A few brief logistics before we start. Participants and members of the executive team in today's event are Peter Blume-Jensen, Chief Executive Officer, President and Founder of Acrivon Therapeutics; Kristina Masson, Co-Founder and Executive Vice President, Business Operations at Acrivon Therapeutics and President and CEO of the company's research subsidiary, Acrivon AB; Eric Devroe, Chief Operating Officer; and Rasmus Holm-Jorgensen, Chief Financial Officer. Peter will introduce the agenda shortly. Anytime during the presentation, viewers are welcome to submit questions in the window underneath the viewing screen for the Q&A session at the end. Finally, today we will be making forward-looking statements, and you are advised to review the forward-looking statement disclosures on our slide deck and relevant regulatory filings filed periodically and in company presentations. Today's slides will be posted after the event concludes. With that, I am pleased to turn the event back over to Peter. Thank you so much, Adam. Adam Levy, I don't know if you heard it, he is heading our Investor Relations and Corporate Affairs. We have the agenda here. It is a tight agenda today. We are super excited to present our initial clinical data. We'll have a brief Acrivon overview the first 5 minutes or so, and then followed by a main clinical session, followed by a session discussing our very exciting single agent active potential, first-in-class dual WEE1/PKMYT1 inhibitor, and the early pipeline. We'll also have a preview of our AP3 Interactome, our interactive actionable machine learning-enabled database, where we house all our quantitative proprietary data. We'll follow that with that live Q&A session at the very end. The first slide here is just to remind you what our critical challenges are in the biopharma industry that I think pretty much every company is facing. One is to discover in a streamlined manner rationally designed potent compounds that could aim for monotherapy that offers a lot of advantages, simple clinical development, etc. And our proteomics-based platform called Acrivon Predictive Precision Proteomics, or AP3 in short, is really able to, at very high resolution, uncover drug-induced signal transduction pathway effects that are actionable, and one can tune in and use it for biological SAR. We'll discuss that further. That's how we have developed our first fully internally developed molecule, ACR-2316, the dual WEE1/PKMYT1 inhibitor. Determining which patients benefit from therapy is obviously the so-called Holy Grail in precision medicine. And we developed drug-tailored OncoSignature tests. That's one of the many deliverables of our AP3 platform. And we use that too, at an individual patient basis, to identify patients that are sensitive to monotherapy to that particular agent. Finally, a low-hanging fruit for our platform and our approach is to uncover resistance mechanisms. We have a number of publications on that. It's very easy to see that every single time one treats a cell, any kind of cell with a compound or any kind of drug or modality, one will see all the drug-induced resistance mechanisms. And one can, because it's so high resolution and all the nodes are interacting, one can identify druggable nodes on those, and that opens up for rational drug combinations or for generating drugs, like we have in our case, with superior single agent activity. Systems data like this require machine learning and require really the activity states of the disease-driving pathways inside a cell with intact protein-protein complexes and subcellular localizations. One cannot deduce from genetics or transcriptomics all the drug-regulated protein nodes that are really the manifestation of disease. And we can directly visualize that, quantify that, and monitor all the drug-induced effects. And because we do that at high resolution, we can do it in normal tissue versus disease tissue, one can identify drivers of disease and also deduce exactly what a drug does on these disease-driving mechanisms and hence optimize a compound for optimal selectivity to achieve that superior single agent activity. And we have achieved that, we think, with 2316. Acrivon, the meaning of Acrivon is exact or accurate to symbolize that exact or accurate match between what drives disease and the drug's action. We're currently focused on within oncology, but in the not too distant future, you can imagine that we might move outside to autoimmune inflammatory and fibrotic disorders and others, where there are no genetic alterations in the disease tissue to really guide you. So the platform itself will describe that more, but it's broadly applicable, and we use it in early drug discovery as well as we'll hopefully convince you also during clinical development. Here's a snapshot at our pipeline. What you see here highlighted with the red outline is what we are going to discuss today. We are going to show you initial positive clinical data from platinum-resistant ovarian cancer and endometrial cancer in our master guidance protocol that we are showing here. Just a reminder that we have about one-third or so of patients of these three indications are OncoSignature positive, and the two-thirds or so are OncoSignature negative and hence predicted insensitive to monotherapy. But we have uncovered there very relevantly a resistance mechanism to further sensitize to that combination. The registrational intent trial are these longer arms here, and that's what we are also going to cover today, the monotherapy trials in platinum-resistant ovarian cancer and endometrial cancer. Furthermore, you see here in red down here, we have very accelerated timelines now for our development candidate, ACR-2316. Our previous guidance was IND filing Q4. We are now aiming for early Q3, followed by an anticipated clinical trial initiation in Q4, 2024. The final slide in the introduction here is just an outline of how to think about Acrivon, and it's actually a very important slide, and I apologize for all the text. But we are having a purpose, which is really we are executing on a multi-year vision, and our goal is really to transform precision medicine, go beyond the limitations of genetics and transcriptomics. And it's very important that there is a team that has worked together, a core team for now 14 years, just focused on this day in and day out. And this team has pioneered both the two underlying technology pillars: mass spectrometry-based phosphoproteomics, in particular our academic co-founder, who is not participating today, Jesper Olsen, who runs the Novo Nordisk Foundation Protein Center, and the quantitative multiplex automated biomarker platform. So it is an expert system we are executing on here with an expert team that is just focused on this. As it relates to our AP3-based patient responder identification, very importantly, we are going to show you prospective data today with the locked signature. And remember that ACR-368, which is also called prexasertib, is an asset in-licensed from Eli Lilly that was a flagship program because it showed durable single agent activity across all solid cancers, and genomics was insufficient for patient selection. So we in-licensed that asset with three clear key objectives. One was in the proven platinum-resistant ovarian cancer type to increase the overall response rate above the background unenriched response rate of about 12%. Published in the JTJN multi-center trial, they had run a 169-patient trial in 46 centers, 8 countries. We also, very importantly, wanted to identify and verify robust clinical activity new indications. We can do that with what we call indication screening using our OncoSignature prior to initiation of a clinical trial. We can screen across just commercially available intended-use processed human cancer tissues and identify sensitive indications. We did so, and that's how we identified endometrial and bladder as two additional predicted high unmet need for treatment solid tumor indications predicted sensitive to ACR-368. Of course, with that, we validated our approach with the purpose of getting a read-through to other drugs, whether internally developed or externally acquired. We have initial data, positive data demonstrating proof of concept for these objectives. We apply also AP3 for drug discovery, and we are going to show you that. We have generated this potent single agent active WEE1/PKMYT1 development candidate, specifically designed for clinical monotherapy development and aiming for the accelerated pathway. We are generating an OncoSignature for that as well, primarily to do this indication finding there prior to or earlier in the clinical development. It was a very streamlined development of that molecule, and that's why we have moved our timelines up now to filing expected in the first half, first part of Q3. We also have a new program, undisclosed target, a cell cycle regulatory