Reprogramming Cancer And Competing With China With Earli's Cyriac Roeding

0:06Ben Comer Welcome back to the Business of Biotech. I'm your host, Ben Comer, Chief Editor at Life Science Leader, and today I'm speaking with Cyriac Roeding, a serial entrepreneur and investor in the fields of technology, AI, and science broadly. He built CBS's mobile unit. And as entrepreneur in residence at Kleiner Perkins, he co-founded Shopkick, a shopping app sold to a Korean telecoms company for $250 million.

0:33Ben Comer He's now CEO and co-founder at Earli, a tech bio company that uses genetic constructs to turn cancer cells against themselves, which I'll get him to explain. Cyriac is a keen observer of technological trends and startups and predicted in 2016 that China would become Silicon Valley's biggest rival. He's an investor in a bunch of companies, uh, including OpenAI and venture capital funds, through his family office, Roeding Ventures.

1:02Ben Comer And I'm excited to have Cyriac on the show today for two primary purposes. Uh, to talk about Earli's story in cancer and what it takes to make a startup succeed. And second, what China's five-year plan in biotech means for biotech companies everywhere else. Cyriac, thank you so much for being here.

1:24Cyriac Roeding Pleasure to be here. I'm I'm looking forward to the conversation.

1:28Ben Comer Me too. And I want to start off uh with a couple of background questions. Uh, as we do here on the business of biotech, I read uh in your bio, Cyriac, that you sold your first computer program in Germany when you were 15 years old. Uh, what was that program? And and is that you know when you were first bit by the entrepreneurial bug?

1:49Ben Comer Or did that come later?

1:51Cyriac Roeding Well, actually, that probably came before then. But I was um I was working for this newspaper publishing house doing my spring break, I think, at uh at my school. And I was uh assigned this really, it was my first job, and it was I was assigned this uh this this really uh tedious job of looking whether in a newspaper whether uh an ad had appeared or not.

2:17Cyriac Roeding And if if so, you had to scratch it out. And and and I was sitting there the first day doing this for eight hours, a stack of paper. And at the end of the first day, I was ready to jump off a wall. And I thought, well, that would probably be a bad way to start the first job. So I went through with it for three weeks, and I thought, wow, man, if this is how it is to go to work, I'd rather go back to school.

2:45Cyriac Roeding First time I wanted to do that, and I went to my my boss and I said, Why don't you automate this? And he said, What do you mean, automate this? Mind you, this is like uh stone ages of computing. And he didn't really know. And I said, You don't use computers to do this this whole thing. So then uh a few months later, the phone rings and he calls me and he says, Why don't you, you know, remember you said something about automation?

3:12Cyriac Roeding I said, Yeah, sure. He said, Why don't you come back and do that? So I I picked up my own computer, went there and started coding. And by the end of the summer, I had written the code for him and sold the program to him. And I guess I basically got rid of my own job.

3:33Ben Comer Well, yeah, somebody uh was probably thankful that they didn't have to do that job after you. So you must have been into coding at a pretty early age. You you kind of took to computers early on.

3:42Cyriac Roeding Yeah, I think I started coding at like 11, uh, teaching myself how to do it. And yeah, I found it fascinating that you could literally take control of a machine and tell it exactly what to do, and then it would spit out exactly what you asked it to do over and over again. And and that was my first thing that really fascinated me.

4:03Cyriac Roeding And the other thing that I found to this day still absolutely impressive was this idea there was this like small robot self-made produced out of little pieces of equipment at my school and a computer attached to it, and it couldn't do much, but you could program it, how to move. And I will never forget this.

4:29Cyriac Roeding When I first saw this and I started using it, and I started coding it the idea that I could use software to manipulate the actual real world, the world we live in, that a piece of code would change how the real world would look like, because something was actually moving and being moved, that never left me.

4:53Cyriac Roeding And to this day, everything I've done since in my startups has been on the intersection of the physical world of atoms and the digital world of bits, and where the two meet, that's why I'm at home.

5:09Ben Comer Interesting. Um, I I also read that in college you were a talk show host and a reporter, and I was curious, you know, what your beat was, like what you talked about.

5:20Cyriac Roeding Well, that was that I actually love the radio, and I like um I also like media in general. So when I was in college, I could not take the idea that I would walk in in the morning when it was dark, and I walked out in the evening when it was dark again. It's like, what the heck am I doing? So there was no attendance requirements uh at my college in Germany, so I decided to do this offline sometime and rather work nights at a radio station.

5:47Cyriac Roeding And first I was a reporter who was literally driving around reporting from their life from the road. Then I did reporting on political topics uh and did uh did stories about them. And then after a while, I literally be I became an a DJ host at night, uh a radio host. And I I was supposed to only do a music show, but over time, you know, I turned this into a talk show more and more with music and interruptions, and I ended up having my own reporters out on the street as well, talking to people, and it became a very interesting combination.

6:28Cyriac Roeding And you know, for a while I flew under the radar that I actually changed the whole format of the show. Uh, and at some point I had to answer to my bosses about it. But it was a great, it was a great experience.

6:39Ben Comer Well, I imagine all of that experience in storytelling is helpful, right? As an entrepreneur and as an investor, uh, I that it's something that I don't talk about a lot on this podcast, but um it, you know, as someone who's worked in media my entire professional career, you know, the importance of storytelling, regardless of whether it's reporting on news, talking about a company, talking about a product, whatever it is, it's an extra the narrative, you know, it's a very powerful form.

