Episode 21 - Tamar Schlick: A mathematician's path to biology, RNA frameshifting, and why scientists (should!) run

In episode 21, Tamar first explains how her unique background impacted the way she approached and conceptualized problems in biology, and how her research projects were born in the first place. We talk about the more recent research coming from Tamar's group on frameshifting, a fascinating process by which the readout of the genetic code can be offset by one or two letters to produce multiple proteins from a single viral gene. We then move on to discuss whether it's algorithms or compute that have contributed more to the growth of computational biophysics. Tamar's textbook and the multiple perspective articles she's written over the years serve us to dwell on the importance of writing for clarity and interdisciplinary collaboration. We end on a people-centric note, talking about the social bonds involved in both running and experimental collaborations.

Transcript

Milosz: 0:01

Welcome to the Phase space invaders podcast. We're back with episode number 21. And my guest today is Tamar Schlick, professor of chemistry, mathematics, and computer science at New York university. The countless contributions Tamar made to computational biophysics come from a number of directions, including large scale modeling of chromatin structure. development of algorithms for molecular dynamics, free energy methods, or a graph theory based classification of RNA molecules. Originally a mathematician, Tamar applied her unstoppable transdisciplinary curiosity to the most fascinating questions of the last decades, now centered around the functional roles of RNA in viral replication. So we start by discussing how Tamar's unique background impacted the way she approached and conceptualized problems in biology. how her research projects were born in the first place. We talk about the more recent research coming from Tamar's group on frameshifting, a fascinating process by which the readout of the genetic code can be offset by one or two letters to produce multiple proteins from a single viral gene. We then move on to discuss whether it's algorithms or compute that have contributed more to the growth of computational biophysics. Tamar's textbook and the multiple perspective articles she's written over the years serve us to dwell on the importance of writing for clarity and interdisciplinary collaboration. And we end on the people centric note, talking about the social bonds involved in both running and experimental collaborations. Hope by now you're anxious to hear more. So let's go. Tamar Schlick, welcome to the podcast.

Tamar Schlick: 1:55

Thank you. Great to be here.

Milosz: 1:59

So most of my guests, and, perhaps it's true of most of biophysics, but they eventually admit to being physicists by training, right? And of course, all physicists are mathematicians to some extent, but since your formal background is in applied mathematics, you know, I find this quote, you brought up once from Sydney Brenner, that mathematics is the art of the perfect and physics is the art of the optimal while biology, according to him, is the art of the satisfactory. Do you think there's a difference in how you perceive biomolecules compared to people with an explicit physics background?

Tamar Schlick: 2:32

I think we all approach problems differently based on our training. I think, having a math background, especially applied math, and that was my, my first degree and, Ph. D. degree was, um, were both in applied math, and it gave me a really different perspective, not just to appreciate, concepts in mathematics, but the way a mathematician thinks, approaches a problem. And you want to collect a lot of information, think deeply about a problem, but approach it, especially a very practical way as an applied mathematician, we have a saying as applied mathematician that, pure mathematicians solve ideal problems exactly, while applied mathematicians solve real problems approximately. So the fun is in the approximation. You have to, in order to solve real problems, you have to make approximations, you have to simplify a lot of, issues, parameters, you have, but the key is to find the right the key variables that will allow you to really get to the heart of the problem in biophysics. So I think coming at problems in biophysics from a math angle certainly gave me some disadvantage that I wasn't as familiar with all the biology and that people from that field know, but it gave me a different perspective and also community of people that are familiar with the algorithms and the latest mathematical research, which was very advantageous when I started to work on algorithms. But, clearly now our field of computational biophysics is It's highly multidisciplinary, and there are people from all fields, mathematics, computer science, and certainly as AI enters more into it, chemistry, biology, physics, engineering, bioengineering. I've had lots of postdocs that come from bioengineering. They were the most. resourceful in terms of developing methods and applications. So there's great synergy now, and I think it's very common to see computational biology labs with an amalgam of people, from all kinds of backgrounds. And that, makes it more fun and I think more productive.

Milosz: 5:07

Right. think we still tend to think of those fields in the kind of ladder like way, right? With mathematics, perhaps on the bottom, and then other fields building on top of that. Some people would take

Tamar Schlick: 5:19

the tree of knowledge. Yes.

