 The thought is that if something comes along and it kills all of the non-resistant cells, all the fast cells, then the slow cells can repopulate that population. Whereas if you had no slow cells, the population as a whole would be able to grow a little tiny bit faster in the absence of any bad things happening, in the absence of stress. But if some stress coming along, it would kill off the whole population. It would kill off everyone. And then you have zero cells and now you're evolutionarily dead. Biologists want to understand things. They want to understand how cells work. They want to understand why we look and behave the way they do. And we want to understand things too, but we want to do so in a quantitative manner that I can write down my predictions for the future and see if they come true or not. If I have a population of a million cells and I give them some drug that kills 99% of them, I want to be able to tell you before giving them the drug which cells will survive and which cells will die. If we can do that, then we can really understand the system. No, someone's opinion may contradict yours. Where's my friend Alan? It's all about your perspective. Who are we and what is the nature of this reality? Ni Hao, everybody. Welcome to Simulation. I'm your host Alan Sokian. We are still in the beautiful Beijing, China. We are at Peking University School of Life Sciences. We are now going to be talking about Mechanistic Prediction in Biology. We have Dr. Lucas Carey joining us on the show. Hi, Lucas. Thank you so much for coming on the show. Really appreciate it. Love your work. It was so cool getting to know it deeper and hanging out a bit more last night, talking about all our cool ideas and stuff like that. It was great. I'm so pumped for this show. For those that don't know Lucas's background, he's the PI, Principal Investigator of the Carey Lab at Peking University Center for Quantitative Biology. His lab is focused on understanding and predicting how individual cells and populations behave and you can find his links in the bio below to the Carey Lab as well as his Twitter profile. Alright, Lucas, let's start things off with one of our favorite questions to ask our guests. What are your thoughts on the direction of our world? I think generally it's getting better. The rate of change is decreasing these days, I think, but I am confident that this is temporary. I certainly hope that this is temporary and that the relations between China and the US will get better. I'm not worried for the long term. Yeah, that's why we're here. We're here to hopefully open up more friendship and collaboration and global cohesion between the US, China and the other countries in the world. What would you say about the rate of change? Because so many people come on the show and they talk about, oh my gosh, the rate of change is increasing. There are so many exponential technologies unleashing into the world and it's being democratized. So where do you see it actually slowing down? Well, I see more protectionism coming in in the US. We see this with decreased immigration, with making it more difficult for foreigners to get green cards and to convert their PhD that they get in the US into a job and a family in the US. I see this also in China that it's more difficult than it was three years ago, five years ago for foreigners to use payment systems and rent bicycles and things like this in China. And so this is I think slowing down. Oh, okay. I think I see. Although the exponential technologies are continuing to boom, there's still some sort of like a slow down of like, there's some sort of conservatism that is still happening at the same time. And I think that immigration in all directions is one of the things that gives you technological development. A lot of the founders of major companies in the US are immigrants or children of immigrants. And so if we want the technological, the rate of technological progress to continue to increase as fast as it has been, the world needs to remain more open in my opinion. Yeah. I was mentioning that so many times now it's like, I can't email some of the Peking University professors because my Gmail or Yahoo emails land in their spam boxes. I can't get a WeChat pay or an Ali pay here because I don't have a Chinese bank account. I can't link my visa to it. So people look at me like, why are you giving me this Archaic cash scan my QR code? So it'll be interesting to see how much easier it is like it took a long time to get a visa to come to China. It's a quite a long and tedious and multi-day process. And so you wonder what could be the future of being able to transit between countries safely vetted in a vetted way. What could be so easy as to just have money not need conversions and all this type of stuff. But it's interesting the cryptocurrency revolution where that could lead us and to make it easy for people to be able to create creatively collaborate across the world. Like if you want founders of the most massive companies in the world to continue for that process, continue happening, making it easier for people to be creative is a core aspect of that. Lucas, who were you growing up? It was Woodstock in New York where you're born and then you went to Stony Brook. But how did you get interested in genetics and in science when you were a kid? So I don't know the remember the exact moment, but the sort of milestone is in fifth grade. I did a genetics project for the science fair where I did crosses of hamsters. And so we ended up with 34 hamsters in my house, the offspring of the parents that I was doing the cross with. And my, you know, hypotheses. So we got more offspring that were white than were golden would look more like the female than the male. And my hypothesis was that they spend more time in their mother, and therefore they have more maternal traits. So that was completely wrong. But that was kind of the first memory kind of created by my parents keeping the poster. So I don't know how much I really remember and how much I remember from the poster, but definitely I was doing science then doing genetics then. Cool, cool. And I love hearing about the stories like that. It makes me want to enable more children around the world to have access to science projects like that and mentors that can help them with that because those moments are so pivotal in picking up what you care about. I can also just like 34 hamsters in the house like the parents. Yeah, the young scientist. And then now teach us about the trajectory to Stony Brook and through Stony Brook, which is on Long Island on the northern part of Long Island. Yeah, yeah, cool. On North Shore. Yeah. So Stony Brook was great. I got there. I had done some research in biology and also in computer science as an undergraduate. And so I started my PhD and I ended up just kind of hanging out with the guy who was going to become my PhD advisor at the first year of PhD students like a retreat. This was when alcohol restrictions were a little bit looser in the US and the Stony Brook would take all the PhD students out to a hotel in kind of farther east on Long Island and get all the students and professors drunk for the weekend and get some mingling. And so I thought it was great. And so I started talking to this guy who I had no idea