Good morning. I'm Roy Smythe, the CEO at SomaLogic, and it's a pleasure for us to host you virtually for our first Investor and Analyst Day as a public company. Joining me today on the program will be Shaun Blakeman, our Chief Financial Officer, Dr. Steve Williams, our Chief Medical Officer, and our President, Melody Harris, will also be joining the three of us for the final Q&A session. I'll be making forward-looking and time-sensitive statements during this presentation and ask that you please refer to our SEC filings for more information. Our agenda today will include my general overview of SomaLogic, comments from Shaun Blakeman, our CFO, and our financial profile in 2022 guidance, followed by a great KOL panel discussion moderated by Steve Williams, our Chief Medical Officer. Following our first Q&A session, we'll wrap up with a discussion with Joydeep Goswami, the Chief Strategy and Corporate Development Officer at Illumina, a formal announcement regarding our oncology project partnership with Imperial College London, and then our final Q&A session. Now starting on slide four. We've known for decades that proteins, the structural and functional molecules of life, could provide us with the most information about human disease and biology. While obviously important, genes are static and do not change. RNA is more dynamic, but only correlates with protein expression about half the time. Proteins, on the other hand, are incredibly dynamic. The proteome turns over about 40,000 times during our lives and changes with age, illness, medications, and environment. This makes them both contextual and an incredibly powerful real-time and future indicator. While we've waited to make this pronouncement for decades, I'm strongly suggesting to you now that the era of proteomics is here. The question is, why now, and what's taken so long for this to happen? At baseline, our bodies have 20,000 basic protein structures, what we refer to as the canonical proteome, and a much larger number of modified forms. The reason it's taken so long for proteins to be as important as they are now, just now, is that we were unable to measure a sufficient amount of this basic 20,000, and we also had to develop the computing capabilities to make sense of the data generated from the measurement. SomaLogic is the first to meet these two crucial criteria. Measuring more than 2x more proteins than any other enterprise, and with 10 years of experience analyzing the data with sophisticated machine learning tools, we're now using that data in ways such that we are singularly demonstrating both the current impact and promise of proteomics across the spectrum of applications. Now on slide six, this places SomaLogic in a unique position, one that's at the nexus of the needs of researchers, clinicians, and patients, the ability to measure the most proteins, and unparalleled experience working with proteomics to create impactful applications. It also puts us in a position to capture a significant share of the rapidly growing proteomics market and to capture that share across the spectrum of proteomics life science tools and diagnostics opportunities. As the number of companies entering the proteomics space grows, SomaLogic remains uniquely positioned. Let's look at the five areas in which we're different from others on slide seven. Moving from the left to right, first, we have a platform and not just a technology, one that includes a measurement identification technology, but also the world's largest proteomics database and a bioinformatics stack. Second, while some in our space are pre-commercial or early in their efforts to commercialize, we've been heavily market-validated. Next, the technology component of our platform is the world's most powerful, capable of measuring and identifying more than 2x more proteins at scale than any others with market-leading technical specifications. Fourth, our database and bioinformatics capabilities render us the knowledge leader for proteomics and the creation of clinical applications with both financial and human value. Leveraging this knowledge base is not just aspirational for us, we've accomplished this. Lastly, it might be seductive to think that a 20-year-old company's technology is becoming senescent, but that's far from true. The fact that we use synthetic nucleic acid constructs to identify proteins means we have a very wide vista of technology and business opportunities ahead. In addition to these differentiators on slide eight, I'd like to also invite you to consider the fact that unlike some others, we don't have to invest to make our technology work or to get initial market traction, and therefore, should be able to apply the proceeds from our public transition financing to a number of business opportunities across both life sciences tools and diagnostics. While it might be an understatement to say that the market has changed since our public transition, we've not only developed a fundamentally sound business that's executing very well, but due to our decision to use the SPAC pipe structure for our public transition financing, whereby we had less than 3% redemption and incredible support from a very large number of investors, we also have more than $600 million on the balance sheet. On slide nine, I just want to go through quickly what we mean by a platform, the first differentiating characteristic I mentioned earlier. SomaLogic, rather than just a technology, has a platform. While that term is often overused, we actually do, in reality, have one. Possessing a protein measurement identification technology is table stakes in proteomics, but content is king, and we measure more than twice the number of proteins than all others. Our database is the world's largest, and that in combination with our bioinformatics stack, has allowed us to create first-in-class clinical applications. These three things create a platform and not just a technology. Let's quickly look at these differentiating characteristics in a bit more detail on slide 10. In regard to market validation, we've had more than 300 customers on our platform over the last several years, including those who trust us well enough to enter into long-term agreements like Novartis, and we increased this customer base by more than 300% over the last two years. We are the measurement leader for proteomics with an ability to measure 7,000 and soon 10,000 proteins, industry-leading specifications and high throughput. On the far right, we've leveraged this ability to measure more proteins than anyone else in the world to grow the world's largest database with 1.5 billion protein measurements and 675,000 years of patient follow-up data. This has allowed us to create a number of currently validated high-plex protein pattern recognition laboratory developed tests or LDTs, and a very large pipeline of many more. On slide 11, our plan to leverage these distinguishing characteristics is really clear to us. We're going to be the comprehensive proteomics provider across the spectrum of life sciences tools and diagnostic opportunities, and we'll also leverage our unique modified aptamer reagents to not only power both of these businesses, but become a business of its own, partnering with and selling into a number of areas where antibodies are currently being used. Importantly, as depicted in the center of the figure on the left-hand side, the flexibility of our technology allows us to work across a number of platforms, NGS, arrays, next- generation arrays, chip-based and others, and the provision of both life science tools and diagnostics products and solutions. On slide twelve, let's discuss the use cases for life sciences tools for a moment and how SomaLogic has a distinct advantage. The bicolor hexagons on the left are the common use cases for protein identification and measurement. Most customers are using proteomics to better understand biology and to leverage that understanding to develop and validate new therapeutics, and they're doing so with either protein data alone or protein data combined with genomic data. As we power the measurement identification of many more proteins, we can support these efforts in a more impactful way. The three use cases in the box at the bottom of the slide are things only we can do at present, provide bioinformatics support and increasingly products built off of our database and bioinformatics capabilities, and the development or use of high-plex proteomic diagnostics for clinical trials and other uses in the life sciences tools market. We also believe that the market for our life sciences tools reagents business to be substantial based on our current conversations with potential partners and customers, the ability of aptamers to work in many contexts where antibodies do, and our ability to see thousands of proteins that antibodies currently cannot. However, as slide 13 depicts, you can't have a successful life sciences tools enterprise without sufficient numbers of talented people. I like to tell our team at SomaLogic that while great technology and products are necessary, they are by no means sufficient. Companies are foundationally nothing more than the people that constitute them and the things they agree to do together. Growing the team at SomaLogic has been important, and an ability to grow an effective team is an important indicator of our company's imprimatur and success. Our commercial group grew from 14 to almost 50 individuals in 2021, and our intentions for 2022 are even more aggressive, with plans to go north of 100 individuals. In this second full year of the pandemic, when many were having trouble attracting talent and in the time of the Great Resignation, we hired more than 160 new employees and doubled the size of SomaLogic. Our team and an ability to support customer use cases in ways others cannot have allowed us to substantially grow and diversify our life sciences tools customer base over the past year, as depicted on slide 14. More than 60 first-time customers came on board in addition to our larger, more established customers such as Novartis, Amgen, and others. Signaling both significant validation of our life sciences tools technology capabilities and our dedication to diversify the platforms in which we deliver products, on slide 15, you can see that we are incredibly excited to have signed a new and impactful commercial partnership deal with Illumina, the first of its kind. I'll be speaking with Joydeep Goswami, the Chief Strategy and Corporate Development Officer for Illumina, a bit later in the program, but I will quickly mention that this is an incredible all- upside deal for SomaLogic. One that provides Illumina with the capabilities to create a co-exclusive, co-branded, NGS-based deployed protein identification and measurement solution, a number of economic upsides for SomaLogic, and one that frees up capacity and capital for us to work on other platforms for life sciences tools, products, delivery, as well as our diagnostic efforts. Let's shift now on slide 16 to talk about our diagnostic business. As a former clinician, I am particularly excited about this in the incredible array of diagnostic use cases that high-plex pattern protein recognition proteomics tests will now support, and those that will facilitate over the next few years. As we're laying the foundation for clinical practice, clinical trials, and population health markets development, you'll see us talking in the near term about partnerships and business combinations with others to either give them access to our already substantial existing diagnostic assets or develop new diagnostic assets for them, de novo from scratch. The ability to measure thousands of proteins concurrently and find machine learning models of between 15 and 350 proteins that characterize conditions in real time, but also to predict what is likely to happen in individuals moving forward in time in regard to the development or evolution of disease will change medicine. On slide 17, you can see the SomaSignal tests we've already launched into our demonstration market on the left-hand side, those that are coming in 2022 in the middle. While our initial efforts focused on cardiovascular disease, in 2022 we'll add tests to evaluate risk of neurologic, oncologic, and renal diseases. Notably, on the far right, you can see that over the next 18-24 months, in addition, to new tests for Parkinson's and inflammatory bowel disease, we plan to develop an entire panel of the world's first cancer predisposition tests, working with our colleagues in the U.K. at the Imperial College London. We'll talk more about that later in the program. Based on work we've already done, we believe that rather than detecting cancer in early stage, these tests will actually characterize an individual's risk of developing cancer prior to its occurrence. Literally changing the way we screen for cancer around the world, and perhaps as well our understanding of how not only to treat cancer more effectively, but to prevent it. As we seek to move from our demonstration market to the larger world, we're putting the building blocks in place for commercial acceptance, payment, and regulatory clearance where we deem necessary. The Proteomics for Precision Medicine initiative depicted on slide 18 is the way in which we're laying this foundation. The world's first prospective clinical evaluation of high-plex protein pattern recognition diagnostics have been initiated at CommonSpirit, Intermountain, University of Colorado, Emory, UPMC, and University Hospitals Cleveland. Trials are already enrolling in four of six of those sites. We hope not only to use these relationships for commercial purposes, but also to create a clinical proteomics data consortium in the future, and we'll look forward to talking more about that soon. At the outset of the presentation, I mentioned our fifth differentiator was the trajectory we have for this technology and platform, even though we're a 20-year-old company, it's by no means senescent. On slide 19, we show you that over the next couple of years, we'll be expanding content to 10,000 and beyond, developing and characterizing multiple proprietary aptamer reagents for each protein instead of one, investing in next-gen array and chip-based platforms, enhancing our manufacturing capabilities, and taking the sample prep front end of our assay and boxing it up, such that it will not only work in front of arrays, NGS, and chip-based platforms, but in front of mass spec as well. We'll also be augmenting our data use capabilities to support the development of new data products, as well as customer interfaces for life sciences tools and diagnostics, and developing a number of new SomaSignal tests across a large spectrum of diseases and conditions. Lastly, we'll entertain inorganic growth opportunities in both life sciences tools and diagnostics, and are contemplating many opportunities there and options in real time. On slide 20, you can see that we've already translated our capabilities and strategy into a fundamentally sound and high execution business. As you can see, on the table to the left on this slide, we had a remarkable number of accomplishments this past year. We had a successful public transition on Nasdaq and put more than $630 million on the balance sheet. We had significant revenue growth. Our original projections of around $67 million were exceeded significantly, and we're now coming in at north of $79 million, as we have discussed previously. We launched a 7,000- protein identification platform product, which, again, measures more than 2.5x more proteins than anyone else in the world. We grew our field team substantially from 14 to almost 50 individuals. We had significant new customer growth. We changed our business model to be one that is much more customer centric. We had early access launch of our site of service array deployed kits. In fiscal year 2022, we believe things will be even better. With a number of milestones, we plan to achieve a number of on the table on the right, and these will be good landmarks for you to watch for this next year in regard to our ongoing execution success. In the first week of this new year, we've already checked the first box. With a new business partnership with Illumina, we anticipate a revenue growth that will be substantial over the next year. We will complete the R&D components of our 10,000- protein identification platform product, with hopes to launch that in early 2023. We expect to grow our field team to north of 100 individuals and also expand in EMEA and APAC. We obviously anticipate another great year of new customer growth. We are entertaining strategic M&A options. We'll have full commercial launch of our suite of service array kits, and we will develop a number of new SomaSignal tests. Hopefully as well, have some early data from our clinical evaluation trials to share with you as well. Companies are nothing more than the people that constitute them and the things they agree to do together. On slide 21, you can see that I have an incredibly experienced and talented leadership team I look forward to telling you all more about, and to whom I owe all of our current and future success. We've also been fortunate to attract a number of great new board members who have already augmented our efforts, and a few of them are listed here on slide 21. In closing, while they've forever been the goals of medicine, we are finally leveraging the power of proteomics to relieve human suffering and to prolong meaningful life. This company's contribution to these most important efforts inspires me and is what frankly gets me out of bed every morning. The era of proteomics is decidedly here, and we are proud to be actively ushering that era in with the crucial and necessary support of all of you. Thanks very much. Now I'd like to pass it over to Shaun for our financial profile and outlook. Thank you, Roy. Hello, good morning. I'm Shaun Blakeman, the CFO of SomaLogic. I've been with SomaLogic just about six months now. I'm very excited to be part of the