Martin has had an illustrious career at iBio. He’s leading the way inย usingย AI for groundbreaking drug discovery.
Welcome, Martin. It’s such a pleasure to have you.
And for me, it was really kind of an exciting career for the first 15 years to see how it all fit together, how you make medicines from A to Z. But at one point, and this is not the fault of any given company, it’s just by scaling, there are certain things that don’t work for early discovery. So, for early discovery, you need to be fast, you need to make decisions, and you need to be aligned.
And the larger the organization, the more challenging this gets, right? There’s conflicting interests, and that means decision making is prolonged, and sometimes politics start playing a role. And what I always do is I balance this with my passion for making medicines.And when I felt that the processes and the politics got in the way of my passion making medicines, that’s when I decided to jump into biotech. Biotech is something that really supports my passion for making medicines. This is why I get up every morning.I’m basically working every single day with a group of people that in their respective fields are no less than geniuses. And so there’s always a stimulating conversation to be, there’s always science to be driven forward. And just being there with like-minded people, having a common goal, being fast in decision making, moving science towards really helping people in the clinic.
For us, it’s always the second we actually dose the first time a human being, that’s for us the breaking even point, if you will. It’s a long, long journey from that point still, but that is what we look forward to. And this, I think, is all provided for me in biotech.
If you will, I often thought about myself, I live close to Las Vegas, driving distance. And I always felt like I’m not a gambling person. This doesn’t excite me because I cannot influence the odds.
But in biotech, I can actually influence the odds of winning by putting together the right team, putting together the right technology, and also putting together the right strategy. And that’s what kind of drives me.
Host: Okay, absolutely. That sounds very exciting and inspiring. So moving forward, transitioning from R&D in a big pharma to running a biotech must have been a great shift in your career. What’s been your biggest surprise as an entrepreneur and as a CEO?
Martin: So honestly, I’m a scientist at heart. I love biology, drug discovery. I like technology.
I have an engineering background from way back when and combining all of those is my passion. But the biggest surprise in stepping in the CEO role was how much I actually like being a CEO. I can get bored very easily.
That’s why, you know, biotech is a great ground for me because you never get bored. And the CEO role requires so many different skills, requires me to shift from understanding and communicating the science to working with my team internally. We have a huge focus on developing our team internally.
So the mentoring takes place all the time. Just, you know, interacting with our board of directors, it’s very, very diverse what I have to do. So there’s never a dull moment in being a biotech CEO. Granted, some of the tasks are not exactly my favorite, but no job is perfect. But I still think I have the best job in the world.
Host: Absolutely. I mean, every day must be very dynamic.
Martin: Okay.
Host: So moving forward, we hear a lot about AI-powered drug discovery often. So could you share how your platform uses AI to discover new antibodies and how that compares to the traditional years-long methods?
Martin: Yeah. I think this is really important that we actually kind of create a realistic picture of AI, right? So it’s a hype word. It’s used by everybody rightfully or not rightfully so. And I think it causes a lot of misinformation. AI is a dramatically important technology that will transform drug discovery.
But if I see people saying, I made an AI drug, it’s kind of laughable. It’s actually ridiculous because it takes more than 10,000 steps to make a medicine. And AI, in our case, influences three of these steps. These are very, very critical steps. But there’s a ton of other things that need to come together. So I want to always be very careful.
That’s why we’re not calling ourselves an AI company. What we do really well is we integrate these new AI tools with existing cutting-edge biological techniques. That’s our niche, if you will.
So we’re looking not to build these AI palaces, but what comes out at the end needs to be something that can become a medicine. So just to level set how I think about this, because if you follow those thoughts, we would call every drug we have today on the market a so-called molecular biology drug, because molecular biology in the late 70s changed how we do drug discovery pretty much as much as AI is doing now. But we don’t call our drugs molecular biology drugs.
So that’s for level setting. But I’m obviously a true believer and early adopter in AI in discovery and in biotech. And the way we use our tools is really to, number one, create precision. So our antibodies need to bind to very specific regions. We also use it to accelerate the drug discovery process, like you indicated. And we have hard numbers for this, right?
And the last thing is, and this is what is often an afterthought if you have in silico designed molecules, for example, can they actually be developed to become a drug? And it’s not just safety aspects of those molecules. It’s also the aspect of you have to actually manufacture those molecules.