who he was and then ended up doing my PhD with him. So we worked on cell cycle control. So yeast cells and also mammalian cells don't divide into from one cell into two cells until they've reached some volume. And so they have some way of, a single cell has some way of sensing how big it is. And we don't know how that is. Like a cell doesn't have a ruler that it kind of puts up. So what is it measuring exactly? Is it measuring diameters, it measuring volume? Is it measuring the amount of protein? And so I spent my PhD trying to work on what do cells actually measure to know how big they are. Whoa. And what did you find out? So we found out that, found out a couple of things. It was a long time ago. That's why some people say that. Like, damn, what was it? Yeah. So one of the things we found out is that the, so we're basically looking for something that's increasing while a cell is growing in size. And if you had a ruler, that ruler might remain constant as the cell grows in size. And so one thing that remains constant, so yeast cells in particular and also mammalian cells, this measurement happens before cells replicate their DNA. So while they have just a single copy of their genome in what we call G1, before DNA replication. And what remains constant is the amount of DNA. And what increases is the amount of everything else inside the cell as it grows bigger. And so what we found was that the amount of these unstable activators that promote progression into S phase and through the cell cycle, they increase very quickly through this period. But the number of binding sites for them in the genome of course remains constant because the DNA replication is not occurring. And that if we increase the number of binding sites, we delay progression through the cell cycle. If we decrease the number of binding sites, we promote progression through the cell cycle. And so it seems to be this, the ruler is kind of the ratio between the number of these promote protein molecules inside the cell and the number of physical binding sites for them in the genome. Whoa. Okay. Okay. So there's a quantitative mechanistic code to the cell when it decides to divide. And it has to do with the amount of proteins that are inside of the cell. Yeah. And then to the binding sites on the DNA. Yes. Okay. And as those get to a certain threshold, then it can divide. Exactly. So you can imagine, there's some constant number of binding sites, and then the number of proteins are increasing. And then eventually they reach some kind of ratio, threshold, interest. And then the cell can divide and be able to sustain. Yeah. And then, okay, interesting. So that could be then that moment where it's like, okay, there's enough proteins, there's a certain amount of binding sites. What if we divide now, we'll be able to sustain, we'll be able to live. And that's, yeah. Exactly. My goodness. When we were talking about stuff like this last night, we were just going off about how there's just so much quantitative mechanisms in biology. And where you are basically like just diving deep into the, how these things are actually codified and how we can explain them and gain insights from them. And it's cool that you ended up actually starting to do that so early in your, even in your PhD work. Okay. So then this was cool. The Wiseman Institute in Israel. So this is where you did your postdoc. Okay. What were you doing there? And what is that institute about? All right. So, Weitzman is, I think, it's a smallish research institute. There's maybe a thousand people or so. It's about 20 minutes outside Tel Aviv. And it is, in my opinion, it's one of the most creative places for science in the world. So it's a little bit outside of mainstream, right? It's a long flight to get there. It's not in Europe. It's not in the US. It's a little bit isolated. And it gives it, I think, a sense of freedom. So science there is very good, very high level. But also they like to do things a little bit off the beaten path, scientifically. And so there's a lot of really fantastic, really creative scientists there. There isn't the sort of money to do, let's sequence 100,000 people or this sort of science that is interesting, but is just driven by infinite sums of money that there are at some places in the US, for example. And so they asked to say, okay, with plenty of resources, but not those kinds of resources, what is the coolest stuff that we can do? And so I was in the computer science department, but I was doing biology. And we were working on the group I was working in of Professor Aaron Segal was working on gene expression. So how does the cell control how many molecules of mRNA, how many molecules of protein are produced? And how is that information encoded in the genome of that cell? So Eric Lander, who led the human genome sequencing, has this great quote about the human genome, bought the book, difficult to read. And in Aaron's lab, we were trying to understand how does the cell read the genome? And so in particular, I was working on cell to cell variation. So if we measure a bulk population of a million cells, we can get some average amount of protein per cell that's produced. But if we look at any one cell, one single cell might have one tenth of that, or 10 times that amount. And so is this cell to cell variation also encoded in the DNA? What are the sequence features in the DNA that affect the amount of cell to cell variation? Okay, so then the cell has a constant query process that it's running to its mechanistic parts. And it's always knowing what amount of these different parts of the cell that I have, how many mitochondria I have, or how much ribosomes I have, all the proteins I have. And then so the cell kind of has like a ledger in a sense. And it's constantly updating its like ledger. And then there's moments where the amount of proteins in a large amount of cells, the average will be pretty just like in the human race, the average will be a specific given average. But when you look at one cell, it could be one tenth of that average in terms of the amount of proteins it has versus 10 times the amount of proteins in the individual ones. So just like with maybe overall on the planet, we could have maybe a specific amount of wealth on the planet in total in the sum. But then one person could have like 3% of the total wealth. And another person could have 0.00001% of the total wealth. So yeah, okay. And wealth is a nice example. So we don't look at the mean, the average when we think about wealth, because it's so influenced by the top point or 1% of the population. So that really increases the value of mean income, for example. We look at the median when we think about wealth of countries, wealth of people. And similarly in cells, the average can sometimes not even exist. So there are cases where you have a population of cells that have lots of some protein, a population of cells that have very little amount of this protein. You take the average, that's in the middle, but there are no cells that are in the middle. They only are all on or all off. And so sometimes the average is completely misleading value. So either they'll have a lot of the proteins or a little bit of the proteins, but there's not really a lot in the middle. There's cases like this that exists. And of