Anders online? Great. Hi, Anders. Great to have you here. Hello. Thank you so much. Excited to be here. Perfect. I'm gonna go straight to the question. Anders, you were one of the very early adopters of our technology. I'm just curious, from your perspective, can you take us through this journey and, I mean, elaborate a little bit how you've been applying Olink's platform over the years and maybe also see what you see has been the major research achievements on a high level since the first experiment to where you are today? Yes, for sure. I still remember the excitement that we had when we received the first data from Olink. So it's some time ago now, but the dataset at the time were, I think 992 proteins in 3,500 patients. That was a significant amount of data, and especially at that time. The journey really sort of began there and then because we saw that there were proteins on that panel that had been well-researched in biomarker research, like predictive or diagnostic biomarkers like NT-proBNP or IL-15. We saw that these proteins show sort of all the expected links with disease that we well would have been worried if we hadn't seen, but we saw all of those sort of relationships and that was great, of course, because, I mean, up to that point, we had done single protein studies, and then all of a sudden we had this fantastic tool at our hands with multiplex protein measurements and in also large clinical samples. That was super exciting. I guess the other thing is that we also identified these so-called cis-pQTLs, which you might ask a bit more about for a large proportion of the proteins that we have measured. This, of course, was a great testimony to the specificity of the assay, and we could sort of say that, yeah, it picks up what it says on the label, basically. Yeah, I remember that study as well. That was a big thing for us back then as well. You were not only an early adopter of PEA, but also the pioneer who first introduced us to the field of proteogenomics and pQTLs, which led to the foundation of the SCALLOP consortium. Maybe take us back to 2017 when you reached out to me, I remember, and asked for other users that you could collaborate with, and then the hope was to form this global consortium. What made you see the potential in using our technology for this purpose? What was the main objective back then? I mean, the driving force behind the SCALLOP consortium was really to allow us to do better and larger studies by joining up with other collaborators with similar datasets, you know, which had both Olink proteomics and genomic information. At the time, there were not that many studies, but we did do an early pilot to see if these relationships between either a clinical factor and proteins or genetics and proteins, if those sort of were consistent in different studies. These were actually, I mean, these were Olink data on completely different sets of clinical cohorts generated in independent labs and so. We found that especially the genetics to proteomics integration, so the pQTLs were really quite consistent across the different studies and that I think really supported the idea that this is something that is very scalable. I think really sort of journey began to expand SCALLOP and invite additional PIs to work with us. Of course, I mean, it is also a numbers game because, I mean, the larger datasets we have, the more subtle relationships we can pick up. These may still be very meaningful relationships, but maybe you can't find them in a small study of a few thousand. That's why we sort of saw this need for scale. Mm. Yeah, I think that I mean also connects to a sort of a good research example and sort of what we have been able to do with Olink data. It has really opened up this field of pQTLs as a research field because without multiplex proteomics this actually would not exist. In turn, I mean, the SCALLOP results and what we've had there I think has really helped us to develop a better understanding of disease drivers at the molecular levels. Yeah, it's all of that. Great. Maybe now it's been 5 years, can you give us an update on what SCALLOP is today and share some highlights maybe? Yeah, we began in 2017 and now 5 years on, I think it's fair to say that SCALLOP is a global network with a variety of different sub-projects going on within the consortium, each led by different PIs that are part of this network. We have a sort of well-functioning structure for managing the projects and for starting new projects as well. We keep welcoming new members into this network, and currently there are actually over 45 different clinical cohorts and approximately covering data on 65,000 patients or healthy individuals within the sort of same umbrella. It has really sort of snowballed over the past. I would say 3 years. There's a lot of excitement also now that, you know, we can tap an even larger proportion of the proteome with the Explorer assays. Just thinking about sort of some of the, some of the highlights and, I mean, I think. We were probably the first to show that most proteins in fact are under genetic control, around 90%. You remember that number? We've been discussing that quite a lot. But also that we can really use these insights that we get from pQTLs to basically sort of turn it into a target candidate sort of engine. We're generating hypothesis based entirely on human data. I think that's of course super exciting. We showed that, I mean already a few years ago, but I mean, that work has really continued. We have other examples of where this sort of systematic pQTL mapping and causality testing can really provide retrospective evidence for targets we already knew about. That's of course good because it shows the concept, but at the same time, we find new strong leads. I think that's one of the highlights of what we've accomplished within SCALLOP. We've done other things too, like mapping the genetic regulation of the coronavirus receptor ACE2. We're probably one of the first groups worldwide to have the scale necessary to understand all of the different contributing factors to why people, for example, have different levels of this important receptor in the circulating blood. Great. Then speaking of pQTLs, it's such a powerful but also quite complex concept. Maybe you can in layman terms describe the differences and values of using cis versus trans pQTLs for drug discovery purposes so we can understand why to go with the both. I will certainly try. I mean, the instructions for making proteins that is really in our DNA. The other thing, each protein that can be measured with Olink or any other technology is really encoded by a gene in our DNA. The cis-pQTLs, these are the variants in our nucleotide sequence, in our DNA sequence of the gene that encodes the protein. That's why we call it cis because it's the gene that encodes the protein but also harbors the variants that are related to the protein abundance. That's the sort of starting point. Trans-pQTLs, they are located in other genes in the genome. We like cis-pQTLs. This is important to understand, I think. The cis-pQTLs are really critical tools for doing the causality testing. We use them to identify proteins that are causal in disease and as such, great drug target candidates. Now, the trans-pQTLs can be biologically interesting, but there we often run into more difficulty in trying to interpret, so how is this gene connected to the protein? Whereas for a cis-pQTL, that relationship is much more clear, and we're basically using cis-pQTLs as models of pharmacology and that's how we infer causality. I think that's. Yeah. Happy to elaborate more, but that's my sort of short explanation. Great. I guess the take-home is that both cis and trans are valuable, and needed. If we then move into your role at Pfizer, you're part of the UK Biobank Pharma Proteomics Project, which is the largest proteomic study run to date, as we know. What effect do you think this type of large-scale study will have for future studies? I guess it's before and after this study almost. Yeah. No, we are very happy to be part of the UK Biobank Pharma consortium. I mean, overall, I think, as I say, it marks a sort of change, a significant change in the field. The data that are generated within the consortium will likely yield a lot of new insights across a really wide array of questions for several years to come. The most obvious use, I think, or the sort of low-hanging fruit, maybe a better way to put it, for pharma and biotech will really be to run the systematic Mendelian randomization and colocalization studies to find this causal biology and generate target hypothesis. I mean, that's sort of happening already now, and I think that sort of activity will only increase. At the same time, I think the UK Biobank data set can also be seen as a sort of really big baseline study. These are apparently healthy individuals mostly, or it's a mixed bag of healthy individuals and those with disease. Ultimately, I think what we are interested in are proteins and protein profiles in patients and we also have longitudinal sampling, ideally relatively dense, so we can sort of follow changes over time. With the UK Biobank as a sort of starting point, it will be easier to understand what sort of what deviates in patients compared to a general population. If we focus a little bit more on your work at Pfizer, can you give us some example on how Olink is currently being used at Pfizer? Yeah. I think, I mean, certainly, what I've just mentioned with systematic causality testing, that's certainly sort of one big use and important use as well because it's one of the few tools that we have where it's a sort of human data first approach to discovery and as opposed to running in vitro or preclinical models, this is like human first approach. That's certainly an important use of data for Olink and pQTLs in particular. In addition to that, I mean, there's also the really exciting space of profiling patients in clinical trials, especially important likely at the early stages. I think that's another great example of where a better understanding of the proteome in patients and how it changes over time can sort of help us provide basically better treatments at the end of the day. Great. Then over the years of using PEA, would you say that among you and your colleagues at Pfizer that your mindset has changed around applying proteomics in both the drug discovery and for clinical trials? If so, why and maybe in what way has it changed? Yeah, I think, I mean, firstly, targeted proteomics such as Olink, I mean, has become this new great tool that we didn't have before. That's the sort of starting point. The awareness of Olink within Pfizer has certainly increased and also what the method can actually deliver. I mean, that awareness has also increased. Yeah, I think overall that's a big change and a great start. The greater coverage now with also with Olink Explore has also sort of made a huge difference, I think, because, I mean, in people's minds, we can now capture a larger variety of biology, and I think that's an important aspect. When you think of drug discovery overall, I mean, it's a challenging endeavor. My sense is that, I mean, both Pfizer and industry will make good use of any tool that can really help de-risk a discovery program, 'cause I think that is a lot what it is about. There is much more awareness also that you know with dipping and really analyzing the proteome can sort of provide a sort of de-risking decision-making kind of data. Based on your experience and expertise, do you think that proteomics will be implemented more or less like a core strategy across all pharma in the