In the third episode, Wojtek Kopeć and I discuss the overlooked yet exciting social aspect of scientific collaboration, and what the world of academia could learn from the industry and statisticians to avoid falling behind. Wojtek also shares his personal experience of starting a lab as a new PI, and we reflect on the role of honesty on social media in addressing the hardships of being a scientist.
In the third episode, Wojtek Kopeć and I discuss the overlooked yet exciting social aspect of scientific collaboration, and what the world of academia could learn from the industry and statisticians to avoid falling behind. Wojtek also shares his personal experience of starting a lab as a new PI, and we reflect on the role of honesty on social media in addressing the hardships of being a scientist.
Welcome to the Phace space invaders podcast, where we explore the future of computational biology and biophysics by interviewing researchers working on exciting transformative ideas. My guest today is Wojtek Kopeć, an old friend of mine who has been very recently appointed a lecturer in computational pharmaceutical chemistry at Queen Mary University of London. Wojtek has been working on the biophysics of potassium channels and most notably on the mechanisms of ion permeation. But many of you might also know him from Twitter, where he gained a large following, sharing the ups and downs of his scientific journey. So obviously we talk about the challenges of starting a new group and about the role of social media in talking about science. But we also discuss the shift in roles that academia and industry play in contributing to science, as well as the social, aside from social media, nature of modern computational science, where we increasingly have to leave the comfort of computer labs and engage with people, well, out there. We emphasize that while science is a great endeavor. It also has its mundane or outright obnoxious aspects, and it's fine to acknowledge them instead of perhaps over romanticizing stress and personal struggles. So if you can't wait to listen, I won't hold you back any minute longer. Let's go. Wojtek Kopeć, welcome to the Phase space invaders podcast.
Wojtek:Hi, Miłosz, hi, hello, everyone.
Milosz:Okay, I think it's been close to 10 years since we first met, right?
Wojtek:Probably
Milosz:I gotta say I was always happy to hear from you as someone who's very relatable to me, you know, coming from the same country, but a few years ahead. And every time our paths crossed, I felt reminded of how small the world of computational science can be uh, which is pretty surprising and amazing at times. But then, it has always been great to see, you know, your consistency, which is something I myself always failed as, uh, as I feel. And last but not least, you gained quite an impressive following in the scientific circles on Twitter. So I don't remember what your Kardashian index was
Wojtek:Yeah, it was very high. It was very high indeed
Milosz:perhaps because of that, many of us feel that we know you better than we actually do from this social media side. But your tweets have often been, you know, both entertaining and talking about important issues. and I was wondering if it was somehow a conscious decision or just your personality kind of took over, to go in this direction.
Wojtek:No, yeah, great question. And, uh, it indeed happens oftentimes that at conferences or other meetings, people stop me and say, I know you from Twitter, which I would never think that that might happen. And I'm always surprised. Uh, when I hear it, it was definitely not a conscious decision. I think I haven't realized, and I still, haven't realized how big this community might be, although I don't know actually how many people read those tweets. it, basically treat, Twitter as a, as a space to vent. and I think this is something common in the scientific community that we are. Very often facing very different issues that for our friends, colleagues, partners outside of academia, even if they have best intentions, it's sometimes very difficult to explain what is happening because some of these issues are quite unique, and I think at times they also sound ridiculous, but yet, uh, they happen over and over again. so I think there was a time when the scientific community in Twitter was kind of close, and prospering that I think also has changed. so, so definitely I felt, oftentimes that when I vent there. That first, it helps me because well, I can just share it with someone or, or with the world, and also that, other people being in a similar situation can see, okay, it's not only me dealing with this problems, there are also people. dealing with it. And I think, yeah, maybe at some point, it actually has become more conscious in a way that I thought, okay, this is a problem. Maybe it actually makes me vulnerable. But if younger people or junior people can see that I'm struggling with these things. I'm not saying that, that I'm a person they should aspire to be, but if they see that other people are struggling with it, maybe it will help them somehow.