program where we follow the same AP3 approach, and then we have the preview of our AP3 Interactome. The clinical data I'll show you now, which we are very excited about, are initial positive data from the ACR-368 prospective patient selection phase II trial. Very importantly, this is with the locked OncoSignature test with registrational intent, and it's with the data cut as of April 1. A reminder before I hop into that: we have our mass spectrometry-based phosphoproteomics step, the machine that enables uncovering of resistance mechanisms, identification of rational drug combinations, identification of thousands of pharmacodynamically regulated biomarkers and biomarker candidates, and drug mechanism of action in diseased and normal tissues. One of the deliverables, just one of the deliverables of this platform, are these drug-tailored OncoSignature. We use our machine learning-enabled scripted steps to generate these through a functional definition of three functionally orthogonal classes of biomarkers that really functionally measure the tumor's dependency on the activated drug target signaling axis that the drug acts on. If the tumor depends on what the drug acts on, it's very simple: we predict sensitivity. This is obviously agnostic to underlying genetic alterations, agnostic to prior therapies, agnostic to where the tissue is taken from. It's simply in the particular tissue we are measuring. We are measuring the biomarkers here in red. This is actual high-grade serous cancer, ovarian cancer biopsies. We can see where the biomarkers are, in this case in the nuclei, it's intact tissue, and all three are necessary, have to be elevated at a minimum level to predict sensitivity to the drug. They are very quantitative assays. Here you see examples of just biopsy samples where you have, again, an example of patients that is a predicted responder with a very high upregulation of the biomarker and one that's a predicted non-responder. And you can see biomarker 1, 2, and 3, and this is how it works across different patient samples. Just a quick snapshot or look at what the therapeutic bar is in these two gynecological cancers. Both are extremely high unmet need, as you know, platinum-resistant ovarian cancer. Once you go beyond second line, you are looking at an overall response rate of about 12%. There are a couple of trials there that were actually investor-initiated trials, the AURELIA trial, the CORAIL trial. We are looking at about a 12% response rate here with a duration of response from everything from 3.5-5.7 months or so. We have the recent approval of Elahere or mirvetuximab, which is approved post first to third line. It is approved for folate receptor positive platinum-resistant ovarian cancer. The VENTANA test is developed, so it really is a folate receptor high, and different publications would estimate that this fraction of patients is about anything from maybe 35%-45% of patients. The folate receptor positivity is defined as more than 75% of cells being positive with moderate or high intensity of staining, and about, call it, 35% or so benefit. So ultimately, there is still a huge fraction of patients, whether it's 80% or 85%, doesn't really matter, that do not benefit from Elahere, which is an ADC. In endometrial cancer, we really have just after carboplatin, pembrolizumab, and lenvatinib as a second line that is the best option here. Once one needs to move to third line, the standard of care is down to just about 10% with a duration of response just a little over 3 months. So both huge unmet need. We are, of course, extremely excited about endometrial cancer, in particular also where actually there's less happening. In the past trials in ovarian cancer, it was found that without patient selection, as I mentioned, we have about 12% unenriched response rate with a duration of response starting at 5.6 months, and there are other studies that have shown them up towards 10 months or higher in this particular indication. We have the duration of response relatively well established from over 200 patients treated here. I think a reasonable soft target product profile is to say if we can separate with a lower bound of the confidence interval about 15% in these tumors, we are in very good shape. Here is, again, the status of this particular trial. This is with the logged OncoSignature thresholds. It is prospectively conducted. We did build into the clinical trial for good ethical purposes the ability to adjust the thresholds in our OncoSignature in up to the first 12 OncoSignature positive patients, and we incorporated that in the early part of the second half of last year. So the data we are now showing you are entirely prospective with this locked threshold and the locked biomarker signature. You can see as of April 1, we have a total of 92 patients enrolled, and those that received at least one dose were 26 and 41. So these are the intent to treat, so 67 patients. What we have scans on with at least one 8-week scan, meaning they are efficacy evaluable, are 10 OncoSignature positive patients and 16 OncoSignature negative. A total of 26 patients, which is those that are efficacy evaluable. And very importantly, in these patients with this OncoSignature, we are seeing exactly what we predicted also in the tumor types that are not previously ever evaluated with ACR-368 endometrial and bladder. We are seeing an OncoSignature positivity rate of just about the 35% that we predicted on just commercially available samples in the past. Here is a waterfall plot of the OncoSignature positive patients that you can see here. We have these are confirmed responses. There's one here with an asterisk where the data had been entered, where the data had not been entered, but the imaging had been done. So we have updated this one day after the cutoff, that one unconfirmed PR was confirmed, and that brings the total confirmed to 5 out of these 10 patients. So all these are confirmed responses, and you can see this is 3 out of the 5 endometrial and 2 out of the 5 ovarian patients that have this confirmed partial response. So that is summarized here, the 10 patients, the 50% response rate. It is possible to calculate an aggressive confidence interval on already this small number of patients. As you are aware, the higher the response rate is, the more you start seeing the fewer patients you need to get the same statistical confidence about the response rate. So so far, it's looking very promising, these 5 confirmed responders as of data cutoff on treatment, and of course, median duration of response has not been reached yet. And as I mentioned, for ovarian cancer, we know that pretty well from a vast number of patients treated in the past unenriched. So this is where we are with that. Very, very probably the key slide that really starts validating the prospective OncoSignature patient selection is this one. This is incorporating all 26 patients treated with the OncoSignature at recommended phase II dose to date. What we are doing here is not a clinical protocol's pre-specified analysis, but it's one that is essential to validate whether the OncoSignature is able to do what it's developed to do: segregate the responders from the non-responders. If you look here, we have the biomarker positive arm here with ACR-368 monotherapy, and you can see all the responders are here. There are, at this point, no PR or CR confirmed responders in the biomarker negative arm. I will show you that we do have activity in these. The ultra-low dose gemcitabine does actually have increased the sensitivity to ACR-368, which is absolutely dead without a low dose gemcitabine, but we still don't have responders, which allows us to use this as a negative control with a bias against ourselves. And even despite that, we are able to calculate a p-value, and as you can see by various methods, non-parametric bootstrap, Fisher, other tests that give very similar results, we have a third decimal confirmed p-value for these responses. So we are very excited about that, which is obviously what this test is developed to do, to segregate responders from non-responders. I want to remind you also, very importantly, that endometrial cancer with the confirmed responses here is an entirely AP3 predicted tumor type, not literature predicted, not tested previously. This is one of the value propositions of this approach, is that we can screen a vast number of tumor types and hundreds of samples in each just obtained through commercial sources prior to initiation of clinical trials. This data here is actually something we showed Lilly back in early 2000 when we started discussing with them the in-licensing of this asset and brought it over the finish line. We showed them why we don't think they should try to treat squamous non-small cell lung cancer. At the time, there was not a single predicted responder in this tumor type, and we showed them that we