Behind The Tech with Kevin Scott - Dr. David Baker: Director of the Institute for Protein Design, University of Washington

Episode Date: May 24, 2021

Technology is influencing protein design research in ways that would have been science fiction a decade ago. Find out how engineering and biology are working together today to improve our health and how researchers are designing new biological compounds to fight chronic disease - and prepare us for the next pandemic. Click here for transcript of this episode.  Kevin Scott Rosetta@home Foldit

Generally, in engineering and biology, if you want to solve a problem, you go look in nature for a protein that already does something similar, and then you modify it a bit. The relationship between sequence and structure has been really mysterious, but now we can actually design proteins with intent to do new things. So I think it's just getting to Behind the Tech. I'm your host, Kevin Scott, Chief Technology Officer for Microsoft. In this podcast, we're going to get behind the tech. We'll talk with some of the people who've made our modern tech world possible and understand what motivated them to create what they did. So join me to maybe learn a little bit about the history of computing

and get a few behind-the-scenes insights into what's happening today. Stick around. Hello and welcome to Behind the Tech. I'm Christina Warren, Senior Cloud Advocate at Microsoft. And I'm Kevin Scott. Our guest on the show today is Dr. David Baker. Dr. Baker is a professor of biochemistry and the director of the Institute

for Protein Design at the University of Washington. His research group is focused on the design of macromolecular structures and functions. Yeah, I really do think that Dr. Baker is doing some of the most interesting work in the world right now. I for sure, I mean, without question, I would be choosing to study computational biology if I were a graduate student again today. It's just so fascinating what he's doing. And I can't wait to hear him talk about what he's doing and to hear the two of you interact. So let's chat with Dr. Baker. Our guest today is Dr. David Baker. Dr. Baker is a biochemist and computational biologist who has pioneered methods to predict and design three-dimensional structures of proteins. He is the director of the Institute for Protein Design and professor of biochemistry at the

University of Washington. Dr. Baker has received awards from the National Science Foundation, the Beckman Foundation, and the Packard Foundation. He's published over 500 research papers, been granted over 100 patents, and co-founded 11 companies. Welcome to the show, David. Thank you. Happy to be here. So for listeners who aren't familiar with your work, can you tell us a bit about the University of Washington's Institute for Protein Design and the work that you do there? Yeah, sure. So in nature, proteins carry out essentially all the important functions in our bodies and in all living things. And they've come through billions of years, millions of years of evolution, and sort of been optimized to solve

the problems that were at hand during evolution. And what we've figured out at the Institute for Protein Design is how to make brand new proteins and to design them not to address problems that were relevant during evolution, but to address modern day current problems. And that's really what we're focused on at the Institute. It's such cool work. And I'm interested how you got interested in the field. Were you interested in science and math when you were a little kid? Were your parents scientists or engineers? Like what sparked the interest that has carried you to where you are today? Yeah, it's interesting. I actually was not terribly interested in science as a kid, perhaps because my parents were scientists. And when I went to college, I actually did not

initially major in science. I was initially a social studies major and then wanted to be a philosophy major. And then it was really my last year as an undergrad that I switched to biology. And then I hadn't really done research before, but I got really excited about what I had learned, what I had been learning in biology, which was developmental biology and neurobiology. And so I decided I would try out graduate school to see what it was like. And I found I liked doing research and I liked, you know, interacting with other people and sort of to solve hard problems. But as time went on, as I was a graduate student, then the next stage in my career, postdoc, I got more and more interested in sort of more basic questions.

And when I had taken my first biochemistry class when I was a senior in college, they had talked about the protein folding problem. And it seemed really interesting, but everyone said it was too hard to work on. And so but then when I as a postdoc, and then really, when I came to the University of Washington, that I decided to focus on trying to solve that problem. Yeah, I remember, I think in 1993, I was a research intern at the NCSA at the Beckman Institute, working in the computational biology group that Shankar Subramaniam was running there at the time. And he was working on protein folding. And I remember how hard it was in 1993.

