Phase Space Invaders (ψ)

Episode 5 - Paul Robustelli: The dissolving barrier between industry and academia, embracing our artistic side, and new models of funding

• Miłosz Wieczór • Season 1 • Episode 5

Share episode

Send us a text

In the fifth episode, Paul Robustelli and I discuss how the once very sharp boundary between academia and industry has been becoming increasingly porous, and what implications it will have for future careers in science. Paul shares his experiences and observations about the competing models of funding scientific research, and makes a point that they can eventually synergize rather than compete. We also talk about the often neglected creative or artistic nature of scientists, and how we can use it to better connect with other scientists in a more genuine way.

Milosz:

Welcome to the Space Invaders podcast, where we explore the future of computational biology and biophysics by interviewing researchers working on exciting transformative ideas. Our today's guest is Paul Robustelli, an assistant professor of chemistry at Dartmouth College, who works on modeling intrinsically disordered proteins. Paul became renowned in the community for his contributions to the force field modifications that significantly improved the simulations of unstructured proteins. And that was while he was doing his postdoc at DE Shaw research, a private institution that made headlines around the biophysics world with their bold goals and an unconventional funding scheme. However, as he explains, his fascination with disordered proteins was born out of his experience with partially unstructured systems studied by NMR and a desire to push the boundary of what can be modeled with simulations. So we talk a lot about the interface between academia and industry, how the once tight boundaries are becoming fuzzy and porous, and how we can use it for everyone's advantage. But then we move on to discuss the cultural dimension of academia and how the stereotype of a boring and strict professor hides the complexity of us scientists as human beings with our often quirky hobbies and passions paul makes the point that we can all benefit from a change of perspective where we'd see ourselves tongue in cheek, of course, as overpaid artists rather than underpaid office workers. So in an effort to emphasize some positive aspects of why we're all here, let's all appreciate the similarities between academia and other creative ventures. And let's put more effort into making it our advantage. Hope you enjoyed the conversation. Paul Robustelli, welcome to the podcast.

Paul Robustelli:

Uh, thanks a lot for the invitation

Milosz:

so your scientific path started with inferring structure from NMR data, but then you joined the side of chaos and decided to move into disordered or unstructured proteins at DE Shaw research. And the big breakthroughs that came out of that were the Amber Force field and Associated water model that I think revolutionized that field at the time. in a way it was a peak hype moment, uh, where people saw the power of big money thrown at molecular simulations perhaps for the first time. though now we have many more examples. at the time it's kind of insanely boosted, our confidence that what we are modeling has some reality to it. And, for the first time we approached the promised land of simulations, which was you put things together and they figure out what to do where to bind, now that you're back in pure academia, how do you feel about the different dynamics of these two worlds?

Paul Robustelli:

Yeah, so let me uh, go back to the start I guess'cause, uh, disorder, proteins and, and motion has sort of been a, a slowly evolving part of my career. But I actually,

Milosz:

actually.

Paul Robustelli:

an undergraduate, um, when I was 19 years old, my first chemistry research project, I'd done some biology research, was looking at these flexible organic molecules and, uh measuring NMR data, scaler, couplings and RDCs and relaxation parameters, and then coming up with, an ensemble of confirmations to reconcile that data. So it was something that, was like one of my very first exposures to molecular structure and research. And then that sort of colored where I would go to grad school. So then I was excited about that and I saw that the Vendruscolo lab, uh, and the larger Dobson group was, was at Cambridge. Were, uh, taking these approaches for proteins. And so that is sort of what brought me there. And that is really where I kind of know, the huge center of, understanding of misfolding and aggregation. And, and that's where I was, uh, exposed to IDPs and got interested in IDPs. And, and I always sort of viewed disordered proteins as we started doing structure and dynamics as, um you know, the extreme test of that getting the roters and ubiquitin, which was like the, I feel like, uh, when I was a PhD student, right, like every computational integrative paper was another model of ubiquitin, and it was like two rotamers and a hydrogen bond you know, so that's where these ideas were sort of developed. And then, you know, I sort of viewed, all right, once we get out of this sandbox, like what's the hardest thing we can do? And I was lucky to sort of work on a, a cool protein that was half folded, half disordered when I was doing a, a postdoc in our Palmers lab. Um, and then. the, by the time I decided to go to DE Shaw research, it seemed like the real blind spot of force fields. We all kind of knew they weren't working well. They sort of had been starting on this, tip four PD water model that was in the works, when I joined and yeah. And so that was kind of the most fertile region I think for additional like rapid development I would say that for me the appeal of sort of simulation has always been this interface where, It's absolutely essential. Like you can't, it's not just like a shortcut for experiments, but it's, it's something that needs to be hand in hand with experiments. And so for that disordered proteins was like the clear, um, it was always going to be a key piece of this. The, the

