Phase Space Invaders (ψ)
With the convergence of data, computing power, and new methods, computational biology is at its most exciting moment. At PSI, we're asking the leading researchers in the field to discover where we're headed for, and which exciting pathways will take us there. Whether you're just thinking of starting your research career or have been computing stuff for decades, come and join the conversation!
Phase Space Invaders (ψ)
Episode 12 - Vlad Cojocaru: Simulating transcriptional regulation, returning to our home countries, and the scales of the genome
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In the twelfth episode, Vlad Cojocaru and I discuss how the simulation field can integrate further layers of complexity in modeling transcription control and genome organization in humans. Vlad shares his insights on the computational biology of transcription factors, including DNA-mediated allostery and pioneer transcription factors, which are crucial for cellular reprogramming. Our discussion then shifts to the underexplored topic of researchers' returning to our home countries as group leaders. We examine the hopes, fears, and conditions for success in making this transition, aiming to rewrite common narratives from our unique perspectives.
Welcome to the phase space invaders podcast. Where we explore the future of computational biology and bio physics by interviewing researchers working on exciting transformative ideas. Today, my guest is Vlad Cojocaru. a researcher working on the computational biology of transcription factors. affiliated with max Planck. in Muenster with Utrecht University. And with Babeș-Bolyai University in Cluj-Nap oca. To make a connection with a previous episode that is also original BioExcel ambassador for Romania. For a good while we have both be working on similar research questions. With that recent work, exploring the question of DNA mediated. allostery. Or pioneer transcription factors that are key for cellular reprogramming. So we start the conversation from the purely scientific side, talking about how they feel might go about integrating farther layers of complexity in modeling, transcription control and genome organization in humans. It's just in recent years, the experiments start to provide high resolution data. About this incredibly complex in a hierarchical interplay. And a lot of work will be needed to turn it into molecule. Develop insights that will one day translate into therapeutic modalities. We'll move on to discuss, I think a severely under explored question of scientific returns, our home countries. As two researchers who sooner or later plan to take this step, we're dissecting, the hopes and fears that come with it. Trying to outline some conditions for success and the rewrite, the common narratives, at least from our very particular points of view. Okay enough of this introduction, let's go.
milosz-wieczor_1_06-10-2024_130413Great. Vlad Cojocaru. Welcome to the podcast.
vlad_1_06-10-2024_130413Thanks a lot. That's a great initiative. Congratulations for it.
milosz-wieczor_1_06-10-2024_130413Thanks a lot. so Vlad I know we followed each other's work for a good decade now overlapping in the field of protein, DNA interaction and recognition. And when I reflect upon it and thinking about your Morrison paper, I'm kind of impressed by the increasing scale and a degree of biological context in the ways we are beginning to address scientific problems. Where do you think we are on this journey? And, you know, going back to your papers again, what might be the next challenges in transcription control you think we'll be able to tackle? I'm thinking, you know, epigenetics, the transcription machinery, full sized proteins, chromatin structure, RNA, and so forth.
vlad_1_06-10-2024_130413Uh, yeah. So, um, I think we are, approaching the moment where we can tackle, much larger systems using this kind of very detailed atomistic, simulations, right? Because that's what we are doing. We are doing computer simulations. I'm trying to understand the biology of it, and this comes back to the times when I was in Munster, and basically I got fascinated in this stem cell field, where you have cells that you can convert from one type to the other, right, you can, you can go from skin cells to pluripotent cells, where you can make like an embryo from, from a skin cell, right, if you think that's crazy. I got really fascinated into that, and there are these molecules which are called transcription factors that actually, are master regulators of this process. So you can just put three or four of them into the skin cell, for instance, and you can completely change the fate of that cell. And that was for me absolutely fascinated at that time. But I wanted to understand how these transcription factors work, right? So, so basically what we did, we, uh, I came from the computer simulations point of view, I said, well, you know, we can start, using the developments that we recently have in simulating macromolecules, try to tackle this question. How do these transcription factors are able to change the entire landscape of a cell, right? we started with, small things, transcription factors on DNA, that, that works relatively well, the current methods are Accurate but then we got the idea, okay, what if we go one scale up and now, you know, maybe you know that, uh, well, for sure you know that, in the, in the cell, the DNA is packed, right? So you have, uh, this massive two meters of DNA is packed in a