Forecast 2050 Episode 10 - Viswa Colluru & Pablo Lubroth - Transcript

Enveda Founder & CEO Viswa Colluru believes we have been looking for new medicines in the wrong places for decades.

In the Season 1 finale, I sat down with Viswa and Pablo Lubroth, VP of Special Projects at Enveda, to discuss why 9 in 10 drugs fail in clinical trials despite Big Pharma spending $276 billion in annual R&D, how AI is finally unlocking 99% of nature’s chemistry that we’ve haven’t been able to read, why the incentive structure of biotech capital markets is actively bad for patients, why insisting on understanding how a drug works actually slows down drug discovery, why aging may not be inevitable, and why mental health is going to be the defining medical crisis of the next 25 years.

Watch the full conversation here:

Transcript:

The Quintessential Problem With Drug Discovery

Finn Murphy: Guys, thanks so much for joining us here on Episode 10 of Forecast 2050. I’m joined today by Viswa Colluru, CEO of Enveda, and VP of Special Projects, Pablo Lubroth, also co-founder of Decoding Bio newsletter. I suppose the topic to talk about now for 2050 is we’ve covered demographics, we’ve covered physics, we’ve covered longevity and fertility. We haven’t talked at all about medicine. Today, only 1 in 10 drugs that we take into trials ends up being approved. Why are we so bad at understanding which drugs work and which drugs do not in this whole drug development process?

Viswa Colluru: You know, that is the quintessential problem. And it was precisely that insight that was the beginning of the journey that led to Enveda. If the core problem is that things that work in the lab don’t work in people—which is why, you know, we spend millions of dollars and spend years trying to identify effectively a hypothesis, right? A chemical and a biological hypothesis rolled into one—let it into the vital, into the complexity of the real human body in this messy world that we live in, and then check to see if it mediates the effects that it so consistently and promisingly did in the lab. And it doesn’t work.

Then a very tempting, I think, modern fragment of behavior is to say, well, let me just understand the biology better. But Enveda was born out of the realization that we had drugs that still exist today that are highly successful, that touch 1 in 3 people in the world on average, every single day, before we had molecular biology as a field and any biology labs. We got our first medicines when our ancestors walked the Silk Road and decided that they should drink the milk of the poppy because it helps their fellow travelers walk longer and sleep better at night. That gave us morphine, taught us opioid biology, and despite their shortcomings, are our only mainstay for breakaway pain today.

That made me think: wait, if everybody is really focused on trying to deconstruct biology in all its glorious complexity in a lab system with plastic dishes, is there a better way to think about it where we say, what is the best chemistry that will ultimately yield a drug? Not only is this idea intuitive because our ancestors used this chemistry, evolution has selected a very small pocket of possible chemistry to create out of a very large set of exploration space that it could have. But it’s also validated in that half of all FDA-approved medicines come from molecules that first started as a metabolite in a living system.

But last but not least, we’ve barely scratched the surface. By some estimates, we know 10% or less of human blood when it comes to chemistry. About a fourth of the chemical composition of a tomato, even though 2 billion people will eat one today. And less than a fraction of a percent of the chemistry of the living world total.

So we had this beautiful, intuitive, contrarian idea that was simultaneously validated and untapped. And that’s exactly why Enveda was founded to say: can we make that ratio better? But by asking the question slightly differently, because at the end of the day, the molecule is a chemical that’s ultimately a drug. And the idea that choosing better chemistry will ultimately yield better drugs was as simple as it was radical and as exciting as it was untouched.

Pablo’s Journey to Enveda

Murphy: Pablo, you were an investor at Hummingbird prior to joining Enveda. You’ve chosen to take the rare investor jump into a company you work with. What was it about the chemistry approach, and how did your perspective on the world lead to why this was the thing to spend time on?

Pablo Lubroth: Yeah, I think it was a couple of things. So first of all, it was meeting Viswa and the team and really having seen how much they had achieved in such a short period of time, and actually having discovered so many potential drug-like molecules that had an effect in animal biology. And that was really rare. This was 2021, and with a few million dollars, they had already achieved what companies couldn’t have achieved with tens of millions. So that was extremely impressive.

In terms of the biology, I think, you know, looking back—I studied pharmacology, neuroscience. I was fascinated with psychedelics and their impact on mental health. And seeing how in the current paradigm, we would never have really developed a psychedelic. It was completely contrary to what we knew about the human brain. We wouldn’t have been able to come to the conclusion that psychedelic was the answer to major depression. And Viswa deeply understood that there’s so much wisdom, to put it metaphorically, in natural molecules. And being able to bring that back with a team that was exceptionally capable at executing that mission—there was no other option, basically, for me to join.

Insisting on Understanding Biological Mechanisms Slows Down Drug Discovery

Colluru: And, you know, Pablo touched on something that we believe deeply at Enveda that is also contrarian to even the contrarians in biotech, which is that we tend to put human reasoning, interpretability, and understanding at the center of the scientific method—for good reason. We’re a curious species, and it’s allowed us to do extremely powerful things to shape our environment and the world we live in.