were predicting tumor types like endometrial and bladder, and we are showing you endometrial cancer here with that predicted ballpark 35% of positivity. In absence, this is how such human tumor samples look. You can see there's a massive addiction to the Chk1/Chk2 activated DNA repair axis here with premature DNA replication in these tumors. That's how these three biomarkers work. In absence of a clinical trial, we have followed up with genetically non-modified patient-derived xenograft models, both for both endometrial and also not shown here bladder, and we are just showing you a small subset of the endometrial tumor samples here. And just as we predicted, a subset of them are completely sensitive to the drug and others less so, and we can predict which ones are sensitive. So this tumor type is just completely predictable with the OncoSignature. That's also why we are developing an OncoSignature for ACR-2316, our novel dual WEE1/PKMYT1 inhibitor, to be able to do that. We are showing you here the spider plots for the gynecological cancer patients, and you can see if anything, there's a tendency to even deeper responses in endometrial. The patient that now has the confirmed fifth response is, I think it's this patient here that's now out on the third scan. But at this point, obviously, we are just showing you the data cutoff here, and as I said, duration of response is maturing. In the OncoSignature negative arm, remember, this is an arm where the OncoSignature is predicting no sensitivity to monotherapy, but where we use the mass spectrometry-based phosphoproteomic approach to uncover the dominant resistance mechanism and could see that could be overcome with small amounts of nucleotide synthesis inhibitors, which we then tested and demonstrated that ultra-low dose of gemcitabine, one such, could resensitize. We had a poster at the recent AACR meeting where we showed and demonstrated that in rodent models and in preclinical models across the board, we see an upregulation of the OncoSignature in OncoSignature negative patients upon ultra-low dose of gemcitabine induction. And so we see that there is a mechanistic rationale completely consistent with the way these OncoSignature tests work. The patients we are enrolling are all- the requirement is that they have aggressive progressive disease. So stable disease restaging is actually a meaningful activity, and you can see we have some activity in these patients. We are showing here that indeed a good fraction of them are just kind of borderline at a resistant response. So there's clearly clinical activity, but we have just entered the recommended phase II dose, as we reported in our latest filing for ultra-low dose of gemcitabine, which is 10 milligrams per square meter, so 1% or so of standard dosage. And so we are continuing this trial and letting that mature, but this is the status of that right now. The treatment-related adverse events, finally, are exactly as we have seen in previous trials, except if anything, maybe better in the sense that here we have the monotherapy arm, OncoSignature-positive arm that receives ACR-368 at recommended phase II dose. And you can see that the encouragement of usage of G-CSF, a way to address mechanism-based neutropenia, is paying off. We, in the past trials reported by Lilly, they have seen about just 38%-40% of grade 3 or higher transient reversible neutropenia primarily, but we have it really well under control. There's an anemia rate that's a little higher here. 17% is still this maturing dataset with a data cutoff of 8 March, which is where we did the latest safety data cutoff. And so predominantly mechanism-based heme as seen and completely consistent with previous monotherapy trials. The message with ultra-low dose of gem is very, very similar, slightly higher, but still very acceptable profile and, again, primarily reversible and manageable. We have very, very limited non-heme AEs that are those that tend to be irreversible, where even grade 2 AEs can sometimes be very interfering very much with the quality of life of a patient. We did negotiate to receive a lot of drug product from Eli Lilly, but of course we have put our own manufacturing initiative that, as you can see, is well underway, and we have both active pharmaceutical ingredients produced in a very big amount as well as successful registration campaigns completed with a very good yield so far. So we have drug product manufacturing well under control as well. So that concludes the clinical part, and we will now slowly turn our attention over to the preclinical program, which is on our AP3-based drug design of the first fully internally developed, potential first-in-class molecule, highly selective ACR-2316. And I'll give the word to Kristina Masson in a second, but basically want to remind everyone that the way we use the AP3 platform for what we call streamlined biological SAR is shown here. Every single experiment we do is with a throughput of about 12 days. We generate a massive amount of proprietary quantitative phosphoproteomic data. We profile compounds, lead candidates across classes of compounds very, very rapidly. We do it in 48-well plates and at very high resolution. We have structured analysis. We always look at heat maps of quantitative heat maps of pharmacodynamically regulated, drug-regulated biomarkers. We do principal component analysis. We do volcano plots. Then the most important for us is to assess the drug-regulated activity state alterations in the intact cell signaling network, as well as directly look and visualize through omics visualization and other methods the exact phosphorylation sites and up and down-regulation of these, followed by functional annotation in the signaling network. We can also functionally annotate our proprietary data against publicly available data. This is a proprietary pipeline with a lot of scripted algorithmic steps, and it is generative AI enabled. We truly have proprietary data that require generative AI. The data are tight. As I mentioned, I will not go through the details of this, but it's a very, very systematic approach, again, all scripted. You can see here is a sample correlation matrix. This is really, really tight data. The team is just completely specialized in doing these types of analyses, and as I mentioned, we have a very high throughput. So you can imagine with about 65 million quantitative data points per 12 days or so for a couple of years now, we have terabytes of data in the cloud accessible through our AP3 Interactome. It's modular. We can scale it. We can do many compounds, and it's broadly applicable. It doesn't matter what the modality of the compound is. Of course, it doesn't limit one to oncology. It's very exciting. Most importantly, we see every single time we treat a cell of any kind, diseased or normal, with any kind of compound or ADC or anything, one sees drug-induced resistance mechanism. Actionable drug-induced resistance mechanism is always taking place without exception, regardless mechanism of action. That is the basis for how we can optimize drugs towards superior single-agent activity. On that basis, we set out specifically to overcome the limitations of current single-agent active WEE1 and at least partially single-agent active PKMYT1 inhibitors. There are three things specifically we focused on. One is to achieve superior single-agent activity. By looking at the minus-plus drug-induced effects within the cell signaling phosphoproteome, we could see the drug-induced resistance mechanisms induced by a potent WEE1 inhibition and could see that a majority of those emanate from PKMYT1 and build into that same molecule through our co-crystallography insights and rational drug design, the ability to quench them in a balanced manner. So by quenching the dominant WEE1 inhibitor-induced resistance mechanisms, we achieve that superior single-agent activity. And that's a general approach we can use. We call it reciprocal quenching. I'll show you a slide on that later. We also wanted high selectivity to limit the AEs to just the most on-target-driven transient AEs. We know for all DDR inhibitors that's the case, that the target organs are where we have dividing cells like the bone marrow precursor cells and myeloid progenitor cells and other dividing cells like villus crypt epithelial cells in the gut, etc. We have achieved that, and we have completed GLP tox studies, obviously, and it does seem to show that and confirm, as we have in our corporate deck, that reversible, actually fully recoverable on-target heme AE and recoverable between each cycle of our chosen dosing regimen so far in our dog and rat studies. As I mentioned, we are developing an OncoSignature to do indication finding. We think we don't need patient selection with this broadly active agent based on what we know about the existing V1 and PKMTY1 inhibitors, but we develop it in case we need