Episode 20 - Rommie Amaro: simulating viruses, cross-disciplinary complexity, and the brain drain

In episode 20, I start by asking Rommie how their research on the SARS-CoV-2 virus first unfolded during the early days of the pandemic, and from this story, we move on to discuss her original motivation to study large complex systems. We touch upon the exciting experimental developments that enable the realistic modeling of systems as big as entire viruses, and highlight some unexpected findings that came out of the large-scale simulations. Rommie also shares her reflections about the collaboration-driven nature of her lab and the complexities involved in interdisciplinary communication, and we wrap up with a few thoughts about the AI-driven brain drain from academia to industry.

Transcript

0:00

Welcome to Phase space invaders. I'm happy to say we got to the 20th interview on the podcast. Today, I'm talking to Rommie Amaro, professor of chemistry and biochemistry at University of California, San Diego. Who works at the intersection of computational biophysics, drug design and microbiology. Her research was particularly impactful in the context of the COVID-19 pandemics. When the Amaro lab made use of their extensive experience. With viral glycoproteins to model and investigate the structural role of glycans in the spike protein, responsible for cell recognition and invasion. After that they kept delivering fascinating results within the field of computational virology, including the model of the entire SARS-CoV-2 virus. So out of curiosity, I initially asked Rommie how this enormous project first unfolded during the early days of the pandemic. And from that story. We moved on to discuss her original motivation to study large complex systems. We touch upon the exciting experimental developments that enable the realistic modeling of systems as big as entire viruses. And we highlight some unexpected findings that came out of the large scale simulations. Rommie also shares her reflections about the collaboration driven nature of her lab. And the complexities involved in interdisciplinary communication. And we wrap up with a few thoughts about the AI driven brain drain from academia to industry. We're again, digressing a lot. I'm sure you're going to enjoy the breadth of Rommie's insights. Let's go.

Milosz: 1:47

Okay. So Rommie Amaro welcome to the podcast.

rommie_1_09-26-2024_121302: 1:50

Hi! Thanks, thanks for having me.

Milosz: 1:53

So Rommie, in recent years, you've been known for your impressive work on large molecular systems. And a large part of that relates to your efforts in understanding the structure of the SARS CoV 2 virus that caused the COVID 19 pandemic. And that, as I understand, was the consequence of your previous research on the influenza and tropical diseases. Can you share the story of how you and your lab turned this, you know, unprecedented global crisis into a nevertheless impressive learning and dissemination opportunity to the point where, uh, the figures from your simulations would make it to the newspaper covers.

rommie_1_09-26-2024_121302: 2:31

sure, yeah, that's a good entree to the discussion, I suppose.. So, gosh, reflecting back right. So my, you know, my group had been working with viral glycoproteins for some time. I actually sort of fell in love with influenza neuraminidase when I was a postdoc with Andy McCammon. I just thought that it was such an interesting, protein with its dynamics and, you know, all that it did. Um, Uh, so continue to study that and also and then, you know, have always been interested in sort of pushing the boundaries of simulation and sort of like trying to get towards complexity. So, naturally evolved to working towards You know, the whole very on for influenza. Right. And then we were sort of, plugging along and getting close to publishing our first big paper on that. And that was right around the time when, when the pandemic hit. It's sort of just, I think it all came about fairly organically. I mean, as soon as we start, I mean, it was just, I remember in December hearing about this, you know, virus, and Wuhan and people already sort of, you know, becoming to it, sort of watching it and then in January and so forth, it getting worse. And then, in beginning of February was when we started thinking, like, you know, maybe we probably should start pivoting towards this. And then the structure came out, from the group of McLellan um. We started working with it immediately. And in fact, like, as soon as the structure, was available, actually, I the director of the Texas advanced compute center. Dan Stanzione, who, you know, it's been a long time colleague, sort of anticipating that we were going to need, um. big compute because it was to be a beastly system and we know most of these are and spike in particular turns out to be large, the largest of the one of the glycoproteins that we've studied. And we studied a whole bunch now. So, yeah, and he was just like, super supportive. From the get go, from the very first correspondence was like, absolutely, we'll, you know, do whatever we can to help you. And I think that, you know, That whole attitude was something that carried through for the first, know, number of years of the pandemic, at least for the first two or three, you know, it's kind of changing now but it was just a remarkable opportunity. And I mean, one of the things that I always say, or that I've said before, academia in general and science, sort of the pursuit of science, has a lot of weird things about it, you know, in terms of the ego and the competition we sort of set up that we construct and just the whole way the whole system is constructed is Toxic in a lot of ways, and and sort of like, what is incentivized and what is disincentivized, uh, even if not directly indirectly and so forth. but what the pandemic really did for folks was, you know, it sort of put us all on in some sense, even footing. I think in some, in some way it like opened people up to, really solving a problem together as opposed to, I think for the most part. And of course, I'm sure there were people who, you know, really, we're still thinking about it, maybe possibly as, a way to just like leg up or something, but, um, I think people really genuinely, for the most part, all the ones I interacted with, I should say, were really genuinely interested in trying to make progress for humanity at that point. And that, that was really special because it disarmed people in a way that, I think I had never experienced before. and all these things, you know, all these incentives and all these reasons why people otherwise sometimes find themselves doing science, these ways that people work solving a problem were then just dismantled and then at its core, what 1 is left with is just. human beings, you know, trying to, do really kind of cool science and to work together. and then not only did you have that attitude, but then it was also the mobilization any resource that you needed, know, because it was really the whole world wanting to move forward. So, know, yeah, we had just a really sort of a remarkable moment in time to like, press on a scientific problem in ways that, I'd never really done before. And so, yeah, I don't know. It was, um, it was crazy at the beginning. It was, I mean, it was exciting, but it was crazy. It was really exhausting. I think that I know for sure that people in my group have PTSD. Um, and me too, and me too, um,

