Titans of Science: David Baker

0:00This BBC podcast is supported by ads outside the UK. BBC Sounds. Music, radio, podcasts.

0:09Hello, welcome to 5 Live Science. I'm Chris Smith from The Naked Scientists. Coming up, a look ahead to the year ahead, 2025. We'll be asking experts in the fields of health, AI, astronomy, marine science and archaeology what we should look forward to and expect to hear about in the next 12 months. The Naked Scientists on 5 Live. In 2025, the UK is set to introduce strict new laws to restrict and control vaping. The aim is to create what's been called a smoke-free generation

0:43and a large part of that will include a ban on the sale of single-use disposable vapes in England, Wales and Scotland. Here's Linda Bald, who's a professor of public health at the University of Edinburgh. So what the bill is going to do, it was previously proposed by the last UK government and has been brought back with the same powers, but also some additional amendments. So it's quite a long list, I'll be brief. I think the headline one is phasing out the sale of tobacco to anyone born after 2009. But there's also quite a range of powers in this bill, which has passed its first readings.

1:18It's starting that stage into the other parts of the legislative process to also look at vaping. So vaping, as you and I have discussed before, is an important public health issue. Large numbers of young people are vaping. So the bill is aiming to have powers to curb youth vaping. And that means doing things like prohibiting branding on vapes that appeal to children, some of the names and the colours, and stopping handing out free vapes, importantly regulating vape contents and flavours,

1:49and also looking at how they're displayed and promoted in shops. And then the last bit of this, which is actually of particular interest to me in my research, is a whole bunch of nicotine products that are not medicinally licensed, that are not covered by our existing vaping legislation, like nicotine pouches, snus as they're often called, and they'll fall within the bill as well. Are they thought to attract the same kind of health disbenefit as vaping and smoking, these other forms of nicotine use? Or should we be turning more of a blind eye to those and bearing down on the elephant in the room,

2:22which is the fact that vapes seem to have crept in, especially in the youth market, and seduced and basically hooked a whole legion of new nicotine addicts? Yes, they have. And you're absolutely right. We have a continuum of risk. That's how I describe it. Actually, some of the other nicotine products, particularly nicotine pouches, so-called snus, although snus is something else, but that's what people call them colloquially, I actually think are really quite low risk, because all you've got in there is the nicotine,

2:52and some of the other ways that the product is packaged, these are things that we know are not harmful. They're also, it's also not breathed in, in the way you do with vaping, which is one of the reasons why we're concerned about vaping far less harmful than smoking, but still risks we don't understand. So nicotine pouches are probably not something we should be terribly worried about, but the reason I think it's good to bring them into the legislation, or at least some of the measures in the legislation, is because they're very addictive. And the other thing about the pouches is that the level of nicotine in them isn't regulated,

3:25and at the moment there's actually no age of sale. So we needed to do something on the nicotine pouches, but you're right, if you look at where harm begins, nicotine replacement therapy, a licensed medicine, very safe, but, you know, some people can get a bit hooked on it, through to things like pouches, then through to vapes, and then through to the smoked cigarette, which is the most deadly. There's different risks associated with all of these products. Are we not slamming the door after the horse has bolted on the vapes, though?

3:55Because it strikes me, we're coming to this party so late, they've already got their tentacles into so many young people, and we've actually seen rates of uptake go up, go flying up, actually, and we're not even at the point where we've banned these things properly yet. China, even China, did it ages ago. We do actually have a lot of legislation around vaping already, but clearly it hasn't been enough, and the problem has been that the products have evolved. So we've banned most forms of vaping advertising.

4:26We have an age of sale. We have enforcement. We have limits on the nicotine content. We have a health warning on the vape. You know, we did all of that stuff in 2016. But then disposable vapes came on the market, which are very cheap, very attractive, and disposable, as the name says, environmentally harmful, and they actually have been a major driver of youth uptake. So the horse has definitely bolted to the extent that we didn't do enough early enough. But I think taking additional action now will be okay. I would say banning them probably wouldn't be that successful.

5:00What the government aims to do with this bill is ban smoking in terms of selling it to children born after 2009. That's the deadliest product. But countries like Australia, which have made vapes only available on prescription, have youth vaping rates almost as high as ours. So I think it's about proportional approaches. I'm trying to keep ahead of industry evolution and the evolution of the products. I hope we can do it, but we have to be realistic. Totally different topic, but health related.

