transcribe

Powering AI-Driven Drug Discovery with President of Isomorphic Labs Max Jaderberg

No Priors: AI, Machine Learning, Tech, & Startups · 39m · transcribed 10d ago
More from No Priors: AI, Machine Learning, Tech, & Startups Business
𝕏 Share ▶ YouTube 📥 PDF 🤖 .md

Section Insights

# 0:00

Challenges in Drug Development

What are the main challenges in the current drug development process?

The drug development process is heavily empirical, leading to failures in clinical trials due to unforeseen toxicities. There is a significant opportunity to enhance predictive understanding of disease biology before testing drugs in patients.

  • Current drug discovery relies on trial and error, limiting complexity and innovation.
  • A shift towards computational models could improve predictive accuracy and reduce failures in clinical trials.
  • The speaker expresses disappointment if advancements in predictive understanding do not occur in the next five years.
# 7:51

Complexity in Drug Design

Why is drug design considered a complex problem?

Drug design involves high-dimensional and combinatorial challenges, making it difficult to find effective solutions through traditional empirical methods. The industry has largely exhausted low-hanging fruit, making it harder to innovate.

  • The complexity of biological systems and drug interactions limits the effectiveness of current empirical approaches.
  • Historical context shows that the biotech industry has not seen significant new players in decades, impacting innovation.
  • Regulatory and bureaucratic factors may contribute to the slow pace of change in the industry.
# 15:42

Advancements in AI for Drug Discovery

How can AI improve drug discovery processes?

AI can create predictive models that simulate drug interactions and biological processes, significantly reducing the time and resources needed for experimental validation. This allows for more iterations and higher success rates before real-world testing.

  • AI technologies like AlphaFold can expedite protein structure prediction, saving months of traditional research.
  • Achieving experimental accuracy through AI can streamline drug design and increase the likelihood of success in clinical trials.
  • The goal is to minimize real-world testing by maximizing computational design and iteration.
# 23:33

The Role of Automation in Drug Development

How does automation impact the drug development process?

Automated labs and robotic systems can significantly reduce the time required for experiments, allowing for rapid feedback and iteration. This can enhance the efficiency of drug discovery and development.

  • Automation in laboratories can lead to faster experimental cycles, improving the overall pace of drug development.
  • Closed-loop systems that provide immediate feedback are essential for efficient iteration in biological research.
  • Collaborations with innovative companies and CROs are crucial for generating the data needed to power AI models.
# 31:24

Balancing Safety and Innovation in Drug Development

What is the relationship between safety regulations and drug development speed?

While patient safety is paramount, an overemphasis on safety can hinder innovation and progress in drug development. Historical examples show that removing regulatory obstacles can accelerate the development of critical treatments.

  • There is a need for a balanced approach to drug safety that considers the risks and benefits for patients.
  • Historical precedents, such as the rapid development of antibiotics during WWII, illustrate the potential for expedited drug development under certain conditions.
  • The COVID-19 pandemic has shifted patient risk appetites, demonstrating that context matters in regulatory considerations.

Transcript

0:00 the large part of the industry makes medicines is this very empirically driven process. Where things really fail is in the clinic because we never could see the problems until we get to the point where we put these drugs in patients and we say, "Oh no, like there's toxicity that we didn't account for." And so the really big opportunity is to have much more of that understanding before you go into patients. How can we bring all of that drug discovery work onto a computer? So how can we get out of the real world?

0:24 I'd be really disappointed if we fast forward even 5 years time like the predictive understanding of disease biology has not step change in like a massive way. Do you have a prediction on when you think for example 80% of human cancers will be treatable today on no prior we're joined by Max Yadberg. Max is the president and part of the founding team of Isomorphic Labs. Isomeorphic is backed by Alphabet and a variety of other investors and its goal is to solve all disease which is very exciting as a topic. We're going to talk a little bit about the role of AI in the future of biioarma, human health longevity and Max of course has a background in machine learning. He has a PhD in the topic working on machine vision eventually sold a company to Google deep mind rose up the ranks there and now is president of isomeorphics. So very excited to have this conversation. Max, thank you so much for joining us today. No priorities.

1:22 >> No, it's such a pleasure to be here. Thank you. >> Isomorphic, I think, is is one of the more fascinating companies right now that's happening in technology in general. Alphabet is a major shareholder and sort of sponsor. Demis, who pioneered a lot of, models at DeepMind and other places, which I believe you sold the company to before, you know, really helped with the inception and setup and now running of the company. and you folks are tackling one of the biggest problems that's left for humanity to tackle which is human health. Could you tell me a bit about how you came to work on this company and what was the motive force to starting it?