So it's, I mean, you picked a challenging problem to get inspired by. Yeah, well, I always tell my students, even when you're in a group, you have to pick hard problems that aren't solved. Those are the only ones really worth trying to work on. And so at that moment when you were thinking about being a philosophy major and you decided to switch to biology, was that a magazine article you read, like an inspiring teacher, a movie you watched, a book, or just plain curiosity? It was, well, I think it was a little bit sort of getting fed up. or a movie you watched, a book, or just plain curiosity? It was, well, I think it was a little bit sort of getting fed up. I mean, I started realizing that more and more of those questions were sort of issues about language and language games and that there wasn't really, it started seeming to me

that there wasn't really a way to make consistent forward progress like discovery. I've always liked trying to discover things. And then when I took the biology class in contrast, there were just all these new discoveries right and left. People were just working out how the principles of developmental biology, and it just seemed like this huge unexplored territory that was much more ripe for exploration than philosophy would have been. So you made the switch and you went to graduate school at Berkeley and UC San Francisco. And how important was the computational part of biology when you were in grad school?

Did you have to learn not just biology, but a little bit about computer science while, you know, while you were a student, or did you have to learn that stuff later? Oh, yeah. Well, so I was a graduate student in the lab of Randy Sheckman, who worked on sort of cell biology, how proteins get moved, sorted around in cells. And at that time, well, I was famous and outspoken for ridiculing anyone who sat at a computer because, you know, what people use computers for, they was pretty rudimentary. You know, it was kind of word processing and, you know, you could you could waste time on computers the same way you can waste time on them now.

So I did not do any computer programming. And really, my major my major interaction with computers, like I said, was ridiculous to anyone who touched one. So, you know, in a little while, I really do want to get into like what changed between then and now, because, you know, I think we'll talk about soon how much of the work that you're doing happens by influence of some sort of digital technology. It's still, I mean, like still the biology is what's driving, but computers play a big role, right? Yeah, absolutely. I mean, yeah, we're probably one of the biggest users of computers in all of academia.

Yeah, which is so cool. But so what did things look like after you graduated? Like what were you working on? And like what did that stage of your things look like after you graduated? Like, what were you working on? And then, like, what did that stage of your career look like? Well, let's see. At the end of my PhD, I sort of discovered how to recreate in a test tube a very complicated biological processes, the ones that were involved in sorting these proteins to where they need to go in cells. And it seemed to me it was going to be a very long, complicated process to sort of purify all the proteins out and figure out their mechanisms. And I had been getting interested in sort of more basic questions about, you know,

the structures of proteins, which I really didn't know much about. And so then when I, for my post doc, I went to David Agard's lab at UCSF, and he studied structural biology and protein folding. And the first day when I went in, there was this computer terminal on my desk. And I asked, what was that for? And he said, it's for computing. So then I had to kind of learn what that was. So that was one of the things I did. But I do have another sort of funny story about that time, which is that whole area

of the building was really devoted to crystal structure determination, determining the atomic coordinates of proteins. So one of the first days I was there, I went into the room where everyone was sitting at their workstations trying to trace a protein chain through an electron density map. And so I said, Oh, can I try? So I said, Sure. So I sat down on the screens and tried to trace the chain through. And I remember then that I have horrible 3d visualization capabilities, I was totally incapable of visualization capabilities. I was totally incapable of doing it. So everyone in the room turned to me and said, David, shouldn't you have checked out whether you're any good at this stuff before you went into

structural biology? That is funny. So maybe you could explain to the listeners a little bit about what exactly that looks like. So like you get a protein, you try to crystallize it, like at the time you were shining a bunch of x-rays through it and sort of looking at the diffraction patterns of, I mean, like I'm probably describing. That's right. No, you're absolutely right. But so like that's just an incredible amount of work just to figure out what a protein looks like when it folds itself up, right? Yeah. Yeah. It's a lot of work and it's very difficult too, because you have to coax the proteins to crystallize. And I don't really, I don't think I, this is probably would be,

it's probably not exactly how it was, but at some point around there, I decided I really wanted to learn how to predict protein structure from sequence so that you wouldn't have to do all this crystallography and chain tracing and all this stuff but yeah and so yeah so so like talk talk about that a little bit like it's sort of a big leap i i guess from this idea that you're using a bunch of scientific tools in the physical world to try to spy into the world of biology much of which we can't see at all, you know, at this molecular level that you're, you know, that you're working at to trying to go from this sequence of amino acids and say, like, if I understand the sequence of acids, like, what is this thing going to look like?