5:30We've had various measures that have been brought in. I think something like 170 policies in the last about 30 years by successive governments of all colours to try to control obesity, sugar tax included in them. That has got to be one of the health priorities for any year in the years ahead, hasn't it? Surely. Absolutely. And actually, I think you're going to see some changes in 2025, but they're not going to be enough for most of us in the public health community. What you're going to see in 2025, if I just start with Scotland,

6:02we're committed to taking action in legislation on multi-buys and temporary price promotions of junk food. The UK government is finally going to be bringing in its restrictions on advertising online and on television, particularly on TV, in relation to the watershed. The problem is our food supply has changed. We've got big, powerful companies, and it's changed both our diet, but also things like rates of cardiovascular disease, which we know some of the improvements we saw with productions and smoking

6:33are largely being reversed because of overweight and obesity. So I hope in 2025, we see more action. It's a public health priority, and we need to begin to tackle it. Linda Bald at the University of Edinburgh with her public health wishlist for 2025. We're now to tech, and we're going to look ahead at what we might expect to hear about from an artificial intelligence perspective over the next 12 months. This year, we anticipate the release of OpenAI's ChatGPT 5, the next generation of the AI language model that powers ChatGPT.

7:06But will it come to pass? And what might it be able to do? Here's Mike Wardridge, who's an expert on machine learning at the University of Oxford. So just recall the timeline. GPT-2 was the first notable release in the GPT series. That was around about 2018, 2019. GPT-3 was the breakthrough. GPT-3 was the first really, really impressive large-scale, large language model that met a large audience. And that was the technology, basically, that goes into ChatGPT.

7:38And ChatGPT is November 2022. Then we see GPT-4 rolled out fairly quickly after that. And everybody has been waiting for GPT-5. Now, I've had lots of conversations with colleagues and friends recently, possibly fueled by a few glasses of wine in some case, about where exactly GPT-5 is. And I've heard flatly contradictory, very confident, but flatly contradictory statements from people about where they think it is.

8:09So one view is that the technology has flatlined and that GPT-5, we haven't seen it yet because it just isn't that impressive. And the people who believe that, their beliefs are fueled somewhat by Sam Altman, the CEO of OpenAI, the company behind this technology. His statements that, you know, the race to scale, just making things bigger, was not delivering the same level of advantages that it did with GPT-3. And so that's fueling this idea that actually GPT-5 hasn't been released because it's a disappointment.

8:43So at the same time, I've had very confident statements from colleagues that, no, no, GPT-5 really is going to be mind-blowing. And it's just that OpenAI and co. are busy frantically testing it to make sure that it's really, really, really fit for the market so that when it's released, it really does dazzle. So I honestly don't know which of those two statements is true, but a lot, I think, is riding on it because the expectations are so high. Do you think that we will see it in 2025?

9:16And what do you, in fact, see in 2025? Is that big correction that we're considering must be on the cards with this level of investment, this sort of scattergun approach to investing in so many of these technologies, which is going to bring some companies financially to their knees if they don't discover the golden nugget of this is the use case that everyone wants? Is the reset coming in 2025 or something else exciting? Well, I think we will see GPT-5. I think it's unimaginable that we wouldn't see GPT-5 in the next year.

9:46What it will do and what the scope of its capabilities will be is anybody's guess. My guess would be right now that its language capabilities will not be hugely advanced. I think it will still hallucinate. I think it will still get things wrong. We will still start to see, you know, a few days after release, you will start to see on social media people say, look at this stupid thing that GPT-5 has done. That's my guess is that will all be there. But at the same time, I think what it will probably embrace is different modalities.

10:19So where I think that one of the big battlegrounds for generative AI is over the next over the next couple of years is around multimodal AI. And that means not just text, but text and images. And of course, in GPT-4, we've already seen text and images come together. But things like sound, music, video, being able to upload a video and actually having a an explanation of what's going on in the video, not just a transcript of the voices in the video.

10:50We can do that now, not perfectly, but we can do that now. But actually what's happening in the video, this is a video about a man taking a golf shot and falling over, you know, this kind of explanation. So multimodal, I think, is where it's really going to be. At some point in the next few years, and maybe even next year, we're going to have generative AI that will be able to take a prompt and generate a TikTok-length video to order.

11:20And, you know, TikTok is one of the biggest social media platforms out there right now. And when we have that kind of capability that we're in a different world, I think, of a new world, a completely new world of entertainment. So multimodal, all of those different modalities, you know, being able to produce music. And again, we've seen prototype systems that can do this to order. You know, I would like a mashup of Joy Division and Ed Sheeran, you know, something like that. And produce a song that's a mashup of Joy Division and Ed Sheeran or an Ed Sheeran song in the style of Joy Division, something like that.

11:55All of those different modalities coming together. And what people are going to do with that is going to be very weird and very, very wacky. But it's going to take us into a completely, a completely new era of entertainment and media consumption. Mike Wildridge from the University of Oxford. And a Joy Division slash Ed Sheeran mashup might not be our cup of tea, but the technology is all the same pretty impressive. What might be in store for space travel in 2025?

12:26Well, the answer very much hinges on how much Donald Trump's incoming administration wants to spend. We put in a call to space scientist and author David Whitehouse. In the Trump administration, which is no respecter of tradition and for which success and achievement counts more than heritage, it may well be that Elon Musk is able to overrule many of the directions and programs that NASA has, particularly in its human space program.