Milosz: 7:32

groups have gone through this. Yes,

rommie_1_09-26-2024_121302: 7:34

groups. Yeah, it's probably true, right? Like, I think we kind of all and are sort of still in some ways kind of like dealing with it or grappling with it, and the change and sort of like moving on. But like, you know, in my group, it's also kind of weird because like, know, we were studying viruses before, and we're just like, genuinely interested in viral glycoproteins. Right? So,, we're so interested in how do these things work? You know, there's so many questions about membrane fusion and intermediate states and there's all this good science, but in some ways to people, you know, you hear people say like, oh, SARS 2, you know, it's over, like, people are really kind of getting tired of it, but, you know, still in terms of the science, it's like, oh my gosh, there's still so many cool things that we don't know. So now we're kind of in that phase, which is also a little weird, like, are we moving on? Are we going in deeper? Like, what are we doing? I guess that sort of strays from your original question, but

Milosz: 8:27

that's okay.

rommie_1_09-26-2024_121302: 8:28

like, Yeah, this amazing kind of confluence of different factors

Milosz: 8:32

Yeah, I think it was an amazing moment, right? Because. People could see both the kind of nascent science of it, right? Because there was no, or very little science of, of the virus,

rommie_1_09-26-2024_121302: 8:43

hmm.

Milosz: 8:44

aside from maybe the previous, virus that was there before and also that's, you know, for virologists, it's such an unprecedented opportunity to trace something on so many levels, right? Like the genetics of it, the structural evolution. So if you happen to be. From that field, like computational virology, which I remember it was a very kind of tiny and cryptic field. Back in 2000, what, 16, 17, I mean, couple years prior to the pandemic. It's like maybe it had a session at the at at the BPS at the conference. Right. But then there were a few people doing that and suddenly trust the field into the mainstream. Right,

rommie_1_09-26-2024_121302: 9:21

Yeah, no, definitely. I mean, it was neat, too, though, like, You know, you had people who maybe, maybe they thought about viral glycoproteins, or maybe they didn't, but that,, there's so many different types of people, even in computational biophysics or computational chemistry or computational biology. However, you want to define us, there are with so many different skill sets, right? Like, the force field developers, the algorithm developers, like, the people who know whatever it is, like, but it was a time where again, people were coming with all these different skill sets, like towards one problem and being like, just super open to collaborating you know, and like the model checking. I mean, when we built even just that part of it, like, when we, when we built the initial system, I mean, it was really a nightmare because it has like. all the protein and some of the protein was missing. And so there was a whole issue with like protonation states and which, you know, cysteines are going to be bridged and which aren't. And maybe it depended on the confirmation of a loop and this and that. And then you had, yeah, all the glycans and everything. um, it was a lot to make sure that it was done correctly. And it really uh, I think we benefited. Tremendously from the input and the care of colleagues, like, for example, Carlos Simmerling, you know, I mean, he was like, looking at our structure, like, at 3 in the morning and saying, like, actually, I think you, I think you missed this 1 position over here, you know, but it was like that from, like, so many different angles and from so many different people and that also just, like, help to advance the science a lot. So, yeah, it was a crazy time.