Milosz:Right. I remember your slogan, which was kindness in academia, which is something I think very much undervalued at many points in science history. So yeah, I think getting those things out there, it's actually great at, starting the discussion and perhaps fixing things, which is also what this podcast is trying to be about, like trying to highlight some issues that our community might have and, making people think of ways to fix that.
Wojtek:Yeah
Milosz:but then. very recently you moved from Göttingen to London and you're just starting to set up your lab there. And I think it's a good moment to ask what the life of a fresh PI looks like for all our listeners who perhaps aspire to be there one day.
Wojtek:Yeah. Okay. So maybe first a few disclaimers maybe this is not is often being talked about of course. that I'm going to talk about is more or less, singular experience. So yeah, be advised not to, not to think that this is always like that. And I think yeah, when we talk about this general questions, like how is the life of a PI how it is being a postdoc, how is it being a PhD? We have to remember that these things can vary a lot between countries, between systems and being a Ph. D. in Denmark, and being a Ph. D. in the U. K. or in the U. S., you can face very different problems and very different barriers. So, this is just something to keep in mind that I'm going to talk only about my experience, which is mostly Europe based. And now it's indeed starting, position in the UK, which is again, a little bit special where you have a little bit more, those semi permanent positions. The P. I. Positions lectures so, and they are usually is no term tenure track usually, but they are can become permanent. But they come with, oftentimes with their own issues, which is very limited funding, for example also another disclaimer would be that I made a step from being a postdoc and later project leader at the Max Planck Institute, which is purely research focused to a UK university, which primary focus is actually education. I mean, we are supposed to do research, but, I would say that the most important thing for the administration. Is actually education and students, so that was definitely for me specifically a very big step that for for a couple of years. I haven't been dealing with undergrads specifically with the year 1, 3. Undergrads, with bachelor students with bachelor thesis. And here, basically from day one, I was assigned a number of students for example, three bachelor students. I had to come up with projects for them, which was already very much, defined how the project is supposed to go for the whole year. So, so it was already a module or a course, if you wish. There were checkpoints defined, assignments for them. So, uh, that was a huge difference from an environment where if someone wanted to do research, we were just basically trying to accommodate them as much as possible and this person could just take a deep dive into research here. Um, this, this is still within the, education program, within the uni program, and so on. So, so, that was a huge change, Another major change, uh, would be the fact, that I, I just faced and I'm facing a lot of new things that I have to figure out how to do myself people will give me advice, but that's it. I mean, you don't have a, now a supervisor or PI who you can usually ask, okay, how, how do I do A, B, C? It's. Well you have to do A, B, C, so there's a lot of trying things and not knowing things and figuring out things, and these are things not really related to research per se, but mostly to university or research management, so this is the major part of my day now, and obviously There is very limited, if any, time to actually do research myself. So, this is a huge, huge change. Of course, partially, I, I knew it was going to be this way what I was maybe not prepared for was overhead that comes from thinking about this thousands of little, little things that you're facing to, set up things and get them running. from the administration point of view, from the technical point of view I think the additional challenge for me, and this is true for, anyone that, that does this kind of move at different stages of an academic career. But I think especially at this stage where, when you, uh, start an independent position, if you do it in a different country. where you never lived before because people will start talking about things that you have no idea and they are in the system for years. So they start using acronyms about the funding sources that you have no idea about. They are using acronyms about What high school pupils are learning before coming to the university, they are using acronyms about the modules and you have absolutely no idea what they're talking about. And you're, as I said, my primary job is to teach. So they hired me to teach. And, uh giving I'm given the module. the course to teach and now it's my responsibility. So,, there are many different things that I think that that would be the primary ones that are, are the most challenging indeed.