Milosz: 5:21

and then the distance between mathematics and biology is, quite sizable, right? So. I think a lot of mathematicians start with this assumption that, oh, biology, you know, there's just so much empirical facts that you have to incorporate that is so unpure. And did you ever deal with such questions of like, why is there so many facts that we have to memorize or so many, accidental aspects of it rather than this pure abstract and nice concepts?

Tamar Schlick: 5:50

Well, when, I started my work, in the Ph. D. and I wanted to develop a project, my advisor asked me, what would you like to work on, and I said, well, I want to work on DNA. I've always been interested and I just found the simplicity and complexity in DNA, the ladder like structure, but, many levels of folding. So you have from the simple, boring molecule that may appear in Bio 101 to epigenetic regulation and an effect on human disease. So, as I learned more about DNA, I thought that that was just a beautiful problem to work on and you, you learn about the folding, how the DNA needs to fit in the cell that's many orders of magnitude. smaller than the size of DNA when stretched out. you think of knotting and topology and there's so many exciting areas of mathematics that can be approached to the problem. So, that's how I got started in the field and I learned everything about DNA that I could, actually from not just textbooks because textbooks and articles can only take you so far but talking to people I had co advisors, in the biology department and also math and computer science. And my advisor was very, very good about, who are the leading people in the field? Let's invite them for seminars at Courant. Uh, you know, that's the NYU

Milosz: 7:24

Right. Okay.

Tamar Schlick: 7:25

Institute of Math. we did that. And that's how we met, uh, Scheraga and Karplus and that got me to do the postdoc with Shneior Lifson at the Weizmann Institute in Israel. um, I remember Sharaga coming and telling us about the multiple minimum problem, and that was,

Milosz: 7:46

Hmm.

Tamar Schlick: 7:47

to, you know, to us, and that's how we brought the computer scientist in Courant and I worked on this truncated Newton method, which got me in the computational chemistry community, so, that's how I think a mathematician gets into the problems in biology, and, you get interested in a problem, you go to conferences. I remember Gordon Conference in Computational Chemistry was a great way to meet people working in the field and in a very informal way and I was just asked tons of questions. And so That is, I think, how I entered the field and of course, I stayed working on DNA from modeling DNA to supercoiled DNA dynamics, which started later with Olson and then into chromatin and epigenetic control, which we're very interested in now.

Milosz: 8:51

Right. So it's wonderful if we have this shared vocabulary that we can translate biological problems right into things that mathematicians recognize as problems they can work with I can see. And you keep working on those, questions now also with RNA.

Tamar Schlick: 9:06

Yes, and RNA has also provided a great opportunity for application of mathematical tools. so we had the idea about 20 years ago of applying graph theory to represent, enumerate, and classify RNAs. Now, of course, this was started, in the 1970s. Other mathematicians And, biologists realized that secondary structure of RNA could be represented by graphs. But, It was mostly used to represent 2D structures, and we had the idea of why can't we go beyond that, beyond representation to actually use graph enumeration, results in graph theory to enumerate the motif universe of RNAs. And that, imagine if we could do that for a protein, so it's not possible, but for RNA you can actually up to a certain approximation, enumerate that motif universe and learn a lot of things. what fraction of that possible motif universe is actually known, you know, solved in nature. So that's how we developed the RAG RNAs graph approach. And turns out that graph theory is very, very useful, not just to enumerate and describe these RNA structures, but also to analyze and design, RNAs. And that, that comes from clustering and other, I won't go into too much technical, details. that's in our papers, but, all this can be very useful when you're interested in specific RNAs. So, when the pandemic hit, we had all these tools available. We had just been working on what's called dual graphs, which are. ways to represent RNAs with pseudonots, those are RNAs with intertwined,

Milosz: 11:12

Right.

Tamar Schlick: 11:12

base pair segments. So that turned out to be crucial for the frameshifting element of SARS CoV 2, the virus of COVID 19 and with support of a rapid grant from the National Science Foundation, we launched into this project and we've been working on it for the past few years. And that has been just a gold mine of opportunity for, Advancing the structure, understanding the mechanisms of frameshifting, developing mutations. We work with experimental collaborators and they have, shown how effective our mutant designs are for abolishing frameshifting. So, I think this is, um, of a great way to advance, viral RNA studies and, open new therapeutic avenues.