And what a protein looks like dictates a lot of what its function is in an organism, correct? Yeah, you're absolutely right. There couldn't be anything more different than the two processes which have the same end goal, which is each protein, each gene in our genome encodes a unique protein that carries out a unique function. And it does so because the DNA sequence in the gene encodes a brick, their sequence of amino acids, which then folds up into a unique three-dimensional structure. So it had been known since the 60s that the amino acid sequence of a protein determines its precise three-dimensional structure. But like you said, the way that people actually have been

figuring out what the three-dimensional structures are is not reasoning from the amino acid sequence, but instead building hundreds of million dollar equipment and x-ray beams and all this complicated stuff to try and work out what the coordinates of the atom, where the atoms are. But all the information we know is in the amino acid sequence. And so then when I moved to the University of Washington, that was really the problem that I decided to focus on, like to really look at the simplest possible cases of protein folding and understand how they work and do that both with experiments to try and understand what the steps what what the process the key key aspects of the process were and what the determinants of protein folding were and then on the computer to try and develop methods for actually going straight from the sequence to the structure and so why is it

important to be able to predict the structure of proteins using a computer or any other mechanism. Yeah. Well, the same reason it's important to be able to determine the structure of proteins is because, like I said earlier, proteins carry out essentially all the important functions in our bodies and in everything in life. And so if you want to understand how biological processes work or how disease comes about, You have to understand the interactions between proteins. And those are kind of, a lot of them are kind of like Glock and key where things are fitting together very precisely. So you have to understand the geometry, how these structures fit together. And also if you want to understand how they work, like how they, how they generate,

they capture solar energy and convert it into formation of chemical bonds, all those things, you really have to understand the structures in the same way that if you want to know how a machine that does any arbitrary thing works, you really have to know what it looks like. Yeah. So we may be at a really unusual point in time where people probably know a little bit more about like protein structure and what its implications are than they ever had before because of the COVID-19 pandemic.

Correct. So, you know, maybe in terms of SARS coronavirus too, like, can you describe? Yeah, that's a great idea. Really good suggestion. In fact, now when I give talks, I explain protein design in the context of coronavirus. So let me just spend a couple minutes describing what we've been doing at the Institute with regard to coronavirus. So the genome sequence was determined and made available at the beginning of last year. So we took that amino acid sequence and used the methods we've been developing to predict the three-dimensional structure of the protein on the surface, the spike protein. Of course, you're right. There's higher literacy about this now

than there ever was. And we knew that the spike protein found the ACE2 receptor on the target cells. So starting initially with that model and then shifting over to the x-ray crystal structure when it was determined of the spike ACE2 complex. The first thing that we did was to design small proteins that we predicted would fold up into such a way that they'd have a shape and chemical complementarity to the part of the spike protein called the receptor binding domain that binds the ACE, binds ACE2. So these are like, I talked about sort of lock key interactions. So if you imagine the ACE2 is the key and the RBD is the lock, so it's sort of the spike protein goes and binds to the ACE2. We basically made things that would compete away that interaction that is bind more

tightly to the virus than ACE2. And we were able to make compounds that bind to the virus about a thousand times more tightly than ACE2. And this was really cool. They were just completely made up proteins, completely unrelated to anything that had been seen before. And with our collaborators, we were able to actually determine experimentally how these small proteins bind to the spike. And they bound basically exactly like in our computer model. So that means we could go from essentially from the sequence of a virus to these very, very tight high affinity binding proteins. And the next thing we showed was that those proteins block the virus from getting into cells. And then we showed with collaboration that they protect animals from infection by the virus. And I think this was kind of a real aha moment for me because we'd been developing these methods for designing proteins over the years.

And here in the midst of a pandemic, we were actually able to apply them to make therapeutic candidates. And those are now headed for clinical trials. It's been slow because this is a completely new modality, this whole idea of computational design proteins. So there's been a little bit of a pushback because these are completely new things. No one knows how exactly how they'll behave. But for the next pandemic, we're going to be ready. So we have all the methods worked out. And I think we've gotten a lot over a lot of the sociological issues to actually using these as drugs. And there's nothing really that can be as fast if you can go from the amino acid sequence to actually computing a

protein which fits perfectly against the virus. So that's the first thing we did. The second thing we did was to design, again, completely from scratch, little molecular devices that emit light, luminesce, when they encounter the virus. And those are pretty neat. We're developing those now for not only for detecting the virus, but also for monitoring responses to vaccination, like how good are my antibodies against the virus? And so rather than that being just like a fixed key that fits into a lock, that's actually a device that can undergo changes in its state when it encounters the virus. And the third area, my colleague Neil King at the Institute has been developing sort of a next generation of coronavirus vaccines using designed protein nanomaterials that we've created at the Institute, has been developing sort of a next generation of coronavirus vaccines

using designed protein nanomaterials that we've created at the Institute, which self-assemble into big things that look like death stars. And we can put the parts of the coronavirus spike on the surface. And when Neil does that, it finds it gets very, very strong immune responses, stronger than with the current vaccines. These designed nanoparticle vaccines are now in clinical trials. So that sort of illustrates some of the key areas in protein design now, being able to design very precise shapes that can block, combine very tightly to targets, being able to design molecular devices that can undergo, that can basically do logic calculations, and being able to design nanomaterials like these protein death stars. Yeah, to me, this is some of the most incredible stuff that I've ever seen in my life.

