Scroll Top

How Pangea Is Using AI to Find New CNS Drugs in Nature

The Harry Glorikian Show 

Pangea BioJohn Bogossian, co-founder and CEO, Sona Chandra, co-founder and head of AI  

Final Transcript  

Harry Glorikian: Hello. Welcome to The Harry Glorikian Show, where we dive into the tech-driven future of healthcare.  

I’m always looking for reasons to be optimistic.  

And one of them is that the combination of better data and more powerful computing is helping researchers reinvent the process of discovering new drugs.  

Getting a new drug to market takes a long time, so we may not see the results right away.  

But I think it’s a good bet that within five or 10 years, we’ll see a huge wave of new medicines that were either discovered or designed using AI. 

Those drugs will finally help us get control of our most stubborn health problems, from cancer to cardiovascular disease to obesity and metabolic disorders to neurodegenerative diseases.  

That said, there are a lot of startups working in this area. We’ve had quite a few of them on the show.  

And I think the successful ones are strong in two very different areas.  

First, they’ve gathered a lot of valuable, proprietary data, whether that means genomic or proteomic data or imaging data or what have you.  

Second, they’ve come up with original ways to use AI and other techniques to sift through that data to find promising new drug molecules.   

My guests this week are from a startup called Pangea Bio that’s working hard on both.  

They specialize in gathering data from the natural world, especially data about compounds manufactured inside the cells of plants and fungi.   

But not just any plants or fungi. They narrow down the possibilities by working with indigenous cultures to find the plants or mushrooms that people have already been using for centuries in traditional medicine.  

Then they’ve built three separate computational platforms that filter through all that data, to single out the small molecules that have the biggest effects in the human body, especially the central nervous system.  

Here to tell us a lot more about that whole effort are the co-founder and COO of Pangea Bio, John Boghossian, as well as the company’s president of AI, Sona Chandra.  

I really feel like the combination of big data and big computing means that CNS drug discovery is about to take off like a rocket ship.  

And frankly we really need that kind of acceleration, especially in areas like neurodegenerative disease, where there’s a serious shortage of drugs that actually work.  

So I can’t wait to see what Pangea comes up with. Here’s my full conversation with John and Sona.  

Harry Glorikian: John, Sona, welcome to the show.  

John Boghossian: Thank you. Nice to. Nice to be here, Harry. Thank you for having us.  

Harry Glorikian: Yeah. No, it’s great to have you guys on the show to talk about the company and what you guys are doing. But. You know, I always like to start the show sort of like at a ground level, so that everybody listening sort of gets an idea of what we’re talking about, and I almost want to start talking about, um, the problem that you guys are tackling. Right. So sort of just setting the stage. We have millions, maybe billions, depending on how you, you know, you diagnose someone, suffering from mental health or maybe some other neurological conditions, and a severe shortage of good ideas about how to treat those conditions. Now, maybe that’s just my opinion, but, you know, generally. Right? So can you guys paint a picture of the current situation from your perspective, and maybe that will lead to an explanation of how and why you were drawn to nature as a potential source of inspiration for new treatments.  

John Boghossian: Great. I’m happy to do that. Harry, just to to to start by saying thanks a lot again for having us on, on your show. We’ve been listening for the past couple of years and it’s a pleasure and an honor to be featured as well today. So thanks. Thanks a lot for that. And so to your question about neurological disorders, um, I think the the problem, I mean, just to to go back to what Pangea does, we develop transformative medicines for neurological disorders, and we take an AI driven approach. As my president of AI, Sona, and I will will try to convey to you today in terms of the problem that we are focused on. It’s large and growing. So if you look at neurological disorders, the unmet need is high, simply because there are very few, if any, disease modifying treatments for these disorders. If you think about it, it’s because the the biology of the brain. It’s the most complex, um, organ that we have in our, in our body. And so our level of understanding of the multifactorial aspects that drive disease and create symptoms of, disease is only nascent.  

John Boghossian: And so as a function of that, you know, we have, uh, tens of millions of people living with dementia. We have no approved disease modifying therapies for Parkinson’s. And more than one third of schizophrenia patients, for example, are treatment are resistant to all types of treatments. The other part that is, contributing to, increasing the problem, is that we are living longer. And with those, longer lifespans, the health span is not necessarily increasing commensurately. And so our brains are becoming kind of diseased and kind of symptoms of cognitive decline are kind of setting in. And because we are able to live, live longer through kind of medicinal approaches, we are we are seeing much, much, much more of that. And I think this this problem will continue to compound over time. And that’s it’s that unmet need that really inspired us to start Pangea, and also the previous companies that I’ve been involved with. Um, I don’t know, Sona, if you want to add anything to that.  

Sona Chandra: Yeah. No, I mean, I think Johnny did a great job sort of framing the problem and really the unmet need in the space of neurological conditions. Um, maybe I would just complement that with, um, a more optimistic look on the opportunity that we identified, um, which is really that we have reasons to believe that nature is really one of the richest untapped reservoirs of potential therapeutic compounds, specifically for CNS conditions. Um, and that’s really evidenced by things like something like 84% of the, um, small molecule therapeutics for CNS conditions on the market today are actually nature inspired. Um, so that means that they’re either kind of natural products themselves or derived from a core natural product scaffold. Um, but at the same time, actually a very small fraction of small molecules in nature have been effectively mapped and explored. Um, so more than 99% is is some of the figures that are, that are tossed around in the research. Um, and so that kind of got us thinking, well, what would actually happen if we could explore that 99% of unmapped natural compounds that are out there? And I think that really becomes the foundational thesis for, for what we’ve built at Pangea, um, where we are harnessing the power of data and AI to actually probe that 99%, and deliver a scalable discovery engine of novel therapeutic compounds for CNS conditions.  