it and also to do this indication finding prior to entry of the clinical trial. So with that, I'll hand the word over to Kristina Masson, who runs our subsidiary. All our biomarker identification and drug discovery is out of our subsidiary in Lund, Sweden, right next door to the world's brightest synchrotron called MAX IV. Kristina? Thank you, Peter. Happy to take over here. So we have, in less than 17 months, managed to generate more than 40 high-resolution co-crystals of various leads across multiple selective novel series. And we achieved a very potent molecule with superior antitumor efficacy and complete tumor regression across models. This is really enabled by potent V1 inhibition and balanced PKMTY1 inhibition to overcome resistance. On the next slide. Thank you. This is just a quick summary of our development candidate, ACR-2316. It's a highly active molecule and about a log scale more potent than a clinically relevant comparator, as shown here in this table. It has low single-digit nanomolar potency against V1, and we also have a potency but balanced inhibition of PKMTY1. Of course, not as much as a dedicated PKMTY1 inhibitor, lunresertib, below here. But it's actually very critical that the inhibition of PKMTY1 is not more potent than V1. It's through this balanced ratio that we achieve the synergy, yet keeping an exquisite selectivity of the molecule. And this, most importantly, translates to superior antitumor efficacy, both in rodent models and in the cellular preclinical studies that we've conducted to date. What you see here in the top middle figure is some of our in-house generated molecules. Most of them are more selective than any of the leading benchmark comparators. The y-axis here is a selectivity score, so the fraction of kinases that are inhibited more than 65% at 1 micromolar. So the lower the score, basically, the more selective the compound. Each compound is also mapped for WEE1 and PKMYT1 target engagement, and this is visualized by the circles. The inner is PKMYT1 and the outer is WEE1. And the orange color indicates maximal target engagement. So as you can see, 2316 is not the most selective to date, but it was the one with the best properties overall and is more selective than any of the comparators. At the bottom middle panel here, you can see the effects on the cell cycle. WEE1 and PKMYT1 inhibition activate CDK1, which releases massive mitotic signaling, which translates into S/G2/M block. Now, the cells can't repair themselves, so they stall here and undergo cell death ultimately. You can probably appreciate the effect of ACR-2316 here. One of the very unique things with our AP3 platform is that we can measure all drug-regulated activity states of signaling pathways. So we bring in activity as a novel orthogonal dimension, which we obviously cannot assess using genomics or transcriptomics or even panproteomics. What we're looking at here is 2316 versus a previous candidate that was very similar as measured by traditional drug discovery. But as you can see in the heat map on the left here, the profiles are very different when we use our high-resolution mass spectrometer method. As shown in these clusters, we get a massive on-target activation of CDK1 and 2. So red here is up and blue is down, and the size of the circle indicates the magnitude of the activation or inhibition. Also importantly, we see a very significant activation of PLK1, which is essential for driving cells into this mitotic catastrophe. And when we try to mimic the same ratio of combining V1 and PKMTY1 inhibition, we don't see the exact same phenotype as 2316. In this case, we get less CDK1 and 2 activation, but we still see PLK activation. So all our mass spec experiments are very high resolution and machine learning enabled. So basically, by a click of a button, we can go in and see all substrates for specific kinases and how the drug regulates those. A testament to the massive activation of CDK1 in response to 2316 is shown here. So on the y-axis, we are looking at over 300 statistically significantly upregulated CDK1 consensus substrates. This means that we have an unbiased ability to quantify sites, and most importantly, these are actionable PD markers and potential biomarker candidates. Also, this is an unbiased way to directly get full insight into the MOA of the molecule. But most importantly, this translates. So here what we see is the potency of 2316, and this is very superior, both across human tumor cell lines to the left and PDX ex- vivo tumor models to the right. So the left panel, important to point out, is genetically non-selected cell lines, so truly a random panel of cell lines of different indications. And what is noticeable is that 2316 is a log scale more potent in these molecules compared to comparators. Again, in this slide here, this is a testament to the very potent cell death inducing mechanism. And this is a cell tech tox assay where the green dye will penetrate membranes of dying cells. So you can see that 2316 is much more potent both in a dose-dependent manner to the left and also in a time-dependent manner shown to the right in these figures here. On the next slide, we also confirm that this pro-apoptotic mechanism is caspase 3-7 cleavage, again, strongly induced by 2316 and superior to comparators. So all in all, we have done a number of studies where we have data and science we're very data and science-driven. And without exception, across all these different studies and models, we have superior potency compared to benchmark molecules with 2316. And of course, most importantly, this translates to antitumor efficacy. We've done a number of tumor-bearing mouse model studies with different dosing regimens, and we're just showing you a snapshot here. But the key message is that 2316 causes complete tumor regression at doses much lower than any other clinical comparators. We see complete regression, but also in some models, a nice dose response here in the top left. You can also see at the bottom right model that multiple different dose levels cause complete regression, which speaks to the therapeutic index expansion that we are achieving with this molecule. We can also come in and establish tumors in the top right model here, causing complete regression with a little more than half the dose and a 5-on-2-off regimen versus the highest tolerated daily dose of adavosertib here. And at the bottom left here is another example of complete cures, really. In this model, we stopped dosing after 24 and 26 days, and we achieved a complete cure in this 3-on-4-off dosing arm of mice. So the key message is that regardless of the dosing regimen, we see very similar robust antitumor efficacy. You will find more details in our corporate deck, but we have completed the GLP tox studies, as Peter mentioned earlier. So with that, I'd like to end on that we have this rationally designed molecule that has a very superior profile. And most importantly now, with this monotherapy activity, we are aiming for streamlined clinical development. We have accelerated our timelines, and we are advancing towards IND and IND filing with an estimated filing date in Q3. And we are anticipating first in human in Q4. As mentioned, we are developing an OncoSignature test, not because we necessarily think it's needed for patient responder identification, but to do indication screener prior to our clinical trials. Also, Biomarker 2 is always a very carefully chosen PD marker that we can use, and we intend to do so, to be aligned with Project Optimus and achieve dose optimization by looking at drug-target engagement rather than MTD in our clinical studies. And with that, I'll give the word back to Peter, and please feel free to add anything that I might have missed, Peter. Thank you, Kristina. I know this was excellent. So maybe it's worth emphasizing now with our, if you want to call it initial validation of both the OncoSignature patient selection method, the AP3 indication finding, and so forth that has successfully identified sensitive indications for ACR-368, you can hopefully appreciate the importance of doing this prior to clinical development. We actually expect this very broadly active potent agent to be active in a number of either two untested tumor types, and hopefully we'll be able to share more with you in the not too distant future on that. But solid tumor indications with high need seem to be primed for this particular active agent. And of course, that is on top of the well-known sensitive endometrial and cancer tumor types as well. We also can, very importantly, we can use the second biomarker in our three biomarker OncoSignature tests are always very carefully chosen and optimized pharmacodynamic biomarkers with a dynamic range, intensity, and specificity in intended use process human cancer tissue that really allow us to assess drug-target engagement. So aligned with Project Optimus, we