AI is Transforming Drug Discovery. Just Not The Way You Think

Monisha Saldanha: Welcome to Colorado Tech People, Today I'm joined by James Brewster, CEO and founder of Cime Therapeutics, a biotech company that is redefining drug discovery. James, thank you so much for joining us James Brewster: Monisha thanks for having me. It's an absolute pleasure to be here and get to chat with you. Monisha Saldanha: Yeah, I'm looking forward to our conversation.

is the problem you created Cime Therapeutics to solve and what is the solution? James Brewster: Absolutely. So it's a really exciting problem. to ⁓ face and a daunting one as well.

So we're focused on human health and more specifically trying to explore the unknown chemical universe. So the number of drugs that we have today, while amazing and treatments are available, there's that many. So we've only drugged less than 10 % of the proteome. ⁓ so is built to really expand the medicines that we have.

And we're really focused on oncology. ⁓ So new treatments and new cures in cancer to give people a better life and a longer life. And we've already actually initiated some programs, in triple negative breast cancer and pancreatic cancer. So we're really trying to go after these kind of devastating diseases and hopefully make an impact there.

Monisha Saldanha: What is the role that AI is playing in your drug discovery? James Brewster: So quite a bit, and it's a really exciting time to be in new medicine development and drug discovery and chemistry in general with AI. So we have what we're calling an AI controller for some of the automation. Claude and these other tools are excellent at generating Python scripts.

And so you can actually use that to then essentially control your instrumentation and then have it run chemistry for you. And then it can iterate. So based on the that comes out of that experiment, you can plan the next steps of the reasoning side and then continue to build on that. So what chemistry do we need to do?

Reagents do we need to where? And things like that. And then going into the reasoning side a more well as well too. So it does a good job of synthesizing from a lot of data, kind of pulling out these key pieces to then guide next steps, hypotheses as well, or help scientists make hypotheses.

⁓ really leaning into this human in the loop aspect as well. So these skilled scientists that have dedicated their lives to ⁓ helping people. And ⁓ we found is AI is excellent for that. Bouncing ideas, pulling papers, and like I said, guiding next steps.

Monisha Saldanha: Are you using Claude what AI are you using? James Brewster: All kinds of tools. So everywhere from kind of generative aspects associated with small molecules. So think of just building off into chemical space.

You can imagine this like filling in various pockets or vectors, you will. So we use it for that. So enumeration, trying to pick up key interactions to drive potency. So we go back to the early days of high school or college chemistry, where we talk about bonding and things like that.

So we try to really bring in specificity, what we call it, into the pocket all the way to ChatGPT and Gemini and these other amazing tools. ⁓ a lot born out of really the boom of Nvidia. So they have an amazing amount of tools as well that we've been ⁓ to help chemists and scientists do more. Monisha Saldanha: When talk about the hype, what do see as being the hype around using AI for science?

James Brewster: So there's a lot of hype around AI for science, said. And some of it is amazing. It enables us to move so much faster and synthesize some data, as I mentioned, a lot faster than previously where we're using Spotfire, JMP, more common tools to glean information. On the flip side, it still hallucinates a lot.

⁓ And so we still have these problems where we don't have the ability to make mistakes it could lead you down a rabbit hole. It could cause problems and waste time and waste money. And that's difficult for a startup. So going back to, we really have this human in the loop aspect that guides the AI tools to what we should be making, what's even possible to make in the chemistry realm.

You find that these tools often hallucinate molecules that will never exist in ⁓ ⁓ ⁓ on the laws of physics. Monisha Saldanha: Shifting gears a bit, can you tell us about a hard decision that you had to make? James Brewster: So drug discovery and science in general, separate from business and even employing these kind of AI tools is constantly filled with difficult decisions. where do you...

staff resources and put your money to focus is it on the AI automation side? When do you have to hold projects and cut projects? And then on the med chem or the medicine side, had some really interesting programs that start and we have these these interesting hits, but you quickly find that it's going to become too expensive or too much of a kind of a difficult And so we put those ⁓ on back burner. And it's unfortunate, right?

The ultimate goal is help people. But you're faced with and I think everyone can appreciate that these these decisions that you have to make you hope they're right from business sense to advance you know one way or the other. Monisha Saldanha: What lessons have you learned? James Brewster: That chemistry, and this is a lesson time and time again, that chemistry is really difficult.

