Night Science
Where do ideas come from? In each episode, scientists Itai Yanai and Martin Lercher explore science's creative side with a leading colleague. New episodes come out every second Monday.
Night Science
2 | Tzachi Pilpel on channeling other people’s minds for creativity
In this episode, your hosts Itai and Martin talk with Tzachi Pilpel, Professor of Genome and Systems Biology at the Weizmann Institute of Science in Israel. Tzachi eloquently describes his creative process, the role of language, the freedom of data analysis, the imagined channeling of other people’s minds for new ideas, and scientific fearlessness.
Tzachi’s research focuses on complex networks within cells. His lab applies systems biology and genomics experimental strategies to the study of genetic circuits that process and transmit information in cells. A central goal in his lab is to define entire pathways through which proteins affect changes in gene expression. Among his many awards are an IBM Faculty Award, the Michael Bruno Memorial Award, the Hestrin Prize of the Israel Society for Biochemistry and Molecular Biology, the Morris Levinson Prize in Biology, and the James Heineman Research Award. In 2011, Tzachi was elected a member of the prestigious European Molecular Biology Organization.
For more information on Night Science, visit www.night-science.org .
Martin
A few days ago, I did an exercise of walking in the street with my eyes closed. Have you ever tried that?
Itai
Yeah, we call it night science.
Martin
That's nice. Welcome to the night science podcast,
Itai
where we explore the untold story of the scientific creative process.
Martin
We are your hosts,
Itai
I am Itai Yanai,
Martin
and I am Martin Lercher. Today we have with us Tzachi Pilpel, the head of the Department of Molecular Genetics at the Weizmann Institute of Science in Israel.
Itai
Tzachi studies the function and evolution of genetic regulatory networks. In other words, he's interested in the grammar of the genome language. Tzachi also happens to be a great old friend of mine.
Martin
Tzachi, thank you for being with us today. Well, we know you're one of those very creative scientists. Can you tell us something about what aspects of your training as a scientist, influenced your development of becoming such a creative person?
Tzachi
The first answer was surprising to me. To be honest, it was programming, the ability to write code in computer.
Martin
That is surprising.
Tzachi
It is surprising. I was surprised and then I asked myself, why did I give myself this answer? And immediately, my mind went back, you know, maybe or more years ago, to the first days that I started to program, I just learned how to do that, and I got biological data at my disposal. It was sequences of factory receptor genes and the ability to program opened a whole new world to me. And it allowed the creativity to be satisfied at an amazing pace, because I would ask something, and then I will learn how to write a code to ask that question in a way that very quickly, I could answer it, and it started a very positive feedback loop of raising questions getting answers, and again and again. So programming, surprisingly, which is often considered to be a very technical and non-creative, you shouldn't be too creative, you know, within your code because that would lead to obvious syntax errors. But you know what I'm saying? It allowed me a very active avenue for creativity. So that's one I think.
Martin
So it sounds like you almost learned a new language in which you could ask questions much more efficiently than than with the other languages, you knew?
Tzachi
Yeah, yeah. Yeah, I think language is indeed a key word here. And maybe, you know, maybe like a child that learns a new natural language. And all of the curiosity of the child can find a very fruitful avenue because you can start asking questions.
Itai
You know, these days, when people explore the data, they're filled with notions of how you have to pre-register hypotheses. And you know that there are a lot of kind of like rigorous rules are built around it.
Tzachi
I am trying to give my students the legitimacy to explore the way I love to explore, at least, you know, in a little heaven that we are building for ourselves here, I think it still works.
Martin
I mean, exploratory data analysis is a fantastic tool. And I think it's cool that he encourages students to do that. But then some scientific circles don't really like that, right? And these kinds of works, if you publish them sometimes get derided as fishing expeditions. Have you yourself ever been accused of doing fishing expeditions?
Tzachi
I actually don't recall being accused of doing fishing expeditions. And it's not because we don't do them. I was prompted by you to think about that, in retrospect, to see, you know, what would I consider a fishing expedition among the works that we have done in the past maybe the ones that I appreciate the most. So, I'll give you two examples of something that started as a fishing expedition and turned to be something very profound. Several years ago, I think it was eight years ago, Anders Lund from Copenhagen was here at the Weizmann and he showed us his microarray for non-coding RNAs in the human genome. And incidentally, he had a few probes for tRNA genes that he hybridized to hundreds of cancer samples from patients. And I asked him, can I just look at the data because I care about the RNAs? And he said, Yeah, sure, take a look because we couldn't make sense of what we see with the tRNA, maybe antisense but not sense and we started to look at the data and maybe one of the most important discoveries from my lab came out through a collaboration with Lund. It even culminated in a cell paper in , that we discovered duality of coding of human genes and how tRNAs take advantage of this duality in cancer.