SomaLogic family, and, you know, for all the reasons that Roy so eloquently explained. For those of you that don't know me, I was also most recently the CFO of Cantel Medical before they were acquired last year by STERIS. I'm happy to share our financial outlook for 2022 with you as we continue our accelerated growth story following a transformative year in 2021. Looking at slide 24, the significance of this slide is twofold. One, SomaLogic has been over 20 years in the making, and is now truly transitioning into a commercial stage enterprise that has been enabled by a significant influx of capital in recent years. Throughout most of our history, SomaLogic has been primarily a research organization, pioneering an approach with the potential to make a true impact in the field of proteomics and how it can impact how we treat disease. This early and sustained focus on R&D has resulted in the development of our core technological platform, which we believe is market leading in scope and applicability. In recent years, management began making key strategic changes to lay the foundation for SomaLogic to transition from an R&D driven organization into a meaningful commercial enterprise. Those decisions indeed have made an impact which we are seeing today. In fact, since 2018, as you can see, we've increased our full- year revenue to over $79 million, driven by seamless execution, offering a premium product and becoming more customer centric while aggressively expanding our commercial team and sales force. Secondly, as I just mentioned, we are extraordinarily well capitalized with cash on the balance sheet exceeding $650 million at the end of the year, before the $30 million influx from Illumina as well. With the benefit of our 20+ years tenure, where we focused on technological development, we have the flexibility to allocate capital to where we believe it'll be most impactful. In this case, we don't have to use funds to develop or validate our core technology into a revenue generating product. Instead, we will spend that money on initiatives that accelerate our growth, investing in the areas Roy highlighted during his presentation, including the continued commercial team build- out, content expansion, expanding our product offerings, international expansion, and opportunistically pursuing inorganic opportunities that shore up our position as the leading proteomics life sciences tools company. In parallel, we will actively pursue commercialization opportunities of our diagnostic SomaSignal tests, either through internal development or in partnership, which has the potential to have a meaningful clinical impact and consider an attractive market opportunity which we believe has a TAM in the order of $40 billion. On slide 25, as disclosed in this morning's press release, we anticipate another outstanding year in 2022. I'm excited to announce that we are initiating full year 2022 revenue guidance in the range of $105 million-$110 million, representing substantial year-over-year growth compared to 2021, which we previously stated will be above $79 million. Please note that this guidance does not include any potential revenue recognition from the Illumina partnership as we are still working on that model. As we previously discussed, we have a lot of room left to grow. I'd like to point out a few key levers that we believe will underpin this profile and can potentially accelerate it. Commercial team growth. We will continue to aggressively recruit and build out and enhance our commercial team. We're excited to arm them with the most comprehensive proteomics platform on the market, and look forward to seeing this ramp translate into additional growth. We've tripled the commercial team from 14 to 47 through 2021, which means that we're starting 2022 with a considerably larger sales force combined with a stronger pipeline. As we continue to expand our team in 2022, expect to see more and more wins to feed that pipeline, further diversifying our customer base. Two, diversifying product offerings. Further expansion of our platform is key by being able to service customers that, for example, prefer a kitted solution and enhancing our ability to penetrate wider adoption of proteomics tools in the life sciences space. Another key point here is that we don't view expanding our product platforms as an either/or proposition with our existing service business. As with genomics, we believe there will continue to be many different modalities customers choose, either by preference or for specific applications. Our view for the foreseeable future is our services business will continue to see strong growth. International expansion. SomaLogic has been a U.S.-focused company historically, and we view growing our presence in the global marketplace as a large and untapped greenfield opportunity for us. We've already added some initial key commercial resources in both Europe and Asia, and we anticipate that we will achieve analogous success as we have in the U.S. as we continue to add resources internationally. Lastly, repeat business. In 2021, we saw or we had much success in adding new customers to SomaScan, often for pilots and then seeing them come back for further endeavors. As we continue to add new customers, this is also a potential accelerant to our near-term grow th. I'd like to conclude by revisiting my commentary on how to think about our gross margins. To reiterate, I would think of our core services business as a mid-50% margin business that should benefit from continued customer diversification. I also expect that our investment in cost out in our service business will expand margins in the near and mid-term. Our service business is healthy, and we expect it to continue bearing fruit for many years. But even more exciting is that as we expand our reach into the proteomics tool space and expand our product offerings, we anticipate materially accretive business to our margins. I plan on providing more details on our 2022 financial outlook during our year-end earnings call, and I look forward to updating you on the continued progress at that time. With that, I'll hand it back to Roy. Thanks, Shaun. As a reminder, we'll take Q&A related to our business and financial profile at the end of our event. We next are going to have a KOL panel moderated by Dr. Steve Williams, and we're going to take a quick intermission and allow those KOLs to join us online, and we'll be right back. Thanks. Welcome back. I hope you had a chance to grab a cup of coffee or take a bio break, and we're excited to have the next part of our program, our key opinion leader panel, moderated by Dr. Steve Williams, our Chief Medical Officer. Joining us for this session are Dr. David Fajgenbaum. David Fajgenbaum is the Assistant Professor of Translational Medicine and Human Genetics at the University of Pennsylvania, Founding Director of the Center for Cytokine Storm Treatment & Laboratory, and Associate Director for Patient Impact of the Penn Orphan Disease Center. He's also Co-founder, President of the Castleman Disease Collaborative Network, that position will be cogent to what he talks with us about here shortly. We also have Dr. Richard Head, the Director of the Genome Technology Access Center at Washington University School of Medicine. Jason Laramie, the Vice President and Global Head of Translational Medicine and Data Science at NIBR. Peter Ganz, a Director of the Center of Excellence in Vascular Research at the Zuckerberg San Francisco General Hospital, where he's also a Professor of Medicine at UCSF. Steve Williams, our Chief Medical Officer, again, will moderate this session, and we'll begin our discussion with Dr. Fajgenbaum to discuss his very personal experience with SomaLogic technology. Steve? [inaudible] his choices of technology were not simply abstract, but they affected his own life. David, please go ahead and kick us off. Thank you so much for having me today, and it's just really an honor for me to share a bit about my journey leveraging the SomaLogic platform and chasing after my cure. I wanna start out by just sharing a few images that I think will help to somewhat bring my story to life. This is a picture of me as a healthy third-year medical student. I wanted to become an oncologist and treat cancer patients in memory of my mom, when out of nowhere, I became critically ill with a disease that we didn't know what it was. I literally was experiencing multi-organ system failure. My liver, my kidneys, my bone marrow, my heart, and my lungs were shutting down and no one knew what it was. I was so sick, I actually even had my last rites read to me, and my doctors didn't think that I would survive. I fortunately eventually received a diagnosis called idiopathic multicentric Castleman disease. This is a picture of me after spending nearly six months hospitalized. You can see due to this awful condition, I gained fluid all over my body, and as I mentioned, in fact, nearly died three times in the first six months. I was so thankful to be alive, and I was on an experimental drug. I believed that that experimental drug may be able to keep me in remission. Unfortunately, I relapsed and nearly died for the fourth time, and I learned that there were no more drugs in development for Castleman disease. There were no more promising leads, and that if I had any hope for survival, I would need to get involved in research. I committed the rest of my life to trying to find a drug that could maybe save my life. I knew I would need to leverage technology and platforms that could help to figure out what was wrong in my body, to really try to identify a drug that could maybe be repurposed to save my life. Finally, this is a picture of me as an advocate. I knew that I couldn't do it alone. I would need to connect with others in this space because I needed this to be a collaborative effort from folks from around the world. Thanks to technologies such as SomaLogic, I was able to perform a series of experiments, including the SomaScan assay, to look for the proteins that were elevated in my blood, and eventually found a signal that a particular communication line called mTOR was highly activated. I did a final experiment and eventually confirmed that mTOR seemed to be up, and actually started myself on a drug, an mTOR inhibitor, based on this proteomic data, and I'm alive and well today. It's been over eight years that I've been in remission. I hope that this message of showing how SomaLogic has literally come to life in giving me life helps to give you guys a sense for the power of this platform. David, a quick question. I mean, what a terrible experience you're going through. What did you know about omics platforms when you started to think about, you know, the search for a cure or a treatment? What did you know at that stage about different technology platforms and what led you in the end to your choice of the SomaLogic platform? Well, I knew that I didn't have very much time to find a platform that could work, and I knew that I would need to find and utilize a technology that could give me a sense for exactly what was happening in my immune system at the moment that I was sick. For me, that meant that proteomics made the most sense. Transcriptomics and genomics could give me a more broad picture, but proteomics could give me a more specific insight to what was happening in my body and potentially involved in my disease. I decided to move forward with the SomaLogic platform and also another platform that's smaller than SomaLogic in terms of analytes, and there was really strong consistency between the results from the two. That led me to say, "Okay, I'm gonna really go all in on this bigger platform." Yeah, that's a question that often comes up. You know, how many proteins is enough? Couldn't you have just guessed which proteins to measure up front? Why did you need to measure it? Why did you need to try and find the biggest platform? Well, with Castleman disease and with so many rare diseases, so little is known about why the disease does what it does. Without knowing what to look for, you need to take a really broad view. That's what we did with the SomaScan, is we took this very broad view, and then we could identify what is specifically going wrong in my immune system. Now you've identified this pathway that's elevated. You looked at the disturbances that the disease was causing, and then you looked at well, what possible drugs might be available today. Tell us a little bit more about what your choices were and why you ended up with mTOR, and then is that the end? Is that your—i s that it? Well, when you do a large omics platform like the SomaScan, you're going to get a number of hits. What you have to do after you get a hit is figure out how you can validate that hit. Fortunately, I had a lymph node sample that I had actually had resected from my last relapse. From that sample, we were able to do our confirmatory test. We were able to look for mTOR activation in the tissue, and we confirmed it in the tissue, and that made us feel much more confident that the signal we had seen from the proteome was actually now being shown in the tissue in situ. That sort of confirmatory step doesn't guarantee that an mTOR inhibitor is gonna be effective. There are many times in medicine where you find something's activated and an inhibitor of it doesn't actually work. I really had run out of options. As you know, I actually was engaged to my fiancée at the time. I dreamed of maybe getting married and having a family and practicing medicine. I really was up against the wall and had no other options. Between the proteomic data, the tissue data, and really this desire for a future, I talked to a number of doctors and shared the data, and we agreed that really an mTOR inhibitor was my best chance. As I mentioned earlier, it's now been over eight years. I don't like to round up because I don't know how long this drug is gonna work for, but I'm certainly so thankful. We've now moved this drug forward to clinical trials for other Castleman’s patients, and we've also shown that this pathway is activated in other patients as well. Putting on a kind of the economic hat, people could look at what you've done and say, "Well, it's a rare disease, as not many people are affected, it's not economically very important." How would you answer that? I would say that, all rare diseases, are certainly rare on their own. When you combine them, they're incredibly common. About 1 in 10 Americans has a rare disease. Unfortunately, 95% don't have a single FDA-approved therapy. If you can find a drug that's effective for these rare diseases for patients like me, I've become very cheap on the healthcare system the last 8 years. I was very expensive on the healthcare system, the first 3.5 years I had my disease. There's certainly economic arguments that could maybe go both ways, but, there's nothing cheaper to the healthcare system than someone who's healthy. If you can find a drug that can make someone healthy and you can utilize omic data to do that, I can't imagine anyone who would argue against that. With that, could you just please show us your next slide? Absolutely. Tell us about the slide, and then we'll move on. These are three more recent pictures. This was the picture from right around the time that Caitlin and I were engaged to get married. This was my fiancée at the time. It's when I made this discovery that sirolimus might be effective based on the proteomic data. I didn't really have any other options. I dreamed, as I mentioned, of being able to get married to Caitlin. I started taking sirolimus, and a few years passed. We were able to get married, and this is our daughter, Amelia, who was born three years ago. This is a more recent picture from this Thanksgiving. We had a little baby boy three months ago. I look at these pictures and I think about the transformation that's occurred over the last eight years, thanks to proteomic technology that helped me to identify a drug, literally saved my life, and I think about my two incredible kids and my amazing wife, it's just incredible. You know, it means so much to me that I'm able to be here with you today and be a part of this panel. Thank you so much for doing that. You know, no one with an ounce of compassion can look at that picture without, well, smiling or crying at the same time. Again, this is an analyst meeting, and the reason that this picture's important is because platforms and tools make a massive difference to even a small number of lives. Th at's something that humans do. It's something that motivates researchers. It's something that motivates government. It's something that motivates advocates. So it's important for everybody to think about making big differences even in small numbers of people. What we're gonna move on to now, though, is to say, well, can we also make big differences in large numbers of people? David's explained how, yeah, there are a lot of rare diseases, but we're also going to think about things which are not so rare. And maybe the other side of, you know, maybe genetics has made a big difference to a small number of people. It hasn't really made a big difference to a large number of people. We're gonna see if proteomics can do the other side of the coin as well as helping people like David and bringing joy to his family. Can proteomics also have much more generalizable applications? I think we can stop the slides right now, and I'm gonna move on to—o f course, David will be involved in the discussion later as well. Let me move on to Rich, who's at Washington University. Rich, tell us about how you're using this—o h, sorry, I'm gonna go back and introduce this slide. We're gonna go through roughly three categories of applications here. We're gonna look at life sciences applications and how the features of the platform relate to the use cases. We're gonna look at some clinical use cases, and then we're gonna talk a bit more about personalized medicine, the Proteomics Personalized Medicine Initiative today. Yeah, back to you, Rich. Tell