Milosz: 10:59

right. It's integrative angle was impressive. Right, because there's always this question, I think it's. It was a big question back before the pandemic. What are we actually learning from large systems, right? So, okay, we can build systems, we can run simulations on them, but what is the information we're getting from the simulation, right? That we didn't explicitly put in. And so that's my question to you. how do you see that these days? So what, what are the new things we have learned from running those huge systems,

rommie_1_09-26-2024_121302: 11:27

Yeah, I think that, first of all, that's sort of like an interesting question. It has in, in many instances, I have to say, it's a question that can be almost weaponized against folks who have had an interest in big systems, you know, and that goes back to Klaus and others, you know, where it's like, oh, but what are you really learning and I think actually, there's been a lot that has been learned and I can go through, you know, the details, but I think it's also, you know, and just because I have to do this, because, you know, I went through Klaus's group and I worked with him and, um, he's not here to say this, but, so there's also something to being able to do something just because we have an ability to do it and not necessarily knowing in advance. This is the exact hypothesis. I want to test, but understanding that when you can. Bring things, bring all these different components together, put them in the mix that you're going to learn new things. You're going to have new vantage points. I mean, it's very similar to sort of climbing a mountain. I mean, why do it? I mean, it does. In some sense, you to a new level and gives you an entirely different perspective on problems, you know, that you have maybe been looking at from, you know, one perspective and so I think, I think it is just important to have folks around who, and I, it does take all types. So I'm not saying like everybody should do big systems, but I do think the pursuit and study of ever larger and more complex biological systems is something that, important and sort of should always have a place and hopefully, you know, is appreciated by folks. But I can say now too with much more confidence as we know more about structural biology and about proteins and molecular machines is that, the dynamics that these systems undergo in situ in cells is different than what they can study in the bench in vitro. It's the stuff, all the single particle cryo EM. It's amazing, but is that landscape, are those conformational dynamics the same as what's actually happening when it's in the ER or whatever? I mean, actually pretty sure it's not. it gives you a lot of, you know, it's a wonderful model system. It's really important for understanding of structure. There's all these amazing advances, right? But putting it into context. is critical for actually understanding true biology.

Milosz: 13:58

Right. It's interesting that you bring it up because I think there's this kind of, Well, I, I hesitate to call it a paradigm in science, right? But there's this other mode of approaching science, which is by observation of something that's a complex system. And this is complexity. Science is becoming a proper branch of science these days, I think. So like, I think what you're advocating for is, is exactly this, right? So put together something that's complex our simplistic imagination, and then observe it with a different toolbox. Without the prior hypothesis or prior expectation of what it's going to do.

rommie_1_09-26-2024_121302: 14:36

Yeah. I mean, I think that's certainly a large part of it. And, you know, the funny thing is just again, how the whole scientific system is constructed. I mean Like, review panels, you know, we tend to be more narrow. We tend to not want to do the edgy kind of high risk stuff. What are they going to learn? it really going to be worth it? I mean, but now, you know, looking back I don't know. It seems to me like, yeah, it definitely is worth it. Um, you know, all the work that we did to construct, for example, influenza, and before that, that Klaus had been doing, you know, at the all atom level, and there were certainly others, you know, I shouldn't. Just, like, focus on him and my efforts, you know, understanding, of course, that we're part of a very large community, none of this would be possible without all the different

Milosz: 15:16

of course.

rommie_1_09-26-2024_121302: 15:17

that make it work, um, know, who are deeply appreciated for, just, that all that work that had been done with HIV influenza, it was because of all that, that took a decade. develop the tools where people really questioned, you know, and like the stuff with influenza, we had real trouble, even getting the compute time for that. And it wasn't until basically said, you know, he kind of took me under his wing and he said, uh, you know, Hey Rommie, I believe in what you're doing. I've been fighting these fights for years and I'm going to give you some of my discretionary time. this big fancy machine we have here at NCSA. It's called Blue Waters. you know, you can't have all my time because I needed to study this and that with HIV. We'll work with you. We'll support you. And that was like, just super critical to getting the preliminary Compute that, you know, allowed us to develop the tools and stuff and whatever and by the way, we did learn amazing things about influenza, which I'm happy to chat about. And it does have to do with dynamics of these glycoproteins that we don't sample necessarily, or in the same of, states, um, in a crowded environment

Milosz: 16:28

Sure. We can dive into that.

rommie_1_09-26-2024_121302: 16:30

Yeah. Versus the, versus the sort of dilute solution type stuff. Um, but it was all of that, that allowed us to, you know, be ready to go with SARS-2. And so, you know, a similar system that had taken us 8 years to actually, like, build and get to production dynamics. We literally did it in, God, we, it was running by, it was running by November, so, so it was like, what, nine months? We did it in eight or nine months? I mean I don't know.