course, if you were to take a population level average, you would get a value that's in the middle, even though there are no cells like that. Okay, cool. Okay, and this was the main project when you were... Yeah, this was the main one. Do you think that that ledger analogy is a decent way to look at it, like the cell keeps a ledger of its parts? I think for some proteins, yes. It's going to be a dynamic ledger. The cell is basically going to be constantly checking, do I have enough of this? Do I have too much of this? And then modifying gene expression, modifying other things if something becomes too much or too little. For other things, I think the cell doesn't care so much. What other things do you think it doesn't care so much? There's quite a lot of proteins in the cell. So different proteins, one thing we found in Israel is that different proteins have very different amounts of variability. Some proteins have lots of variability. Other proteins have very little variability. And this amount of variability is encoded in DNA. So evolution has selected to say, okay, this gene, I really care how much there's going to be. And I'm going to kind of keep track of how much there is and very tightly regulate it. And when I'm making it, I'm going to make always exactly the same amount. And other proteins, whether you have 10 times too much or 10 times too little, the cell doesn't really seem to care so much, at least in the experiments that we can do in lab. And maybe that there's some environmental condition where the cell really does care, or our experiments are not sensitive enough to pick up the cells caring. But certainly this cell doesn't care that much. So it could even be that the cell has a ledger of its own internal mechanistic parts, but also its environment that it's in. So it's constantly dynamically adjusting the ledger and then expressing genetically to dynamically adjust. Interesting. This cell definitely cares about the environment. Whoa, the ledger internally and the ledger externally, kind of like we have one too, like we're like, okay, my ledger for my stomach is I haven't eaten yet today. But my ledger for my outside world is like, it's whatever, it's like 2 30 p.m. right now. And it's little Sony out, you know, so like that's not raining, you know, which is so these are like different environmental ledger. So I wouldn't need to bring an umbrella today, you know, so yeah, yeah, yeah. Interesting. Yeah, there was, there was a quote from from when we were hanging out yesterday where I said something along the lines of like, when we were hanging out, it was like, like Lucas goes, Alan, this is how biology works. And then I go, well, Lucas, is it possible that civilization works in a similar way? And, and so it was great when you were like, yeah, maybe or maybe not like that quite like that. And so there was just this constant process of trying to find relatable analogies or metaphors or storytelling methodologies that could then get other people around the world to better understand biology. So important to be a good communicator. Okay, so then how about post Israel? How did you accept this position in Barcelona to be a PI and a professor? That was for five years. Yeah. Yeah. So I went to Barcelona. I never been to Barcelona before, but I liked I lived in Tel Aviv during my postdoc. I like this kind of Mediterranean lifestyle. Tel Aviv is a fun city. And I went to Barcelona, my office looked out on to the Mediterranean. I went swimming during a lunch break during my interview. And then I went back to the city for a week or so to check it out. And I said, I could live here and the science was very good also, of course. But the city is fantastic. Yeah, I love how your pitching points are like the Mediterranean, the city, the life, the science. I love that. Okay. And then, so then what did you end up, you know, signing on to investigate? All right. So when we in my postdoc, we worked on self-discipline and gene expression. And at the end of my postdoc, I guess we were still doing experiments on this, but I was we're trying to think me when I say we me and a friend of mine who's now a professor at Yale, David Van Dyke, he was in the same lab as me, we worked together. And we're trying to think, okay, like, what is more exciting than gene expression? What is where's their self self variability? Where's it interesting? But where's their field that people don't really know so much about? Because people know a lot about about gene expressions. We worked on for a long time. And so we thought, well, what about growth and proliferation? So cell division, division after division after division. So not so much the length of time it takes one cell to divide, but the rate of growth of a whole population of cells. And if we look at single cells, we see that some of them are proliferating very rapidly. And other cells in the same population that have the same DNA sequence and are in the same environment are right next to each other are growing much more slowly. And so he said, okay, well, there's lots of cell to cell variability there and noise there. Let's try and figure out what's happening here. And we didn't know anything. So the first thing we did in Barcelona was we developed some techniques to measure cell to cell variability and proliferation and to understand how this relates to gene expression. So we developed some techniques to get the slow proliferating cells from a population and also the fast proliferating cells from a population. And then we could do RNA sequencing on these cells and other sorts of experiments, because we were able to grab the slow and fast proliferating cells and to physically separate them from this single population. And so this led us kind of better and just generally understand what's going on in the system. And then I spent most of the rest of the time in Barcelona trying to understand where this cell to cell variability in growth comes from. And also what are the consequences for the cells? Well, okay, in proliferation, you also see great variability. Some cells are proliferating quite healthily and rapidly and others are much more slowly. And so then it's okay. Let's take a single cell from both of those, a faster growing and slower growing, and then sequence the RNA. And when you found, when you did the RNA sequencing, what were the findings? So we found a couple of things that were sort of expected and some unexpected things, which I think is always good. We know something about biology and so you should always be able to find something that you look back on and say, oh, this makes sense. So we found that fast growing cells need to make more proteins, which totally makes sense, right? If you're growing more quickly, you need to be producing more cell, producing more biomass. And so fast growing cells have more genes for producing more biomass for producing more cells. Slow growing cells seem to be very stressed. They were expressing more genes, different types of genes, just kind of expressing more stuff from their genome. And so it seemed to be sort of maybe diversifying themselves a little bit. And at that point, we weren't really sure whether it was every single cell becoming diverse or whether our slow cells are a population of individual cells. And we're still not really