1:56 >> I was part of the you know the founding team of isomorphic labs. I was a deep mind for a long time beforehand. You know personally I was working on early days of generative models, language models, deep reinforcement learning, all of the big you know game playing challenge problems. those were my jam, like Alpha Go, Alpha Star. and it got to the point where, you know, we were we were going after these like, you know, quote unquote impossible problems at that point in time. You know, this is this is probably 2018 19.

2:24 That's >> like gameplay and can AI beat video games? >> Exactly. You know, at that point in time, you know, we were still trying to prove to the world that AI was even a thing, right? >> and so we were going after these impossible problems and I could see, you know, firsthand like how to like just systematically how to solve these. Mhm. >> And it got to the point for me personally where it's like, okay, sure, we're not at AGI yet, but we clearly have this amazing, you know, set of technologies, you know, in deep learning and reinforcement learning. And, you know, what are the problems we really need to be focusing this stuff on? And I was, you know, I was thinking all of this, of course, Deus was thinking all of this. the same time we're seeing the alpha folder results, you know, the first embers of that starting to emerge and really starting to see for us this was like the first proof of concept that you could get a deep learning model that generalizes in you know a meaningful space in biology. And so that was really the proof of concept that gave us conviction that actually you know now is the right time to go after this you know the big mission of you know changing the way that we think about human health you know going after solving all disease and that's going to take like a really substantial effort and that warrants its own company its own culture you know specific people of different disciplines coming together for this and the ability to run really fast on that. So that was the genesis of isomeorphic labs in in about 2021.

3:48 >> Oh, amazing. Your background really started off more on the deep learning side and AI modeling side and I believe you sold a company to Deep Mind and was it 2014? >> Yeah, that's right. >> So quite early. I mean you were at this stuff before people realized how important it was or many people realized how important it was. Tell could you tell me a little bit more about sort of the journey you took, how you got there?

4:06 >> Yeah. What got you interested in it to begin with? >> Yeah, I mean I I I I come from an engineering background. ended up almost by accident doing a PhD in computer vision and this was actually just at the turning point. I think I started my PhD pre- AlexNet. >> Oh because Alex Net was 2012 was it? >> Yeah, exactly. And then I saw that emerge. I was like wow this is awesome. I remember speaking to Yan Lakun about convolution neural confets and I was like we need to translate this into the core computer vision space that we we're working on in Oxford. and yes ended up developing all of these methods. We ended up having the image net winning network of of 2014 best object recognition text detection all of this sort of stuff. and I had a company like built a company in this space computer vision as a service and trying to raise money and at that point in time you know Demis and Deepmind had just sold to Google. So trying to raise money very very quickly ended up just speaking to Demis asking for angel investment and he was like well why don't you just come to Deepmind >> to solve AGI >> I was like what's AGI >> but so began the journey into Deep Mind and everything there.

5:22 >> Yeah. Amazing. And your focus now is basically a vertically integrated company that's doing everything from models to the biology. And the idea is to find drugs that can help people different aspects of human health. and so if you look at the pharma industry, one of the things that I think is really striking like I' I've graphed this out in the past where if you look at the number of new drug approvals per year, it's roughly flat.

5:44 >> Yeah. >> But that's despite an explosion in knowledge. So all these scientific papers have sort of exponented in terms of the number of papers. You know, the numbers have gone up 10 20x during that period. the number of new modalities, the type of new drugs that you can do has gone up. So there's antibbody based drugs, there's small molecules, there's proteins, there's peptides, there's all these new ways you can treat drugs. The price has gone way up, right? It went from a hundred a few hundred million dollars per drug to 1.5 billion. It went from eight or nine years to 15 years. So every metric has gotten worse while the number of things have gotten flat. The knowledge base has exploded.

6:17 The number of types of molecules that you can do has gone up. Why is that? Why has it been so static despite all that progress in other ways? >> Yeah, it's crazy to see just how it keeps getting harder and harder to make new medicines, more and more expensive, longer and longer. and you know, I I I think there's probably a number of reasons behind this. Like this the thing that I really think quite a lot about is that you know like the way that we actually make medicines in general in the industry and as a species you know forever has been a very very empirical process right at the very beginning it was you know and and and and this was you know a long time ago like a hund 100 years ago you know the medicines that we came up with they came about by chance we just like happened to stumble upon something that people realized has you know, physiological effect and people adopted it. Things got a bit more sophisticated. It's a bit more empirical trial and error.