But I don’t think that it is necessary to build systems that have interpretability first, or understanding as we understand the term first, in order to create useful things. So, for example, if you had to first spend the entirety of your budget, energy, time understanding the brain before you could make a molecule that moved the brain in the direction you wanted, we wouldn’t have had any psycho-therapeutics until now.

So our bet is that you can actually create models—physical, digital hybrid models—that as long as they are sufficiently predictive, that is enough for us to be able to transform drug discovery with technology, again, digital or physical. You don’t need to put it through this egocentric, anthropogenic step that you have to understand it. People think, and especially scientists, are extremely, I think, motivated by the idea that they are the agent of discovery and that their comprehension is what’s most important.

But we made the steam engine before we understood thermodynamics. The Wright brothers flew before they understood aerodynamics fully. We created and used electricity before we understood the flow of electrons. We made the first vaccine before we knew that germs cause disease. And so I think that we’re entering another era where there’s a fork in the road. If we have access to incredible intelligence, are we going to bottleneck it by first saying, no, no, no, you need to be able to explain everything to me first? Or are we going to simply use it and say: as long as a model is sufficiently predictive, it can be used to create massive utility and good in the world, especially in biotech and drug discovery.

Murphy: You used the example of the Silk Road. But I think most people are more familiar with this heuristic, trial-and-error-based approach to discovery. Like, I think most people will be familiar with Fleming and the breadcrumbs and “Oh, this is how we discovered penicillin.”

Colluru: Turns out there’s lots of things about Fleming’s story that don’t add up.

Murphy: Fake news.

Colluru: Yes. Including how he described this experiment, how it was not possible for it to have gone down the way it did, and nobody really knows what actually happened. But if you pieced together his story, apparently it doesn’t make sense the way he told it.

Science Has Been Divorced From Human Experience

Murphy: Why did we lose that heuristic, trial-and-error-based approach to science? Why did we move towards that more egocentric point of view?

Colluru: I believe that it is largely our definition of science. Like, for example, you can say science is the incremental acquisition of knowledge through observation. But we tend to divorce human experience from science. And it’s almost frowned upon in the modern scientific enterprise.

There’s a beautiful book, actually, by a theoretical physicist, a particle physicist, and a science philosopher. It’s called The Blind Spot, and it essentially argues that our picture of science is as if there’s a disembodied observer from nowhere, looking impartially and completely unemotionally at a system to observe a real truth. And it goes on to examine all of the laws that we’ve created and then actually say that they’re all—we append the word “ideal” to them, like the ideal gas law, because no gas actually behaves like that.

You could say, oh yeah, we have an extremely incredible understanding of mechanics and how water flows. But if you ask a surfer, an expert surfer that’s at the crest of one of the world’s toughest waves, and he’s about to scale it in that moment in time—you can’t really separate the experience of being on the wave, perhaps, and be able to dispassionately say, well, here are all the forces that are acting to be able to predict it.

And then from a pure, like, maybe philosophical level, we already know that experience itself manipulates systems. Like the double-slit experiment shows that photons are conscious. They know if we’re counting them or not, and they know how to behave either as a wave or a particle. And so I think this has very much been an emergent phenomenon in the post-Industrial Revolution era, where science is supposed to be pure and the real truth, but as an extension of that was very militantly forced to separate from human experience.

But the reality is—and we know from everything from the double-slit experiment in physics to the placebo effect in humans—that maybe a lot of reality exists within experience. And that’s why you can even tell people that they are on a placebo, but if they believe that they’re going to get better, they frequently do.

Scientific Reductionism Has Not Led to Better Drugs

Murphy: When you think about this as applied to biology and applied to drug discovery, there’s this Napoleon quote about “the pains of precaution often exceed its benefits. Sometimes one must abandon oneself to destiny.” We’re obviously as far away from that for many good reasons - on putting novel drugs in people. But how did we get there? And what do we need to modulate to put ourselves on a better path?

Lubroth: We went from really finding drugs based on what they caused phenomenologically. So saying, oh, this had an effect on a petri dish—or maybe not a petri dish in Fleming’s case—but it’s like, okay, let’s explore what molecule caused that. And we developed many useful drugs that we still use today through that method.

Then, you know, the molecular biology revolution came. We started seeing things as: oh, we can actually reduce this phenomenological problem to its constituent parts. Let’s see if we can find the thing that causes the specific disease. It was a way, in part, to reduce the number of variables in order to find an effect for a new drug. But in a way, I think there’s also something to science where you want to find the purest form of the answer. And there is something maybe unsatisfactory to “it just kind of works, I don’t know why.” So it’s like, well, it’s this protein that has this specific mechanism of action and these are the biomarkers.

And I think there’s some utility to that. Like, you can increase the safety of a drug if you know exactly what’s happening—which rarely occurs. Rarely. There’s always going to be things that we don’t know. You can maybe understand the population subtype that that drug is going to work in. But you’re also missing a ton of complexity. And that complexity is an opportunity for many reasons, in that you can potentially find drugs that affect many more types of patients rather than a specific subtype, or even treat a disease in a completely different way that we haven’t even discovered molecular biology for.