Itai
That's a brilliant paper.
Tzachi
So that was just a fishing expedition. In contrast, if you asked me if I was actually blamed of anything, I was blamed of the opposite. years ago, one of the brightest students I've ever had, Amir Mitchell, who is now an assistant professor at the University of Massachusetts, started his PhD with me and he came to me with an amazing idea. And that was a hypothesis. He said, I hypothesize that when organisms respond to environmental changes, they do not merely respond to what they encounter at the moment, but they try to predict and prepare in advance to the next change.
Martin
That's a work I love. I think that's beautiful.
Tzachi
It was Amir’s hypothesis, it was his original idea. And immediately, you know, I'm still excited when I recall that moment, because I said, “Wow, I'm hearing now a very, very cool idea”. And then the question was, what do we do with that? And for four years, we have attacked the question from various angles. And, you know, a very nice work, I think, came out of it, etc. But when I showed it to at least, I remember one response from a very prominent scientist in Boston, who said, you knew what you're going to discover from the very first moment.
Itai
You can't win, you can't win.
Tzachi
If you are on a fishing expedition, at least from this critique, you're immune.
Martin
Yeah, absolutely, absolutely. But that's funny if you come up with an idea and then you can think of a way of how to show that this idea is actually true. That's, of course, awesome. It's not always that straightforward.
Itai
And Tzachi, do you find that in your own work, you consciously or unconsciously distinguish between night science and the more sort of executive aspects of science?
Tzachi
Yeah, I think I'm making perhaps less of a distinction than some people do. I mean, if you don't do a great distinction between the day and the night, it's either because the day penetrate into the night or the night penetrate into the day. I think, for me, the night is welcome within the day, it's not so much in the opposite direction. And I tell my students, especially when practical problems are encountered, I love to give them an advice that I've gotten once in an analytic, creative thinking course that was given here at Weizmann when I was a student. If you see a challenge, if you see a problem, just imagine that there is a little magician that solves it for you and just keep going. I don't know how to, by the way, programming is all about that, right? Because if you write your code properly, you have it in sub routines and functions and procedures, etc. So even if you have no idea how to do a Fourier transformation, you just call a function that does it and then maybe you have to write this function. But it's when you think about the whole flow, you don't need to worry about that. So, I think it's a cornerstone of creativity to believe that all sorts of problems, somebody will solve them for you. And maybe that somebody will turn out to be yourself, but not now.
Itai
That’s right. Like you you take this sort of irrational step to say, you know, let's just assume for a second that it is solved. Is it even interesting? Exactly, you know, Tzachi, we're talking in a really abstract way. I'm wondering if you can walk us through an entire project that you once did, any project you choose, and just sort of give us not the published official version, but the behind the scenes.
Tzachi
So, when Amir Mitchell came up with this idea, we said, “Wow, that's amazing”. It feels as if we are trying to import some concepts from cognitive psychology into microbiology, because in Pavlovian psychology, that's exactly what the dog does when it hears the bell and knows to condition it with the arrival of the meat later on. So the first thing that we have given this idea is a name. We said that would be Pavlovian microbiology or genetics or something like that, and that excited us very much. So then we said, “Okay, what do we do about that?” and I mean, how do we discover anything useful about it? And we realized that we should be doing three things. It wasn't that obvious right from the beginning. But we thought one thing would be to see if organisms in nature, have the ability to condition and to prepare in advance for a change before it occurs. Second thing we said, let's see if we can evolve in the lab a new type of conditioning and what was nice is that we could inherit a lot of knowledge not that we had, I had a lot of knowledge in psychology but psychology had a lot of experience and knowledge in investigating conditioning. And it was nice because we could take the analogy between psychological conditioning and microbial conditioning all the way through, because what's a bit less famous? Is that following the establishment of conditioning with the Pavlovian dog, is that Pavlov then took his dog to two additional stages of extinction and reinstatement. Extinction is when you in that case, you know, you would ring the bell, but don't provide the meat. So, the dog learns that the rules of the game have changed, and it stops salivation for the bell, but only keeps it for the meat. So, we said, okay, maybe we can do that too. Maybe we can evolve our microbes to extinct the connection.
Martin
Yeah, you think it's awesome that you use concepts from psychology in your genetical research?