12:57Because there was a study just released which showed how America plans to get to the moon. The astronauts at the moment, it's going to go to the moon on the Orion capsule. They're going to transfer to the starship, go down to the surface and then come back, back up again and transfer back to the Orion capsule to come back to Earth. When that plan was actually put in visual terms, we saw pictures of it. People have started saying, but Orion is tiny. Starship is enormous. This doesn't look right. This doesn't look compatible.

13:29This looks very strange. Now, either that's the way to go to the moon, or it's telling us something about meshing the government, expensive, long-term space effort to go to the moon with Elon Musk's nimbler, faster, more risky approach. Elon Musk has launched it more than anybody else put together into space. His starship could be launched five, six, seven, eight, nine, ten times a year as you get going.

14:00The space launch system that launches the Orion capsule, that's to take astronauts to the moon, probably cannot launch more than once every two years. So these are two very incompatible systems that people are beginning to realize that perhaps, even though power is too much with Elon Musk, one would argue, he might be the way to get back to the moon before the Chinese. You've written some really powerful words, because you said 2025 will be the most important year in human space flight since 1969.

14:36Very, very powerful. What do you think is going to happen then in 2025 to justify those words? Well, I think this is the year of decisions. This is about how to go back to the moon.

14:50Because since we've been to the moon, we've had basically the space shuttle and the space station. And the space shuttle we know was a flawed system. And we're going to have to face, Elon Musk's, one of his big tasks he's going to have to face is the same problem as the space shuttle had with putting people on board his starship. Because the starship is a capsule. And when he does put people on board, and it seems that he wants to put, you know, design it for a crewed flight relatively quickly.

15:22His main problem is that at the moment, there doesn't seem to have an abort system in the sense that the crew are inside the capsule. And the rocket and the capsule are next to each other, as was the space shuttle. And that, you know, when Challenger blew up, the crew had no way to get out. So he has got to address when he puts people on board the starship. And that's part of going back to the moon, because at the current plan, the crew will transfer to the starship to go down to the moon's surface and come back again.

15:54He's got to work out the safety issues of that. And that is going to be very difficult, no matter what his ability to send the starship around the solar system. And we expect to send him an uncrewed one to Mars in the next couple of years. That is a fundamental human safety problem that he is going to face. But success is everything at the moment with the Trump administration. And the decisions for going back to the moon for what capsules America uses and how it explores the solar system will probably be made this year.

16:29Because at the moment, you have you have the starship, the up and coming, what could be a very versatile, very powerful spaceship to take people into space, to the moon and possibly onto Mars. You have got the Dragon capsule, which is also SpaceX, that takes people up to and from the space station. You've got Boeing's Starliner capsule, which had all these problems and wasn't able to bring the crew back from the space station earlier this year because they had problems with the thrusters.

17:00And those are still not resolved. And you've got the expensive and capable Orion capsule that is going to take astronauts to the moon and is complicated and has problems with its heat shield at the moment. So those overlooking the American space program, and Trump will say this to Musk, Musk will say, how come we have got Starship, Starliner, Dragon and Orion? Do we need four ways to put American astronauts into space?

17:30Surely, with wanting to save money and be more efficient and achieve more, which is what SpaceX has shown us it can do, something's got to go. So I would argue big decisions this year, next year and the year after. And that will set the agenda and the goal, the technological direction to go back to the moon, which I think is going to be much later than we thought now, and possibly onto Mars. David Whitehouse, watching this space for us, and what lies ahead and overhead in 2025.

18:05Now, the year ahead is predicted to be the hottest on record, which has been something of a trend in recent years, let's face it. But what effect do we anticipate it having on the marine realm? And what, if anything, are people doing to protect us and marine life from the harshest of the effects of climate change? Will Tingle's been speaking to marine science communicator and host of the Out of Our Bubble podcast, Liberty Denman. I think it's going to be all eyes on COP. And that's not just the COP for climate, but also for biodiversity.

18:36This is something that I think they've suffered a lot of backlash, both in the contents of what comes out of COP, but also the locations that they've been hosted in the last couple of years. I think ultimately what we're waiting to see is that action, both the biodiversity and the climate side of it, and seeing the countries having the political courage to make the moves to do that. I think conservation suffers so much in the political spaces. Political terms are short, and so politicians think short term. But unfortunately, the natural world and conservation doesn't work on four-year terms or two-year terms,

19:06depending on where you're looking. It's something that needs to have that long-term view. And unfortunately, I think that's something that it really suffers from, because when you're constantly changing things, it's very difficult to see that long-term impact. So I think that's really where we want to be looking for the ocean. It's seeing that commitment to long-term protection, whether it's for specific areas of ocean under that kind of MPA framework or targeted areas for specific species protections for different areas, different species, depending on what's necessary, I think is really where we want to be turning our attention to in 2025.

