Everyone, and thank you for joining us for our 2026 TD Cowen Neuropsychiatry Summit. I'm Joe Thome, one of the Senior Biotech Analysts here on the team at TD Cowen, and it is my pleasure to have with me today two members of the team from Seaport Therapeutics. We have CFO Lauren White, and we have CSO Michael Chen. Thank you both for joining us, and to the investors listening in. Maybe just to kick things off, Lauren, if you want to start at a little bit of a high level, maybe just briefly recapping what the company has achieved since the IPO and maybe what should be on investor radars into the end of 2026 and first half of 2027, and then we'll dive into the programs. Sure. Well, first, thanks for hosting us, Joe, and a warm welcome to everyone joining on the webcast. I will preface this by saying during today's conversation, we are going to be making some forward-looking statements. They're all based off of our current expectations and beliefs. So I'd encourage you to take a look at our SEC filings for any additional details. So happy to talk through upcoming milestones, which we have quite a few of, but maybe just to start by taking a step back and framing the opportunity ahead of us. More than 1 billion people are living with mental illness right now. That includes over 300 million each with depression and anxiety. And as you can imagine, this touches almost every family. Existing psychiatric medicines have helped millions of people and are among the most widely prescribed and commercially successful drugs of all time, but they have limitations, and there remains a clear unmet need for new therapies. Seaport is a neuropsychiatry company that exists to turn the tide on mental health and deliver new therapies for patients with a validated platform, a proven team, and a pipeline that's based on clinically validated mechanisms. Our most advanced program, GlyphAllo, has shown initial proof of concept and is now advancing in a phase II-B trial in major depressive disorder, which has the potential to be registration enabling. Our second program, GlyphAgo, also recently demonstrated clinical proof of concept, and we're preparing to initiate two phase II trials in generalized anxiety disorder beginning later this year. And our third candidate, Glyph2BLSD, is currently in preclinical development. This program really builds on the recent momentum in the psychedelic space and aims to harness similar pharmacology, but without causing a trip, which is exciting. All of these programs are built off of mechanisms that have already shown human efficacy, but they were held back by some sort of a limitation that our Glyph platform can now address. Glyph can overcome issues such as poor oral bioavailability, high PK variability, and other side effects. And the great thing is every time we Glyph a molecule, we generate a new chemical entity with novel composition of matter IP. We've got a seasoned leadership team with a track record of bringing new psychiatric medicines to patients and driving successful business outcomes. You're probably aware that our CEO, Daphne Zohar, and our Board Chair, Steven Paul, founded and led Karuna, and the approach that we're taking at Seaport is quite similar to Karuna. For instance, at Karuna, the team knew that xanomeline was an effective antipsychotic, but it was sitting on a shelf due to GI tolerability issues. They were able to address that with the invention of KarXT, which is now called COBENFY. At Seaport, we're taking a similar approach of identifying new drugs with proven efficacy, addressing the limitations that held them back, in our case, with the Glyph platform, and then designing and executing studies with an experienced team. At the end of the second quarter of this year, we had approximately $427 million on the balance sheet. We expect that to provide runway into 2029 and through some really exciting clinical catalysts for us. First, top-line data from our phase II-B BUOY-1 trial of GlyphAllo in MDD. We expect that in the first half of 2027. Next, top-line data from our phase II-A trial of GlyphAgo. That's in early 2028. Finally, top-line data from our phase II-B trial of GlyphAgo by year end 2028. We're really pleased with our progress to date, and we're excited about the road ahead. Perfect. Excellent. Maybe to drill down a little bit more just into the platform itself before we jump into GlyphAllo, can you just maybe give us a brief overview of the platform? Maybe how does the platform help therapies avoid first pass metabolism and maybe some of the benefits that the platform can offer to maybe the underlying therapeutic? Sure, Joe. Glyph is a prodrug platform. It's a prodrug technology, which essentially cloaks a molecule, a drug, so that the body recognizes it as if it were a dietary fat. What this does is allow it to be transferred through the lymphatic system, which bypasses the