Milosz:Right, I guess it's division between pure academia, I mean institutes, research institutes, and university where you have a mixture of teaching and Scientific responsibility is one of the most important ones when it comes to planning one scientific future, right? I also came from doing a PhD where a large part of that was teaching and I remember how exhausting it was Because most of the time anyway, these were new subjects that we had to come up with And uh, yeah now it's for me as a postdoc It's just pure research and I understand that at some point this will also switch to perhaps 50 50, or I don't know, as we, as some people say, it's 80 percent teaching and then another 80 percent research. Uh definitely something important to consider. And then, as you say, every country has its own cultural environment when it comes to doing research or teaching that's important to have in mind. Yeah, I remember, I remember from Twitter, again, that you struggle with insomnia, so I won't maybe ask which Scientific, breakthroughs keep you up at night. Uh, but I will definitely ask what are the things that you're looking forward to in, in the scientific world right now.
Wojtek:Yeah. Yeah. So the, so it's just a short note. Yeah. The my experience with the insomnia. That's where partially caused by working too much and burn out. they really kind of showed me the dark side of working too much and forced me basically to keep some level of work life balance. So we can talk about it later. Yeah, what, what keeps what, what scientific advances. definitely with the amount of every day. these days, I would think that, enthusiasm towards science is maybe dipping right now, just to put it, and I think that happens to everyone. You cannot be all the time excited about every single thing. You will find times where, where there is something coming up and everyone want to participate and maybe you too and there will be times where, where you will start asking yourself where, what's the point? And, and I think this is also normal and we, probably should also talk about it and normalize it a little bit. Definitely if we talk computational biophysics, I mean, this is eventually what, what we do. If we talk computational biophysics, we are living indeed exciting times. And for me, what is exciting is the fact that we are closer and closer in actually connecting directly experiments uh, with calculations or simulations. So, there's more and more examples where simulations are actually either guiding the experiments or providing, uh, observables or quantities that are the same that can be measured in experiments. And I think for me, when I think, when I think about science as a, as a kind of joint endeavor. This is exactly where we should, go, or, this is what I personally find, uh, exciting because actually forces people, or, It makes more and more sense to collaborate. And, for me doing science has also kind of a social activity. I think the times sitting alone and trying to figure out one or two things, I mean, this can be very rewarding. But I think, the questions now or the challenges that we face, they, they have to be addressed from a very different angles. And for me, the, the, the most interesting moments in science were really those where there were sets of different experiments. There were sets of different calculations, simulations, and you start seeing very similar thing, or you can think of an interpretation that gathers these different approaches together and provides a very clear picture of what might be going on. And I think these moments are really, are actually living for this moment to, see these things coming up together and, and kind of, uh, filling the gaps. The gaps in knowledge. So I think this is, where are we going as a community? And, this is definitely very exciting.
Milosz:I think it goes a bit against this, um, stereotype, right, of a lone scientist. I mean, we all grew up hearing those stories of people solving things in a almost single handed way. I mean, of course, even back in the day, there was plenty of collaboration, plenty of people talking. Well, Göttingen is definitely a place where all those talks happened for, say, quantum mechanics back in the day, right? But I think we are not being told enough about this necessity to train the social muscle. in academia, there is,
Wojtek:Absolutely.