us about how you and your researchers are using this platform at WashU, and with the example of malnourished children kicking us off. Certainly. So as you know, we've been using the SomaScan platform for quite some time across a very diverse range of research here. One of the more intriguing projects that we've worked on over the last few years is with Jeff Gordon's lab, who some may know as often called the father of the microbiome. We worked with his lab investigating growth stunting in children that experience malnutrition and chronic GI infections, often due to lack of clean water sources. This is also known as EED or Environmental Enteric Dysfunction. The combination is known as EED or Environmental Enteric Dysfunction. This particular study was part of a larger collaborative that was funded by the Gates Foundation, and then while it was multi-omic in nature, with the goal of trying to understand how malnutrition and gut health were linked to growth, in the past, numerous studies had looked at using nutritional supplements to overcome the growth stunting that was often seen in these children. While they were able to see changes in weight, BMI, very few, if any of these studies, ever showed any impact on height, all the way into adulthood. It's really, you know, the nature of this is really never been very well understood, and we've never really even had good biomarkers, if you will, to be able to follow this. When the study was run, Using the SomaScan platform was part of this, as I mentioned. This is this larger sort of multi-omic study. In addition to changes that were observed in the gut microbiome, one of the most meaningful findings that we had came from the children using the SomaScan analysis of the plasma proteome. We had expected to see proteins associated with infections and nutritional status, the sorts of things that are very common in EED. What we didn't anticipate was that we observed a fairly strong signature for proteins associated with bone growth. This obviously caught our attention immediately when we saw this particular signature show up in the middle of the proteomic data. You know, we were hopeful we might see something like this, but we really didn't anticipate it. These markers were found to be lower in the children with EED versus the controls. This was really key because it demonstrated that the breadth and the sensitivity of the platform would allow us to detect in a very minimally invasive way. As you can imagine, anything that's really invasive is gonna be very difficult to use in this population because in addition to the growth stunting, some of these kids can actually be fairly sick as well, and biopsies are often very difficult to get. This really demonstrated the power of the platform being able to detect these markers from the plasma in a relatively minimally invasive way that we otherwise probably would not have been able to detect. I mean, there's picking up these types of secreted proteins in any other way, in a very high- throughput fashion, just really wouldn't have been terribly feasible. This was really eye-opening for us at the time in this ability to detect these key biomarkers associated with growth in this study, and basically give us a foundation to do further research. Thanks, Rich. I'm gonna ask you the same question as I asked David a minute ago in around the number of protein measurements that you need. Couldn't you have taken an educated guess and measured the smallest number? Why was it important to, you know, go with the largest available platform? Couldn't you have guessed about what was gonna happen? In this case, no. If we had been doing this study with what we thought we knew about what we expected to see, we would've spent our time tracking things associated with nutritional status and probably infection, and would not have come across these particular markers. The likelihood that we would have thought ahead of time to put these particular markers into a smaller panel is exceedingly low. Okay, thanks. I'm gonna stick with that theme for a moment and move on to Jason and Novartis. Yes. We've worked with Novartis for, I don't know, around a decade now, and Novartis has been a great help to SomaLogic in building the content of the assay. At one point, Novartis provided us with thousands of proteins in order to make the reagents that we've used to expand the content. The commitment of Novartis to building and expanding content has been there and is still there. Tell us, Jason, you know, why is it that Novartis is interested in measuring many proteins? Yeah, sure. Thanks for having me. You know, in Novartis, we view the platform, at least right now, as a discovery platform. The ways that we employ it within both discovery and development lends itself to the size of the platform and how many proteins can be measured. There is a lot of unknown biology, the network of biology, which pathways are active in diseases, as you've heard before. There's a lot of unknown biology of why adverse events happen when we perturb pathways, so we can use this platform as a hypothesis-generating tool as we're bringing either targets to be prioritized or therapeutics into the clinic and get first looks at what happens in a human when we perturb these certain pathways. That gives us a lot of information and confidence to continue to move either the drug program forward or mitigate potential issues that we're seeing. Then on top of that, we're in our late-stage programs, we're starting to or have been for a while now, profiling a lot of the patients that come in to start to disentangle the clinical definition of diseases. For instance, in heart failure, you have clinical definitions of preserved ejection fraction or reduced ejection fraction. We can break those down into sort of molecular definitions that allow us to reposition drugs into places where we see pathways are overactive, and we might have a compound that can mitigate that overactive pathway. How does this technical work affect the economics or productivity or the decision-making in drug development? The economics of it is it's a relatively easy platform to use within the clinic. It doesn't require a lot of extra site freezers and other things of how we handle the samples. Really, the economics comes down to basically mitigating unforeseen issues that we may see in the drug development pipeline. What I mean by that is, if you have a safety event appear that because of the biology wasn't 100% known around the target, that can be very costly later on in the different phases of development. By having this platform, having a broader look at the proteomic signatures that are coming out as we perturb these pathways, we get a higher level of confidence in moving these programs through safely. Thanks. Thinking, you've seen, we've all seen these very large publications using the SomaScan platform in tens of thousands of people, the deCODE study in 36,000 people. That's great that we can show that we can scale the platform to run that kind of capacity. Do the studies all need to be that big? How big are the studies that you're using the platform on generally? Yeah, you know, we originally felt the same way. Because of the size of the platform where you're measuring thousands of proteins in one shot, there is a statistical measure that you come against, which is a multiple testing correction. You're running all these statistical tests, and you have to correct for that number of statistical tests that the larger the platform gets, the more power you need to be able to detect the difference. What we are observing when we're running sequential tranches of patients is that even at the lower number of patients that we could run through the platform, say 30, that list of proteins that are changing while only a few of them are considered statistically significant, quote-unquote, "passing that multiple testing correction." As we moved up from 30 to 1,000 patients, that list of proteins just got more significant, but the list order, all things being equal, never changed. That gave us confidence in moving and actually starting to profile in our much smaller first-in-human studies, where we can get a first look at a therapeutic in a human with only 20 patients. We're starting to see some interesting biology come out of there as a first look. Because we're using this as internal decision-making and a hypothesis generation, we can then, and casting that net really wide, we can take those individual proteins that look interesting and actually test them clinically as the programs are moving along, where we would then be looking at statistical significance in a much more rigorous way. That is to say that, while true, you may need larger numbers to reach a statistical significance. It doesn't mean that the biology doesn't show itself at less samples, if you will. We've seen that [crosstalk]. You can get proof of mechanism of your drugs early, because of the precision of the platform. Absolutely. We also get a list of proteins that are either upstream of the target or downstream of the target that we can then monitor as we're moving the therapeutic through the pipeline. The importance of having those extra biomarkers around the pathway allow us to moderate dosing. It allows us to have confidence that we have modulated that pathway as strong as we would like to try to meet a clinical efficacy endpoint. Yes, that is absolutely true. Thanks, Jason. We're pretty proud of the precision of the platform. The coefficient of variation across 7,000 measurements is 5%. Bouncing back to you, David, how many samples did you use? You relied on the precision because you didn't have very many samples. How many samples did you take from yourself and use to reach these conclusions? When I was having my fifth relapse of this disease, and in the weeks leading up to it, I'd been collecting blood samples, every three to four weeks on myself and storing them away in the freezer, in the event that I would survive that relapse, thanks to chemotherapy and maybe be able to search for another treatment. I had been storing a lot of samples. Importantly, I also had a previous flare as well, a previous relapse of the disease I had to sample. When I did the proteomic platform, I could both look at what pathways are enriched in flare number five, and also look at what pathways were enriched in flare number three. What was exciting in this particular case was that mTOR was the most enriched pathway or potentially activated pathway across two independent flares in the same patient. You know, we're talking really small numbers. Earlier, we were talking tens of thousands of samples, right? I'm talking four or five samples right now. But when you can see replication within the same individual, different disease flare separated by two years' time, it makes you feel more comfortable that what you're seeing may be real. Of course, as we have thought more about mTOR and Castleman’s across other patients, we've now done these sorts of studies in dozens of Castleman’s patients: 88 Castleman’s patients, about 200 control samples. It's certainly, you know, a different sort of analysis when you have bigger numbers, but we've found it to be incredibly powerful. Thanks. We're gonna switch themes slightly now and think about quantitation and predicting, in quantitative terms, risk and actual phenotype and actual disease. I'm gonna jump to you, Peter, that you've been involved in quite a large number of discovery studies using the proteomic platform. Many of them have actually endeavored to create a quantitative disease or risk predictor. Tell us about how that comes about. How do you develop a quantitative prediction model from proteomic analysis? Good morning, everyone. Before I answer the question, I want to say I'm also one of the panelists, someone who is well-funded by the NIH. I have eight NIH grants at this point, which is quite a bit. What we do in part, as Steve mentioned, use proteins to predict the risk of diseases. To accomplish that, we have to find a group of patients who've been monitored for the occurrence of a disease. The disease could be heart disease, such as heart attacks or heart failure or strokes, or it could be cognitive impairment, Alzheimer's disease. It could be progression of chronic kidney disease to dialysis. It could be cancer, or it could be skin disease. I'm just involved in a project on atopic dermatitis or eczema. In each case, we have a blood sample within this group of patients. After the blood sample was taken, these patients are monitored for the development of the disease. Looking back at the blood sample and the protein patterns in the blood sample, and using computers and machine learning to analyze these blood samples, we can develop multi-protein risk models for each of these diseases. The diseases affect the human body from head to toe. As I said, it could be brain disease, heart disease, kidney disease, skin disease, cancer. I've called this blood test and the proteomic analysis of these samples that can inform the risk of so many diseases the executive physical based on a single blood sample. I'm thinking of democratization of healthcare, where anyone at a reasonable cost in the future should be able to have an executive physical through a blood sample. Not only can we inform the risk of diseases, but we can also through these protein risk models, monitor the effect of therapies to reduce the risk. Whether it's lifestyle changes or specific treatments, we can actually see whether a patient who was deemed to be at high risk of a disease was able to reduce their risk for appropriate interventions. I think this is very exciting, and I hope that in the future, this will have a major impact on how we take care of our patients. I'm also a clinician, by the way. Thanks, Peter. In your experience, when you've attempted to make quantitative risk prediction models on just from pure proteins, how well do they compare with available alternative risk scores, for example? Yeah. In each case, when we develop the multi-protein model, we use the best clinical model for comparison. Goal is for the protein model to be at least as good as the best clinical model. In virtually all instances, the protein model exceeds the performance of the clinical model. Not only do you have the convenience of assessing the risk from a single blood sample, but in each case, the protein prediction is as good, but typically better than the best available clinical model. That's our metric, that's our goal, and as I said, in virtually every case, we exceeded the clinical model's performance. Partly, you know, I'd argue that that might be due to the precision of the assay, or it's certainly a contribution. I know that some of the studies you've been involved in have hidden duplicate samples within a large dataset. What has that shown about the precision of the assay on duplicate samples? Yeah. We have actually, I would say at the urging of the NIH, we have sent SomaLogic some blinded duplicates where only we had the code to know what the samples are. By split duplicates, we mean we took the same sample, and we divided that in two. We would see how well SomaLogic was able to give us protein levels for the same sample that was divided in two. That gives you a measure called coefficient of variation. I would say for at least 50% of our proteins, the coefficient of variation was under 5%, and I would say for at least 95% of the proteins, it was better than 10%. These are, I would say, exceptionally good numbers in terms of how reproducible or how precise the assay is. We are actually publishing some of this, and I would say the NIH was pleased to see the good precision of the assay. Thanks, Peter. Also because you've been involved with SomaLogic for longer than I have, you know, longer than 12 years, you've seen different generations of the assay. You've seen it go from 1,000 to 1,129 to 1,300 to 3,600 to 5,000, and now 7,000. Have you seen over those generations what have you seen in terms of ability to make more, you know, the performance of these predictive models, has it changed? Is it better measuring more things? Yes. Let me tell the investors that I've been with SomaLogic since 2008, but I've accepted no financial remuneration. I'm in this collaboration for what I think it can do for patients, but not for any financial gain. As far as seeing the evolution of the assay, when I first got involved, the assay had 505 aptamers, quickly proceeded to 1,129, and now it's at 7,000 proteins. I would say with each increment in the number of aptamers, we are able to improve on our models, on the performance of the model. With the older version of the assay, which had 1,129 proteins measured, we had a nine-protein model. Currently, our best model—t hat's a cardiovascular risk model. Currently, our cardiovascular risk model, as Steve knows, consists of 27-proteins. I think that model is closer to being what I would call a universal model, meaning it can only predict the risk of heart disease, but it can predict the responses to treatments fairly universally, as far as we know, no matter what the treatment is, no matter what the biological mechanism of the treatment is. I think increasing the number of proteins gives us more predictive models, which may be more universal in terms of being able to capture the effect of therapies. Thanks. Then, shifting themes a little bit again about exactly what the modified aptamers are measuring. University of Cambridge recently published an example showing a protein called GDF15, which is pretty important protein. They showed a genetic variant was affecting the SomaScan measurements through changing the shape of the protein without changing the concentration of the protein. That shape change was related to phenotype, to outcomes. This brings us to the extent of protein biology. It is concentration, it is shape, it is interaction with other proteins. I know that you've been involved in some of the complexities of that conversation around what should we be measuring. What