Milosz:

Right. So Pushing D Frontier, uh, it's a, it's a very common theme here so far. Everyone, I think has said that hand in, hand in experiment is the, the new way to go, which is just the last 10 years, right? But, uh.

Paul Robustelli:

Yeah. But in, in terms of what you mentioned for these new, finance models or ways of funding science? I think Dhar research was, it is really interesting for someone to just devote a, a large chunk of their personal fortune to funding things. And I think. Protein folding was, they were sort of lucky to have hired some great protein folding scientists in, in Kristen Lindorff Larson and Stefano Piana

Milosz:

early

Paul Robustelli:

on. And, and so that, that became a very natural application of these supercomputers. And, and I think that was kind of, uh a big moment for the field in terms of being able to systematically look at folding that we have data on of many proteins with the same model. And I think maybe there was always this feeling that well this lab can fold this protein with this force field and this lab can fold this protein with this force field. And in a sense that it was like, how much of this is cherry picked? You know, how robust is this really? And I think those, those early science papers from DR research I think did play an important role in sort of the. taking people outside of the field to take it seriously. I think like, uh, there's always been a good segment of structural biologists, or particularly biologists who, who like, think we're just playing video games all day. like

Milosz:

Yeah.

Paul Robustelli:

who's an anesthesiologist who refers to my research as video games. So I think d it's Oh, how the video games going. Um, I, so I think, yeah, that folding those early results, we're an important certain milestone for recognition from the broader fields, from broader structural biology and biologists.

Milosz:

Yeah, I I repeat that. It's, I think equally important for us to believe in what we're doing and for other people to believe in what we're doing, right?'cause there can be two disconnected questions.

Paul Robustelli:

Yeah. And then in terms of the model so that was really interesting model, right? Uh, that, that you have one person devoting their own personal finances to this. And then he sort of set it up in a way that he was at the top of this organizational infrastructure and kind of everything went through him. And, and we got to shape the vision to an extent, certainly, but it, it sort of not necessarily the Bell Labs here's your money, go do what you want, write whatever papers you want sort of scenario. But yeah, it did expose me to sort of as someone at that career phase where I was like finishing a postdoc at Columbia and, and sort of thinking about, all right, so if I apply for professorships now and, and you know, I get a startup of this size, like what is gonna be tractable? Then to sort of look at, all right, well, if I have these resources at you know, what can I accomplish in that same amount of time and not necessarily wedded to that institution for my whole life or that sort of model. But it, it was, yeah, an early time where it was. I could look outside the traditional structure of your startup in your first NIH grant and sort of make a calculus about what is it I could accomplish? And, and sort of one of the things I was really, you know, I was really excited about disorder proteins and, the notion of, drug discovery for disorder proteins and that, that was kind of kicking around the, the Cambridge labs and Vendruscolo and Dobson and, and looking for aggregation inhibitors. You know, at that point that was done largely in the dark, right? We didn't have any structural models of these binding events. So my goal at DE Shaw research was always, all right, can we improve the physical models? Can we use this to demonstrate that we can model ligand binding for disordered proteins? And then having demonstrated that, or, or, you know, determining whether or not it works, what can I then do with the next step of my career? Like that, like to take that demonstration and then it becomes easier to write NIH grants, to study these things. And so kind of like using this one resource and model of funding to enable another path. Um, so as you mentioned, you know, that was kind of early on. Really new, but we, we've seen a, a number of, sort of, even within that sector of, of people who made their fortunes and hedge funds, uh, making large donations to institutes. You have the Loeffler at Stony Brook a, a huge investment there. Uh, you have the Flatiron Institute, you know, which we're all slightly different models where you could be money to a department and letting them do what they see fit with it. Or you can sort of create an institute outside of a university hierarchy. Uh, where maybe you have a, a little more say in what the sub regions of funding are but you know, a different model from DE Shaw research where you have one person who's an author on all the papers and it has been really interesting. And then I think along that timeline, know, in my period since leaving DE Shaw research to start a lab at Dartmouth there was, and maybe some of this was just a zero interest rate phenomenon, but there was this huge rush of funding for computational startups. I mean, there's a huge rush of funding for all startups. But you, you were seeing a lot of people you knew doing postdocs or starting labs, maybe, starting a company on the side or switching to start a company and, you start to think about this practical question of, well, if I have my startup and my NSF grant to work on this problem I'm passionate about, or you know, there is a VC who is willing to kick in some certain amount of money, which may be substantially larger than your startup in your NSF grant. You know, is that something you can do for a little while? And then also have these things like Altos Labs where you're sort of moving, buying labs wholesale to come outta the university into a private funding environment, my friend Andrew White, who maybe, maybe you've reached out to for this, but uh, right, he has this company that he's been a part of Founding Future House, which is looking at AI agents and that has some, backing from Eric Schmidt, you know, private philanthropy in a more hands off way. And I, I think. It's kind of a very exciting mixture right now. And it, I, I'm kind of optimistic about watching these things evolve and, and having many different models.'cause on the one hand, when