Milosz: 12:10

I see. So What would be the biggest strength of this approach? Would you say it's functional prediction or there's some conservation of topology that is evolutionarily preserved or what was perhaps the most surprising thing that came out of this

Tamar Schlick: 12:24

yeah, so what is very intriguing about these frame shifting elements? They're very small RNAs. But they have many, uh, folds. So, not just a single fold. And, of course, it depends on the length. So, as you change the relevant length, and that's the length of the mRNA transcript changes depending on where the ribosome is coming along. And, um, These alternative folds exist, biologists know about them, but their role in the frame shifting mechanism is not well understood. And usually from experiment, we have cryo EM structures, we have X ray crystal structures of the frame shifting element of SARS CoV 2, and it's thought, it's always found to be this, one structure, which is a pseudoknot, a 3-6 pseudoknot, in our dual graph notation. it turns out, other chemical experiments, DMS and SHAPE, show a variety of structures. So using, our graph approaches for mutational design, it was very easy to design mutants. our idea was, what if we change that pseudonot to something else, or we eliminate the ability of this RNA to fold into multiple folds, in other words, have transitions? What if we block those transitions? So we designed, three particular mutants that would lock the frame shifting element at one fold. and Our experimental collaborator tested those. This is a paper now that appeared in the RNA journal. And they showed nearly abolished frameshifting for those mutants. So I think we're beginning to understand that these transitions are very important for frameshifting. And this could potentially with, RNA editing techniques, maybe in combination with small molecules, could lead to new therapeutic avenues. And I think the interesting thing about working on, the RNA of viruses is that we're going to the heart, of the infection rather than the viral proteins, which are the products so if we could stop the virus from entering or replicating, that Could go a very far way. And of course, many, many viruses have this frameshifting element, not just SARS CoV 2 and other coronaviruses, the flu virus, HIV, and many others. So this is a very important mechanism that's also relevant to human genes, by the way. this is an exciting area, and I'm sure we will learn a lot more as we continue this exploration.

Milosz: 15:22

sounds beautiful. I think the conformational heterogeneity of RNA will keep us employed for some time. And it's great to have better tools to describe that or to approach that, right? such a plentiful such a great field to

Tamar Schlick: 15:39

Yeah. And, of course, DNA continues to amaze us, too, with a genome structure and function. And there, there's so many interesting things. areas to explore using models at different resolutions. So we've been working on a nucleosome resolution model on the kilobase level so we can understand the effect of epigenetic marks on the folding of, the chromatin fiber and its effect on higher order folding. and many in the community are working on different levels of coarse grained models from nucleosome resolution, of course there is also the base pair resolution higher order polymer models, and together with a wealth of experimental data and from Hi C and other experiments is really creating a very exciting rich, opportunities for collaboration and advances on understanding how this epigenetic control of genomes is responsible for human diseases. And, From something very theoretical and abstract, you know, you have this image of this chromatin fiber folding in many, many levels, and forming all these contacts with proteins and RNAs, it's a very exciting, rich, dynamic environment that is yet to be explored fully.

Milosz: 17:16

Mhm. Yeah, absolutely agree. I mean, we have this wonderful data now coming also from cryo electron tomography, right? That starts to look exactly at, the path that DNA makes in the nucleus. Mm So this is going to be a challenge for people to integrate into what we know about. Chromatin structure, and perhaps we'll soon start getting data looking into that under different conditions as well, as you say. So yeah, that's, that's going to be a goldmine.

Tamar Schlick: 17:46

Yes, and this is where I think modeling is very important because in experiments, you can only explore one condition, but that's the nice thing. Once you have a model that's verifiable and reliable, you can explore different densities of linker histones or salt concentrations and that, that turns out to be very useful and for specific applications.

Milosz: 18:12

Absolutely. And alluding to what you mentioned already, you worked on algorithms. Do you think this new era of integrative slash large scale simulations will have its own challenges when it comes to developing algorithms for modeling for simulations?