future, as we maybe have seen with it on the genomic side, that we will do the same with proteomics? Yeah. Well, I think in the preclinical space, it should basically have been implemented already. All companies will need to have a strategy for how to best utilize proteomics as a discovery. Yeah, like a discovery tool, or discovery and validation of your hypothesis. I think with the trial space, there will be certain disease areas that will hugely benefit from using Olink, for example, as a stratification tool. Probably all disease areas and any drug would benefit from understanding all the potential downstream actions of a drug. Perhaps there are some disease areas where you see even more benefit, like infectious disease, for example. I think is one of those that will benefit the most. Yeah, I mean, in thinking about the various omics technologies that we sort of have at hand, there's of course genomics, proteomics, metabolomics, transcriptomics, and I think of those, it's probably transcriptomics that's the closest to being a sort of general tool that most or everyone has adopted. Genomics is probably a little bit more long-term. It's a different nature. To speculate a bit, I think that proteomics has a really good chance, and Olink has a good chance of becoming very sort of part of the standard toolkit. Great. I have some more, but I'm just going to check if we have any questions from the audience. Nothing online. Yeah. Great. We have one question here, Anders. Anders Yes. Thanks for all the insights. I just wanted to get your opinion on how you think about sort of concordance across different platforms. You know, Jon and I were talking about this earlier as well, around, so just curious as to your thoughts on NPX, and do you think the industry at large is at a stage where that's now quickly becoming the accepted standard? My second question was, you know, to your point on proteomics helping sort of pharma in terms of the preclinical space relative to in vitro or perhaps, you know, animal models, do you envision a future where animal model sort of testing just goes down? I mean, there's the FDA Modernization Act, et cetera, which contemplates some of that, and does proteomics sort of play a critical role in that future? Yeah, great. Great questions. So yeah, I think firstly, on the different platforms and our, I guess, still somewhat limited understanding of the proteome, I think it has just become clear that different methods seem to pick up different parts of the proteome and measure different aspects of the proteome. So far, I think we've sort of seen the most alignment with the antibody-based methods and what we can actually do in vitro a lot due to that we actually use antibodies as detection in Western blot and other sort of standard lab technologies. Then you have this sort of close alignment. I still think that there is space and interest for looking at proteoforms or looking at post-translational modifications, and it's possible that other methods can sort of pick that up as well. I guess the sort of boring answer is that probably more research needs to be done to understand the different proteoforms and exactly what sort of different assays can also pick up. So far, I think at least, you know, our experience is that with the Olink assay, we see a strong concordance between the gene variants associated to protein abundance. So we see that clear connection at least. Yeah, it's a great question. To the second question, I think the hope is really that we can move away, at least from preclinical models as a sort of discovery tool. I think there will be a need for sort of back translation at some point because having access to good models, both preclinical and in vitro, are sort of enablers to take a drug program further. Yes, I really think that this sort of approach for finding our target candidates is going to reduce the need to model disease in animals, which would be great. Anders, unfortunately time is out already, but thank you for taking the time. Very much appreciated. Thanks, Anders. Thank you very much. Thanks. Thanks. Okay, so now I think we're ready for a break. We're back in. One. At 1. Okay, great. Ladies and gentlemen, we'll resume in 3 minutes. I can see the fire in your eyes. Great. Welcome back. We have three more KOL sessions, and I have the pleasure of introducing the next one, Professor Charlotte Teunissen. Charlotte is full professor in neurochemistry. Her mission is to improve care of patients with neurological diseases using body fluid biomarkers. Studies of her research group span the entire spectrum of biomarker development, starting with biomarker identification, followed by assay development and clinical validations, and implementation in clinical practice. She's responsible for the Alzheimer Center Amsterdam Biobank. She leads several international biomarker networks, such as CSF Society, Alzheimer's Association Global Biomarker Standardization Consortium, the MIRIADE Marie Curie project, and the Coral Proteomics Consortium. Welcome, Charlotte, to our Investor Day. It's great to have you here. Yeah. Nice to see you too. Yeah. Great. You will first share some slides that give us an introduction to your research. Please, go ahead, Charlotte. Yeah. Thank you very much. I just want to give a brief outline on how we perform our research, and the best illustration is a paper that we just published last Friday, which was a large body of work, and I will show in the next slides what we did. Our main aim is to develop novel biomarkers in for dementias by the cerebrospinal fluid, which is the brain fluid that immerses the brain. We also have a focus on blood analysis. Can I have the next slide? In our study that we just published, our goal was to understand the CSF proteome changes in AD, and we did so by analyzing a large number of CSF samples of AD patients in different stages, and also in other dementia patients, we call them here non-AD dementia, and in controls. You see in the yellow brain MCI patients who are in early stage of Alzheimer's disease. In total, we had a cohort of 797 individuals of whom we had CSF samples, and we analyzed these samples using the Olink PEA technology, including more than 900 different proteins. Our first aim was to identify and map the changes in mechanisms by analyzing the differences in proteins between the AD patients and the controls or in other non-AD dementia patients. You can see by the pictograms below the graph that we identified several mechanisms that were changed in AD patients compared to the other groups. It helped us to identify those different mechanisms and how relevant they are in Alzheimer's disease. Can I have the next slide? What's more, we identified panels because we are interested in ultimately translation of our biomarker findings to clinical use. We want to use those different proteins and biomarkers to help the patients in their diagnosis and differential diagnosis, and also for prognosis. Here you see the results that we're very proud of. We, in the proteomic study where we started off with, identified a panel of only eight different proteins through which we could optimally discriminate AD patients from healthy controls. In such diagnostic analysis, we take advantage of using what we illustrated by areas under the curve, and an area under the curve of 1 is the max that you can get. Here you can see that we were almost at that high area under the curve. We here obtained area under the curve of 0.96. We next developed, together with Olink, custom panels, so for those 8 different biomarkers out of the 900 that we used in the discovery panel. We validated those custom panels of 8 proteins in an independent cohort of patients, so again in CSF samples. We here could confirm our really very high area under the curve, and now it was even higher, so 0.99. Can I have the next slide? In parallel, we are interested in translating those different panels, here a panel for discrimination of AD from non-AD, to use in clinical practice. For that, we developed single assays for each of the biomarkers within the panels. We were successful to do so, and already within one year, we obtained assays for five out of the nine panels for this panel that we use for discrimination of AD from other dementias. That was an unprecedented success that we never obtained before with mass spectrometry analysis, and it was a confirmation that our strategy is the right one. To use Olink proteomics for the discovery that allows us for downstream later on validation of our findings into using more practicable assays. That's what I wanted to illustrate you today. Great. Thank you. Happy to answer any questions. Great. Thank you. I should also take the opportunity to say congratulations to the beautiful paper published last Friday in Nature Aging. Charlotte, to start off, can you describe why your field of neurodegenerative disorders has had such significant challenges with identifying new robust biomarkers and also developing successful drugs, especially maybe Alzheimer and Parkinson disease? That's a very good question. It's really difficult in our field, because we cannot take a biopsy from the brain. The brain is hidden by its skull. Therefore, we have to rely on alternative methods and also very sensitive methods. We have to measure our biomarkers preferably in the brain fluid, in the cerebrospinal fluid. But that's a little bit more difficult to obtain than, of course, a blood collection, as you can envision yourself also. That's really hard. Due to this brain being encapsulated so well within the skull, we're also not able to really look in vivo what's going on in a tissue, so what's going wrong. That's also really relevant because if you know what's going on and what's wrong in, really in the affected tissue, then you're able to identify the drugs and try those. Another challenge is also that there is a relative lack of funding compared to the cancer fields. There has been far more investments into the disease in itself than in the brain diseases, and that's also why we're a little bit lagging behind. Great. A follow-up question to that. What do you think is needed to change this negative trend, and how are you addressing these needs? Well, for the biomarkers, I really am very convinced by our approach. One of the approaches is using the Olink, so it looks like that I'm advertising now. That's not my purpose, but I'm very enthusiastic and also because it's supported by our recent results. In the past we really had a lot of problems when we did our discoveries within the cerebrospinal fluids using mass spectrometry also because it's a very labor-intensive method so we could analyze only small sample sizes, so small cohorts of patients, like 10 or 20 in each group. You can see from our study that really a large scale is needed to obtain more reliable results. In our example, it also shows a very good proof of concept of that using a large cohort and then technology that is amenable to make smaller panels. This combination leads to more success. We need to do large-scale analysis, and people have to collaborate also for that to reach those large cohorts of patients that you can include in your studies. That's really helping to shape the future of the biomarkers. I'm just convinced by that. The next step is to go from CSF brain fluid to blood. That's just the next step. Now we have very sensitive technologies, not only Olink but also other sensitive technologies through which we can now analyze the really small, low concentrations of brain-specific proteins within the blood. Great. Thank you. In the MIRIADE study that you just presented for us, you