It's just amazing to me that you can go from a published sequence for this virus, simulate a compound that has this incredible binding affinity to the RBD, like in this computational domain, and then, you know, just be able to go from there to like diagnostics, vaccines, you know, like potentially therapies, like I'm guessing even that you could use these same techniques to try to assess how effective a human antibody response might be to the variants of the virus. Yeah, we're actually doing that now. Yeah, that's it. Yeah, you know, it sounds kind of science fiction-y to me, to you. It also still sounds that way to me, which is why, I mean, this field has been moving so fast. We've been able to

make so much progress to the things we can do now that still are a lot of work, but I don't think I would have thought would be possible just even, you know, several years ago. And that's what's been exciting. I mean, people always ask me to predict the future of the field. And I always say, well, my biggest hope is that it will move so fast in such new directions, I can't predict it. And that's actually what's been happening pretty continuously. Yeah. And so what are the big things that are moving progress forward right now?

Like what is, I mean, like you can sort of take a long view, right? Like I think the landscape was very, very different when you started than it is now. Yeah, it's totally different. What are the big things that have? There are just, there's so many things that are coming together. And I think when you have technological revolutions, that's generally what's happened. There's just been a lot of things that come together. And I think we just happen to be at the right place at the right time. So what are they? Well, our understanding of the basic principles of protein folding have been improving over the

years. And we've been developing this computer program called Rosetta to model these for 20 years now. So we've been doing a lot of work and sort of trying to get more and more of the details right. Second is computing power. These things are, despite my ridiculing people as a graduate student, these calculations now are, these systems are really complicated. And the basic principle is that proteins fold to their lowest energy states. So if you want to design a sequence that folds to a new structure, then you have to search through this huge landscape, and it's very time consuming. So the fact that computers have just gotten more and more powerful continuously has really

opened up. And we couldn't have not have done what we did for coronavirus in that amount of time with computing even of 10 or 15 years ago. And the third that's really completely independent is because of the genome project, there's been this huge advance in gene synthesis technology. So we design on the computer, we can compute, you know, millions of different possible proteins that are, you know, all possible creations, but actually bring them into the lab. We need to encode them in synthetic pieces of DNA, synthetic genes. And after we do that, we can put them into microorganisms. Who will then produce them? But synthesizing DNA is still expensive, but it used to be hundreds of times more expensive.

And so now we can routinely design 100,000 brand new proteins for all kinds of different applications. I mean, the range of things we're doing now in the group, you know, trying to create plastic, trying to capture solar energy, trying to do exotic chemical reactions, all sorts of things. You know, this just would not have been possible, again, with the technology of 15 years ago. So, being able to manufacture DNA very rapidly and cheaply is really enabling us to move forward quickly. And these things all go together because since we can compute, since we understand the principles, we can compute very large numbers of designs

because we can manufacture so many genes. We can then bring them all to life and then we can do measurements to see which ones work and which ones don't. And then we can fold that back in to improve the methods. And that brings up the fourth thing that's coming in, which is deep learning. And deep learning has just advanced by, you know, in the last 10 years has just

been incredibly, has advanced as fast as the other areas, if not faster. And so now we're complementing the Rosetta picture folding, which is sort of this physical model where you have this protein chain folding up with more of a deep learning approach, which is basically looking for recurring patterns and relating sequences to structures using those. And that's turning out to be very, very powerful. We can also use machine learning to try and interpret, to basically relate all this data that we're collecting to improving the computational models. And the fifth thing that's happening is that for the first time, these de novo design proteins are being developed as drugs. So there were probably four or five different proteins that we've designed, which will be in clinical trials this year. And so it's,

we're kind of at this inflection point in all these different areas. And so, you know, we don't really know how they'll turn out, but if one or two of them, you know, actually work pretty well, then you can imagine that, I mean, we'll have de-risked this whole kind of platform. And I think that the things that way engineering and biology has worked up until now is not intentional in this way. If you want to make a drug that binds to something, you either just screen a huge random library of compounds or you try and coax an animal to make an antibody against it.