liver on its way into systemic circulation. This side road essentially allows us to avoid the liver, and because you avoid the liver and less drug gets metabolized, more drug is actually able to reach the brain. Also by avoiding the liver, this bypass overcomes issues like liver-related hepatotoxicity and also PK issues, low oral bioavailability, variable PK. For example, allowing us to reach therapeutic exposures, but with lower doses, or turning a drug that is only available as an intravenous drug into something that can be dosed orally. As Lauren mentioned, every time we Glyph a molecule, it's a new molecule, and we've generated an initial proof of concept clinically with our two lead programs, GlyphAllo and GlyphAgo. Perfect. That answers a little bit of the first question on the GlyphAllo platform in terms of what GlyphAllo can offer over allopregnanolone. Maybe if there's anything else specific about the molecule, GlyphAllo itself, a reference to allopregnanolone would be great. Then also you have had a phase II-A proof of concept study that showed that GlyphAllo blunted salivary cortisol results. Just interesting why take this approach, and maybe what did you learn from that phase II-A study? Yeah. For those who are not familiar, maybe a word on allopregnanolone and GlyphAllo, then glad to talk about that phase II-A. GlyphAllo, as you alluded to, Joseph, is actually based on an approved drug. Allopregnanolone is actually an endogenous molecule, like we all make it in our brains right now. But it's actually an approved drug as an antidepressant, was approved as a rapidly acting therapeutic with anxiolytic and sleep-promoting effect. But allopregnanolone itself has very low oral bioavailability. Over 95% of it doesn't make it past the liver into systemic circulation and into the brain. What that means is, even though allopregnanolone was approved for the treatment of a form of depression, postpartum depression, it was approved as a 60-hour continuous intravenous infusion. Now, we've shown now in the clinic that GlyphAllo, our Glyph prodrug of allopregnanolone, is able to overcome these bioavailability limitations of allopregnanolone and deliver systemic exposures that are therapeutically relevant, but with once daily oral dosing. For example, in preclinical studies, we had shown about a 9-fold increase in relative bioavailability compared to unmodified allopregnanolone. In phase I, GlyphAllo was not only achieving the PK profile that we're looking for, but generally well-tolerated. The pharmacology is retained. That is, for example, the most common adverse event we observed was somnolence, which is actually consistent with the pharmacology of allopregnanolone, which is a GABA positive allosteric modulator. So we've shown now that we can overcome these limitations of allopregnanolone, and in addition, we demonstrated the pharmacology in a phase II-A randomized double-blind placebo-controlled study in a clinically validated model of anxiety called the Trier Social Stress Test. Now, for those of you unfamiliar, this is a test where you take healthy volunteers, you put them into the clinic, randomize them, dose them, and then give them a stressful task. In this case, giving a speech, preparing and giving a speech in front of a live panel of judges who are judging the speech, and then also doing math live in front of the panel. So 2,024 - 17 over and over again, and the judges say, stop, you're wrong. Start over. As you can imagine, this is stressful and actually causes salivary cortisol, the body's primary stress hormone, to spike. As expected, individuals with placebo on board showed a big spike in salivary cortisol. But we showed that a single dose, a well-tolerated dose of GlyphAllo at the 375 mg dose that we're actually taking now into phase II-B, achieved a statistically significant reduction versus placebo in the increase from baseline to peak in salivary cortisol. So essentially blunts that salivary cortisol response. The only neuropsychiatric drug to demonstrate a similar result on the TSST is actually alprazolam, a benzodiazepine that's better known as XANAX. But allopregnanolone is not a benzodiazepine. It's, again, an endogenous neurosteroid. So taken together, we have phase I data supporting the PK and tolerability profile, as well as phase II-A data showing that we have not only drug in body and drug in brain, but actually well-tolerated and active doses that we've now taken into an ongoing phase II-B study. Perfect. Seems like an intense study that I'm glad maybe I wasn't in front of the class on that one. But you also ran a study on next morning driving. You, I guess, indicated before the potential of a somnolent signal with this class, I guess. Can you talk a little bit about why run the next morning driving study? Was it because to rule out some of the somnolence related AEs, or what was the