 proteins are essential for life on the planet. They're tiny biological machines that perform important functions in our bodies, things like repair, digestion, production of energy, bolstering our immunity and so on. I find it hard to believe that the amazing array of proteins we see in nature makes up just a small fraction of all possible proteins. I'm here to meet a team who are striving to give nature a bit of a helping hand. Hello, I'm Esme, nice to meet you. I'm Karlie, nice to meet you. I'm Chris, nice to meet you. Nice to meet you guys. So tell me about this place, what are we doing here? So we're standing in the Edinburgh Genome Foundry and this is where we make the DNA instructions that we use to make our custom proteins and then we go upstairs to the EPPF and we express that we actually make those proteins and characterize them and understand how they behave. Incredible. Now this might sound like a bit of a silly question but what's a protein made of? No, that's not a silly question at all. Proteins are chemicals, long molecules that are made of building blocks called amino acids and this is an amino acid here, this is a representation of it and there's 20 different amino acids with different chemical variants here. In nature, or you can do this chemically as well, you can link together, thank you Karlie, you can link together these amino acid building blocks to make long protein molecules and on average proteins have about 200 to 300 building blocks inside them and so if you think about that, with 20 different options at each of these positions, even if you had 100 building block protein, that's 20 to the power of 100 possible combinations and that's 10 with 130 zeros after it, this is an absolutely staggering amount of possible proteins and only a tiny, tiny little fraction of those have been explored in nature despite the amazing complexity of biological systems. That's a huge number, so are you saying that nature's been a bit lazy when it comes to making proteins? Evolution is definitely a much better optimization algorithm than it is a sort of innovation engine, so it's much easier to make existing proteins better than it is to make totally new proteins but we don't have the same types of restrictions that evolution has, so we can make giant leaps instead of small steps. How hard is it to make a new protein? Well if you think of a protein as a tiny molecular machine that's trying to carry out a function that is incredibly important to model the physics that goes on in this, without considering the physics it'll be like a toddler trying to build a really tall stack of Lego blocks without considering stability. The frustrating thing about this is that many of the experiments will result in nothing, they will just come down crashing. Well I've heard you've got a demonstration that might clear this all up for me. Absolutely yeah, do you want to come and see? Let's go. So I'm seeing a lot of people in coloured tops but what actually are we doing here? So we're going to do a protein folding simulation but we're going to do it with people rather than molecules. Interesting so shall we see it in action? Yes let's go. Okay so what we're going to do now is we're going to make the polypeptide, could we have the end terminus up at the front? So what Chris is doing here is he's trying to mimic a computer simulation. Each person here is going to be pretending to be an amino acid which is a component of a protein and as you can see they're lined up with either left or right hand out and in a minute what we're going to do is ask them to play the simulation out and they're going to try to find other amino acids to either move towards or move away from. That's it this is the way this compact no no we've got a broken chain there we go there we go we've got a solid bridge we've got one solid bridge right we can rearrange can we rearrange? Oh can we break you up temporarily? Yes right compact compact everybody together right here right and then compact yes this is it we've got a stable right and stop this is great. What have we just seen here? So we ran our simulation and then the polypeptide chain the protein folded up into a compact three-dimensional structure or two-dimensional in this case. It seemed very chaotic is this a good structure? So yeah absolutely I think this is a really nice a representation of what happens on a molecular level you know when the forces that fold up the protein they make the chain wiggle and jiggle until it gets and it folds and unfolds and goes back until it gets to this energetically stable state and this is what we've got here basically hopefully you can see that we have all of the black amino acids in the middle these are hydrophobic they don't like water so they want to be hidden out from this environment here and we've got all of the blue and red and white which like water all on the outside and you can see that the blue and red have found each other and paired up so this is a good structure. What this simulation shows you is how a protein falls but one of the problems that we're excited by is the inverse problem where you give us the shape this shape for example to design a vaccine for a virus and then ask the question what sequence of colors should we line up in order to end up here all right thanks very much amino acids well that did seem quite difficult in the end absolutely so hopefully the demonstration showed you that it's difficult to design a sequence of building blocks that will make a protein that forms into a compact structure repeatably and this is really this difficult problem is the reason that Carti can I collaborate between the School of Informatics and the School of Biological Sciences to make new computer algorithms that we can use to make proteins that go beyond nature computer algorithms are very effective at detecting patterns and what Chris and I are trying to do is try to teach these algorithms to detect patterns that result in consistent successes or perhaps even failures now Carti I heard that you actually worked at Disney before you returned to a life of academia was it a similar skill set you used there well that was magic of a different kind about 10 years ago the animation and special effects industries rely heavily on approximate physics model but what's really cool about the work that Chris and I are doing together is that this goes beyond entertainment we're trying to solve problems maybe even speed nature up so you're speeding up nature with the help of technology now is this just science for the sake of science or is it science that can actually help us one day it's going to help humanity there's no doubt about that natural proteins are incredibly important we use them broadly already in all sorts of areas but we can design new proteins that go beyond nature then we can start to solve problems in medicine and agriculture related to sustainability in the environment and it's starting to happen now we really are at the beginning of a revolution when it comes to protein design well thank you so much for chatting to me guys and there you have it nature has so many proteins left undiscovered and experts are using AI to try and discover them and hopefully these discoveries will help us all out someday I'm so excited to see how this science progresses in the future back to you guys at the Royal Society