There's no, generally in engineering and biology, if you want to solve a problem, you go look in nature for a protein that already does something similar, and then you modify it a bit. The relationship between sequence and structure has been really mysterious. But now we can actually design proteins with intent to do new things. So I think it's just getting to be very exciting. There's so much that you just said that I find fascinating and I want to follow up on, but like one of the things that as a computer scientist, that wasn't super obvious to me as I started diving deeper into biosciences is the extent to which you are,

we're now at this point where you can actually use biological machinery to go do work. So like what you just described a minute ago is you want to synthesize a protein, but you don't synthesize the protein by like having some 3D printer that just lays out a bunch of amino acids. Like you program a little piece of genetic material to use the cellular mechanisms that turn DNA into the protein that you want. And like, that's sort of like, that's just fascinating to me that we understand enough

about some of the basic biological machinery that you can leverage that to do some of the work for you. Yeah, I left out, there's a sixth thing I should have added to my list, which is exactly what you're saying. The whole recombinant DNA protein expression thing, our advances understanding basic biology, which we play a lot

of, we use biology right and left when we're, when we're making these proteins and then we're screening them for activity. So you're absolutely right. That's another thing. It's like all of these different technologies have come together. Yeah. And so I'm, one of the things that you, you know, that you've described is the, the end state of is uh de novo drugs uh like being able to design therapies for diseases like covet 19 and like in the limit like what you would be able to do what you'd like to be able to do you know for cancer therapies for instance is you would like to be able to look at like the very specific uh cancer mutation that you have, like making someone ill and be able to custom tailor a therapy to like the very particular illness that you have. And in order to

like, there's sort of two axes that I think are interesting. Like one is upper respiratory viruses are probably a little bit better understood than cancers. But like the trick with them is like you need to go very fast. So you need to go from new virus to, you know, therapeutic for it to drug that you can start delivering. And like the amount of time that that takes dictates like how bad, you know, a public health crisis you're going to have. For custom tailoring a therapy to an individual, like what you're trying to drive down is cost. So like we have some of these therapies right now, but it may legitimately cost a million dollars to synthesize the treatment just because it's very complicated. And, you know, you, you have five people who have the sickness a year.

So like, do you see both of those things getting better? So like time to, you know, deliver a therapy or like the cost of delivering a therapy or those both getting better? Yeah, I think there's also the precision of the therapy. So you brought up a number of really good points, but one of the issues with current protein drugs, which is why they're so expensive, is their antibodies.

They're very complicated proteins. Antibodies are what we make to defend ourselves against disease. And so naturally, in the sort of spirit of sort of emulating what's in nature, when pharmaceutical companies want to solve a problem now, they try and make a new antibody. And then that's the drug. The problem with that is that antibodies are very expensive to manufacture, and that contributes to the high cost. These proteins that we design, in contrast, can be manufactured for a hundredfold lower

cost. They can be made in bacteria, not in complicated mammalian cells, which are much more sensitive and much more difficult to grow. As far as precision goes, antibodies, the way they work is they're kind of blunt instruments. They hone in on a target. They're basically just, they're like the coronavirus binder I described. They just bind. We can, with design protein, we can actually design logic systems that actually can go

into the body and pick out cells that have combinations of proteins on their surfaces that are indicative of disease. Because in some cancers, there isn't really a single distinguishing mark. There's a number of things or some things are higher and some things are lower in abundance than they would be in normal cells. And you need to be able to resolve those differences. And so you need more sophisticated types of more sophisticated drug when they, like I said, can sort of do logic calculations in the body to really approach that. And in terms of time, you know, we're not there yet. But since everything's going on in the computer, then in principle, the reaction time should be much

faster, say to a new pandemic than it is if you have to. So with the antibodies, the antibody therapies for COVID-19, they weren't designed on the computer or designed by anybody. Instead, there was a lot of searching after and in the bodies of people who had been infected with the virus or animals for antibodies that happened to bind to the virus and really be effective at blocking it. And it's pretty rare that you find those antibodies. So if you can design things by intent, I mean, it's kind of like the way that biological drug discovery and engineering has worked. It's sort of like the way that biological drug discovery and engineering has worked. It's sort of like you're trying to build a building, you keep throwing a pile, you know,

bricks into a pile and you hope it assembles into a building. Well, it's much better if you understand the principles of construction and can just build it. So I think the future will be very bright. Yeah. And I'm just always reminded how complicated these biological systems are. Things are very rarely as neat as you as a computer scientist or a mechanical engineer have. You think, I will say for myself as a computer scientist, I often think that I'm dealing with complicated systems.