7:14 >> Yeah. A lot of medicines we just guessed, right? That like aspirin I think for a long time we had no idea how >> aspirin we didn't know how it worked. Paracetamol we didn't know how it worked. and the way we've been making medicines and still like how the large part of the industry makes medicines is this very empirically driven process. empirical as in you need to you know you have a hypothesis on you know how a disease might work. You have a hypothesis on what a good molecule design could be but you have to make it and test it in the real world and do this very iteratively.

7:44 >> That's slow. It's really really difficult to get signal out of that and you know the space of biology is really really big and complex. These are like big comp you know the cell is a big complex chaotic system >> super highdimensional >> and then molecule design whether it's small molecules antibodies peptides you know they're really again really highdimensional objects combinatorily large design problems and the only tools we've got available you know to us you know is is basically trial and error screening processes >> and so that limits to me the complexity of solution that you can find in in a finite period of time or finite amount of money that you spend on that empirical work.

8:27 >> and so I think you know what you've seen over time is that the lowhanging fruit of medicines that can be found >> with our empirical process of designing medicines are being found. We're finding the low hanging fruit, but then it's getting harder and harder to sort of ratchet up the complexity and and find things that are out of reach from this empirical process. >> That's interesting. But don't you think so? If I look at any other industry, often what happens is the age of the industry impacts the rate of change in it.

8:56 >> In other words, the older the industry, the slower it moves. >> Things become more bureaucratic. There's more regulatory capture etc. I think the last time there was a biioarma company that was worth over $50 billion besides Madna and I think Madna was almost an accident of COVID was the 1980s I think with Regeneron. >> Yeah. And so it's been almost 40 years since we've had a new tens of billions of dollar biotech built. >> Yeah. Regeneron, Vert.Ex, these you know >> Amjen was that era too early ' 80s.

9:27 >> and >> and so how much do you think is just like regulatory issues or you know bure bureaucratic organizations or just like that natural aging I think that happens everywhere you know that was automotive until Tesla that was you know >> yeah we you know we talked about maybe the the science side of it. I think there's a big cultural aspect of you know just how how can you like keep innovating as an industry how how can you keep moving really really fast and I think as well and this is related to the sort of like science and technical part is what like what are the mechanisms that actually give you economies of scale on the innovation side on the research and development side so if you go after you know a particular drug target and you're trying to make a particular medicine and you're a pharma company or biotech you're going all in on that. What do you learn from that that helps you work on the next medicine much quicker and then the next medicine even quicker after that?

10:21 because if you could get that compounding of learnings over time, you could maybe get the benefits of scale and build something you know really amazing and big as a company. but I think a lot of the techniques and the way we go after medicines or the industry goes after medicines is really treating each one as like a segmented you know it is like a bespoke frontier science project >> and you learn so much about that project and there's maybe not so much that's transferable to the next like there's of course many many different learnings of disease areas >> and this is I think where things can shift quite a lot in the world of AI and machine learning where if you underneath You know, if you have this platform underneath going after these medicines that allow you to like, you know, really suck up and understand the learnings and help impact your core foundational understanding of disease biology and how to design good molecules and exquisitly engineer these medicines, then you can hope to actually get compounding improvements the more medicines you go after.

11:19 >> Because I guess like one of the striking things that has emerged from deep learning and the sort of foundation model era that we're in right now is the generalizability of some of these things with scale. So it sounds like perhaps in biology as well if you have sufficient scale and a sufficiently thoughtful approach you can generalize some of these problems that were bespoke before. what approach are you all taking? >> Well I I think that's the big shift that we're taking with isomeorphic compared to you know we're not the first people to try and change the way we do drug design and drug discovery with AI. One of the things that's really fascinating about isomorphic is a degree to which you have a very diverse space of people all working together towards a common cause often who come from backgrounds that are very disperate. Can you tell me more about what sort of people you hire?

12:01 What are their profiles? Who you're looking for currently? >> Yeah, I mean it's an interesting space because the the the sort of mashup of you know frontier AI and drug discovery, chemistry, biology hasn't really existed for very long. So you can't kind of go out into the world and say, "Hey, I want to hire the world experts at AI for drug discovery." Because they basically don't exist. And if they do exist, they're not using our models using different techniques. So we think about going after literally just the best people in core AI and deep learning who maybe don't even have a background in life sciences or biology or chemistry, but really want to see their work on deep learning translate into amazing impact for for humanity. So we hire just core people in AI. We have core people in biology and chemistry maybe who haven't had much exposure to AI and then you know we work hard internally to build that culture that everyone talks and learns from each other and then internally we get these sort of cross-disciplinary experts expert scientists emerging which has been really exciting to see.