19:37And the elephant in the room, whenever you talk about conservation, is, of course, climate change. 2025, some people predicting it is going to be, again, the hottest year on record. That kind of feels meaningless at this stage. Consistently beating ourselves every year. We love the competition with ourselves. I mean, I feel like this is an obvious yes, but how will that potentially have impacts on the ocean? I mean, let's just imagine you're stuck in a bath where the water is really, really hot. How's it going to impact you? You're going to be sweating, feeling a bit sick, dehydrated, probably also a load of other things that I'm not thinking about.

20:09But ultimately, it's that same concept. The animals and ecosystems within the ocean can't go anywhere. And that ocean is warming. And that can look very different depending on what areas you're looking at or what species you're looking at. I mean, generally speaking, as the ocean warms, that will cause a sea level rise. But it will also cause a change in the chemistry, making it hard for things like crustaceans, crabs, to maintain their shells. To put my pessimist hat on just for a moment, obviously, I'm sunshine and rainbows for most of the day.

20:40But when we talk about marine protected areas and we wax lyrical about them in the look back at 2024, how can they even hope to reconcile a fraction of the damage that we know is going to be caused by this behemoth of climate change? I mean, how you define like a vulnerable area to protect does, again, completely depend on the classic science cop-out answer. It depends. Depends on what you want to be protecting. But to cover the broad strokes, what we're actually talking about is the fact that nature-based solutions are actually really where we're starting to look.

21:12As we kind of realised, especially over COVID, as we kind of left things alone, nature kind of started coming back. There was all those headlines like we're seeing dolphins in these rivers and all of these different things were starting to reappear as we kind of receded our impact on those areas because we weren't allowed outside for any more than an hour a day. And so I think what that really did show us is that nature-based solutions, and if we can leave nature alone in that way, it will recover. It's just how do we actually make that work in a society that isn't functioning under the power of COVID?

21:44And so marine protected areas are obviously a super important part of that, literally kind of just designating an area to leave alone to recover because then that allows the spillover effect into those areas that aren't protected, which means that we will have increased fishing abilities, increased ecotourism, perhaps more resources to use, all of these different things. And so ultimately, that's what we need to be doing is for that collaboration across different sectors, showcasing the mutual benefit of protecting these regions and how it's going to be useful to everyone. And also understanding that you don't just have to be in marine conservation to be able to make a difference in this.

22:17You know, blue finance is becoming increasingly something we speak about. How can we actually inject money into the conservation sector that is so chronically underfunded? And therefore, difficult to protect, because it doesn't have the same financial benefit. Protecting kelp forests and seagrass doesn't have the same financial benefits. It does tearing mangroves down for a hotel. And so that's what we need to understand is where that middle ground is so we can ultimately be benefiting from nature in the way that we do every day. And perhaps even putting a financial value on that so that we can kind of understand that in the Western world of how we hold value.

22:51So a great deal of cooperation and cohesion is going to be required. Definitely. I think that is going to be the biggest thing that we all need is collaboration moving forward. Absolutely. Liberty Denman in conversation with Will Tingle there. Finally, we're going to look forward to archaeology in 2025. As we heard earlier, AI is set to have another huge year across many scientific fields. So how much of a game changer is it in archaeology? I went to see Emma Pomeroy at the University of Cambridge to find out that and to hear what else she's getting up to in 2025.

23:25I think this is a really exciting development and I'm sure we'll see more of this in the years to come. But essentially by using things like LIDAR, so where you're using kind of light, how it kind of bounces back from the ground when you're sort of flying over it or sending a drone over a particular area, it can almost let us see things that we couldn't see before. So using this sort of technology where you send up a balloon with a camera or something, I mean, that's been done for a very long time in archaeology and then you can see structures on the ground.

23:57But some places that's really hard to do, for example, in tropical forests where you've got very heavy plant coverage or, for example, at high altitude where it's not so easy to get there. And there's been a number of really interesting studies that have used this technology to kind of see through those tropical forests or to look at places that are much more inaccessible and show us really exciting things about sites of kind of scale that we wouldn't necessarily have predicted or been able to study before.

24:29And what's in your diary for 2025? You've normally got somewhere exotic to go and something ancient to dig up. More of the same? Yes. So hopefully we will be going back to Shanidar Cave in Iraqi Kurdistan for more excavations there. We've been working extensively over the past kind of 10 years now finding Neanderthal remains, but also trying to understand the behaviour of Neanderthals and more recent modern humans. And we're hoping to go back there to focus on this period where we have a transition between Neanderthals using the cave and the first modern humans turning up

25:06to really try and understand what that sort of transition looks like. Are they there completely separately from one another? Are they perhaps overlapping in time, alternating in their use of the cave? So that's one of our big targets for the coming year. Emma Pomeroy at the University of Cambridge there. Well, now it's time for the news and sport. But after that, Titans of Science continues with the man who helped to crack one of the hardest problems in chemistry and biology. And that's predicting the shapes of proteins, including the ones that make our bodies look and work the way that they do.