Tamar Schlick: 18:29

Yeah, I mean, clearly AI algorithms will need a lot of statistical guidance in application reliability and so on. I think it's a very interesting story on algorithms, actually, because early on, I think we, we thought algorithms would solve all the problems. so I remember in the 1980s when I started working on algorithms, and it was algorithms for the time step problem in molecular dynamics. And a lot of mathematicians worked on that. And there was thought that these, uh, symplectic integrator. These integrators with favorable energy conserving properties would just solve this type step problem. But, of course that wasn't the case because, even though there was lots of great papers and we learned a lot about resonance and coupling, mode coupling, and we proposed many approaches that would mitigate some of these effects. It turned out that the method, the leapfrog method that people were using, which is velocity relay, has been, it's symplectic, it's very effective, and small time step is really the way to go. And I, think, Algorithms have a very, very important role to play in our, in the advance of computational biophysics, but I think that what is really important is the computation. and I think a lot of computation is what is going to help us solve real problems. I remember in, um, 1989, I read this editorial by John Maddox, who was the editor of Nature, and he said, One of these days someone will begin a paper, here's the Hamiltonian of a DNA molecule, and will then after a little algebra, tell us how the stop codons work, and so on. And I remember writing about this quote, and I said, I don't think it's going to be a little algebra. I think it's going to be a lot of computation. And I think, looking back, into the progress in molecular dynamics, that's what turned out to be really crucial. Not fancy algorithms, but just a lot of computer time and doing things. Of course, you have to do things right. Algorithms have to be reliable, but computer time has made a huge difference, and computational biologists have exploited computational resources extremely, extremely well. my perspectives, I have these charts showing how the field of computational biology compares to, progress in other fields or the fastest computers available. And our landmark simulations in the field are always near the top. So, I think we have We've made a huge effort into utilizing the best available resources out there. And algorithms have had an important role to play in guiding us and, uh, certainly giving us a lot of ideas, enhanced sampling, and as I mentioned, integration, guidance into resonance into long range interactions., Capturing those integrating, uh, say, PME with multiple time stiff methods. ultimately, lots of computation is, is really key to solving important real problems.

Milosz: 22:24

Right. So maybe this comes back to this concept of computational irreducibility, right? I think it's Stephen Wolfram's idea that very often you cannot know the answer to a problem until you actually go through the steps, iterate and see the outcome.

Tamar Schlick: 22:39

Yes.

Milosz: 22:40

So maybe we're suffering from the same, issue there.

Tamar Schlick: 22:43

And then you iterate again as you find more information and it never ends, right? Because there's

Milosz: 22:49

Right,

Tamar Schlick: 22:50

more and more to explore.

Milosz: 22:51

right. And there's, of course, there's this promise of AI, which is kind of starting to pay back that maybe this can be somehow alleviated by advanced AI systems. But as you say, there will always be a limit where it's, it's not the case.

Tamar Schlick: 23:04

Right. I think we're definitely at the hype of the AI phase and an expectation curve of the field. We're definitely just like, I think biomolecular modeling and simulations were in the late 1980s that they thought that will be the solution to everything, to drug design, but it's not. But there, and there's no doubt that AI will continue to make astounding progress, but I'm sure there will be more downs before we go up again with these methods. And there's so much to explore. Large language models are also beginning to help us understand certain problems, including RNA structure and transitions and, so I think AI and all these algorithms will be very important coupled to lots of computation and guiding and taking us forward to, to important problems.

Milosz: 24:04

Yeah. I think we need to stay on the ground there. Cause I personally go through the cycles of hype and doubt as well, from like, Oh, everything is now solved. And we will be left jobless to like, Oh no, it actually doesn't work at all. And probably the reality is somewhere in in the

Tamar Schlick: 24:21

it's

Milosz: 24:21

sense that

Tamar Schlick: 24:22

because every idea that you have, you find that someone's written a paper about it. Right?

Milosz: 24:28

yes, that's true. Okay.

Tamar Schlick: 24:31

and you, you sometimes think, well, what can I contribute if these people do did that? And then you find that people did that with such level of detail. How can you possibly approach it? I think that There's always opportunities for new perspectives I think that's why I like to work with undergrads the most because they're, they don't know anything and they're so open to new ideas and they say, well, why aren't you doing it this way? And they, you know, some little thing they heard in a class or, and it's refreshing and sometimes. That's what you need. So, even though it's easy to feel overwhelmed, I think, by what exists in the literature, when you really get to the heart of the problem, you will always find some niche where you can contribute something, and that is either unexplored or not done in a very good way.