is the extent of the expansionist view of proteomics? Is it just the traditional view of concentration, or is it shape and interaction with other proteins? I know you've been involved with the GDF11/8 story from the beginning. Can you tell us a little bit about what we learned about those proteins and what we learned about measuring shape? Yeah. I think the reason that Steve you must be asking about shape is that aptamers are sensitive both to the epitope as well as to shape, whereas antibodies are more sensitive to the epitope, to the linear sequence of the amino acids, and they are less sensitive to shape. Aptamers they may have something that's fairly unique in terms of detecting shape as well as the epitope. Question is, does shape provide any additional information? I would say rather, I think the GDF-11 and GDF-8 story was interesting, but we have even I would say more interesting story with another protein called tenascin. For tenascin, actually, SomaLogic has five different aptamers. In a study where we were using aptamers to assess cardiovascular risk, and we used five different aptamers for tenascin in the same assay, we were surprised that the machine learning process actually selected three of the five tenascin aptamers for the risk prediction model. I would have thought, well, the tenascin aptamers should be redundant, the machine learning process would only select one of them. This part actually gave us the same— Just to clarify, Peter, machine learning only chooses proteins for inclusion in a model if they're providing unique information. This was a surprise— Correct. That we've got three aptamers to the same thing. The implication is it's something different. Correct. I think what the machine learning does is it chooses aptamers or proteins in terms of providing complementary information, not redundant information. We thought, "Well, this will be redundant information. You're measuring tenascin with five different aptamers." Actually, three of the aptamers were selected into the risk model, suggesting that they all provided independent biologic information that machine learning was able to detect. We do think that the shape actually does matter and that it provides some unique information for each of these aptamers. We've seen it for tenascin, and we've also seen it for the growth differentiation factor proteins 11 and 8. Thanks, Peter. We heard Roy earlier mention that one of our, you know, relatively new initiatives, based on some of the findings that you and your collaborators have shown us, that actually developing multiple reagents to the same protein is something that we're going to pursue. Uniquely, again, for aptamers, the scale economics of developing multiple reagents to the same protein that bind to different parts of the proteoform. The scale economics of doing that for aptamers are very attractive. It's relatively low cost from the same selection process to get multiple aptamers to the same protein. I'm sure that's something that you'll hear more about in the future. Moving on, we're gonna talk about collaborations with the National Institutes of Health, and Peter, you've mentioned a couple of those already. Obviously, it's important for a technology platform like this to be welcomed and accepted and to attract grants from government entities, and obviously the NIH is one of the most important ones in the world. Rich, I know you've had some experience of using the SomaScan platform to attract NIH grants. Tell us about one or more examples of where you've seen that work and why you think it worked. Certainly. Yes, we've actually had numerous groups here with NIH-funded grants that have used SomaScan as a major component of their research. One in particular was just recently published in Nature Neuroscience back in the fall from Carlos Cruchaga's lab. They were looking at neurological disorders with a particular emphasis on Alzheimer's. This was actually a proteogenomic study, if you will, where they identified what's called pQTLs or protein quantitative trait loci. This is where they took basically the combination of protein data along with genetic data looking for connections, if you will, between mutations in the genome and proteins that were associated with different outcomes. Within the study, they processed about 2,000 total samples from cerebrospinal fluid, blood, and brain with SomaScan. Again, kind of speaking to the abilities of the platform. While they did find some proteins that were specific to certain tissue specimen types, they did find a number that were in common and were very interestingly related to different forms of neurological disorders. Like I said, AD was one of the primary neurological disorders they looked at. They also looked at Parkinson's and a few others. They were able to find some very intriguing proteins and potential new targets and pathways for new targets in multiple neurological disorders in that one study. That was just an example of a study that was NIH funded. It was where SomaScan was actually the centerpiece of the study and produced some very impactful results. Thanks, Rich. Peter, you mentioned earlier that you've been the PI or collaborator in eight different NIH-funded studies. What is it about the SomaScan platform? What features helps it to attract a good score when it's being peer-reviewed, when a proposal, a grant proposal is being reviewed for innovation and attractiveness? Well, the proteomic research prior to SomaScan consisted of measuring a handful of proteins. We would call it rounding up the usual suspects. You would measure a handful of proteins, and you would see if they related to the outcome of interest. You're basically testing something that you already suspected. That's limiting you because you had to have already suspected the relationship. What SomaScan has allowed is for us to measure thousands of proteins, so we can learn something that we didn't suspect. I could give many examples of that. We are also involved in Alzheimer's research and published in Nature Aging paper on some proteins that weren't suspected of being related to Alzheimer's. We've identified proteins that may be related to the progression of chronic kidney disease. Again, that relies on a combination of proteomic and genetic approaches. I think part of the excitement is that we can learn about things that we simply didn't even suspect. That's really the key advantage of having such a large platform. I would say the other thing that the NIH has found helpful and actually required, and SomaLogic has been very forthcoming, is making all the proteomic data available, not just to the initial investigator who got their grant funded, but making the proteomic findings available widely to other investigators. For example, I have a grant on progression of chronic kidney disease that use SomaScan in a population of patients who are monitored for progression. Yesterday, I was contacted by an investigator from another university. That she was just funded to study cognitive impairment in these patients using the data that SomaLogic has made publicly available to any other investigator. This openness that the NIH requires, but also that SomaLogic has completely adhered to, is what I would call a multiplier effect. You're not relying on a single investigator to make an advance, but making the data publicly available, many investigators will be able to take advantage of this data, again, for the benefit of patients. Thanks, Peter. Yeah, by doing that, then their platform becomes a de facto measurement standard across different conditions. On Wednesday, Peter, SomaLogic announced a collaboration with the NIH in the MESA study. Congratulations to you because you're the PI for the proteomics part of that study. It is a landmark study. Tell us a bit more about why it's a landmark study, why it's gonna help proteomics, and why it's gonna help patients. MESA, which stands for Multi-Ethnic Study of Atherosclerosis, is a study that is funded by the NIH and looks at the predictors of atherosclerotic heart disease outcomes, meaning heart attacks and strokes in a multiracial cohort. It looks at Whites, Blacks, Hispanics, and Asians. The other strength, particular feature of this cohort is that they not only have these heart outcomes, the heart attacks and the strokes and the like, but they also measure subclinical disease before it becomes clinically apparent. By subclinical disease, they measure coronary artery calcium. They measure cardiac function by doing cardiac MRI. We can develop protein models that not only predict outcomes, but we can develop protein models that can actually predict subclinical disease in its earliest stages when the disease may be most treatable. We can do it based on race, and we can do it based on gender or sex. This is, I would say from my perspective, among all the NIH-funded cohorts in cardiology, the MESA cohort may be now the premier cohort. It used to be Framingham, but I would say today it's MESA. I'm excited that SomaLogic is able to collaborate on it with us. Yes, thank you for helping to create that collaboration, Peter. Broadening back out now to, we're gonna hear a bit more later about the business side of the collaboration partnership with Illumina. Rich, thinking about as we go forward, you've been using the hybridization array readout, and it's been very successful and very precise. How do you think the addition of another readout method, you know, that of sequencing the Aptamers, as well as having at your disposal the hybridization method, how do you think sequencing will actually help you and your ability to run SomaScan in your laboratory? There are really a couple of reasons that this complementation is kind of attractive to a lab like ours. You know, we have substantial processing capabilities here for the array, and we'll probably continue to grow those. Really that's probably the only thing that will limit the amount of samples that we can process here is physical space and equipment. You know, we're primarily a genomics lab, but SomaScan is a very, very important proteomic component to what we do here. You know, the ability to be able to use my sequencers to process even more samples than we can process with just our array processing platforms alone is tremendous. Because again, I think that's—t here are people and equipment and space that are probably the only things that are going to limit the amount of SomaScan work that we can do here, from past experiences. The other is probably a technical component. We see this in the transcriptomic world, where there are occasionally cases where people are looking for very high sensitivity for very low expressers. While SomaScan already has a very impressive dynamic range for an array-based platform, the ability to sequence really deeply for something that could be a rare species, or you know, changes in rare species, is attractive. We can certainly see it complementing our existing workflows on a number of levels, moving forward. You know, I have about 30,000 sq ft of space for my laboratories, but we're pretty much landlocked, so the ability to be able to use multiple processing workflows for the system is really attractive to us. Thanks. Jason, I think Novartis already has some experience in combining sequencing and proteomics. Yeah, absolutely. Rich already touched on this around looking at pQTLs, the protein quantitative trait loci, where you combine, in this case, genetic information with the proteomic data. We find that the two datasets actually complement each other very well, and actually, we can make additional discoveries by the combination of those two. What I mean by that is, if you were to just run genetic association, you end up with a potential list of mutations that are associated with your disease. When you go down to try to figure out, well, what is the functional outcome of that association, the pQTLs help you co-localize into a specific space of what the functional mutation is. We've been leveraging that. Also, the combination of the two datasets to do some much more sophisticated machine learning called Bayesian graphical models, where we can look at causal chains of pathways using the genetic data as a fixed variable in the system, but the proteomic data can be modulated along with the clinical outcomes. They are very complementary to each other, and we internally call it, as Rich, said, proteogenomics. It's this proteogenomics efforts on all these different clinical trials that we have. Thanks. Sticking with the clinical trials theme, in 2003, the FDA published a Critical Path Initiative where they asserted that biomarkers and surrogate endpoints could lead to more efficient clinical trials. That's quite a long time ago. We haven't seen new surrogate endpoints emerge in the last 17 years. Peter, tell us why you think that happened and what the recent progress has been on developing a candidate surrogate endpoint for cardiovascular outcomes. Well, as you know, the concern that the FDA had is really the inefficiency and the expense of developing new drugs, which typically can be in the ballpark of $1 billion-$3 billion for a cardiovascular drug. They suggested it would be important not just to rely on expensive clinical outcome trials, but perhaps there's other indicators, some surrogate endpoints that could guide the development of drugs. We worked together with SomaLogic, with Steve and colleagues, in developing a cardiovascular risk prediction model that consists of 27-proteins that I alluded to earlier may meet the FDA requirement for being a universal model in the sense that it could capture the effect of drugs, whether they are beneficial or potentially harmful, regardless of what the mechanism of action of the drug is. What we've done with this 27-protein model is we've first shown that it predicts cardiovascular risk across different populations, across different races, across both sexes, across all continents. It was already universal in its performance across various groups. We also looked at about six different pharmacologic or lifestyle interventions and showed that a blood sample early in the study, the study was a pharmacologic study, a lifestyle study, a blood sample early in the study predicted the clinical outcome several years down the road. We could show whether the outcome is beneficial, but in one case that we've already published, where the drug in development turned out to be harmful, a drug called torcetrapib, developed by Pfizer, using a previous version of this cardiovascular risk model. Already at three months, we could predict that this drug was going to be harmful in terms of increasing cardiovascular risk and increasing mortality. Having a proteomic tool allowed us early in the trial to say, "We told you so. Two or three years from now, you'd be terminating your trial based on adverse events, and we could have told you that much earlier in the study based on these surrogate protein models." I think that's one application of these models, is to give an early indication to pharmaceutical companies whether their drug is likely to be helpful or potentially harmful. It can also be used to enrich the population in the trial for being a high-risk population. We can select patients based on these protein models to be particularly high risk, which reduces the sample size. It makes the trial much more financially affordable if it can focus on patients who are likely to have a large number of events. Both from the standpoint of predicting the outcome of the trial and for making the studies more affordable in terms of enrolling high-risk patients, I think this 27- protein model and similar protein models will make a big difference. I think that's the reason that the FDA has been so interested in this work. Thanks, Peter. Jason, I know that we were thinking about using protein measurements t o detect whether drugs work or not. I know you've been thinking about Novartis using this to evaluate products that you might be going to in-license. Tell us a bit more about that. Yeah, absolutely. I'll start with a study that was run a few years back called CANTOS, which tested the hypothesis that lowering inflammation in cardiovascular disease would be beneficial with a drug called canakinumab. In that trial, we used a standard way of measuring inflammation in those patients, which is C-reactive protein, CRP. It's accepted that it's not the perfect proxy for what the outcomes of that cardiovascular trial will be, and so we could use the platform, the SomaLogic platform to test, as I mentioned, the large fishing net to cast and say, "Could we find a better alternative to CRP to run in future trials?" Along that path, last year, I believe, the year before, we in-licensed a molecule for targeting the inflammasome, which is a way to modulate inflammation. When you marry those two efforts together, we start to basically take a study that we already had an outcome on to try to find a better proxy for inflammation in cardiovascular outcomes. We can then carry that hypothesis forward into now our in-licensed compound as we're beginning that journey from first-in-human through different clinical trials, like the inflammasome inhibitor we used for looking very quickly at whether it had a beneficial effect in COVID. Now we're looking at a bunch of other therapeutic areas while carrying that hypothesis forward. We find, again, that platform of casting that wide net. We can now narrow down our search into a very specific space, trying to find a biomarker, in this case, a proxy of inflammation that we can use as inclusion, exclusion criteria in a trial using a similar compound but a in-licensed one. Thanks. While you have the stage, Jason, I know that Novartis has also been interested in NASH and used the SomaScan platform to home in on a particular protein. Tell us a little about how that happened and then what you're doing with that individual assay. Yeah, absolutely. We, like other pharmaceutical companies, focused in on NASH over the last three years or so. In that effort, we set