12:59 >> How do you think about the ratio of those things? So I started a company a long time ago called color and I was co for the first couple years. My co-founder took over a decade ago now. So this was a long you know ways back and we were basically using ML and data science as applied to genomics and we were building these genomic pipelines and what we found is you could have like five ML people and then one research scientist from the traditional field and that was kind of the ratioish >> that we adopted. What is it like now or how do you think about how many people of each discipline do you optimally have in a pod working on a problem? Yeah, for some of these core AI, you know, superpower projects, the the AI modeling projects, I think that's actually not too dissimilar. You know, you have your core modelers like engineers, AI scientists, and then you just need some key domain experts, people who really understand the application, the nuances of the data, like what do the metrics mean? make sure that we're not, you know, chasing smoke metrics and these are real things that are going to translate into actually making medicines. so yeah, those sort of ratios, one in five, like that I think those resonate quite a lot for what our teams look like.

14:06 >> And are there specific roles that you're actively hiring for currently or just across the board, you know? >> I mean, we want the best AI researchers coming into ISO and helping us push on this mission. And I think there's there's literally so much work to do on the >> on the like core simulation models like Alphold++, but also on the core reasoning models like taking all of the frontier models and and open source models and and adapting them to our space. so that's I think a really important area for us to be hiring on.

14:35 >> What are the components that you tackle or what are the main areas that you all are focused on in terms of how you approach new drugs or you know you do you have your own labs? Do you run robots that do pipetting? Like what what is the approach we're taking? >> Yeah. So you know if you if you go into our like sort of our core AI team and like our north star and the thing that we're really trying to achieve is how can we bring all of that drug discovery work, all of that drug design work, understanding of you know disease biology onto a computer.

15:07 >> So how can we get out of the real world? you know I know a lot of people are building ways to be like interfacing with the real world really really efficiently so you can do lots and lots of real world it iteration >> that's great and it's complimentary but fundamentally the the big time bottleneck the big cost bottleneck is synthesizing real atoms >> and then testing them out in real assays and you know experiments and you know just like we've seen I think pretty much every other area of engineering move into a computer computer AED design scenario and that's just changed the way that you know change what we can design change what we can engineer >> >> we want to do the same thing in this space >> and so really we like when we think about the core AI drug design engine development side this is really focused on well how can we achieve experimental accuracy >> for our simulation models our world models if you can get a predictive model you know let's use Alpha fold as an example. Alphafold looks at structure prediction. The the the structure of proteins and how they interact with each other.

16:17 Normally that takes, you know, months or years to do X-ray crystalallography and and and get a structure out. If you can do that with a simulation model like AlphaFold, then you don't have to go into the real world and do that. You save months and months of time. And there's many many you know drug discovery is a is a combination of many many different really hard scientific problems and engineering problems and if you can bring that you know create predictive models instead of the real worlds then you know you're saving tons and tons of time. So internally we have you know a dotted line on all of our leaderboards that is experimental accuracy >> and we you know the goal of all of these like superpowers that we're trying to create is to get to a point where it's experimental accuracy. At that point, you don't need to go into the real world. You can do so much more design and iteration on a computer. So that when you do go into the real world, when you make your molecules, when you test out your scientific hypothesis, the probability of success is much much higher than it would have been traditionally. And so you don't have to do as many turns as the real world and things actually become quicker and problems fall easier.

17:20 >> Well, so when I look at drug development, I think it's about a 15-year process now. >> It cost $ 1.5 billion. My sense is the small molecule and the in vitro work, the tissue culture and m mouse work is basically say two to three years and then it's say $50 million of that order of magnitude and then the other 1.45 billion and 12 yearsish is the clinical trial and you know the different phases and safety and how much of that do you think you can actually encapsulate in a computer? You know, you have the discovery phase. So where you go from idea to you think you've got a good like a good molecule that you could put in patients or before that point you have to do pre-clinical, you know, animal studies to make sure that this is safe and working as intended. That takes another year. Then you go into clinical studies. You have phase one, phase two, phase three ratchets up. And so, you know, when you look at the timeline and when you look at the failure rates, and I think these are the things that really exacerbate the costs on average, >> right? Why do things cost so much? It's because so many things fail.

18:24 >> Then you have and and where things really fail is in the clinic. >> Mhm. >> It's not because these are problems. it it they fail in the clinic because we never could see the problems until we get to the point where, you know, the rubber meets the road and we put these drugs in patients and we say, "Oh, oh no, like there's toxicity that we didn't account for." Or >> we have a molecule and it hits the thing that we've been designing it to hit, >> but it hits something else, which is a really negative effect.