7:09Ben Comer And people who are good at telling stories, uh, I think it's you know, it's one of those tools for success. I mean, have you found that to be true?

7:18Cyriac Roeding That is absolutely true. I I just think we need to clarify something for the listeners. Storytelling doesn't mean you make things up. Yeah. You you can, you can make you can tell a story that's completely fictional, but you can also tell a story completely based on facts. But your job is to connect the dots and explain it in a way that first it can be understood, but beyond that, it's actually, I don't know if you do you know the three stages of of actually getting people fully convinced of something.

7:55Ben Comer No, I don't know, I don't think I do.

7:57Cyriac Roeding So there are three levels, right? The first one is um basically it's it's called persuasion, where you basically tell somebody, hey, can you just do this? And I'm just gonna do it because I kind of owe you something, or I think you know, I just want to be nice to you, but you're not really convinced of it.

8:14Cyriac Roeding The second level is conviction. This is where your brain, your mind tells you, okay, I think this makes a lot of sense, I should do this. But the third level is enrollment, and enrollment means both your mind and your heart are fully engaged and are bought in. And when you reach that, anything becomes possible.

8:42Cyriac Roeding The war the walls tumble and people become engaged, and they actually take ownership of the thought, of the mission. That's what we're what we're in the business of doing when we are building something. When you truly want to build and you don't want to build it all alone by yourself, it's absolutely critical to not just persuade or convince people.

9:11Cyriac Roeding You need to enroll them. And once you do that, some really powerful forces begin to work where you don't have to be in officially in charge anymore. People take ownership, they run. And that's beautiful to see. And I think what we call storytelling is essentially a way to enroll people, but not by making stuff up, especially in biotech, that's really important.

9:36Cyriac Roeding You can't just make stuff up. Everything has to be based and grounded in facts, in hypotheses that are declared as such, and in results accordingly. And once you have that, you can then combine it into a flow that people understand, not only intellectually, but also with their with their minds and their hearts, and bring that together and get people uh on the mission.

10:00Cyriac Roeding And then you are really on to something.

10:02Ben Comer That makes uh a lot of sense to me. Um, it seems like the you know, like the the emotional aspect or the heart part would be the most difficult. You know, you can get the rational part, but but how do you how do you get the the heart in? Yeah, I guess it depends.

10:18Cyriac Roeding Yeah. I I think the heart comes from recognizing that regardless whoever is listening or watching you, they're all humans like you and me. Yeah. I always find it amazing that people when they talk to really important people, and there's all this like uh the deference, and all of that is building up walls.

10:42Cyriac Roeding But when you break down the walls, we're just humans, we're just sitting in front of each other, we're talking with each other. You have your story, your personal story. And I'm sure there's a lot of interesting things that happen in your life. And when you understand, uh believe that every person has something really interesting to tell and to share, then there are a lot of connection points.

11:05Cyriac Roeding How you can actually build a bridge into their world? Cancer is one of those, unfortunately.

11:13Ben Comer Well, let's uh let's talk about Earli. Um, biotech can be a turnoff, I think, for some Silicon Valley investors, primarily due to the long RD timelines, regulations. Um, before co-founding Earli, what did you have previous experiences with biotech or with the life sciences sector? Absolutely none. None.

11:36Ben Comer Okay. Well, let's maybe you can tell me the story then of how you got connected uh with the late Dr. Sam Gambhir at Stanford, um, and how that turned into Earli.

11:49Cyriac Roeding Well, by background, I'm an industrial engineer. I'm not a biologist, and I'm not a scientist. So, what happened? I told you before I love to be on the intersection of the physical atom and the digital bit, right? Right. And my last company did that in mobile. And mobile is essentially like a a mobile phone is like a digital overlay on top of the real world.

12:13Cyriac Roeding And that company uh was doing that in in the retail world, and it was super interesting from that point of view. So then the next company I wanted to build was on the intersection of biology and software, also known as AI. And that intersection, I believe, was going to change fundamentally how we live.

12:35Cyriac Roeding Because if you think about AI meaning the real world, this is called physical AI. And physical AI comes in two forms. One is one that everybody knows and thinks of immediately, which is called robotics. Robots are essentially the non-living atom combining with the digital bit with AI. Right. And then there is a second arm, and that's the living atom combining with the digital bit.

13:04Cyriac Roeding And that's called biology, meets AI. I call it BioAI. And BioAI is an intersection of, to me, the ultimate frontier. It's the living atom, it's our own bodies. And when you can start manipulating that world, you're talking about fundamental changes to how we treat diseases. Perhaps one day we will be able to treat many diseases by using software.

13:32Cyriac Roeding And that's how was my first premise. So now, then I was looking for my next thing to do in that field. And I was looking at many ideas. I think it was over 200 in total. I have a very dense filter when it comes to ideas because most of them are either features or they're products, but they're not company ideas.

13:52Cyriac Roeding So finding a company idea is rare. I also believe there is an ocean of good ideas out there, but there's a dearth of outstanding ones, and only outstanding ideas are worth pursuing. It's very hard to find. So what happened was I looked everywhere, and since I live only five miles from Stanford University, I started stumbling around the campus.

14:13Cyriac Roeding And I talked to all these scientists there. I talked to over 100 of them, I think. And almost every single one was telling me they were working on the world's most important idea. Sure. By definition, 99 of it out of 100 were wrong. Or all of all 100 were wrong, right? Right. The problem was I didn't know which one was which, because I'm not a scientist.