Milosz:

Mm-Hmm.

Paul Robustelli:

coming up through the system, there was all this sort of, the giant increase in the number of PhD students in the early two thousands and the steady amount of faculty positions, and no one was retiring and there was kind of this clut of scientists and for a long time it was, you know, everyone was kinda like, oh, the goal is to be a professor. And that was like mathematically, statistically impossible now. And so I, think that this in some ways, right, the. Funding agencies have been like, what's the solution? And, but in some ways I think maybe we're looking at it sort of evolving where you have many

Milosz:

Right.

Paul Robustelli:

models of funding and science.

Milosz:

It bursts the pipe of supply in a way. Uh,

Paul Robustelli:

Yeah.

Milosz:

yeah. It's, it's, I think a lot of people are now on the crossroads, right? Thinking, finishing the PhDs and thinking where to go. And as you alluded to with the different models, there's probably going to be different. Degrees of liberty in terms of what to study or how to approach your, your problem and different rewards. So there's, there's a sort of trade off. Do you think there is a sort of trade off in how free I am to approach an interesting project versus, uh, how free I am to, you know, uh, achieve something develop my skill as a future pi. If someone thinks of be becoming a PI one day, um, is it, uh. Consideration, you know, when choosing whether to go to industry or to stay in academia.

Paul Robustelli:

One, one of the things I'm, I'm hopeful for that these boundaries are gonna continue to become more porous. And, and I think that that is kind of best for everyone in terms of not having this stark dichotomy and, and Right. So I'm someone who has gone to sort of a weird industry funded thing and, and you know, but I've kind of always had my. Eye on the prize of being, you know, publishing papers and, and you know, the time I worked at DCR research, I was going to the same conferences. I was interacting with the same academic network and, you know, very much considering myself a, member of that network and trying to cont doing things that would value that network as much as it would, you know, the overall pursuit of integrative structural biology in the, the community of academic studying disorder proteins. As much as it would benefit, uh, the company I was working for. And I would say that, I thought that transition was great. That it, you know, that experience I got, I got really informed, you know, my network grew that whole time. Uh, I got to try a lot of different things with different amount of resources. It really informed sort of the, the grants I started writing right away. I think I had sort of a mature. More mature understanding of, mentoring, you know, being in a environment where you have, you know, they're hiring a lot of early career scientists and so I think it was a, a great help it's interesting now as my first PhD students are looking at jobs and they're sort of applying to a smattering of, they're thinking, all right. If I did an academic postdoc, what would I maybe do? But also applying to various biotech companies at different stages. Uh, you know, some larger series defunded, biotechs, and then sort of also hooking around and looking at some of these series A, series B companies. And, and it, guess compared to when I was finishing my PhD, I think it was, uh there's just so many more options, I think, and, what is nice is that yeah, we, not everyone can get an entry level computational chemistry position at Sanofi or Merck or whatever, but I think lot of these biotechs, maybe earlier phase are gonna be right a, a a much better route into sort of developing an industry than

Milosz:

Mm-Hmm.