Harry Glorikian: Yeah. I mean, I think we’re at that point where, and we keep moving in this direction, it’s not like it’s static, but technologies to measure things that we couldn’t measure, you know, 5 or 10 years ago, you know, and now you have the data and now you have. Systems that we’ve developed in AI that actually allow you to sort of do that analysis and glean unique insights. Let’s say, I’m not going to say you’re going to necessarily find the answer right off the bat, but insights that sort of lead you down a path. I think we’re at that point where — and this is funny because since we did the genome, I keep saying biology is at that inflection point and it’s moving up and to the right. And it just keeps, I think now in the next 3 to 5 years, it’s going to move like a rocket just because of we can now probe much more multifactorial views of a biological condition than we could even 2 or 3 years ago. But before we go down that that hole. What led you guys individually to Pangaea. Like what got you here? What was your personal path?  

John Boghossian: Yeah, great. I can I can get us started. Maybe, uh, since, uh, Pangea has been on my mind, uh, and very much fitting my days for the last three years. Um, so, um, you know, I’m, I’m co-founder and CEO, and before building this, this company, um, I was at another, uh, biotech company that I had joined as employee number one called Compass Pathways, um, based in the UK, now Nasdaq listed um, and is focused on mental health, uh, innovation. Um, and back then, you know, we were focused on psychedelic compounds, um, which are naturally occurring. And we had great successes actually with the lead asset there, which is currently in phase three clinical studies, and I led global operations there, which, you know, meant very different things in the times of pre-seed all the way to the times of IPO. You know, I set up several functions and grew them, um, before being able to hand them over to a subject matter specialist over the years when we had the means to, to hire them, which means that I had, you know, uh, I was able to touch a lot of the different parts of biotech company building and the process and was lucky to be part of a story that, you know, quickly was able to garner steam, you know, with, uh, um, uh, with the ability to raise large, large, large private rounds and then also rounds of, of funding on the public markets, um, from, from 2020 and at the time, um, post IPO, uh, to me, it became clear that I wanted to and I wanted to I wanted to continue to build companies to improve human health and diseases of the brain, specifically like mental health kind of disorders, where my, uh, initiation to this CNS space or like central nervous system space.  

John Boghossian: Uh, but I wanted to build another company that could occupy me, really for the, you know, the next 20 years. And together with the co-founders, uh, what we realized was we had been exposed to a couple of different companies that painted a picture of what was possible. One was Compass Pathways. I mentioned it. The other one was Atai Life Sciences, who my co-founder and CEO, Lars, had also co-founded, which was also focused on, um, uh, naturally occurring scaffolds as inspiration for mental health innovation. And we were also close to the team at another UK based company called GW Pharma, that had also gotten a couple of different compounds across the FDA approval finish line. And and we sat down and we thought about what was common about all of these projects and why had none of them failed that to date in the in the clinic? It was certainly not because we were you know, it was certainly I’ll speak for myself. It was certainly not because I was an amazing and, you know, outlier drug developer. 

John Boghossian: You know, I had been relatively new to the space in, uh, biotech time scale. Um, but it was, it was because we were starting from a place of knowledge. And that place of knowledge was, you know, came in various shapes and forms, but traditionally in, um, anecdotal human evidence of specific phenotypes that had been changing. And then, uh, the, the biotech companies being built around data generation to verify those, those, those claims and put them in the construct of, you know, an FDA approvable data package. And so that was the idea that germinated in our minds, that led us to start to start Pangea, uh, to to build a scalable engine that, which Sona mentioned, as an AI driven approach or machine learning driven approach, I should say, um, to to be able to effectively mine this data set and scale the approach of, uh, finding novel therapeutics from nature. So it gave me a lot of excitement to be able to scale the approach to many more scaffolds. And also, to be honest, the prospects of being able to build another company with the same set of people that had been working successfully in the past was also very exciting. Um, and so we kept the band together and went back at it, and here we are.  

Sona Chandra: Yeah, maybe. Maybe. No thanks Harry, and maybe I can also just add to that. So, um, in terms of sort of my, my personal story or venture into this space, um, a lot of that comes from kind of my, my personal history growing up, um, where I grew up in an Indian family, um, but in the US, um, and had sort of a long history of exposure to herbal and natural remedies growing up, um, things like ashwagandha, you know, I had sort of, um, palpable or noticeable differences in my cognitive health. Um, started taking berberine. Another, another, um, very powerful natural ingredient, um, for many years and also noticed some, some meaningful kind of palpable differences in, in, in my blood glucose levels, and neuro health as well. Um, and all that kind of got me thinking. Well, you know, in my previous life in Big Pharma, I became intimately exposed to and familiar with the problem of failure in the clinic. Um, that kind of lives across the pharmaceutical industry. Um, and I sort of wondered, well, we have this kind of plethora of herbal and natural remedies that we sort of know works in human biology based on centuries of, of medicinal wisdom. And on the other hand, we have a huge kind of problem of failure when it comes to, um, translation of our of our pharmaceutical products into human biology in the clinic. Um, what if we could actually unlock the chemistry that’s found in these natural remedies, in these plant based medicines and bring that into pharmaceutical CNS drug discovery? Um, and so bridging those two pieces of the puzzle is ultimately what kind of got me thinking about this concept and seeing if we could actually apply technology and things like data and AI to unlock the potential of these medicinal plants and drive the development of novel therapeutics that have a higher chance of translation into the clinic, because they already have that history of medicinal use and efficacy.  