But compared to biology, our artificially engineered digital systems that we're building are like nowhere, even remotely as complicated as a biological system. One of the things is, you know, I think we've all looked at these vaccines as you have two types of vaccines now that are prevalent. You have sort of the mRNA vaccines that we are very excited about, and you have things that look more like classical vaccines. So they use something like an adenovirus to carry the spike protein into the body to try to produce an immune response. But one of the funky things about adenoviruses, right, is like if you already have antibodies

for the adenovirus itself, like the antibodies may swarm in and kill the vaccine before it can produce like the new immune response. So I mean, I'm just sort of wondering, yeah, one of the things that you said that's really fascinating to me is I think we very quickly, especially by classical vaccine development standards, went from the genetic sequence of this virus to vaccines or therapeutics. How long did it take you to design this thousand x binding thing once you had a sequence like a month two months it was probably it was probably um it was a few months but that was because we hadn't done it before and we had to sort of learn by doing and i think now we're working on methods for really doing that what we'd like to do is be able to do that within two weeks so the sequence of a new pandemic threat comes out and then two weeks later

we have a really high affinity antidote. That's aspirational, but I think it's possible. Yeah. And like, you know, everything that I've seen seems what, like, you know, far better than I do, but like, it sounds like a reasonable aspiration to me. Like the thing that I don't understand as well is then the hard part starts, which is trying to assess the safety of the thing that you just synthesize and like that's you know even with the mrna vaccines that we uh that we made like we had a pretty good

idea that they were going to work uh like we just didn't know that whether or not they were going to be safe yeah um and so how how do you i mean like can we use the techniques that you are describing to like do some of that safety assessment to drive the time? Yeah, it's a really good question. And that is probably one of the biggest question marks currently. So what we can do is on the computer design in properties that we think will correlate with safety. But humans are very, very complicated.

So it's hard to predict exactly what will happen. And so I think that is evaluating safety is the thing that is slow. And that actually is what has been taking all this time much longer than the design. So it's a good question. And there, we'll just have to learn from experience, I think. So what are you most excited about over the next handful of years? And like, what do you think we as citizens, folks who are very encouraged by what you're doing, who want to help things go faster,

what can we do to get some of this goodness that you're building moving faster? Well, let's see. I've been very interested in involving the general public and scientists at large in what we're doing. So we have an online game called Foldit where we post a lot of the current problems we're working on. For example, now one of the Foldit puzzles is that we've designed small proteins, which bind to different parts of the coronavirus. And now Foldit players are being challenged to connect

them in just the perfect way. So you get sort of the much stronger binding. So Foldit is one way. And then we have a project called Rosetta at Home, which is how we do a lot of our computing. If we send out jobs to people's computers and they compute new design proteins and send them back, and then we select from those which ones to make. So those are two of the ways in which people can get involved in what we're doing. Yeah. And like, in my opinion, like that's a much more beneficial thing for like our collective human wellbeing than like using your spare GPUs

to do cryptocurrency mining, for instance. Yeah, that's right. So, you know, just in terms of your work, like what are you really excited about over the next handful of years? Like how do you prioritize your energy among like all of the many, many, many things you could go explore? Well, there's a magic ingredient to all of this, which I haven't really talked about yet,

but that's part of the answer, which is that the most brilliant people in the world, graduate students and postdoctoral fellows are now coming to the Institute for Protein Design to sort of push the next wave of discoveries and make their fortune and start out exciting new independent careers in this area. So what I actually do every day is I talk to these absolutely brilliant people in my group about, you know, new people come in and, you know, we sort of brainstorm different ideas. They go and talk to everybody. I have sort of a theory of scientific creativity, this idea of a communal brain where everyone's talking to everybody all the time and like connected neurons, you can get these really emergent things.

So in terms of the new areas that we're going into, there's really, it's very wide and it's driven in part by the interests of people come in. But some of the really exciting areas are, well, really sort of pushing ahead on this very rapid therapeutic design, making better vaccines, biological machine, you know, designed machines, nanomachines that can do work. And for example, you know, we're making rotary motors now, kind of couple them to energy sources.