Paul Robustelli:

postdocs.

Milosz:

Now I think it's an incredible thing that's, I mean. One incredible thing we can learn from the industry is the structure, is the management and the accountability, that they have you can also ask whether, you know, the merging or the blurring of lines between academia and industry has any perils, like conflicts of interest or different incentive structures you know, it's, I think it's a valid question we could ask if. It's fine to blur the line in the sense of most people being somewhere in, between or is there any value in preserving the distinction between academia and industry? I don't know. I'm, kind of, I agnostic there, but it just comes to my mind as a natural question.

Paul Robustelli:

Yeah, it's a, it's an important question, I think. Yeah, it's a good question I would say if we think sort of, of the old model and I, yeah, and I kind of wanna bring up, the Open Force Field Consortium and the, the Molecular Software Foundation as models of, of sort of a hybrid. But, so in the old days right, Merck would've hired their computational chemists and they would have worked closed doors with Merck Resources to build something like the Merck Molecular Force field. And, and that was eventually released. but kind of the, the lessons along the way were sort of kept firewalled in the, you know, and a lot of that was proprietary. And so I think if all the IP remains safely in its legally protected buckets, the transfer of expertise back and forth is gonna be valuable for everyone. For people who have learned the lessons of what were done with the resources at companies to then have that inform the problems you're posing academically. And then for me, so even since I, you know, started my lab at Dartmouth, I've been involved with a couple companies in terms of sponsored research and uh, some consulting. And for me it's super important to understand like the boots on the ground challenges. Like what are the scientists working on today? Like, what tool could you develop? That a company trying to drug IDPs would use next year or right now. and so I like to, be close to that and that informs the questions I ask. And then I also have like very, like much more obscure kind of academic questions that are just fun. Like can we use topological descriptors of disordered proteins to make models and, and, and things like that that are just. Emerge from coffee breaks at conferences, talking to

Milosz:

topologists

Paul Robustelli:

and are fun to do.

Milosz:

Yeah, I love this because you are kind of making the point that you can keep both, right? So keep doing meaningful science that has real life applications and pushing some boundary in theoretical physics or Theory of simulation without necessarily committing to either.

Paul Robustelli:

yeah, and we've seen, you know, you never know what basic science. Discovery is going to rapidly become hugely important in, industry applications, right? Like I

Milosz:

Yeah, totally. That's.

Paul Robustelli:

of the ingredients of alpha fold are probably seen as like pretty obscure, machine learning research at some point. So I, I think we're all sort of aware and I, and I think the companies are increasingly aware, right? Like, so things like open force field, like they're getting. These are companies who want, more transferrable models and they're, they're buying into these efforts and understanding that it, it benefits everyone. And

Milosz:

All right.

Paul Robustelli:

I, I, I can only hope that that consciousness continues to grow, and we see more of this, but I think you can do both. Right? I don't, think there, one thing that I do feel like is a, anything there's a little too much I pearl clutching of ip, which like I get if you're a lawyer, but as scientists, we sort of, I think you have an understanding, like if you work in these organizations, it's like, all right, well we could have our legal, our legal team spend however much time safeguarding some possible thing that arises out of this. But we sort of all know that this isn't gonna be that important or like, you

Milosz:

Yeah. Yeah, definitely.

Paul Robustelli:

is not gonna be the difference between a drug or not a drug. maybe this boundary is best understood by scientists and not lawyers in terms of what's important. I, I get it right. Why they have to do this. Like, you can't raise money with venture capital and then give the, the seek the structure of your compound away to an academic to like make a patent for, I think there's a lot of great things that are seen as a gray area now that are not actually gray areas that in the hands of scientists you would be like. just doesn't

Milosz:

Right.