Milosz: 17:02

Yeah. sounds like part of serendipity, right? That you can be lucky, but only if you're well prepared from prior, the prior work you've done. Right. So let's talk about what you learned about influenza. Cause that's a great case for those big systems, right. To be made

rommie_1_09-26-2024_121302: 17:20

yeah. I mean, the, what we learn and which I think we mostly already know, but the issue is that when it comes to structure. I think the issue is that when we come to structure, we are just only now developing the capabilities to actually see what happens to these systems. Again, in situ with all of the complex interactions of the environment, and even then it's still static. Okay. And I'm thinking of, like, you know, cryo electron tomography, for example, with the milling or where you can do some, tomogram averaging, on these systems that, are really sort of in cells. but anyway, so what we learned with the flu, you know, and so, and my group had studied and others. For years, a decade, had studied these hemagglutinin systems. so hemagglutinin is the class one fusion protein for influenza. It's a trimer. It has a lot of similarities to the SARS 2 spike protein, to HIV envelope protein. you name it and, you know, it's interesting because it's, a target of vaccines and so forth. on top of being, mechanistically interesting, but anyway, so we had studied these systems for years and looked at, I don't know, binding of various groups and so forth but when we put it in the context of the crowded, Virion and, you know, all of these viruses, they have different propensity, different, like, numbers and density of the glycoproteins. So, like, SARS 2 has maybe about 40. HIV has 10 of these glycoproteins. Influenza is like the opposite. It's super packed. Like, it looks like super, super crowded. It's got like 200 and some hemagglutinins

Milosz: 18:59

Well,

rommie_1_09-26-2024_121302: 19:00

like another. Yeah, it's got a bunch. Like, these guys can basically hardly move. It looks like on the surface. So we built it, got it all in there, simulated it. And we, what we saw was just like stuff we, we were seeing motions that we had never seen before in these, like in all of the many, many, you know, probably now hundreds of microseconds of dynamics of single particle dynamics. We never saw these types of motions and these types of motions, what are they? They're like a really kind of more extreme tilting that happens relative to the membrane normal, and then there's also breathing that happens and the breathing, it turns out, you know, has to do. I mean, so I think these molecules, as we know, molecules, they do actually breathe on their own. But when they have, you know, if they're just like, again, in dilute solutions, of course, you're going to have these different intermediate states, but the structurally resolved state, the 1 that they could get to the high resolution was always the tidy. Upright.

Milosz: 19:59

yeah.

rommie_1_09-26-2024_121302: 19:59

packed, trimer. And so I think that colored how people, that data, that was just like such a beautiful picture. You know, it totally biases scientists About everything about how they think about it, about how they interpret, how it's interacting with things. But it turns out that, it's just much more dynamic than that. And when you put it in the crowded environment, we actually see sort of very quickly evidence of this additional dynamics, including tilting, but also a breathing open the protein itself. And in the case of. hemagglutinin, influenza hemagglutinin. What was interesting about that is that when it breathes open, there's a quote unquote cryptic epitope, you know, cryptic being a fun word for things discovered through simulation that weren't there in an experiment, a hidden epitope inside Of hemagglutinin that, actually could be a viable strategy for neutralization, but that, sort of really was a curious finding, because they weren't sure, like, well, how does it really get access to that? And so forth. And it turns out it's part of it's sort of like ordinary breathing,

Milosz: 21:05

I see.

rommie_1_09-26-2024_121302: 21:06

and which was actually substantiated by experiment also, but at

Milosz: 21:09

Do

rommie_1_09-26-2024_121302: 21:10

where you can just see sort of like the blobs opening. Okay.

Milosz: 21:14

you also have knowledge of the higher level? Yes. I recently had a conversation about like, do we know what happens to a virus when it dries up, let's say you have a, I don't know, a droplet, right. And it evaporates.

rommie_1_09-26-2024_121302: 21:26

like in,

Milosz: 21:27

Why does, why does a virus?

rommie_1_09-26-2024_121302: 21:29

interesting

Milosz: 21:29

Yeah. Why does a virus get deactivated?

rommie_1_09-26-2024_121302: 21:33

Yeah. So, I mean, that's a question that we're trying to understand and why, and we don't have, I mean, we have hypotheses only because we have these big simulations, but you know, we're still trying to understand that and I think we're getting closer, but, I don't want to spill any additional beans, but it looks like, I mean, I think things that I've said before, know, it does look like for stars to that, in these, respiratory aerosols, right? Which are pretty much the predominant mode of transmission for this virus, and that whole sidebar of interest is like, why are we so reluctant to admit that it is airborne? And, it's just. Mind boggling to me, but, in any case, understanding the physical chemistry questions of what's happening to these biological things in aerosols. That's like a whole frontier field that is just beginning to get cracked open. And one of the also kind of weirdly serendipitous sort of things was in 2018 at UCSD, where I'm a faculty member, there was this really big NSF center. It's called CAICE, center for aerosol impacts on chemistry of the environment, which was studying sea spray aerosols. So, you know, like, when we're really close to the ocean, gratefully, it's beautiful here. And the ocean, the waves crash, and then like small bits of the ocean actually become aerosolized. And these bits of ocean, they're not just like salty water, right? Because the ocean is like filled with all sorts of critters and so many biological things. So there's all sorts of things that actually do like, Get transferred from the bulk ocean into the atmosphere, and they can do things like seed clouds react with atmospheric gases. But so we had studied separately ways to use the quote unquote computational microscope. In other words, these really large. larger scale MD simulations to study what happens inside these aerosols with this sort of like complex mixture of different kinds of fatty acids and ions and lipids and all manner of things. So, when COVID hit, we were not only studying, you know, the bulk of my group is studying, like, Biomedical research type questions viruses. We also work in cancer. You know, I have a whole drug discovery branch in the group. But there was a small bit, a few researchers who were working on aerosols, the sea spray aerosols, which were kind of viewed as being disconnected from health. And,

