Can understand exactly the impacts we've seen over decades, which is low probability of success, mostly mechanisms that just follow each other in a kind of me-too fashion, and not really understanding why things work or don't work, which creates a lot of risk and ambiguity. What we have done over the past 15 years across both our work at Stanford and the work at Alto is be systematic in understanding the biology of individual patients with a range of psychiatric disorders or focuses on depression, bipolar depression, and schizophrenia. Right now, we've also done work on PTSD in order to understand how their brains function differently and how that different function leads to the ability of certain interventions to work better than others. The way we look at it is it's not, for example, take depression, one diagnosis, which we currently make based on clinical grounds, but a range of individual sub-diagnoses that clinically look similar but biologically look distinct. We develop drugs for those distinct populations as defined by tools that measure brain activity, either directly, things like EEG, electroencephalography, or indirectly looking at behavioral performance in tests of cognition or emotion or reward processing or wearables that tell us about daily rhythms in sleep and activity patterns, all ways to get at what is a true underlying biological structure so that we can take a more targeted development approach, much like oncology and immunology have done quite successfully to date. Got it. Before we dive into some of the upcoming clinical milestones across your pipeline, maybe you could tell us about the data recently presented at the Society of Biological Psychiatry meeting and how it supports your precision psychiatry approach. Yeah. So we presented six posters. Let me highlight four of them, which really are two themes. One of them uses a machine learning approach, something we've done now over years and really finessed our ability to implement, focused on using EEG to predict placebo response in depression. We know placebo response has been a big deal in depression and that it has risen through the years. It's been something that's created a big problem for differentiating drug from placebo effects. The placebo response, like active treatment response, is something that can be predicted and is something, in fact, that you see shared across all treatments. It's a nonspecific response when you expect a positive outcome. Turns out we could predict it with EEG. We found a signature that does that. We could replicate it across multiple prospective data sets, both placebo-armed from randomized trials and other treatments, and then use that as a way to essentially downweight people who are high predicted placebo responders and upweight people who are low predicted placebo responders, regardless of what they got, to enhance drug placebo differences. That was pretty cool. The other side that we were presenting on is taking, again, a machine learning EEG approach that we'd previously described to find a biomarker that enhances response to ALTO-300 or agomelatine, a drug that's an approved antidepressant in Europe, and we have in a phase 2b right now guided by EEG. What we did is we asked that though we have found and replicated this biomarker entirely data agnostic through machine learning, can we use the understanding of the EEG signal itself, which you can take all the way to animals, to get an understanding of the molecular mechanisms of this biomarker? What we were able to show both in animals and in humans is that the biomarker that we found that predicts response to ALTO-300 does that because it's tied to the molecular mechanisms of action of that drug. ALTO-300 includes a component that inhibits 5-HT2C receptors. When you activate 5-HT2C receptors, you induce that biomarker and you make animals depressed looking. When you deplete dopamine, which is the opposite of what the drug does, again, you create that biomarker and you can show that even in humans. Across the two of these, it strengthens the argument that you can find signals that differ between people that are robust, that are reproducible using tools like machine learning and EEG, and that through that you get really unique insights into mechanism that, frankly, I've not actually seen something discovered through machine learning in humans in the brain before that has gotten down to the level of a mechanistic, a molecular mechanistic understanding. A lot to be excited about in the science and obviously, as we're going to talk about shortly, in the clinical trial readouts using this information. Yeah. If we kind of take some of those readouts in order, I think one of the great things about what you guys are doing is you've ensured that there are lots of readouts across your pipeline. A nice steady diet of catalysts there. We're expecting the proof of concept data for ALTO-203 in MDD this quarter, I believe. Maybe you could provide us a brief overview of the asset and its history first. Yeah. Alto-203 is a histamine H3 inverse agonist. It blocks the activity receptor and reduces the basal activity further. The interest in that molecule has been through its effects, through the H3 receptors' effects on dopamine, and in particular, dopamine within the reward