Tzachi
You know, I'm a biologist, but I like to take some inspiration from other disciplines, other sciences, or maybe not even a science, language, or psychology. And when I do that, from the other discipline, maybe I'm being less disciplined or more, you know, more night scientist. I know, I know, psychology at a very, very superficial level that might suffice for a night research. You know, biology is something that I'm more invested in. So maybe the distinction is this.
Itai
So, you were saying how you were using the analogy, those Pavlovian terms, then what happened next?
Tzachi
One thing is to see if microbes can do that. The second thing is to see if we can evolve a new conditioning, extinction, reinstatement to follow all the way through the Pavlovian paradigm. Third thing was to model it mathematically to see how realistic is to expect that organisms would have this capacity in terms of the fitness reward for being able to predict relative to the cost, etc. And then started the whole very, very long process that, nevertheless, if I'm very honest, I think it was exciting, almost throughout, of investigating in parallel, both yeast and E. coli. It was when my lab actually transformed to become an experimental laboratory. I had to learn experimental science while discovering something quite amazing, which was a very heavy, I must say, mental burden, because I felt that I don't know enough about the technicalities while I'm trying to discover something very interesting.
Itai
What you're describing is actually amazing, because what you're saying is that your first experimental paper, after many years of being a computational biologist, was actually published as an article in Nature. That's something pretty extraordinary.
Tzachi
Yeah, it didn't happen to me too often later on. So, I don't know if that promise is fulfilled.
Martin
Maybe you should switch fields more often.
Itai
Given that you've transitioned, also do wet that experiments in addition to the computational stuff, what did you learn about that transition? Is it a sort of different kind of creativity? Or, you know, is the night science the same?
Tzachi
Yeah, you know, there's one obvious thing that frustrates your creativity a little bit when you do experiments and this is the waiting time between raising a hypothesis and seeing results. This is amazing, you know, you I just started, you know, talking about programming, how immediate is the reward that you're getting, and how, you know, it puts you within a positive feedback loop that is very immediate. And in experiment, it's not the case you decide to do an experiment, and you know it very well to, and then you need to wait sometimes half a year before you're beginning to see some results. So so in that respect, it was a bit disappointing, I would say. But on the other hand, being able to ask questions within the lab is amazing, is totally new ground, and I am still very, very excited about it.
Itai
Have you found different ways to think about how to integrate the theory and the computation? For example, do you generate the data first and then explore computationally or, or use the experiments as a kind of final test?
Tzachi
I think we are exploring as appropriate for each project every possible way. One of the things that I'm very excited about now is trying to think about a good experiment to test an exciting theory. This is something you know in physics, they were there all the time, right? I even find myself recently establishing or trying to establish new experimental system to test other people's theories. Something that I would never Imagine what happened to me.
Martin
So, one thing that can clear in the examples that you gave was how important language is in this, you first give a good name to something, right? And then that kind of helps you think about it, you know, like the Pavlovian effects that you talked about earlier. So, it seems that language is an important night science tool that you use. Are you aware of any other techniques that you use in your creative scientific process?
Tzachi
One which I love the most, is that when I think of an idea, I oftentimes like to think of myself discussing it with another scientist. And it needs to be a very particular scientist, a particular person. It could be you. It could be, I don't know, Naama Barkai in our department is a constant visitor in my mind when I have ideas. And then what happens is that I can think about things that I would not have thought of, if I was just thinking about them myself because, you know, I spent so many hours discussing science with her or with other particular people, and inviting them into my brain for an imaginary discussion allows me to harness the brainpower and different ways of thinking of other people. You asked me by the way, if there's a particular paper that I want you to look at - the story behind this paper is quite amazing. I worked on it in , when I was a PhD student. And at that time, I was talking quite often with Michael Levitt, and got to appreciate his way of thinking and got a little bit of his vast knowledge in protein structures, etc. So, he was at Weizmann, back then he was a professor at Weizmann actually the work that he got the Nobel Prize for what he did, when he was with Shneior Lifson and Arieh Warshel here at Weizmann, at least some of the work was done here. We were quite close. And I got to know about how he thinks. And then I'm reading a paper on something totally different from what we discussed. But I think I was reading the paper from his mind, kind of, his mind represented within mine.
Itai
You're thinking, oh, what would he say?
Tzachi
Things like that, no split personality, but just one mind seeing the world through the eyes of another mind, all of that happening in mine. And then I immediately saw something that I think I wouldn't have seen if I thought about it myself. And that led to this paper, because what I discovered there is a way to model energetics of membrane proteins. And since then, I've been using that technique all along, always recruiting particular people and you could ask who owns the discovery.
Itai
Mike Levitt would be pleased to know he owns this discovery too.