Milosz: 24:11

Right.

rommie_1_09-26-2024_121302: 24:12

that gave us the tools then to converge respiratory aerosol chemistry with viral chemistry, which turned out to be turning out. Of course, it's like, really critical for stars, too, but also many other pathogens. So yeah, that gave us a platform to develop our, like, uber crazy, insane, big simulations that were, like, over a billion atoms. And we're just. Crazy, you know, where we had all the complexity of the respiratory aerosol that has, like, mucins and different types of things coming from lung surfactant and all that,

Milosz: 24:45

That's an impressive, coincidence I was just thinking if you use coarse grained models, how you simulate evaporation. But you're saying you're actually doing this at a fully atomistic level

rommie_1_09-26-2024_121302: 24:56

Yeah. So we did it, evaporation is, of course, interesting. So we generally tend to assume a type of relative humidity, which gives us

Milosz: 25:06

stability. Okay.

rommie_1_09-26-2024_121302: 25:08

Yes, a local stability of the aerosol, you know, and there have been folks who have developed fully atomic sort of ways to look at evaporation. And I should also say That in these, so we have the aerosol, which is like, all packed, but then it's also, we're not necessarily using periodic boundary conditions per se. Like, there

Milosz: 25:30

Right.

rommie_1_09-26-2024_121302: 25:30

water interface in this box and you do see escape. Of water molecules. but it's not, but it's, it's, it's a minor, minor fraction. And so it's, you know, it's not actually evaporating, in the sense that,

Milosz: 25:42

I see. So you're

rommie_1_09-26-2024_121302: 25:43

dramatically

Milosz: 25:44

you're mostly looking at the partitioning between the bulk and the interface.

rommie_1_09-26-2024_121302: 25:49

Yeah.

Milosz: 25:50

In

rommie_1_09-26-2024_121302: 25:50

at, morphology, what's going to the surface and then in particular also, you know, so for questions of sea spray aerosol, we're interested in the reactions that are happening on the surface

Milosz: 26:01

uh,

rommie_1_09-26-2024_121302: 26:01

aerosol particles or micro droplets. is interesting because the chemistry

Milosz: 26:05

yeah,

rommie_1_09-26-2024_121302: 26:06

changes at the surface versus in the bulk reactions get dramatically accelerated in some cases, or like, you know, for various real sort of detailed chemistry reasons you know, in the case of aerobiology, which is maybe how I would call what we're doing now with SARS 2 in these droplets and aerosols, it's more, you know, the questions are more about what is happening to the structure of this thing, because the, these are fairly small. Particles, right? They're basically on the same order of magnitude size of the virus itself. And we know from structural biology that at the air water interface structure will be degraded. It will be warped or whatever. Right?

Milosz: 26:49

yeah. I

rommie_1_09-26-2024_121302: 26:50

word right now, but don't get messed up. So there must be something protecting the virus in flight. or at least some proportion of the viruses enough that, you know, they actually can remain infectious. And so these are the kinds of questions that we're curious about what's happening in aerosols. Anyway,

Milosz: 27:06

have people studied those droplets like aerosol particles with, is this even studyable with let's say in situ cryo EM.

rommie_1_09-26-2024_121302: 27:15

I know that's something that well, so there, there has been 1 study actually also coming out of UCSD, where they took a sea spray aerosol, and they trapped it and were able to visualize it. This was sort of a, I think, 1 of the larger. It was probably more the size of a droplet, not an aerosol. So aerosols are like the little ones that are, you know, they have to be about a micron or

Milosz: 27:37

yeah.

rommie_1_09-26-2024_121302: 27:38

in size and smaller and then beyond that are droplets and those guys are bigger and then they fall. They fall off, but it's aerosols that keep floating anyway so there are folks who are, we're trying now to develop techniques to trap and visualize these, actually using tomography, but that's still something that needs, uh, sort of methods development.