Milosz: 25:32

And, there's also truth to the fact that even genius ideas are relatively cheap, right? There are so many genius ideas floating around. It's, the, Sitzfleisc h or the act of sitting on an idea and perfecting it to some extent that actually matters that, um, absolutely

Tamar Schlick: 25:49

Yes. Ideas are very, very important in guiding us, and I think it's even harder to execute them, and that's where most of the work, but without ideas, our field would go nowhere. So it's always good to dream and to come up with new, avenues for research, I think, and there's lots of room for every. A new person who comes is intrigued by the problems in computational biophysics.

Milosz: 26:21

Absolutely, people are welcome to come in. And to switch gears a little bit, I like to ask about the writing that people do. So you, had a book out, which is an interdisciplinary guide to molecular simulation. And you also, uh, wrote a lot of perspective articles. I also noted in one of the comments of one of the suggestions for students that mastering good writing was one of your top advices that you give. How do you approach writing? Because what struck me was that your interdisciplinary guide was in fact, very non mathematical in the sense that it didn't contain a lot of maths compared to many other textbooks that would approach this subject. what is your recipe for good writing?

Tamar Schlick: 27:04

Well, think I love writing because it helps me think and get my ideas down. And often I talk to my students and they hate the writing part, so they love the science. And I say, you know, until you don't start writing the paper. You won't really think and go beyond just the results and the data. if, if you're a good writer, you will find so much more to do once you put together the data. but I, I think, a textbook is, is a little different. Every new, teacher often wants to have the material available exactly for the course they want to teach, for the audience they have. at their institution or whatever. And when I started teaching molecular modeling, I had a very interdisciplinary audience from math, computer science, who knew nothing of biology and biologists, chemists who knew nothing of mathematics or, or some very basic, information. So my goal in writing the book was to just excite the mathematical people in biology. Why is it so exciting? And I know when I entered the field, What was exciting is that it's problems are tangible. You could talk about, nucleic acids or proteins that are actually relevant to nutrition. Or, you know, the amino acid content and vitamins and things, and, and the folding actually is related to things, uh, the DNA and genetics and disease and protein

Milosz: 28:47

huh.

Tamar Schlick: 28:47

workhorses of the body. so I wanted to excite those mathematical students in entering the field, but there's also very detailed. chapters on time step methods and, long range interactions, PME, and fast multiple methods and so on, and, the potentials, just the basics. I thought that it's very important for students to understand the basics. I know I had a hard time collecting that when I was starting out because the papers I read were so focused on the results, and, uh, Not so much on introducing the basics. so my main answer to you, I think that good textbooks teach you how to think or how to approach problem. And I think I was very inspired by Gilbert Strang, who wrote, uh, linear algebra books, which is my favorite math subject a very informal way. It's not stuffing people with information, facts, results, theorems. It's teaching you how to think. Why are things important? How are things related? How do you approach problems? And Strang is magnificent at doing that and linear algebra is a beautiful field. It really teaches you. It teaches you how to think abstractly and connect vector spaces with,

Milosz: 30:23

Oh,

Tamar Schlick: 30:23

practical things, and that's very useful in our field from

Milosz: 30:28

absolutely. I agree. If,

Tamar Schlick: 30:29

to optimization. Yep.

Milosz: 30:32

if someone hasn't seen his lectures, I mean, there are lectures also available on YouTube. So I

Tamar Schlick: 30:36

Yes.

Milosz: 30:37

recommend to everyone who ever struggled with linear algebra. This is a great resource. I haven't read the books, but I I know his style. I can imagine.

Tamar Schlick: 30:47

Yes, it's really good. And, um, besides books, I think, the problem with books today is they become obsolete, once you're done with them, in that sense, the material itself may be out of date, but I think the approach and the way they teach you how to think is always important. Also in making a stamp on the field, like the early book of McCammon and Harvey, on molecular dynamics was important in establishing that this is a field that's going to grow in its importance, I think. and that's why I think there should be more appreciations for perspectives and reviews that are very important in,, helping especially young people enter a field.

Milosz: 31:39

Right. Absolutely. If we want to be an interdisciplinary field, we need to have some sort of.

Tamar Schlick: 31:44

that's.