14:40Cyriac Roeding So that was very frustrating. I could probably tell the difference between a bad idea and a good idea, but I couldn't tell the difference between a good idea and an outstanding idea. And only outstanding ones are worth pursuing. So then I got very lucky. I was living with my wife and my at the time uh tiny micro children in another person's house that we had rented because our own house was being remodeled.

15:12Cyriac Roeding And so we got their their subscription magazines and all the stuff that they usually order, but they weren't there. So Thanksgiving Day comes around, and I was actually at the time very disappointed with myself because I had spent the whole year looking for my next thing and I still hadn't found it. And when Thanksgiving comes around, it's usually the end of the year in the in the US, there's not much happening anymore.

15:37Cyriac Roeding So I thought literally, I was sitting there thinking, I I absolutely didn't do it. The year's over and I have nothing. And while I was sitting there thinking that, the door opens, and my wife Angel comes in with a magazine in her hand. She says, Cyriac, this just came. I think you should look at this.

15:58Cyriac Roeding She hands me the magazine, and it was from Stanford University, a medical school, and it was the cover story, which was called And Yet You Try. And that story was about Sam Gambhir, one of the world's most preeminent early cancer detection and treatment experts in the world at Stanford University, and about how he lost his only child to cancer at age 16.

16:31Cyriac Roeding And I read this story and I didn't know him. But I felt a gripping sense of grief just reading this. But I was also very inspired by the work he was doing on early cancer detection and treatment, including his son, who died, by the way, who helped him in the lab at age 15. That's how brilliant he was.

16:54Cyriac Roeding So I made a decision to send him a cold email that day. And I said something along the lines of I can only imagine how hard this day is for you today, this holiday. But I read your story, I was very touched, but I was also very inspired by the work you're doing. I'm not a scientist, I'm not a biologist, I'm a serial entrepreneur, I want to build my next company.

17:24Cyriac Roeding I'd like to meet you. Two months later, he finally found time to meet with me on a Saturday morning at a small restaurant in Portola Valley, which is where I live. It's like 4,000 people close to Palo Alto and Stanford, right? Well, it turns out he also lived in Portola Valley, and we had no idea we were neighbors three minutes away from each other.

17:45Cyriac Roeding Oh wow. So my first question, and this gets back to your question. My first question to Sam was I'm not fishing for compliments when I ask you this. Should someone with my background, or better lack thereof, bother the world of biology with my presence? And he looked at me and he said, There, I have two answers for you.

18:11Cyriac Roeding The first answer is the world of biology needs more people like you, not fewer people like you. We have a lot of experts and very few generalists. We even we have even fewer people who know how to build a startup company. And the second part of the answer is the world of biology will always find a way to screw you over.

18:32Cyriac Roeding It will take longer than you want and it will be harder than you want. I took the second part of the answer as the cost of doing business and and kept moving. And before I knew it, every other Saturday I was sitting at his kitchen table in his own house where he was trying to teach me biology, largely in vain, but I give him credit for trying it.

18:51Cyriac Roeding And then he started introducing me to a lot of people. And I met all these scientists, and unfortunately, you know, the odds of finding the right person and the right idea combined is like probably one out of 10,000 times one out of 10,000. You can do the math. So after three months, I went back to Sam and I said, I'm so sorry, Sam.

19:09Cyriac Roeding I still don't have it. May I ask you a simple question? You have seen hundreds of ideas in your own lab, outside of your own lab, probably just in the last year alone. Which one has the highest potential of all? And he looked at me in his scientist underselling type of way. And he said, There is this one idea that I believe has some real potential.

19:41Cyriac Roeding And that became a really wow.

19:44Ben Comer Okay. And I, you know, getting back to AI, it's a major tool that Earli uses, I think, to create these genetic switches that that turn cancer cells into um immune therapy factories, so to speak. Can you can you describe like what that process looks like, how how AI is used, and then how the genetic constructs work, you know, when they're delivered to cancer cells, with the understanding that you're not a scientist, neither am I.

20:14Ben Comer So we can keep it high level. But give me a sense of you know how the how AI is applied and how you know the the eventual candidate that I assume you'll be moving into to clinical trials will actually work.

20:27Cyriac Roeding Great. So that we are two two non-biotech people talking biotechs. That's awesome. Exactly. Yeah. Okay. So let me uh first explain why we're doing what we're doing. What are we trying to accomplish here? Because who cares what we do when there is no real problem to solve? Sure. The main problem we're trying to attack is what we believe is the number one problem of all cancer drugs, which is when they go into the body, they cause major toxicities.

20:56Cyriac Roeding These toxicities, also known as side effects, are not just uncomfortable for patients, like hair falling out and all of these things, but they're actually dose limiting. It means you can't give the patient as much of the drug as you actually want to because there are too many side effects. And these are almost cynically called on-target off-tumor effects.

21:19Cyriac Roeding It means your drug hits its target, but unfortunately, your target is not just on the cancer cell, it's also on healthy cells. And that drives uh toxicities you don't want. At the same time, what happens is you inject the drug, it goes everywhere in your body, you have five liters of blood volume. And by the time your drug reaches the tumor, there's this much left of the initial drug, and this much elsewhere in the body where you really don't want it.