Milosz: 28:00

Yeah. I was going to ask, is there any kind of secret experimental source that you can use these days beyond. Cryo-EM tomography that helps you study those things. Like, what are the latest improvements in experimental techniques that make it doable?

rommie_1_09-26-2024_121302: 28:16

yeah, there's, um, for you mean, for aerosol chemistry in general,

Milosz: 28:21

Well, it's virology broadly. Yes.

rommie_1_09-26-2024_121302: 28:24

I think 1 of the most exciting places right now to do this type of research is actually at the University of Bristol. There's this there, Jonathan Reid, has been working in aerosols for a long time and they have ways to basically using something called like an electrodynamic balance, they can basically like trap individual aerosols and then like interrogate them under different

Milosz: 28:46

Oh, wow.

rommie_1_09-26-2024_121302: 28:48

Yeah. And what's really to me is some, I mean, the technique is really cool you know, it's, it's an amazing experimental technique, very unique.

Milosz: 28:55

So they're balancing the

rommie_1_09-26-2024_121302: 28:56

also,

Milosz: 28:57

with

rommie_1_09-26-2024_121302: 28:58

yeah.

Milosz: 28:59

fields. Right? Yeah. Okay.

rommie_1_09-26-2024_121302: 29:03

Yeah. and you can make this sort of like almost direct connection. They can't see structure per se, but they can look at so many other things. And to me, that's also exciting because I think, We don't always have to see the structure necessarily, right? This is where computing actually, and the computational biophysics actually should be really complimentary and useful, to sort of poke at and go into places where experiment can't go. That's some of my favorite things to do. And, you know, maybe the experiment will catch up later. Maybe we'll be right. Maybe we'll be wrong. But like, our hypotheses stand up to different types of experiments and, you know, advance our understanding in the now. And I think we're able to do that. But anyway, yeah, those methods in particular, being developed by the Reed group and so forth are really cool. And the other thing I want to say is that they've done a really good job at intersecting the disciplines required to make those successful experiments. Because I'll tell you what, one of the, I think the hardest thing about science. not, and here's getting a little bit on a soapbox, like, what do you want to talk about? That's not necessarily science, but, like, I'll say this, you know, I don't think it's like, it's like the technology development. That's like, the hardest part. I think it's like the people. That is the hardest part of science and I can come in different ways, but I think. Like, when we were starting to work with aerosols, we stepped into, you know, we were always working in computational biophysics and looking at enzymes and proteins in cells, blah, blah, blah. then we go over to this other part of the building. we're starting to work with really large groups of really smart people who have a completely different vocabulary and are coming from a completely different, like, view of systems. So, like, I say substrate, and I'm thinking of 1 thing, and they say substrate, and they're thinking of something completely different. I say residue, you know, there's just like, it took us two years to even be able really to have the kind of conversations where can go back and forth and you can really understand each other and you can trust each other enough to do the experiment. And, you know, and that whole aspect that's. That's actually really hard. And it's also hard to just have that in a science session. It's even harder to have that in a review panel. So what happens is science becomes siloed into these, like, niche groups where everybody can understand each other. But then, then you do have people and groups who are working at these interfaces. I personally think, and many others will agree with me, know, that these interfaces of science is where it's like, Okay. Is where all the interesting problems are, you know, it's where the new solutions are going to happen and, and where we can actually, you know, address sort of frontier things. But it's like, it requires really hard work just from the people side.

Milosz: 31:55

No, I absolutely agree

rommie_1_09-26-2024_121302: 31:57

oh,

Milosz: 31:57

is,

rommie_1_09-26-2024_121302: 31:58

Good.

Milosz: 31:58

is there is a case to be made for, you know, like a transformer style translator between subfields of science, right?

rommie_1_09-26-2024_121302: 32:05

Yeah.

Milosz: 32:05

would see,

rommie_1_09-26-2024_121302: 32:06

So

Milosz: 32:06

would see what the other, you would see what the other person is saying in your language, like computational biophysics language in real time.

rommie_1_09-26-2024_121302: 32:14

Like, yeah, like, well, had some kind of software that would like, so, yeah, real time

Milosz: 32:18

Yeah. I think that will be doable

rommie_1_09-26-2024_121302: 32:20

my. Yeah, maybe like a decade from now or so.

Milosz: 32:24

now. I've experienced it myself. Yeah. So many times that after many, many meetings you spend another two hours, it's like. Going through your notes and trying to see what these words mean in that. Community, right? Okay. Talking to AI people or talking to visualization people or talking to, well, proper experimental people. Yeah, absolutely.

rommie_1_09-26-2024_121302: 32:43

Yeah. And

Milosz: 32:45

Someone should work on this,

rommie_1_09-26-2024_121302: 32:48

they should, they should. There's so much, there's so much good things to be done, you

Milosz: 32:53

but on the people side, yeah, I actually asked people, you know, Oh, what should I ask Rommie when I talk to her? And a few of them, yeah, I was, I'm always curious about people's inputs and a few of them pointed out, and I was similarly curious about it. How does it happen that so many of your group members, gain so much visibility, right? I'm talking to, let's say social media, because people like Fiona, uh, Kearns, so like Lorenzo, like Mia Rosenfeld, like Mohammed Shehata, I've, I've known these people from Twitter, for example, right. And then I see, Oh, they actually connected to your group. do they just get inspired from you or do you select people who have this kind of personality that they're, outgoing or,

rommie_1_09-26-2024_121302: 33:32

Hmm.