Colluru: Pablo said something that is also discussed in The Blind Spot that I think is exceptional, which is: we tend to not just divorce experience from observation, but we tend to define things as fundamental if they’re smaller. So we started by saying, oh, do I have a model that faithfully replicates gastric reflux or hypertension? And do I have a molecule that is safe that can ameliorate those phenotypes or that phenomenon? And for every place where there existed a model, we created a drug already.

And then we have the molecular biology revolution. And we said, no, no, no, no, no, we’re going to study disease not at the level of the organism, but actually at the level of the organ. No, we’re actually going to study it at the level of the tissue. No, we’re actually going to study it at the level of an individual cell—and quickly jump that boundary wall to say, I’m going to study it at the level of a single protein. And in the last ten years, we’ve said it’s going to be about specific residues on a specific protein.

Now, we all know that almost everything that matters is emergent. Consciousness is emergent. No doctor will tell you that any disease has to do with a single mutation, unless it’s a rare disease. Most chronic diseases we care about are multi-organ. Obesity is as much about the brain as it is about your adipose tissue and the pancreas.

And this reductionism, I think, is something you can see everywhere in science. I had the opportunity to go see the Large Hadron Collider at CERN, and the rhetoric around how they are discovering a science that is fundamental to existence is everywhere. The word “fundamental” is fundamental to their messaging. But do we really care about subatomic space? And if we did, do we believe that life as it exists today in all its beauty could be explained purely by that? No. Maybe one day, if we had infinite computation, we’d be able to say how the atoms that were once stardust that now live in us would then go on to behave and rearrange in a way that is divorced from consciousness and experience. But we’re still far away from it.

Now, if you say, well, I’m going to try and break drug discovery and/or biology down to very simple things—a single protein, I’m going to call that a target—I only know and have discovered, thanks to nature, a small proportion of things that nature does to modulate this protein. So activate it, inhibit it, or degrade it, causing it to essentially go away, recycled in the cell.

Over the last 20 years, we’ve reduced all of complex biology, physiology, and disease to saying: I’m going to target a single protein. And then I reduced all of how nature orchestrates biology during the course of the normal living process to three things. And we already know that most of biology doesn’t happen at the level of individual proteins. Most of their regulation doesn’t happen through these three things. But what we’ve done is to say: I’m going to break this down into these parts where I identify a target, identify a chemical that is able to do one of these three things, and then I’m just going to keep going. And then I’m going to put this molecule that does one of three things to one of whatever 20,000 proteins, and I’m going to put it in people, and I’m going to hope that it works in a walking, talking, complex human being to reverse a multi-organ physiological dysfunction.

And for what it’s worth, the last 20 years have given us the false confidence that a lot of biology and life can be manipulated this way, which actually, I think, explains the 10% success rate. I’m actually surprised it’s 10%. And that’s how we do drug discovery.

As you might imagine, if biology and the underlying phenomenon are a distribution of complexity—and let’s just say they’re normally distributed—if we have 70% of our proteome spending 70% of their residence time in complexes, then you would expect that only a small proportion of proteins that we manipulate individually should do what we want. And that’s, I would say, lower than 10%. So we’re actually kind of lucky that we hit 10%. This formula, this really reductionist formula of doing science, is what we decided was the right way.

Once we mapped the genome, and then we also said: but it’s too much hard work. So as we learn it, we’re also going to engender this cottage industry in China called contract research organizations, where they’re just going to do the work for us, and then we’re going to own the intellectual property, and then we’re going to reap the rewards when the drug works. So that’s effectively what has happened over 20 years.

Lessons Learned From Big Pharma Missing Out On GLP-1s

Murphy: The biggest gangbuster for every large pharmaceutical company at the moment between Eli Lilly, Novo Nordisk, is GLP-1s, which of course the gangbusters application of them is not at all what they were developed for. And that’s where we’re at in terms of hoping for the best. Is there so much industry structure around pursuing proteins that breaking that inertia is just almost impossible?

Colluru: You asked me the question about why we make science so egocentric while positioning it as dispassionate and disembodied? I think it’s similar dynamics at play, because if you are at any large company where there’s status quo, the status quo is that we think that all drug discovery should be manipulating a target protein. It is really, really hard for you to be able to go against that, whether it is a scientific bet—to say, oh, I actually think that we should first start with phenomenology that has an interpretable positive predictive value—or whether it is, you know, you say, you know what, I actually think the world would benefit from and would pay for a drug that caused body weight loss, which, by the way, the pharma industry did not believe for the longest time.

Murphy: Was even possible.

Colluru: No. Even if they believed it was possible, they didn’t believe the market existed. There’s multiple analyst reports you can dig up that reinforced pharma beliefs, that reinforced analyst beliefs for three decades, that suggested that obesity was never going to be a big commercial market because people can just eat well and exercise. Why would they pay for it? And as late as, I think, the last decade, really sophisticated leaders from top companies in metabolic disease were throwing the number out that it’s about $100 million. Today, we know that it’s at least $100 billion. So they were off by a thousandfold.