up numerous academic collaborations to get access to samples. We had internal samples in our own human tissue network that we started profiling specifically for the purpose of profiling with the SomaLogic platform, specifically with the purpose of identifying a protein signature that could help with diagnosis of NASH, which is an under-diagnosed disease because it's very invasive to actually diagnose, but also to identify those patients that would be potentially faster progressors. With the combination of all that data, there was a signature that was identified, a set of proteins that seemed to mark fibrosis and progression. Over time, as we were bringing compounds into the clinic, we furthered that validation study in the sense of narrowing down what that signature looks like. Since then, we've started to step back a little bit from NASH, what has happened is that signature has been transferred to an external company to develop as a diagnostic signature for NASH that we would then use opportunistically if we have other NASH programs coming into the clinic. Thanks. Rich, I know that you told us earlier about how J eff Gordon and your collaborators had looked at proteins in malnourished children. How are they using that information to look at what therapies might be useful? Yeah. As you might imagine, that protein signature that I mentioned earlier was not just useful for trying to get some understanding as to how EED was working, but also for looking at potential interventions, right? That signature that we saw for bone growth was actually used to look at the impact of different nutritional supplements, first pre-clinically. Gordon Lab actually has a mouse model that replicates some of the aspects of EED. What they discovered was that of the different nutritional supplements that they used, even though all of them led to a weight gain, actually only one of them modulated the bone signature that we observed in the blood with the Soma platform. That has led to follow-up studies now in children with that particular nutritional supplement. We're obviously—i t'll be interesting to see how the results of those different studies turn out. This was a case where basically the nutritional intervention or therapeutic, if you will, in this case was picked based on the profile that we observed from the children in the original study. In the interest of time, I'm going to move to the personalized medicine initiative in a second. Rich, while you're speaking, I know that you had also another piece of work at your institution where you had worked to redefine disease molecularly in different subgroups of heart failure. Maybe you could just touch on that before we move on. Yes. Doug Mann and I co-senior authored a paper about a year and a half back in JACC, where we looked at the plasma proteome in heart failure subjects. In heart failure, left ventricle ejection fraction has been used as a way to kinda categorize individuals for follow-up and treatment for quite some time, looking at whether they have preserved ejection fraction or reduced ejection fraction. We did a study with SomaScan on about 200 subjects from plasma. There's also kind of this group called mid-range or borderline. They're kind of in that gray zone of ejection fraction as to whether they're preserved or reduced. What we discovered was not only was there considerable molecular heterogeneity within some of the traditional groups of HFpEF and HFrEF, as they're called in shorthand. We discovered that the mid-range group was molecularly quite distinct. Even though there weren't a large number of subjects in the trial, it basically led us to the conclusion that we needed to look even further and deeper into these cohorts because this could significantly impact treatment decisions down the road. To your point about precision medicine, this was an initiating study. There's more work to be done on larger cohorts. If these results were to hold up, it would potentially point at groups that had been categorized at broad strokes before and allow us to use molecular signature to better define them in the future for treatment. Thanks. The final few minutes is about SomaLogic's proteomics personalized medicine initiative. Peter, as a member of our medical advisory board, you've been helping us develop this strategy for a number of years. Now we have collaborations with six different health systems, and four of those have already started using the cardiovascular risk predictor that you described for us, Peter. In this case, it's an interesting application because it's taking people with diabetes who are eligible for enhanced cardioprotection per guidelines, but they're not on it. This is pretty typical in that the uptake of these drugs is often less than 20% of the eligible population. We're determining whether this providing the practitioners with enhanced risk prediction information will enable a more individualized and cost-efficient application of these drugs and will enable people to be on them who weren't on them before. Peter, tell us a bit more about this context of use and about, you know, how important is this kind of application? Well, Steve, as you're describing, the SomaScan is able to identify high-risk individuals for whatever the disease is, whether it's diabetes or other diseases. One is able to then show the so-called report card to the patients or their providers. Hopefully identifying patients as high risk will encourage the providers and the patients to be compliant or to undertake therapies that have been shown to be effective to reduce the risk. Again, having a system where we can, using proteins, precisely define what the risk of each patient is and target the high-risk patients in particular, I think will be a good way to allocate resources within the healthcare system, but also it will encourage individual patients and providers to be compliant with guideline-directed therapies. I do think it's very important both for individuals and providers and healthcare systems. Thanks. Just to finish up on this subject. David, so you have done literally personalized medicine. Do you think this kind of approach is expandable outside of Castleman's and outside of cardiovascular disease? Is it possible that this could be a large impact to a large number of people? Absolutely. Beyond just personalized medicine for myself, we've used SomaScan to identify a predictive biomarker for patients who respond to the only FDA-approved therapy, siltuximab, so we can predict whether someone's likely to respond. We also identified an indicator that they actually are responding or aren't responding. We found a repurposed drug, another drug called ruxolitinib, that we started to give to Castleman's patients, again, based on proteomic data. Even a new therapy that targets CXCL13 that's never been used in humans that based on proteomic data in Castleman's, we're now developing. You know, we've expanded from me to the broader Castleman's field. As you said, I think what's so exciting is to think that for so long in medicine, we haven't had tests that could ask the question, "What's wrong?" We've had tests that could ask, "Do you have this or do you have that?" I think SomaScan provides this possibility to ask the question, "What is wrong and what can fix it?" When I put my patient hat on, I just can't think of anything that's more exciting. Thank you. Thank you to all the panelists. I really appreciate your presence here and the work that you've done that underpins the opinions and thoughts and conclusions that you've reached for us today. Now I think we have the next 15 minutes of questions for the panel from the analysts, which I believe Roy will moderate. That's right, Steve and listen, and thanks for, as usual, your expert moderation of a great panel discussion and sincere thanks to David, Jason, Richard, and Peter, you know, for leaning in as well. Well, now I'll turn it over to the operator for our first Q&A session. Operator. Thank you. We will now be facilitating our first Q&A session with questions from SomaLogic's covering analysts. We'll take a moment for your questions to be submitted. Please press star one to get in the queue. While the operator is collecting those questions, I have a whole ton that I would like to ask. I want them to have most of the time, but I do have one quick one for Peter. We've noticed in our first 20 LDT tests that there are 1300 proteins in the models that no one else currently measures. What do you think this suggests about the importance of understanding the expression of the entire canonical proteome and beyond the diagnostics based on pattern recognition? You referred to the importance of this earlier, but what do you think the importance of moving up in content will be to diagnostics? When we think about drug development or understanding of diseases, we measure individual proteins that associate with the disease of interest, and then we organize these proteins using various software tools along biological pathways. The more proteins we measure, the greater the confidence we have that we've identified the correct pathways. I won't get into the statistical details, but again, if we could measure every protein in every pathway, which I think hopefully is where SomaLogic will be at some point, that would give us the best prediction as to which pathways are involved and what diseases and what the appropriate treatments might be. I think the content is important. Great. From the phone line, sir, we have a first question from Dan Brennan at Cowen. I'm sorry, Dan, please queue up star one. I lost you. Okay, Dan, your line is open. Thank you. Great, thank you. Sorry for that, everyone. Thanks for the day. It's really informative. A ton to cover. Maybe to start off with Jason. Just would love an update, really exciting everything that Novartis is doing. Just could you give us a sense of just how extensive across, you know, Novartis' biomarker discovery programs like SomaLogic is being used today? And what would cause the company to expand the usage? Is it anything on the technological front that you're looking for? Or is it just as time goes on, you know, impact is growing that, you know, maybe you'll get more confident? That's the first question. Then the second question is just on clinical trials. I think you discussed you are using SomaScan, you know, profile patients in different ways. Just kinda how routinely is SomaScan being used today at Novartis in clinical trials? Just, you know, how do you see that, you know, progressing over time? Yeah, absolutely. Thanks for the question. If I recall, I think SomaLogic has touched on every therapeutic area we are looking at oncology, gen med. Within gen med, it's broken down into seven different therapeutic areas. It's touching on almost every program in there. Part of what we had established with the SomaLogic partnership a couple years ago was around doing this activity by default. The idea being every Novartis patient that is recruited into a trial would have SomaLogic data generated on them. Over the last couple years, where our rate limiting step is availability of samples. Because Novartis recruits somewhere, you know, between 30,000-40,000 patients a year. In part of the collaboration, we actually went back into our biobank, our human tissue network, and sampled programs that were of strategic interest. We had collected the samples that are in our freezers, and we were generating data on those strategic areas of interest like heart failure, et cetera. We've also done some collaborations in kind as COVID and the pandemic was starting by contributing basically our SomaLogic assays that we have through the partnership to academic institutions to help further the science of COVID and what's going on there. That’s really the breadth that we’re going after, which over the last couple of years is basically touching every therapeutic area that we have. I think when we look out at the future, we’re actively, as we talked about, actively looking at ways to both use it in early discovery to de-risk programs, prioritize programs, but also now, as the last part of this conversation started moving, was how do we employ it or the combination of it and other data modalities as ways of targeting subpopulations that we think are much more efficacious for a drug. We started using this data to look at combinations of proteomic data with digital endpoints, proteomic data with genetic data, as I talked about already. We sort of see in the clinic. That's the future we're going after is really how do we narrow down into populations that we believe would have most benefit from our therapeutics. Hopefully, I answered your question, two questions. Great. No, thank you. Maybe to follow up, I'll switch gears and pose a question to Rich. Rich, just wondering, you know, the Illumina relationship is kind of an exciting one in terms of this NGS readout and something I think Wall Street is, you know, trying to figure out what the opportunity is. From your own experience thus far, you know, being on the genomic side and, you know, seeing the benefits of, you know, kind of the multi-omics, how do you think about, you know, the applicability of, you know, conducting multi-omics experiments as this NGS readout plays out? Like, is this something you think will be just maybe help us think through, you know, some of your peers and colleagues across the market. Just how broadly applicable we'll be doing this multi-omics experiment. You know, for an NGS readout, I guess it sounds like the throughput is a significant advantage for you and utilize your instruments in the lab. Any other kind of capabilities or benefits of using NGS versus other, you know, the array-based approach? Thank you. Sure. Yeah, I mean, obviously this is, as I had mentioned, a very significant announcement, especially for labs like ours. I fully expect that within the next 18 months, we will probably be approaching with proteomic samples the numbers of samples that we run currently with transcriptomics, which has had about a you know 20-year head start in the omics space. I think that's huge. The ability to use both the array-based format as well as the sequencer-based format will be tremendous for us. In regards to the complementarity of the systems, multi-omics is where everything is heading. Almost every study I'm collaborating on right now with most of my major collaborators is all multi-omics. Everything that we're doing is multi-omics, because no one technology, no one omic technology gives you the complete picture, when you're trying to understand disease, interventions, any of it. Biology is extremely complex. There's no other way to say it. Without multi-omic platforms to be able to investigate, you literally, you know, you just don't know what you don't know in a lot of cases, right? This is huge. I mean, in nutrition, inflammation, immunology, cardiovascular, neurosciences, virtually every major collaboration we have now, most of the NIH grants that are going out are multi-omic in nature. Did that answer your question? There were a number of questions in there. I'm not sure I caught them all. Yeah, no, I think that was terrific. Maybe just a final one, then we'll hand it back to the operator. Peter, the 27- protein model sounds very exciting. Just, can you help us think through, like, what are the next steps here in terms of what the implications of this could be? You know, it sounds like this could have broad applicability, but just wondering, maybe this is gonna take a long time to really play out more broadly in the market. Peter, I think you're on mute. We're not hearing you. I apologize. I would say the next step would be to increase the number of therapies that we have examined using these 27-protein models. We, you know, we have some, I would say, glaring omissions there, if you will. We were not able to look at drugs like PCSK9 inhibitors. We were not able to look at SGLT2 inhibitors. I think we have to keep expanding the number of drugs that we have, that we will have looked at with this 27-protein model to see if the model is truly universal, if it's truly able to predict the effect of a drug irrespective of the mechanism of action. We still have some homework to do on that side. I would say on the clinical application of this model, can this model be used for prediction? I think we've done good work in terms of showing what will be expected by guideline committees, which is to pay attention to races and sexes. I think the ultimate goal would be to actually convince the guideline writers and the guideline committees that a model like a 27-protein model would be suitable as a test to be applied in a clinical setting. There's always work to be done with incorporating a model like this into practice guidelines. That, in my mind, would be an ultimate goal for patient care. Thanks, Peter and Dan. Operator, do we have others on the line? Thank you, sir. Next question's from Kyle Mikson at Canaccord Genuity. Please go ahead. Great. Thanks. Thanks everyone for the great day, and the panelists, really appreciated hearing your stories. Very, very powerful. I had a two-part question. The first for Jason. Jason, as we get closer to the clinical, like I'm thinking translational bio from a clinical trial research, you know, we've heard that typically researchers know the set number of proteins that they're looking for, and usually it's not, you know, thousands. Is there a reason why, you know, very high plex would be necessary or preferred, or maybe the ability to de-plex, like starting from, you know, a high number and then scaling down to lower plex to home in on those proteins would be kind of a unique and attractive attribute. If you could talk about the value to your research that the super high- plex platform adds compared to the lower- plex assays with potentially higher sensitivity, that'd be great. Then the second question I have is for Peter. Peter, you mentioned the example of three aptamers per protein that you found. Very interesting. What would, you know, more than