Milosz: 31:45

grounding in the basics. I agree with that. And part of the goal of this, podcast is to give people an overview are the different concepts or ideas we all, grapple with. So, I hope it facilitate some transitions between fields and, entering and deeper into other subfields. But yes, I think especially in this era where people from AI or statistics come into biology, I'm often worried that we need more resources like that to prevent people from reinventing the wheel. Maybe that's also something because everyone. All

Tamar Schlick: 32:23

to write another molecular mechanics program or integrators. There's good resources. So that. That's also important that before you work on a problem, I think it's. It's good to do really good homework, and make sure before you invest years in a project that it's not going to be wasted. But even, even though you do the homework, sometimes it does happen that what you discover is perhaps not as successful as you'd hoped, but I think you still learn a lot. nothing is wasted in research and training, even from negative results, maybe even especially from negative results. It always goes somewhere, and we never know when we may need to use what we've done. So I think the main thing is to work on something you're really passionate about, and, you can never be sorry.

Milosz: 33:29

right. And do you think the incentive structure is there for people to provide those resources, let's say, right. If, you, talk about perspectives, all this integrating, knowledge, do you think well, let's call it the system rewards those efforts, or is it something that you have to reward yourself internally? Yeah,

Tamar Schlick: 33:50

think that it's not as appreciated as it might be. I mean, there, there's a lot of special volumes, for example, in biophysical journals, and we always invite perspectives. And Like to do them. I think, but, most scientists I think at least the ones that are very comprehensive, the one it by physical journal are relatively short, the comprehensive ones like an annual reviews or, current opinion. Also, they have to follows certain guidelines, only the past couple years of papers, so, you're a little bit limited in what you can bring to the table. I think there should be more appreciation for those kind of reviews and a different way maybe to rate them or to highlight them because I know I've learned a lot from writing them. In putting together lots of activity in a certain subfield but I don't know if there is, you know, widely read as they could be. I mean, maybe we need to do more social media advertisements. I'm not, I don't use that for science, but I think younger people are good at promoting their papers now on social networks and that, that can

Milosz: 35:21

but I definitely have the impression that research papers. Kind of spread more rapidly, uh, than perspectives. There's much less recommendation of like, Oh, have you read this perspective on something?

Tamar Schlick: 35:35

Right.

Milosz: 35:36

Maybe we should do this more and,

Tamar Schlick: 35:38

it would be great if there, we could do it. Yeah.

Milosz: 35:42

and have more high level discussions about what things mean rather than just what came out of the simulation, right?

Tamar Schlick: 35:49

Right.

Milosz: 35:50

if we're at higher level conversations, you have just completed a marathon in New York. We had this discussion before, if it's real. I actually even asked chatGPT but apparently there's no statistics on that, whether scientists, especially computational scientists are more prone to running, what got you into running? was it? A reaction to something or some internal need

Tamar Schlick: 36:12

have? Because my husband jokes that runners love to talk about running. So, well, as scientists, we sit a lot, you know at least I do. I have extra energy that I need to expand other ways And running is a great way get out there and, uh, you know, be with nature, either run solo or with a group, think about things. I remember when I started running, a few years ago with Adidas runners, it's a group that all over the world did. branches, including in New York, Berlin, Tel Aviv, Shanghai, And I would go to their runs, the community runs, and I would say, why am I here? These people are so different from my background. And they're people from all kinds of professions, ages, political beliefs, you name it. And. I start talking to these people and I find, you know what, we have one thing in common. We're all these type A personality with goals and challenges. we like to set goals for ourselves. Training plans of how to get there, I think. Oh yeah. And I learned to really enjoy these kind of communities and the people and talking to different, not, not just about running, but about other things. when you're out there in July, in the middle of the heat and in New York at 5 a. m. doing long runs, you know, that takes dedication And those people, So I really appreciate my fellow runners that, you know, we're done before most people get up the summer morning. And,, so there's something very nice about that. Now, another thing that I thought about is these communities are, really a model communities. And in the sense that they learned about, inclusivity, equity, diversity, you know, way before. It became an issue I remember when I started running and we would have these tempo runs and it was, you had to push yourself and speed and so on. And the pacers would, you would end and I remember I was struggling one time and the pacer would push me, you can do it, you can do

Milosz: 38:48

Mm-Hmm. Mm-Hmm.