system. When you inhibit H3 receptors, you lead to an increase in dopamine. What differentiates this drug from others is that it leads to an increase in dopamine not just in cortex, but also in the reward system subcortically, which for things like motivation, which can affect both mood and cognition, are particularly important. What has been shown in the past for this drug in healthy people is that it can lead to an acute increase in positive subjective emotion. That is with even single doses. That is part of what we are going to be looking at in the readout to come this quarter. That is single doses that have been given to patients with depression and anhedonia, but also looking at the broader aspects of what increasing dopamine does. Those things are, for example, cognition and motivation, brain changes, behavioral changes, and so forth. It is an early stage clinical study. It is focusing on a pharmacodynamic response, but one that could guide an understanding of how to develop the drug, given our biomarker platform, given that we are not just, as in the typical study, just looking at maybe symptom changes and trying to kind of guess at clinical effects, frankly, too early in the case of most early stage phase 2 studies to really understand them. We're using the biology here measured as cognition, as various behavioral measures, as subjective measures, as EEG measures, as wearable measures to understand what exactly the drug does and put that in context of understanding of various illnesses and their biology. Got it. I mean, obviously, without giving things away, what do you think investors should be watching for in the ALTO-203 data? It seems like, as you said, not major indications of efficacy at this point, but where should we be watching? I think what they should be looking for is a clear story on what the drug does so that you can look at the range of data presented and say, "Yeah, you know, with all of these ways to describe the pharmacodynamic effects, I understand what the drug is doing and now understand much more clearly how to develop it." Because that's really the argument here is that by using these biomarkers early and often, as it were, we can really de-risk development by understanding exactly the brain effects, dose effects, and so forth. Frankly, we see that as well in other stories we'll get to, for example, ALTO-101, which is a readout upcoming at the end of the year in schizophrenia for cognition. All of this is to take this stepwise from early stage using biomarkers as outcomes to understand what the drugs do and later stage to understand who the patients that are responsive are using biomarkers as patient selection approaches in order to de-risk this enterprise. Great. I think that would be a great opportunity to talk about ALTO-101. As you said, I think the proof of concept readout is expected in the second half of this year. What can you tell us about the asset? Yeah. Alto 101 is a PDE4 inhibitor. It's a mechanism that obviously has been very successful in the immune system with drugs like Otezla. The mechanism has long been of interest for the brain, the idea being enhancing neuroplasticity by enhancing the levels of cyclic AMP and that improving cognition. The problem has been historically twofold. One is not knowing how much to give in order to hit the right brain targets. That's partly needing the same kind of pharmacodynamic outcomes that I just talked about from the Alto 203 perspective. The other has been a tolerability challenge. That is that for all PDE4s as a class, as you give more of them, you get nausea, vomiting, and diarrhea as dose-limiting effects. We actually addressed both of those before we launched the study that's ongoing in schizophrenia. On the how much to give for what kind of brain effects and quite frankly, using that information to understand indication, what we did is we did a number of things. Number one is we understood from the perspective of schizophrenia in a large data set that we also prospectively replicated the findings in. We found which EEG biomarkers are best indicative of the disease, cases versus controls, and cognitive impairment in patients. Cognitive impairment associated with schizophrenia or CIAS is really the feature of the disease that starts earliest, is the most predictive long-term of people's functioning much more than psychosis and has no treatments whatsoever. Understanding the biomarkers of the pathophysiology is critical. We then found that ALTO-101, when given as single doses to healthy people, led to an increase in those EEG measures. In particular, we highlighted what's called theta response or a low-frequency response to auditory stimuli, so low-level sensory processing abnormality. That's actually the best correlate of these cognitive deficits in patients. We saw that as a dose-related effect. We also saw dose-related improvement in cognition in these individuals and that the theta response was correlated with cognitive improvement. It gave us the understanding mechanistically that we're on the right path. Actually, last week at Biological Psychiatry, we reported that even in animals, you could show that in models of schizophrenia, you have a reduction in theta and then ALTO-101 rescues that. That sets up our proof of concept