Harry Glorikian: So I know you guys, like, now, have this incredibly strong belief in ethnobiology as sort of a neglected field of science. Right? And I was looking at the website and it says, you know, we’ve only mapped less than 1%. And so I guess one of my first questions is. How’d you guys come up with the 1%? Because I’m not. I’m not sure we know the complete, uh, library to to come up with that number. So I’m just curious if you guys have more insight into that. And then I guess just to keep going down that path is, do you think there’s a built in bias in the medical and scientific establishments against sort of, I don’t know, indigenous cures or against the world of plant-based medicines that would have a more, uh, neurological effect on people.  

Sona Chandra: Yeah. It’s a great question, Harry. Maybe I can just quickly take the 1% piece. Um, so this is a number that is, is sort of, um, in some ways kind of, you know, to your point, a bit vague. Um, but I think on purpose it’s vague because the reality is we simply don’t know the total number of, of natural small molecules that exist in the universe. Um, but we do know that we we have only probed a very small fraction of that. Um, and so, you know, how kind of ethnobotanists and people working in the natural product space sort of think about that 1% number is, you know, we have hundreds of thousands of of plants and other kingdoms of life. Um, and we have decoded roughly a few hundred thousand of those natural small molecules. So the number is also very variable depending on exactly how you define that. Um, but let’s say roughly 500 to 600,000 are sort of established small molecules. And so if you just kind of take that as a portion of the potential, um, number of small molecules that exist out there in, in the hundreds of thousands of life forms, um, that really only covers a very small percentage. And so 1% actually could be, um, overestimation. In fact, it could be much less than 1% of small molecules that, that we’ve, we’ve managed to decode.  

John Boghossian: Yeah. I mean, the point being that that that most of the space is white. Um, and so and so that’s, that’s that’s what drives our deep belief as you, as you rightfully called it that, Harry. And then to your question about concern from pharma or pharma having gone away from traditional from, uh, natural products based CNS drug discovery. I think there’s some truth to that. But there’s also, um, I think some evidence to the contrary. So if you go back to history, you know, until the 1970s and 80s, most of the way in which we made drugs, uh, was actually very much based on nature, you know, we would we would observe something through trial and error and then through medicinal chemistry and, you know, other, other approaches, we would eventually figure out what is the bioactive moiety that is responsible for the pharmacological action of interest, um, in, in humans. And then from that chemical structure, we would then construct back all these experiments to then prove that in animals and then later in clinical trials. So that’s really what what leads to the 84% number that Sona that Sona mentioned, which really drives most of the therapeutics that we have at our disposal today, you know, both clinicians and patients alike. I think in more recent decades, what has happened is pharma has moved away from natural products based CNS drug discovery, you know, to focus on, um, uh, other technologies, you know, using high-throughput screens and with the with the ability to basically engineer, uh, specific compounds, um, and then trying them at a very high volume in these high throughput screening approaches.  

John Boghossian: The idea was that we didn’t need to necessarily focus on natural chemistry as like a starting point. And we could really explore are these all these other parts. And I think what that has generated, you know, a lot of novel therapeutics. And I think there are great kind of technologies that we can cite to this effect that we are using today. Um, what that has also created is a lot of different, uh, ballooning of costs and timeline in order to get there. Um, and so I think that’s on the one hand, I think on the other hand, uh, what is important to still mention is that there are still major contributors to novel approvals that are actually originally inspired by nature. It’s just that pharma doesn’t package it in that way. So if you look at blockbuster products of the last, you know, 15, 20 years, there are a really large number that actually can be traced back to an original discovery of a particular natural compound that had an effect. And then researchers may be optimizing on that compound, and eventually that IP being bought by a small biotech that, you know, took it to phase one or phase two, and then being bought by a larger biotech and eventually ending up, you know, on the shelves of pharma, you know, for these late stage trials.  

John Boghossian: And by then, you know, they’re branded in a certain way, and they’re certainly not necessarily a, you know, big focus on their natural origins in the branding. But they might, you know, but they might very much come, come, come from there. And I think some examples that I’ll just briefly mention are, you know, a multi-billion dollar franchise that Pfizer has around a drug called Lyrica, used for epilepsy, pain and anxiety. Botox, uh, which is a huge franchise for AbbVie, you know, around chronic migraine, but also spasticity etc. is also derivative of a of a, of a natural of a natural compound. So there’s actually a lot more than, than than meets the eye really. Um, so our, our job is to is, is to get the next generation of compounds. And I think you did mention something I don’t mean to jump around, but you you ended the section prior to us introducing each other or ourselves and, and talking about why why now? Right? I mean, why are we starting this company now? And I think it’s perhaps worth saying a couple of things about, uh, about that, because I think in previous companies that we co-founded, you know, that that question of, you know, why not five years ago? Why not in five years, um, is really at the, at the core of, of, um, of the thesis for the company.  