21:48Cyriac Roeding So, how could you change the whole game? What if you could flip it? What if you could no longer try to target it to tumors and hope for the best that it really Only activates there. What if you could inject it, but it's only actually produced, made into a drug where it is actually needed. What if the drug could be produced in the tumor by the tumor against the tumor?

22:27Cyriac Roeding If you could do that, then you would suddenly have very limited systemic toxicity. Right. If you can keep it in the tumor, you could have very high local activity inside the tumor where you want it. You could even potentially turn cold tumors hot, meaning make them susceptible for immunooncology drugs by inflaming them from the inside.

22:53Cyriac Roeding Now, that would be a significant deal because the biggest drugs, the most powerful immunodrugs right now are these checkpoint inhibitor drugs. I don't know if everybody knows about them. So very briefly, like it's the biggest drugs called K-Truder per Merck, right? It's I think it produces like 35 billion in revenue a year.

23:15Cyriac Roeding But all of these checkpoint inhibitor drugs have one major problem. They only activate in about 35% of all patients. Meaning two-thirds of all patients show no response at all. But these are very powerful when they work, they're amazing drugs. So how could you turn this up, this percentage? And one of the major problems is that these tumors are cold, they're not responding at all to immune activity.

23:43Cyriac Roeding And that's where Earli could potentially drive a major act, major impact. So now how do we do it? We use AI to create the first fully synthetic genetic switches made out of DNA code that get injected into your body. They go everywhere in your body, but they're designed to only turn on in cancer cells.

24:10Cyriac Roeding And when they turn on, they program the cancer cell to become their own drug factory to make a drug against themselves by activating the immune system locally to attack and kill the cancer. So now step back for a moment and imagine this actually worked. If you could have a genetic switch that really only turns on in cancer cells, and then you can program these cells to make drugs, what would that unlock that we cannot do today?

24:46Cyriac Roeding And I think the answer is you could make extremely potent therapies that are otherwise today too potent. They're too toxic. But if you could make them only in the tumor and keep them there, you could be much more toxic against the tumor, which is what you want because you want to kill it. And that's precisely what Earli is working on.

25:14Cyriac Roeding And that's why we chose the poster child of toxicity. You know, for decades, people have been wanting to use this specific drug, interleukin 12, a cytokine that activates the immune system to attack. IL-12 has been loved by pharma companies for its toxicity against cancer, but also hated by pharma companies for its toxicity against the rest of the body.

25:41Cyriac Roeding Right. And that's why it's the perfect poster child for Earli technology as the first drug that we're taking to the clinic. And if all goes well, we want to dose the first human patients in late-stage lung cancer by the summer of next year.

25:56Ben Comer Excellent. And so the uh the genetic constructs are delivered to tissues beyond the liver using cancer lipid nanoparticles. Are you making those uh in-house?

26:08Cyriac Roeding Yes. So that's the second big thing. You just brought up the key second requirement, requirement to make this work. If you inject DNA freely into the body, the body immediately destroys it. The immune system discovers there's a foreign object and it destroys the whole thing. So you need to somehow wrap it and hide it from the immune system.

26:31Cyriac Roeding And that is being done by a container. And that container is a lipid nanoparticle. So for the for anyone on the call uh here who on the podcast who doesn't know what a lipid nanoparticle is, most of us have had it in their own bodies already as part of the COVID vaccine, because the COVID vaccine was mRNA that was transported in a lipid nanoparticle.

26:56Cyriac Roeding So now, lipid nanoparticles are great FedEx trucks that transport DNA or mnRA, mRNA cargo through the body. Here's one problem, though. The FedEx truck knows only one destination it takes you to the liver. Right. And nowhere else, there's it's always the same route. And within 10 minutes, your cargo gets flushed out into the liver and out of the liver.

27:25Cyriac Roeding And that was it. And that's why is that? Because your liver is responsible for getting rid of anything that doesn't belong in your body, from alcohol to lipid nanoparticles. So if you want to treat any cancer except liver cancer, what are you going to do? So we had to take on the challenge to design completely new lipid nanoparticles that can take you to other parts of the body.

27:52Cyriac Roeding We intentionally started with lung. Lung cancer is the biggest killer by far. It kills twice as many people as the second biggest, which is colorectal cancer. So, how do you get to a lung tumor with a lipid nanoparticle? This is called the billion-dollar bottleneck in the industry. Why? Because if you solved it, it would unlock so many therapies.

28:18Cyriac Roeding From gene therapy to what Earli is doing, cancer therapies. But nobody has done it. So what people have tried is to stealthify these lipid nanoparticles from the immune system. So the immune system wouldn't destroy them immediately or flush them and transport them to the liver and get rid of them. The problem then is if you stealthify them, then they are so stealthy, so compact, that they also don't get into a tumor anymore.

28:51Cyriac Roeding Right. That's called the so-called peg dilemma. Peg because PEG, because PEG is this the stealth coat around the lipid nanoparticle. So you have this dilemma. You got a really good peg that hides it, but you can't get in anymore. So how do we solve this problem? Was the question. Most companies have solved it or tried to solve it, I should say, because it hasn't really worked so far, by using a different ionizable lipid in the inside of the tumor, uh, of the LNP.

29:23Cyriac Roeding However, unfortunately, um, there's a lot of IP, there are a lot of players involved, but it hasn't really worked. So Earli went down a completely different route. We looked at the formulations, meaning the composition of all the different materials that make up an LNP. We created a whole assembly line, a factory.