sure. So you've got two possibilities. One possibility is that every cell is diversified as much as it possibly can be. And the other possibility is that you've got this population of slow cells that look diverse from a population level. But at the individual level, you would see that one cell is doing task A, another cell doing task B, a third cell doing task C, and you're kind of mixing them all together. And it looks diverse from there. But certainly that slow growing population is much more diverse and much more stressed, but also much more stress resistant than the fast growing population. Okay. So faster growing is more stress resistant, less stress resistant. And slower growing is more stress resistant. Okay, so you're more vulnerable if you're growing faster. Yes, you're much more vulnerable if you're growing faster. Okay. And then within the faster growing, you saw that with RNA sequencing that there was a higher production of proteins. Yeah. And that was a pretty given like, okay, that makes sense. Yeah. And then you also saw that, in specifically the faster growing or maybe in, maybe it was just in the faster growing, but that if you saw the whole proliferation, the whole culture, that there was just the organism was going in a specific direction. But then when you went and took individual cells out, you could maybe see that there was like a type A or a type B execution process it was focused on. Yeah. And especially for the slow growing population. Oh, especially in the slow growing population. We think that in the fast population, there's less diversity. The cells are probably just kind of maximizing their growth rate. There's one major way to do that. And all the cells are growing as fast as they possibly can to do this. And in the slow growing population, we see that there's more different types of cells. There are kind of different types of slow growing cells. And some of these are resistant to heat, for example, but where some of them are resistant to drugs, to antifungal drugs, but they're sensitive to heat. And so maybe that so there's this idea of bad hedging that there's that the population as a whole wants to grow very fast. It does this by having individual cells that are growing very fast. But if some stress comes along, it will kill the fast growing cells because they're very sensitive to to stress. They're very easy to kill. And then the slow growing cells, you can imagine two strategies. One would be to have some slow growing cells that are resistant to everything that could come along. But maybe this is difficult or impossible, the by the way the cells are wired. And so what we see is that there's different types of slow cells and some of them will be resistant to some things and others resistant to other things. And the thought is that these if something comes along and it kills all of the non resistant cells, all the fast cells, then the slow cells can repopulate that population. Whereas if you had no slow cells, the population as a whole would be able to grow a little tiny bit faster in the absence of any bad things happening in the absence of stress. But if some stress coming along, it would kill off the whole population would kill off everyone. And then you have zero cells and your evolutionarily dead end. This one was really interesting when we were diving into it yesterday, this idea of bet hedging. And this is quite a bit of what you're doing now at the carry lab. Analysis of this, proliferation and bet hedging and single cell RNA sequencing, stuff like that. And so you were doing quite a bit of this in Barcelona then. Okay, cool, cool. All right. So in the slower growing population, let's actually go to the faster growing population is more vulnerable to death from some sort of an environmental stimuli, like maybe a drug like an antibiotic. Meanwhile, that's trying to grow in our, the bacteria is trying to grow in our body. And so then it's more vulnerable because it's proliferating faster and faster. And it may not have the hedged bets placed on some of its even a very small population within it that is able to be antibiotic resistant. Whereas in maybe the small population, there has been a bet that's been, you're doing some bet hedging in the smaller population where there are at least a percentage of the cells that are expressing themselves in a way that is not quick proliferation, but is at least saying, we're going to be drug resistant so that we can make sure that we evolutionary stay alive longer. Gosh, that part is so nuts. Okay. And then now let's go back to that first part. So in the small population, you see more like type A, type B, type C, type D, like, and maybe in a sense we could compare this to like how someone is right now like, like literally bringing watermelons onto campus at PKU and then like, and then they're like, you know, cutting the watermelons and giving it to the students that are paying for it and stuff. That's like type A and type B is like a PI. Okay. But like this is in the population of the planet, like humans. Okay. So then one note is growing watermelon and selling watermelon to students. One note is doing some cutting edge biological research. So like different nodes, cell types are expressing themselves in different ways. There's like thousands of PI's across the planet. There's like tens of thousands or hundreds of thousands of farmers around the planet. So there's maybe a little bit more of the type of cell that is a farmer than there is a PI cell type. I should think more about this a little bit. These like sociological comparisons are so interesting. I can, we talked a little bit about this comparison with like investment portfolio. This is a good one. Yeah. The hedging. This is for hedging. Yeah. That you, there's the simplest thing would be to invest in, you know, index fund tracking S&P or something like this. And this will do reasonably well. But in economic downturn, you're screwed. And so maybe you want to put aside a little bit of money into something that's counter cyclical that will, you know, that won't, well, you won't lose all your funding if the economy crashes. And there are many different things that won't lose all their money if the economy crashes. So you could put a little bit into shorting the housing market and a little bit into some bonds in Japan or something like this that's relatively stable. And so none of these will grow as fast, but they won't lose all their value in the case of a disaster. Yeah. This diversified portfolio analysis is really interesting. Everyone in their mother is pouring money into real estate and that's like the fast proliferating. And then all these other types are then maybe in the slower proliferation are bonds or the different types of slower growth, diversified aspects of the portfolio so that if there is a big economic downturn for real estate that you don't just go to zero. So you have bet hedging being done. Even in cells, in cell populations, there's bet hedging being done. So the other alternative and there's not really good evidence for one or the other is that cells simply can't all grow this fast and they're growing as fast as they possibly can and they're growing so fast that they have some chance of tripping and falling. Like if you were running hurdles as fast as