John Boghossian: Maybe I’ll talk about the biology side of it and then maybe Sona, You can also touch upon the technological side of it. Um, and really, you know, our understanding of the brain has really increased to a degree that it has crossed a threshold where we we have a deeper understanding of neurobiology and specific targets that that could be helpful. Uh, we now also have biomarkers. And through neuroimaging, we can actually test the effect of various molecules in the brain prior to waiting for these later stages of, uh, development. And we have a really exciting science called, for example, the organoids kind of approach whereby we can even grow mini brains, um, by taking blood samples and, you know, like differentiating stem cells into specific organ types. And then we can test, uh, specific compounds into these petri dishes, uh, with minor renditions of the brain, um, to get insights prior to really like, trying them in, in humans. So I think those are some of the biological innovations and advancements that have enabled our company to be, uh, really cost effective in its kind of development right now. So, Sona, maybe on the tech side as well.  

Sona Chandra: That’s right. Um, totally agree. And I think that on the tech side, there are a number of advancements that have really allowed us to, um, apply technology to this problem. Um, so think, you know, really we reached a time point where there’s an explosion of data, um, the cost and efficiency of actually generating and making sense of large quantities of life sciences data has progressed significantly. Um, and, um, there are new kind of modeling approaches that have continued to grow at a breakneck pace that allow us to actually retrieve and process that data at an unprecedented scale as well. Um, so things like large language models and generative AI, um, really allow us to actually make sense of, of this data and really combine, um, heterogeneous sort of biomedical data sets to actually generate insights at an unprecedented scale.  

Harry Glorikian: So continuing sort of down this technology theme because I think I know you guys are naking therapeutics. But you’ve had to create some very specific technologies to help you go down this road. And I was sort of reading about a few that you guys have. I mean, one is you guys have a knowledge graph called Sage. Um, a second one is, uh, a computational metabolomics component called Kava. And then a compound activity profiling product called Lila. And I, if there’s others. I’m sorry I missed it in my homework, but, I don’t know if you can walk us through, what what do these three components do? You know what sort of knowledge graph are you trying to build? What are you learning from the metabolomics. I mean, you know, and I can keep going with some questions, but that sort of gives you an idea of the where I’m trying to figure, you know, dig in a little bit here to give people an idea of what you guys are doing and what you guys are seeing from it.  

Sona Chandra: Sure, sure. Happy to give an overview on that. Um, so maybe giving giving a high level sort of review. Um, we have developed an AI platform that aims to accelerate the discovery of novel natural products from plants and other kingdoms of life. Um, and these three sort of modules, if you will, that you described, um, create that kind of scalable discovery workflow that allows us to move very effectively from plan to hit compound. Um, so now double clicking on each of those. Um, so the first piece is Sage, which is our knowledge graph is really a graph database where we’ve integrated data across four major domains. Species, their chemical composition and their impact on disease biology as represented by targets and how those targets relate to indications. And that knowledge graph is embedded with AI, including graph neural networks, to actually predict links between any two domains of data. Um, so we can query the graph and we can say, hey, for a given set of plants or compounds, what could be relevant targets or diseases. Or we can go the other way around and we can say for a given set of targets and diseases, what could be relevant plants and compounds. Um, and so that knowledge graph is really kind of the heart and soul of the platform that allows us to generate very meaningful hypotheses to direct our discovery programs. And we continue to enrich and sort of grow the data within our knowledge graph, um, through a variety of streams of data, including structured data. Um, we apply large language models to actually extract and structure large quantities of data from unstructured sources. And we actually generate experimental data prospectively to enrich and grow underlying data streams. Um, so that’s the first piece is really our knowledge graph.  

Sona Chandra: The second component that you discussed there is is our Kava platform, um, where we’ve developed, um, a portfolio of modeling approaches to actually translate the language of mass spectra into structures. Um, so historically, trying to understand the chemical composition of a complex biological mixture or a plant in this case is something that was very challenging and really elusive to scientists or analytical chemists. It requires a lot of sophisticated expertise. Instrumentation can take weeks, months, you know, up to a year even, um, to actually sort of deduce the structures or the, you know, the chemical composition of something like a plant. Um, and so we’ve actually set out to address that problem by applying machine learning techniques, where we’ve developed, um, models that can actually take untargeted metabolomics data. Um, so data representing complete small molecule chemical composition of a plant or another life form, and actually translate that into chemical structures based on the fragmentation patterns of the input spectra. Um, and that allows us to very effectively and very efficiently decode the chemical composition of a given plant.  

Sona Chandra: And the final piece is actually the Lila platform that you mentioned, um, where we can actually we’ve developed models, um, leveraging cheminformatics based approaches to actually predict the activity of those compounds in the human body. Um, so we’ve developed models that can actually predict on- and off-target effects given a chemical structure, as well as our implementing models to predict neurobiologically relevant endpoints. Um, things like neurite length that really allow us to deduce the potential of a given compound when it comes to neurobiological activity. Um, and this really creates this very scalable framework, um, to move from plant. So nominating and selecting a potential plant, um, to a potential compound, um, to a, to a hit molecule. Um, that we can then carry forward into preclinical development as a part of our pipeline development efforts.  