Milosz: 33:33

because it's interesting that

rommie_1_09-26-2024_121302: 33:34

question.

Milosz: 33:34

it happens a lot, apparently

rommie_1_09-26-2024_121302: 33:36

Um, well, I think I'll I mean, I don't think everybody is necessarily the same. I Fiona had like a big following and was super. She's super active on social media stuff. I think like from well before I only joined social media. Type stuff, um, in 2020, randomly also, right before the pandemic. right, and yeah, and then there's like Mo and he has like a whole YouTube thing. I don't know, are they drawn to my group? I don't know. I mean, Lorenzo joined before I mean, he might've been on a social media thing. I don't know. So this is an interesting question cause I've never thought about

Milosz: 34:10

it's not a lot of policy that everyone gets social media training is

rommie_1_09-26-2024_121302: 34:14

God, no, no. And I

Milosz: 34:15

coincidental.

rommie_1_09-26-2024_121302: 34:16

just, no, I'm hanging on to social media by a thread because like this whole transition from

Milosz: 34:22

Oh, yes.

rommie_1_09-26-2024_121302: 34:23

to X has been a nightmare and I've become kind of like just less active on the platform. But, I think that for sure in my group, we do a bit of methods development, but we, we do have the bulk of our projects require extensive teamwork and not just like a collaboration with a group, but like, if you're working on a project in my group, it's very likely that you're going to have 6 different collaborators with a different experimental techniques converging on the problem that you're studying. And so you have to be able to be, able to work in teams. And that means, in some sense, usually very good communicator kind of probably a little bit more on the assertive side., and maybe that's why it's sort of also trends that way, uh, but

Milosz: 35:18

Okay,

rommie_1_09-26-2024_121302: 35:18

too, I'm sure it it attracts a certain like type or kind of person. maybe that's like

Milosz: 35:23

the chicken and egg. Yes, I see. But yeah, that's a great point. Point because it's not always that common, right? Putting people in this, collaborative environments. Not all groups do that all the time. So

rommie_1_09-26-2024_121302: 35:34

Definitely not and definitely not academic groups and definitely not chemist academic groups., yes, it can happen, but, um, I feel like I'm like, my group is a little bit, like, a little bit different than, um, some of the other groups that you would find in a traditional, say, chemistry department where the questions are, maybe a bit more tightly focused,, we sort of bleed across from small molecules to proteins to larger systems now. And so there's, you know, there's interplay, um,

Milosz: 36:06

Yes.

rommie_1_09-26-2024_121302: 36:07

with other groups.

Milosz: 36:08

More room for exploration as well for individual.

rommie_1_09-26-2024_121302: 36:11

I have to say I also have, definitely have people in the group who you've never heard of. Ha ha!

Milosz: 36:16

Yeah, I guess

rommie_1_09-26-2024_121302: 36:17

have, like, zero presence,

Milosz: 36:19

I can imagine.

rommie_1_09-26-2024_121302: 36:21

yeah, and,

Milosz: 36:22

you have a big group, so I don't know everyone for sure.

rommie_1_09-26-2024_121302: 36:24

yeah, and it's not like, it's, it's not like everybody is the same, for sure. I mean, we have, one of the things that I've always taken a lot of pride in is the diversity of my group I sort of was raised thinking that diversity is a huge strength. and it doesn't matter if you're a sports team, or if you're a family, or if you're a lab, you know, having different perspectives is like, 1 of the most and I think beautiful and interesting parts of life I think, I think the Amaro Lab is a place where you, you find that type of diversity, whether it's geographic, whether it's religious, whether it's even just like kind of almost also like what interests you, you know, I have people who are like really interested in something very deep and narrow. And then others who are like, Oh my God, how does this amazing thing work in this big sea of complexity? So I don't know.

Milosz: 37:18

Yeah, many voices. That's

rommie_1_09-26-2024_121302: 37:20

group

Milosz: 37:21

that coincides with what Syma said before on the podcast. So I'm happy to

rommie_1_09-26-2024_121302: 37:25

that

Milosz: 37:25

to spread. Yes, to say it to spread this message,

rommie_1_09-26-2024_121302: 37:29

yeah, she's, she's someone, you know, it's funny you know, there's also all types of colleagues, right? Syma is someone who, you know, who I've enjoyed meeting and who I think we share a bit of a kindred spirit. You know, you, you sort of like, you run into folks and you feel like sometimes like either I've known you a long time or somehow, you know, like you, Just really kind of get on with them very well. She's certainly one of those folks. Yeah, she's a keeper. She's a gem. Yeah.