really have an opportunity here to optimize drug-target engagement based on what we know how translates from cellular models, the cellular drug-target engagement Kristina showed earlier, through viability assays and toxicity and caspase assays to animal models with allometric scaling to humans, and really avoid to just go by MTD, but also bring in based on strong target rationale, drug-target engagement in these studies. The next couple of slides, I want to kind of just show a little bit. Oops, what happened? Try again. It says, I hope people see the slide? Yeah, it works, Peter. Okay. I'm not sure what that clip was. So I just want to remind everyone that we have this universal method to identify the dominant resistance mechanism for any compound, any modality. And an example of that is shown here. It is also in our corporate deck in a slightly different version. But basically, for all our proprietary data sets, which are interactive in the cloud, accessible for all scientists, where we can pull out data in a second, pushing at buttons and say we want to look at pathway interaction signaling networks or heat maps or substrate consensus motifs, etc. We can also, for each experiment with proprietary data, functionally triage them against publicly available data sets. And when we do that, we can, for any compound, ask what functional pathways and networks that are regulated predominantly by a drug. We can zoom in on what is, for example, up or down regulated within protein sets. Each circle here is a protein set. The larger the circle, the more proteins in that interconnected protein set. This is just from publicly available data, but based on triaging our own proprietary, our own compound-regulated data into that. That way we can make sense out of this. As an example of how we can deduce the dominant resistance mechanism, we can go through this real quick. Here you see, in a sensitive and resistant human tumor setting, an example of—in red indicates up and blue down—what are all the V1 inhibitor up-regulated phosphosites and phosphopeptides and nodes in the signaling networks. So when you zoom in on these, you'll see a particular enrichment in the area of DHR, DDR, and mitotic regulation, and not so much any other places. Very interesting and very relevant for the mechanism of action of V1 and makes sense. The most significant regulation happens in the biologically relevant processes, and that's where the cell kicks in ways to counteract the insult of a drug. That's at least a hand-waving interpretation of that. By looking at the actual amino acid sequences, the primary amino acid sequences, and the phosphorylation consensus motifs, one can deduce what their dominant mechanisms are. And it was very clear through this unbiased approach that one can see that many of these originate as PKMYT1 driven. We could then verify that by simply conduct a control experiment and show that a PKMYT1 inhibitor on top of these would down-regulate many, many of these nodes, this inverse diagonal pattern here. When we extract these oppositely regulated, hence the word reciprocal quenching, oppositely regulated putative resistance mechanisms to a dominant WEE1 inhibition, it's this small subset indicated by this orange bar here of all these mapped quantified phosphopeptides in this particular experiment, where you can see that we have up-regulation, we blow it up here, up-regulation by WEE1 inhibition, and then counter down-regulation by PKMYT1 inhibition. The proof of all this being really relevantly driven by PKMYT1 is here. This is the only known dedicated direct consensus substrate motif for PKMYT1, threonine-14, one of the two inhibitory phosphorylation sites in CDK1 that is specifically this motif here. We're also sharing another one, but not the others because we think they are actionable. But there's an interesting one here which we are willing to share, which is an autophosphorylation site in Chk1 that brings that places Chk1 downstream of PKMYT1. When we inhibit with PKMYT1, this goes down. And that means that this Chk autophosphorylation site is somehow regulated by PKMYT1. This means there's a significant synergy between the WEE1 dominant potent inhibition with a balanced PKMYT1 inhibition. And we see with BRAID and other types of Bliss and other synergy scores, we see a massive score of approximately 45 or so across many, many cell lines, much more than, for example, with, as we can easily see in our networks, a combination of our Chk1/2 inhibitor with WEE1, which also creates synergy and is well-known as well. And you can see there's a Pearson correlation. The less sensitive that a cell is to WEE1 inhibition, the more synergy there is with the WEE1/PKMYT1 inhibition. A preview of our interactomes here. We are going to share that hopefully at a future event. It's all structured data. It's the same type of Google Venture-friendly, if you want, data that we are generating, storing in the cloud. And these are proprietary data, actionable across all experiments. We can look at all this that we have talked about. Here's an example of consensus motifs. What is it that's phosphorylated at what frequency? What is activated in red or inhibited in blue? And what are the substrates as rectangles versus the circles that are kinases? We can zoom over them and see the PTM-SEA scores, etc. All this is interactive and directly telling you what drives the disease at a functional level. Obviously, something that is simply not possible with transcriptomics, genomics, or even panproteomics. We're measuring the allosterically altered activity level of enzymes and proteins through this approach. We look forward to sharing more in this in the future. This is what we call the AP3 Interactome. Finally, we have a new undisclosed target in our pipeline. It's a very exciting, we think, cell cycle regulatory target that kind of has been overlooked and is perfect for a proteomics-based approach. There are different isoforms that are very similar of this target. There are proteolytic splice variants. There are, sorry, proteolytic products of this. There are splice variants, etc. It's an exciting cancer drug target and really with, so far as we can see, with minimal competition. We have done all kinds of public analysis and so forth, and it certainly shows essential, it seems to be essential, for cancer cell viability and also proliferation. Hence, this data alone provides for very strong mechanistic target rationale for an important role in oncogenesis. We also know it based on the family this molecule belongs to. Very importantly, by serendipity, a tool compound was developed against another target that happens to be a very, very selective inhibitor of this particular target. And of course, we've already completed phosphoproteomic profiling of this tool compound in both sensitive and resistant settings in the same experiment. We call it differential global phosphoproteomic profiling. So we already understand the role of this molecule. And very importantly, this tool compound, while not having biodistribution PK and biodistribution properties amenable for oral or even injection delivery, it can be delivered as a liposomal formulation. That and other experiments, including genetic knockdown experiments and other, has really shown in preclinical models that there seems to be an essential role for this target in tumorigenesis. You can see here basically complete regression in these particular tumor-bearing models. Based on this, we have now launched this new clinical program, which will pursue exactly the same way as we've done with our previous programs through rational drug design, leveraging the MAX IV synchrotron right next door in Lund, as well as using the AP3 profiling for what we call biological structure-activity relationship, SAR. The chemists can, instead of looking at chemical structure only and thereby create selective compounds, you can actually achieve an optimal target selectivity profile and try to leverage the co-crystals to dial that in rationally in the molecule through medicinal rational chemistry. And so we expect to have a development candidate for that next year. And with that, and there you can see how well this fits into the pocket here. And with that, just mentioning that we obviously recently achieved to raise a PIPE financing. And you can see here we have now with that added flexibility, added runway, we have a lot of flexibility in our execution. Our history is capital efficiency, and we are still operating at an extremely capital-efficient level. And so you have our pro forma financial highlights here. Rasmus, anything to add? You covered it very well. The team is very capital-efficient, and the AP3 platform is very capital-efficient. So we have a lot of flexibility from a finance perspective also. Thank you. Thank you. And with that, I think we come to the final slide. Let's go through quickly the