you could, there would be some threshold where you might gain a little bit of speed but you increase the chance of you tripping over hurdle. Like increase the chance of a poor mutation maybe? Either a mutation or something going wrong biochemically like building up too much of something because some components had cell that becomes toxic because you don't have the ability to get rid of it fast enough or we're not really sure. Interesting. But this is, we also have and that's called autophagy when the cell gets rid of the component that is. So you could grow so fast like a hurdle that's running that they trip over the hurdle like a cell would be proliferating so fast that some biochemistry goes wrong and it's an autophagy process ends up failing and then there's a uh-oh moment. And now we're going slow. And so this is sort of the evolutionary neutral point of view. In this point of view, this slow growing state has not evolved to be such, it just is a consequence of the cell trying to grow as fast as it possibly can. And it could be both of the things that you described. Yeah, okay. Interesting. Okay, and those two things again are just like if you're growing too fast and I like the hurtling analogy so it's like you could be growing too fast and you could trip so you want to prevent that that could be one of the these quantitative mechanistic biological components of cell proliferation is hedging against that and then and then the other one is that you want to diversify this portfolio for all different types of an environmental potential malevolence and yeah interesting. And you can imagine a situation like where it only matters if you get gold, silver or bronze and so you want to run as fast as you can because coming in fourth versus tripping and coming in last are the same. Wow. Okay. Wow. Yeah, if that's the case then that adds a whole new layer of complexity if only first and third place matter and fourth is the same as tripping and falling and placing zero at their last because then you would want to put as much of your logic towards quantitatively proliferating putting the mechanisms for proliferation prioritize that put the foot on as much of the gas as possible. Yeah. Because we want first, second or third. Yeah. It's kind of in a sense it's maybe it's somewhat similar and someone different in our like in our sociological hierarchies because we in the sense of in the entire eight billion hierarchy let's say of wealth that if you come in fourth you're still doing ridiculously well. So you might want to just put your foot on the gas but yeah play some you might hedge some bets and stuff but if you're maybe if you're only in a group of like like 10 PIs in the life sciences department at a university then you know placing fourth isn't so much on the grid it's it's it it lands so like placing fourth is like okay well the the fourth best PI out of 10 at like the life sciences you know department so versus being fourth in like this whole the wealth hierarchy interesting there's always trying to find these sociological analogies okay and then I'm wondering what thinking about this if it's somewhat similar to the way that the tech industry is working now where you burn through insane amounts of cash to become first and in your market that's right and it doesn't matter if you're there's no difference between zeroth place and you know the fourth biggest you know e-commerce platform in the market nobody's buying from them everyone's buying from olibaba and amazon no one even knows who is the second most like ebay you know you definitely don't know the third or fourth because you yeah yeah yeah interesting and you don't know the third uh rideshare company either you know didi in china and you know uber and lyft in the us yeah and other places in the world but that's it yeah uh so we only remember one two and maybe three maybe so fourth is almost yeah the same then in a sense as wow so that's why it's uh pour as many resources onto the fire of the tech companies growth as possible yeah what an interesting analogy so we have some uh some of our economical uh processes that some of our quantitative mechanistic biological processes related to economical growth are actually somewhat similar to the way that sells proliferate i think that's fair to say you get the same sort of exponential benefits so if uh pop if two populations are growing exponentially so the the bigger they are the faster they increase um then very small difference in initial growth rates gives you very large constant you know advantage towards the end and this is why tech companies blow through all this money to become first in their market is because of the thought is that once they have kind of this dominant dominant place this gives them advantages that are not available to second third fourth place okay so when you were teaching me about this you were saying it from a perspective of the the predictability of evolution and at least to have a better version of understanding and making that more quantitative so what exactly can we say about the like predicting the the fates of single cells and of organisms and like why are you so passionate about understanding the codified processes of that okay um so we i use predict because if i say if i told you i understood something you would basically just have to take my word for it but if i tell you i can predict something i can write down my prediction and you can test me on that and you can see okay can you predict this uh and so but i'll just want to understand things they want to understand how cells work they want to understand why we look and behave the way they do um and we want to understand things too but we want to do so in in a quantitative manner that i can write down my predictions for the future and see if they come true or not and so i i want to if i have a population of a million cells and i give them some drug that kills 99 percent of them i want to be able to tell you before giving them the drug which cells will survive in which cells will die if if we can do that then we can really understand the system and there might be some complete randomness in this factor so i might be able to tell you this cell has a 90 chance of survival and this cell has a 10 chance of survival that might be the best i can do if i knew all information um but we're now very far from being able to do that um and so this is what we're really interested in doing can we can we predict which cells will survive um a drug treatment can we predict which cancer cells will survive chemotherapy this sort of thing um and then can we predict uh we know that in in hospital settings for example cells tend to evolve antibiotic resistance very quickly also inside of patients cells tend to evolve antibiotic resistance and especially antifungal resistance pathogenic fungi evolve antifungal resistance cancer cells evolve chemotherapy resistance very reproducibly um but sometimes they don't and in the lab we can do experiments and we see that some types of cells are able to evolve this drug resistance very rapidly and very reproducibly and other cells just don't or do so much more slowly and so this cell tells us that evolution should be predictable that we should be able to predict kind of which cells are more easy to evolve than other