Now, the reason I bring that example up to answer your question of why do we stick to a method that maybe won within the matrix and two, clearly not working, is, I think, because the incentives for you to individually take risk are completely lopsided. Because if you take what is a contrarian idea and you champion it within a large company, then you actually have all of the downside risk on your shoulders and none of the upside benefit, because if it works, everyone gets credit. Your boss gets credit for signing off on it and doing the deal. But if it fails, you’re no longer at that company.

Capital Markets Disinventivize Risk-Taking and Harm Patients

Murphy: A lot of the incentives we’re talking about are ultimately: where is the capital coming from, and what is that capital’s appetite for this type of risk? What works about the capital markets as applied to biotechnology and drug development now? And what does not work? And how do we spend the next 25 years fixing that?

Lubroth: You know, if you’re running a drug discovery fund, you see all of your investments as a portfolio. So you have at least one out of ten investments that will have an approved drug or will get acquired, and it will make up for the rest of the portfolio. So the power law is even greater in that sense.

However, typically, you know, more traditional biotech investors will want to have a very good insight into why specific target, why specific management team, and control as many risks as possible just to leave the inherent risk of science as the only kind of leftover risk. So that works relatively well for those funds. But up to a certain point, most of those investors won’t really invest in companies that don’t follow that specific formula.

But if you look at history, you look at Genentech, you look at Vertex, you look at Regeneron—all of these have pretty non-traditional ways of being founded. Genentech, which pioneered the biotech revolution, was started by a VC and then an academic. But a VC that was not really—they weren’t traditional biotech investors. And it was similar to Vertex. You know, Josh Boger left the company because he wanted to implement a completely different way of carrying out drug discovery. And those are the types of companies that I feel the capital markets aren’t as ready for.

Colluru: I think an easy symptom to sum it up is: the capital market ecosystem, as it has evolved and ossified in the last 20 years in biotech, has produced really good drugs—maybe not quite generational drugs. GLP-1 came from a pharma company, but has not produced really good generational companies.

That’s because there’s two local maxima. One that maybe evolved 15 to 20 years ago, and one that is evolving in the last 5 or 10 years. The first local maximum is that of a single-asset biotech company, as Pablo mentioned. Because biotech is so hard, a single drug can be so valuable and so risky at the same time. And it’s an illiquid, whole lot of value that boards, investors, and the markets are like: just make one. Even if you have a few, let’s just pick the winner, invest in that, get that to show a Phase II result, and then let’s sell it. The IRR works, the math works. The behavior gets reinforced. So that’s the first local maximum.

The second local maximum is: because we’ve struggled to discover what we call first-in-class drugs—so drugs that work by a novel mechanism—and because the failure rate is so high, in the last 5 or 7 years we’ve gone to saying, well, we’ll just make the same version of that drug, but one that you can take once monthly instead of once weekly. So it’s iteration instead of innovation. And that’s because, again, as long as large pharma still buys this company, as an investor you’ve made a return. So now your downside risk of looking stupid in an IC meeting is lower, and your upside was about the same. But the net losers are the industry, the scientific enterprise, and the patient. Because the patient now gets iteration instead of innovation.

And these two local maxima are at play and work against any company that’s trying to do a novel approach, completely novel biology, and trying to be the next SpaceX, Apple, or Amazon in biotech.

Murphy: Maybe the way people think about it is drug versus platform. What is it about a platform company versus an asset company? And why do the platform companies typically, historically—say, Recursion—become an asset generation company? Why is that the case?

Colluru: It’s really simple. Because the time to value, the risk to value, and the lack of liquidity in that path means that value—net value—can only be underwritten in the industry with the product. In other words, that investor, we say, a drug company must have drugs.

As of today, there’s really no other model that has been shown to be successful. Eli Lilly is worth $1 trillion. 70% of their EV is forward sales multiples of Mounjaro. And that power law is active at Lilly, Novo, Sanofi, Merck with Keytruda, AbbVie with Humira and now Skyrizi, and so on and so forth.

So we said: look, if we truly believe that our combination of thesis, technology, and team can produce better medicines faster—and cheaper is a bonus, but better medicines—then we should put our money where our mouth is and just try and make medicines, and make them in the most competitive areas. So our lead drug is for atopic dermatitis and asthma. Our second drug is for obesity. Our third drug is for inflammatory bowel disease. Our fourth drug is for essentially alcoholic to metabolic liver disease. So together we touch billions of lives with these indications, and we set the bar really, really high.

Now I have to say that we’re not uniquely smart in being able to think about this. It’s just that because of the work that I think companies like Recursion have done in creating the path and bringing in a new kind of capital base to the industry, that we were even able to afford to do this. And that’s why I think science and progress and disruption happens in invisible cycles. And if Recursion hadn’t done what Chris did, I don’t think Enveda would have been able to raise capital from the unique capital base that sits outside of those two local maxima and do what we did.

How to Prioritize Which Illnesses to Cure

Murphy: How do you think, as the leaders of a company developing a platform to tackle such a wide range of illnesses—how do you prioritize? What do you think about impact versus commerciality? Is there a trade-off, or is it just finding the thing that works?