described your work into two steps. The first, the broad discovery, and then the verification of the selected top markers. What do you see the strengths in this approach? Yeah, I do think it's a very powerful approach due to the large sample size that we can include. It's a nice intermediate step if you are ultimately interested in clinical assays. It's still a risk if you do a large discovery, you have to choose the biomarkers that you will work on downstream. Probably we also need to analyze multiple biomarkers for those complex diseases. Probably one marker is not enough. That's why we think that using such a multiplex approach goes along with the risk. In our stepwise approach where you go from the discovery next custom panels for the analysis and the development of single biomarkers. I do think that. Oh, now I've forgot what I wanted to say about it. For drug discovery, I also think that it's relevant to use this approach, proteomics per se, because it can help us to identify relevant subgroups within the population of, for example, AD patients. We've shown that before and published it before that, within the AD population, so a group of Alzheimer patients are subgroups of individuals that have a more inflammatory subtype and others have more a profile of neurodegeneration. It's conceivable that anti-inflammatory drugs that are also being tried in Alzheimer's disease would have a better effect in those individuals with an inflammatory profile. Using proteomics, we can identify such profiles. Now you already touch upon the clinical use, so I have some questions around the implementation of proteomics for clinical use in your space. First, how do you think proteomics will contribute to clinical trials in AD? Yeah. They will contribute, for example, as I said, the subtyping, so you can do a stratification, and by that an enrichment of the trial with individuals that are more likely to react to a drug. Again, the inflammatory subtypes who will react probably more favorably to anti-inflammatory drugs. The proteomics and the Olink proteomics are also very relevant for endpoint analysis, so to measure trial effects. There's always a target engagement of course, and the target is then usually identified before you start a trial. We also want to understand the downstream effects of a drug, wanted or unwanted side effects. The proteomics analysis can help us map those different biological effects in an individual. Because Alzheimer's disease, but also many other diseases, they are biological diseases, so the biological biomarker endpoint will become more and more important as a readout, mostly in phase 2 trials, but when we can show the surrogacy also in phase 3 trials. I mean, specifically for clinical implementation of new drugs, how do you see proteomics will play a role there? For novel drugs, they can help us identify the individual to track and monitor the responses of an individual towards a drug, and whether this drug is effective or not, or whether it's wise to change the drug. It will help us in defining the personalized profiles, biological profiles, and also personalized medicine. Because when there is just one drug, you have no other choice, but the future is rather that we have multiple drugs to choose in cancer and also in multiple sclerosis. There it's very relevant to know the early biological effects, and if there is a biological effect. Otherwise, this can help the decision making for changing to a different drug. Mm-hmm. You already mentioned a little bit of mass spec applications and other platforms. You have experience of using alternative tools as well. Maybe the audience will be interested also hearing how you see our platform compared to the others in respect to overall technical performance, and also maybe the flexibility. Yes. I can comment on that. There are differences between the different platforms in the types of binders that they use, so that's a technical difference, and in the number of targets. If you compare with, for example, with SomaLogic, they have also high-affinity binders and have a large number of targets, but Olink is also increasing their number of targets. Their binders, called SOMAmer, they are in my view not very easy tools to use downstream in clinical tests. We were not successful in using those SOMAmer in the more ELISA, so the single assay formats that we want to use. With respect to the mass spectrometry analysis, I do think that it's really problematic that it's a different platform and a different type of chemistry. At least we were, and with us also others, were not very successful in translating our CSF proteomics findings by mass specs to a clinical assay. Mass spectrometry by itself, it's not high affinity or the. It measures largely the high abundant proteins. We do think that we are in need of the low abundant proteins, so those with the low concentrations. For that you need a binder assay. Either antibodies like in Olink or with the SomaScan. My preference currently with the flexibility to build your own custom assays, and also that Olink has already panels that you can choose from, so you don't have to measure everything at once. If you're interested only in inflammation, you can choose only the inflammatory panels. Yeah, that's really the flexibility of Olink. And it's also really nice, I think it's a strong advantage that you can build your custom panels together with Olink. Yeah, in my experience it was also very nice way to communicate and a nice collaboration. We got good input. Yeah, very open way and I really experienced that having a good quality analysis is what Olink is interested in, and that coincides with our interest. Probably that was also part of the success. Thank you. I have one more quick question, then I'm gonna see if anyone in the audience have any. You're also the founder of, and the driver of the relatively new, neuroscience network, CORAL. Can you please tell us a little bit about this consortium? Yeah. CORAL is a consortium. It's like a network or a framework with the aim to accelerate the identification and of mechanisms and also proteins in the neurological disease areas. Ultimately we want to translate those findings and offer biomarkers to the clinic. It coincides with my own aims of my research groups. The focus is in neurological diseases. What we do want in this network, it started a year. We have meetings to share experience, but also to share samples and data, and to perform meta-analysis. For ultimately improving the quality of the studies, and by collaborating, taking advantage of the large data sets that people build within a neurological disease area, and to share those experiences to ultimately reach our goals, in a very effective way. Great. Thank you. We have one online that's very similar to what Charlotte's already covered, so we can see if there's someone in front. Okay. Any questions from the audience here to Charlotte? Yes. We have one here. Hi. Evie with Goldman Sachs. Could you talk about the switching costs and how you're benefiting from staying on the same platform as you move from discovery to targeted? Yeah, I think that's the core of the success, that you don't switch between platforms if you go from discovery to validation. I really think that's essential. Yeah, if I talk a little bit in detail, in mass spectrometry, you have a different type of chemistry. You cut your proteins in small pieces. Then next, we wanted to use immunoassays. Why do we want to use immunoassays? Because that's the type of assay that's present in routine clinical laboratories, and that's really common to use. Ultimately, we want to have immunoassays. Then you have different chemistry, so you look at the proteins not cut in small pieces, but they are in a 3D structure to those different parts of the protein. With Olink, we have the pair of antibodies already on their arrays, and a pair of antibodies, so two antibodies, is the holy grail in immunoassay development. You cannot use only one, that's less specific. That's also the advantage of Olink. The discovery also is highly specific for the different proteins. The pairs are already available. The proof of concept is provided that such binding reagents are there, and you can take advantage of that. Our study just published in Nature Aging shows that this works. I hope I answered your question, or did you mean something else? Yes. Maybe you should have asked earlier. You did, Charlotte. Time is up. Okay. Thank you so much for taking the time to be with us here today. Thank you, Charlotte. Have a good evening. You're welcome. welcome. Yeah. Thank you. Thank you. Have a nice day. Thank you. Now I'm gonna hand over to Carl to take the next Q&A session. I'll come up, Carl. Great. I think we're gonna have Lori without. Without video. Yes without video. All right. Great. All right, now we'll be joined by Lori Turner. She's Director of Strategic Alliances at Azenta. Lori, hello? Can you hear us? Hello? Lori, can you hear us? Hi. I can hear you. Sorry, it seems like I'm challenged today and having some technical difficulties. Oh, oh, fantastic. Thank you for joining us today. We appreciate it. Even audio is fantastic. Great. I'll jump right in. Lori, since you come from a commercial background, unlike a number of the KOLs we've had here today, would you mind providing a quick introduction on yourself, your experiences, and your responsibilities at Azenta? Sure. First, thank you, Carl, very much for the invitation to be a part of your investor day. Hello, everyone. My name is Lori Turner. I am the Director of Strategic Alliances at Azenta Life Sciences. I began my career on the bench, working to support the Human Genome Project, and I find it's pretty amazing to think that today we can sequence a human genome in a matter of hours to when the first genome took 13 years. It seems we're very close to hitting that $100 per genome mark. I previously worked at Beckman Coulter Genomics, a division of Beckman Coulter, which was acquired by GENEWIZ in 2015. In 2018, Brooks Automation acquired GENEWIZ. About one year ago, Brooks Life Sciences and all companies acquired were rebranded as Azenta Life Sciences. My role is to work to further Azenta's footprint within the medicine arena. My team is responsible for driving innovation by identifying strategic partners and technologies, onboarding those, then building strong relationships with our partners by closely working together to achieve our long-term goals and objectives. Great. Thanks, Lori. Can you tell us a little bit about Azenta's customer base? Sure. Azenta is a global life sciences company with locations in seven countries and supporting customers throughout all phases of therapeutic development, from research and discovery, preclinical and clinical trials, through to supporting diagnostic and manufacturing distribution solutions. Our customers are those that need a partner that understands the importance of sample integrity. We provide that one-stop shop solution for anything dealing with samples, from pre-analytical and multi-omics services to sample transportation and storage. In fact, we have the most extensive biospecimen collection capability in the industry with millions of customer samples in storage globally. Great. Impressive. Thank you. Lori, if we go back in time to 2021 when you were looking at the evolving landscape of omics, your existing portfolio, and your business goals, how did you think about sort of next generation proteomics as a potential fit into your offerings? You know, what were those drivers for Azenta to commit to that direction? Right. Okay. We had previously considered onboarding a proteomics platform