18:54 >> Or the thing that you do hit, we thought it's going to solve the disease. Maybe it does even, but it actually has a really negative effect in and of itself. >> And so you like people only learn about those problems when they start introducing it to patients, to real patients. >> And so the really big opportunity is to have much more of that understanding before you go into patients. >> So if you have a better understanding of disease biology, how cells work, you can maybe understand how the thing you're trying to hit and modulate is really implicated. So you're actually modeling all these other components and not just like what molecule will bind to what >> exactly we're modeling all of these components.

19:31 >> How do you get the data to model those things? Like where does that data set come from? >> You know, we have this like ridiculously comprehensive, you know, data strategy. There's data all over the place, but I think fundamentally what we've learned over the last, you know, 5 years is there's a lot of opportunity to generate this data yourself. >> there's of course like big repositories of data. Sometimes that's really good for deep learning. oftentimes these repositories of data you know whether they're public or in pharma or you know in companies they've been the byproduct of many years of passive you know data collection from doing a you know the activity of of making medicines. So they can be very very biased >> which often doesn't really help you if you're trying to make a general solution to drug discovery. If you really care about generalization out of these models there's maybe a different distribution of data you need. So we go out and generate this ourselves you know and we've seen some really really amazing data scaling laws as a result.

20:30 >> How do you measure whether some of these things will have these side effects because ultimately there's the old saying I used to be a I have a PhD in biology and I used to work in a variety of fields cancer no science etc. And on the cancer side we cured cancer in mice hundreds of times and those molecules wouldn't translate into people >> and so a lot of the data you actually needed you had to generate in human clinical trials. How do you go about generating that data that actually informs it? Because if you're using animal models, often the data that comes out of that is incorrect relative to human biology.

20:58 >> Well, exactly. It sounds like you experienced a lot of this firsthand. It's very painful. >> It's very painful. And you know, if you think about >> you know, toxicity like sources of tox like the gold standard that you know you need to get through for regulators is you need to prove that your molecules aren't toxic on rats and other animals. but still even when things are approved and they've been proved on rats, they end up being, you know, often toxic in humans.

21:25 >> And well, why is that? Well, okay, a rat isn't actually a human. It's not actually very predictive. I think, you know, choose your study, but it's about 30% predictive of of of human clinical tox like in a rat. and you know, if if one of my like AI scientists or engineers came up to me and said, "Hey, we've got this amazing model. It's it's it's 30% predictive." >> You're like, "Okay, maybe do a bit better than that."

21:45 >> Yeah. and so I think there's a lot of opportunity to actually rethink how we you know like what are those predictive endpoints for human clinical talks like forget animals you know they've been a proxy for the last you know 50 70 years >> >> but we can maybe do better we can do better by >> you know bringing together >> in vitro endpoints so things you could measure quite easily in the lab together with you know AI predictive models also together with there's some amazing new experimental technologies coming out in the space of organoids, organs on a chip, like synthetic human tissues, >> which sort of approximate and create these like little microcosm environments of human organs.

22:32 >> And so you can imagine there's probably an opportunity to bring all of this stuff together >> and then directly try and model human clinical talks. Now there's not much data on that side because >> you know there's there's only a finite number of clinical trials that have been run. >> but I think we have some really good methods now. >> And for these sorts of approaches or techniques, do you tend to partner for them? Do you bring them in house? Like I'm a little bit curious as a company how you operate relative to because there's so much that could be done.

22:57 >> Yeah. >> And a little bit curious how much you want to just emphasize the computer science side and then you know biology often there's these things called CRO or these contract research organizations that people use. Unfortunately, the quality tends to be pretty mixed in terms of the output of these types of organ. I'm just a little curious like how much you depend on external parties versus internal. >> Yeah. So like making our drug design engine all of the AI models, we do that all in house. and you know we see a big opportunity of generating proprietary data and owning that ourselves but you know biology and you know experimental platforms like this is really complex and longtailed and there's some amazing people working on some fantastic platforms here. So you know we work with the whole world whether that's you know stuff we do ourselves contract research organizations CRO's other innovative companies to actually help us create the data that is you know powering our models.

23:55 >> so yeah >> that's cool. How much so Enthropic recently announced that they're doing a wet lab in San Francisco and you know reading between the lines without any special knowledge just reading external papers it kind of looks like there's some potentially robotic system or closed loop system or something else they're trying to build and in general if you look at AI models as you know better than me you want closed loop systems to be able to iterate quickly in other words if you can get rapid feedback like if you're writing code and it can instantly get tested then you have a fast way to test if the model's working well or not right and So in biology, one of the issues is you set up an experiment, you wait 3 weeks, then you get the results, you have to interpret the there's a big loop and so that slows down progress in the ability to iterate on data and building these these robotic platforms or other things may expedite that. how do you approach that problem or do you think that's part of the solution that you all will be pursuing in terms of >> Yeah, I mean it's been really exciting to see like how much better these you know automated labs how how much better these robotic solutions for experimental labs have been getting and you know we we we've been working with these sort of automated labs for a long time now everyone should and we should keep driving down that like iterate iteration time with the real world and fundamentally I want to get out of the real world right And and you know, can we actually because there's going to be just like you're going to hit the finite laws of physics at some point like how long does it take to synthesize a molecule?