Tzachi
Who owns the intellectual property,
Martin
I think it's beautiful that you have those discussions in your head with people. So, it's kind of an in-between between doing night science all by yourself and doing it together with other people where you discuss things, right? So you have these imaginary people that you talk to, I mean, they're not imaginary people, but they're not actually there when you do discussion. I think it's a great technique. And actually, you know, it reminds me of a book that I and Itai read while we were writing our own book, The Society of Genes.
Itai
Are you gonna talk about Stephen King?
Martin
Yes, I'm talking about Stephen King on writing, a beautiful book on how to write but Stephen King said about his own writing processes that when he writes he always has a particular reader in mind not not just some arbitrary reader but a person right and which happens to be mostly his wife, but he has that one person he's talking to in his books. You know, what you said reminded me of that nice.
Itai
And Tzachi, how do you know who to channel when? I mean, when is it a problem for channeling Naama Barkai. And when is it one for Michael Levitt?
Tzachi
I don't know. But preparing for this interview when I thought about that I immediately imagined talking with you Itai about this, about this notion. And you would never believe what you discovered.
Itai
Well, I can't, I can't take all the credits.
Tzachi
No, I’ll give you some because what you told me before I tell you what you told me and tell you about my second technique. My second technique is to walk while thinking. I love to walk, and I love to walk while thinking. So, for example, the Weizmann campus is beautiful. I love to walk in campus and to think or to talk with people either in reality or by imagination. And it's a very interesting technique. Because sometimes what happens is that thoughts get anchored into particular places. You walk many, many times in circles, every time you get to the same corner of the grass or the lawn or something, the same thought gets surfaced again, but maybe slightly modified. I think it's a very interesting mental process that takes place in the mind when you somehow combine an abstract thinking with spatial, you know, anchoring. And when I told you, Itai, in my thoughts about that, you realized something quite amazing. Maybe we realize it together, I don't know. But, you know, I noticed, or you noticed that in neuroscience, two, very prominent recent discoveries have to do with two types of neurons. The first is called mirror neurons, and the second is placed neuron spaces. So, neurons, as you know, are the neurons in your mind that allow you to see the mind of somebody else. It is for empathy and places are those cells in your brain that are activated whenever you are in a particular place. And I thought, it's interesting that my two tricks for this night science have a particular neuronal correlate into very interestingly studied types of neurons, one that allows you to reflect the mind of others and the other to combine spatial information into your thoughts.
Itai
Absolutely, and I think also the Weizmann Institute, landscape gardeners will be very pleased that you're taking advantage. You know, it reminds me of the book by Daniel Kahneman, where he would talk about how him and Amos Tversky would take long walks and, and really have a discussion. So I'm wondering, also Tzachi, you're telling us a lot about night science that you essentially do in your head. But, you know, it's very clear to me that you also have long night science discussions with members of your lab. So maybe you could tell us a little bit about that? Do you have a place where that happens? Is it necessarily in your office? Is it with just one member or multiple? How do you do it?
Tzachi
I think never at the same place for too many times. I really try to embrace the fact of different places on the mind. By the way, now, in Corona times we have to meet outside, which I think makes a beautiful effect on the way we think and the way we reason and the way we try to be creative. There is a beautiful deck, for instance, in our building, and we love to sit there and yeah, but I'm just trying not to be hooked to the same place when we're thinking.
Itai
Same place, same ideas. Yes, thank you.
Martin
So do you typically, when you when you have this night science together with people who are actually physically there and not just in your head? Is that typically also one person? Or do you prefer to do it with a whole group of people?
Tzachi
I think both. But what I really really love is when very spontaneously within the lab, a few students, we gather together, me and some of them my students, and then a discussion begins. Oftentimes, it's done in front of some of the students computer or something. And then a few more people gather, and we just brainstorm. These are the best moments in our science. So, it's in groups of, you know, four or five, six people. Sometimes it's just one person.
Itai
You know, the question, going back to what we were speaking about earlier, regarding language, I think also about our discussions with Doron Lancet and how much he likes puns. And I think that's something that you and I sort of inherited, too. I'm wondering if language itself can be such a great source for getting new ideas. You've talked about the analogy to Pavlov and about social, socializing to change mutation rates.
Tzachi
Yeah, yeah, I think so too. And I think I want to think about it more even after this interview, because you're raising a very interesting point that I was only partially aware of, but yes, I think I love this and I got it and I know that you got it too from Doron. I would love to think more about the possibility that is word games and horns, etc. effect. I think they do affect my and I think yours way of thinking too.
Itai
You know, it's very obvious to me that you're interested in language you recently organized a big conference at the Weizmann that was truely interdisciplinary.