Milosz:wonder if, if there's something to be fixed about how we are being taught about the necessity of collaboration or the necessity of keeping, let's say social cohesion within the field
Wojtek:yeah, I mean, definitely. but one can look at those things from a different angles, right? So even if we look at the scientific publishing, for example, and still referring to papers as the first author asking who is the first author. Uh, so we again put this one person on the pedestal in a way and kind of forgetting there might be 20 other people that contributed to the work, but we kind of forget about it and are still okay, this, this is the lab of X and Y person. They are super smart and very good. So, yeah, so we should, I don't know, give them money or something yeah. so I think this, the stereotype of a lone scientist solving things or, extremely smart. Persons being able to really move things forward. Of course, they are a lot of extremely smart people. as you said, the social aspect is still kind of forgotten and undervalued. In my view, and and we definitely should think about the ways where, these multidisciplinary collaboration works, are valued even more. And Yeah, as you said, this is the, the, the perception hasn't changed that much or it's, changing slowly and, all the things that follow from the perception, right? The, the Nobel prize. is again a very good example that, we come up with very important discoveries or fields actually these days and, are able or we are allowed to honor up to three people for a given discovery, where in reality, those discoveries might have taken. 20 years and, and sometimes the Nobel Peace Award is a list, actually, the, the number of people working on that. And it's like a hundred names in those lists but still, yeah, then, we hear, okay, well, this person got a Nobel prize. We still, there is nothing else in scientific world that even remotely compares to this word Nobel prize. especially outside of academia, right? Like if you talk with anyone on the street, they will know what Nobel Prize is, and they will probably think that scientists with Nobel Prize are extremely smart. I mean, this is, and I think this is everywhere. And somehow we haven't. come up with anything else that would promote and value science in a similar level, in a remotely similar level course, one can say that it's a human nature, right? That this is how everything else in the world works, that certain people are being recognized as very important But I mean, we are scientists, right? So in theory, we should be able to recognize these things and come up with a better system. Um,
Milosz:Yeah.
Wojtek:uh, clearly it hasn't happened yet.
Milosz:Yeah. I mean, I, I kind of understand the point of publicity, which is just to promote. People in the sense of, I think we all somehow grew up with role models, right? I mean, becoming a scientist usually means that you followed some sort of role model and maybe this is harder to do if the role model is a group Or if it's a mix of people who you don't really know who contributed what So, I mean, I understand that for keeping the public interest in science, perhaps Highlighting the individuals is important, and then, as you say, it translates into so much, let's say, power and recognition that it has become a much wider social issue in the sciences. For sure, everyone who knows a scientist, and is not a scientist themselves, think that you're gonna win the Nobel Prize one day, right? This is the,
Wojtek:Yes.
Milosz:uh, the cliche idea that you get from every family member or every friend Right, so that's, I guess that's one thing we might aim to work on in, in the sciences. You also mentioned the growing influence of the industry, right, on how science is being done and how it affects, well, purely academic endeavors as well. Uh huh. Uh huh.
Wojtek:briefly. Yeah. Again, these are, these are the experiences or observations that kind of come from our bubble. So this probably works differently in different scientific fields. We can also Theorize about it a little bit But let's say we talk about computational biophysics, which often Touches on things like computational drug design computational drug binding protein folding a bit, so these things have been done largely in academia for a number of years, and one would say that the progress was pretty good in the 90s, before the 90s, in, in early 2000s, and then a couple of bottlenecks were hit. Or walls that happens all the time. But I think in in recent years we see, that some of these bottlenecks that have been bugging scientists for four years are started being solved by commercial institutions or companies and I think this is particularly visible in anything that is related to computational science, because, let's say computational chemistry, computational biophysics, it's not particularly strong, uh, it's a well established field, but sometimes I have a feeling that we have to justify our existence, uh, whereas if you look at the power of computer based companies. These are actually the biggest companies in the world. things like Google even Twitter mentioned now or, uh, Apple and so on Microsoft now. And on the other hand, you have, of course, pharma companies that are interested in these questions that we mentioned at the beginning, like computational drug design, where, uh, and of course, pharma companies, they do have quite a bit of cash as well. So if you look at these two perspectives, and if these big players get interested in, in similar questions to computational science in academia, it's perhaps not a surprise that there will be topics where, there will be no chance for academia people to compete purely and eventually it boils down to money, obviously so, to be more specific. Yeah. So protein folding