rationale behind it? Then we've seen data now, so maybe what did you see from the study that gives you some comfort in taking this forward? Yeah. Maybe just to back up, we know that allopregnanolone is effective as an antidepressant. It has these interesting qualities, including sleep promoting and anxiolytic properties, but it's no longer marketed, it's no longer available. Now, a synthetic analog of allopregnanolone called zuranolone was actually advanced and showed strong and consistent efficacy in major depression, which is where we're developing GlyphAllo. So, for example, zuranolone, this oral synthetic analog, showed efficacy on a primary endpoint in five out of the six studies. It was actually approved for MDD outside of the United States. But it was held back by some clinical trial issues that we can talk about. One of the things that zuranolone, which is approved for postpartum depression, also has is a black box warning, label restriction for 12 hours on driving. Now, we feel that based on our phase I data, GlyphAllo is differentiated from a pharmacokinetic profile perspective. We also know that driving simulation studies are very commonly required for CNS drugs, especially those that have mechanisms that have been associated with somnolence. We conducted this trial a little bit earlier than maybe is typical in order to not only inform the design of future clinical trials, but also look at this very important point of differentiation from zuranolone, which is in the same class. Zuranolone is not approved for major depression. However, it is approved for postpartum depression with that black box warning. We conducted a randomized double-blind trial with placebo, a positive control called zopiclone, and GlyphAllo in healthy volunteers. This is a very standardized driving simulation study that is often used in CNS and other drugs. We assessed the effects of GlyphAllo dosed once daily at bedtime, which is how we are dosing it in BUOY-1, our ongoing phase II-B study. We assessed next morning driving. This study met the primary and key secondary endpoints, essentially showing that GlyphAllo did not impair next morning driving performance, either at single dose or after multiple doses, and was well tolerated. As I mentioned, these data are actually consistent with the PK data that we have shown in phase I with GlyphAllo, which showed that allopregnanolone levels after GlyphAllo dosing are reduced to about 15% of its peak at Cmax by eight hours after GlyphAllo dosing. This is notably different than zuranolone, which remains at about 50% of its Cmax all the way out to 12 hours after dosing. Altogether, we are very pleased with the results of this driving simulation study, which again, we feel adds to the overall profile of GlyphAllo, as well as the potential differentiation profile. Perfect. As you mentioned, the agent is progressing in a phase II study, with data expected in the first half of next year in MDD patients, both with and without anxious distress. Can you talk a little bit about why you chose that specific patient population, maybe to investigate in the initial phase II, and any comments on how enrollment is tracking? Maybe if the company is going to be able to, at some point, narrow from H1, I guess how you are thinking about that would be great. Yeah. I will just start by saying enrollment is on track. We have seen strong enrollment, and we continue to expect top-line data in the first half of next year. The point about anxious distress is actually a really interesting one, because as I mentioned, allopregnanolone not only has this rapidly acting antidepressant activity, but in trials in, for example, postpartum depression, had this anxiolytic activity and sleep-promoting activity. MDD with anxious distress is a specifier within the overall major depression diagnosis. It is not necessarily based on a cutoff on an instrument. It is actually part of the diagnostic workup, just like postpartum depression is. Our decision to enrich and look at this question of anxious distress is rooted both in the clinical data around allopregnanolone, as well as the unmet need. These are patients which represent, frankly, the majority of patients, 60% or more of patients with MDD have anxious distress. These are patients that respond less well to frontline SSRIs with worse quality of life, take longer to respond. It is a huge unmet need, and there are no drugs that are specifically approved for MDD with anxious distress. Now, because of this biological overlap, that is allopregnanolone having not only the depression effect, the antidepressant effect, but this anxiolytic effect, we feel it is very legitimate and interesting to test this hypothesis on whether GlyphAllo above and beyond a potential antidepressant effect may actually have an anxiolytic effect that would be very relevant in anxious distress. We saw this also