29:44Cyriac Roeding We tested thousands of them, many thousands in vitro, and I think close to a thousand and different ones in mice. So we did a lot of work over years. It took six years of work. And now I'm really excited to say we developed this new technology called FlexCore, which is a very different formulation and composition of the P and of the LNPs.

30:10Cyriac Roeding And it allows us not only to get to other tissues, but to get into tumors, because that's another challenge. And once you're in the tumor, you also have to get into the nucleus of the tumor because that's where the DNA cargo has to go. And now we're at a point where we're our LNPs are like, I think, 22 times higher transfecting these cells, these tumor cells, than what's currently market standard.

30:36Cyriac Roeding And with that kind of power, we can now actually deliver our genetic switches into the place where they need to go, it to the nucleus of the cancer cell and make sure they only turn on there. And that's where AI comes into play. Because the genetic switches, they are literally, there are trillions of possible combinations, man.

31:00Cyriac Roeding How you can design these. I'm not making this number up, it's it's trillions. And that's because what we found was when you try to build it the traditional way, it won't work. Um, maybe I should explain. The traditional way is normally by using endogenous pieces, meaning what nature provides. You find it somewhere.

31:20Cyriac Roeding We were looking for this. And what Stanford did in the originally was to use an endogenous promoter. A promoter is like a switch in in the genetic world. And we it worked really well on cell lines in the lab that never changed. They're always the same for 50 years, right? Lung cancer cell lines. Wonderful results, 95% activation, look great.

31:44Cyriac Roeding Well, until we actually took real cancer patient samples out of the operating room at Stanford, they they very nicely, uh graciously helped us get it get these samples. And we tested it on these real human samples. And every person's cancer is a little bit different because these are mutations, right?

32:06Cyriac Roeding Right. And the result was we went from 95 to 26% activation. It was so bad that it would have been better for you to bet against our result than for our result. So then we thought, what are we gonna do next? And the natural next step is to think about okay, what causes cancer? It's mutations of genetic mutations.

32:29Cyriac Roeding So now you're gonna look at all the mutations that can cause lung cancer. Well, guess what? There's over a hundred of them that can cause lung cancer. And that's just non-small cell lung cancer. There are other types. So if you use a hundred different mutations and you interrogate each one of them with a genetic switch, your whole switch, total switch, becomes so large that you can't get into the nucleus of the cancer cell anymore.

32:53Cyriac Roeding So either you're accurate, but you can't get in, or you can get in, but you're no longer accurate. So that was our situation. Rock and a hard place. You choose. So then the next logical step is to think about the 80-20 rule. Where maybe 20% of the mutations cause 80% of the problem. Unfortunately, that's not true either.

33:14Cyriac Roeding And so now what? And this is where I've learned once again the the the wisdom of this of this uh statement the most important step to solving a really hard problem is to ask the right question. And I think it took us over a year to ask the right question. We had literally which was what the question was if we can't interrogate every single mutation that can cause cancer, what if we didn't interrogate any one of them?

33:52Cyriac Roeding None. What if instead of looking for the upstream causes, the mutations that cause cancer? What if we looked at the downstream consequences of such a mutation? The results, the symptoms. Okay, and those are called the hallmarks of cancer. Panahan and Weinberg described them over 20 years ago in the literature.

34:18Cyriac Roeding These are rapid cell proliferation, lack of cell death, rapid blood blood vessel construction, angiogenesis, and so on. And when you look at these results, you don't have to look at all the causes that can cause it. You only look at the results. And then we used machine learning on thousands of fully sequenced cancer samples from the lab, from real patients, and sequenced them and looked at them and analyzed them with machine learning for patterns.

34:50Cyriac Roeding And we found these master transcription factor binding sites that were dysregulated in cancer. And we used a lot of techniques to filter out the right ones. So that was a big step, but it was only the beginning. Because the next step is to say, okay, we know some of these transcription factors. When you combine them, and you not only look at one, you can't just look at one, it will not work, it won't be specific.

35:14Cyriac Roeding But when you look combine them and look at multiple, you can get a very high specificity rate. But when you do that, you have many ways to space them directionality and order, and that leads to trillions of combos. So then we applied some other logic to trim that down. Now you need to test it in the wet lab.

35:35Cyriac Roeding And what do you do? Normally you can sequence like 50 a week in your lab. Well, if you have millions of possible combos, you're going to sit here in 100 years. Right. So then we went to the University of Austin in Texas and found the super expert at building massively parallel reporter essays for the wet lab.

35:56Cyriac Roeding We hired her, brought her on board. Took nine months to build this very special essay that allows us to genetically barcode these sequences and test 250,000 in one batch in the wet lab. Wow. Okay. We went from 50 to 250,000. Now we were talking. When you test these sequences, 99% of them are useless.

36:23Cyriac Roeding It's like a power curve. They are not much better expressing than the rest of it. But then there are some outliers, and these outliers are really interesting. You can take those and test them in mice and so on. That used to be the end of the process that was technically possible. Today, it's the beginning of the process.

36:44Cyriac Roeding Because we're taking this very high-quality data and turning into the training data for an AI large language model. That proprietary data flows into an LLM that's trained on DNA. And then we build a proprietary oracle, which is a prediction model, to say which of these news sequences that get spit out by the LLM are going to be interesting or not interesting.

37:15Cyriac Roeding By the way, 99% are not interesting. But even in the 1% that we do select, most of it is not the useful. So we filter as much as we can. The best players we're putting back into the wet lab that I just described. We run another cycle. The best ones out of those go into mice, and all the data that comes out goes back into the LLM for the next round.