25:40He's the 2024 Chemistry Nobel laureate and biochemist David Baker. We all can agree that housing is expensive. Rent, mortgage, it doesn't matter which one you're paying, it stings every month. But BUILT can make you feel a little better. Let me explain. BUILT started out rewarding members on their rent. Now, as of 2026, BUILT members can also earn points on mortgage payments wherever they live. Every housing payment earns you points you can use toward flights with top travel partners like United & Hyatt, Lyft Rides, Amazon.com purchases, and so much more.

26:15And here's what you might not expect. BUILT members can also get access to a neighborhood concierge. It can make restaurant reservations, book fitness classes, and find new local spots, all while being rewarded at more than 45,000 merchant partners. It's like having a personal assistant baked into where you live. It's simple. Being a renter and now owning a home is better with BUILT. Join the membership for where you live at joinbilt.com slash podcasts. That's J-O-I-N-B-I-L-T dot com slash podcasts. Make sure to use our URL so they know we sent you.

26:49This year, experience the NHL playoffs your way. Want to watch a single? Do or die Game 7. A full week of heavy hitting back and forth series. Then check out Slings 1-3 and 7-day passes. Watch the chase for the cup on ESPN and TNT, starting as low as $4.99. No long-term contracts or overpaying. Just live hockey at home or when you want it. Sling lets you do that. Visit sling.com to learn more. Welcome back to 5 Live Science with me, Chris Smith.

27:24And today, I'm in conversation with Nobel Prize winning biochemist and protein guru, David Baker. David Baker was born to two parents who were themselves both scientists in Seattle on the 6th of October 1962. He attended Garfield High School in the city before he read biology at Harvard University. On graduation, he then began working on how proteins are transported around cells. Later, he would go on to pioneer methods to design proteins and predict their three-dimensional structures.

27:54And that helped to earn him a share of the 2024 Nobel Prize in chemistry. David's co-founded several biotechnology companies and he was included in Time magazine's inaugural list of the 100 most influential people in health. He's now the director of the University of Washington's Institute for Protein Design. Welcome to the show, David, and congratulations on your Nobel Prize. How did your interest in science get started? Well, perhaps surprisingly, I really became interested in science relatively late.

28:27I'm not one of those people who was fascinated by science from an early age. In fact, when I got to university, I initially declared my major to be social studies. And later, I got interested in philosophy. And it was really not till my last year of college that I decided to switch to science and focus on biology. But proteins. I mean, I remember my biology teacher at school, I wasn't very old, obsessing about proteins. I really didn't get what all the fuss was about. Why did they matter to you?

28:57What drew you into that? And why do they matter to anybody? Well, in fact, at the time I was in university, I had no idea what proteins were either until I took a biology class. And it seemed interesting to me. But it wasn't until quite a few years later that I really started becoming obsessed. And I'll tell you why. In nature, you know, biological, you know, organisms, animals, humans do all kinds of really, really amazing things. And if you look in detail how those things are accomplished, at the heart of everything are proteins.

29:30So they're specialized proteins that mediate the electric currents through our brains while we're thinking and talking. There are proteins that allow us to move around. They're proteins that enable plants to capture solar energy from the sun and use it to make molecules. Basically, everything that goes on in biology is done by proteins. So the way that biology works is there are all these different jobs, thousands and thousands of jobs in any organism. And for each job, there's a specific protein.

30:02So you can kind of think of proteins as the miniature machines which do all the important things in life. And what's their structure? How would I recognize one? Well, you wouldn't recognize one if you saw it because they're extremely small. They're just one nanometer across, which means that you need a trillion of them to get to a meter. But each protein has a very well-defined shape. That's one of the really kind of miraculous things about proteins. If you think about the machines that we're used to encountering in real life, each machine has a very defined shape, which is really important for it to do its job.

30:37Like a car has wheels so it can roll and it's got an interior compartment you can get into. And in the same way, every protein has, you know, a very defined shape. And those shapes are really what lets the proteins do what they do. How do they come by that shape, though? Proteins carry out all the work in our bodies and all living things. Like I said, the instructions for making proteins are in our genomes, in the DNA. And so the DNA in our genomes specifies what the chemical structure of each protein is.

31:09That is, what the sequence of amino acids is going to be. Proteins are made out of amino acids. There are 20 different types of amino acids. And a protein is a linear chain of about 100 to 500 amino acids. And the sequence of amino acids of a protein completely determine what its shape is. And that sequence, as I said, is specified in the genes in our genomes. And those amino acids are all different chemically. So they can have different shapes, different structures, different sizes, different electrical

31:42behaviors. And so that means, depending upon which ones you slot in, it's a bit like building a wall with different shaped bricks. You're going to get a different shaped wall or a different colored wall or a wall with interesting properties if you put different amino acids into it. Yes, that's a very good analogy. It's kind of like having a universal building block kit, kind of like a child's construction toy, where you can basically make any shape by having the right combination of amino acids. And it's not just the shape. A machine also has to interact with whatever it's operating on in the right way.