Drug discovery is really difficult. being a startup company is really difficult. Things break that you don't have the funding ⁓ or the desire and time to fix, but you still have to fix it. Again, going back to as well, you want to explore everything as a scientist and you have to be really lean and smart about what you focus on.

There's also difficulties with the ever-changing environment of AI, to keep up with everything that's going on and making sure that you're on top of it and utilizing the best tools ⁓ what you need to do. Monisha Saldanha: What do you see as ⁓ difficult? Could you go into the details? Like, what's difficult about being a scientist and about chemistry and about working in the startup?

James Brewster: So being a scientist, you're, and chemists especially, every day is constantly filled with failure, and a lot of it. So you'll set up reactions almost with the expectation that nothing works, ever. you become accustomed to that kind of lifestyle. But not easy, right?

So you come into work knowing that likely whatever you did yesterday didn't work, and you're on to the next one. And I think you've, you over the past decade plus, I've been able to kind of get over that, but it's still tough sometimes, right? You want to see success, especially when you're developing a medicine that has the potential to impact people. see what we call like idiosyncratic talks or just something weird happening and you can advance that compound and, you know, the next steps.

And we that lot from the startup perspective. think a lot of it's just. the limitations associated with the budget, right? In Big Pharma and in any big company, you have a pretty large budget to kind of do what you want and explore things and make mistakes.

⁓ of mistakes you're allowed to make or that room for error is almost gone when you're in a startup. ⁓ so again, you have to really think about what you're doing and we pretend it's being smart, but a lot of it's luck with a little bit of intelligence too. Monisha Saldanha: So not much room for error when you're working in a startup. you're based in Louisville.

What is it about Colorado that attracts you? How is it working as a startup within the Colorado ecosystem? James Brewster: I absolutely love it. Colorado is home.

It's been home for the last five or six years now. And a lot of it comes down to the people. I think what attracted me here initially and what's continued to keep me here and want to build a company and continue building a company here is the quality of person that you get. They're genuinely good human beings.

They're kind. They also work really hard. And not only do they work really hard, but they deliver. It's not fluff or this promise of something amazing.

It's the promise of something amazing and then they go. in our case, into the lab and they go and deliver on the chemistry and the biology and the experiments that really start to advance medicines. And it's not just us, you see it. You the entire biotech community here, we have some amazing companies, Cogent, Edgewise, etc.

, that are delivering medicines to patients. They're already in phase three. Cogent just submitted their NDA for the FDA for one of their medicines for their kit inhibitor. And so I don't think you find that anywhere else.

It's really, I think, unique to Colorado, at least my experience being on the coast and then coming here. Monisha Saldanha: you tell us about a pivot that you've made? James Brewster: So quite a few. And I think that goes with every startup and every everyone that working at a startup can appreciate.

got to find that product market fit. It's a little different from us versus like SaaS or B2B ⁓ things like that. But, you know, initially we'd started out with we had this unique platform and we were thinking about potentially running an outsource team ⁓ really early We had some from some of the venture groups here, ⁓ specifically, Range and Service Provider and DYDX, folks that funded us, ⁓ to build a team here and establish a lab and start to build out the technology. And so the pivot from things ⁓ if you will, to an on-person was a big pivot and ⁓ building out the for that.

And then we're even seeing now this push towards AI, ML, and especially automation and where you can enable humans to do more. ⁓ And so we've really leaned that aspect and going into the platform and our ability to do chemistry really well, but start moving that and enabling, I said, humans to do more. So making more compounds, more chemical space, expanding our target list. many medicines can we develop?

Can we develop a nice pipeline to help not just in triple negative breast cancer and pancreatic, but in liver cancer and colorectal cancer and maybe even brain cancer and beyond. ⁓ Monisha Saldanha: Wow, amazing. are the hardest scientific challenges you've had to overcome so far? James Brewster: So chemistry, as I mentioned, is a constant fight.

works ever until it does. And so you'll set up dozens of reactions to screen conditions and sometimes even more than that. Unfortunately though, it's something we've all been trained to do really well. the chemistry side is not too bad.