question, although I would say the protein folding question is not solved, but it's not even the progress. I mean, the, it was really a paradigm shift that came from AlphaFold, uh, basically, uh, prediction of a structure from a sequence. has been tried for many, many years in either pharmaceutical or setting. Google basically solved it one way or another, but it was a huge paradigm shift. and probably it would have been very difficult to be achieved in academic setting unless we think about one of the few places in the world that get very, very good funding, so, these things happen and I guess, uh, more and more, these computation based companies are getting, interested in similar things. then, uh, it will be more difficult to compete again. Question is, do we really need to compete? Maybe we can actually work together or combine our, expertise or actually maybe the role of the academia will be more. To, uh, to train people, to teach people, and then the research will be done in some sort of a collaboration. This is, uh, definitely something worth exploring, I think. Because, I don't, if we think about science or research as a mankind endeavor, is there a point of, of having hundreds of little academic groups working on similar things? Or try to consolidate and try to get the best players and try to really solve things that are important. In solving, so, so that would be one aspect if we again talk about computational science. so that, that what I was mentioning now, it was maybe more from a software perspective and approach to largely software development and, and software usage, but there is also a big component of hardware. Uh, so many, many simulations. uh, computational work depends, for example, on GPUs. And if we think about machine learning, AI, they depend on the TPUs or even different specific architectures, which are now being produced basically by one company being NVIDIA. Uh, so we think that all this research being done even in academia depends on, on one provider. This also doesn't paint a very good picture. or it shows how much we, even if we don't want to, how much we actually depend on industry. So there, there is, uh, I think there is no way around it unless the funding for general in science going to be increased because with players one with one of the most richest companies in the world, I mean, there is no easy way to compete. And as I said, maybe there is actually no point competing
Milosz:Yes, I hear that in many cases, uh, scientific projects are being outbidden even by, the industry when it comes to buying GPUs, for example, for the latest cluster around here, they have to wait more because someone else paid more for the GPU. So they are just standing in the line waiting for the GPUs to be physically produced. They're not off the shelf products anymore, especially if you're thinking of a big, big cluster, big, big center. So that's definitely true. And as you also mentioned before, there are many scientists who are taking up parts, academic part, industrial positions, or fully switching to industry, but still doing academic work.
Wojtek:Yeah.
Milosz:And I think in a way it works both ways because I'm also starting a collaboration with a local company and I see that their mindset is also. Very much academic. So they do very much basic research things as well. They publish they don't just keep things proprietary. So I think that's also a welcome mindset shift that commercial companies are starting to do open source or open science. In a way, of course, it's not fully open as we would do in science, but it's again, it's a welcome. paradigm shift, perhaps.
Wojtek:Yeah. Yeah. And on the other hand, as you say, it's maybe not fully open. That's probably not very good but on the other hand, there might be industry standards that might be actually useful. So if we again, look at the computational science and I mean, this is also like a private thing that really my nerves is. Eventually in, if you want to really, advanced your science, computational science, you need to produce code yourself. And in academia is completely undervalued, partially because of the system itself that promotes short contracts, few years contracts. So there is no continuity. You cannot hire a full time software developer in most of the groups. So how can you be sure that you're that whatever you're working on is actually continuously developed people that are Able to do it. They do tremendous job and they do all sorts of acrobatics to keep these people Hired for, for as long as possible. And yet these people get so little recognition in the whole
Milosz:Mm hmm
Wojtek:ranges from salaries. Again, typically a good scientific software developer these days, if decides to switch to a company, the switch in salary is just tremendous.
Milosz:twofold or threefold. Yes,
Wojtek:exactly, exactly. So, that. That results, uh, very often in code producing in academia being being sloppy, and this is not necessarily the fault of the researchers is just they trying to operate in the system that they're being put. You have three years to do a PhD. You have to write the code. You have to use it. You have to write the paper. And if you put the code on GitHub and it's well documented, well, it's good for you, but you not going to get extra credit within academia, which is just ridiculous. But this day, uh, the institutions don't understand this.
Milosz:Yes,
Wojtek:so, so sorry, I mean, I I can run on this for a long time,
Milosz:I know, I can attest that I have some bug reports pending in one of the popular libraries for around two years now, and still nobody has fixed that. So that happens a lot. And I think the last thing you wanted to mention was the question touched upon in one of the well, paper exchanges, right? Which was about statistics.