Colluru: I think for us it’s really simple. Is this a problem that you and your loved ones would wish you had a solution for, or a better solution for? Historically, the pharma industry has struggled with what we call the existing multiplier problem. So, you know, 15 years ago, 10, 15 years ago, everybody was convinced nobody would pay for a drug for eczema because it was a disease of young children. You grew out of it, and it was just a rash that could be treated with almond butter. And this decade, we obviously thought that about obesity until it hit us in the face that, of course, you could just eat better and exercise.

That’s because, again, I think commercial risk-taking in pharma has this same problem as scientific risk-taking. So what we do is instead say, you know, from first principles: do we or someone else we love have this problem? And would we be willing to pay for it? So we look for markets by just looking around the world and paying attention.

So yes, we’re working on a non-steroidal, once-daily drug that’s safe and effective for atopic dermatitis and asthma, which we think is the second-biggest market in the world today. We also think that, yes, obesity is a huge market, but GLP-1s are just the beginning. We had the former global head of long-term obesity medicine and obesity prevention at Novo Nordisk join us recently. And he said we’ve done surveys of tens of thousands of people. And one of the most religious users of GLP-1 are perimenopausal and menopausal women, and they don’t want to lose 25% as much as the stock market would lead you to believe. They actually just want to stay at 48 how they were at 38.

So if you think about real-world obesity struggles, it’s chronic—unless you’re special and very different from me. Obesity is something that you struggle with on a daily basis. You don’t become obese overnight. And you want something that, even if you use medicines like GLP-1 to get thin once you’re fat, to stay healthy, metabolically healthy. And there’s only two other markets with features like this that affect hundreds of millions of people: hyperlipidemia—so high cholesterol—and high blood pressure, hypertension. And both of those are actually served by a once-daily oral pill, so a statin or a beta blocker, that your doctor can put you on in the five-year journey between you being completely healthy to your LDL starting to go up a little bit and then a lot.

We realized 18 months ago that a big need would be a long-term, durable bedrock of care—like a statin for obesity—that can be used to prevent you from getting obese ever in the first place. Or if you’ve used GLP-1, for you to lock in those gains. Nobody believed that. They were like, no, no, no, we want you to make a medicine that can get even more than triple G. Now all of a sudden they’re like, oh wait, people are going off of GLP-1s. The insurance companies that paid for them are losing the benefit. Now everybody is in for maintenance.

So that just gives you an example of how we think about markets. And we think that there’s no shortage of markets today. So, for example, we’re working on medicines that can prevent migraine, that can treat Parkinson’s without rebound tremors, that can treat depression without substance abuse risk, that can treat rheumatoid arthritis without you having to cycle through multiple medicines. And because as long as there is unmet need for healthcare, there’s willingness to pay. The key is: can you produce medicines that are scalable, that you can price medium?

I think for a long time it became vogue for pharma to charge millions of dollars to treat really rare populations. And that I don’t think is a sustainable model for healthcare. There’s a place for that. But we should fundamentally aim to do what GLP-1s have forced the industry into, which is medium price for a large number of people.

Lubroth: Sometimes you just have to observe what’s around you. If you’re at an airport and you look around and everyone’s on their phone. Everyone is slightly overweight or obese. Clearly not everything is solved. Sure, we have benzodiazepines for anxiety, but people are so anxious. Or similarly, you know, with depression. SSRIs don’t work in a large portion of patients. So there’s still a ton to be developed for those people that don’t have specific solutions to such important diseases that reduce their quality of life.

And so I think that’s an important aspect. And I think how we see it as well is: okay, what are the trends that are changing in our society and what is going to continue to happen? And I think—and this maybe goes more into the 2050s side—but as we become increasingly interconnected with technology, that’s just going to create new problems. Anxiety rates have increased massively.

Colluru: Cognition and attention have gone down.

Lubroth: Yeah, exactly. And I think that’s just going to continue. So how can we prepare and develop drugs for those types of maybe even diseases that haven’t even been quite defined as a disease yet? Maybe they’re not in the DSM-5.

Prediction and Prevention Is a Better Strategy Than Treatment

Murphy: These are the biggest problems in the world. If we do not figure out how to keep people happy, healthy, and strong into their 80s, that’s the biggest opportunity in investing right now. Do you guys believe that by 2050, with enough resources, we can actually have solved all those problems?

Colluru: I’ll give you a practical thing we’re working on right now, and I’ll give you a philosophical answer, and maybe one of my most recent exciting solution spaces or solution sets that I think can help us get there.

So first and foremost, I think you need to be able to predict and prevent rather than just intervene. And an extremely key piece of that is being able to administer safe medicines where the risk-benefit ratio for someone that isn’t yet obese, for example, but whose trajectory belies that they will be obese very soon—for example, you know, there’s 11.5 million US adults who are not obese in 2026 that will be obese in 2027. We just don’t know who they are. But they show up in the census.

But the reality is that we know that if we can measure, we can predict. Then the question is: well, what’s the use of predicting unless you can prevent? Because you can tell someone, hey, your blood parameters, your behavioral patterns that we can tell, and your weight trajectory puts you on a high-risk metabolic trajectory. Good luck. What is that going to do to people? Nothing. And usually that becomes the argument for “let’s not invest in infrastructure.”

