In this episode we explore the vastness of the dark genome and why "junk DNA" has been overlooked for so many decades.
For the third season of Theory and Practice, we wanted to ask: what lies ahead for the intersection of life sciences and data sciences in the next ten years?
In this episode, we explore the vastness of the "dark genome" and why "junk DNA" has been overlooked for so many decades.
Our guest is Dr. Rosana Kapeller-Lieberman, A GV Fellow and the CEO of Rome Therapeutics with over 25 years’ experience in science and therapeutics. Rosana discusses her team's scientific approach, and how they have tackled investigating the 60% of our genome that we previously thought was just filler, or repeatable DNA.
This season we'll dive deep into the languages of life through explorations of the "dark genome", genome editing, protein folding, the future of aging, and more.
Hosted by Anthony Philippakis (Venture Partner at GV) and Alex Wiltschko (Staff Research Scientist with Google AI), Theory and Practice opens the doors to the cutting edge of biology and computer science through conversations with leaders in the field.
Hello, you're listening to the third series of Theory & Practice all about science for the future. I'm Anthony Philippakis.
And I'm Alex Wiltschko.
In this series, we'll be exploring research that we think may be impactful in the next 10 to 20 years. We focus on research in the life sciences, particularly where it intersects with the data sciences. Today, we'll be exploring a new area called the Dark Genome. And by that, we mean the part of our non-coding DNA, which is the repetitive filler sitting between our coding genes, and the regulatory elements that control them. For a long time, people have referred to it as “junk DNA,” or the colorful phrase, the repeatome. It showed up as long dark patches that were hard to analyze with existing methods.
Since the development of long read genomic sequencing combined with machine learning, some extraordinary findings about this part of our genome have come to light. Around one in 12 of those repeated segments of non-coding DNA are recognizable as ancient and current viruses that have been incorporated into our DNA. They're passed on through the generations and picked up as we live. We are all part-human, part-virus. Complex mechanisms lead to these parts of our dark DNA being reactivated under stress. For example, when the chickenpox virus lies dormant in the peripheral nervous system, under stress, it is reactivated as shingles, at least partly by the Dark Genome.
We know that about 60% of our Dark Genome is composed of things called “transposable elements.” These transposons can duplicate themselves hundreds of times and move around the genome. There's more and more evidence emerging that these transposable elements are implicated in the regulation of our innate immune system, and hence our initial response to attack by viruses and bacteria. Ongoing environmental stress can also challenge this part of our DNA, and potentially promote the development of cancers.
To help us work through the complexity of this topic, distinguishing between hype and real potential is Dr. Rosana Kapeller-Lieberman. She is Google Venture (GV)’s, first Entrepreneur in Residence, CEO of ROME Therapeutics, a medical doctor, and she has a PhD in Molecular Cellular Biology. She also has a deep expertise in computational approaches to biology.
Hey, Rosana, welcome to Theory & Practice. It's great to see you here today.
Very nice to be here, Anthony. It's a pleasure.
So, you and I are both on the team at GV. Over the last few years, we've had some wonderful talks about the Dark Genome. And now, I think it's a chance for our audience to hear a little bit more about what you think it's about and why you think it's important. So maybe you can start off and say, what is the Dark Genome and what is the repeatome?
So, the Dark Genome, the best definition I have seen is: everything in our genome that does not encode for proteins or regular proteins. And if you think about it, only 2% of our genome is responsible for basically encoding for the proteins that we are aware of, and about 20% of our genome are part of the regulatory mechanism that basically modulates the expression of these proteins. So, if you think about that, it leaves about 70% of our genome that's considered the Dark Genome. We don't really fully understand the function of it. And for the longest time, this Dark Genome has been ignored; it has been thought to be: “Oh, maybe it just plays a structural motif. You know, it's important for chromosomal structure; it doesn't do anything. It's dormant, you know, we just carry it along for the ride.” And that doesn't really make a lot of sense from a biology evolutionary perspective. And I think in the past five to 10 years, with the advent of new technologies, especially next gen sequencing, long read sequencing and machine learning, etc, we are starting to understand that this Dark Genome is not as dark as we think; there is light there; a lot, actually. And we see bursts of it in adult normal, healthy cells; it's usually dormant. But during certain conditions, it gets reactivated and plays a very important role in both human health and disease.
I just want to be clear about what we're calling this. So, you didn't mention it, because I think you're avoiding the term, but I don't think you like the term “Junk DNA” for the Dark Genome.