And in the space of obesity, where we’re at the moment focused on, it means that 10% of the total worldwide population are 100 million people. So how do you manufacture drug for this large group of people? So all of these things are built into our systems. And this is how the integration of AI really drives us forward. But let’s get a little bit into, yeah.
Host: Okay, that surely sounds like a great complex process, as you’re saying. Yes, you were saying?
Martin: Yes. So let’s get a little in the details. So we start our process with our so-called engineered epitopes or engineered antigens. So think about a drug target. It’s a large protein, but only a small region actually has a biological function. If you find antibodies against the entire protein, you find antibodies against every region.
And they bind to the target, but they don’t do anything, right? So if you will, huge haystack and a few needles, we’ve removed the haystack by creating small representations of the region that we want to target. So we basically build the same surface.
And then there’s a scaffold underneath that stabilizes that. It’s high level, but that’s basically how we can eliminate the haystack. Think of it as Google Maps for antibodies.
Imagine you’re in Paris, France, and you need to find a very specific apartment in a very specific house, very specific street, very specific around a small in the city of Paris. And now I take the map away from you. This is how we do antibody discovery in the old days.
So currently most companies do. It’s random. It’s trying to do something. And I’m not saying you can’t find that apartment. You just walk around long enough, but it can take years. What we do is we punch in the address of the apartment in Google Maps and you can choose, do you want the fastest way?
Do you want the most fuel efficient way? You get there basically however you want to get there in basically no time. And this is how we have this, describe this epitope steering technology of ours.
The second part where we use AI is we have built a so-called naive library. So these are billions of antibodies that have been derived from human antibodies. So providing some safety when we actually put these molecules in humans, but also provide the diversity to, you know, mostly, most of the time, some of those billions of molecules binds to the target that we’re interested in.
So that has been created with, and I want to be particular with a machine learning tool, not an AI tool, not a true AI tool. And then basically we can do the same thing by then taking our engineered epitopes, injecting them into llamas or into mice, and then having the immune system of these animals create antibodies against that target. Now we still don’t match human and, or mammalian evolution, if you will.
It still creates a better diversity if we do this in an animal than if we do this in, in silico. There’s still some gaps. We’re narrowing the gap, but we still use both technologies, but we use them for screening purposes. So think of our engineered epitopes as hooks on a fishing line. And we dip them into the big lake of antibodies, either in an animal or in the naive library. And then we basically pull something out that binds to this hook.
Once we have this early molecule, they need to be optimized, right? So they’re usually not ready to become a drug. And there’s two ways of optimizing those. The first way is you create cousins of that molecule that is already working, right? And you create billions of them so that you might find among these cousins, better molecules, right? They bind tighter to the target or have other characteristics.
Usually what you do is a so-called phage display. You put an antibody on top of a bacterial cell, if you will. The advantage of this is you can create billions of cousins. Your diversity is high, but the downside is what is ultimately coming out is not a fully assembled antibody. And there lies the problem. Often these molecules don’t actually make it to a drug.
So you have to check for many of them to find one that ultimately becomes a developable drug. The other way of doing this is really kind of using a so-called mammalian display technology. There you actually make one antibody on top of one cell and it’s the same cell type that you actually use later down the road for manufacturing of the antibody.
And it has two advantages. Number one is you can actually look for multiple characteristics at the same time. And what comes out at the end is an antibody that can be optimized already for manufacturing.
But the downside of the system is you can only cram about 1 million molecules. So orders of magnitude smaller than in a phage library. And that’s where we actually integrate true AI into the process because what we do is we take this one molecule at the beginning that works just a little and then we not only teach our algorithm to make cousins that look similar, we also tell it, look, there’s billions of cousins that have not worked.
So exclude them, right? And this helped us to get to a much, much higher success rate when it comes to designing molecules that should be better. Not only optimizing for what we already have, but also basically teaching the system to ignore what has not worked.
And so we could increase the hit rate, which we could allow that to compress this massive diversity of a phage library in a much smaller library. And that combination actually was very successful. Now speed component comes from this, right?
Because the traditional approach to optimize an antibody is a sequential process that takes eight months. We have compressed this to three weeks. So this is one part where we really dramatically accelerated the preclinical development.