to complement our genomic services with the goal to be that one of a kind multi-omics solutions provider. We conducted an intensive analysis of the market. In fact, it took us well over a year to investigate the proteomic platform market to determine what platform would complement our services business model, as well as add the best value to our biorepository customers. Most of the technologies we considered at that time did not fit our model in one way or another. Just as an example, mass spec would have been capital intensive and required a significant investment in personnel that just did not make sense for us. In late 2020, when the UK Biobank brought in the Olink platform to complement the whole genome sequencing of the participants of the program, is when we became aware of Olink. This was a major turning point for Olink, I feel, and brought the technology to the light of the greater genomics community. That's great. I guess in that decision, what was appealing to Azenta, in the decision to work with Olink, obviously, aside from the visibility from the UKB study? Right. Why did we choose Olink? Well, there are many reasons why we partnered with Olink, and I'll highlight a few. A significant driver for us to choose Olink was that your readouts are traditional genomics technologies, next-generation sequencing and quantitative PCR. Azenta has been using these platforms for more than a decade in many of our laboratory locations, so this allowed for a seamless integration into our already working service lab. There was no need to bring on a new platform or to learn a new workflow, very advantageous for a service lab. Additionally, we like the scalability of the platform. The PEA technology allows our customers to easily look at a single protein or up to 3,000 proteins. They can then bring a biomarker from discovery to clinic, again, using the same PEA technology. Last but certainly not least for Azenta, Olink has done a great job in doing a lot of the upfront work for us. Olink has validated every protein assay on both Olink Explore and Olink Target platforms, and Olink has qualified many different sample types. Real work that a service lab usually has to do themselves, which made onboarding the Olink platform straightforward integration for Azenta. Oh, great. Coming back to your customer base, can you discuss the appetite for multi-omics among your customers and the value that they, especially the proteomics aspect of that, brings to the equation? Yes, we have seen the trend where the general scientific community realizes that integrating various omics methods can generate insights that cannot be found from single omics method alone. Obviously, including protein data can reveal valuable functional information that's just unattainable from genomics alone. But, you know, proteomics has lagged behind genomics and transcriptomics since analytic platforms of proteins were not as robust as those for nucleic acids. Olink's technology has really shifted that paradigm. In general, among our customers, we have two basic profiles, those in research and development and the clinical and translational scientists. Our R&D customers seem to need education on how to incorporate proteomics data into their existing genomics and transcriptomics workflows. Our clinical customers are raising their hands, wanting to add Olink to their prospective and retrospective trials. These folks are familiar with including proteomics within their studies, and they see the value that Olink mid-plex assays bring. The addition of Olink to our portfolio has increased our customer satisfaction and retention because we can now bring them the complete and cutting-edge solution, in one package. Great. Appreciate that clarity. What areas of demand for proteomics does Azenta expect to see from its customers in the future, thinking sort of 2023 and beyond, including both, I think you've touched on this a little bit, but the high-plex and the low-plex offerings? Well, Carl, I'd say at least from the past year of Azenta offering the Olink platform, we have seen every facet of our customer base interested in bringing on proteomics. Our R&D customers are looking to use the high-plex Explore platform for biomarker discovery. Our clinical biomarker customers are looking to add target panels for their clinical trials as an exploratory endpoint. These clinical folks are thrilled with the mid-plex target offering. They're assaying not just one biomarker, but up to 92 with just a single microliter of their precious sample. I would say for future, we are very excited about Olink's new Flex offering, where our customers can pick from a large number of protein assays to create their own unique panel with an absolute readout. Also, for our discovery customers, they're very anxious to know when Olink will come out with the new 4K or 6K biomarker Explore offering. Great. What are some of the ways that Azenta is supporting customers who are expressing interest in expanding their work into proteomics? I'll share a couple of ways that we're addressing the demand of our customers' proteomics needs. Many of our customers are not that familiar with the specifics of the proteomic workflow, so we support them with any sort of sample sourcing, sample collection, or pre-analytics as needed. We're very comfortable in working with the traditional proteomic sample types such as serum and plasma, but we're also very familiar working with tissue and cells if that is what the customer would like to assay. On the back end, we support those customers from an analysis perspective, not only delivering the proteomic data, but also aiding them in visualizing their data and statistical analysis, whether it be straightforward or a more customized approach, including helping them with multi-omics integration analysis. Finally, we're all set to support the clinical customers by offering the entire proteomic workflow in a CLIA/CAP-validated laboratory. Fantastic. Last, what would you consider some milestones of success, that Azenta hopes to achieve with its presence in proteomics? Yes. Okay. Well, there are many ways to measure success, and of course, metrics change as strategies evolve. However, at this point, the way we would define success would be to bring the platform global so that we can support our customers that have global clinical trials. Also Azenta is a company that stores millions of clinical samples for our customers. Many of those samples already have genomic data associated with them. Our goal is to educate these customers on the value of adding proteomics to their dataset, through using the Olink platform. At the end of the day, true success would really be to have the demand for proteomics equal to that of genomics. Excellent. Thank you, Lori. Now we'll take a moment. I think we have time for a question or two from the audience. Yeah, we can try that. Nothing online, so just. Nothing online? Okay. We have 2 up here. Hey, Lori. This is Tejas from Morgan Stanley. Can you just walk us through which of Azenta's customers, you know, are using your service versus and how they approach the decision of bringing, you know, the Olink sort of platform in-house versus perhaps using, you know, Olink service lab? Then, my follow-up was, you know, earlier this year you guys had talked about sort of the average order size being down within genomic services. Was that mainly sort of Sanger and NGS sort of related, or is that a dynamic that you see impacting the proteomic side of things as well? Sure. Okay, the first question is in terms of customers wanting to either bring the proteomics platform into their lab versus the Azenta service lab. I would say, you know, for us, we wanna be that one-stop shop. If we have a sample that our customers are looking to have multi-omics analysis on that, whether it be proteomics, genomics, transcriptomics, epigenomics, then it really makes a lot of sense for those customers to use Azenta for that multi-omic analysis where it would include proteomics as well as genomics. It, you know, Azenta has been a service supporting services for greater than 20 years. We're fully aware of a lot of these customers who wanna bring on platforms within their own labs, and they're absolutely, you know, can do that. We offer that, our research use only as well with clinical services. One stop shop from an Azenta perspective. It just makes sense to send that out. Your last question was something about our sample numbers. We're seeing a great increase in our services that we're supporting our multi-omics customers on. I see that number only growing. Great. We'll squeeze in one more question from Puneet. Yeah. Yeah. One quick question please. This would be security. Just what do you think is the barrier at this point in time for customers or especially NGS customers to adopt Olink and the platform? Because obviously there is the familiarity with the mass spec, some familiarity with the mass spec side of things and the ELISA side of things. Just what is the barrier for Olink to be adopted more widely? Well, I think from a discovery perspective, a lot of these folks have their budgets in mind where they're bringing on transcriptomics, genomics, and now they have to kind of take a step back and say, "How am I gonna increase that budget to bring in proteomics?" Because since the Olink platform is really an intuitive platform for them, they now need to question how are they going to add that to their budget. At least from a clinical perspective, we see customers they know the value of Olink. They love the mid-plex being able to multiplex multiple proteins in a single assay. They're already expanding their budgets to add Olink. Great. All right. Lori, thank you very much for joining us today. Much appreciated. Have a good afternoon. Thank you, Carl. Thank you, Lori. With that, I'll turn it back over to Ida. Great. Thank you, Carl. Now we come to the last, but certainly not least, Professor Mathias Uhlén. Mathias is professor at the KTH Royal Institute of Technology in Stockholm. His research is focused on protein science and precision medicine, and ranges from basic research to more applied research, including clinical applications in cancer, infectious diseases, cardiovascular, autoimmune, and neurological diseases. He leads an international effort to systematically map the human proteome, creating an open access resource, the Human Protein Atlas, which is now one of the most visited life science databases in the world. Mathias' research has resulted in more than 750 publications, and he's also a co-founder of 20 biotech companies, of which 4 are publicly traded. Mathias, great to have you here, and thanks for staying up a bit late. Yes. Thank you, and I'm very happy to be here. Great. You're gonna start sharing some high-level overview of your ongoing research. Please go ahead and start. I will then just show you a few slides. Can I have the first slide please? I don't know, I can't really see it, but can you see the first slide? At the title slide, but I think you can go to next now. Now, Mathias Okay, perfect. I will talk about the next generation pan-cancer blood protein profiling that we have been doing using the Proximity Extension Assay. If I can have the second slide then which it says system mapping of the world. I just want to remind you that we are in a very exciting place right now and time in the history of science, where we've gone sort of from mapping the chemistry in the 1800s, the physics in the 1900s, and now we're actually mapping the building blocks of life, and medicine. We are doing that, looking at the proteins. The proteins are the building blocks of