key takeaways. I think we're right on time here to have our Q&A session. We have hopefully shown you that we have demonstrated the prospective validation of our AP3 platform, and we have achieved our key objectives that we described in the beginning with this initial clinical data with a cutoff of April 1. We show a 50% confirmed overall response rate in patients with OncoSignature-positive gynecologic cancers, the known ovarian and the predicted endometrial. We show an enrichment of response rate in ovarian cancer, and we show that we can not only predict prior to entry of clinical trial start sensitivity in tumor types, but we can also go in with the individualized OncoSignature patient response identification in these new tumor types and achieve response rate enrichment here. In this case, we are currently having a good initial data set on endometrial cancer, a very, very promising tumor type. We're extremely excited about that. So these 5 confirmed responders, as we mentioned, as a data cutoff are in treatment, and duration of response is still maturing, but we have a good grip on that. We hope to report more on that. We have the initial prospective validation. Very importantly, that's kind of the whole foundational concept of the company is around the AP3 platform and demonstrating with the segregation with the p-value of 0.0038, the segregation of true resist responders, and actually, very importantly, a complete segregation of all those that responded over to the OncoSignature-positive arm, where we have a high disease control rate as well. I think there's only 2 PDs out of these 10 patients versus the 16 negative where there are 0 responders is obviously a very clear segregation of sensitive versus nonsensitive. And because of that, we achieved that very significant p-value. And we have that streamlined AP3 infrastructure now where we use co-crystallography, rational drug design in parallel with AP3 phosphoproteomics profiling to streamline preclinical inhibitors internally. And we have new candidates on the way. And as mentioned, we have demonstrated that superior single-agent activity truly across models. We are extremely data and science-driven and will remain so as long as this team is involved. And we have a ton of data we haven't shared yet, but we had a poster at AACR on this molecule that was very well-visited, I should say. In fact, the team was very busy showing there. And as mentioned also, we now have accelerated timelines with this, I should say, very well-behaved molecule and are anticipating filing in next quarter and then the first-in-human in fourth quarter. We have the pro forma cash and marketable securities of $234 million with, I would say, a comfortable runway with a lot of flexibility into second half of 2026. Just a final closing slide that while we are focused in oncology here and especially the DDR space, this method is very broadly applicable, modality agnostic. We are aiming to move outside in the next phase of Acrivon. Beyond the new cell cycle regulatory program, we will most likely start moving over to fibrotic inflammatory and autoimmune diseases where one does not have genetic alterations on an individual patient basis to guide precision medicine. With that, I'd like to thank everyone for listening in and close off the session here with our Q&A session. Thank you so much. Thanks, Peter. A number of questions have already come in during the presentation. It's great to see many of you excited about the positive data. Thank you for that. We are too. There are some overlapping themes. Let me try to distill those, Peter, in real time into three initial questions for you and the team. The first is whether we can elaborate a little bit more on the importance of a proteomics-based approach, why that is differentiated. The second, whether that approach is truly independent of underlying genetic alterations and whether that is critically important. And specifically, whether certain genetic backgrounds are more suited for this approach than others. And that question is also accompanied by the observation that genetics-based approaches have not worked well in the DDR inhibitor space previously. And third thematic question is. Hang on, hang on. Can you repeat the first one, and then I'll try to answer that? Yeah, sorry, Peter. Sorry. Yeah. Can we elaborate a little bit more on the importance of a proteomics-based approach and why that is differentiated? Yeah. So that's a key question, obviously. Our method is completely agnostic to understanding in an annotative manner what's going on in a tumor cell or any other disease cell at the genetic or transcriptomic level. Because we can measure the activity states of what drives the disease, it's very easy. We have publications on that. You can look at the disease tissue versus normal and see what's upregulated at an activity state level. And we can match it with the drug mechanism of action. We are completely agnostic to what the underlying genetic alterations are. For example, in our patients, we have more heavily pretreated and less heavily pretreated patients, BRCA mutant, BRCA wild type, all kinds of prior therapies, our endometrial cancer patients, all three responders there have all had pembrolizumab and lenvatinib, etc., progressed on these. This shows both the power of the approach, the OncoSignature platform, and of course, also of the drug. But our approach, the value of it is because we're just looking at what drives the disease and we can match it with the drug effects on that disease-driving core axis, we have a very, in a way, simple way to predict sensitivity. So that's the importance of that differentiation and why it goes way beyond simple single-gene driver mutated diseases or cancers and, say, diseases with a simple synthetic lethal context, where it's very easy to just annotatively link that with prediction of response to the much more complex diseases and heterogeneous diseases to which we also believe that inflammatory bowel diseases, fibrotic diseases, kidney, liver, lung, etc., belong. So that's the importance of that. It's completely differentiating that and agnostic to underlying genetic alterations. Yeah. And I think your last comment addressed the third part of the thematic questions we've received about how broadly applicable AP3 is and our drug-tailored OncoSignature technology. Yeah. I would say that very importantly, right, that's why we're interested in moving outside oncology, which was really our first proof of concept here to get started. There's a much broader potential, and our ambitions are very big to tackle the, in a way, just as, if not big on, med need outside oncology. Autoimmune inflammatory diseases are really, in a way, underserved. There's a lot of revival in those with new technologies. And I think this year is really prime for such an area where we cannot use genetics very well, at least. Great. Our next question is from Roger Song at Jefferies. He has a three-partner. Back on the 368 clinical data, what is the median follow-up for OncoSignature-positive and negative cohorts? How will you move into stage 2 for OncoSignature-positive ovarian endometrial cohorts given the promising data? And what will be the registration path for OncoSignature-positive and negative tumors? Okay. So you might try to take notes here. Onco-positive plus negative. Okay. So median follow-up, as of data cutoff, I've shown that. So we lock down the threshold from the moment one consents a patient till you actually have a patient on treatment. It can be within a month or so, but it can typically be towards a month with biopsy taking, processing, etc. We have the first imaging at eight weeks. The second imaging that confirms the response or efficacy is at 16 weeks. So the patients we are showing you now are those that have at least that 16 weeks plus, and some of them are out further. They are on treatment. We are describing here as of data cutoff, and they are on treatment as of that. That's all we can say at this point. This is with that data cutoff. That's how we have shared the data also recently. So we do know what it is, the duration of response from over 250 published recommended phase II dose treated monotherapy ACR-368 ovarian cancer patients from past trials. And we know it's in the anything between 5.7-5.8 months, medium duration of response up to over 10 months as published in different single-center phase II studies as well as a multi-center 46-center, 8-country study in 169 patients. The stage 2 of the so very simple. When one powers the registration on 10 trials, obviously, simulate stage 1 and 2 trials. Typically, one has a futility analysis after the stage 1, which allows one to seamlessly transition into stage 2 if it goes well. And one has to power that based on a target product profile and a bar and a lower bound of the confidence interval separating. And we are obviously using