trial, which is 10 days of treatment in a crossover fashion in patients with schizophrenia and prominent cognitive abnormalities, looking for improvements in both theta responses, the primary outcome, and then cognition as well. Now, the tolerability aspect of it is its own kind of interesting twist. What we noticed is that the adverse events that are characteristic of the class come at Cmax in a dose-related manner. We hypothesized that by slowing down absorption, we may be able to get people to the same level, in fact, increase their overall level of exposure, but without the adverse events. What we did there is we reformulated the drug instead of an immediate release oral to a patch, which gives you, as a transdermal patch, a nice steady state level that comes kind of slow to reach its peak and then over a couple of days reaches that steady state. What we found is that not only does that fairly dramatically increase the total exposure, reaching roughly the same level of exposure at any given moment as you would otherwise get with the oral at Cmax, but that by doing so, you dramatically reduce adverse events. You take something where to give that dose would not be well tolerated at all orally to now actually being very well tolerated when given continuously. That is a pretty exciting setup. What we have in this trial, as I mentioned, is a crossover, so well-powered study. What we will be looking for is those brain changes, be they at the EEG level as the primary outcome that we think is the operative kind of biomarker to track, but also looking at cognition with 10 days of treatment. It's not the full course that you ultimately expect for the phase 2b or phase 3 program to look at both stable cognitive and functional change. That's actually the next step if the trial succeeds, is taking that longer treatment path in a registration-type design. Got it. I think you spoke a bit about why we should have confidence in this when you alluded to the mechanism of action and some of the animal studies. Maybe can you talk a little bit about the development and treatment landscape for CIAS? It seems like there is some white space and opportunity for any assets that can do well there. Yeah. Huge white space. Again, if you think of the disease itself, right, this is the thing that starts often in your teens long before you have psychotic symptoms. Most patients will spend their time not psychotic, but cognitively impaired. The prevalence of cognitive impairment, of course, is very high and it predicts long-term course. The need for something A that works and B that you could take on a chronic basis is massive. There's also very few programs going on right now because there's been a huge historical kind of effort here that has failed. I mean, the most recent was a large phase 3 program from Boehringer Ingelheim looking at a GlyT1 inhibitor. I think what we've learned from that is critically what we've injected into the sort of DNA of Alto is take this advancement stepwise, show that you're getting engagement in the right brain circuits measured with EEG and cognition, understand your dose, understand how to track your biological effects and how to scale this up. We think that that will substantially de-risk development here. We can track, for example, that EEG signal throughout a therapeutic course of treatment. It's a passive test that people do just listening to auditory stimuli. When you look back at some of these other programs that have failed, oftentimes they've skipped that step or they've just gone right through it, not really showing clear brain effects and not being able to take that into patients to guide how development happens. Again, it's taking a page out of the playbook of precision medicine elsewhere, just applying it here to psychiatry in a new way. Got it. You have other late-stage programs and catalysts extending through 2026. Are there any that you would like to highlight at this point? Within this context, maybe you could speak to some key learnings from the ALTO-100 phase 2b MDD trial and how you've adjusted some methodologies. Yeah. Let me talk a little bit about ALTO-300 and then, as time allows, the 100 bipolar program. ALTO-300 is targeting mid-next year for a phase 2b readout in 200 biomarker-positive patients. It's a large study as well as a portion of patients without the biomarker. As I mentioned earlier, we found the biomarker through machine learning. We understand its molecular mechanisms. The drug itself is agomelatine at 25 milligrams, which is a drug that's approved at that dose for the treatment of depression in all comers in Europe and in Australia. What we found in the biomarker is that it's a particular group of patients that respond particularly well. Now we understand the reasons for that. The drug itself is really well tolerated. It It has always been extremely well tolerated, in fact, better tolerated often than even SSRIs. Here, we're using it as an adjunctive treatment in depression, so on top of a failed antidepressant, which means the comparison set here is a bunch of drugs that are very poorly tolerated. Those are antipsychotics and esketamine. We know the chronic effects, metabolic and movement disorder effects of antipsychotics that, frankly, other clinicians really would