Obviously, because of the engineered epitopes, we can also accelerate the part in the front, but that’s kind of dependent on the target. This optimization process, we can almost guarantee it’s going to be three weeks, has been for many, many of the targets we have been pursuing in the past. Now, why are we so fast with this?
What we do with mammalian display is a so-called single shot multidimensional optimization. So instead of sequentially optimizing an antibody, we optimize multiple characteristics all at the same time. So we shoot these cells through a cell sorting machine, and then can basically tell all at the same time, does it bind to the target?
Does it express really well so that we can manufacture it? Does it aggregate, which is a really bad sign for an antibody, or does it bind to any other proteins in the body, which is also a really bad thing, right? And all of these parameters up to 13, we optimize all at the same time.
And that gives us this speed component, but it also gives us this developability because every molecule that comes off this platform is then basically ready to become a drug. So we increase the probability that this molecule is highly developable in manufacturing later down the road. And that is a critical component because every academic lab can actually make antibodies.
They’re actually quite good at this, but those antibodies are not necessarily drugs. And we heavily focus on, can they later be developed to become a drug?
Host: Okay, that truly sounds amazing. I mean, I understand how reduction of time in this entire complex process can make a great impact. Okay, so moving forward, let’s talk about your work on obesity, which aims to prevent muscle loss. Unlike many treatments, how does your approach promote healthy weight loss while improving overall quality of life?
Martin: I think that’s a really pressing question that becomes more and more focused. Here in the US, I think there’s been more than 50 million people now on GLP-1s. And so, first of all, I want to say GLP-1s have really started a revolution in how we treat obesity.
If you think about it, before GLP-1s, you had to get bariatric surgery, highly invasive procedure to see the same or similar weight loss. Now we can do this with a drug without the scalpel. And that is kind of a really big change, right?
But like every molecule, every drug, GLP-1s have adverse effects. One is nausea and vomiting. It can be pancreatitis, gastroparesis. So again, drugs are always kind of a compromise between how sick you are and the side effects of the drug. But what GLP-1s also do, and this is not the fault of the GLP-1s, if you go on a very low-calorie diet, the same thing happens. It’s really the caloric reduction.
If you lose body weight really rapidly, you lose not only fat, which is the desired tissue you want to lose, you’re also losing muscle and bone. And so unfortunately, if you take GLP-1s or go on a very low-calorie diet, you lose for every kilogram lost, you lose about 0.3 kilograms of muscle. The problem starts when you come off of the GLP-1s, because you regain body weight, about 80% of it within a year.
Unfortunately, for every kilogram regained, you only regain 0.08 kilograms of muscle. So that means at the end, you have the same body weight, but you have a lot less muscle. So what happens, we’re humans, right? We go back on GLP-1, we lose another round of muscle and another round of body weight. And guess what? We’re going to lose more muscle.
And this is called weight cycling. And there are studies that have been done, not with GLP-1s, but with low-calorie diets. And it actually, they showed that your risk for fractures due to low bone density and low muscle tone increased dramatically.
So if we’re not careful, right, we might not get cardiovascular disease, but we might end up in a nursing home in a bed because we don’t have muscle mass left or break our bones every time we go out. So that’s something we’re very, very cautious about, because our goal ultimately is to kind of extend the health span of humans towards the end of the lifespan, right? You don’t want to be 20 years in a hospital bed. You want to actually lead an active life. And for that, you do need to lose body weight or fat mass, but you also need to retain muscle. And that is our approach.
And we’re differentiated in this because when we started in April, 2024, to build the strategy, we thought, why compete with the Eli Lilies and the Novo Nordisks, who are the experts in making GLP-1s, instead of let’s look at what has to come next. Every drug leaves an area for patients open where there’s still kind of an unmet need for these patients. And that’s exactly what we have decided to focus on.
And now, you know, a year and a half later, it actually starts to really be beneficial because now the focus is actually shifting to exactly that, more healthy weight loss that leads to kind of a more productive life.
Host: That sounds like a great help for people who are on a weight loss journey, but do not want to compromise on their quality of life and muscle gain. Surely, that’s truly amazing. So, continuing as we go to our next chapter, your ShieldTX and EngageTX platforms seem highly adaptable. Beyond your current focus, what future applications are you most excited about for the tough-to-treat diseases?