So then you ask the question: what is the bottleneck? The bottleneck is if I can give you a pill that is cheap, that is convenient—so you form a habit in a good way—and that you are incentivized to take, that nudges your own internal behavioral characteristics and gives you the lift that you need to make changes. Then you will have the power to essentially say, well, let me just not become obese in the first place.

And so we work on small molecules—easy to manufacture, easy to distribute. We do believe that, as pharma is finding out right now, obesity and metabolic health will be an increasingly and overwhelmingly self-pay, direct-to-consumer option. So we said: look, if we can make a once-daily oral pill that’s a small molecule, that’s cheap to manufacture, facile to distribute, and achieve global scale—today, less than 1% of the people that are eligible for a GLP-1 can access it or afford it—then you can create a class of medicines where you can intervene in chronic disease earlier and earlier, and you can preserve healthspan.

In the case of obesity, for example, there’s a metabolic set point. So if you wait for someone to become obese, then your whole life, your body will try to get you back there.

Lubroth: Maybe to adjust on the prediction aspect—we’re already there. We can already do it pretty well. Just from your selfie camera, you could correlate BMI with a DEXA scan on body adiposity. We just need to implement it in terms of prediction and treatment in order to lead to prevention. But that is extremely hard. And it’s the same reason why diagnostics is extremely hard as a business model.

Colluru: Yeah. We’re so excited about it because one of the things we’re finding is that we can finally take all of this latent energy where it’s been so obvious. You can tell whether someone’s going to be obese from even their blood test, from your phone, facial recognition. In fact, Apple had to learn what weight gain looks like because if you’ve been to a couple of Indian weddings and then suddenly your facial recognition doesn’t work, you’re pissed off at Apple. But that’s because you put on a couple pounds on your face. So Apple’s had to learn to subtract obesity out from your face this whole time. So it’s already all been there.

So then you say: well, how can I create incentive and infrastructure to close the loop? And the missing piece has been safe, highly scalable medicine, which we argue is not cold-chain injectable peptides or enzymes or biologics, but the oldest scalable pharmaceutical technology in the book, which is a small molecule that can be in a pill.

And if we bring that—we’re starting to do that for obesity, but we think we can do that for mental health, cognitive health, we can do that for inflammation. We can begin preserving people’s essentially biological entropy before it goes out of control. And then, practically, through an existing low-risk business model, create healthspan intervention without sounding kooky.

Drugs to Make Us Limitless Like Bradley Cooper

Murphy: I think everyone has seen the movie Limitless, and they imagine: wow, drugs. Imagine you could create a drug that just turned me into the smartest person in the world. I spoke a little bit with Noor Siddiqui about the genetic side of this, which is the very third rail. How does chemistry work in that world of enhancements?

Colluru: So the short answer is we see tremendous possibilities. For example, you know, we’ve discovered a number of new exercise-regulated hormones that can be made into pill form that basically mimic exercise. For example, not only do they improve cognition, metabolic health, cause reduction in weight, reduce substance abuse liability in mouse models, reduce visceral pain—but also they seem to reinforce exercise. So when we give this molecule to mice, they stay on a voluntary wheel twice as long. Oh wow. And so that shows you the power of being able to look within that 90% of unknown chemistry and discover the next testosterone, the next melatonin, but one that regulates everything from mood and pain to exercise and appetite, because we’ve barely scratched the surface.

Murphy: How is the current regulatory regime set up for an enhancement versus an ailment? Is it very similar?

Lubroth: So for something to be a drug, it has to have a therapeutic use. Of course, you can use it off-label. I mean, everyone uses GLP-1 for enhancement.

Colluru: Or Adderall during exams.

Lubroth: Or Adderall, modafinil. There’s so many drugs that are already being used off-label, but there’s no specific path for enhancements. I think the closest thing could potentially be the supplement path. You can still claim a structure-function claim. So you can say “reduces stress,” but you can’t say it’s for generalized anxiety disorder. I think that is probably going to be the regulatory space where most companies are going to—they’re either going to go down the supplement path where they can claim something around that, or hope that the drug also makes it out into kind of the enhancement area as well, just like GLP-1s and all the other drugs that were mentioned did.

And I would add also, in terms of the chemistry: as we mentioned a couple of times, there’s just so much chemistry in our bodies that we don’t know anything about. We’ve discovered one metabolite from an amino acid that has similar potency in animal studies to diazepam, but it’s totally natural. And so it’s like, how many of these exist that we could replace with people with certain deficiencies?

Colluru: Our scientific co-founder nearly doubled the size of the known human metabolome or chemistry map code, if you will. And along the way, you know, discovered that there are tens of thousands of bile acids, not a few hundred as originally thought and as written in biochemistry textbooks. And it may be how our body does cross-organ communication of which organ needs nutrients.