Tamar Schlick: 38:49

And then at the end When I'm done, all these people say, great job, great job. And, you know, we need more of that in science. And science is so competitive even though we, tend to hide it pretty well. We're competitive about papers, grants lots of things, awards maybe conference participation. We don't. tell each other enough. Good job. Great job. And runner communities, you know, you compete with yourself. You don't compete with other members of your community and I think that's, there's something very attractive about that and also, as someone who goes to a lot of conferences around the world, running is very easy. You can get up, you know, all you need is. Running shoes and you, you go out there, you explore, it's a great way to meet people and to just be out there with nature and get a perspective. And then you come back with lots of energy, ready to sit down the whole day, listen to lectures.

Milosz: 39:56

Absolutely. There's a sort of like mindedness, as you say, of a good kind, right? It's not necessarily having the same views, but having the same mindset. And

Tamar Schlick: 40:06

yes.

Milosz: 40:07

you're right that we could probably use this um, we'll appreciate this intrinsic motivation more and more in science and recognize it for what it is.

Tamar Schlick: 40:17

Yeah, we do it in a way. We do it for our students, but we don't do it in enough at different levels of, science, you know, as, Yes. people need it at all levels. We all are insecure on some level. You know, today it's hard not to, there's so much great science everywhere. so we need to be more, I think, appreciative of what each person brings to the table rather than combative or competitive.

Milosz: 40:49

Right. It's a great conclusion coming from running. One would think that it's a very distant connection, but, yeah, it's actually a very good point. Thank you.

Tamar Schlick: 41:01

Yeah.

Milosz: 41:02

And then I guess that should come from the communities we build as scientists.

Tamar Schlick: 41:07

I just wanted to, to emphasize the importance of. Just working with people you like and that you understand, how they think. And I started to work with different experimentalists, I met different kinds, and I think experimentalists see computational people as we'll just tell them what we want done, and they will do it, and bring it to us complete, right? those are people who just don't understand what computational biophysics is. And you can't work with these kind of people. But then when there are experimentalists who are really broad minded and they say to you, let me understand how you approach it through modeling. And then you have a real iterative process of, Okay, so you're doing it this way. Why can't you introduce this energy term, you know? And so, those are the people I work with. Not the former kind. Because that's not a collaboration. That's an exploitation.

Milosz: 42:21

Yeah, I agree very much.

Tamar Schlick: 42:23

Yeah, so I think that's, that's really important to work with people you like and you can learn from and they can learn from you and, once heard a great, definition of a good collaboration that both parties feel that they're exploiting the other.

Milosz: 42:41

That's a great, goal to aspire to. In a way.Think, yes, hearing the other side and doing the homework of understanding what they mean by that is important. And I agree that, many people approach computational science as a sort of magic. But, you know, whatever comes out,

Tamar Schlick: 42:57

And

Milosz: 42:57

Or maybe,

Tamar Schlick: 42:59

a black box. Yeah.

Milosz: 43:00

or maybe just a visualization, right? Oh, like, can we have a figure that shows this and that's it? Yeah, I agree. Hope some experimentalists are listening to that and paying attention.

Tamar Schlick: 43:15

well, I think this is a very exciting field for young people. There's so many great problems to work on, and one should not get discouraged by the vast amount of literature on everything and anything. And, you know, there's more than AI to our field, and there's lots of room for new methods and innovations and applications. And the main thing is, find a great problem to work on

Milosz: 43:46

I think one thing that physicists and mathematicians are always surprised by is how easy it is to have a conversation with a biologist and to come to the conclusion that we actually don't know this, right? Because in physics, especially like fundamental physics, there's always an explanation from the 1930s or so.

Tamar Schlick: 44:04

yeah, I've learned one of the early things I learned about biology is it says, is it A or is it B? It's always both.

Milosz: 44:12

also true. Yes.

Tamar Schlick: 44:14

So that's been a very important realization.

Milosz: 44:20

Okay. Tamar Schlick. Thank you so much for the insights and for the conversation.

Tamar Schlick: 44:24

Thank you so much. And, uh, it's really great that you're doing this podcast and, you know, great service to our community.

Milosz: 44:34

Well, we've got one great conversation more. Thanks to you. Have a great day.

Tamar Schlick: 44:38

Thank you. You too. Bye.

44:42

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