In a studio interview at the JPMorgan Health Conference, Christine talks with Mike Nally, CEO of Generate Biomedicines, about the company’s mission to create a generative biology platform that can design novel proteins and medicines using AI. Nally explains how traditional biophysical approaches were brittle, leading Generate to adopt machine‑learning‑driven, diffusion‑model techniques (the “Chroma” paper) that have already generated three million previously unseen protein concepts. He outlines the rapid scaling of the organization—from 30 employees at inception to 244 today and a projected 375‑400 by next year—driven by five core values (audacious ambition, insatiable curiosity, unapologetic scientific rigor, fierce unity, and intrinsic digital) and 15 supporting mindsets that shape hiring, promotion, and day‑to‑day behavior. Finally, Nally shares a bold vision: within the next three to five years Generate aims to have its first approved therapeutics, dramatically shortening the traditional 10‑15‑year drug‑development timeline by operating as a computationally native drug‑discovery company.
Chapters
Welcome back to the flagship studio here at the JPMorgan Health Conference in San Francisco. We are here with Mike Nally from Generate Biomedicines. Mike, great to be here.
Great to see you, Christine.
There's actually a lot of talk about generative biology at this conference this year. It's something that you've been working on from the beginning at Generate. Can you talk about generative biology, generative medicine?
Yeah, the team at Flagship and at Generate started this journey 4 years ago, and in many ways we were trying to solve a problem that, you know, has befuddled scientists for a few decades. You know, if we knew how the code of life in the form of DNA dictates protein function, we could create novel medicines almost on a programmable basis. Unfortunately, when we tried to take biophysical-based approaches and build up atom by atom, those results lacked robustness, were brittle. And so in many ways, to try and solve this fundamental scientific question, we needed an orthogonal approach. At the time, we saw the promise of machine learning and we said, you know, if we could do this and actually understand these principles in a really profound way, we're going to be able to transform how new drugs are discovered. And, you know, over the past 4 years, the results have just been absolutely remarkable. We continue to learn more and more about nature. You know, biology, despite the fact that its majesty has stared at us, you know, for all of our existence, we really don't know much about it. And through the complement of complementing human ingenuity with technology, we're starting to understand biology in such a more profound way. And from a generative basis, you know, I think a lot of times people had always thought of AI as, you know, something to augment analytics.
Yeah, exactly.
But the reality is, is AI is foundational in terms of how we think about creativity and augmenting creativity now. And that's what the generative revolution is really unlocking is, you know, can we augment human ingenuity with computational creativity?
Speaking of computational creativity, Generate published a paper earlier this year that got a lot of attention and showed just the amount of mastery in terms of form of proteins. You want to speak a bit about that paper?
Yeah, the Chroma paper was a real groundbreaking paper. Obviously, we've seen this trend in diffusion models. That help us understand complex fields. But translating that to the complexity of proteins required significant innovation that our machine learning team really spearheaded. What it allows us to do now is, you know, over the 3.5 billion years that the planet has been around, we basically have had 20,000 protein concepts.
20,000.
Okay. This technology basically allows us to come up with new concepts that are mathematically correct almost instantaneously and on a perpetual basis. So we're expanding that opportunity set from basically one drop of water into an entire bucket of potential protein concepts. And if you think about that from a biologics and a, you know, human therapeutic perspective, these could be the new medicines that cure some of the greatest diseases in the world.
And so— So you're going from coding for medicine to coding almost directly for function.
Absolutely. Exactly. And so, you know, over the 4 years the company's been around, what we've tried to do is learn not only the protein sequence to structure interrelationship, and that's what we, some of the stuff we could learn through the genomic sequencing revolution, as well as kind of the protein databank had 200,000 high-quality protein structures. But since the inception of the company now, we've generated 3 million novel proteins that nature hadn't discovered and learned on the function of those proteins. And all of that data then is fed back to refine the computational approach.
3 million. Wow. Mike, you mentioned the inception of the company. How many people at Genentech How many people were at Generate when you joined?
30.
And now?
244.
End of next year?
375 to 400.
You think a lot about culture.
Absolutely.
How do you go from 30 to 350+ and maintain a culture of exploration, of boldness, white space? How do you think about culture? I know you think a lot about it.
So, you know, for us, it was really imperative, you know, almost from the day I joined We spent time as a leadership team defining what are the values that we most represent as generators. And so there are 5 core values. There are then 15 mindsets that support those values. The 5 values are: we are audaciously ambitious.
Audaciously ambitious, okay.
Right, so I think to your point on going into white space, the courageous activity that that requires, we want to always have that audacious ambition. We think about an insatiable curiosity. You know, we're trying to do things that have never been done before. We're going to have to look for inspiration from a range of different fields. And we want people who are in many ways polymaths, who understand the intersection of different fields, because I think that's where a lot of the promise of the future will come. We're unapologetically scientific.
Unapologetically scientific.
Yeah. Which is one of my favorite ones.
Yeah, interesting.
Because, you know, in some ways machine learning can be a black box. We need to have the highest standards of scientific rigor in everything we do, and I think that the company really evolves around that. You know, the, the next one is we're fiercely united. This is a team sport. Collaboration across the team is super, super important.
Do you call yourselves generators? Oh yeah, totally.
Okay, cool. And then lastly, we're intrinsically digital.
Intrinsically digital.
And the reason that's really important for Generate is that We recognize the scalability of the platform is one of the most amazing elements of it. We believe we are going to create many, many medicines. And as we migrate drug discovery from this artisanal craft that is down to individual genius and make it a much more engineerable endeavor, we're going to want to have a system that can scale with it. And if we don't think about technology, if we don't think about data from the outset, we're going to have a really hard time. And so, you know, these 5 values then translate into the mindsets. Those mindsets then are embedded in everything we do from our hiring guidelines, they're embedded in our promotion guidelines, our year-end reviews. We talk about this at all of our town halls in terms of who are the employees that best reflect these values. And ultimately, the people that succeed within the company reflect these values. And I think most importantly, I think we also give everyone the license and liberty to call anyone out who's not reflecting those values. And so, you know, our frontline scientists will go to me and say, Mike, you know, you're not being ambitious enough on that, you know. And so it's a, it's a culture where people can actually express themselves and we want to stay true to those values as we go forward.
You talked about what lies ahead. We're all talking about bigger leaps this week. When you think about bigger leaps for Generate, if we're here at JPM 3, 5 years from now, you're a veteran of this conference. What are you talking about at Generate? What are you sharing with investors and others that Generate is accomplishing?
I think what Generate's doing is, you know, in many ways it's accelerating the pace of drug discovery and development. And, you know, I think we are going to be in a position where we're talking about our first approved therapeutics over that time frame. And if you think about the traditional 10 to 15-year window, this would be a monumental step forward. And so, you know, we're continuing to think creatively and innovate not just in the kind of, you know, machine learning to protein engineering domain domains, but how do we become a computationally native drug discovery company that thinks about how we apply these sort of computational techniques to make sure that these life-saving medicines reach patients as quickly as possible?
Computationally native drug discovery. Move over, Terminator. The generators led by Mike Nally are coming for you. Thanks so much for joining us in the studio.