Martin: So, let me first talk a little bit about Shield and Engage. So, EngageTX is a so-called T-cell engager panel. So, these are antibodies that bind to certain targets on immune cells.
And they’re meant or were meant originally when we were still working in immuno-oncology to treat cancer, right? Because what these molecules do is you have an arm of the molecule that binds to the tumor cell and the other arm binds to the immune cell. You bring the immune cell close to the tumor cell and it gets destroyed, the tumor cell.
And so, that was the reasoning behind EngageTX. Now, as we know from the past, these T-cell engagers can be pretty toxic. We see some really dramatic adverse effects. Now, this is why we developed ShieldTX. ShieldTX actually renders these bispecific molecules inactive while they’re circulating in the bloodstream. But once they enter the tumor, basically a mask or cap that sits on top of them, makes them inactive, comes off and the antibody gets activated.
It’s called conditional activation. And that allows us to very specifically target a certain tissue and avoid generalized adverse effects of a bispecific molecule. Now, as I mentioned at the moment, probably the most straightforward thinking is oncology for this.
But we are in a cardiometabolic space and we’re we talked about expanding the health span. So, there’s a growing field of aging research that is yielding now pretty interesting results already. I was never a fan of aging when the endpoint was death.
Now, we can actually determine biological age and have other parameters that we can look at. And there’s a so-called senescent cell population in an aging population that grows and grows. And that basically prevents tissue regeneration.
And senescent cells, pretty much like tumor cells, can be targeted and eliminated. And that should actually help with increasing tissue regeneration. There’s some early studies that are taking place in non-human primates that has been shown.
But these are small molecules that are not very selective. So, that could be a really, really intriguing application for both technologies in the future. If you think about it, any disease that has a cell type that needs to be eradicated, might it be overshooting immune cells that cause autoimmune diseases?
Might it be senescent cells that cause aging? All of these are very, very intriguing paths for these two technologies going forward.
Host: That sounds truly impactful for the industry. Next up, the drug development is a long and risky process. How do collaborations like your recent collaboration, the AstralBio partnership, help speed things up and reduce risk?
Martin: So, let me take one step back. So, generally, collaborations are most successful if both partners bring something to the table the other partner doesn’t have. And it’s not no different than AstralBio, but AstralBio is different.
So, for example, we’re a small company, right? And if you think about an approval trial, a phase three clinical trial to approve an obesity drug that will have to enroll 5,000 patients and will go for three or five years, this can easily be $300 to $800 million. For a small company like us, this is very unlikely that this is going to happen.
So, you need a partner in this case, likely a large pharmaceutical company that moves this molecule then ultimately to an approval process. We’re never saying never. If the company grows, we will be able to do this.
But from where we are right now, this is a good explanation how successful collaborations can work. Now, AstralBio, a year and a half ago, was seeing that there’s movement in the obesity market and there will be the need for the second generation drugs. And they have been looking all over the planet for companies that A, have the right technology platform to move quickly, but also have kind of the background in the obesity cardiometabolic space.
And this is how AstralBio found iBio. And so, while iBio provides all of the capabilities to move an antibody from early discovery all the way to clinical development, AstralBio came with the initial idea of the second generation drug, and they came with a very strong group of investors in tow. So, this combination of their foresight, plus their engagement in the investment community, plus our technology and team and skills, that was basically a perfect combination.
So, we were able to raise five times our market cap at the time. We were a really, really small company at the time, $3 million market cap, and we were able to raise 15 million. And that tells you, if these components come together, you can actually make things happen. And we utilize this money really wisely. We have created a very strong portfolio that’s highly differentiated. We have molecules that others have tried to make, but did not succeed.
So, the whole conversation, can any biotech scoop you at one point? We’ve put a lid on this for now, and there’s always somebody who will ultimately succeed, but we have created some distance, some barrier for entry, utilizing that money that came in. And that led us to our recent round.
We just raised $50 million to actually move two of these molecules into clinical development. So, we’re really excited that should happen sometime next year. And as I mentioned, the first time you treat a human being with this molecule is a big milestone.