cells and that this information is somehow encoded in the cell and probably encoded in the DNA of the cell um I see I just want to say I see especially with the um um the the oncological the benefits here of doing your experiments in the lab is that so many of us have parents or grandparents that are literally going through processes of cancer and their health deteriorating from that and if we can do this science faster we can understand and more effectively build better tools for it we can understand then there are specific genetics encoded within cancerous cells that enable them to fight better our treatments than other ones that just seem to die off from our from our treatments so we want to now what would we do that would be optimal would we want to uh figure out how to compete against the the uh the the the cancerous cells that have a stronger robustness um towards a fighting against our treatments so it's a good question so one one thing we've thought about doing with the bacteria is are there non-toxic things you could give the cells that would interfere with their ability to evolve or would interfere with their ability to be in this resistant state um so these are kind of non-obvious treatments that you wouldn't normally you think okay well maybe if I give two different drugs then that'll kill the cells better indeed that usually does work but cells can evolve resistance to this also um but um but if we could put this cell in the state where it was less evolvable or or we could put the population of cells into a state where there are fewer of these resistant cells and that might even be making by making them grow faster right because as I said cells that are growing fast are more sensitive to stresses into drugs and so if we could get rid of the non-growing tumor cells by making more by making them grow and of course this would be very risky but if we could make all of the cancer cells grow then all the cancer cells could be killed by chemotherapy a one two punch of sorts where you give them this little uh artificial boost and you make them think aha we got it and then you come in with the second hit and you're like aha you thought because you were growing proliferating faster you became more sensitive vulnerable and now we really wiped you out exactly so you move you move them from slow proliferation state to a fast proliferation state and then you wipe them out you trick them that's one option okay and the other option for this um this evolution of drug resistance um here you give the cells the drug and it kills most of the cells but some of the cells have some mutations and they're able to gain more mutations and evolve drug resistance um if we you can only give a and right now we don't know you know if we have a choice of a couple of different treatments for a tumor for an infection we give one that the cells are sensitive to and end of story and we hope that it kills all the cells if we could predict hey this infection is going to evolve resistance to drug a but it's not going to evolve resistance to drug b if we could predict evolution and if these evolutionary paths were different enough then we could give treatment based on the predicted future of the infection or of the cancer instead of just based on the current state as in there is a you could map the cancer's future trajectory and then give treatment based on where it's going yeah instead of its current state so if you think about climate change we're not worried about climate change because of this one degree temperature rise we're worried because of the future predictions of climate change yes which we're quite confident in and we're less interested in telling people to burn less fossil fuels we're more interested in being like let's obsolete all of the old technology with this new more efficient more effective cleaner more sustainable technology but all this is based on predictions of the future and not so much what is the current state of the earth yeah so we as future prediction machines are thinking about our grandkids and the planet that they inherit and that's why we're caring so much and there's actually an indigenous principle called the seventh generation principle so literally them thinking prior to acting anything okay how is this going to affect seven generations out prediction is hard that far in advance but i think it's also you're right it is very hard that far in advance but it also can teach us something with uh it's not just i'm going to fish here or i'm going to harvest this or i'm going to burn this fossil fuel because it is not that's a very immediate we it's like overly reliant on immediacy and we've developed a cortex that enables us to do something like think about a seventh generation and be like if i fish here what happens next generation if i harvest this what happens next generation how do i replenish the soil or replenish the fishery or produce a more sustainable way of producing energy for my town yeah interesting so then the ledger the dynamic ledger of a cancer cell is also understanding its internal processes understanding its exterior environments and also trying to predict some sort of future where it can basically take over that organism potentially this is hard stuff yeah so there's a big open question in in biology of like can cells evolve evolvability can you can a cell have selected for the ability to change in the future um i'm tempted to say no can cells evolve evolvability yeah because right so for a cancer cell its current state is okay i need to grow as fast as i can or for a population for any organism it wants to grow as much as it can um maybe it we're kind of complicated enough to realize that well if i grow as fast as i possibly can then my grandchildren are screwed but microbes are not certainly um or we would at least think that maybe maybe they are and that's why they do this bad hedging who knows um but now you're asking can they can a cell have evolved so that it can more easily change its genome in the future in response to something or to allow it to grow in some new environment um and that i'm hesitant to say that the organisms can do that yeah interesting we can certainly design them to be like that yeah this is this like designer organism idea could we make the proliferation populations more prediction machines about the future more evolvable to have them develop evolvability more maybe that and maybe that has to do with the running so many different engineering so many different simulations of the proliferation for them to build more resistance to all of the different environmental stimuli we throw at them and then they encode those better and better yeah huh wow yeah this is really biotechnology which uh and as we become designers of the biology is a it's a whole new opening up of potentials what about you gave us the uh oncological assistance this work in the lab can help with that where else do you think the work in the lab can help with um so with infectious diseases which i would sort of put in the same category as cancer in the sense that we do some treatments you have fungi that infect rice and wheat and things like this and you give them at you spray antifungals onto the fields and they just become resistant and same thing viruses that are killing some crops um so all