25:20 >> At some point you have to just wait for the reaction to happen and that's how long the reaction takes. How long does it take to culture a cell? >> At some point in time you just have to wait those days for the cell to get to the right state before you can test things on that. And so yes, we can, you know, I think do a lot of work and we should and we are as a as a species as well is on on on you know improving automation and improving the way >> you want to bring it in silo. It sounds >> but I think we should also be driving really hard at bringing that in silicon on a computer >> so that we can reduce the amount of real world time so that we can do much more predictive work on a computer so that when we do go into the real world >> yeah it looks like we go into real world to validate >> not necessarily to learn and iterate. We still need to go into the real world to validate because we're coming up with novel, you know, scientific hypotheses and we want to test them out in people at the end of the day. So, you have to make sure that they work and they're safe. So, you have to go into the real world to validate. But the question is, how much can you reduce the amount of time you go into the real world to experiment and to iterate?

26:21 >> How much of your approach is focused on partnering with existing pharma companies versus developing your own drugs? >> We make the AI drug design engine in house and then we we actually, you know, don't sell the model. We we don't license out the model. We don't sell tokens in that way. We use it internally to to make medicines. >> And then we do that in two ways. We have programs, drug discovery programs that we fully own and fund ourselves and work on ourselves, our internal pipeline. And then we have programs that we work on with, you know, amazing farmer companies, biotech, where we partner on these on these projects. And those are really exciting as well because often those companies and those teams they've got you know amazing ideas and amazing expertise in specific disease areas and so they come to us with like these are the hardest problems maybe problems that they've been working on for sometimes 10 20 years not making much progress and then together using our drug design engine we can make a huge amount of progress. What are the indication areas or areas of human health that you all are focused on?

27:23 >> Internally, we're focused on, you know, much more on oncology, immunology, and inflammation. >> but the cool thing is, you know, the drug design engine, all the models, they're general. They're disease agnostic. They're target class agnostic. They're modality agnostic. So, they work across multiple >> modities. You're starting with like cancer and is it autoimmune disease or something else or >> Yeah. So you know immunology inflammation they're often quite related so it's that sort of space and when we go after these things internally whether it's oncology you know cancer inflammation you know I think there's a big opportunity to go after some of these root node problems the the sort of things that could affect you know positively if you have a solution could affect you know lots and lots of different patients for different downstream indications >> an example of that >> I If you if you think about cancer, >> like this is a really really complex and longtail disease and it shows it itself up in in patients in many many different ways.

28:24 >> But we also know there are some like maybe there are some core driving mechanisms that are common to many different things. you know, we saw this amazing example earlier this year. you know, Durax and RAS, like Kraz inhibitor, standing ovation at a big cancer conference like drastically improving >> you know, the the the the lives of of of sort of like latestage pancreatic cancer patients from Revolution Medicines. And, you know, again, great example of like going after some sort of root node.

28:55 >> I see. Yeah. RAS has been popping up in the cancer literature for like 30 years, 20 years. Yeah. But these these problems are devilishly hard >> from a like drug discovery and design point of view like just >> and also has a natural function. So being able to >> modulate. Exactly. So you really and and of course then you know in cancer things mutate very quickly. So you think you've got a solution but actually clinically it doesn't work out because it just mutates around the solution.

29:20 It's actually an interesting point because for a lot of different drugs, for a lot of different areas of health, this also was HIV, right? You end up with a cocktail of drugs that you give eventually and that's actually what allows you to suppress AIDS or HIV. In cancer, similarly, a lot of drugs are sort of developed as as individual drugs versus cocktails. And to your point, often within a tumor, you have a spectrum of mutations that you need to suppress simultaneously if you really want to cure the thing because otherwise it'll sort of >> evolve out under given the sort of substrata of mutations that exist in a distribution across the cells. Are you thinking about those more complex approaches of cocktails or multi-drug approaches or >> so? So why aren't we giving more cocktails drugs? Well, the problem is like often these molecules they are they work but they have some toxicity >> or they exhibit some effect on the liver such that if you're taking multiple of these molecules these medicines at the same time they have these drug drug interactions that can be really harmful for patients.