one aptamer per protein do for specificity, clinical utility, and some of the other metrics that we care about in proteomics? Thank you. Sure, I can start off. Thanks again for the question. I view this as the way that the other sort of omics technologies went. I was involved in very early days, actually at Washington University in St. Louis, around the microarray explosion back in the late 1990s. One of the drives was how do we get everything on a single measurement platform so that we could get a look at as much biology as possible. I view this platform heading in that direction specifically for the purposes that I had mentioned. There's so much unknown biology out there that as you're measuring as much as you can, you may discover new biology. You can mitigate, for instance, for us, off-target effects that we had no idea would even show up because we can actually take a look at those pathways that were modulated with a platform like what SomaLogic has. We also through high-content screening can identify targets without necessarily really understanding the mechanism of action in a platform that allows you to look at as much biology as possible, helps us understand what the patient outcome might be, what safety risks might happen, etc. The farther that you can expand the platform and the bigger that you can make it makes for a very nice application to do, as I mentioned, the sort of hypothesis testing. I also mentioned that one of the ways that we're using it is to cast that net very wide so that when we do get potentially to later sort of registration trials, we already have a smaller set of proteins that we can then take through clinical validated assays and measure those, as I mentioned with the CRP alternative. We view that the platform, the bigger is the better as we're going after what is the biology that is happening both pre-clinically but also in the clinic with it, with our patients. Hopefully that answers the question. Regarding the question, Peter Ganz here. Regarding the question of allowing multiple or utilizing multiple aptamers for the same protein, would that improve risk prediction? Should there be an effort to have multiple aptamers for each protein? Because right now most proteins for SomaScan have a single aptamer. We've actually gone through the exercise, as I mentioned, for two of our models, the cardiovascular risk prediction model that we developed in patients who have chronic kidney disease, and a second model that predicted progression of chronic kidney disease. The machine learning process selected several, as I said, tenascin molecules. It also selected another protein called Sushi, von Willebrand factor. It actually has a very long, lengthy name, I'm just mentioning the first few names in the protein. We've gone through the exercise of actually forcing the machine learning process to only include one aptamer for each protein. What we found was that the risk prediction slightly degraded, meaning it wasn't as good when we forced the machine learning process to only select one of the aptamers. We do have some numerical evidence that allowing several aptamers into the model may offer some advantage, and that's really been limited by the fact that most proteins only have a single aptamer. At this point, we don't know the full potential behind allowing multiple aptamers for each protein into the SomaScan. Based on the data we already have, there is at least suggestion that multiple aptamers for each protein would provide additional information for risk prediction. Well, listen, thanks again to the panelists. Thanks again to the panelists for, you know, for a great discussion. Thanks to Steve for moderation. Thank you, operator, whom I just interrupted for moderating our Q&A session as well. We're now going to move on to take some time to go deeper on our recent announcement with Illumina. We're grateful that Joydeep Goswami, the Chief Strategy and Corporate Development Officer at Illumina, has agreed to join us. It's been a pleasure to get to know Joydeep and his team over the last few months leading up to the announcement. Welcome, Joydeep. Thanks, Roy. Hopefully, you can hear me. I can hear you great. I've got a couple of questions for you, and if we have some time, you can quiz me as well at the end. First of all, there's obviously a number of companies that Illumina could have worked with to develop this new NGS solution, and obviously other reagents that could have worked in this system, maybe not quite as smoothly, but could have worked as well. What were the factors associated with your desire to work with us? You know, obviously there are a lot of different solutions. Proteomics is not a new field, but what we were looking for, and there was clearly an unmet need in the market, was for a very high plexity, which I mean thousands of different proteins at the same time or tens of thousands of different proteins, eventually moving to full proteome coverage, married with high throughput, which is a lot of samples to be analyzed at the same time, right? And with high specificity, sensitivity, and dynamic range. That's the solution we were setting out to solve that unmet need. We looked at a range of technologies, but we found that SomaLogic's SOMAmer technology really could best address this. You know, here we could start with greater than 7,000 proteins, rapidly iterate to get more protein coverage. The technology was very compatible with NGS. We could move it, we thought, to a simple automatable workflow, and, you know, eventually make it more amenable to, you know, what the last speakers were talking about, right, a more of a proteogenomics multi-omic workflow at the end of the day. That clearly excited us. We spoke to a lot of KOLs that have used SomaLogic's technology before and endorsed it. We were able to validate the technology in our own hands, right, in collaboration with people at SomaLogic. Honestly, we found your team to be very collaborative and deeply rooted in science, which is something we look for and appreciate. You know, we felt that there were a lot of complementary skill sets and a willingness to jointly address the market, right? All of us, all of this is important. Not to mention, right, the shared vision of improving human health through this multi-omic approach. You know, of course, Roy, after that, it was working with a brilliant CEO, which would seal the deal for us. Of course. The most important consideration. Joydeep, we view this deal as an incredible win-win, you know, on a number of levels for both Illumina and SomaLogic. It's actually the benefits being relatively equally weighted between the two. Do you agree with that, and why would you agree with that? You know, I do agree with that, right? Look, I think, this is—it's a win-win-win. I think, you know, I would call out this is a win, not only for Illumina and SomaLogic, but for customers, too. Now, our platform is an open platform, so, any player in the market is open to develop proteomic solutions on our platform. But we do believe that this solution, that we will create jointly, uniquely addresses this high-plexity, high- throughput need, that I talked about. You know, I think, between us and this, you know, I think there is, as many of the previous speakers mentioned, right? There is an unmet need to address, proteomics, but also multi-omics needs for experiments that customers are trying to do at scale, right? That's an important piece. I mean, you could always do things at low scale before, but this is taking it to a whole other level. You've seen this before. You've seen this story unfold in the genomics and in the transcriptomics world, which really changed the speed at which new discoveries can be made. We're very excited to bring that to customers in the very near future. Now, for Illumina, I think, you know, the win clearly is to be able to leverage SomaLogic's incredible knowledge of proteomics and aptamers and start with this current base of more than 7,000 SOMAmers and rapidly create and innovate more SOMAmers, but also then to translate that into an NGS workflow that is simple and automatable. Making it easy is a big part to reducing the barrier to new innovation and new discoveries. In addition, we look forward to working with you to solve the [back end], right? It's very easy to get the data, but then you have to analyze the data, and I think we believe we could work together to develop bioinformatics tools that could make analysis of the results easier for customers. You know, for us, this sort of a partnership continues to improve the utility of our instruments for our customers, who can now go from DNA to RNA to protein on the same instrument and eventually in the same experiment as well. I mean, for SomaLogic, I do see this as a way it opens up, you know, a way for you to leverage our global install base very quickly, and easily and really access customers that are familiar already with how to use NGS to address problems. You know, you do get access to our experience in developing an NGS assay and on kit development. It really is the fastest way, I believe, for SomaLogic to reach a wide array of customers, and the best way for you to leverage the scalability of your technology, right? I think you and others have mentioned that, you know, it is really the best marriage of technologies in this high- throughput, high-plex area, both from your technology and ours. Those upfront fees hopefully help to you know put money back into your core technology development and the diagnostics business. I think the downstream royalties will continue to enable you to get broad participation while leveraging our infrastructure in on a global scale. Yeah, I think also there's an advantage I think as you get access to more customers some of the proprietary protein pathway databases that you have talked about right? Will be easier then for these customers to leverage and for you to you know you to gain access for them for therapeutic and diagnostic discovery purposes. Truly a win-win-win situation. Great. I agree with that, and I appreciate you calling out the, you know, the win for the customers. Of course, downstream from that, are the wins for the patients that are at the very end of the value chain, of what we both work on, and those who wish not to be patients in the future as well. That's right. You mentioned diagnostics a little bit there at the end, so you know, while we consider our foundational business to be our life sciences tools solutions, we have developed a new class of diagnostics that leverage our ability to measure thousands of proteins at a time and use the bioinformatics capabilities we developed over a decade to create these incredibly powerful diagnostic models. Of course, we are laying the foundation for the development of a diagnostics market in addition to our life sciences tools. What benefits does Illumina potentially provide to us at SomaLogic in the development of this diagnostics market moving forward? I think, you know, Illumina has obviously made a commitment already to support distributed diagnostic solutions on our platform, right? This allows the solution to reach a wide variety of customers across the globe. We've already made our MiSeq, which is our low- throughput platform, and our NextSeq, our mid- throughput platform, available as DX instruments, and, you know, set up an installed base across the world. You know, these have been widely adopted. We intend to introduce our NovaSeq 6000Dx, which is our high-end platform, or high- throughput platform, into the market this year. SomaLogic can instantaneously access this install base of instruments globally, and of course, the support network and infrastructure that goes along with it. We already have a track record of well-recognized DX players like Roche and QIAGEN who are developing content on our platform. Of course, that's on the DNA side, but still, right, there is the well-trodden path that SomaLogic has access to. In addition, we continue to make progress on our Illumina Connected Analytics platform, which allows for partners to be able to support customers who prefer a web-based informatics infrastructure to support their assays and provide much more flexibility on that. Lastly, you know, I think SomaLogic will be able to tap into this infrastructure whenever it decides to kind of take some of the work that you've already done and decide to move it into an IVD model, right? Which then allows you to reach ever more customers that don't wanna ship for whatever reason solutions to a centralized lab. You know, while this might start with proteomics tests, I mean, you know, I think you should be really excited about the opportunity to then maybe combine this into multi-omics tests with other suppliers or other partners perhaps, right? I think those are the key advantages I see, Roy. Yeah, we're excited about all of those things as well, Joydeep. You know, the ability to run our diagnostic platform, our diagnostic products on a variety of platforms is important. I particularly like NGS for this concept of combined, you know, diagnostics and what we all believe that's gonna drive in the future. In the last couple minutes, you have a question for me? I do. I may have two questions. Well, let's see, you can pick them in any order if we run short of time. I did want to ask you the reverse of what you asked me first. You know, what attracted you to Illumina as a partner? Then maybe the second one, since we've talked about diagnostics a little bit, what's your vision for diagnostics based on your platform and based on some of the work that you've done? Yeah. Well, the decision to work with Illumina was really pretty easy for us. You know, while obviously the science of next-generation sequencing and sequencing is advancing and other players are coming into the field, Illumina is the premier provider of these services. You control 70% of the world's capacity of NGS. You guys are scientific innovators on that platform. Then I would also agree with all the things you mentioned earlier about the collaborative spirit and the collegial relationship we've developed. We've really enjoyed, in the last few months, our discussions with your group and the early discussions between the technical groups about, you know, getting the technologies combined and formally launched. We are extremely excited about the diagnostic potential of high-plex proteomics. The ability to use pattern recognition in this new way by measuring thousands of proteins at a time to not only characterize conditions in real time, but, you know, in almost an amazing way to predict what might happen for individuals in the future, some of which you heard from, you know, David earlier, is incredibly exciting to us and exciting to me as a former clinician. Again, we feel fortunate to be working with Illumina, in regards to your experience, in this space as you so clearly articulated. Thanks, Roy. Thanks so much for your time, Joydeep. We look forward to deepening the relationship and having other discussions like this so that the investors and analysts across the world can understand how impactful this is gonna be. Thanks, Roy, and likewise. Thanks, Joydeep. We're really excited to begin this journey with you and excited about the journey itself. Take care. All the best. Great. Take care. I'd now like to introduce Elio Riboli, joining us from the Imperial College London, to speak with Steve Williams, and for there to be this formal announcement of an important project for which we're partnering with the Imperial College. Our discussion of this partnership comes on a very fitting day. Whether or not everyone is aware of this is World Cancer Day. As a former surgical oncologist, I mark this day every year, having taken care of thousands of cancer patients myself. World Cancer Day falls on the tenth day each year, and exists to raise awareness of cancer and encourage its prevention, detection, and treatment. This effort's led by the Union for International Cancer Control and very strongly supported by the United Nations to support the primary goals of reducing illness and death caused by cancer and to end the injustice of preventable suffering from this class of diseases. I'll now turn it over to Steve and Elio. Elio, welcome. Thank you. Can you hear me? Yes. We can hear you, Elio, and welcome from me as well. Thanks for joining us. We're here to announce this Cancer Proteomic Center of Excellence at Imperial College, which revolves around you and your leadership for the past 20 years of the what was and still is the world's largest cancer surveillance study, EPIC. We're not limited just to that study, but that's the core. Take us back 20 years and, you know, you led it then and you lead it now. What were you thinking? This is before UK Biobank, this is before any of those big things, and it's complicated and transnational. You know, what were you thinking back then? What were your objectives? Well, first of all, Steve and Roy, thanks for having me here. Yes, if we go back 20, actually more years, a big question in cancer research was whether cancer basically needed a chemical or physical carcinogen or a biological agent to develop. There was a fixation, basically, the idea that it was absolutely necessary. On the other hand, we were doing more and more research on diet, nutrition, metabolic factors, I would say physical activity, and so on, that were challenging this vision that came from the lab studies in the 1950s, the 1960s, and the 1970s. EPIC, which stands for European Prospective Investigation into Cancer, was really meant to provide the data, the biological samples for investigating what I would call today metabolic carcinogenesis. The fact that we could, you know, understand that breast cancer, prostate cancer do not come from exposure to chemical carcinogens, but come from internal processes, complex interactions between genetic predisposition, lifestyle, and metabolic factors. EPIC was indeed a big challenge when I was forced to set it up. We aimed to include 350,000 