the competitive landscape. The Elahere approval was extremely illustrative for us on where the boundary is for the lower bound, what FDA agreed with in that regard. One can imagine that's what we're using for our trial powering as well. Now, with response rates like we're seeing here, which I forgot to mention, are very, very consistent with what we saw, 47% and 58% in our past pretreatment tumor studies that were conducted entirely blinded, prospectively designed with a third-party biostatistician. These were done on pretreatment tumor biopsies taken from metastatic lesions in very much the same patient types as we have in ovarian cancer here. Of course, endometrial was never done. That was a prediction with the AP3 platform. These patients have all kinds of genetic alterations and underlying mutations. That response rate obviously is such that the number of patients required to achieve the same confidence interval boundaries is much, much smaller than the number of patients the conservative powering in our Simon stage is done with. Such that with such response rate, your Simon stage one number, we have listed it as 23 conservatively, could be much, much smaller depending on how response rates hold up. At this point, we are not saying that we need to get that. We certainly believe we know that we don't need that. But we also are saying this is what we are seeing so far. This is initial data. The third question was the regulatory pathway for OncoSignature-positive and negative patients. I think for the OncoSignature-positive, this is a monotherapy trial, obviously, Simon two-stage. It's very clear that that aims for the accelerated pathway with an accelerated approval. Those typically take a matter of years, and we are aiming for approval, and we are well underway with that, the locked threshold, the fact that we got clearance for recommended phase II dose by FDA in not only ovarian but also the 2 predicted tumor types. And very importantly, we could show our confidential blinded prospective assigned biomarker data. So despite it's a pioneering test, we got also clearance for usage of that in that registration on 10 trial. That is obviously with the goal of registering that through an accelerated pathway followed by a confirmatory trial. The OncoSignature-negative patients, that's what we call an exploratory phase 1b/ 2 trial. We have literally just filed and achieved the ability to transition in that exploratory trial into the phase II. The difference between the drug-tailored OncoSignature where positive OncoSignature predicts sensitivity to monotherapy is in OncoSignature-negative patients; it's our AP3 platform that uncovered the dominant resistance mechanism, which could be overcome with ultra-low dose or gem. So that's not individualized patient response prediction. We predict a proportion of patients to become further sensitized, and we are seeing that already in the initial data being confirmed. We need to continue in the phase II trial now to see how that translates with the recommended phase II dose into potential true responders or whether we just have some disease control, which we are already seeing now, which is clear evidence of activity, but not obviously with PRs or CRs reported yet. So that one is depending on the outcome of that could be relevant for a future confirmatory trial where one could imagine that we leverage the sensitization and the OncoSignature upregulation mode for low dose gem by combining that to both OncoSignature-positive and negative patients. That could be a basis for a creative trial assigned in the future. I hope that answers that, Adam. Thanks, Peter. And thanks for the question, Roger. Our next question is from Marc Frahm of TD Cowen. And Mark also has a multiple-part question. And several others have asked these same questions. Can you speak to the demographics and baseline characteristics of the OncoSignature-positive patients and the responders in ovarian and endometrial cancer? Yeah. Were the ovarian cancer patients, were they platinum sensitive? Question around prior PARP therapy and prior PD-1 therapy. I can answer that indeed. So let's start with the ovarian cancer patients. All our patients are platinum resistant. None of them are platinum sensitive. They're truly platinum resistant. They must have aggressive disease. The median line of prior therapies is four. I think in case of these responders we have now and in general, we have seen usage of PARP inhibitors, and we have both BRCA mutation and BRCA wild type. Most of them have been on angiogenesis inhibitor treatment as well. For endometrial cancer, all of them have had pembrolizumab, lenvatinib. I can mention that several of them have progressed. Best response to pembrolizumab, lenvatinib has been progressive disease. We saw the deep responses there in endometrial with our drugs. Demographics of ovarian, it's very similar to what was reported in the Lilly multi-center JTJN study, which had four cohorts. The three of them were platinum resistant. They were heavily pretreated and less heavily pretreated BRCA wild type. Then they had all lines of prior therapy, a BRCA mutant that must have had prior PARP. It's very much the same we are seeing here. That was the first question, I think. Well, that was two questions, Adam, more than one. Yeah. No, that's it for that set of questions. Let me turn to Joe Catanzaro from Piper. Joe has a few questions also. The first one being, is the monotherapy efficacy you've observed in ovarian and endometrial sufficient to fully expand to the full target enrollment of the Simon two-stage design? Again, that's Joe from Piper. Because it's initial clinical data and the numbers are still somewhat limited. Where we are, we want to see more patients coming in. So it's driven by statistics, but it's also driven, obviously, by getting a sufficient number. So while in theory, at least in endometrial, one could already start thinking of things like that, it is too early to start considering that. But thank you for the question. We're obviously very excited. We want to see how it holds up now. We're off to a good start. I think the key thing for us as a company and for patient benefit is, frankly, to show that we have a way to enrich for those patients that truly can benefit from this extremely active drug, actually. And we have benchmarked ACR-368 against our own WEE1, PKMYT1, and all kinds of other inhibitors. It's a very, very active drug. And that's why there's been so much clinical activity in past studies where Lilly was looking. This was a flagship program for precision medicine approaches for patient selection. And the reason we have this compound in-license, remember, that's from Eli Lilly, is because we demonstrated that with those heat maps that show these are actually historic data we showed to them that we could identify the responders. So at this point, while there is a broad activity of this agent, we still want to see in this new predicted tumor type endometrial, want to see a little bit more how it holds up. And that's how we go about it. Great. And Joe also asks, given AP3 selected endometrial as a potential indication and the clinical data you're observing thus far in endometrial, how much added confidence do you have that urothelial will see a signal in OncoSignature-positive patients? Can you tell Joe that's a fishing question? Joe, thanks for the question. Basically, it's very simple. The enrollment has gone slower so far with the oh my God, it's flashing down here. In GU cancers. We simply have fewer patients there right now. You can expect an update on that later. I cannot comment on that. But we have already given our initial soft clinical readout last year where we confirmed activity in all three indications. Let me just stick to that. It's premature to draw that conclusion, Joe, based on that. Let's turn to 2316, a question from Matt Biegler of Oppenheimer. Can you speak at a high level to the need for 2316 now, given the very encouraging data from 368? Ostensibly, there will be some overlap in indications. Do you think hitting PLK1 is a bad actor? Let me pause there. He has a secondary set of questions too. Okay. So great question. So first of all, the question relates, I think, in part to mutual redundancy or exclusivity of patients that have been treated with either inhibitor. And I think I want to bring up quickly a slide to answer that question. Do you see the slides again? Yes. Okay. So Matt, look at this slide here. So this is activity states here. When you inhibit with 2316, you see the on-target driving. It's not hitting PLK. It's activating PLK. And it's by design. This is the third most across studies, this is the third most activated kinases, which is what we have specifically designed it to do. So in these types of experiments that Kristina showed here, if you dial in on PLK1, you'll see a massive number of substrates as well that