John Boghossian: So I wanted to, uh, to maybe add one, one thing to this very eloquent overview, um, about our platform, which is actually a bit that always makes what makes me laugh and get really, really excited, which is the the middle piece that Sona mentioned. So Kava, our plant metabolomics platform because I think it’s, uh, it’s a little bit, a little bit of a nebulous concept for, you know, uh, most, most people, it certainly was for me, you know, a few few years ago. But the way that, uh, that this actually works, which I think is maybe interesting for your listeners, is that, you know, you will traditionally, you know, it would take, you know, a few months, even, like sometimes years to really figure out what is the, the, the, the compound responsible for the activity of a given plant. And many a PhD thesis would have been written in order to figure that out. And now we can figure it out in minutes with these metabolomics algorithms. And the way that it works is you take a plant sample, you liquefy it, and then you shoot protons at it. And those protons will then predictably break up the molecules at specific bonds that tend to break with specific activation energies. And then those, those molecules or molecular fragments will then, um, drop and create a peak on a 2D graph, which is called the mass spectrogram. And these are the 2D spaces of data that are then fed into the algorithms, some of which we have developed, which will then, with a 70 to 80% accuracy, give you the 3D molecular structure, you know, by way of a SMILS, Um, for for those like chemistry friends of ours, um, that that that are likely to be present in that mixture. And so that’s the that’s the really exciting bit um, of the, of the platform that was just, you know, like simply not possible, you know, a few years ago.  

Harry Glorikian: And so I’m laughing. I’m laughing because like when I was at Applied Biosystems, the facility I was in here in, uh, Framingham was the mass spec facility. So I was living I was I was surrounded by time of flight and everything else that was all around me all the time. I think I was one of the few genomic people in that space. So no matter what, you were constantly interacting, talking about these spectra, what could we see? What could we do? How do we make it better? Uh, so I’m sorry if I’m, I’m chuckling in the background, but, uh, it’s something that, that you, you know, you realize as you’re going through your career, all these things you get exposed to at some point start to add up over time.  

Sona Chandra: Yeah. Yeah. No, totally. I mean, so then I mean, you you must be then intimately familiar with some of the challenges when it comes to actually analyzing that data. Um, it really is, is quite an intensive process that requires a lot of sort of expertise to sort of make sense of those fragmentation patterns and sort of translate that into chemical structures. Um, and what we found was that when we looked at the history of natural product research, um, as we mentioned before, this was the main way that pharma did CNS drug discovery for many years. And a lot of the reason that pharma kind of moved away from natural products, even though it is a very promising source of potential drugs, was the challenges and actually the, the analytical chemistry side of things. Um, so what they found was that, um, duplication or the rediscovery of the same compounds over and over again, um, is something that was was very limiting and frustrating for researchers. So they would take this they would go through this very lengthy process of bioactivity guided fractionation. They would take a plant, they would have to physically fractionate and test that plan, um, in various, um, sort of assays in order to finally kind of land on the part of the plant that they believed was truly driving the activity.  

Sona Chandra: That process alone could take months. And by the way, you’re limited only to testing things, one use case at a time because of that sort of experimental limitation. Um, and then if you were sort of lucky enough to sort of identify the part of the plant that was driving its activity, um, you would then sort of try to characterize the compound and what happened. There was a lot of the times, um, researchers would realize that they had spent all this time and money actually rediscovering a compound that had already been previously known about. Um, and so with our kind of computational approach, we can very quickly and efficiently kind of, um, de replicate all of the known chemistry so we can match, um, all of the fragmentation patterns to the, the structures that they, they reflect. And we can narrow in on the part of the, the plant that is truly representing novel chemistry. Um, so we can very effectively sort of streamline our resources and create a very efficient discovery approach. 

Harry Glorikian: So I’m going to go out on a limb here and say, is that what you would identify as the special sauce of the company? I mean, just thinking about. What is unique and defensible about your approach to CNS drug discovery.  

Sona Chandra: Yeah, maybe. Maybe I can just take the first answer at that and John can fill in. Um, so yes, I would say the computational metabolomics platform is one, one element of our secret sauce. Um, so our ability to really effectively decode the chemical composition of the, of the, of the plant and then apply our compound profiling models to identify their potential activities. Um, so that’s one angle. Um, the other angle that I would want to highlight, um, is our ethnobiology approach that you kind of alluded to earlier. Um, where, you know, a lot of what we do is based on the historical or medicinal use of a plant. Um, so that we’re really optimizing our chances for translation into the clinic by, by really ultimately decoding a chemical structure that’s in a biological mixture that already has that signal of safety and efficacy in humans. Um, the key point I want to highlight there is that, um, really most of that knowledge today lives in very disparate sources. So books, peer reviewed journal articles, field notes, etc. it’s all over the place. No one has really ever effectively gone and brought together that knowledge into a scalable, queryable search engine form. Um, and so that’s one of the kind of core sort of ambitions of our, of our efforts is to actually apply things like large language models to go out and structure and retrieve that knowledge from this disparate source material and bring that together into our knowledge graph so that we can apply data driven and AI enabled approaches to selecting plants based on their potential for therapeutic relevance. I don’t know, John, if you wanted to, to add on to that at all.  

John Boghossian: Great summary, Sona.  

[musical interlude]  

Harry Glorikian: Let’s pause the conversation for a minute to talk about one small but important thing you can do, to help keep the podcast going. And that’s leave a rating and a review for the show on Apple Podcasts.  

All you have to do is open the Apple Podcasts app on your smartphone, search for The Harry Glorikian Show, and scroll down to the Ratings & Reviews section. Tap the stars to rate the show, and then tap the link that says Write a Review to leave your comments.  

It’ll only take a minute, but you’ll be doing a lot to help other listeners discover the show.  

And one more thing. If you the interviews we do here on the show I know you’ll my new book, The Future You: How Artificial Intelligence Can Help You Get Healthier, Stress Less, and Live Longer.   

It’s a friendly and accessible tour of all the ways today’s information technologies are helping us diagnose diseases faster, treat them more precisely, and create personalized diet and exercise programs to prevent them in the first place.  