in that phase II-A I mentioned. It is a stress test, and these are patients with anxious distress. While we are fully powered for overall major depression, drug versus placebo, we will be able to analyze in a more exploratory fashion the difference in effect of GlyphAllo on patients with or without anxious distress. I will also just mention there are some other interesting facets of this trial that are less related to targeting of anxious distress and actually what we would just call good clinical trial hygiene. These are, as Lauren mentioned, takeaways from Karuna Therapeutics on essentially clinical trial design and execution facets that we believe can help increase the probability of technical success. For example, we are randomizing patients one to one, drug placebos, so we do not have three, four, five active arms, as this has been shown to correlate with the extent of the placebo response. We are also looking very carefully at controlling the number and frequency of clinician-administered assessments, something that we have shown in meta-analysis to also correspond to the extent of the placebo response. Overall, we are looking at anxious distress, with or without, we are overall looking at depression, but we are also looking at other design features that we think can help improve the overall success of the trial. Perfect. The primary endpoint is the HAM-D, and companies pick whether they want to use the HAM-D or the MADRS and tend to track overall in line together, I guess. Is there anything specific about the patient population enrolled or GlyphAllo itself that might make the effects more apparent on the HAM-D versus the MADRS, or any thought going into that decision? Maybe just related, what's the study powering to show in terms of a delta on the primary that you planned? Yeah. So maybe on that second point, we're really focused on demonstrating a statistically significant and clinically meaningful difference. As you know, the bar for approval in this space is stat sig on a trial versus a placebo. We often get asked what good looks like, and we haven't guided on a specific effect size, although we'll note that therapies for major depression have ranged in effect sizes from things like 0.2 -0.4, but are often defined actually by their differentiation in terms of their ability to have an effect on the symptoms of depression. So for example, we feel that GlyphAllo could be positioned to uniquely address, for example, the onset of action. We know allopregnanolone is rapidly acting. We know that standard first-line care like SSRIs, SNRIs can have side effects like sexual dysfunction, weight gain, and insomnia. These are areas where we feel GlyphAllo could also differentiate, given that we know allopregnanolone doesn't necessarily carry the same risk, for example, weight gain or sexual dysfunction. On that third side effect, that is insomnia, we know SSRIs can make actually sleep worse. Allopregnanolone has this sleep-promoting effect. So beyond the primary efficacy endpoint, we're also going to be looking very carefully at how the overall profile differentiates and why HAM-D versus MADRS. As you pointed out, they're highly correlated. They're both gold standard essential endpoints that are used for registration. They tend to behave very similarly. But to your earlier point on anxious distress, the HAM-D has built-in subscales, for example, on anxiety and somatization, as well as insomnia. So we feel it could be somewhat better suited to capture some of the known activity of allopregnanolone. Perfect. Maybe just as we think about the potential registrational path here, was the phase II designed to be potentially one of two pivotal studies you needed for licensure? You need the data and the sign-off from the FDA, but maybe what is the base expectation? If this is successful, what would your phase III's look like? Would it again be sort of an all-comer MDD population with maybe some subset analyses for anxious distress? How are you thinking about that? Maybe just one more related point on, maybe for Lauren on the cash runway. Can you just remind us where you get in terms of timing and if a registrational program for the agent is included in that? Yeah, absolutely. Again, this is a little bit from the Karuna playbook, where a phase II-B that is adequate and well-controlled can look a lot like a phase III program. We believe the phase II-B is a potentially registration-enabling study. The exact nature of potential future studies, including phase III studies, I think will be dependent on the data that we see, the effects, overall population, anxious distress. I will also mention, and I know we are running a little bit short on time, but we do have a second program, GlyphAgo, that we recently had exciting data, and we are also advancing that into multiple phase II studies. Very similarly, the phase II-B that we are