We're really excited about making new types of materials. So in nature, there are things in nature, we have examples like bone and tooth and seashells where it's proteins interacting with inorganic compounds to create all sorts of, you know, hybrid structures. And I think that's a really rich area. What if those things that you were interacting with were semiconductors rather than, you know, calcium carbonate, you could, you know, the sky's the limit there. Catalysts, trying to catalyze chemical reactions, which don't exist. We have a big effort now in exploring molecules that really nature couldn't explore because they're made out

of more exotic stuff than just the 20 naturally occurring amino acids. So new classes of compounds, we can take the same computational methods we've been developing and apply them well beyond outside proteins. And we're making systems that can sense the environment, respond. You know, I could go on and on. I'm excited about it all. Well, and it is sort of super exciting. And like going back to the intro like your your bio like you've started you co-founded 11 companies and so you know i i think beyond the yeah the fact that this is all scientifically some of the most interesting stuff in the world right now and it has this huge potential for you know impact in the same way that great scientific discovery usually does like you

also are like at the epicenter of this new entrepreneurial engine so like you're like in a lot of the ways like silicon valley uh like you go to a silicon valley school and you know like a bunch of professors there starting companies and like it's this vibrant ecosystem like i'm really excited and uh to see this happening with what you're doing yeah that's a really important part, because I told you with all the many brilliant people in their 20s are coming here to do great things. Many of them, it used to be they all wanted to go on and be professors. Now many of them come with the idea of starting companies, and many of them are doing that. I mean, I think

we'll probably be spinning out three more companies this year. And it's just, it's sort of accelerating. There's so many different things that proteins can do. And that's great because it is creating this whole new ecosystem. It's people and, you know, the people who don't want to start companies or take faculty positions instead are taking jobs at these companies. And the companies can then really go much deeper into the various application areas. We're kind of like a discovery engine, the Institute for Protein Design, but then actually bringing things out

in the real world is a whole other effort. And so that's what our spin-out companies are doing. Yeah. And then I also really have to think, the reason you didn't really touch on this, but what's really also fueling this is philanthropy, because, you know, what we're doing is so new, it's almost impossible to get grants to do this. And so a lot of a large fraction of our work is supported by private philanthropy, people sort of see this as that this really is the, you know, a technological revolution, and that pushing it forward rapidly will, you know, lead to all kinds of great new things for society. And we're really completely indebted to the people who are supporting our work.

Yeah, which I wish we could channel more funding. Like if I had a magic policy wand to wave, I would 10x, 100x the level of collective investment from philanthropy to government funding to corporate funding that was happening with this stuff. Like the return on those investments, both financially and in terms of goodness for society, like better health care for us all, for instance. I think it's just tremendous. So what would be your advice to students who are thinking about this as a career path? What should they study? What should they learn? It's it seems very interdisciplinary to me that, you know, it's

chemistry, biology, computer science, entrepreneurship, it's like just a bunch of things. I would say be a generalist is I mean, that's what you can sort of see from my trajectory. There's not, you know, the tech, this field is changing so fast that even if you were an expert in protein design today, five years from now, you would probably have to relearn everything. So yes, I think getting it's exactly the same. There's biology, there's physics, there's computer science, there's some sociology, you know, we live in the world. And so I think it's important to, yeah, to get a broad education. And, you know, and actually, when I look back at my own education is for a long time, I thought that, well, what the hell does social studies or philosophy

have anything to do with what I do now, but actually, it's turned out to be very, very useful in a lot of ways. So anyway, so I think that's in terms of a broad education. And then, you know, to actually, if you actually want to start doing this kind of work, being joining a research group, where in some capacity where this where such work, joining a research group where in some capacity where such work is going on is really, there's no substitute for that. Yeah. Well, so we're almost out of time here. The last question I like to ask everyone, and it's a weird question because everyone I chat with has such interesting work that they're doing in their professional lives, But I'm always curious

what you do for fun or what you find interesting outside of your professional life. Oh, yeah. Well, I love the mountains and are just going out, getting away from things, whether it's on the water or in the kayak or climbing or skiing or hiking. It's really what I love to do. During the week, there's a lot going on and a lot coming in and I need a little and having time just out in open space to process every once in a while is really great so it's yeah and you live in the pacific northwest like one of the most beautiful places in the world uh to be outside hiking or biking or kayaking or whatever you're that's right that's right and that's that's that's part of the reason I'm here. So, yeah.