32:15So by having certain types of amino acids on the surface of a protein, that will enable it to interact with other proteins or with DNA and carry out the job it's meant to do in biology. By the time you were obviously doing your studies and getting into these sorts of questions, it had been nearly half a century since Watson and Crick worked out what DNA did and that it was storing the code that told cells how to make these magical things, proteins that were going to go on to dominate your life. So what was it at the time that intrigued you?

32:46What was the unknown or the unanswered that you wanted to explore further? There was a very important observation made in the 1970s or late 60s, which was that if you took a protein and you pulled it apart, it would fold back up to the same shape. And that was what really proved that the shape of a protein was determined by its sequence of amino acids. But nobody knew how that worked, how that code for going from amino acid sequence to three-dimensional structure worked.

33:17And no one really understood what the process of this, what this folding up process was. And I became fascinated about that when I joined the University of Washington as a professor many years ago. I think I found it fascinating because it's kind of the simplest case of self-organization in biology. So if you think about all the non-living things around us, they're not really organized. They're kind of all random. But then you look at your pet, a dog or a cat, or your brother or sister, animals and plants

33:48are really highly organized. That's how they're different from the rest of the world. And proteins are really the simplest case because proteins are made out of thousands of atoms and you would expect them just to be completely random jumbles of different shapes. But yet they just have one shape. And so I was really fascinated about it from that point of view. And also, like I said, proteins carry out all the important jobs in living things. And so I thought if we could understand how proteins fold up, we might be able to actually

34:18make new proteins at some point, which could then have a huge number of possibilities. So we would think, well, I want to have a miniature machine at the scale of atoms capable of doing job X. And I think it would have to have shape Y to do that. And therefore, I would design a bespoke protein that would do that. That was sort of the end goal we had in mind at the time. Yes, exactly. Well, what was stopping you doing that? Well, it's very complicated, the process of going from amino acid sequence to three to dimensional structure or three dimensional shape, or going backwards from a three dimensional

34:53shape to a sequence of amino acids that will encode it. As I said, proteins have many thousands of atoms. And so the way we started working on this problem was to try and model all of the interactions between all of those atoms. So to think about the process of protein folding up, we sort of developed methods for kind of actually modeling that folding up of this very long protein chain and guided by all these thousands of interactions between atoms.

35:23And we got to, we were able to make some progress on that problem. And then we realized we could go backwards and take a brand new shape and go backwards to figure out what combination of atoms and what amino acids would have the property that they would actually fold up to that shape. We were actually able to do that about 20 years ago and design a new amino acid sequence that folded up to a new shape. And that was really, that kind of really opened the door to a lot of possibilities because once we could make new shapes completely from scratch, it seemed like we should be able to

35:59design new functions. Make, given any job that you might want a protein to do, we should be able to design a protein to do it. Is it tricky to do this sort of thing because all these atoms are different? They're different sizes, different masses, different electrical charges. And that's going to mean that if you just had two atoms to consider, A and B, and you could quite easily understand how they would probably stick to each other or want to attract or repel each other. Once you've got thousands of them and they're all doing their own thing independently, you've

36:30got thousands and thousands of possibilities to consider. Is that why it's a difficult problem to solve? Yes, that's one of the reasons because there's so many of them and there's so many different possible combinations. The other thing that makes it challenging is we don't know exactly the details of how those atoms or amino acids interact with each other. And so for that reason, there are errors when we do these calculations, which have to be done on a computer because there's so many atoms. And so there's so many different possible places each atom could be.

37:00And the interactions between the atoms, you know, we kept improving our description of those interactions on our ability to try out different number, more and more different combinations of atoms. And that's why we were able to make progress in designing more complicated proteins. It does sound, though, like something that would be more tractable with heavy duty computing, because you can ask a computer to consider all of those different possibilities. And it may take a while, but wasn't that what we built supercomputers to do, to be able

37:32to do loads and loads of calculations and model this sort of thing? That is exactly what we reasoned. So, in fact, we started a distributed computing project called Rosetta at Home, where we enlisted people all over the world who had computers at home and who were running screensavers to run a screensaver that would do these calculations. It would both predict protein structures and it would design brand new structures. And we were able to enlist quite a few volunteers.

38:02And actually, Rosetta at Home became equivalent to a medium-sized supercomputer. And so that, you know, I think community involvement in our science has always been really important. Rosetta at Home then led to, we had a screensaver so you could see the protein getting designed or folding up on the computer screen. And people would watch the screensaver and they would write to us and say, you know, it's really cool, but I think I could do better. And that led us to develop an interactive game called Foldit, where the participant could

38:35not only have their computer fold the protein, but they could get in and actually guide it. So this is sort of like the biochemical equivalent of what went on to become the SETI at Home concept, wasn't it? People downloading bits of data to their home computer, which could then be crunched on their computer, their electricity bill. Very crafty, David. And that led to a resolution or an analysis of a piece of data that when brought back together, you had that enormous wealth of computing power by harnessing thousands of computers around

39:07the world. Yes, that's exactly right. In fact, we didn't have to build up the infrastructure for this because the SETI at Home group and developers were incredibly generous. And they actually helped us to use their entire platform and infrastructure for connecting, you know, many, many, many personal computers together to do these calculations. Except the difference was rather than processing signals, radio signals, you know, the participants in our project were folding and designing proteins.