30:20 >> And so it's actually because we're not great at engineering exquisite molecules and medicines that means that we can't combine them together in a safe way for patients. And so we're maybe limiting like the the surface of applicability. >> Yeah, I guess you only have so many patients to and so many people you can enroll a clinical trial and so the combinatorics of it my sense is also kind of adds up in terms of the clinical trial side >> especially when there's a lot of like risk and uncertainty about how these things will combine and so that's where we see an opportunity. Yeah. If you can actually much more accurately engineer these molecules to be safe, to not have these effects on the liver, to not interact with each other in bad ways, then you can feasibly come up with actually, I would really hope some quite complex, you know, combinations >> that can modulate not just a single target at a time, but a whole vector of targets that span these like very complex biological pathways at the same time and so have a better effect on disease. There's a very good YouTube video of this guy Jansen who started Jansen Pharmaceuticals. It was acquired eventually by J&J and he was one of the big pioneers of drug development and pharma and I think he was one of the they had one of the most productive pharma companies in terms of actual drugs launched and in it he his view that he expresses pretty strongly is that the main constraint on drug development is an overemphasis on safety >> and that regulators kind of curtail progress. And he pointed out how during World War II when St. Churchill tried to find a cure for soldiers of some venerial disease in the field. And so they they rediscovered antibiotics because penicellin was discovered in the 20s by Fleming and then kind of put on a shelf and forgotten. And then in a 9-month period they rediscovered and commercialized penicellin from scratch specifically to treat this and it was because he removed all regulatory obstacles. So that was an example that Jansen pointed to one could argue the COVID vaccine was similar in terms of being able to rush it. how much do you think there's too much of an emphasis on safety? Like Jensen's point was it's about riskreward.

32:22 >> You know, you need to make trade-offs and you can't just monolithically say it's only safety. You should also say, well, it's about benefit and in some cases risk should be taken as well. Like how do you think about that? >> I think it's like centrally important to just always have patient safety in mind to be honest. and but then you need to understand like what are the risks and the appetites of patients and that can be different for different indications.

32:45 you know COVID's a good example where the risk appetite for patients because it was so pervasive and so dangerous just changed dramatically and that changed I think the demand and the force for doing things slightly differently. You even see this across different disease areas. >> If patients have you know latestage cancer >> and not many options in front of them there's a very different risk appetite. >> Yeah. compared to if you're trying to treat a chronic low-level, you know, inflammation indication. And so I think it's actually really really important that like we keep focusing on patient safety, >> but that we can modulate that, you know, depending on what the patient need is.

33:28 So ultimately, you know, ultimately we want patients who are living healthier lives >> and they're receiving that treatment as quickly as possible. Mhm. >> And you know we we we speak a lot to regulators here you know FDA here but also all over the world and I think the one thing in common is like you people really want to get patients the best drugs and the best solutions possible. and I do think that over time we will see a difference in our you know pro you know the success rates and the safety profiles of molecules coming out when they're designed with much more understanding about how they work >> with you know we're be able to search the design space to engineer something you know much better than you could do from an empirical empirically driven process and I hope to prove out over time that this is very stat statistical different >> and that will change the way that we think I think about about regulation and and getting drugs to patients.

34:26 >> When do you think you'll have your first clinical trial? >> I would love to give you like a a hard day. This is like really hard frontier science. >> we've got some really exciting medicines in the works. and these are going really well. >> So we're pushing really hard to get these to patients as quickly as possible. We've been building up our clinical development team. So we're getting ready for this. in there. >> Do you have a I know these are always impossible questions. Do you have a prediction on when you think for example 80% of human cancers will be treatable >> or you know some other metric of why this is so important for the world? I'd be really disappointed if we fast forward, you know, even five years time, if we don't have the base tools like the like the predictive understanding of disease biology has not step change in like a massive way. Mhm.

35:15 >> our ability to design molecules that are much more precise, much less toxic, has not step changed in such a massive way that you know still drug development, discovery and development takes years >> and that'll reduce over time but it takes years now. but we should have the tools in place to really start be like start cracking these large large portions of disease in a different way. if you remove if if you removed cancer from society and heart disease, I think the stats suggest that the average person would live I can't if it's five or seven years longer. And so the actual impact to longevity for two of the biggest killers is on some in some sense quite large and another point of view is kind of small. there's a number of companies that have tried to focus on aging or longevity as a more general process which seems to clearly be genetically programmed. Right? If you knock out certain genes in seaggans or drosophila or in mice, they'll live 20 to 30% longer. Or in the case of seelagans, the the little worms, they'll live three times longer, right? If you knock out certain genes in certain pathways, I actually worked on those pathways as part of my PhD.