all life on this planet. We started, as Ida was saying, a Human Protein Atlas program about 20 years ago to try to create an open access knowledge base for all human proteins in cells, tissues, and organs. We spend about 2,000 person years on this, and we're very happy with the funding from a nonprofit organization. Can I have the next slide, which says building an international core resource? Just as Ida was saying, this is one of the most visited biological databases in the world. We have more than 15 million web pages, which is updated annually. We're very proud that we have used this resource to create more than 10 million tissue images and as Ida was saying, actually more than 750 publications. Every day, this atlas is cited more than, actually more than 30 times. So that's of course very nice. What we are doing now is moving very much towards precision medicine. This is I think one of the most important changes in medicine, in a very, very long time. What we try to do is to do a profiling to do early detection of disease, stratification of patients, and also monitoring of treatment. Next slide, which says, should say a revolution for blood profiling then is that we, as I said, we have been working with Olink Explore now for several years. We have started by looking at all the major cancers, so we have hundreds of patients for each cancer, and then we have identified the profile in these cancers from minute amounts of blood, less than 10 microliters. Then we use this, and I'm not gonna go into the details there, to use AI-based prediction models to actually say what patients have what cancer. Without going into the details, this will be published in a few weeks. We will then get a very good both precision and efficacy for these to determine these diseases. We are releasing a new version of the protein atlas in a few weeks, in December seventh, and it will then contain a human disease blood atlas, where you will be able to explore then the results from Olink Explore in different cancers such as lung cancer, colorectal cancer, breast cancer, and so on. With that, I think that's a very short introduction to what we are doing now, and then I want to hand back to Ida for a Q&A session then. Thank you. Thank you very much, Mathias, for a great overview of the extensive work that you have. You are considered for sure to be one of the most senior thought leaders in the field of proteomics, and since the beginning of the Human Protein Atlas, you said 20 years ago, how have you seen the view of proteomics change during these decades? Well, it's been an incredible progress. Back 20 years ago, we were focusing very much on genomics, the DNA, and all the incredible tools to do that. But now we are moving into the real building blocks of life, which is the proteins. And these have now been systematically mapped by us, but also by others. And this of course forms a base both for applied research in the pharmaceutical industry, but also for all of us that tries to understand human biology. Great. You have also done a very ambitious work on a wellness cohort where you monitored healthy individuals longitudinally. In that study, you basically applied all types of omics that are out there. Maybe you can elaborate how you see the value of the different omics for better phenotypic understanding. Yes, we have, as you say, we have been not only looking at diseases, but also as you say, wellness. We have followed a bunch of people for two years now, and we're continuing taking samples every three months, and we are throwing all the new analytical tools that are available today to look at the microbiome, look at the proteome, the genome, and so on. What is very clear from this is that there is so much that we can gain by using the Olink Explore, but also other omics technologies to try to understand what's going on. Each of us has a unique protein profile which is following us throughout life. This sort of fingerprint then is very important when we are looking at both wellness and disease. Great. A follow-up question to this wellness profiling. As you said, you sample individuals quarterly to monitor the changes in health and disease over time. Do you think for future proteomic studies that this longitudinal sampling will become more or less standard to capture the full potential of the proteome? Maybe not every quarter, but my vision is definitely that every one of us, I think, should actually do a protein profiling once a year, and then see if you have changes, and then to actually compare your own profile with what you had one year ago. By doing that, we will be able, I think, to spot cancers much, much before than we do today. But we will also have a feeling for what the wellness state of you as an individual. I'm a very big believer that one should do this kind of protein profiling maybe once a year in the future. If you go to the more technical perspective, maybe with the incredible development we have seen on the sequencing side with increased throughput and decreased cost, what effect do you think this will have on future studies also with our platform? Well, as you say, there has been an incredible development on the sequencing side and next-generation sequencing. The Olink platform are using next-generation sequencing, so obviously I expect the cost of the Olink platform to actually go down also in the future simply for the fact that one of the most expensive part of the assay is actually the sequencing part, which is now being technically developed so rapidly. By automation and development in next-generation sequencing, I can see that we are actually entering a revolution in blood protein profiling using Olink, but also maybe some other similar techniques. As we know, you know, you have always many projects ongoing in parallel. Which of the ones that you have ongoing now are you currently most excited about and why? No, as you say, we are releasing a new version of the Protein Atlas in three weeks, and we will have a lot of new features, but I think the most exciting feature is the new Disease Atlas, which has been using the Olink Explore technology. We will release the pan-cancer, the cancer version of this, but we are following this up with other diseases. We have already done cardiovascular diseases, various infectious diseases, autoimmune diseases, and also neurodegenerative diseases. This, what we expect will happen in the next year or so, is to have a flood of knowledge and resources that will actually help researchers around the world to actually look at these different diseases and what's happening with their blood fingerprint, when you have a disease and when you're treated with, and monitoring of the disease. Regarding your pan-cancer study that is already out in preprint as well, what are the next steps there in terms of validation of this protein panel or signature that you have identified, this pan-cancer signature? Maybe a follow-up to that question, what do you see as the long-term vision of that project? In the preprint that was published just a few days ago, what we are showing is that with a small amount of maybe 5-10 microliters of blood, we can actually identify if a person has a cancer and what cancer that person has. What we need to do now is that we will have to validate this in larger cohorts, independent cohorts, more patients. Also, sort of confirm that we can actually do early detection of cancers, even in prospective cohorts. We also have a dilemma which is sort of a very big one for all biomarkers and screening projects, and that is what I call the false positive dilemma. Because even if you have a fantastic assay, and that's what we have now, you will always have false positives, and therefore it's very important that you actually also develop second, additional, validations in order to remove these false positives. If you can do that, I think that we have a screening method that will revolutionize the cancer diagnostic field. Great, very valid point. Speaking of this disease atlas, what are your hopes with the atlas maybe from an open science, we are in an open science environment now, what effect do you think it's gonna have for the scientific community to have access to this atlas? The Wallenberg Foundations that has funded, which is a nonprofit organization, they have been very, very focused on that all the data in the protein atlas should be open access. One of the reasons for that is that it is very important that we can put together data from different parts of the world, and especially when you're working with biomarkers as we do, it's important that they will be validated by independent researchers. By putting all of this into a disease atlas, which is open access with the results from our sort of, well, I would call it almost pilot studies, we will then hopefully get all the data then confirmed or rejected from other people around the world. Then obviously we can move this very important field of early diagnosis of cancers forward. Great. We have had the great pleasure working with you around this disease atlas project, and we have also implemented it into our newly launched Olink Insight platform that I will have an introduction about very soon. But you have had a sneak peek on this. Would you mind sharing some feedback on your view of Olink Insight and how you think it will become a useful resource? No, I think what you have developed and is now launching the Olink Insight is an excellent way for people then that are interested in different diseases and even wellness to actually go in and look what are the protein levels which are important to actually monitor wellness and monitoring disease. We will then launch more or less at the same time as two versions of this: the Protein Atlas, Disease Atlas, and the Olink Insight. I think they are very complementary, and both of them are then available for researchers around the world. I hope this will be very valuable for everybody interested in precision medicine. Great. I agree. I have one more question, then we can see if anyone else. When do you think precision medicine based on proteomic profiling will become a reality and what applications do you think we will see first? Oh, that's a very good question. I have been around for at least 20 years where people have been talking about precision medicine and biomarkers, and we have seen very few actually reaching the clinical routine. I think it is different now. I think that we are now having the technology and the quantitative assays that allow us then to actually explore the profiles in people. Of course, how fast this is going, I think it will be incredibly fast on the academic side to actually start using these tools. There are thousands of publications already. On the commercial side, obviously this very much depends on regulatory, the sort of timing. I absolutely think that we will see a lot of applications, and one of my favorite ones is of course cancer because obviously here to actually see a cancer maybe even before you have symptoms, that will save so much cost for society, but also will save thousands and thousands of patients. I hope that cancer will be the first commercial and clinical use of this new technology. Thank you. We have some questions here from the audience. Puneet, I think will go first. Hi, Mathias. Thanks for taking the question. Puneet Souda at SVB Securities. So, one question, just a high level if I could. If you look at the last 20 years of proteomics technologies, where do you think we are in the Olink sort of continuum? And when you if you can make an analogy with maybe the PCR, is it the sequencing or mass spec or ELISA or other platforms that