very, very tiny nucleus. And then you have, this nucleosomes, which are kind of they are kind of structures, which, which pack like 100, about 147 of, of base pair of DNA. some histone proteins. So now when these transcription factors, if you remember them, when they bind to the DNA, they have to recognize the structures. So now we started going into that direction. And basically, at the moment we are simulating the binding of these transcription factors to these structures, which are nucleosomes, which now you get in the range of millions of atoms, basically in a single run. Which is great. And, we have a number of difficulties here. So, um, the way these transcription factors bind to nucleosomes, they are not exactly well defined, especially not on native nucleosomes. So that's a major challenge. That's why we have to go through the experiments. And we have to validate every single binding event, every single binding locations, and so on and so forth. So it's not an easy thing to do, but the question that we start to be able to answer is how the structure of the nucleosome, the structure of DNA wrapped around the histones, is changed when the transcription factors bind. So now this is the level where we came to, but. Now, the question is, these transcription factors don't change one nucleosome, they change an entire genome. So how does that happen? So now my dream would be that in the next, in the future, we'll be able to use even fully atomistic simulations where each atom is represented. actually tackle the question how, how these transcription factors change the arrangement of multiple nucleosomes and stuff like this. But we are I think we are still far from that point. At the moment, we mostly basically try to understand how each nucleosome is changed by the transcription factor. And of course, then you have the epigenetic landscape, right? Each nucleosome is modified in certain positions to actually, give this layer of epigenetic regulation, and that we can all include into the simulations. And that's the beauty of it, because in the simulations you can actually have things that are very hard to tackle experimentally. So these proteins that wrap the DNA, they have these very floppy ends, which are called the tails. And they are very, very tricky because they are very, very flexible. They move back and forth. those in experiments, if you look at all the structures of nucleosomes, only one single one of them has these tails. And in simulations, we can actually use that. We can have them in every single simulation we do. And, well, the accuracy is, an issue, of course, but we are at the moment quite accurate in describing these interactions. So I think we are at a level where we can start tackling basically larger arrangements of nucleosomes. also proteins like these transcription factors that interact with it.
milosz-wieczor_1_06-10-2024_130413Right. That's amazing. I think many people take this picture, especially in the case of transcription factors, take this picture from bacteria, right? Where all those things, all those complexities don't even exist. And there's just some sort of sigma factor binding to a promoter and then everything goes smoothly. And we've got this hugely complicated embryogenesis that relies on so many things, right? We have, for example, the, um, remodelers, right, that are machines that are working on chromatin and remodeling reorganizing nucleosomes with ATP derived energy. And then we have polymerases that also have their own pushes, pulls. So is there any hope for us even to, I'm just thinking about this, you know, multiscale approaches as well, right? At which level we'll have to model the biology with some sort of multiscale approach. methods, including grained or even, like, metascale, uh, methods. Mm hmm. Mm
vlad_1_06-10-2024_130413we will always need multiscaling because, at this point it's, it's, it's very tedious to simulate all these motions that are involved. So you can imagine if you talked about the, the remodeling, right, I mean, this involves practically, evicting the histones from the it's a round kind of shape, helix, if you wish, with the DNA wrapped have to actually remove the histones from it. then you have to slide the DNA and these are motions that are very, very hard to tackle from with the atomistic simulations. Now, for these type of motions, you do need coarse grained simulations. I think we already have quite a number of, uh, in fact, I would say that in the field of, genome organization, coarse grained methods still, more used and, they, they have upper hand on the atomistic ones, just because you can tackle much bigger problems, right? And much
milosz-wieczor_1_06-10-2024_130413hmm.
vlad_1_06-10-2024_130413types of motions. Having said that, of course, you lose the details of the atoms interacting, which are very important. but with coarse grain, you can, tackle this type of large scale motions and large scale systems, right? So. So I think will always need it. This is a genome organization coming from the pair of DNA the whole chromosome. a typical multiscale problem you will always need different types of models to actually. tackle different questions up to the polymeric physics, right? Which, which are used to tackle the organization of the chromosome fascinating. So the way that the genome is organized, it's absolutely fascinating. And, and, as you go up in scale, of course, you, at some point you will not be able to do it atomistically to have all the atoms there.