Paul Robustelli:

of like open force field is an example of that. It's like this is a model that if everyone uses, we all win. And I think the more scientists make those decisions, the sort of the will go back and forth.

Milosz:

Yeah. Perhaps scientists and and lawyers don't really mix well together. To switch to the other question, well, scientists and scientists should mix together well, and I loved it when I reach out to, to talk to you on the podcast when you asked well, can we talk about metal music and, uh, yeah. The point being that we as scientists should find non-scientific points of interest or non-scientific, topics on which we can, we can find common ground. Right scientists being humans. We're not just one dimensional people thinking, doing one thing only in our lives. Actually, many, many people are polymaths. So yeah, you had some reflections on how, you know, how we can enhance the non-scientific or not purely scientific collaboration between scientists.

Paul Robustelli:

Yeah, absolutely. And, and I think i, I, I just remember as an early career scientist having this, you know, as a, as a, as 19, 20-year-old kid feeling like I had to have this like, kind of stoic professionalism with my mentors and it probably did sort of a purpose in a, in a way, I don't know. It maybe kept me a little more like business focused, but. I guess I would say that well as I moved further in my career, like as a PhD student and, and then as a, as a postdoc and, and was fortunate to have some, you know, very like, kind, advisors and, and colleagues who were interested in, in like the, personal interactions or, I, I guess I would, I would put it another way. I, I realized that like, not just the intellectual content, but one of the awesome benefits of science that maybe some of my friends in more conventional industries, like work in New York City and you work in advertising or you work in finance, like you, you often get sort of a, a, an archetype and a, a lot of people. Uh, and those professions are gonna be maybe like socioeconomically a little more homogeneous and one of the huge benefits of science is that like, it's a crazy mix internationally and background wise. And, and, you know, we all come to and interest in models of proteins from, in very different, unique, idiosyncratic ways, from different nations, from different educational systems. it sort of, became clear to me right when I was working in Midtown and, seeing sort of the homogeneity of in Midtown Manhattan when I was at DC research and the people around me that, you know, when I would go to these conferences and, and hang out with these, like you know, if you can get some of your older colleagues to like start telling stories about when they were in graduate school in the seventies, in Berkeley or whatever, or, you know, when they were living on a commune in Vermont, you realize, I guess I got a, I had a better appreciation for how diverse of a group of people participate in science and, and how important that was to me. And that's sort of one of the main interests. But in, and then I guess so for that, knowing that this is what is exciting. And appealing for me and just made me sort of feel having that be a big advantage is is as I continued to sort of move on in, in my career, you know, to the point of now being a A PI at a university. It became clear, right, that I don't have to keep a stoic professionalism with my students. I can tell them about my interest in, in sports and in music and hip hop and metal and, mountain biking and skiing and, and I think

Milosz:

Right.

Paul Robustelli:

of inclusivity, you do your students a, a favor to show that like we are not just like stoic equation crunchers that like we all. you sort of wanna sell science on that, on that diversity. And thing that has been fun is as I started, as it's just sort of come up conversationally as, as found, you know, I like, uh, a lot of different types of music. I like metal, I like punk, I like hip hop. I like jam bands. Um when you just start bringing these things up, you realize that a lot of your colleagues also, you know, who maybe you'd only talked about. Uh, meta dynamics with also like fans and stuff, and, and uh, it deepens kind of the connection and, and I think it sort of helps

Milosz:

Yeah, I think Twitter was one of the catalysts as well, right? When people started sharing their personal stuff on Twitter and now I have, now I actually have two accounts. I have a personal account for all the political stuff and I have a professional account for sciencey thing but I think many people inevitably mixed some personal life with some professional life there. And then you get this idea that, oh, these people, you know, have many dimensions, have many things going on in their lives. And it's, I have to give shout out to my institute, to the IRB for, giving us a space to do all different things. We have a football league here. We have a rock band in the institute my point is people would want to spin it as, you know, something that increases productivity and so on. But I think it has a great value just in keeping us sane from most of the initiatives that we have, the social things here. We don't create new scientific projects while playing football or while practicing a song but it's just a great way to, to, talk to other scientists in a way that it's not a hundred percent scientific. well, it's to talk about work life balance, but yeah, I think it's, it could be underappreciated

Paul Robustelli:

yeah, I think this can really be a strength of a career in science. having. More outlets to sort of appreciate the varied texture of personalities involved

Milosz:

We often read biographies of past scientists, right? I dunno. Einstein or Fermi or whoever, and they are in those biographies, they're very often quirky and, uh, all the interactions they have with people around them are non-trivial very often. But then this, I think this complexity gets lost in the public perception of a scientist as someone very rigid and proper and, and businessy.