have emerged over time, where are we with Olink and the adoption of the technology? You have obviously used it widely in your experiments and atlas, but when you look at the broader research community, where do you think we are right now today and where it could be? When we are publishing papers on this, and we've done that in the last two or three years, we call this next generation blood profiling. This is because I think to the analogy on the genome side, I think the whole field changed a lot back in 2006, 2007 when we got the new platforms for genome sequencing, which sort of has revolutionized the DNA technology. What I think now on the proteomic side, we have been working with antibodies and one protein at a time. We have been working with mass spectrometry, which is a fantastic technology, but it's not very sensitive, and for blood profiling, it's not really adequate. I think that we will actually see a revolution, and we will see thousands of applications and researchers digging into this new technology. I am very excited actually. There is not only Olink, there is also alternatives, but this way of taking very small blood proteins have a dynamic range of tens of billion, and therefore very sensitive assays which are also quantitative. It really is a gigantic leap forward for science. Great. Thank you. Do we have time for one more? Yes. Hi, Mathias. This is Tejas Savant from Morgan Stanley. One quick follow-up for you. Can you just talk to us about, you know, sequencing as a% of the total cost of your workflow today? With the savings on, you know, the NovaSeq X, for example, how do you envision allocating those dollars? Will it just be sort of doing more proteomic samples, or will it be something else? As a separate sort of follow-up, if you can talk about the importance of PTMs, particularly in the clinical context, that would be great. Okay. The first question, the assay, as you all know, is rather expensive, so everything we can do to cut down the costs, and the sequencing cost is of course a substantial part of it. I don't wanna go into the details exactly how much, but it is a substantial part. We would, of course, like to cut down the cost for the assay in order to then do more. The reason to do more is that it's better to do tens of thousands or even hundreds of thousands of patients to actually get even better precision in what we are doing. When it comes to PTMs, I would basically say if the assay goes down, I would like to add more samples and analyze more samples. When it comes to PTMs, this is a very difficult area, and this has traditionally been handled by mass spectrometry. The problem with mass spectrometry is that it's not very sensitive, so it works on tissue samples, but it's very hard to. It can only work on abundant proteins in the blood. I think one of the technical challenges for us is to be able then to spot the PTMs, the modifications of the proteins, the isoforms, and then develop Olink's assays for these PTMs. That will take time, and it will be rather time-consuming. I think probably you have to then try to do that for those which are clinically valid, which are yeah for different diseases then. I hope that I answer your question there. Great. Mathias, thank you so much for taking the time and sharing your valuable insights. Thank you very much, and I hope the meeting will go fine. Thank you, and it's nice to be able to talk about this fantastic technology. Thank you very much. Thank you, Mathias. Great. Okay. That was the last Q&A session. Now we gonna quickly move forward to the next part of the agenda. Jon already said that I'm gonna blow you away, so I feel some pressure here, but I'll do my best. I have the great privilege of introducing our newest development, Olink Insight, which is a digital knowledge platform for proteomics built uniquely for our customers. At Olink, we wanna continue building an attractive ecosystem of products and services to accelerate the adoption of proteomics in the research community, and this includes essential softwares as well. Here, Olink Insight will lead the way with by providing modern scientific tools that are closely integrated with the rest of our product offering. We see that this together will create this seamless Olink experience throughout the complete customer journey, and that's gonna make our community reach their insights about human disease much faster. We're gonna go in and take a look at Olink Insight. Before that, I just want to quickly take a step back and just outline the main objectives with this new product. First, it's a knowledge platform. We wanna drive a new standard to unveil and consolidate complex data demonstrating how proteomics helps solve the most urgent scientific questions. To address this, Olink Insight offers apps like the Pathway Browser, allowing better understanding of biological pathways and proteins, and we also provide open access to unique datasets that enabling shortening time from discovery by basically educating users on proteomics. We have developed the go-to platform to plan, analyze, and execute complex research into tangible results for more efficient and efficacious drug discovery and patient care. Next, best practices. We wanna ensure the best research design and outcome by guiding the selection of products by Olink, the global standard of modern proteomics. Here, we provide user-friendly tools to help select the right products, 'cause understanding, buying, and using our products will be simple. 'Cause this also in turn will empower the users to ask the right questions as early as possible. Also, with the launch of our new Olink Flex today, we are also introducing this Olink Flex panel builder, that we will have a look at, that allow customers to design their own panels, share designs with collaborators, or they can modify templates that we have created. Finally, we also present high impact publications based on the thousands, more than thousands that we now have on PEA with concrete insights on the reported findings and significant proteins. That will be covering all like disease areas. We see that we have a unique offering for design of successful studies aligned with our high throughput and high multiplex portfolio, which will transform the basic research into targeted prospective clinical decision-making precisely and accurately. Moving forward, it's also an analytics platform. We wanna accelerate the strategic use of proteomics in the scientific field with faster and more accurate results, shortening the time to the next experiment. On the analytics side, we are already offering a wide range of tools for power calculations, data visualizations, and statistical analysis. In addition, in Olink Insight, we're also providing resources such as the open databases, mechanistic insights on a protein level, as well as automatic annotations that allow our customers to also further enrich their Olink data with information such as tissue expression, gene function, and also how the proteome varies in a normal population. We see that this will enable the understanding of the dynamic proteome for more immediate interpretation of the results in health and disease. Finally, open science. We create an open access platform for the global research community to share data and insights to accelerate proteomics. Like we were talking with Mathias Uhlén, open science is the new mindset now with stronger incentives for sharing data. With Olink Insight, we are breaking new ground really by offering this complete access to unique datasets. We base that on studies that we have done in collaboration with our customers, and we are also openly sharing the tools and methods that our own data science team is using. We also have co-written data stories that we have done together with customers. We see that this will build loyalty across the customer journey by establishing this forum for community-driven data sharing with insights and results that will improve drug development and patient treatment. That was that. Hopefully this overview gave you a better understanding that we saw and we see a great window of opportunity here to also in this digital space, to further democratize proteomics, really to accelerate precision medicine and next generation healthcare. That was that. Enough talking. Now we can go in and have a look, and I'll go here. This is gonna be a sort of a demo with different chapters. Gonna show the different steps of what you can do at Olink Insight. Starting off, we go in, insight.olink.com. The first thing that you can see is the range of applications and datasets that Olink Insight currently offers, and I'll take you through almost all of these features. I'm gonna start with the Pathway Browser. Let's say that I'm a researcher. I'm interested in understanding disease mechanism, and I want to explore how Olink can address my needs. First here, if we go to the next, then the browser quickly can display all the pathways, and if the nodes and lines are colored means that they are covered by our library. As you can see, hopefully all pathways are colored. We can then look at coverage by product, starting with Explore here, which immediately changes the view as you can see. You can go to our target panels and maybe then go back to the browser, and then we can select a few panels, and then provides me with an update. I select a few that are of interest to me. Metabolism, I can zoom in and out and go through. I add cytokine, I can zoom out and then go for example, signal transduction. Then I can also look at the sub-pathways in this main pathway. We can also go from the other direction, so starting from my biological process or perhaps a specific signaling pathway that are important for my drug development program. I can learn more about the proteins involved in that pathway, and I can go and have a better overview of which panels I should use, so and I can read more about the different proteins and get a full insight basically into this specific pathway. You can see here the range of the proteins, you can read more about them, and again, see which products that they're in. Perhaps I have already run, let's say in an Explore study, and I have a list of some interesting markers that perhaps separating disease from controls. I can just simply bring my targets into Insight, then just go there. I can convert the gene names to proteins, and then I can just paste them in and then immediately get a better understanding of which mechanisms that are underlying this condition, as you can see. Here I can go from proteins to pathways, or I can go from pathways to protein and zoom in and out and quickly get the mechanistic insights through these interactive tools. Also can mention that it's also useful that I can save all of these visualizations, of course, so I can have them for my presentations, reports, and publications of course. That was that. Let's say that I want even more information about my proteins. I know which ones that are separating disease from health, but also want additional information. That I can go in and I can create a list. I create the list based on these top markers. I do that here. I just name it something relevant. I already have them, and then I add annotation sets. Here I can add information on, let's say, tissue specificity, molecular function from Gene Ontology database, and maybe most importantly, how these proteins vary within a normal population. That could be extremely valuable information for me that will help to prioritize which proteins that are most robust that I should prioritize. I should say that based on customer interviews and the feedback that we have received, that