39:39What were you sending to the users? And what were they when they were first doing this before you got onto the game that you developed that you just mentioned? But when you were just showing them data that was being crunched, what were they picking up from your server? What was their computer doing? What was it showing them that some people were then latching onto and saying, I think I know how to improve on this? Well, in when we were trying to figure out how a protein folded up, as we said, there are many, many different shapes that any protein could have. So we would send out just the amino acid sequence of the protein and then the participant's computer

40:15would fold it up and it would send us back the folded structure. And what we would get back from this is hundreds of thousands of different possibilities for how the protein could fold up and from which we could identify those that were the most likely correct solution. And the principle we use to evaluate that is similar to expectation when you have a ball rolling on a bumpy surface that it will eventually end up in its lowest elevation point.

40:47Well, similarly, as I said, we work in these calculations where we look at the interactions between all the atoms and we calculate the energy of each protein in that way. And so we select out those proteins, those shapes for which the amino acid sequence had the lowest energy. This was about the turn of the millennium, early noughties, wasn't it, that you were doing this? So that got you a bit further along, but there was still clearly a gap because this didn't immediately revolutionise our ability to predict proteins or you would have won the Nobel Prize

41:18a lot longer ago than you have. So what was still the stumbling block at that stage then? Well, the stumbling block for both structure prediction and design were just the ones that I described, that proteins are very complicated and they're made out of many thousands of atoms. So really doing accurate calculations. It was really hard to get really accurate structure predictions, for example. On the design side, we were able to design more and more powerful proteins doing a wider

41:49and wider range of jobs. But we had to try a lot of different designs to find one that really worked well and solved the problem that we intended it to solve. So the real game changer was the advent of deep learning. And that was really demonstrated in a spectacular fashion by the DeepMind team, my co-laureates John Jumper and Demis Hacibas, who showed that database of protein structures was sufficiently large that one could learn from it the rules of protein folding and go from an amino acid sequence

42:24directly to a three-dimensional structure. So I have to tell you one thing, though, just to put this in context. Before it was possible to predict the protein structure, the structure of a protein from its amino acid sequence, scientists around the world spent many, many years, and actually still do, determining the structures of proteins experimentally. That means figuring out where in space each atom of a protein is. And they do this in a number of ways. For example, one of the most powerful is shining x-rays at a crystal of the protein and figuring

42:58out how those x-rays scatter. And that gives you direct information on the position of atoms. Now, tens of thousands of scientists over 50 years, at an expense of tens of billions of dollars or more, spent their careers determining the structures of proteins. And many scientists, great scientists, are continuing to solve the structures of more and more complex proteins. And so what this led to was a database of about 200,000 different protein structures. And each protein structure specifies exactly where each atom in that protein is relative

43:31to the others. So it's this incredibly rich storehouse of information. And what the DeepMind group showed is that this information store was sufficiently detailed and rich that you could really learn the rules and predict structures of proteins from their sequence. You feed in to the artificial intelligence all of that wealth of information where people have painstakingly worked out where the atoms are in three-dimensional space in each of those proteins so it can then learn.

44:03And that presumably means you can then feed it an unknown protein, an amino acid sequence. These are the building blocks of a protein you've never seen before. And it can apply the same rules to then work out what it would look like. That's exactly right. So the program that the DeepMind group developed is called AlphaFold. AlphaFold was trained on all the amino acid sequence of proteins of known structure. It was trained to predict the structure. And so now you can give a new amino acid sequence to AlphaFold and it will generate the predicted

44:35structure for it. One of the things that the award committee said was that you achieved the almost impossible feat of making new proteins. So this was essentially upstream of what we've just said. You proved that you could make a new protein from scratch. You could come up with a concept and design it. And I suppose what the DeepMind team then did was to equip you with a way of doing that far faster. Well, yes. So as I described, when we started designing proteins long before deep learning methods

45:08had, before deep learning was even, you know, a well-established field. And we used this sort of atomic description that I described earlier where we had to model all the interactions between pairs of atoms. And we used that approach to design completely new proteins. And that was what was cited by the Nobel Committee. That was back in 2003. After DeepMind showed that protein structure prediction could be greatly enhanced using deep

45:43learning, we naturally were very, very quickly moved to apply deep learning to protein design. And what we found is that we were able to develop very powerful methods for designing brand new proteins that were much better than the previous sort of methods based on sort of this cloud of atoms I described earlier. And using these new design methods, we can design proteins that have a very wide range of different

46:13functions. And we have made these methods freely available to anyone in the world. And so we're, it's very exciting now because we're seeing many different research groups designing new proteins using the deep learning methods we've developed. So there's going to be maybe, you know, 10, 15 years ago, the idea of making a, trying to solve a problem in biotechnology or in sustainability with the design protein just sound totally crazy on the lunatic fringe. But now there's really great interest in designing new proteins to solve problems in medicine and

46:49sustainability and technology. So it's a very exciting time. Could you, for example, to think about how we might deploy something like this, could you say, well, look, ocean and marine plastic pollution, that's a major headache. I want to design an enzyme that has never existed in nature. It's a protein that can attack plastic in the ocean and get rid of it. Could we throw that sort of problem at this sort of solution now and begin to build protein machines that would do that sort of job for us?