milosz-wieczor_1_06-10-2024_130413Yeah, absolutely. But on the other hand, as you mentioned, we have the histone coat, right? The histone, the very tiny modifications. Sometimes it's a single methylation that absolutely changes the whole landscape of macroscopic behavior. This is also Kind of working on the face separation level, right? So the physics of face separation is also getting into the field
vlad_1_06-10-2024_130413Yeah. So
milosz-wieczor_1_06-10-2024_130413quite prominently recently.
vlad_1_06-10-2024_130413Kind of novel concept, right? at this phase separation, how especially, interesting for this, I mentioned to you this, terminal unstructured, right? Uh, floppy tails of the, of the histone proteins wrap the DNA. And there is emerging evidence that basically this kind of help creating this kind of phase separation where, Basically creating regions in the face space where transcription actually is active. and this is one of the fascinating, uh, parts of, the newest developments. And the other part is, of course, genome organization and transcriptional control. It's, a field where AI is still at the beginning, right? There is so much data but I'm sure that in a few, in the next years, it will start picking up. If I went recently to a conference about the multi scale genome organization, and, and then you really see, if you compare it to the protein structure field, where AI is like, here, we don't have much AI. And that's the reason for that is that all the machine learning, deep learning algorithms, they need a lot of data and the data is still scarce. Right. So I'm looking forward also for this development because I'm really certain that a few years from now we will see a lot of AI prediction tools aiming to predict the structural dynamics of nucleosomes, arrangements of nucleosomes, and so on and so forth. So very, very exciting times from that view.
milosz-wieczor_1_06-10-2024_130413Absolutely. And then we also have the experimental techniques such as cryo EM tomography, right? I recently saw a paper, I don't know if you saw it too, where they were able to show the 3D organization of chromatin. assigning, you know, nucleosomes, chromatosomes, so distinguishing between the H1 bearing and the H1 free nucleosomes in the actual cell nucleus. So, it's pretty amazing that we're gonna have this data soon to, to train our models to reproduce that.
vlad_1_06-10-2024_130413That's absolutely fantastic. That's amazing data, basically, where you, where you really can look at, how a few number of nucleosomes basically are associated together. And then you can also look at, full fibers where you have Uh, genome organization has been right in the middle of a debate, this kind of, 30 nanometers fibers very packed. and now we know that that's not really the case, that even the closed chromatin is very dynamic and, and being able to visualize it literally with this kind of cryo E. T. techniques or electron tomography techniques, really gives us the data that we can then use in the simulations. We can use, we can use this data to build models, to refine dynamic models and so on. so I think this is also, um, fascinating now and I, I mean, all my career, I, I loved working with experimenters, right? So this has kind of it's a stamp on my career. I would say even from the beginning of, my PhD to today, we're basically really working close, close by, uh, with experimenters, even if they are like developmental biologists. And this has been always, always very nice. And I hope I can, continue doing that basically
milosz-wieczor_1_06-10-2024_130413Yeah, I think we should all stick to that. I mean, it's a very common theme in this, collaboration between physics and biology that, for example, physicists will come up with something like the zigzag model or the solenoid model where everything is nicely arranged, right, and symmetric. And so when they look, you look at the actual biological structure, and it's like a complete mess that's maybe 30 percent of one and 20 percent of the other, and everything in the middle is just completely, indescribable by simple human readable patterns. So we need to keep, both mentalities of like order and chaos in check in this way.
vlad_1_06-10-2024_130413made. we'll be able at some point to actually find new patterns that can describe this, uh, chaos as you name it, right,
milosz-wieczor_1_06-10-2024_130413Yeah. There's statistics for Everything. Of course. Absolutely. Then to the other topic. So you are now moving back to your home country of Romania. And we were discussing this question of returns academic careers, right? So how do you see the pros and cons, the ups and downs, you know, of going to, to have a career, in, let's say Western labs, and then bringing the expertise I mean, I'm planning to do the same. So I'm asking you as a fellow, person on this journey, um, How do you, how do you see this story unfold? Right.