these sort of fall into this general class of single cells that are growing kind of as fast as they can and they're trying to grow and we're trying to kill them well and then what would a farmer do for a crop rice wheat that is currently has the solution of just spraying antifungal but then obviously they develop the the counter the properties that counter that antifungal spray so how would your lab produce the solution so here i think that if we could predict which strains which pathogenic strange which can um will evolve resistance to which drugs we could say okay spray fungicide a but not fungicide b because this particular genotype will evolve resistance to a more much more easily than resistance to b even though the initial killing will be the same the evolutionary kind of paths will be different this i think would be really cool to be able to do and is there an even more maybe golden key solution to it where there could be a process of engineering the genomes of the rice or the wheat to be just super duper combative to all of the different fungicides yeah well this is this is sort of how the bt from unsanto works or bt corn bt soy if you have crops that express their own toxins these end up being much more effective if i spray fungicide on the field i'm spraying the workers on that field i'm spraying the soil i'm spraying the air and i'm spraying the rivers around it but if the crops produce that toxin it's much more targeted to the fungi that are growing on the toxin um and you can have the crops produce very toxic things so plants are nice and that they're not human right so things that are very toxic to plants and very toxic to um to plant pathogens can be completely nontoxic to mammals and we would just need to have some sort of like longitudinal tests of like is it really nontoxic yeah yeah so if we engineer the rice or engineer the corn or wheat that to compete combat against the um the infectious toxins more than would to compete thinking could and also to combat against pesticides and against uh drought and all these types of things or maybe just yeah make them grow in different conditions so all this type of stuff the question would be then if i eat that for 20 years am i still good and uh we would just need to do the longitudinal tests to do so okay okay and then um let's talk about how like overall this field as exponential technology continues to increase and our computational capacity continues to increase our microscopy increases all these types of things are kind of like coming together sequencing abilities all these types of things coming together do you see us unlocking some sort of basic source code of cells and biology in general um i think we're getting much better so we have the source code um right we have the genomes let's see that's the easy part um it's reading the source code you know the source code is all assembly from uh i don't know and it'll be a mainframe or something like that it's very difficult to read um it's source code that has been worked on by you know programmers over the past you know million hundreds of millions of years uh so it's a complete mess there's no comments anywhere there's no comments hello yeah um so we're getting much better reading it that's for sure there's no repositories of like the older bills exactly yeah we can go back well we can go back 10 000 years right we can get resurrected into dna from about 10 000 years ago um but that doesn't do us nearly good enough right we yeah okay so source code is there with the dna yeah okay and then it's specific short sequences of dna create amino acids with amino acids which then create proteins and so that part we're still figuring out as well which sequences create which amino acids which create which proteins that we've got it's more all of it pretty much yeah it's wow um and how many total proteins are made by like 25 000 25 000 depends on what you count but 25 000 and we have that all down yeah the question is how does the cell control and how's our body control these so how does the cell tell cell tell you cells to become eye cells and to become no cells and this sort of thing and how does the cell know how much you know how much melatonin to produce to create a pigment or how much hair to produce or whatever so these kind of quantitative things of how much of this gene how should i make how much of this protein should i make that we're still working on yeah melanin for yeah skin tone and melatonin for sleeping yes and then digestion too like if you you know how do you know when the cell regulates to produce digestion abilities when you eat food and yeah that type of stuff it's like yeah the pigment of the color your eyes blah blah blah there's so many things why is there no hair on your forehead yeah so all this is encoded in dna and this we have no idea so somewhere in that dna is the statement no hair on foreheads yes for most people yes sometimes that goes wrong and you do get hair on foreheads yeah um but we have no idea where that information is in the genome okay source code how that produces amino acids and proteins got it how the cell decides when to call for that building block of amino acid to gene expression amino acid to protein is like figuring out yeah okay okay okay and then do you then think that it's possible for us to figure out how the cell keeps that dynamic ledger of its internal state of its external state of its like future possibilities i do um i think that we can do this in simple cell systems we can do this in engineered systems that we've so if we take a system and we design it and then we make changes we're pretty good now at predicting the outcomes of those changes if we take one thing that we do in my group is we will design for example a million different variants of a single protein um and then use all that data to to understand what is the effect and to predict what is the effect of individual mutations and of groups of mutations on function so once we have a million data points if i make a new mutation the million first data point i can use those previous millions predict pretty well um what the outcome of that new mutation is going to be even though i haven't measured it yet um and i think eventually we'll get there with humans i think it will take much longer humans are complicated um but right the ultimate goal is if i i should be able to take a new patient and say okay this patient has the following thousand mutations in them relative to the population as a whole what is the impact of each of these mutations on their health um and we're moving there slowly i'm confident that we'll get there yeah actually yeah okay given the rapid pace of exponential technology and democratization of the tools and our ability to probe with science that we think that we could basically have a better understanding of the ledger of biology and all of the quantitative mechanisms of biology to the point where it could be that um when you come in for you don't even need to come in you just have a constant stream of your biometrics being processed and that you can uh and that's literally being pattern recognition on that compared to our vast library of other patterns that have been recognized and it's just like okay lucas you're we're starting to detect a specific pattern with your