That's awesome. Well, this is amazing. Like, thank you so much for taking the time to chat with us today. And more importantly, thank you for what you're doing. Like, I don't know whether you think about your work is like this great public service, but it very much is.

And so on behalf of many of us, like, I just want to express real appreciation and gratitude for the work that you and your students and your colleagues are doing. Well, thanks, Kevin. This was a lot of fun. And honestly, I'm not a surfer, but I just feel like we're riding the wave and we want it, you know, the longer we can ride it, the better hope that we can make the world a better place. That's really what we're trying to do. So yeah, this was a lot of fun. I enjoyed our conversation. Awesome. Thank you so much. Well, that was Kevin's conversation with Dr. David Baker. So what stood out to me about your

conversation was just how fascinating everything that's happening in this field is. Like, I honestly, I'm going to be honest with you, even after you said that this is what you would study in grad school, if you're in grad school right now, in my mind, I'm like, I'm not a traditional sciences sort of person. I don't know how fascinating any of this will be. And I'm just riveted by all the cool stuff that is happening in this space. Yeah, it's really crazy how fast things are moving right now. I think one of the things that he said that was really interesting and very, very true is that scientific revolutions, technological revolutions tend to happen when you have multiple things that are moving very quickly and have had their own transformations come together at the same time.

And he named off six of those in the course of our conversation. And yeah, it's one of those places where because it's moving so rapidly and because every day we're learning more and more and more about how to push things forward even faster, this is just where so much of the interesting things are going to be happening in the world over the next couple of decades is in this particular discipline in digital biology, synthetic biology, computational biology, where we really are trying to get this like high resolution and more accurate picture of what happens in biological systems and then being able to engineer those systems to solve really, really tremendous problems. Yeah, no, I mean, that's what struck me. I mean, like you said, he was talking about how all these different things kind of need to coalesce to have this real innovation. And when he was saying that some of the stuff that the industry has been doing and the scientific community has been doing with coronavirus wouldn't have been possible with

computers and technology from a decade or 15 years ago really struck me. And when he was talking about how the way that the engineering and the technology is working now is that it's intentional. Whereas before, you know, you'd look to nature and now you are intentionally and having this intentionality about saying, these are the problems we want to solve. And we're going to engineer a solution rather than looking for something that might already exist that can solve that problem. Yeah, I mean, in a very real sense, like many of these areas in biology now have a real engineering discipline that goes along with them in addition to the normal scientific practice.

Like we understand enough about the systems where we can start designing them in ways that serve purposes that are different than the ones that nature strictly designed them for through the process of evolution. And like, to me, that is the most fascinating thing. interestingness about this particular moment is we are all at a heightened level of awareness about both the upside and the downside of biology because of the COVID-19 pandemic. Like we know what one little molecule can do to impact hundreds and hundreds of millions of lives, and, you know, just sort of trillions of dollars worth of economic damage. And like, we also know that we can harness this greater understanding that we have about biological systems to, in an unprecedented way, design therapeutics and vaccines for the pandemic itself.

And like, all of that just gets better in the future. It does. And it makes it exciting. I mean, it's a little bit harrowing in some senses. And if you wanted to become really dystopic about it, it could be, you know, sort of concerning in that respect. But it's also so exciting, because it really does feel like the things that we've thought about in the past of being science fiction are really going to be things that not necessarily, right? Like that actually could be reality, which is, I'm going to be honest, I think if nothing else, I mean, I think it's exciting, but if nothing else, it's really interesting.

Yeah, I do think it's exciting. You know, I keep saying that I'm an optimist, even though, you know, my wife would look at me and say, I'm a grumpy a grumpy old cynic. But I'm very optimistic about the choices that we will make about how these tools and this greater scientific understanding of biology will get used. And I think you can absolutely see it with COVID-19. We have just had our scientists and engineers come together in exactly the way that you would hope to just pour time and energy and expertise into getting us past the misery of a global pandemic as quickly as possible. And I think not just the science has been a shining light, but the way that we've come together as a species has been a shining light for me. I agree. I agree. It's actually been pretty remarkable to see and to see what's happened in this relatively small period of time, all things considered. Anyway, this is a great conversation.

I love hearing about what Dr. Baker is working on in his work. That's it for our show today. We are so grateful to Dr. Baker for joining us. And you know that we love hearing from our listeners. So contact us anytime at BehindTheTech at Microsoft.com. Thanks for listening. See you next time.