And along the way, just discovered that a big part of that is how we interact with our environment, for example through the microbiome, and discovered some really cool things. No big deal. Like, oh, why if you take ibuprofen as an ultra-marathoner, it takes you a longer time to recover? Because all of the succinyl-CoA in your body essentially gets soaked up by this one metabolite, and it acts as a sink. Or, you know, this 70-year-old drug, mesalamine for IBD, actually doesn’t work at all the way that people thought. And it actually relies on a gut microbe that needs to conjugate it to create this active metabolite.

So we’ve barely understood human physiology. And the last thing on the case for chemistry is, I think, exemplified by eusocial insects and naked mole rats, which are the favorite organism for aging. So in naked mole rats and, for example, in bees or ants, the queen has the exact same genes as the workers. But in the naked mole rat, I think the queen lives ten times longer, and in eusocial insects between 10 and 100 times longer. So clearly, plasticity exists outside of just understanding genes.

Aging is Not Inevitable

Murphy: In terms of really using chemistry to control the body more effectively in terms of lifespan, where are we there?

Colluru: I have a fun idea. And this actually was spurred by a book I read on a small beach in Mexico during the winter break called Why Does Evolution Kill? It’s a theory called pathogen control theory by a professor that used to be, I think, at Berkeley or UCSF, now is in Hong Kong, called Peter Lenski.

His hypothesis is that rather than aging happens naturally because we accumulate damage, which is the prevailing view, or that aging doesn’t matter to evolution because you reproduce before you age usually, that aging is actually an active, programmed mechanism for older individuals to die off so that they don’t become sinks of chronic infection. And all you really need is sterilizing infections, the possibility for epidemics, and that you live near genetically related individuals.

And what’s beautiful about this theory is that it’s the most parsimonious explanation for the 500 divergent observations we’ve had for aging. For example, why do birds that fly live 20 times longer than you would expect from their mass and metabolic rate? His hypothesis is that birds that fly disperse far away from their genetic relatives. Eusocial insects—why does the queen live 10 to 100 times longer? Because the hive is essentially the immune system that interfaces with pathogens, and the queen is essentially sterile, so her systems to kill her off have not been activated.

So if that’s true, if Peter was on to something, then we reframe all of known biology essentially into timekeepers. What are the molecules that tell your body it’s old? And then essentially, what are the programs that kill you off? And if you view that—some of the most exciting things we’ve discovered are dozens of these molecules that correlate with aging and aging-related diseases that nobody knew about. Now we can say: what are these things doing to tell the body, “don’t repair yourself anymore”? And I think studying at the level of the organism rather than the single gene, studying through this completely new model or lens of evolution, which I think resources can enable because it allows you to think outside the scientific establishment, will allow us to perhaps do some pretty fantastic things for health.

Everyone Has a Chinese Peptide Guy

Murphy: Biohacking has become much more prevalent. Everyone has a Chinese peptide guy. Everyone’s got a supplement dealer. Everyone’s using the red light. You’ve been in this space, following it for a while. What are the positives around this becoming much more mainstream, and what are the potential negatives for people to think about?

Lubroth: Yeah, I think on the positives, there’s so much in the supplement space that is actually effective. I don’t know if you’ve ever tried lithium or melatonin—obviously, you know it works. Ashwagandha is another plant that you can also take for calming. It works. I mean, at least it could be placebo, but there are several studies out there that prove that there’s some efficacy to it.

So I think in terms of the positives, people can take control of their own health in a way that they haven’t because they need to wait to get the prescription for the drug that they actually want to get. And that, I think, is definitely useful for maybe subclinical indications—meaning ones that are not that severe. Say, you know, just feeling anxious versus actually having panic attacks.

In terms of the potential downsides and consequences, I think there’s a lot of snake oil, and I think that’s not going to go away. Perhaps it’ll become even harder...

Colluru: As people look for it and are willing to pay for it.

Lubroth: Yeah, exactly. And I think there’s space for new regulatory categories that blend things like supplement and drug, or cosmetic and drug. It occurs in some other countries in Asia, for instance, in order to be able to select what is efficacious and safe from the snake oil.

Colluru: As Pablo laid out, if you flip it and say, what’s the capital value equation, you’d be hard-pressed to find a supplement company that’s doing over $1 billion in sales. I think Optimum Nutrition does it, but through everything from protein powders to gummies. Meanwhile, I think a half-decent biotech company that has created a molecule and proven its effectiveness has pricing power and has societal underwriting to be able to value it.

So in the long arc of time, we believe that drugs are certainly not the only go-to-market channel and/or not the only way we should get our inventions to people. But for the investment and underwriting that a company and a technology platform like Enveda needs, we’ve decided to put our molecules to the highest bar. And while we have the privilege of breaking the bounds of traditional biology thinking, we really want to fit them all down that go-to-market channel and prove our science is fantastic. But we’re very eager and excited to see where the world responsibly leads, perhaps on regulatory strategy, such that we can create the next generation of products that don’t have to go through eight years of clinical testing but still have that same bar of risk versus benefit.

Lubroth: Yeah, 100%. I think also on the supplement and longevity space, there’s kind of a Pareto law aspect to it, which is: maybe just don’t have fast food, run every day, and don’t smoke cigarettes, and you’ll probably...