The book is now available in print and ebook formats. Just go to Amazon or Barnes & Noble and search for The Future You by Harry Glorikian.  

And now, back to the show.  

[musical interlude]  

Harry Glorikian: Let’s jump now to. Some of the compounds you guys are working on and I, I found, you know, uh, something that goes by the code name of OT003. And then, John, before we even press record, you said there’s another one called KH001, I guess. You know. How did you guys discover it? How does the discovery process illustrate, uh, say, the nature inspired CNS drug discovery strategy we’ve been talking about. And what neurological disorders do you guys hope that these compounds will actually be effective on?  

John Boghossian: Right. Um. Thanks, Harry. So let me maybe set the stage. So when we started the company, um, it was very important to us to have early proof points of the thesis, and we didn’t want to, you know, create this platform tech and just, you know, be a platform company for quite a while, you know, until we were, uh, let’s say, mature enough to, to start to wet our feet and, and try to edge towards the clinic. So from, from the beginning stages, we thought we would kind of pursue both both strategies at once. One was to build the tech, um, which Sona and her team have been really laser focused on, um, over the past year and has recently been launched and also now has gotten the first signs of validation, which we’re very excited about. And then simultaneously, we had a couple of ideas that we that we had, uh, committed to really researching, let’s say, in the manual traditional way to prove out the thesis. And some of these threads created what today are, uh, you know, two flagship programs and assets, uh, one around OT003 and the other around K001. 

John Boghossian: So for OT003, um, the way this, this, this came about is we were looking at, uh, neuroplasticity, which is, uh, for those who don’t know, the ability of the brain to, uh, regrow itself. Um, and neuroplasticity had been, you know, a central theme of our research focus in past years, also in past companies. And we were looking for the lowest common denominator to a number of different neurological disorders where neuroplasticity played a big role. And it turned out that there was this uh receptor called TRK-B or uh tyrosine receptor kinase B, uh, which, which seemed to be impaired in terms of its signaling in um, in a range of these neurological disorders. And so we set out to figure out if there was a small molecule in nature that would, uh, modulate this receptor. Um, and we found and we found such a molecular class which had some convergent evidence. And, you know, that’s maybe to speak to a little bit of a of our approach is in general, we look for convergence of evidence, you know, like given the data quality and a number of these kind of domains is variable. The idea is to get conversions between various threads. So maybe various regions around the world where specific plants and bioactive molecules have been used for the same purpose. And if we can triangulate that, then it gives us higher confidence that we may be right. And that’s what we found effectively here, which was there was a proprietary whether there was a class of molecules that had been used in traditional Chinese medicine, but also in, uh, in other, uh, traditional medicine, um, around the world, um, which had a compound, uh, which was the naturally occurring version of it that had been responsible for this phenotypical effect.  

John Boghossian: And what we then did is we found, you know, a researcher actually, at Emory University in the US that had had the same insight, um, and had optimized on this molecular scaffold for, you know, about a decade or so and we, you know, licensed that IP from the university. And that became the, the Nexus for OT003, which is a proprietary small molecule of ours, um, that we that we have now generated, you know, a number of different in vivo data sets. So data sets in animals that, uh, make us believe that we have, you know, what’s called a pipeline and a pill. Uh, we have generated really encouraging results across a number of domains of neuroplasticity. And so illnesses like schizophrenia, Parkinson’s, Alzheimer’s, um, and so that gives us a lot of confidence that this, this compound will be able to help reverse, um, some aspects of neurological decline in a number of these, uh, indications or kind of conditions. And we’re now one year away from human trials with that compound. So that’s, so that’s, so that’s one example.  

John Boghossian: And then the second example, uh, which gave us K001 was this idea of, uh, improving mood disorders. Um, in a quick way. So we had been looking uh, we were very we were intimately familiar with, uh, anxiolytic treatments and antidepressant treatments that took multiple weeks in order to, to work. The traditional antidepressants that you’ll be familiar with will have those qualities. And we were interested in something that had a very rapid acting effect. And so we actually found in nature a class of compounds that are originating in a South African succulent plant, uh, called kanna, that has this effect. Um, and when we were looking at the various individual, uh, compounds within the plant with one of them, uh, we noticed a really interesting side effect, which is that it had the unintended, if you will, side effect of, uh, delaying male climax. Um, and so we, you know, we, we then spun out a different entity because that’s a little bit of a different focus from the, from the traditional neurological kind of disorder focus of Pangea. We spun that out in into a different effort with external co-founders on the scientific side, um, and to focus on premature ejaculation, which is a condition that impacts up to 1 in 5 men and their partners and has no FDA approved treatment. And so those are two examples that came through, let’s say, more traditional discovery processes, but very much in the realm of natural products, which we are now hoping to to scale, um, with the AI platform.  

Harry Glorikian: So. Why CNS? I mean. The relationship between, say, plants, fungi, animal cells at the molecular and protein level. You know, I mean, they definitely apply to many, many more types of, of diseases or as is neurology, say, just to start, for you guys, that sort of leads to other drug programs down the road.  