36:22 >> Amazing. >> It was really fun. how do you think about that in the context of human lifespan? Like should we just hit a point eventually where people live three times longer, five times? like what how do you think about longevity in the context of drug development and you know all these tools that will now exist? >> Yeah. No, it's interesting like there there are some of these major diseases that afflict so many people >> that if we could start to make impact on those that that would have a concrete impact on on how long people are living.

36:52 I think even outside of things like cancer and heart disease, you just have the sort of chronic afflictions >> of aging that start to add up >> and the root cause is literally aging. >> In other words, like the root cause is the fact that you're allowing these systems to degrade over time. >> Great. And stem cell compartments. >> Yeah. And they and they combine in a way that means that people's you know even if people live to a lovely long age of 80 or 90 like they >> maybe the last like 10 years of their life >> or even 20 years of their life >> isn't not isn't necessarily as great as it could be because they've you know because of you know your bones start to degrade your muscles start to degrade like you have a fall and suddenly you can't you're not as mobile as you want to be that has impacts on your mental health and suddenly your quality of life has dropped dramatically. You still might live another 10 years.

37:50 >> In the aging community, they talk about that about as lifespan versus health span. >> Exactly. So, we think a lot about that. and >> but all the all the aging models, if you impinge on aging and you delay it, the health span goes with it. In other words, if all lives three times longer, it's actually a healthy adult for three times longer. It's and then it crashes out at the end. But you actually have extended both simultaneously. So, I'm a little bit curious how you think about like life extension and you know that kind of stuff.

38:15 >> I I think it's going to be like it's going to be really fun to see everyone working on this and and yeah, as you said, I think we can attack this from every different angle, right? we can think about improving you know the things that afflict us and and and help maximize our health span >> and also be trying to hit some of the root root causes of aging and degradation of our bodies and cells and organs so that we actually you know can actually extend that over time as well and and I there's yeah I mean there's such exciting work happening in this space and yeah it'll be interesting to see how this pans out.

38:51 >> Amazing. Thank you so much for joining us today. No, it's been a fascinating chat. >> Find us on Twitter at no prior pod. Subscribe to our YouTube channel if you want to see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way, you get a new episode every week. And sign up for emails or find transcripts for every episode at no-bers.com.

Summary

Max Yadberg, president of Isomorphic Labs, discusses the potential of AI in transforming drug discovery and development, emphasizing the need for a more predictive understanding of disease biology to reduce the high costs and lengthy timelines associated with traditional pharmaceutical processes. He highlights the importance of integrating AI with biological insights to enhance drug design and improve patient outcomes.

- The pharmaceutical industry relies heavily on empirical methods, leading to high failure rates and costs in drug development.
- Isomorphic Labs aims to leverage AI to simulate drug discovery processes, reducing reliance on real-world testing and improving predictive accuracy.
- The company focuses on oncology, immunology, and inflammation, seeking to address complex diseases through innovative drug design.
- A diverse team of AI experts and biologists collaborates to create a culture of cross-disciplinary learning and innovation.
- The integration of advanced technologies, such as organoids and AI predictive models, is essential for better understanding human toxicity and improving drug safety.
- Yadberg emphasizes the need for a shift in regulatory approaches to balance safety with the urgency of delivering effective treatments.
- The goal is to develop drugs that can be tested in clinical trials more efficiently, with a vision of significantly increasing the treatability of diseases like cancer in the near future.
- The company is committed to generating proprietary data and collaborating with existing pharmaceutical firms to enhance drug discovery efforts.

Questions Answered

What are the main challenges in the current drug development process?

The drug development process is heavily empirical, leading to failures in clinical trials due to unforeseen toxicities. There is a significant opportunity to enhance predictive understanding of disease biology before testing drugs in patients.

Why is drug design considered a complex problem?

Drug design involves high-dimensional and combinatorial challenges, making it difficult to find effective solutions through traditional empirical methods. The industry has largely exhausted low-hanging fruit, making it harder to innovate.

How can AI improve drug discovery processes?

AI can create predictive models that simulate drug interactions and biological processes, significantly reducing the time and resources needed for experimental validation. This allows for more iterations and higher success rates before real-world testing.

How does automation impact the drug development process?

Automated labs and robotic systems can significantly reduce the time required for experiments, allowing for rapid feedback and iteration. This can enhance the efficiency of drug discovery and development.

What is the relationship between safety regulations and drug development speed?

While patient safety is paramount, an overemphasis on safety can hinder innovation and progress in drug development. Historical examples show that removing regulatory obstacles can accelerate the development of critical treatments.

© transcribe · For agents Built with care and craft by Gokul Rajaram