37:39Cyriac Roeding And with every turn of this flywheel, it gets stronger and smarter and stronger and smarter. And what we're essentially doing with this is turn DNA into a programmable language. Just like Alpha Fold made proteins programmable. We make DNA programmable. And that is so exciting that we can now, we now are working on the next generation of this with some of the largest um top AI labs because this is becoming a top um uh uh quality tool for developing disease-based switches.

38:25Cyriac Roeding And when you can do that for cancer, maybe for autoimmune diseases, you now control the disease cell. And then think about this what this means. If we could stop chasing the latest mutation, by definition, it's always two or three steps ahead of us. Always we're always behind. We have to be, otherwise, we don't know what to chase.

38:47Cyriac Roeding But if we control the disease cell itself and can tell it what to do, we are suddenly taking over. We can take control and tell the disease cell what to do, the cancer cell what to do, rather than it doing what it wants. That's what we're really doing here.

39:07Ben Comer That's uh that is really exciting, Cyriac. And you said uh first in human next summer. Is that right? That's the plan.

39:14Cyriac Roeding We want to we are already manufacturing the drug. It's already in manufacturing. We've already had an FDA meeting. So we are moving really rapidly towards dosing the first humans. This will be uh uh uh our first um time we're taking it from mice and monkeys to humans. We've done a lot of NHP work, non-human primates to show safety.

39:39Cyriac Roeding It's exceptionally safe. This is one of the big problems with lipid nanoparticles containing DNA cargo that often they produce high fevers and so on. And our monkeys are not even showing elevated, mildly, not even mildly elevated temperature levels, because we have engineered this out. You also have to focus on redosability.

39:58Cyriac Roeding A lot of these um LNPs are not redosable because that means they're they're effective the first time, but not the second and third time. So we've done that. The other big thing is a lot of people when they use lipid nanoparticles, they put a moiety on it. That means a connector to the cancer cell. That looks really great in an animal model that has that exact connector, that exact moiety, the counterpart, the data looks great.

40:24Cyriac Roeding The problem is once you go to heterogeneous human cancers where every cancer is a little different, the whole thing falls apart. So we took at least two years more work. It was a very hard road to do this without any moieties. There's zero moiety on the LMP. Everything is very clean and it's built to go as to as many places as possible.

40:44Cyriac Roeding But the genetic switch makes sure it only turns on in cancer.

40:49Ben Comer That's fascinating. And I I think it'd be very easy for us to stay on Earli for the rest of this conversation, but I I have to get to China, so I'm gonna switch switch to Cyriac. Um, and I wanted to ask you about uh the five-year plan that was recently announced by China, 2026 to 2030. Um, through the plan, the Chinese government is designating biotech an emerging pillar industry, targeting over 520 billion in sector revenue by 2030, uh, other goals, um, producing five global blockbusters, a blockbuster being a drug with over a billion dollars in revenue annually, and then 25, a 25% share of the world's first-in-class drugs.

41:34Ben Comer Uh, a bold, a bold agenda, a bold five-year agenda from China. Um, my question for you, Cyriac, is you know, what are the the biggest implications of this plan for biotech companies outside of China?

41:49Cyriac Roeding Yeah, first off, I think this plan will be absolutely fulfilled. You do. I think uh, and there's no to me, I have no doubt about it, because the momentum is so high. You know, some people have always thought of China as the copycat of the world. Right. That is so outdated. It was already outdated in 2016 when I wrote my first piece on this in the tech world.

42:12Cyriac Roeding People were talking about China as the cloning, the cloning engine. You know, at the time it was already visible in in traditional tech hard and hardware, software and hardware, that China was innovating, not just copying. At the time, they already had, I think, 70% share of home uh homemade uh uh home use drones by GJI.

42:35Cyriac Roeding And now you're seeing the same thing 10 years later in in biotech. We're way beyond the idea of just cheap knockoffs. China is innovating. China is in an innovation powerhouse, and that is not a bad thing for the world. It's a great thing that there is another place besides the United States where deep innovation happens on biotech in order to help cure or treat the worst diseases of humanity.

43:02Cyriac Roeding There is nothing nothing bad to say about this. The question is, what does America think about? Somebody else taking over the lead in innovation in biotech and developing the world's most innovative and powerful drugs. What do we think about this? And I want to I want to just turn this into a question that's more personal.

43:30Cyriac Roeding Imagine we get to a point where the majority of the most powerful new drugs come out of China. And for some reason or another, may it be with intent or without intent. Let's just say out without any intent, there is a shortage of production. And we've seen this before. It was called the pandemic with COVID vaccine.

43:55Ben Comer Yeah, absolutely.

43:56Cyriac Roeding What do you think every country did? Every country served its own citizens first, right? As you would expect. So, what do you what would happen and how would we feel personally about this? Imagine um, God forbid, one of us gets diagnosed with cancer. And the most powerful cancer drug against that specific cancer comes out of China, but there is a shortage.

44:25Cyriac Roeding Do you think you and I, you or I will get this drug? No. So now there might be other situations where it's geopolitical. Like, for example, we've known this, we've seen this with um uh with with with rare earths. This is a constant battle between China and the US because China has more rare earth than than the US.