participants from seven European countries to be followed up for 20 years. We ended up with half a million in 10 countries, 23 centers covering north to south, the European continent. We are now conducting the 27 years follow-up actually of EPIC as we speak. All the data are flowing to the International Agency for Research on Cancer. We will have something in the order of 90,000 incident cancer cases for which we have pre-diagnostic data and biological samples, and close to 80,000 deaths, plus we have follow-up on cardiovascular diseases and diabetes. Why did it need to be half a million people? Good question. The reason is that cancer as a cancer, it is obviously perceived as a common disease. When you break down cancer in dozens and dozens of very specific diagnoses you do need large numbers. When I say that we already have 80,000 cancer cases, it basically means that we have only a few thousand cases of each specific cancer which may be of interest, isn't it? Ovarian cancer, we have barely 2,500 cases. We need big numbers and long follow-up to get the statistical power that is required to identify risk factors and predictors of cancer occurrence. Why proteomics? Why go back into those precious samples stored in liquid nitrogen for 27 years and now do proteomics on them? Well, there are a number of reasons. One reason is that when I was setting up the study, I was very lucky to get advice from Renato Dulbecco, who got Nobel Prize for you know, discovering what happens, how DNA mechanisms regulations basically lead to cancer. So a Nobel Prize, we got a Nobel Prize for DNA research. In a talk he gave, he said, "Forget about DNA. Sorry for Illumina." Protein is the future because you need both. You need both. We know that's why we have been doing GWAS study for the past 15 years in EPIC, and this has been of tremendous scientific interest. We really needed SomaLogic to bring us to that level of power of looking at thousands and thousands of proteins with high sensitivity, high specificity all at once. Basically it is a dream that is becoming real. You've mentioned some of the features that attracted you to the SomaScan and the SomaLogic technology. What was it? Yeah, what was the key thing that really convinced you and the other PIs that, you know, these irreplaceable samples should be, part of them should be sacrificed to do proteomics? Now, what was the one overarching thing about the technology or the interactions? What really led it, you know, after all this time? Because this is a unique, you know, sort of public-private partnership, isn't it? And an industry academic partnership, which of course Imperial is used to. But you know, the involvement of the World Health Organization and all the sensitivities around that, what was it that led to a yes decision? Well, perhaps because in one word, you are unique. You know, I've had the pleasure of you know, being somewhat linked with you and your medical advisory committee for years, so I've seen you developing. It's a matter of scientific trust, personal trust, in the unique kind of innovative developments you have put together. I think that for us, working with you has now become one of the best thing that could have happened because we have done metabolic studies for the past decades. We've done omics, as Joydeep was just mentioning, multi-omics. This was, in my opinion, the missing important element. You know, how you go from genetic predisposition to what actually happens in tissues. We don't have the tissues— So let's— We have the blood. Okay. The blood is a very good marker of what happens in the tissue. The samples are already starting to be aliquoted, and we'll be running them in the next month or two, and the subset that we're running is a nice case cohort design with the most prevalent cancers highly represented. Let's imagine that we've done all that, and you've got the data, and the data scientists at the Center of Excellence are looking at it. What do you dream and think and wish will come out of it? How will this help patients or public health? Well, we hope that thanks to this unprecedented power of basically looking at 200 million protein data, the 7,000 proteins multiplied at 30,000 EPIC samples. We will have a unique opportunity to look at proteins that both predict on the short-term cancer occurrence and on the long-term. Because I should have said that we will analyze samples of study EPIC participants who eventually develop cancer from baseline up to 15 years after baseline. We will be able to look at which protein patterns may predict cancer on the short-term, which might, in the future, translate into new ways for cancer screening. At the same time, we can look at what predicts on the long term and might pave the way to better understanding its pathogenesis and processes. Characterize people who are at higher risk of developing one cancer and the other, and help us moving more into, let's say, a cancer prevention that is more targeted to subgroups with specific risk factors. Thank you. Thank you for joining us today, Elio. I think with that, our time is up, and we will move on back to Roy. Thanks, Steve. Right. Thanks so much for joining us, Elio, and we certainly look forward to with great anticipation to all the things we're gonna learn together in this project. We'll now turn it over to the operator for our second Q&A session directed at our business and financial profile, and our President, Melody Harris, will join Steve, and Shaun, and I for this session. Operator? Thank you. We will now be facilitating our second Q&A session with questions from our SomaLogic's covering analysts. We'll take a moment for the questions to be submitted. Next question is from Dan Arias at Stifel. Your line is open. Hi. Good afternoon, guys. Thanks for the presentations here, very helpful. Maybe two for Shaun, and then I'd like to come back to Roy, if I could. Shaun, on gross margins, it sounds like there's some improvement opportunities that you see. Can you expand on those? Do we need to be mindful of kit development and NGS work being impactful along the way? And is there a range that we can kind of think about once you're fully scaled up? I mean, there are some older company filings that have you getting into the seventies, but I realize that's sort of from a prior era from the company or of the company. Is there anything you can kinda help us with in just in terms of the long-term trajectory for gross margins? Well, you know, again, I do plan on trying to provide a little more detail on how, you know, to think about the near term, for 2022. I mean, you know, we've kind of already talked about, and you've heard the levers and, you know, even in your question, you kind of hit the levers, I think, that are gonna drive margins in the future. I mean, you know, we will continue to benefit from customer diversification, in the short term. I think the real, you know, levers that are gonna move margin in the future will just be, you know, our platform expansion or our into new products and diversification. To your point, that will be, you know, a multi-year journey. I do think that the margin potential, I mean, again, I wouldn't refer back to those numbers as specific five-year guidance. I certainly foresee, as we do expand those offerings in the next two to three years, that we should certainly be able to see material margin expansion going into the future. Okay. Okay, maybe another one on the long-term outlook. I mean, I know you don't have long-term targets in here, so apologies for asking when you don't have it, but this is sort of our first chance to talk to you guys about that. I mean, you describe SomaLogic as an accelerating growth story. Your outlook calls for 33%-ish growth, I think, at the midpoint of the range, and that's with a commercial scale-up that's meaningful, a better portfolio, a better pipeline. Can you just talk about, you know, sort of the growth trajectory in the outyears and the potential for acceleration to your point? Well, I believe the 33% is at the low end, not the mid, the mid-range. You know, I mean— Okay. Listen, a comment on that. That's our guidance for the year. You know, as I pointed out and, you know, I was trying to, you know, maybe more emphatically say now is that there's potential accelerants to that, and I listed those, you know, in my presentation in terms of just, right, we do have a bigger sales force, and there is a somewhat unquantifiable effect that could have, I think, kind of to the positive going into this year, where we've got a stronger pipeline, we've got a larger sales force. We're gonna continue to add that. You know, that could potentially have, you know, kind of a virtuous cycle or flywheel type effect on us, you know, in the short to mid-term. We certainly expect to see the same kind of success that we saw in 2021. There's not really meant to be some kind of dichotomy or message behind the guidance in terms of the relative growth rates. You know, I would stand by what we've always said about our growth going forward, that you know, we think what we've done in 2021, that's gonna be repeatable for the foreseeable future, you know, both through commercial expansion in the near term, you know, kit deployment, international expansion coming near the end of 2022 as well and being more material in 2023. You know, obviously opportunities like the Illumina partnership will continue to provide opportunity for us in those outer years as well. Just frankly, all the stuff Roy talked about, those starting to come to fruition as we execute in the next, you know, two, three, four years. You know, I really do think we have the potential to continue to outperform expectations in terms of growth. You know, the 2022 number is the guidance and, you know, as with every company I think in the world, we always hope that we can execute better than that. Yeah, I would just add, Dan. Yeah. Okay. I appreciate that. Dan, I would just add that, you know, pursuant to the outyears, that diversification strategy that we have to have life sciences tools and diagnostics running on multiple platforms to grow a diagnostics business on top of a life sciences tools business that we do not plan to de-emphasize in the outyears, you know, should provide us the opportunity for substantial growth and, you know, and perhaps acceleration of that growth as we move forward and all those opportunities are realized. Okay. Helpful there, Roy. Since you hit on my last one, can I just kind of throw it at you? On the clinical and diagnostic side, what should we look for in terms of evidence of progress there? Do you think there's a set of data that can emerge in cardio or some other therapeutic area that just sort of points to how your partners are pushing towards medical decision-making with the platform? Sure. Obviously, four of the six sites in the Proteomics for Precision Medicine initiative have already begun to enroll patients. These are not huge trials. These are hundreds rather than thousands of individuals. We're hoping to be able to share some early data toward the end of the year from one or more of those studies and a couple of other studies we've got keyed up as well related to that. The first thing to look for is whether or not we're able to provide data from those studies. I certainly wouldn't say that not providing data from those studies is a negative sign, but if we can provide that data before the end of the year, it should be informative for the market. The second thing is that we have this unusual situation. Most diagnostics companies when they launch, have one test. Most therapeutics companies when they launch, have one therapeutic. We have about 20 diagnostic assets now. We're gonna add maybe 10 or so over the next year. We've also thought very carefully about, you know, is it really reasonable for us to take, and then a bunch in the pipeline, 80 diagnostic tests to market over the next few years alone. We don't think that's frankly feasible. We're in this really unique, you know, and sort of first- world problem situation that we have this many assets. I think another thing to look for over the next year is, have we announced any licensing deals with potential partners around the existing assets or development deals for those partners who know that we have this capability that's unique to develop these high-plex pattern recognition tests? I think either or both of those things over the next year or two should give the market another reason to believe that this is very, very real. Okay, fantastic. Thank you guys for the time. Thank you. We have a question from Julia Qin at JP Morgan. Hi. Thanks for taking the question, and thanks for the very informative session today. Maybe just a quick one with the guidance. I was wondering how much of the volume of the guidance increase is attributed to volume ramp from the 60+ new customers that you added last year, versus how much is from new customer acquisitions that you expect to get in 2022. You know, some of the KOLs mentioned earlier that, you know, thanks to the high precision of the platform, you actually don't need to run a lot of samples in order to kind of generate discovery insights. How should we think about the implications of that in terms of, you know, utilization ramp at your existing customers? What will be the main drivers for utilization increase in the near term? Sure. I'll take that at a high level, then I'll ask Melody Harris, our President, to chime in as well as the commercial activities of the company report up through her. You know, I think it'll be a combination of things. First of all, obviously, we don't go into the year with no idea of what the opportunity is. There's a fair bit of you know contracted business going into 2022. We've grown the pipeline substantially over the last year, and Melody can speak to that. I think you know this distinction that you don't have to run as many samples to get as many insights has no impact on the number of samples that people still want to run. What that statement means is that basically underneath that statement is this reality, that proteomics are much more informative than we might have realized they were in the past. In other words, you know, for insights to be developed at times from GWAS studies, you have to have thousands or tens of thousands of patients. What we see is that hundreds or thousands of patients can drive insights using proteomics because the signal is so strong in the proteome, especially when you're measuring as many proteins as we measure. You know, that first point I made early in the presentation about what we've been waiting on is measuring enough of the canonical protein to get a signal. I don't think it has a negative impact at all. I think it actually probably encourages people to run more and more samples as a result, rather than running fewer. It's just that you don't have to run tens of thousands of samples to get the same, you know, degree of insights at times that you have to run from GWAS. We've been surprised by that, obviously, but it's again, a good surprise. Melody, you wanna talk a little bit about, you know, the drivers for, you know, for revenue growth over the next year? Sure. Thanks, Roy. Thanks, Julia, for the question. Between the balance between new customers and existing customers, we haven't reported those specific numbers yet, but we're pretty happy with the balance. An important driver for us this year will be our geographic expansion, and in particular, we've made great strides in putting feet on the ground in Europe. We're expecting good things from our European market this year. We're also expanding into Asia-Pacific and expanding our relationship with NEC that we already have in Japan. You'll see that expanding both on the life sciences side of the business as well as the clinical side of the business. We've really put a lot of effort just into inside sales opportunities, which are, of course, generating those qualified leads for us. We have had an explosion of qualified leads this year. That new customer growth that we really focused on a lot last year, we're looking to actually triple new customers this year. Maybe just as a follow-up, Steve, could you make a couple of comments, perhaps in support of my assertion that proteomics, you know, compared to other data sources, can be incredibly informative even at, you know, relatively modest numbers of patients being evaluated. Yeah, that's right. I do agree with what you said, and I think the reason it comes about is because proteins integrate signals from multiple different genetic variants and multiple different environmental influences into fewer degrees of freedom, fewer pathways. So the impact on a pathway can be measured, and it's strong. I think the concept that a high degree of precision means that people will do smaller studies, I don't agree with that because in fact, in advance, you don't know how big the proteomic change is going to be. So you don't actually sample size your study based on any statistics knowing what the change will be because you're looking for the change. People generally run studies on the basis of how many samples have they got available, how much budget they got available. They don't do it on a statistical sample- sizing basis. Then the other thing is, of course, that when you measure the proteome, when you're looking for a piece of biology, some of those effects are big, and you wouldn't need many samples to find them. Actually, some of them get more and more subtle as you work your way down the list, in order to detect those changes and make them significant, you need more samples. Finally, the other reason you want to run more samples is because people change. It's not like a GWAS study where you run them all once and you've done it. With the proteomic study, you run it once, and then you wanna see, did the intervention do anything? You actually might, within a clinical trial, want to run five, six or seven sequential sample sets. I think that the need for measuring change over time or the attraction of