But we had to build out this new platform technology. So all the way from the material science side, ⁓ we wanted to create essentially what we're calling our chem printer or chem print, ⁓ you can think of it like a piece of paper going through a printer and we're printing molecules on top of that to then screen ⁓ a biological target. So we had to build that entire technology and it took quite a bit of effort with lot of failure. And then, you know, the medicine side as well.

We had some compounds that we thought were really good and when you get them into kind of more advanced models, they just weren't meeting the product profile that we wanted to see. And so then you have to go back to the drawing board again of, how do we fix that? Or do we need to completely pivot into a different, what we call scaffold? Monisha Saldanha: What is the biggest breakthrough your platform enables that wasn't possible before?

James Brewster: This is where I get excited. It's the ability to, if you think of like a room that's dark and you're in there with a flashlight and you're shining your flashlight all over the room to kind of map things out. We're able to do that very fast. And so traditional chemists, so traditional like you and me in a lab would make, you know, maybe 50, 100 if we're really good and we're having a good year, 150 compounds.

Our platform allows us to make thousands per day, you know. Monisha Saldanha: Hmm. James Brewster: all areas of chemical space that we couldn't do before at this speed. ⁓ so you can then imagine going back to we've only drugged less than 10 % of the proteome in the case of cancer, you can then start pushing that number to 20 % and beyond.

And so it's for me to think about how many people we can actually help when we something like that. Monisha Saldanha: What is it about your platform that enables such scale? James Brewster: And so this a lot of time and a lot of effort. And I think a lot of thinking and then a little bit of luck or maybe a lot of luck.

And leaning into advancements of ⁓ science and really been done over the last three, four decades that has enabled us to do chemistry that's more traditional to what you'd see again, going back in the lab. Chemists and scientists in general developed all these methods to make compounds really fast, they kind of required odd conditions or things that didn't work well, it was like a band-aid, essentially. And so we really wanted to go back to how do we make things like you normally would at machine scale and speed.

So it goes in automation and then thinking, how do you print molecules? Because a printer is really fast, right? And so that was really the inspiration for us, was to pull from all of these fields that nobody else had really thought of combining before. Monisha Saldanha: Was that a collaborative, creative process between the three of you, or how did you come up with this new platform approach?

James Brewster: So this was actually going back to my time. I was in Pfizer's oncology group thinking that, you know, we apply automated synthesis. Can we do it faster? And then how do we do that?

And so I spent, you know, quite a bit of time thinking about how do we answer that one key question and then had the bootstrapped some experiments and then raised some money, built a lab. And I've been lucky that Pat and Greg joined and ⁓ hit the at a full sprint. And so, you know, the three of us together have really worked out how this works at scale to make thousands of compounds a day and beyond, you know, as we really want to start mapping out these really difficult targets and jugging them.

Monisha Saldanha: And why start with oncology? James Brewster: From a perspective, I think it's really important to me to make an there. It's something that I've spent a bulk of my career in. So my PhD was in a bio-organic group working someone had really helped develop for cancer.

⁓ went and did a post-doc, same thing, was working on potential cancer medicines, and then my time at Pfizer was in the oncology group, ⁓ working in oncology. I think there's a lot left to be done where we can improve patient of life, not just how long patients live, but ⁓ making that the time have is not ⁓ too Monisha Saldanha: Yeah, yeah, I guess as good as it can be, right? It's important. Quality of life.

What's something about building a biotech company that surprised you? James Brewster: Everything! It's very easy to be, I think, very easy, loosely defined, a chemist, right? Your job is to make molecules and try to advance those into animals and beyond into humans.

When you're building a biotech, there's so much more that you don't appreciate. And I think you see this at any time you're at a big company versus a startup. There's a lot that goes on behind the scenes to keep the day-to-day moving smoothly. And so those, I think are the biggest pain points that we face instrumentation goes down and have to fix it.

So for instance, we have an analytical instrument that went down and I had to take the thing apart and learn how to fix what we needed. ⁓ so it's little things like that that you don't think about. And then, you making sure that you're making on everything, ⁓ organizing and payroll and HR and safety and everything else. then you add in, you know, kind of the health aspects of this.

too, ensuring that your employees are safe and you're safe too in the lab. Monisha Saldanha: What kind of team or culture are you intentionally building at Cime? James Brewster: Yeah, I think this goes into why Colorado is so amazing. The people that we have are the kind of people that when your analytical equipment goes down, they're willing to go stop what they're doing and fix it really quick.