Paul Robustelli:

Yeah, I. That's exactly right. And, and one thing I always try to, to think about, uh, and maybe like a little bit informed from sort of, working in a skyscraper in Midtown for five years is that think it makes more sense almost to view the scientific community, like more of as an artistic community in a way. And, and in some ways like people who devote their lives to studying the motions of molecules have lifestyle wise, like maybe more in common with people who play jazz or people who paint in, in that there is maybe standard paths of success or trajectories in society that, you know, you are interested in. Maybe more self-motivated by your own vision or, or kind of looking for different, and I sort of, when people are talking about doing postdocs, like whether or not you should do a postdoc, I always say this, that we all benefit as scientists to think of ourselves as the most well paid and having the best health insurance of all creative pursuits, right? Like if a painter. Could at age 22 get a stipend in health insurance to paint all day. Like that would be a real privilege where sometimes I think scientists can think of themselves as like the least well paid or the least compensated white collar or professionals. Right? And, and so I, I think when I talk about people, when people ask me about doing a postdoc I almost think of it as a musician, like, this is your chance. To go make an album with like the artist you respect most in the world. And, maybe that's not like the, the shortest path to your most economically gainful scenario, but if this is something you're passionate about and this represents an opportunity to collaborate with people who could be at the top of their craft and if. That experience is meaningful for you? Like almost in the way that like, oh yeah, maybe I won't make that much money on this tour, but I can play backup for this person. Or I can write a song with my musical idol or whatever. Like, I think sometimes that's the, the, uh, right way to think of it. And then in that way, like that makes one of the, the benefits, right? One of the things that jazz musicians like about being jazz musicians is that you hang out with other jazz musicians and not like gray face. Midtown Moneymen riding the commuter train. And so that is something that should be celebrated, that this is a creative pursuit with creative personalities and, and we can talk about metal music and we can talk about avant-garde art and, and whatever it is that, uh,

Milosz:

All right.

Paul Robustelli:

celebrate that diversity and,

Milosz:

I love this point'cause I myself love to hang around artists and I think there's a lot of stuff we can learn from their mindset and like aim for incorporating parts of this creative mindset into what it means to be a scientist yeah, that's a very much.

Paul Robustelli:

I, I had a cool experience, my last year in graduate school at Cambridge. I had a, this shared house, on Jesus Green in Cambridge. Really lucky, like the conservatives of the cam let this house alone on the park fall into disrepair. It was called the, the Lock Keeper's cottage. It was a historically listed building, and we were interviewing people for roommates to fill the last room. And one of them was an art student at Cambridge Art Institute. And you know, and then that sort of circle of it was all PhD students, but we kind of made friends with all these artists and having sort of my eye on that subculture and, you know, I would go visit London and they were all living in these warehouses in Hackney Wick and,, enjoying the fruits of being in an artistic community with, like-minded people. And that was something that I think we can all enjoy as scientists too. And,

Milosz:

let's put it.

Paul Robustelli:

was informative that there was this overlap, I think, and it, like, I think the art students, it was fun for them to, to make friends with PhD students and, and have sort of these conversations across these different worlds. But I think, yeah, there is a similar, uh, creative spirit in both pursuits and, and that should be celebrated.

Milosz:

Right. Happy to put it out there for people to start striving for a more artistic set of values in their lives

Paul Robustelli:

and maybe we could all dress a little cooler, you know?

Milosz:

Yeah, definitely. Let's do that. Okay. So Paul Robustelli, thank you for talking to me and to our listeners. It was a pleasure. you have a great day.

Paul Robustelli:

Yeah. Thanks a lot. This was great.

Thank you for listening. See you in the next episode of Face Space Invaders.