measuring change over time is actually a driver of increased sample numbers. Thanks, Steve. Thank you. Our next question comes from Brandon Couillard at Jefferies. Your line is open. Hey, thanks. Good afternoon and appreciate all the time today. Very informative. Melody, maybe just coming back to you in terms of the commercial build-out. Could you just elaborate on the size of the team that you have in Europe and Asia- Pacific today? Where do you see that ultimately kind of going over time? What do you view as an appropriate sort of level of scale for the organization in those markets? Sure. Thanks, Brandon. We had one European salesperson last year. We are in double digits now, and we anticipate that we will at least double that again yet this year. Europe is a very important market for us, not only for research customers, but we also know that there are a fair number of large biobanks over in Europe, and so we wanna be able to support them in country and help them. The kit strategy is also important to our European growth, and we will be expanding our kits business this year, both in the United States but also ex-U.S. And we will start to ramp that toward the end of this year. You'll start to see our kit business ramp. Got you. Maybe one for Roy on diagnostics. You've highlighted the, you know, pretty big pipeline, 100 or so tests kind of across early detection, risk prediction, disease progression. You know, which of these areas do you see as most commercially viable the next few years? Maybe updated thoughts on kind of regulatory commercial pathways, you know, whether LDT is the most likely, I guess initial preference for, you know, commercial launch. Sure. I'll start out by saying that one of the fascinating things about proteomics, for me especially as someone who used to use diagnostic tests every day in patient care, is the breadth of opportunity, you know, to approach and characterize and predict such a huge spectrum of diseases. You know, genomics has been incredibly impactful in clinical medicine, but, you know, a large percentage of the value has been in cancer because obviously cancer cells, escape the germline imperative and have their own genetic abnormalities that can be characterized and capitalized on. As Steve said, proteomics not only capture those genomic abnormalities but also capture the impact of environment, and I would include things like age, as well as, you know, sickness and medications and so forth into that environmental category. It's really an unlimited breadth of things that can be, you know, more clearly discerned and characterized by these powerful tools. That being said, obviously the two most common disease categories in the world are cardiovascular disease and oncologic diseases, with obviously in the developing world, infectious diseases close behind. Of course, over the last two years, infectious diseases have been a bit more prominent than they have been in the past. Steve and the team had the great vision, you know, five or so years ago to begin work on cardiovascular models. Certainly we've got a number of very powerful cardiovascular predictive tests now. You know, primary cardiovascular risk, secondary cardiovascular risk for diabetics and people over age 65, tests that predict outcomes with CHF, etc. All those tests, I believe, will be very important, both from the standpoint of, you know, financial and human value. Then I think the cancer tests that we're in the process of developing with Imperial, with this work with Imperial College, will also be, I think, very, very important. Just because those two are the most common class of diseases. You also heard, you know, from David Fajgenbaum today that there are a host of orphan diseases and less common diseases, you know, beyond cardiovascular and oncologic and respiratory and renal and infectious diseases, that can be characterized. We'll be working on all of those over time, both ourselves and with partners. As far as the regulatory path, we're still figuring that out. Again, I think one of the benefits of having so many assets to start is that we're not necessarily forced down one single pathway, putting all of our eggs in that basket and then being at risk, you know, for having some major failure. With this many assets, we'll have a number of opportunities. We actually believe that several years out from now, what it will look like is some tests that will stay on the LDT platform because they'll be very useful in that context for things like population risk stratification. Some tests that will be, you know, fully regulated by the FDA and fully paid for by everybody, including CMS. We have, I think, a number of options. The work this next year with the first members of the Proteomics for Precision Medicine initiative will be instructive. Not only are we working to get the trials done with these clinical partners, but we're also in discussions with them about, you know, if you're going to use these tests, which of the contexts would you like to use them in? And you know, which of these would you be comfortable using as an LDT test, perhaps for risk stratification, and which would you like to see fully reimbursed? And so that'll be part of the work this next year too. But the key thing for us is because we have so many of these assets and so many different ways they can be used, not to mention in clinical trials, right? Another set of RUO and potentially, you know, IVD uses in clinical trials, things like should this patient be in the trial? Is the drug working in the trial? I don't have to wait for a cardiovascular drug to know if a cardiovascular drug is working in a trial for MIs and strokes. I can look at the biology of patients and know in real time. Lots of options, and we're likely to pursue, you know, a multiple path forward for the regulatory, and payment structure for these tests. Great. Thank you. Thank you. We have a follow-up from Dan Brennan with Cowen. Your line is open. Great. Thanks. Thanks, guys. Maybe first question would just be on large pharma. Obviously, Novartis, very exciting. I know you've also disclosed Amgen is a big relationship. Just wondering, is SomaLogic already appropriate scale with the other large global biopharma companies, or is there a meaningful opportunity to penetrate them? Can you repeat the last part of the question? Sure. Sorry about that. Obviously, you're, you know, very successful with Novartis, and we know you disclosed Amgen as well as a large partner. Just wondering how you're scaled with other large global pharmas as we look ahead the next couple of years. Is that a meaningful growth opportunity for SomaLogic? Got it. Well, I'll answer at a high level again and then let Melody give you a more sophisticated answer. We do believe that there's a lot of scale left working with large pharma customers. Over the last year, we experienced significant growth. If you remember that bar graph slide I showed with the red and blue bars, a lot of growth both in large biopharma and also even more relative growth in academic basic research. You know, I'm not even sure that all large biopharma organizations have completely turned their attention to proteomics and more of that's gonna happen over time. And of course, we have you know, a substantial pipeline in that space as well. Melody, any more comments about the potential to grow in along with large biopharma organizations? Yeah, absolutely. We've put together a strategic accounts team that is targeting, you know, the top 20 or so large pharma companies. We do deep dives on these organizations on a regular basis and really looking at the opportunities for them along their entire spectrum. You heard a little bit about how Novartis is expanding out. We're taking those value propositions and taking them to other customers. We're seeing a lot of repeat business starting to come from those like BMS, like Novo Nordisk. There's still enormous upside for us there, but we're tackling it in a very strategic way. Our strategic accounts team is up to about 10 folks now, and just really focusing in on the deep dives for each of those organizations so that we can continue to expand those relationships and really grow that market. Great. Thanks for that. Maybe a second one just on the kitted solution on microarray. So, there's something baked in for 2022, correct? Like, have you said when that will fully launch? And then secondly, in terms of, how much will that expand the market versus just, you know, kind of shift business maybe away from your centralized location in Colorado? And then any color on the margin differential. I know Shaun made a comment in the prepared remarks. Obviously, it's probably gonna be higher margins, but any way to think about that. Yeah, we're very excited, you know, about expanding our suite of service kit solutions into the market. You know, Melody, maybe you can comment in a second on some of the specifics related to his questions about timing and impact on revenues. The one thing I'll say that it's probably evident to most of the analysts on the call, but maybe not all the investors, is that a lot of our customers, we are very aware that they use a variety of approaches. We have customers that we will be kitting the assay out over the next year or two that will still want to send some samples in for service, you know, for various reasons. We know that we have customers that are using our service arrays, and will be using our kit arrays that are also using other approaches like antibodies and mass spec for complementary studies. We believe the same thing will happen with NGS. You heard, you know, one of our key opinion leaders talk about how he's excited to use both platforms. We feel like this is a total growth and upside opportunity for us and not one that cannibalizes, you know, one thing from another related to all the various platforms and products that we would have available moving forward. Melody, any more you know specific comments about timing and impact of the kits business over the next year or two? Sure. The timing of it, we are expecting our GA to come in the back half of the year. We're working through some of the equipment that goes in the kit to make sure that we can export it, get the safety clearances and certifications that we need. We're working with friendlies to work some of those things out and get the safety testing and the things done that we need in order to get that whole kit outside of the United States. We also are being very responsible about managing supply chain and making sure that once we do launch into the GA, that we have a sustainable supply chain for our customers. That's important to us as well, so that we don't have a stop-start sort of thing once we go into the GA period. Best I can tell you on timing right now is that you will start to see the GA opportunities in the back half of this year. We do have some dollars baked into the 2022 numbers for kits, but it is not a material or meaningful amount for 2022 to give us the opportunity. Kits for us is really gonna be —the growth period is gonna be in the 2023 timeframe. You know, I would like to add. Maybe last one—sorry, Shaun, Roy. Let me just add one thing to that, Dan, that I think is important, and then I'd love to hear your next question. This is not a high-risk execution project for us. I think another company recently stated that they had the first deployed high- throughput proteomic solution in the world. We had more than 20 kits in the market before 2017. Due to a change in the business model prior to this team being on board, most of those kits were pulled back in. So we've already actually done this at scale once before. We feel very confident, you know, that once we just check the boxes and, you know, pay close attention to the details that Melody mentioned that need to be dealt with, that this will be a successful program. Thank you. Our next question comes from the line of Kyle Mikson with Canaccord Genuity. Thanks. Thanks for the follow-up. Again, thanks for the day, guys. Really good. Really good colors today. I guess, Roy, a question for you on the MESA study, this population proteomics deal. Was that a competitive process, and how long could it take to process all 15,000 samples? I guess importantly, Roy, what's the pipeline or, like, the funnel, like, for these types of projects? Great. I think actually Steve would be the best person to answer that question about MESA because he's very close to them. Steve? Sorry, I had trouble unmuting. Steve? Yes. The MESA study was really a collaboration that was led by Peter Ganz. In order to get the collaboration, the NIH support, he has. He led all of that. Strictly speaking then, it wasn't an open competitive process, although they could have chosen to go with somebody else. It was a biased selection because we had a strong advocate who's proposing that we did it. Some earlier subsets of it had been done, but on the SomaScan platform, on a smaller plex by one of the collaborators in Boston. Was that the whole question? I've forgotten what the other part was, sorry. He also asked, I mean, how long you thought it would take for us to run all 15,000 samples. About two weeks. I mean, once we get the samples in-house. It turns out that the staffing for the biobank where the samples are housed in Vermont is—that they're understaffed. Actually, shipping them to us is gonna take longer than us running them, or we run 1,000 samples a day. It won't take very long to run them at all once they're here. Kyle, just to follow up on your question about other opportunities. Yeah. We really have only focused on this over the last year. To be totally honest, we're sort of catching up in regards to getting in front of many of these large biobanks around the world. I think that the things you need to consider over time, and there are many of them, you know, there's not just one or two, but there is a large number of these around the world, are a couple of questions. One is, why wouldn't you want to have more protein data on each of those samples to be shared among all the beneficiaries of those studies, the pharma consortia and the academic consortia that surround those biobanks around the world? Why wouldn't you wanna have, you know, 10,000 proteomics data points or more on each sample? Why wouldn't you wanna have the options to run those samples on multiple different platforms, depending on what your preference was? We're really only turning our attention when we've really only done this over the last year or so. We began to turn our attention to these large population-based study opportunities. I would keep those two questions in mind as we do so more vigorously over the next few years. Okay. That was great. If I could just ask a follow-up, more broadly. When you think about your financial model, the long-term model, does that currently assume, you know, offering aptamers to third parties for any, antibody applications like single cell and spatial? And, you know, those are obviously rapidly growing markets. It'd be great to kind of participate in those. Where would you sit in the workflow and kind of the value chain given the incumbents? Sure. I'll give a high-level answer and then ask Shaun to comment, and then we should probably wrap up because we're at the top of the hour. We have not modeled those revenues into our financial projections, having only made the decision end of this year to pursue this as a business opportunity, but we do believe the opportunity is substantial. Basically everywhere an antibody works, so in platforms like ours, and others for high- throughput, you know, high-plex proteomics, in single-cell proteomics, in spatial and in situ proteomics that use antibodies in various ways, you know, to signal off of those ligands in those platforms. Even in things like immunohistochemistry in the laboratory, sort of a more commoditized or certainly more time-honored use of those constructs. Every single one of those places, our aptamers could work. There are some potential benefits of using our constructs. Even though the binding sites for aptamers are very similar in size to those of antibodies, they're 10x smaller, you know, in regard to the total footprint that an aptamer has on one of those platforms. The ability to zoom in on spatial biology is potentially even stronger. As far as the approach here, we're going to cast a relatively broad net. We'll be talking to companies in all of these cases. We may be talking to other proteomics platforms themselves that use antibodies in those systems and even channel partners that, you know, distribute those reagents around the world. We do believe this is a substantial opportunity. Don't forget the fact that currently we can see about 4,000 proteins that antibodies at scale can't see. Obviously, there's a benefit to over and above what's currently done and not just something that's purely additive. It's really synergistic. Any comments, Shaun, about the modeling in regards to the reagents business? No, this is just exactly, you know, an example of what I was talking about as, you know, potential accelerants, you know, to our model, as well as quite frankly, again, this is the type of business that would be accretive to our margin. You know, it's exactly what we're excited about. Yeah. Thanks, guys. You bet. Thank you. Great. Well, I think it's time to wrap up. Great. Thank you, operator. Thanks to the analysts and investors. Thanks for your questions. Again, thanks to our key opinion leaders for participating today. Joydeep, much appreciated. Operator, thank you for your help, even though I once again interrupted you here at the end. Sincere thanks to all of you for joining us today. More importantly, thank you as well for your support and interest. We couldn't do what we do without our analysts and investors, you know, full stop. That's not a complicated calculus. We do thank you for your support and interest as we work together to leverage the power of proteomics to relieve human suffering and to prolong meaningful human life. Thanks so much.