advancing for GlyphAgo is a potentially registration-enabling study that could mirror, again, the bar for adequate and well-controlled studies that have been used for registration in the past. And maybe, Joe, just on your last question. We have about $427 million on the balance sheet at the end of Q2. We have guided that that provides us runway into 2029. We cover the BUOY-1 readout in the first half of 2027. We do have some ramp up of phase III activities when you think about the earlier bigger ticket items like CMC, and that also gets you the top line data from our phase II-A and IIb of GlyphAgo in 2028. Perfect. Excellent. I do want to touch on the other parts of the pipeline here. Maybe we will jump over to GlyphAgo. I guess, can you talk maybe just first at a high level, maybe some of the important differentiators versus agomelatine, and it seems like really the safety and potential lack of liver enzyme monitoring could be differentiating here. If you could touch on that would be great. Yeah. GlyphAgo, like GlyphAllo, is actually based on an approved drug, agomelatine, not approved in the U.S., approved ex-U.S. for both major depression, also generalized anxiety disorder, where it separated from placebo statistically significantly in four out of four studies, so it works. It is actually well-tolerated, so you do not get the SSRI side effects, but it is a bit rough on this liver, specifically for first pass metabolism. It also has highly variable PK and drug-drug interactions, and these are all points of differentiation that we showed in a key de-risking phase I proof of concept study with GlyphAgo, our Glyph prodrug of agomelatine. GlyphAgo, as the platform allows, bypasses first pass metabolism, which allows us to demonstrate in phase I that we saw a nearly 7-fold, that is a 6.8-fold increase in bioavailability of GlyphAgo head to head against agomelatine in healthy volunteers in essentially a crossover type design. Not only does that mean that we can have more drug with less drug actually dosed, that is similar exposures that we know are efficacious but less drug, but it also means a 10-fold reduction in variability, as well as avoiding drug-drug interaction effects on PK. Now, why is this important? To your point, agomelatine has a liver function testing requirement due to incidence of liver function enzymes that can occur with agomelatine related to this high first pass metabolism. But with GlyphAgo, now that you can cut the amount of drug that the liver sees, you can retain the pharmacology, retain the exposure, but the liver sees less drug, liver is less stressed out, you have much less variable PK profile, and you have a new drug with new composition of matter IP. Important given that agomelatine no longer has composition of matter IP. On the basis of the de-risking phase I data, we are launching two studies, as Lauren mentioned, a phase II-A proof of pharmacology study. This will really be focused on sleep. We did not mention this, but agomelatine actually is sleep promoting, which could be very relevant for generalized anxiety disorder, where we are pursuing initial development. But the sort of main course, as I like to call it, is a potentially registration-enabling, randomized, double-blind, placebo-controlled study of GlyphAgo in GAD. We are expecting data from both those studies, as Lauren mentioned, in 2028. Perfect. Unfortunately, we are running out of time, but it would be great to ask a last question. Obviously, this platform, the Glyph platform, seems like it could have a lot of different applications here. I guess, how is the company maybe thinking about broader development? If it is not something that you maybe want to keep in-house, is BD on the table to use your Glyph platform with some others? Maybe any sort of commentary around that would be great. Yeah. Sure, Joseph. We really believe the Glyph platform has broad applicability across multiple therapeutic areas, and that we can help pharma companies solve some important issues and generate non-dilutive funding through partnerships while we, of course, as a company, remain focused on neuroscience. We continue to field significant inbound interest from those potential partners. We can also advance funding our programs with grant funding. You are probably aware, we recently announced a grant of up to $15 million from ARPA-H for a Glyph molecule specifically engineered to target and normalize dysfunction in the gut lymphatic system, which plays obviously a central role in metabolic disease, pancreatic cancer. We really see broad potential for the platform and ample opportunities for partnerships. Perfect. Excellent. With that, I think we are at time. A lot to take a peek at here, and we look forward to the data in the first half of next year. Thank you both for joining us, and thanks to all the investors for listening in this morning. Thanks.