Colluru: Be at reduced alcohol.

Lubroth: And reduce alcohol. So that’s pretty much...

Murphy: First of all, come on! We can only do so much :)

Colluru: If this aging as an active programmed organismal death is real and it’s driven by selective pressure of infections, then the biggest lever to dodge would be actually changing your immune system. Your immune system should be the master control for multi-organ aging, which actually five years ago was a Nature paper where they took old immune cells, put them in young mice, young immune cells, put them in old mice. And that alone was sufficient to change biological age.

So what is the only known molecule that can age the immune system? It’s a molecule from a living system called rapamycin that comes from Rapa Nui, Easter Island, made by a fungus. So we think the living world is the richest source of chemistry to modulate biology and physiology in traditional and non-traditional ways, to create drugs, and hopefully lots of other products along the way that improve human and planetary health.

The Loon Shots for 2050

Murphy: What are the loon shots, the big ideas that we should be cultivating and pursuing in biotechnology and medicine in the next 25 years? What should inspire people today?

Lubroth: So I think one of the things that we’ve focused on right now is looking at differential profiles of the metabolome in different humans. So in disease and healthy states—looking at what exists in a higher concentration in the plasma of diseased individuals versus healthy. And I think given the huge amounts of data that we’re going to gather, I think it’s also breaking apart the categorization of disease around: is this an autoimmune disease? Is this a metabolic disease? Is this a neural disease? Those things—the body doesn’t do that. It’s not categorized. So could we find nodes that we can poke at that effect not just in metabolic disease but also in neurological conditions as well.

Colluru: You can imagine that you have your baseline and the average healthy human baseline for your comprehensive chemistry. That is a snapshot of your internal genetics, your personal diet, and your living environment. That’s your chemistry at any point—a full reflection of all of that: your diet, your microbiome, and your physiology and genetics.

You could constantly monitor that, perhaps through an invisible sensor—a small mini mass spec. You could predict and prevent any disease that seems to be emerging at the multi-organ level, because chemistry is also the highway with which your organs talk. So if you just take: 0% of genes are exported outside a cell. Something like 20% of RNAs are packaged and shipped between cells. I think something like less than 50% of proteins, if even, are secreted. But almost all of the molecules you can find within a cell, you can find outside the cell.

So if you think about this, chemistry is the highway for inter-organ communication. So then you say: oh great, whole-body physiological monitoring, prediction, prevention. And then you just take molecules to rebalance your chemical code, proactively keep your immune system young, keep your organs young, achieve longevity and healthspan.

Mental Health Will be The Biggest Challenge for 2050

Murphy: What do you think is the big medicine healthcare challenge that we will have completely eradicated? What do you think is the challenge that we will still be battling with in 2050?

Lubroth: I mean, I think monogenic disorders we’ll be able to eradicate. I think that’s slowly becoming mostly an engineering problem. I think what disease will be in the future—I think as relatively more severe diseases get cured, there will be newer definitions of what disease is, similar to what we were talking about earlier. You know, if you were anxious 200 years ago, that wasn’t a disease. So what is the blending of technology going to lead to? The biggest challenge will be around the psychological consequences of technology. And I think that will be extremely hard to tackle in the future.

Colluru: I think I’m with Pablo on that. I was going to say the thing that we will, I think, get worse at is our mental health. I think technology, constant connection—but maybe not really the right kind—news flow and the attention economy are going to make it so that we’re compulsive doomscrollers. And I think as a long-term meditator, I found that I’m most happy when I’m present, and I’m least present when I’m caught up in technology and devices.

So I think that will be something where we will need maybe chemical and non-chemical interventions in order to help us find a place of stability and happiness. I do think that in addition to monogenic diseases, essentially most physiological bases for long-term health—like maybe obesity and metabolic health—we will have solved. Companies like Enveda would have created globally accessible solutions. And if you can fix obesity, you can fix a lot of things. So we will make a big dent on chronic disease, I think, in the next 25 years, and eliminate one of the big drivers for morbidity.

The AI Revolution in Medicine

Murphy: How transformational has AI been for the work that you do? Is AI changing everything, or are people over-hyping it in the space of medicine?

Colluru: Yeah, for us it is indispensable to what we do because the key piece that we’re doing at Enveda is effectively doing to chemistry what next-generation sequencing did to DNA. So building a sequence of life’s chemistry. And for that, we had to yank analytical chemistry from its prehistoric movers into the 21st century. And we couldn’t do that if we didn’t have LLMs and context-based predictions and the ability to turn these molecular fingerprints, which we get from mass spectrometers, into chemical structures that humans can interpret.

And so AI is both, for us, enabling a completely unfinished idea to be revisited in a powerful modern way, but also on a day-to-day basis, it’s helping us do things that would have taken a very long time much faster. You know, with Claude, we all have an unlimited, infinitely powerful assistant for 100 bucks or whatever a month. And the pace of discovery, the pace and quality of decision-making is directly translating to Enveda being a much more robust, antifragile company than it used to be.

Murphy: This was amazing, guys. Thanks so much for taking the time to talk about everything. I really appreciate it!