biometrics where um it's going to be important for you to get eight hours of sleep tonight or to yeah go exercise or to eat a healthier meal or whatever to make sure you don't have some sort of dysfunctional um pathology it develops or why not okay okay yes whoo okay it also has been hypothesized that this is like a 100 trillion dollar industry over the next like hundred years that biotech and all of its components because you were giving the examples in health care but then there's examples in agriculture and the examples in energy and there's just like that's already extremely insanely massive fields and given the amounts of big data they're being crunched by computation all this stuff like there can be some serious big booming companies and like get creating in that space that's a big space get researching in that space forming companies yeah yeah okay a couple quick questions on the way out um how can we inspire more people around our world to work together I think making it easier to work together um scientists certainly want to collaborate um and also I think many scientists enjoy traveling to different places um and sort of using collaborations excuse to go on vacation and check out new cultures new places um but I I think regulations tend to make it hard right um you weren't able to you know you use the awesome payment systems that we have access to in china because you don't have a national ID number and you don't have to bank out um and likewise there are issues and similar issues when people visit the us yeah um so this I think makes it more difficult um yeah I'd say just increasing mobility yeah doesn't have to use appropriate mobility but just even temporary temporary mobility and making it easier for phd students who go to the us to go home to visit their families um yeah yeah and making it easier for phd students who want to come to china it's really difficult for foreign phd students to come to china interesting degrees of freedom degrees of mobility even if temporary just making it so that you can vet people and through a trusted process more effectively and get increased global collaboration and creativity through that going into this exponential technology age you have a young son as well what do you think young kids should do should learn that's a skill that's going to just be enabling them to be equipped really well I think just general problem solving as kind of in an abstract sense like if you if you know how to solve abstract problems that have undefined solutions where it's not clear how to solve this then you can learn programming if you need to you can learn to interface with politicians if you need to you can kind of learn all the specific things later but if all you've done throughout your life is solve problems that have defined steps to ending up at a predetermined solution where you already maybe know the answer um then you're in trouble once you get to the first to ask where there is no clear answer or it's not clear how to get there and I see this as a problem in in my education and in the education system here I remember in our labs and physics and undergraduate we're solving all these problems and doing these experiments where we know the answer already we're measuring g the gravitational constant but we already know the answer we know how the experiment worked out and I don't think that these are inspiring and I also don't think that they teach you to do proper problem solving yeah yeah here the 17 sustainable development goals figure out how to solve exactly yes stuff like that I love that yeah okay how about what is the ultimate nature of this reality why are we here what is the meaning of this the teleology of the species I don't think there's any inherent meaning for being here I don't think we're here for some purpose um yeah it's interesting seeing how many people view this question from like a very like materialistic perspective versus like a lot of other people are like hyper spiritual and like magical about it and it's like interesting to find out where those two perspectives okay yeah yeah uh consciousness he thinks what do you think is a biological phenomenon or what are your thoughts about it I'm purely materialistic about that too what about I just think we've figured out some things about it's hard to figure out it's hard to study we don't have a good model system for studying consciousness we can't do experiments on people other organisms don't have the same consciousness as we do if you know like the Thomas Nagel essay what does it like to be a bat um like I have no idea what it's like to be a three year old toddler even though I live with one like even though you lived as one yeah so without a good system for really understand and and in spite of that fact we've we know more about consciousness and how kind of emotional states are encoded in in neurotransmitter concentrations and like this and we did a hundred years ago and I think that this will continue to increase what do you think is the role of love I think you got to love what you're doing you got to love who you're doing it with yeah um yeah yeah do you think this is a simulation no um but I don't care so much um so long as it's uh so long as the programmer is not changing the parameters of the simulation while it's running because that would make doing science very difficult yeah what do you think is the most beautiful thing in the world um when you've found something that nobody else in the world knows oh that's that's exciting and you very quickly forget that nobody else in the world knows this and I like even before you get to publish a paper on or something I find that I do it and certainly my students do it we forget how cool this is because we've already known it for six months or a year or whatever but nobody else in the world knows this and this is completely new yeah oh I love that answer it's such a good one I believe that's the first time we've had that one on the show yeah yeah yeah that's a huge one discover what no one else knows and then bring that forth as a gift into the world I love that okay wow thank you so much for coming on our show thank you thank you it's been an honor and a pleasure it's great hanging out with you yesterday and today this is a lot of fun thank you good I'm really happy to hear that it's been so enlightening thanks everyone for tuning in we greatly appreciate it we'd love to hear your thoughts in the comments below on the episode let us know what you're thinking have more conversations to your friends family co-workers people online on social media about quantitative mechanistic prediction in biology have more conversation about that and how it can augment our health augment our society at large and go and build the new tools that help us poke and probe at that and check out the links in the bio below to the carry lab also check out Lucas's twitter profile as well check those links out support the artist the entrepreneurs the leaders in your communities and around the world that you believe in support simulation our links are below so you can tell you and cool things like coming to china for interviews and go and build the future everyone manifest your dreams into the world thank you for tuning in and we'll see you soon peace that's a wrap brother good job