Host: Surely. I mean, collaborations do act as a catalyst in the process of innovation. So, to speak about balance, you are both the CEO and the Chief Scientific Officer. So, how do you balance the data-driven science with a long-term company strategy? Do these roles ever conflict with each other?
Martin: So, it is challenging to carry or wear both heads, right? Science moves rapidly. CEO role really almost exclusively requires my attention. I have a lot of support on the R&D side. We’ve built up three really brilliant young scientists that have moved on in their career really rapidly over the last two years. But in an ideal situation, you have that full alignment.
But to get to this full alignment between science and long-term vision for the company, you have to constantly align, right? And in a small company, this is much, much easier. Yes, there’s definitely a risk for, hey, this gets misaligned, right?
And as I mentioned, I’m a scientist at heart, but we actually took really good care of hiring people that are also very entrepreneurial, entrepreneurially thinking. And that is kind of what makes us very, very different. It’s literally, you know, biologists want to always dive into the biology, understand everything there is to understand.
What we took really good care of, we hired people that said, do I have to do this experiment? Is it adding value to what we need to do? That is step one of not getting into a conflict between science and, you know, long-term vision for the company.
And this goes through every level of our company all the way up to me. And we have a rule at iBio, no egos in the room. So my idea is usually the best until somebody tells me a better one.
And I’m always open to listen. And this is, I think, the power of a team decision or a team coming together, sharing information that allows us to really fine tune the long-term strategy with what we need to do in the lab. And this alignment that everybody’s in there, it’s funny because company strategies are usually not shared all the way down to, you know, your most junior scientists, but this is kind of a religion almost for us.
Our strategy needs to be understood by everybody in the company. Only then can they align what they do in the lab with our long-term vision. And yes, it’s work to keep that balance.
So it’s not something you set once and forget about it. You have to work on that every day, but in an ideal situation, it’s always aligned. In reality, sometimes it gets a little lopsided, but then we work towards kind of getting it back to where it is.
Sometimes it’s not fun because these are all babies, these programs. And if you lose one, or if you have to give up on one, it can be painful. But this is where, you know, we have to be adults and have to think also as if we were investing in the company, and that helps a lot.
So by being able to bring these people in, we have found a really good way of kind of, you know, checking in on ourselves if we do the right thing.
Host: Perfect. I mean, it’s amazing to hear how all the roles align and complement each other together. Okay. So finally, for those inspired by your journey, what advice would you give to the aspiring scientists or the biotech entrepreneurs wanting to make a difference?
Martin: So first thing I think, find something that you’re passionate about. If you’re passionate about something, it’s so much easier not to get deterred because I often compare biotech with the boxing fight, right? You get sucker punched basically every day, things go bad, and you have to get back up from the ground and, you know, dust yourself off and go back at it.
If you’re not passionate about what you do, you’re not going to have a career in biotech because it will hit you. There will be bad days, but then you have great people around you that will help to lift you up again. So do not let people deter you from what you really want to do. What is important is still to listen, right? Listen to your network, listen to, you know, what other people are telling you. You can be wrong, but you can adapt your plans.
So this balance between not being deterred from your goals, but still integrating kind of the outside voice is really important. And then last but not least, don’t go after small problems. If you want to, you know, build a successful company, solve big problems, find a big problem that you feel needs to be solved, that serves humanity, and then the success will come. If your primary goal in life is to become rich, don’t go in biotech. Go to a bank, run a hedge fund. That’s so much easier.
In biotech, financial success only follows the success of you making an impact on humanity. And I think this is something what people often don’t understand. It cannot be your number one priority. If it is, you’re going to be disappointed. If making people’s lives better, if this is your priority, it’s the best job in the world.
Host: That’s wonderful. I mean, it’s truly a very impactful and heartfelt advice.
Thank you so much for sharing that, Martin. It was truly an inspiring conversation. Your journey, your vision for iBio, and your insights into the AI-driven biotech are truly eye-opening. It was a pleasure having you here.
Thank you so much for your time.
Martin: Thank you so much for having me. Really enjoyed the conversation today.
Host: And to our audience, thank you so much for joining us today. I’m your host, Sannidhi, signing off. See you in the next episode of ExtraMile by YourTechDiet with another extraordinary leader sharing their story. Stay tuned.