today to get this type of information from different scattered data sources requires tons of resources and time. This automatic annotation tool is greatly appreciated by our customers. Then let's have a look at this normal proteome database. The first thing when we go into this, that you get more information about the database, the cohort, distribution of age, gender, BMI. If you scroll down, you also get more information about all the protein ranges. Here each line represent one protein. If it's the wideness represent how much it varies. A tight range is low variability and wide range, large variability. We can go in and look at one specific protein, LIF protein that was on my list, and I can get more information about the variability of LIF. I can, if I go down, get more information about the protein itself. I can decide if I wanna view a box plot or a density plot. Also, I can have more information about lifestyle factors, so how it varies depending on gender, BMI, and so on. I can also then add more information. I want to have a broader search with all of my top proteins of interest. I can import my list with markers. Here I have it, and then get directly an overview. Perhaps I wanna stratify by gender, and then I get them all listed here. At least from this set of markers, I can see that they have fairly low variability within normals, and that is very promising from a disease biomarker perspective. Again, this type of validation will save me as a scientist a lot of time and provide me with the insights that will help to make accurate prioritization. That was the normal proteome. We can explore the disease atlas that we also talked with Mathias about. Currently in this first version, it covers a wide range of cancer types, and this first figure just shows the number of patients per cancer type broken down by gender. If we scroll down, we can also see it broken down by age. Next, you see a cluster analysis based on the different cancer types, and then you can get information about the individual cancer when hovering over. Here, next, we do differential expression analysis for each cancer, and we report the number of significant proteins for each cancer type. And then also here we can study biological processes by doing pathway enrichment to see the pathways that are most affected for a specific cancer. We can also select specific diseases. Here, for example, I selected glioma. I can compare glioma with other cancers and immediately identify the top proteins that are separating glioma from the other cancers. I can do different types of pathway analysis in different formats, and finally also select the number of top proteins that are of interest to me. I can also view the top proteins that we have identified in our analysis, and we can actually see how they perform as predictors for classification of, in this case, glioma versus other cancers. You can also go in and browse publications at Olink Insight. You can either start from the publications published based on PEA, and we can share the abstract, the reported proteins, which panels that they are found in, and so on, covering all disease areas. You can just simply do a broader search, again, using the same list of markers that are important for me. I can see the most relevant papers. I can see which proteins they have reported. Again, integrated with the panels that you can use to get that information, and then you can further specify your search by adding, for example, I'm working with leukemia, I can add that as a keyword. I get even more relevant search results on that. This is also important when doing reports, of course, or publications that you need to have the most relevant references to cite well. The last thing that I wanna show you today. Now we have explored Olink's library and products based on pathways, proteins, publications. We were enriching our data based on the automatic annotations tool. We could validate our proteins in a normal population and review how other conditions looking in the disease atlas. Now, me as a researcher, I feel confident about my selection of proteins, and I want to do a follow-up study, and I want to develop an Olink Flex Panel. How do I do that? Well, that's very easily thanks to now this new Flex Panel builder. You upload your list, and we can use the one that we have worked with here. You start with that. You create a Flex Panel. As you have heard, we allow up to 21 markers, so these were just 5. I can add a number of additional proteins that are of relevance and importance for my research question. I can just add a couple of ones here. I can add more. IL-4 receptor alpha is important. Then I simply just select the quantity that I need for my study, and I press Request Quote. Within a few weeks, I will have my Flex Panel and be ready for the next study. Very simple. That was the last thing I wanted to show you. I hope that gave you a good overview of Olink Insight. Then of course, for anyone who wants to go in and play around with it or learn more, just go to insight.olink.com, and it's all available there. That was it. Thank you very much, and I'll hand over to Jon now. Great. Thank you so much, Ida. You blew me away at least. I don't know, but I hope that the audience appreciates that this actually takes proteomics and bioinformatics in this space into the 21st century. If you talk to scientists doing proteomics 10 years ago with a tool like this, how this moves the field forward, how it will accelerate the field to more insights, spread more globally, and get to much more actionable results faster, which we think will drive a next purchase and a next experiment much faster as well. To me, this is a little bit like Apple moving into iPhone, like this is a new world. I'm excited. Okay, I'm just going to wrap up the day, and then we'll open up for a Q&A to the Olink team. We're of course extremely proud of what we have built as a company in a very short period of time. The revenue growth here with the 65% CAGR in, you know, in a few years is of course super exciting. This is driven what we talked about by the actual utility that we support scientists with, and you've heard a lot of examples today how valuable those inputs are. Despite done a lot in a short period of time, still when we sit down and do our yearly revisions of a five-year plan and so forth, every time it's like, feels like we just gotten started. Perhaps you've had that sense here today as well, right? That this is basically just starting. What we also pride ourselves in is that we keep our promises. Actually, we haven't missed a yearly target since we started the company. We had one glitch in 2020. We had to actually revisit, and, like, a pandemic hit. We had higher growth expectations going into the year. I believe in March, we sort of took a revision, and then we hit those targets, managed to grow in a very complicated year. Since we started the company, every year has met our targets and goals, you know, something that we pride ourselves in and with. We also do that now in a more detailed time perspective, perhaps, with the quarterly reporting we do on Wall Street, and really pride ourselves on executing and delivering. I would truly think that that is based on that we really know what we do, we know this market, we understand our customers, we have very good sales processes, funnel management, and so forth, that we can be very reliable and predictable when we follow up on our business. What's driven our success as well and what we and many of the KOLs have spoken about as well. Our read on the market, what has been truly important is to grow the number of proteins that we need to increase multiplexing, we need to increase throughput, and we need to drive costs down. The way we've been able to do this and innovate in this space, I think is quite remarkable from what you see on this slide. Look at how we increased data points through experiments and throughput. It's like over 100x in a very short period of time. In combination how we've driven cost down because we believe that the more affordable and more accessible we can do this will benefit patients, healthcare, the whole society, but also Olink. When we look ahead, we have talked about us being ahead of the 4.5K assays that we have validated. We are. We've also collected a lot of customer feedback on Explore. When we move into next year, and as we see that we will continue to increase multiplexing, increase throughput, increase number of assays, and drive costs down. Innovative steps that we will share with you later next year is a remarkable milestone in our history actually. As I look into our product roadmap moving into next year, I think actually 2023 will be our most exciting product development year across high-plex and mid-plex. We are extremely excited heading into 2023. Another area where we're very clearly communicating and very clearly executing on is the externalization or driving a product business. Actually, as we think about our business from my perspective, we weren't in a situation until quite recently to push that product business. We've always wanted to be a product business, but the first readout platform that we used was the BioMark HD from Fluidigm, so it's quite old, large instrument, expensive. Fantastic technology for sure, but a limited installed base. The first step, right, as we translated our technology onto NGS, obviously opened up a tremendous opportunity to drive product. With the launch of Signature for the mid-plex market, the same thing that we now have a competitive, cost-effective benchtop instrument to be very competitive in that mid-plex space as well. As Carl alluded to, right, I mean, we see here that we are executing very well since we got to this position where we could drive product. Very early days again, right? We penetrated roughly 1% of these sequencers or high-plex or mid-plex markets. Of course, with quite exciting pull-throughs as customers are adopting this. As shown on the right-hand side as well here, how we also now are executing on the kit mix and our product business. When we started Olink back in 2016, we had very little funding. We had to be profitable to continue to invest, make a dollar, invest a dollar. We did that very nicely. We're profitable already year 2. As you see here in 2020, we went to the board and said, "Wow, this technology, our set of products has gone from clarity to clarity. As we look into the proteomics market, it gone from clarity to clarity. We think that this industry and this space will be so exciting looking ahead. So we really think that this organic strategy that we had, that we should change that around. We should invest to ensure that we really are in a position to capture this market looking ahead." That's the driver of the IPO. We raised money. We invested significantly, in particular the commercial organization globally, but also in R&D. That will drive the innovation and what we will talk more about next year. But as we said, we now made those investments that we developed the support functions and so forth internally, and we'll revert back to profitability next year. We should keep in mind that you all know that we have a very nice financial margin or financial profile, and margin, and hence, we will be able to continue to aggressively invest through our P&L. There is no stopping in investing here, but we certainly believe we can do so in a profitable manner. To close the circle and the presentations, this was where we started today, where we hope your takeaways are. At least from my perspective, I think that, as I talked about, that no organization is more valuable than the value that we