Milosz: 37:57

beautiful. So, do you have any final thoughts you wanted to share with us about the drug discovery side of your recent.

rommie_1_09-26-2024_121302: 38:05

yeah, I have to say, I want to stay kind of positive because you know, I've worked for many years at the intersection of simulation and sort of drug discovery and design, you know, had sort of my first part of my career was really all about methods and applications You know, connecting structures from simulations, with, small molecule screening and so forth, started a company, have a lot of patents, I think it's an interesting moment, I think, for. This field sort of looking at the small molecule side with like this AI wave, I think one of the things that I find unsettling. I think I can use that word or that. Sort of sometimes maybe keeps me up at night. I don't know. Maybe it shouldn't bother me. But, like, a number of my colleagues have sort of gone the route of, leaving academia you know, and trying to run in the, In the herd with all the AI venture startup and all this, and I think that that's, you know, normal and good. I also worry a little bit about, the academic side and about methods development and about You know, really investing in problems where the solution is not necessarily right around the corner, but that requires, you know, a real kind of slow chipping away, chipping away, chipping away. I mean, I, there are definitely problems, you know, like that. And with the sort of the money being thrown around it's, you know, it's attractive to jump. I get that and also, It's exacerbated, I think, because in academia, there's been a bit of maybe a tightening, especially, particularly, maybe towards method developers. I don't know. You know, like, it's almost and that's what Vijay was saying. We had a little discourse on X or Twitter that it's like,

Milosz: 39:56

you mean Vijay Pande?

rommie_1_09-26-2024_121302: 39:57

can just, you know, Yeah, Vijay Pande, sorry.

Milosz: 39:59

Yeah,

rommie_1_09-26-2024_121302: 40:00

know that you can, know, you can do everything that he kind of said, you know, everything you think you're doing in academia, you can do at 10x. Or 100x pace here, you know, you know, like you should consider it as a jump that you should make. And, know, I don't know, I think that would be maybe an interesting subject for like a group podcast. You know, you could get a few of the folks who've jumped and a few of the

Milosz: 40:22

I see.

rommie_1_09-26-2024_121302: 40:22

behind and really discuss like you know, what people are thinking about the future and, you know, training of people and that whole interplay, but right now, you know, there is an enormous amount of, of hype, but also probably there, you know, there's hype because in principle, you know, this machine learning and AI together with amazing data sets should absolutely transform biology and chemicals, small molecule discovery, definitely as we know it, but, you know, we got to, there's a lot of work we got to do to get there.

Milosz: 40:53

Oh, yes, absolutely. We have actually, we had a conversation with Paul Robustelli about a similar topic because he was switching back and forth between the industry and academia. His point was that, yeah, it's good to have both because it's like you can keep the best of two worlds, but you don't need to fully commit to the advantages or disadvantages of either.

rommie_1_09-26-2024_121302: 41:12

absolutely.

Milosz: 41:14

the more, the more we can make this barrier kind of available for people to pass through both ways, the better for, for the field in the end, perhaps that. was his take.

rommie_1_09-26-2024_121302: 41:24

to agree with that. Yeah, I mean, I do tend to agree with that. I think similarly to like this principle that diversity is a strength, you know, I, it plays to that sort of same theme. I do think that I guess sometimes I, worried like if too many people jump, like we don't want to just like a mass exodus of all of the. Intelligence in some sense, you know, in a particular field out of academia, in particular, also, because institutions move really slowly. So if you have an exodus, it's areas that have taken kind of years to build up it takes a, it takes a while to

Milosz: 42:00

Oh, yes.

rommie_1_09-26-2024_121302: 42:00

it, it takes a while to get it back. And so, I don't know.

Milosz: 42:04

So many smart people, so many smart people who take their very field specific knowledge to,, intellectual property protected environments I get it.

rommie_1_09-26-2024_121302: 42:14

different. Anyway.

Milosz: 42:17

Well, I hope someone thinks deeply about these issues, and as you say, maybe it's a good moment, a good thing to do to think about a panel or a broader discussion about it. Okay.

rommie_1_09-26-2024_121302: 42:29

yeah. Yeah, Yeah,

Milosz: 42:31

So anyway, thank you for, the thoughts, for the story, for all your insights. Uh, Rommie Amaro, it was a pleasure to talk to you

rommie_1_09-26-2024_121302: 42:38

this was really fun. Thanks for having me.

Milosz: 42:40

and have a great day. Bye.

rommie_1_09-26-2024_121302: 42:42

You too.

42:44

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