vlad_1_06-10-2024_130413um, taking this opportunity to go to, to the Babeș-Bolyai university in Cluj-Napoca, it's with mixed feelings, but with a lot of hope. So I think If we are about to talk the positive stuff first, think, the developments in Romania, but not only in Romania, I'm talking more of the entire Eastern Bloc, if you, if you wish, are really interesting. And I think there is a lot of potential there. So, um, we all know the brain drain that happens in these countries, right? So, so people go abroad very fast, which is not bad. It's actually good. Yeah, I gain amazing experience by going abroad, but the thing what happens is people don't return. So that's at least with Romania, that has been the case and it's still the case, quite dramatic, basically, you have lots of, uh, Romanians leaving abroad and very few coming back. And I think you actually have a lot of potential. So provided you have certain conditions need to be met. one of them is of course, predictability of funding. In the case of Romania, also raising the national budget for research because it's like the smallest in the EU, of funding, trying to offer a little bit, better conditions for the start. with this, the countries like Romania have a very big advantage if they can offer you a, non temporary, a permanent contract, right? Because that's the major issue in the West, right? You have a lot of, uh, nice positions, but they are extremely competitive and they are also, I mean, then you end up spending a lot of time. A lot of years on this non permanent contracts, right? Two years there, two years there, three years there. So one of the major advantages, and what also basically, convinced me about the move was the advantage of having a permanent position. Uh, which at my level of career, I think it's, it's essential because you don't want to end up close to the pension going from one place to the other so that's a big advantage if, the predictability of funding, if the budget is a bit increased, and maybe the most important, if the evaluation is done strictly on scientific merit. Then I think you have lots of opportunities in countries like Romania. And, uh, that's why I like about the Babeș-Bolyai University because I think the mentality, if you wish, is right, uh, in that respect. And, also attracts me for this move is basically that coming from abroad with the experience I have which I might not be top five researchers in the world, but coming with that experience back, I think you can move forward the system there. You can, you can contribute to changing the way the research is done in the country. And I think that's a very interesting perspective. That would be like the positive aspects about it. Now, of course, the negative aspects are exactly the opposite what I mentioned. Funding is very bad at the moment in Romania. It's increasing. There is a lot of movement on that aspect. So there is now a big call for centers of excellence. there is movement, but still the level of the of the GDP that is allocated to research is very, very low predictability is not there. You don't know when the competitions are open. attractivity of the system, it's not there because salaries are, small in comparison to any. not talking about comparison of the absolute values, but even if you take into account the standard of living, that's also, lower if you take the prices into account. So, there are some, cons, but I am pretty excited, actually, about the next step. I think it's, This is going to be very difficult, I am confident that, I still have quite some enthusiasm left. So, after all these years, so, uh, I hope I can, I can bring some of this back. And I think if I find the right atmosphere around the right, mentality, because I think, I you can really move things forward. And of course, research is also, um, for many of my research, I need high performance computing. We don't have a center in Romania for that. so these are going to be like daily difficulties that, you have to, let's say, be smart enough to find alternatives, And stuff like this. So it's not as easy as it's here, where you have access to things relatively fast. but, if you are smart and you can actually, go over those, obstacles. So I think it will be fun
milosz-wieczor_1_06-10-2024_130413I think we have to kind of consciously start talking about on maybe dismantling or rebuilding those narratives, right around the brain drain and the returns, because I think our field especially has this unique. Possibility of having a wide network of collaborators all around the world and not necessarily being, you know, in the facilities. Like, uh, of course it's great to work with experimentalists who work next door, but there's no problem in principle in, in working remotely with collaborators all around the world. The only thing that we actually need to do is to go there, right? And to make those contacts and, uh, and to create a network. And then exactly expand this network because, as people from, you know, nations historically disconnected from global Western science, you only started to realize how big the network is when you, um, You go to the West and you discover that, you know, Oh, wow, these people have been working together for decades and, and now you're just joining this. And so how can we, I'm talking about, you know, all the people from, from Eastern Europe, from India, from Africa, rest of Asia and South America, how can we take this experience and make it kind of international and get people excited about bringing
vlad_1_06-10-2024_130413working
milosz-wieczor_1_06-10-2024_130413networks? Thanks.
vlad_1_06-10-2024_130413been facing with recent interaction with Romania in basically this, this working together. It is so kind of in us, here it's, not always happening there. And this is something that needs to change because, the progress of science, requires that you cannot do things if you're just hiding yourself. in your own, uh, bubble and do your stuff. That's not possible. I mean, the way you progress science is by basically talking to people and, working together. And that's something that I really, want to, to put my stamp on because I think is something that needs to change, uh, dramatically in, in the country, at least in Romania, and I don't know about the other countries, in the Eastern Bloc and that details but in Romania, this, this concept of working together, meeting and chatting about science, um,
milosz-wieczor_1_06-10-2024_130413Right.