generate for our customers, that we heard a lot about the value of proteomics, and in particular, by Olink today. That hopefully that has resonated, and that is the reason to the very strong adoption that we see across both biopharma and academia. I also mentioned that I truly think that we very much understand and know our customer base, what they're thinking about their unmet needs and how they drive their experiments, and that we, with that customer focus and the innovation and the fantastic talent pool at Olink, have put ourselves in the situation we're in and, moving forward based on. Then, as you heard all over, right, we've just scratched the surface. These are very, very early days and largely an untapped market. We strongly feel that we have put ourselves in a driver's seat, and we built out an organization and function and a foundation to really continue to be market leaders and capture this opportunity for many, many years to come. With that, I'd love to see the rest of the team join me here, and we will open up the day for for any Q&A. You must have tons of questions to ask, right? It's been so silent today. Hey, just a couple from me. Excellent day. Thanks for all the KOLs presentations and your presentations as well. You know, I think a high-level question longer term is really from the 725,000 pull-through that you have today, when you think about sort of the elasticity of demand, some of the KOLs talked about they are excited about the technologies, but you know, in order to grow it broadly, maybe cost generally comes down on these technologies longer term. How do you think about the progression of that? Within that context, if you can talk about 4.5 K panel, when do you think you can reach the 6K and when do you think you can reach, you know, 10K? How should we think about the progression towards that? What does that mean for the pull-through? A couple of questions in there. Yeah. Very easy ones too, huh? I can start. Yeah. So as you see, we read the market as well, like you heard today, and you know, cost can be prohibited. So you saw how we've driven cost down very significantly, right, in a short period of time, and that we also agree that needs to continue, and that we have our molecular tricks and tips to do that. Of course, what was mentioned earlier today as well around genomic developments, that how we see that part of the cost for the assay is coming down, so benefiting end customer use as well. So, yeah. I mean, we will very aggressively continue to do that, and we think that will position us, you know, very nicely, that the experiments will be larger, and as they get it right, results will be more precise. We feel good about what we have in the pipeline and moving in on costs for next year. As we talked about on the library size, we are now beyond the 4.5. Next year, we're gonna sort of sail past that and towards higher numbers. We don't want to disclose any details yet for market and competitive reasons. A very nice step forward, and that we think the way that product is designed as well will be very, very exciting. Those were a couple of your questions. Just wanted to clarify in terms of gross margin, you will continue to strive for that current gross margin that you have on kits even at 6K and 10K. Is that right? Is that fair to say? Yeah. Yeah, absolutely. I mean, the kit gross margin is certainly sort of scalable as we go up in plex. Also from a group gross margin perspective, as we drive towards kit, that will sort of, you know, improve the overall. Across medium profile. Okay. Just last for me. In terms of the broader customer base that you have that is using Target today, you know, can you characterize, and I didn't specifically hear it through the day, in terms of how those customers are getting upsold into Explore. Maybe talk to us a little bit about the progression of those customers from 96 to mid-plex to Explore. Yeah. I think I've talked about it a bit before. It's a bit of a virtuous cycle, I think, from you know, we started in 1996, sort of in the middle, right? It's where the company began. You know, as we've stretched the product lines, you know, we see a great transition both from customers in that high-plex area who are then driving down, as you heard from some of the KOLs today, right, down into more targeted sets of markers that they want to look at. Vice versa on the low-plex. I think sort of these 96-plex customers, you know, as they become more, you know. Just realize too, I guess, really going back to sort of the genesis of the company, like even a 96-plex when we started doing this back in, you know, 2016, 2017, was pretty amazing. Like, being able to do a 96-plex was, I think it was just convincing the world that this was, you know, this was real. This was amazing. I think as we've exposed the market to that, then that comfort level is growing with like, wow, like look at the possibility here going way beyond this to thousands of markers. Then of course, you know, stories like UKB and others really just sort of, you know, driving the market forward. I think with the launch of Flex now, I think we're actually going to introduce a whole new audience of customers to the power of the PEA technology, which again, we think will be sort of, you know, an introduction to the technology, an advancement to tens of proteins. Could I do fifties? Could I do hundreds? Maybe thousands, right? I think it's, you know. Again, we'll sort of continue gathering new customers with the extensions of the product lines. Hey, guys. Maybe I'll start with one for Ida. Now that Insight is available, I mean, how correlated is protein expression to RNA expression in terms of the data that you see so far? Well, that's an extremely difficult question, and it depends, I guess, what you're looking at. When you review the thousands of publications that we have, there are many that are representing the correlation and confirmation from RNA to protein, and in some, there are quite good correlation if you're looking at the same matrix and the same matched samples and all of that. There are other cases when they go from a tissue to the circulation, and the correlation is very, very poor. It's difficult to say, but I think what we have seen, both from hearing from the community and also reading these reports, that there are added value from the proteomics with some of the challenges that you see from the transcriptomics space with the extreme robust dynamics with the RNAs that are more unstable and have these burst effects that could be difficult to target, whereas the proteins are actually more robust. Yeah, it's a difficult question that we are asking ourselves as well. Got it. Jon, going back to that cost chart that you showed, you know, going from $2 to about $0.30. First of all, does that include the sequencing cost decline? And how should we think about teasing apart, like how much of that decline is just NGS getting cheaper versus non-NGS costs getting cheaper? Yeah, no. That's total experimental cost embedded in there. We're definitely benefiting from NGS in that lower cost down for sure. I don't know the sequencing cost that you asked Mathias the question. I guess that could be depending on the volume and blah, blah. Maybe it's like 15% of the total cost of running an assay today. It's really not like. It's of course, when you run thousands of samples become quite a lot. The advancements that we see there as well will be, you know, greatly appreciated and benefit this market as well, for sure. Got it. That's helpful. Just one question on biopharma budget flush. You know, this morning there was some news around, you know, one of your peers basically they're not assuming a biopharma budget flush in the fourth quarter. Just curious as to, you know, what you're seeing in that customer base and, you know, given where we are in the quarter today, like how de-risked do you think your sort of end target for revenue is at this stage, Jon? Yeah. I think, Carl probably is even closer to that. Well, first of all, we are almost mid-November, you know, Q4 is a big quarter for us. You know, we have every year had quite a lot to do between November 14th and December 31st, still time to go. Obviously, quite a lot of insights to what we think will happen between now and then. On the biopharma flush, Carl, I don't know what you- Yeah. I think I commented on some of this last week. I think our ear is to the ground, you know, and we're listening very carefully. Right now we're not hearing from our customers that they're pulling back on experiments or budgets are being cut or any of this sort of thing. Although again, in this quarter, it's. The clarity tends to come late in November and early December when you hear the budgets. We're talking to customers about this, but they don't have clarity yet on what these budgets look like. It's still honestly a little TBD on exactly what it's gonna look like this year in comparison to prior years. Again, you know, for now we're feeling good and we're feeling confident 'cause our customers are not reporting that they're cutting back on their experiments. Got it. One final one on M&A, Jon. I mean, you've talked about, you know, Agrisera as being a, you know, a valuable addition to you. As you think about sort of expanding Explore, is that sort of the natural direction in which you'd like to take the M&A engine in terms of perhaps other such assets to expand the target library? Yes. If we are, and I think we've spoken about that in the past too, right? You're spot on there, that if we're looking at adding further capabilities to the Olink family, it's sort of upstream. Also keep you know in mind what you've heard today, obviously expanding the library, but also for where we see this one playing out in the long-term clinical markets, to further strengthen our position, to capture that. Yes, and but, you know, M&A you can't plan for, but we keep our eyes open. Thank you. Yeah. I think we have time for. Hi. Just one for me. How much of your target market do you believe is held back by cost? Meaning, how much of the market can be opened up with either lowering price or lowering costs? Yeah, no, I honestly don't think. I mean, Carl, you can correct me if you think I'm wrong, but I in the mid-plex market, I think we're, you know, very cost effective in what we do. So competitive wise, we should be fairly good there. I don't know if you- Yeah, no, I agree. Agree? I don't think it's there's a price access there that would significantly tip things. I think customers are getting the value out of the platform. I think they understand that, so I think the price points, I mean, look, everybody always wants lower pricing. You always hear that. But no, I think we're at a really price effective point in the market right now. Like, this is not genomics. I mean, if you think about how far modern proteomics has come in a short amount of time versus genomics, which has been at it for more than 20 years, I think, you know, the advancement that Jon showed in terms of the cost per data point is pretty darn impressive. Again, you can only draw so many analogies between genomics and proteomics as well. It's not more bases or a few more reads or a few more whatever. It's wildly more complex like we talked about earlier. You wouldn't see quite the same dynamics, and yet at the same time I think we've done a pretty impressive job in a short amount of time. Great. I think a round of applause for a great team. Great. Thank you. Thanks to organizers for terrific job today, and thanks to the virtual audience as well. We hope to see you live soon. Have a great day, everyone. Thank you. Thank you.