Wojtek:Oh, okay. Yeah. Yeah.
Milosz:That's an interesting point that we as a community have always faced with various degrees of success.
Wojtek:yeah. I mean, depends a little bit how we look at this problem. So if you look at science as the whole thing, people already long time ago figured that, okay, reproducibility and statistics is important. So, we can complain about the journals and scientific publishing, but at least the good journals these days, they realize that and usually they require you to follow a certain procedures or, best practices in the field for statistical treatment of your data. Of course, are able to push poor data through, but I would say for many experimental techniques, this is now. At least certain baseline has been established. It's very difficult to publish a single experiment. Of course, it depends on the experiment. But, uh, typically these is enforced by the journal. So, so I think, that's not optimal, but that's okay. And, particularly in, in computational I have a feeling that maybe because the field is relatively younger. if you compare it to, say, x ray crystallography, there's definitely a few decades that are lagging behind but there were definitely or there are still examples of, things that. are simply a statistical fluke, and it's not even being considered and it's being published in journals that, pride themselves of being in the, top of the field. this is, this is the moment when these things are becoming concerning because it's damaging the whole field. There were, of course, certain examples in the field that, uh, that, uh, recently really, sparked this discussion again. So that was the work of my former colleagues at Max Planck that, commented on a, on a very, obvious case to many eyes in one of the, the better journals in the field where, Single simulations at different conditions were used to, to talk about the trend in this conditions. No error bars, no uncertainties of any case. And if four points seem to be on the line that the, the conclusion was that there is an obvious trend here. And it took good two or three years to actually get it right. And in this case, it kind of worked out, but how many are there that are not being, discussed, or haven't been spotted yet. So, so as I said, it is, it is damaging to the field in the end, because if these problems are being highlighted, they're going to be picked up by, People outside of the field and they will be like, okay, well, if these people themselves don't know what they're doing or, or there is no statistics involved, well, maybe not trust them at all. And the good thing is that now certain communities started to put forward best practices in the field, how to run the simulations or how to do the statistical, uh, statistical treatment in a way that at least something is involved. Of course, often. This is not, just just tick box that you can just tick off. You have to think what you do and. You cannot just blindly follow the, the, the guide and, and be sure that your final result will be correct but at least it is slowly changing, uh, in this direction because the feeling was, at least I had this feeling before. That, yeah simulations oftentimes were used just to, just it was maybe a very expensive visualization technique. Okay, something is moving, something is dancing. So we just do a little bit of hand waving and there we go there is really powerful machinery behind and one can put much stronger hypothesis based on simulations, but, uh, they have to be treated as any other experiments. So they can be. They can be right. They can be wrong and there will be degree of wrongness and there will be degree of rightness. So
Milosz:Yeah, I, I remember many papers that were essentially like we launched one simulation of the wild type and one of the mutant and after 15 nanoseconds something flipped. So we declare this the mechanism and that's, that's the bottom line of the paper. Yeah, it's, I hope it doesn't, repeat much these days, but at least I think nature published this. checklist for reviewers. As you say, checklists are never perfect, but it's at least a step in the right direction, I guess, that people will start thinking of those before launching anything.
Wojtek:Yeah, absolutely.
Milosz:So, yeah, I don't know if experimentalists will start trusting our results, but maybe at least we will start trusting our results eventually that would be a welcome change., So, wonderful. I think that's all we have for today. Again, Wojtek Kopec,
Wojtek:Okay.
Milosz:thank you for being on the podcast. Thank you for taking the time.
Wojtek:Thank you.
Milosz:Wish you the very best with your new lab at the Queen Mary University of London.
Wojtek:Thank you so much. It was a nice to share some of these views and, uh, yeah, hopefully we'll talk again soon.
Milosz:Hopefully, yeah. And I hope this also inspires some people to think deeply about the problems. Thanks so much. Have a great day.
Wojtek:Thank you. too.
Thank you for listening. See you in the next episode of Face Space Invaders.