like to avoid having to use if they had a better choice. You're alluding to what we learned from the ALTO-100 program. What happened this fall is we did not reach statistical significance in a depression study focused on ALTO-100 using a biomarker in that case of cognition for a pro-cognitive antidepressant, putative antidepressant. The reason for that was one around professional patients in a subset of the study. The study looked at both adjunctive patients, which is the kind that we're studying for ALTO-300, and those taking the drug as monotherapy. The adjunctive patients showed a really nice effect size, Cohen's d of 0.47. Despite not being powered, we were nearly statistically significant. The monotherapy patients didn't show an effect of the drug. What it came down to is an issue of compliance, so 100% compliance in the adjunctive arm, but much lower in the monotherapy arm, driven by particular sites that recruited a fairly substantial number of people that brought in a lot of non-compliance. That really highlighted the risk that we'd been aware of, but now really redoubling our efforts around with regard to understanding the site execution risk and the issue of professional patients. We've revamped our entire recruitment approach. We've inserted things like a sponsor eligibility review. We look at antidepressant levels in people's urine. Really making sure that we have the right patients each and every time. We also have done a case review, blinded case review of our ALTO-300 participants, excluded sites with this high-risk behavior that's associated with professional patients. Did an interim analysis earlier this year, which showed drug-like signal and supported continuation of the study as we have been doing, now targeting mid-next year. There are a lot of elements that de-risk this program, the fact that it's a known antidepressant, well tolerated, that we have done the interim analysis, that we understand the basis of the biomarker. As we think about the commercial utility, we also know that historically there's been some risk of LFT elevation at higher doses, 50 milligrams of this drug. It's never been an issue, but led to a requirement for monitoring. That is not the case for the 25 milligram dose that we are developing that has been now through thousands and thousands of data sets in real-world evidence shown to be safe, well tolerated, not to lead to LFT elevation above kind of the standard rate you would expect for any drug. We are excited and looking forward to that readout next year, having hopefully taken a lot of risk points off the table as we move to these big phase 2b readouts, which put us then obviously right into a phase 3 program, which would be a very exciting place to be for a precision psychiatry approach. Got it. I think you addressed some of this here, but I'm curious just if you can speak to how much read-through you would expect from any one of these readouts to the broader biomarker selection strategy given kind of the differences in the assets themselves. Yeah. Read-through is partial in all cases because the biomarkers, as you said, are different. The drugs themselves are different. There is a common approach we're taking in terms of the data science approach, making sure that biomarkers are robust and reproducible and highly reliable. They are all different. It is important to see for any one trial, whether it succeeds or not, what the reason for that was. That is why, for example, we include a small group of patients without the biomarker in these phase 2Bs to understand what enrichment looks like. Everything follows kind of a standard randomized trial approach where it is one-to-one randomization and a number of other things that we try to do to mitigate placebo response. Placebo is always part of this equation that has nothing to do in a sense with the biomarker for selecting treatment response. That's always something to be watchful of in every trial and something we were actually quite successful mitigating in the ALTO-100 study. Each should be really taken on its own under that broader lens of a systematic approach for discovery that opens us up to the possibility of finding biomarkers across programs. Got it. In our last minute here, I'm just wondering if there's any late-stage trials or potential drug approvals that you're watching this year that could have an impact on one of the indications you're working on. Yeah. So we've actually been through a number of late-stage programs very recently from other companies. We've seen there's really been very little success in psychiatry. This is across half a dozen different readouts or so of late. The field is actually quite open. I think a number of the readouts this year are in areas not particularly related, things like psychedelics, which are really more of a kind of a niche area. The landscape is wide open with tremendous, tremendous need that's unaddressed across all of the indications we're looking at. We're just excited to see what our results are and making progress finally in a field that has been needing it for decades. Great. I look forward to digging into the details as all the clinical data comes out. I would like to thank you again for taking the time to speak with me today. My pleasure. Have a good day. Thank you.