So it's a team of doers. And I think that's, you know, again, what drew me to Colorado. It's kept me here and what makes me excited to build here. is that's the culture that we have is at the end of the day, we all want to make medicines.

And that is a really important driver and an important motivator for what gets us out of bed and gets us into work and the ability to explore as well, utilizing these tools in AI ML that we maybe wouldn't have been able to do, definitely wouldn't have been able to, or would have never tried to use when we were at Pfizer, integrating in robots and physical AI to figure out how can we enable humans to do more of the fun stuff, to do chemistry and not cherry pick inventory or go and grab vials of chemicals day in and day out.

And so that's what really excites us as well is the ability to, lack of better words, kind of have this childlike wonder or the ability to explore and have fun while also doing impactful. Monisha Saldanha: So be curious to know like how you have a highly technical product. How do you approach the fundraising? How do you explain what you're doing in a way that people that aren't scientists can understand?

James Brewster: And so that takes some practice as well. It's maybe the difficulties that a lot of us technical folks have is we over explain or we dive into the weeds or we add all these kind of asterisks that are, well, you you caveat things. And I think it's just showing excitement about what you're building, really proving the novelty and the capabilities of what you've built. Right.

And it's really important to show is it's not just kind of a fun scientific thing, but there's the potential to really drive value and make an impact in the world. also how you frame your pitch depending on who you're talking to. I think I was a little more technical in chatting you. ⁓ But it up so it makes it really easy to digest, going back to that chem printer.

Everyone can imagine a printer and feeding paper through, and then the ink depositing on the printer, ⁓ then you change ink for a molecule. And so it's little analogies, I think, that help out a lot. Monisha Saldanha: Yeah, fantastic. Great.

Well, this has been a great conversation. James, thank you so much. I've learned a lot. I think this would be very useful for audiences who are curious about how AI is being applied in science.

⁓ Do you have any last words on AI? ⁓ James Brewster: Yeah, absolutely. So I think we're at this really interesting time in history, or in the present right now, where AI is everywhere. And we're starting to see a lot of it come out in drug discovery.

And it's a lot of hype. And it's pulling, I think, a lot of people from tech into biotech. Unfortunately though, you know, maybe a controversial take, it just hasn't delivered. And so I think we're going to start hopefully seeing over the next few years as this finds its place in the toolbox of drug discovery, just like structure-based drug design and everything else before.

But yeah, it just hasn't delivered yet. And so it makes it this kind of interesting from the business side and pitching to the scientist side where we're still waiting to see the hypothesis be answered one way or the other. Monisha Saldanha: But you mentioned earlier that you're using AI to more efficient in your drug discovery. So you are seeing some benefits from AI.

⁓ James Brewster: Yeah, absolutely. I guess to expand a little bit too, it's more on like the prediction side. So I need to run these experiments and I need to make this specific molecule and I need to go after this exact target. really just haven't seen it delivering kind of those predictive senses yet.

⁓ from a tool perspective, its ability to write Python script is, ⁓ code is amazing. And so we can use the control instrumentation and... ⁓ everything else and I think it does a really good job of synthesizing and we run thousands of reactions at a time and so the ability to pull information from a thousand data points which a human kind of struggles with, does that really well. Monisha Saldanha: So some good benefits from AI, but maybe still a lot of hype.

Wonderful. And then last question for you. What's one book that every builder should read and why? James Brewster: Yeah.

Brad Feld has two excellent books or a number of excellent books. And there's a number of excellent books from the venture groups that are in Colorado. ⁓ But one of my favorite books is by John Roberts. like the right place at the right time.

⁓ I it tells a story we've all heard before of prepared for when the opportunity comes. And then being able to on that is extremely important. And then kind of these aha moments of potential for discovery and advancing things for humankind. Monisha Saldanha: Fantastic.

James, thank you so much for this conversation. I've learned a lot. I'm sure our audience has learned a lot. Thank you so much for being here.

James Brewster: Thank you so much. It was an absolute pleasure. Thanks. Monisha Saldanha: And I'd like to thank our audience for listening to this podcast.

If you found this a useful conversation, please do share this podcast with others subscribe so you can see what comes next in our series on people building great products in Colorado. Thank you and goodbye.