47:20That is exactly the type of problem that we're working on now. So there are several extremely talented researchers in my group who are working specifically on that to design catalysts that will break down plastic. We're also working on ways, new ways to fix CO2, as well as new proteins that will very specifically target cancer cells in the body so you can treat the cancer without systemic effects. It's an exciting time also because we have our first medicines that have been approved for use in humans.

47:50And that's a vaccine, a COVID vaccine developed by my colleague, Neil King at the Institute for Protein Design here. People often say that it gets interesting when things break or don't work. So when you do this, are there any things that trip up these artificial intelligences, things that they consistently get wrong and shouldn't? Because often there might be something interesting lurking in there. Have you noticed anything like that? Well, in fact, in every problem we work on, we only work on problems which are kind of at the cutting edge of what's possible.

48:22Because the really easy problems we figure, you know, people in other places could do with the software we're releasing. And whenever you work on a hard problem, you only understand about 30, 40% of what's going on. And so one of the things, really key thing, is you start working on a problem like targeting a tumor or breaking down plastic. And the first few designs you make don't work or they don't work very well. And then you have to look at what's going on that's wrong. Then that gives you ideas on, you know, what do you need to improve about your design strategy

48:53or the methods to really solve those problems. And so that's really largely what science is about is, you know, having some hypothesis about how to solve a problem, trying to solve it. And then it doesn't work as well as you thought. And then trying to figure out what the basis for that is and improving your method and approach accordingly. Some branches of science are also now going down the synthetic route, where when we began this conversation, you explained a protein is something made from one of a combination of 20 different amino acids.

49:25But we can, as clever chemists now, make amino acids that don't exist in nature. So we can therefore do chemistry that may not exist in nature. Can the artificial intelligences be brought to bear using these novel chemicals, though? Because, of course, we won't have that vast database of proteins that use these new chemicals we're creating to train on. Exactly. So this is where the previous methods that were based on modeling all the interactions between the atoms still are very useful.

49:56So we're trying to do exactly what you described, build catalysts now that incorporate unnatural amino acids and unnatural cofactors into our designs. It's like having our machine now has this kind of totally new, powerful thing in it that will allow it to do more sophisticated chemistry. And this is where combining the new deep learning methods, which, as you pointed out, are really used to just seeing the natural 20 amino acids, with the previous methods that I described, where we're modeling everything as just a collection of atoms using physical principles.

50:31That combination is powerful because those older physically based methods have no problem modeling that unnatural amino acid or cofactor as just a collection of atoms connected by bonds. It's so exciting, all of this. And you can really see how this is going to translate and quickly into really groundbreaking stuff, can't you? But when you go home at the end of the day, is your head spinning because of this? Or do you have crafty ways of managing to relax or sort of get away from it and not think

51:03about protein structures for a while? So I live in Seattle, which is fortunate because I love the mountains. So on the weekends, I try and get up skiing or hiking or climbing pretty much every weekend. And so right now, in fact, it is a little rainy in Seattle, but that means it's snowing in the mountains. So I'm excited to get out and ski this weekend. Good for inspiration, I should think. Because Cary Mullis, who got the Nobel Prize a number of years ago for discovering and coming up with the idea of the PCR reaction to copy DNA, he told me he came up with that

51:36concept driving up to his mountain cabin at Montesino. So maybe your trips into the great back of beyond are very inspirational. Yeah, they certainly helped me preserve sanity, which is very good. Well, thanks very much for telling us all about it, David. Congratulations once again on your Nobel Prize. And I hope that with the Nobel Prize comes a bigger office because it sounds a bit cramped. Well, so far, I would say the Nobel Prize has been remarkably useless in trying to improve our research conditions or resources, but I'm still hopeful.

52:07That's it for this time. But if you'd like to get in touch in the meantime, our email address, as ever, 5lifescience at bbc.co.uk. From me, Chris Smith, thanks for listening. And until next week, goodbye. I'm Matt Shorley, coming to you live from Westminster, Monday to Friday from 2 o'clock. It's politics, but how it affects you. It's the funny stories. The people and the personalities that make up the Westminster village. But I want to explain how it works. Order! Order! I've worked there for 20 years now.

52:39I really know all the people, the quiet corridors. And I know how to find the stories about what's really happening in politics, which affect you. As soon as the news breaks, I can bang on the wall and one of the BBC correspondents will come running in and tell us what's going on. Matt Shorley. Monday to Friday from 2 o'clock. On BBC Radio 5 Live.