vlad_1_06-10-2024_130413from what I experienced so far. And I also have a a trouble with the discussion over all the discussion in the Romanian society, for instance, about the, the, the brain drain, right? So you have the brain drain. It's always like, Oh, how can we make? People stop leaving the country and I always argue this is not a problem. This is not the problem. You actually want young people to leave the country. Maybe, maybe still study there, but but after the university. You want them to leave the country. I think the major question is how do you attract these people that once went abroad to come back? And I think that's the key point. And I, I'm always like, I always try to understand why people focus so much in about how we stop the people from going instead of how we attract people to come back. And also not only, but not only Romanians, so this is also how do you attract people from other countries? I mean you could in principle Romania the economic developments and so on. I think if I speak just about Romania, will be quite good in terms of living and so on if you can offer a decent salary. Not, I think no scientist wants luxury or something I haven't experienced scientists that really want luxury. You just want to have a normal life. And I think with that, you can start attracting people from, other places in the world. So you you actually don't have to focus on only the people who want to return to Romania, but you, you, you should focus on attracting, uh, from other parts of the world. think that that's key. That's where you gain as a country. That's where the people coming to you gain as a person, right? In many respects.
milosz-wieczor_1_06-10-2024_130413That's right. I hope there are those positive feedback loops where, you know, people, people like you or people like my former boss in, in Poland who also came back after a postdoc in Germany, you know, essentially planted this culture of postdocs in his or yours, hopefully, graduates. so this is something that only becomes a cultural standard. When there is a critical mass of people who come back and who want to promote this question and then at some point it becomes the kind of cheap political point for the government I mean, in the end, the resources that the government provides for science are on the level of 1 percent of GDP. So this is really something. compared to the kind of image boost of being a modern scientific society. This is a very small cost for almost any government, right? That you can, uh, pay.
vlad_1_06-10-2024_130413that.
milosz-wieczor_1_06-10-2024_130413Yeah, exactly. So they have to have some sort of again, critical mass of people who want the change, because I think what tends to happen in countries that have this cyclical, internal, um, pedigree of scientists people who don't leave is that people prefer this system over, you know, over this internationalized postdoc, scheme, uh, because it's just easier. You don't need to leave anywhere. You don't need to be evaluated. So I think there is inertia in every country before it actually. transforms to this, but this transformation is exactly driven by people like you who come back and can, you know, can explain the benefits, can explain how this can translate into better outcomes on every level of the economy and society and so on. Right.
vlad_1_06-10-2024_130413having level of stability and tenure track positions and stuff like this, because one of the things that I experienced in the West, is basically that it's very hard to, get a tenure track and a 10 year position. And then, then you end up I mean, okay, if you are the top 1%, which are 5 percent that I don't know, have all these big grants and so on, then you are safe and it's fine. But if you are a good scientist, but maybe. You are missing one of the things in your cv, then, then you have no option, basically. And I think that that's something that also has to change because in life there is also other things, right? So I mean, as much as we love science, like I love what I'm doing, basically. I love the academic system, but there are other things in life that are equally and sometimes even more important. And, and then you also have to balance that. And that's why balance systems where you have both this kind of permanent exchange, especially at a very young level, like PhD, first postdoc, second postdoc maybe, but combined with some stability. Okay, based on evaluation and everything. I mean, it has to be, I mean, you should not lose your productivity because you are now permanent and stuff like this. But I think always balance, balance is always good.
milosz-wieczor_1_06-10-2024_130413I hope it inspires some people to to think about it. And perhaps we have some listeners who, you know, can actually change something in some places that will be a great outcome. Okay, Vlad thank you so much for the conversation, for sharing your experience and your knowledge. And, uh, yeah, amazing.
vlad_1_06-10-2024_130413Thanks. Thanks a lot. Good
milosz-wieczor_1_06-10-2024_130413great having you here.
vlad_1_06-10-2024_130413I hope, uh, had some very nice episodes so far. So, I hope it will continue. I think it's a, it's a great, great initiative.
milosz-wieczor_1_06-10-2024_130413Thanks. I think we have another great episode right now. Have a great day.
vlad_1_06-10-2024_130413I see you too. Bye. Take care.
Thank you for listening. See you in the next episode of Face Space Invaders.