Sona Chandra: I think that, um, the answer to your question as to why CNS is, is really a combination of two things. One is the unmet need that we recognize in the space that we touched on earlier, where, um, there’s a significant increase in the prevalence of these conditions tied to the aging population, etc., and a lack of effective disease-modifying therapies. Um, so that’s one piece. Our team also has significant expertise in actually, um, discovering and developing such drugs. And so we felt that’s where our expertise was best placed. Um, and on the other hand, there is a significant untapped opportunity when it comes to neuroactive compounds in nature. Um, so we mentioned that 84% number. But even when we look at kind of ethnobiology and look at the data, um, in terms of sort of the number of plants and fungi that actually have an established relationship to neuro activity and cognitive health, we see that it’s one of the highest groups of indications that tend to have that relationship. Um, so we recognize that there was significant opportunity, unmet need, and sort of a marriage between where we think that we can add value.  

John Boghossian: One thing that I wanted to to to add on this point, because I think it helps really bring the point home, I think it it also makes sense intuitively. Right. So humans have been able to very easily correlate ingesting a particular plant or or other natural molecule and the impact that it has on their mind state. Um, and if you think about it, it’s much easier to correlate those two things and write about it, therefore creating a lot of a lot of evidence that we can mine then say, you know, taking something and then expecting that it might have an impact on your tumor, for example, and kind of decreasing the size of that. And so the ability to triangulate or to correlate that is a bit a little bit lower in other in other therapeutic areas. And so that’s, that’s why, you know, there is a wealth of, of data that, that we can mine kind of starting there. Um, and, and to your point, Harry, uh, CNS is just the starting point. So the idea is that the platform can indeed be, uh, like, deployed to other, uh, therapeutic areas, you know, for some of the, the stumble kind of discoveries that we’ll be making, you know, we might partner those out with a third party in the short terme, but in the longer terme there are, you know, there are a number of different opportunities that we could also, um, kind of like delve into, um, as we kind of mature and, and get a few first, you know, kind of proof points to the thesis in the CNS space.  

Sona Chandra: Final concluding thought there was that, to John’s point, you know, our platform is built to be disease-agnostically. And we were already in conversations with a number of external partners on, on ways we can leverage our platform to actually deliver candidates across the board and other indication areas.  

Harry Glorikian: Which slips me right into business model. Right. Which is. What is the business model that you guys see? For Pangaea, is it, you take the compounds you discover to market yourself. Um, you mentioned, John, you guys set up a separate company for one of the compounds that you find, you found is, you know, is that the model to spin these out into different entities? Or is it to outlicense these drug leads to larger pharmaceutical companies? Or maybe it’s all of the above. I don’t know, I’m just wondering what your you know, at some point there, the business model has got to add up to to something viable.  

John Boghossian: Fair enough. Uh, and good question. So I think it is indeed a combination of our own development and maybe kind of co-discovery and, uh, kind of research collaborations with third parties. Um, it is definitely not the intent to necessarily, uh, you know, create different legal entities for individual compounds or programs. Um, you know, the, the history of how, you know, K001 kind of like came up. And the fact that it is it is a program that is bucketed in like a different, uh, therapeutic area, like within the pharma space led us to, to do that. And we also had external co-founders for that kind of project. But everything otherwise, you know, lives within Pangea. And the way that we think about our business model is, you know, on the one hand, we have some discoveries that we are really excited about, and the idea is to drive them through the first stages of trials and such that we could get them to a level of proof point that it would become, you know, attractive to either go all the way to commercialization or to partnering with pharma companies on these on these programs and the more traditional, you know, pharma business model, um, that that we’ve come to, to know the other business model that we are exploring is to use the platform as a starting point for specific milestone based, uh, deals with third parties.  

John Boghossian: So Sona has started to allude to the fact that we are in discussion with several potential partners that have say, you know, a particular disease of interest or a target of interest. And that might be then the starting point for our conversation with Sage, uh, to Sona’s, expose earlier to then, you know, go into the platform and, and try to basically, you know, get specific molecules and molecular classes that could act at that target of interest for the third party. And then the way that you would set these out is in a milestone based, uh, kind of fashion. So, you know, if we were successful at a specific stage, we would get, you know, a certain level of, of payment. And then if they were to progress, you would get further payments, but those molecules would then, you know, leave leave the Pangea building and kind of get developed by third parties. And we would then get, you know, some of the economics upside as a, as, as a result of our contribution. So I think we are, you know, interested in both. You know, I think it is it is important to have our own pipeline and be able to showcase some proof points and then also to be able to to have the second kind of business model, to really scale that to more partners.  

Harry Glorikian: So as I was looking through the website, it said that you guys are in the process of applying to be a B Corp. And so for those people who are listening, you know, when companies do that, it’s it’s a way of saying to investors that you’re you’re going to be guided by social benefit, not solely by, say, profit motive or increasing shareholder value, although I’m sure that has got to play a role in it somewhere. Um, as a shareholder of many companies like you do want to see a benefit to to what you’re doing, but what is being a B Corp mean specifically to you guys? I mean, what role, um, do social and environmental impact play in Pangaea’s mission, you know? Take me down this road of why you guys made that decision.  

John Boghossian: So maybe I can get us started on that one, because I think it was it was front and center when we started the company, and it’s still very close to to our hearts. So I think for, for us, what this basically means is reciprocity and sharing benefits with, uh, traditional knowledge holders. And what I mean by that is that, um, you know, as opposed to research teams that are starting from scratch and are making discoveries in a vacuum, uh, say in a lab at Pfizer, you know, and in our case, we are, you know, so-called building on the shoulders of giants, as in, you know, on the collective knowledge that has been amassed over the past millennia, to what Sona have described of this knowledge that is, you know, sparsely distributed around the world. And the idea is to involve those those people that put this knowledge together in the economic process that develops drugs out of that knowledge. And so what we want to do and what we have been able to do so far is to, uh, is to set up a partnerships that are a little bit like, you know, licensing intellectual property from a university and having some sort of, you know, royalty type deal with the university whereby they would get some sort of upside from the future sales of a of a therapeutic here is the same, but you have to identify specific groups that came up with the original knowledge that you’re building on top of.  