44:47Cyriac Roeding Um, and I and I don't actually expect China to do this with drugs because that is very cynical uh uh uh to do this on purpose. Um but there might be simply a shortage of supply. So now I'm asking everyone do we want to be dependent on another country to develop the world's most leading edge drugs? And most people would answer this question with, of course not.

45:14Cyriac Roeding But the problem is we're not acting the way that we're talking. We are facilitating just what I just point painted as a picture, is what we are currently preparing to happen. For the first time ever, since the biotech industry has emerged 60 years ago, we are risking losing the innovation leadership.

45:39Cyriac Roeding And why is this happening? Yes, there are some regulatory issues. We're slower with clinical trials and certain other regulatory issues in the approval process, and we need to change that. But that is not the number one problem. The number one problem is believe it or not, money. Really it's funding.

46:01Cyriac Roeding Let me explain. You mentioned it at the very beginning, Ben, when you talked about biotech. Biotech has long cycles, right? Right. It takes 12 to 15 years to develop a drug from start to finish. And in the US, we have a completely free market, which is awesome. I'm a big believer in capitalism. But there are certain situations where the system is failing us that is otherwise so efficient.

46:29Cyriac Roeding And that is called boom and bust cycles in biotech. The boom and bust cycles in biotech are more extreme than I believe in any other innovation-based industry.

46:41Ben Comer Yeah, and we're just coming out of a long bus cycle, right? Hopefully coming out. Yeah. Thank you.

46:47Cyriac Roeding So the last four years, not this year, but the prior four years, have been the biggest bust, I think, in the last over 20 years. Right. So now we're this happened after the pandemic. It was sort of like an aftershock of this overhype situation, and the all the money went away. But now let me give you three major facts that make paint this picture how extreme this is right now.

47:12Cyriac Roeding Number one, guess how many percent of venture capital dollars this year are going to AI versus to other parts of the of innovation? Oh, I've 10 to 1. 87.5%. Wow. Okay. That means everyone else gets 12 and a half. Now, fortunately, Earli is on the intersection of bio and AI. So that's great. It's special, but not every biotech is, most biotechs aren't.

47:41Cyriac Roeding So 12.5% left for everything else. That's number one. Number two. Five years ago, less than five percent of the licensing dollars from pharma companies went to China. This year, it will be two-thirds.

48:03Ben Comer Yeah.

48:05Cyriac Roeding Two-thirds. One third is left for US firms or anywhere else in the world outside of China. That is a dramatic shift. What are they buying? They're not just buying cheap knockoffs, they're buying innovation. Right. So, how come there's so much innovation to buy? And the answer is a sustained, flat, much flatter curve of funding by strong support, largely from not only from private sources, but also from the government.

48:41Cyriac Roeding It's not that they invested in every startup, they also develop ways to buy things from them. There is there there are all kinds of mechanisms to support these prioritized industries. You mentioned the new five-year plan. This is the second five-year plan in which biotech has been listed, not the first.

48:59Cyriac Roeding Yeah. And with every biotech uh state stating biotech as a major pillar together with AI and fusion technology and and robotics and so on, this means the whole country is behind it. So now we don't want to shift to communism. Last time I checked. I don't. I guessed you don't. So what are we gonna do?

49:21Cyriac Roeding What are we gonna do about this? And the answer to this question, I believe, needs to be a sovereign biotech wealth fund. What is that? What's the sovereign wealth fund? A sovereign wealth fund is money from the government, from taxpayers that is designed and allocated to be invested in innovative companies.

49:48Cyriac Roeding Exactly between this phase of preclinical to clinical, because that's where most of the cut shutdown happens.

49:54Ben Comer The valley of death, that's right.

49:56Cyriac Roeding The valley of death. It's like you get some early funding, you show some promise, but now you need money to get it to the clinic. And you know this beautiful catch 22 statement that every that everybody hears in funding in biotech. When you have clinical data, there will be money. When you have money, there will be clinical data.

50:15Cyriac Roeding Right. So to overcome this valley of death, we need a concerted effort to flatten out this boom and bust cycle. Right now, the remaining 12.5% that might be left over for biotech, if if at all, they go to late-stage companies because they are also desperate for cash after the last uh the last bust cycle.

50:34Cyriac Roeding So they're cheaper and the VCs have lower risk. There was an article in the Financial Times, an opinion piece a few weeks ago. The headline was biotech VCs are becoming like bankers, and it will probably all come to hurt us. Because there's not enough risk taken anymore. And they don't have to because there's so much desperation for starvation for cash at the later stage, too.

50:56Cyriac Roeding So at sovereign wealth fund can fix this. Let me explain how this works because it's actually a beautiful model. Some people might say, well, this only works when you have so much money, like oil, oil-rich industries or company uh countries like the United Arab Emirates or Saudi Arabia or Norway that pump it out of the North Sea, right?

51:16Cyriac Roeding We don't have that. So what are we gonna do? Well, look to Singapore. Singapore started its sovereign wealth fund with no oil and no other resources in the late 80s. Today, they have $1.3 trillion in these funds. They cannot add as much as they're earning automatically every year just by having this much money.

51:43Cyriac Roeding It's like a self-running train. You can't stop it. And when you invest in these high-quality, high-innovation companies, many of them will go belly up and won't make it. But a few moonshots will become so successful that if this system is properly handled, like it is in Singapore, this will more than feed itself.

52:07Cyriac Roeding And the money will go back to taxpayers or to future innovation for the next cycle of innovation to keep the company, the country, at the top edge of innovation. That's what a sovereign wealth fund looks like. And of all industries of innovation, I think biotech needs it the most. And this would we're risking losing our leadership, and we shouldn't do that.