Tzachi
Yeah, it was a marvelous conference. This was a conference that I've been dreaming of doing for many years. And at some point, two years ago, I said, Well, I'm actually going to do that. And I invited people from biology and from language. I asked a colleague from Haifa University, Wendy Sandler, who is a linguist, to join me so that we can attract linguists from around the world, people who are interested actually in language evolution. And we had an amazing list of speakers from both disciplines. Almost everyone that we invited came happily. I wanted to make it a special conference that would be not only because of its science, but because of the atmosphere. So at a lot of culture involved, too, we had several performances, and artists and dance company play big improvisational theater.
Itai
Sounds amazing.
Tzachi
It was so much fun. Yeah.
Martin
No, it really sounds great. And, and also like it would inspire a lot of creativity. Yes, yes. I actually remember a conference, an earlier conference that you organized at the Weizmann where both Itai and I were there, which was extremely amazing as well, for for many reasons. But there was also a performance there by two of the participants of this podcast.
Tzachi
That's right. Yes. Of course.
Itai
Society, society.
Martin
We sang, we sang a song about our society of genes.
Itai
That was, that was a lot of fun.
Tzachi
Yeah, of course. You were, you were wonderful.
Martin
Thank you. Thank you. Tzachi, I have one more question for you. So, you told us, for example, about a great idea from a colleague of yours that you thought you really want to test. But you also had a lot of great ideas yourself. So how do you know a great idea when you hear one or when you have one?
Tzachi
When I hear one? The first impression that I have is – “Oh shit, why didn't I think about that?”
Martin
Great. That's great. Yeah, because that's exactly what Itai told me.
Itai
It's painful jealousy.
Tzachi
How come? I didn't think about it? Yeah. It really hurts. When I have a good idea myself, I tend to be very optimistic about science. So, I judge ideas by the score that they would get at the best-case scenario, I would say, in the best case scenario, this will tell us this, and that, and this is how interesting it would be.
Martin
So you are an optimist at heart?
Tzachi
Yes, I think so. And I tend to disregard many of the problems, the shortcomings, the challenges. Sometimes it's good that they do that. And sometimes it's terrible. We talk too much about successes today. I don't know if you want to talk about his failures. But
Martin
Please, go ahead.
Itai
There's no success, like failure, as Bob Dylan says.
Tzachi
Yeah, I think, I had more failures than successes, for sure. But to be honest, I think that what allows me to survive easily in science is that I don't care so much about failures. Honestly, I mean, if it's a failure that, you know, when destroy the thesis of a PhD student, then of course, I would care very much. And I will do everything to ensure that even if the project fails, the student doesn't. But if a project fails, I think I can take it quite lightly.
Martin
But I think that's one of the prerequisites for being a scientist, right? You have to be stable against this kind of disturbances.
Itai
I think also Tzachi, it's a very important thing for people who are graduate students today and postdocs to hear you say, you obviously, the super accomplished scientists say something that would be very surprising for them to hear that, you've had more failures than successes. No, absolutely. But how do you train your students to have this sort of optimism to overlook the failures and keep their minds focused on a possible success?
Tzachi
I think that's first of all, from example, they never see me, and it's not that I'm hiding it. They never see me taking a failure, I don't know too badly, or in a way that would be that would be destructive and I'm trying always to tell them, this can only be a good project if the chance to succeed is it has to be low. Wanting to build projects that have a high chance to succeed, and I think that my aim is to get them used to this notion and to recognize that it's legitimate, and that they should, should be ready for a failure. Again, there is a difference. I'm trying to secure the student, not the project. So the best way to secure a student amidst projects that are risky risk is to, for instance, have backup and multiple projects and etc.
Itai
Wow, I think that was a very interesting and maybe even profound way to, to draw to a close. This extremely interesting discussion that I think I'm going to be thinking about for much longer.
Martin
Yeah, I learned a lot about night science from you now. So, thank you very much. It was an amazing discussion that we had.
Tzachi
Thank you very much. I think your whole idea of night science is brilliant, and I enjoy very much reading it. And I was provoked by the notion by your questions, and I enjoyed very much this discussion, too.
Itai
Thank you, Tzachi. That was so much fun.
Martin
Yeah, it really was really wonderful for me to, you know, like this idea of this mental discussions that you have. It's not a technique that I use, but I think it's one that I'm going to be using from tomorrow.
Tzachi
Wonderful. You can invite me to yours and I will start inviting you as well, Martin.
Martin
Thank you very much.
Itai
If we all get into the convention of also acknowledging this in the acknowledgement section of the paper, thank you for imaginary discussions with Tzachi Pilpel.