John Boghossian: So for example, for this first, uh, program at um, our affiliated company, Kanna Health, um, K001 has been um, come, come about as an optimized version of this naturally occurring, you know, like succulent in modern day South Africa. And there is a group of human tribes, you know, including the San and the Khoi, uh, which have traditionally used the kanna plant for medicinal purposes. And so what we’ve been able to do is to set up a benefit sharing agreement with those groups whereby they would get some, some part of the upside, you know, a bit like, you know, like they were our source of IP, if you will. Um, and that allows us to have reciprocity with these other groups and to share the value of the innovation that of course, we then take forward into, into the drug development, uh, process, but originate in insights that, that, that had come across through. Initial discoveries. And that’s why, you know, being a B Corp, um, you know, for us, you know, was a bit of like a rendition of, you know, that is something which is, you know, core to our DNA and the way in which we want to do CNS drug discovery.  

Sona Chandra: To quickly add, add a statement to that. So, um. Yes, that, you know, the benefits sharing piece, um, is, is very crucial for our discovery efforts and a lot of what we do. Um, and I think part of our unique approach, um, is that we’ve, we are implementing an ecosystem of partners globally in order to actually access unique breadth of biodiversity. Um, and part of that kind of effort involves actually working with local research hubs and communities to actually probe the local biodiversity, while also building up local research capacity and creating long tum economic wealth in the region so that there is this kind of combined or collaborative incentive to make these discoveries and actually bring those forward, um, as therapeutics.  

Harry Glorikian: So, sort of, bringing this back, sort of, you know, circling back in a sense, assuming which I’m sure it will happen. [Let’s say] Pangea is a huge success, right? How will drug development, do you think you guys will have an impact on how drug development will look five or 10 years from now. Um. Do you think that your approach will help open people’s minds to different pathways on how to bring these things forward or identify them. How do you guys look at that? 

Sona Chandra: Nrro, absolutely. I think I think that, um, we we have a sort of pretty significant grand vision, um, for, for our company in the 5 to 10 year horizon. Um, my personal view on that is that, um, you know, we have a significant, um, sort of reservoir of potential therapeutic compounds in nature. Um, and by applying these approaches, I think that there are hundreds, maybe even thousands of potential drugs that can be discovered this way by actually probing, systematically probing various medicinal plants and other life forms, um, to actually unlock potential therapeutic compounds. And that’s, you know, within the world of CNS, but also outside of it. Um, and also even outside of the world of therapeutics. Um, so, you know, in my view, I think that there’s a significant opportunity for us to, um, actually leverage the platform that we’ve built and the expertise that we will have built in our ecosystem of partners to generate new discoveries, um, within the world of pharma, but also outside of it. So areas like consumer health, nutraceuticals, food ingredients, um, dermocosmetics, maybe even, um, materials and agriculture. Um, so actually creating that kind of category leader in the world of, of natural products and linking that to human biology.  

John Boghossian: Yeah. And in that vein, we are we are not alone, right? We we have, you know, like several peers that are focused on these like various types of, of verticals. And I think there is this emergence of a new class of, uh, call them tech bio kind of companies that are really focused on this general thesis. And I think we have only scratched the surface, to Sona’s point. Um, and I, you know, I think there is really room for, um, for lots of new discoveries. I think, you know, if I look at. The biotech time scales, right? It usually takes about ten years to take a drug all the way to commercialization. And so I think if I, if I want to be realistic and look at the next kind of ten years, if we can, uh, if we can have contributed, at least, you know, a couple to a handful of such drugs, uh, to the arsenal of clinicians. I think that, you know, that will be mission accomplished for me. Um, and if and if along the way, we can enable a lot of others to advance in their own, in their own efforts, then that would be even better.  

Harry Glorikian: Well, listen, it was great having you guys on the show. I mean, I love these, obviously, I mean, you guys know the show, so I love these discussions, right? Um, and, uh, you know, I mean, I always tell people like, I’m not getting any younger, so the more success you guys have, the more I’m gonna personally potentially benefit from these things. So I wish you, you know, the greatest success in in your endeavors.  

Sona Chandra: Thanks Harry.  

John Boghossian: Thank you so much. Thanks, Harry, and thanks a lot for, you know, taking the time to, to speak to, uh, to people hard at work and making, you know, good, good uses of data and to augment human health. Um, it’s been a pleasure to be on the show, um, and looking forward to the next few years of discovery for us all.  

Harry Glorikian: That’s it for this week’s episode.   

You can find a full transcript of this episode as well as the full archive of episodes of The Harry Glorikian Show and MoneyBall Medicine at our website.   

Just go to and click on the tab Podcasts.  

I’d to thank our listeners for boosting The Harry Glorikian Show into the top two and a half percent of global podcasts.  

To make sure you’ll never miss an episode, just open Apple Podcasts or your favorite podcast player and hit follow or subscribe.   

Don’t forget to leave us a rating and review on Apple Podcasts.   

And we always love to hear from listeners on Twitter, where you can find me at hglorikian. 

Thanks for listening, stay healthy, and be sure to tune in two weeks from now for our next interview.