52:30Ben Comer And this is something that would need to be taken up by the federal government, approved by Congress to get started, right? That is correct.

52:37Cyriac Roeding Interestingly enough, the Trump administration actually Trump himself signed an executive order about a year and a half ago to start a sovereign wealth fund. And strangely, but happens very rarely in this country, the other side of the aisle wants this as well. Even Bernie Sanders, except they are they are have no agreement where the money should come from.

53:05Cyriac Roeding Okay. And that's where this is stuck right now. To decide, yes, we want it. We both understand, both aisles understand the benefit. How do we fund it? And there are some interesting proposals that are being discussed, like taking a small percentage of the largest AI companies in order to fund it, who have actually some of them even expressed an openness towards that idea.

53:32Cyriac Roeding Um, but that there is no agreement on this yet. But I guess what we first need to create is a common understanding of how big this problem is. When you suddenly see two-thirds of all licensing dollars go into China from global pharma companies, up from less than 5% five years ago, you know there is an issue.

53:55 Yeah.

53:56Cyriac Roeding When you see some of the best biotech firms go belly up right before they can hit the humans, after having spent years of research, you know what's happening. Some of the potentially best drugs of the future just got buried. I don't think we want that. No, do we?

54:13Ben Comer No, I'm glad you mentioned the sovereign wealth fund. That's a really interesting idea. I uh one follow-up question to that I have is who would be the administers of it? Like who would who would decide which companies get investment? Would it be the party and the party in power?

54:27Cyriac Roeding You know, would it be a great question, Ben? And the answer is hell no. Okay. It has to be an independent body of absolute professional top life science investors who know exactly what they're doing.

54:44Ben Comer And those members couldn't be swapped in and out depending on who's in power.

54:48Cyriac Roeding No, cannot. And and also let me let me also make another suggestion, which I think would be really helpful, especially in the beginning. My suggestion would be that in the beginning, for the first few years, until you see it's going well, the sovereign wealth fund is only allowed to co-invest with the top 10% best performing VCs of the world.

55:12Cyriac Roeding Interesting.

55:13Ben Comer Okay.

55:13Cyriac Roeding So you take a the percentile of the top best performing ones, whether they're American or non-American, doesn't matter. You just want the best investors of the world, and whatever they invest in, the sovereign wealth fund can opt to co-invest into. Okay. And by doing so, you have a really good shot at not wasting taxpayers' money and coming out on top in 10 to 20 years from now.

55:45Cyriac Roeding And then it starts compounding, and that's what you want to see.

55:49Ben Comer I uh I want to be respectful of your time, Cyriac. We're coming to the end here. I have one final question that I just have to ask you because I you're close to this. Uh, OpenAI announced you know yesterday that they would not roll out their newest model due to safety concerns. And I'm curious if you agree with that decision and and you know how much you know you think this idea of you know AI potentially you know eliminating the human species in 10 or 15 years, how much of that is hyperbole versus an actual possibility?

56:25Cyriac Roeding Well, it's definitely a possibility, um, but I think we have to be careful to understand we don't know what we don't know. There is a lot of different scenarios this could play out, and it's very wise to be careful. And yet, at the same time, and both can be true. I would also highly encourage everyone to prepare for a non-doomsday AI scenario.

56:55Cyriac Roeding Yeah, right, where we see AI literally help not only treat but potentially cure some of the worst diseases humanity has experienced. And where we design a future that has more abundance than it has today. In order to do that, we need to be very responsible, we need to do it right, we need to act really well.

57:24Cyriac Roeding We also need to rem understand that humans are generally not very good at constraining themselves. We're not very good at that. We honestly we kind of suck at it.

57:36Ben Comer Yeah, very true.

57:37Cyriac Roeding But you know, we know what we're really good at? We're really good at out-innovating our problems. So, what I would also encourage is a debate about innovation, innovating solutions against threats of AI. In other words, the counter-AI, a defense scheme, a defense AI, because we are probably not gonna always succeed at constraining ourselves, even if we do it, other people may not.

58:14Cyriac Roeding And if the bad guys don't, then the good guys can't if they're constraining themselves. That's always the problem, right? By the way, I don't want to misuse this argument by saying, oh, let's just run and not look. I'm not using, I'm not misusing that argument. I'm just saying it is also really important, just like we do this in general defense technology, we do this in AI too.

58:39Cyriac Roeding We have to develop, out-innovate the threats of AI by innovating against them. And we're as humans, we're really, really good at that. Often much better than at constraining ourselves. And I'm not saying we shouldn't in certain cases, but I am saying let's actually ramp up on the innovation side against the threats.

59:02Cyriac Roeding So we're prepared.

59:04Ben Comer Right. Well, um, I think we're gonna have to end it there. It's been a real pleasure speaking with you, Cyriac. Um, thank you so much for coming on the show.

59:13Cyriac Roeding Thank you, Ben, for having me. It was really fun.

59:15Ben Comer We've been speaking with Cyriac Roeding, uh, co-founder and CEO at Earli. I'm Ben Comer, and you've just listened to the Business of Biotech. Find your find us and subscribe anywhere you listen to podcasts, and be sure to check out our weekly video cast of these conversations every Monday under the Business of Biotech tab at life science leader.com.

59:35Ben Comer We'll see you next week, and thanks as always for listening.