are significantly phosphorylated and activated. Why did we want PLK1 not to be hit but rather be activated, which was also the basis for choosing that? PLK is up here in these red upregulated notes. Well, it's because PLK1 through interaction with Aurora kinases and other things is absolutely essential for mitotic execution. You do not want to inhibit PLK. It's also an essential gene. So you want to activate PLK1, in our opinion. We see that massive cell cycle phenotype here. That's why we are highlighting that. This shows that mitotic catastrophe, the cells are just forced with PLK1 and CDK12 release into that S phase, where they undergo that mitotic catastrophe because at about four hours, which we see in our experiments, that's where the DNA repair tries to kick in, and that's inhibited with V1. Regarding the mutual exclusivity, you can also see a little bit of an answer to that. Notice that we see when you inhibit with a WEE1 inhibitor, you see the compensatory upregulation of other parts of the DDR axis. Here you see, for example, ATM are kicking into trying to rescue. We see that at the level of Chk1 and WEE1 as well. And that's why there's synergy with these two molecules. As shown here, we predict and have shown that there's also synergy. This is similar to what we see with gem and ACR-368 with a WEE1 inhibitor. So that shows there's functional exclusivity as well. But also we know that when we inhibit with either, you get a drug-induced upregulation compensatory mechanism of the other. So in fact, we believe that you can have sequential treatment of patients. What the ultimate overlap will be remains to be shown. That's why we are so excited about that broad activity we have seen of ACR-2316. We believe it's active and will be active in many, many other different types of cancer. And not shown here are pancreatic cancer and others where we know DDR inhibitors are very active. So we have 4-6 major indications, some of them much more prevalent than endometrial and ovarian lined up for monotherapy with ACR-2316. Very good. And Matt has also, as has Roger, asked, did the biomarker threshold change midway through the trial? Looks like after 12 patients. Can you provide more detail on that? It was built in per trial protocol. If you look at the S-1 filing, that's actually why we share the confidential data with FDA. Our team is one of the few teams that have a great experience with what is called aging. So when you go from archive samples that we had developed all the pretreated tumor studies on and established the thresholds on, when you go to freshly fixed samples, there's a slight intensity increase over the first month or so of routine fixed and processed tissues. We were aware of that. We shared that with FDA. It's a well-known phenomenon for them as well. And that was the basis on which they could give us clearance to use this test in a prospective registration-intent trial. Had we not built that capability and to adjust that threshold on the way, we would not have been granted that green light for a registration-intent trial, even if we had RP2D clearance, which we also got. We incorporated it such that in the first up to 12 patients, and it's all in the S-1 filing, if you look at it carefully, that in the up to first 12 OncoSignature-positive patients of each indication, we could use both the OncoSignature-positive and negative patients to adjust the threshold based on the frequency of positive and negative that we had achieved and learned about from pretreatment tumor from studies and pretreatment tumor biopsies and commercially available samples prior to clinical trial entry. We locked that down per protocol. Everything we're showing you here is prospective data with a locked threshold. That's why this is counting towards the registration intent. Very good. Our next question is from Etzer Darout. We are over time. We are over time, Adam. So happy to take more questions, but we probably should maybe take a couple more and then wrap it up here. Absolutely. Our next question's from Etzer Darout of BMO Capital Markets. Etzer asks, as you think about potential opportunities in autoimmune disease down the road, do you see any low-hanging fruit or indications that appear to be a good setting for proof of concept of the platform? I think that there's no other way I can think of. Etzer, thanks for the question, where a function-based approach is more suitable, is more needed than outside oncology. In oncology, we have a subset of tumors, as you know, and I used to spend my entire academic career focused on those, what we call simple single-gene driver mutated cancers. All the cancers with a gain of function point mutation or an amplification, think HER2 or BRAF, V600E, or a fusion, EML4-ALK, crizotinib, etc., where it's very easy to just look at that and predict sensitivity. Outside that, you don't have that. The proof of concept we are generating, hopefully with this data here already, is in very complex cancers. We know very well that in general, in the entire DDR space, genetics has been met with very limited success. There are no recurrent genetic alterations that are useful in the DDR pathways per se on the targets we're working on or in proximal nodes. It's very complex genetic alterations because the cells are typically having premature DNA replication. There's a very, very complex genetic alteration pattern in the genome. So it's very complex to predict. The fact that we can ignore all that and just look at what drives the disease is completely equally relevant outside oncology as in oncology. So I would say that in and of itself really opens up for the diseases outside oncology, where you have perturbed the dysregulated signal construction, which we can measure with our proteomics method. And we can see how the drug acts on these. It's perfect for those disease areas. We've started in oncology. That's where we have our basis. And that's where we think there's the bar is higher to prove what we did because genetics is there as in competing. Outside that, genetics has at best worked as so far so-called classifiers that can slightly enhance, increase, or decrease for sensitivity, as you know. So we think it's particularly suited for that. We think this year is a very strong proof of principle for the disease agnostic approach here that we are taking. Great. If we have time for one more, Emily Bodnar from H.C. Wainwright & Co. Emily asks about the strategy going forward in the OncoSignature-negative low-dose gem combo arm and what the strategies are that you're exploring to get patients to reach a response. So I think I answered the first part of that question earlier. But just to repeat, we have just in our latest filing declared that we have achieved recommended phase II dose of 10 milligrams per square meter for ultra-low dose of gem in combination with RP2D of ACR-368 and 105 mg per square meter every 14 days. We have just transitioned into that phase II. We know there's clinical activity, but it's not individualized OncoSignature-based prediction. This year is OncoSignature-negative. We give ultra-low dose gem. We have shown also, as on our poster at the AACR meeting, that we upregulate the OncoSignature when we give that ultra-low dose of gem in a proportion of patients. But because these biopsies here are taken prior to any low-dose gem or any therapy, we simply are taking the negative now and expect a proportion of them to turn positive. We simply want to. That's why we call it exploratory. What I didn't mention at the slides we went through, but a very important point was that our unenriched, if we look at all the ovarian patients to date, we have a response rate of 12.5%, completely dead spot in line with the JTJN trial. So we know that we'll have some ability here to further enhance. And that's why we know it does sensitize. What that proportion will be and might be remains to be shown. But we want to explore that by continuing the phase II trial for a while. The second question is, can you repeat that, Adam? Yeah, I think you covered it. It was just the strategies to gain a response in those patients. Yeah. I I mean, it's called exploratory for that reason. We don't have an individualized. As I said, the knowledge that ultra-low dose of gem further upregulates the Chk1/ Chk2 DDR stress axis specifically around Chk1/ Chk2 can be deployed towards getting deeper responses also in OncoSignature-positive patients. So now one can imagine that even if it's a small proportion of the negative fraction that gets converted to sensitive, if we gave all ultra-low dose gem, you can imagine you expand the number of patients that can benefit. And that could be a strategy that one can imagine can be deployed downstream. With that, thank you so much to everyone for great questions and a great session. Thank you for your time today. Thank you. Thanks, everyone. Thank you all. Thank you.