No, I don't like that term “Junk DNA.” It's not junk. You know, people call it junk just because they didn't know what to do with it. And to be completely honest, when we didn't have the technologies, we developed ways to mask it, and basically throw it out, so it's like the proverbial “throwing the baby with the bathwater.” I think that has changed and I think people are now paying a lot of attention on what is inside this Dark Genome, that “junk DNA.”
So it's kind of, “One person's trash is another person's treasure.” It's your treasure, I guess!
It's definitely my treasure. You know, I am absolutely fascinated with the Dark Genome. And I think there's so much there.
“Rosana Kapeller, dumpster diver. Genomic dumpster diver.”
I never thought about it this way. You know, I think that from rags to riches, you know, what is? Yeah. Dumpster Diver? Okay. I'll go with that for the moment.
But going back to the science here; this is a major part of our genome by bulk. And we couldn't see it before, until, as you mentioned, long read sequencing was developed, because of how repetitive this is. And maybe also some statistical techniques here too, absolutely. Could you maybe give us a tour of what we have learned about it that's made you excited about the Dark Genome in the past five or 10 years?
Absolutely. So, before I go there, I want to make an analogy of the problem. It's like having a 100,000 piece puzzle or a million piece puzzle with no edges. They all look the same, no picture on the box, and then being told, “Okay, put the image together” - it's basically almost an impossible task. And that's what we're faced with, by the nature of the elements in the genome, that are repeats. So these are repetitive sequences, repetitive elements, that have basically integrated our genomics revolution, through actually viral infection. So in a way you think about this as our genome parasites. So if you go and look into the Dark Genome, 20% of it is LINEs, which are long interspersed nuclear elements that basically are transposable elements. It’s the only active transposable element that we have in our genome. And there is more and more that we're learning about that, in terms of genomic plasticity, embryogenesis, and very much the role it plays in pathogenesis and a variety of diseases. 8% is comprised of HERVs. These are human endogenous retroviruses that also have activities. You know, if you look at these retroviruses, some of them are incomplete. But some of them are complete pro-virus that have gag, pol, and env. What are they doing in our genome? Many groups think that they went there to die. But are they responsible to help us defend ourselves against viruses? You know, what is their role in our genome? I mean, they comprise 8%. It's a big number. If you think that only 2% code for proteins, it's incredible.
So I'm curious about that, Rosana. You kind of allude to many different functions. But, should I think of these sequences from HERVs as scars in our genome? Should I think of them as little badges or medals of achievement that we beat them? Or are they hitching a ride? And you use the word parasite? Is that the right way to think about them?
I think it's all of the above, Alex, because I think it depends when they integrated into our genome’s revolution. The most ancient ones are definitely fossils; they don't do anything. They're like battle scars. But there are more recent HERVs that seem to be active. And the other thing about HERVs is that we have co-opted some of the HERVs proteins. For example, the most famous one is Syncitin 1, which is the envelope protein of HERVs-W, which is important for placenta formation. So, if you think about that, without Syncitin 1 mammals would not exist.
Well, you know, it's fascinating on so many levels. And again, let me step back, especially for some of the listeners that maybe are not as close to genomics. I'm curious; I think of repeats as being like the following: you have two kinds of them; imagine your genomes as a book. And imagine that there's a page of that book that just gets repeated, reinserted over and over again, throughout the genome. And you'll see 100 copies or 1000 copies of that one page, just at random places. And then there's a second kind of repeats, which are like in the old movie, The Shining, where, you know, there's like, “All work and no play makes Jack a dull boy, all work and no play makes Jack a dull boy…” And so there are passages in the genome, where you see the same sort of short string repeated over and over again. So that second class, we've known a long time that they could cause disease like Huntington's diseases is “All work no play, Jack…” But I think one of the things that is so fascinating about your work and the work of a lot of your colleagues, is this idea that these other classes of repeats, the pages that are just randomly reinserted over and over again in the book can actually also cause disease.
Absolutely. And just before I got there, just want to finish with some of the elements just so that we understand that what this is comprised of. So I talked about LINEs I talked about HERVs, you have also SINEs and Alus, and satellite elements, all of them play an important role. Some of them still have high CpG content. So they look like viruses. And I think that the breakthrough that happened is when we understood that these repetitive elements are not always dormant, okay? They're very active during embryogenesis. Then once you're born and you're healthy, these repeats go dormant, that is true. But upon certain stages, like for example, if your cells are submitted to environmental stress, smoking, irradiation, UV, whatever it is, you know, there is an epigenetic control of these repeats, and these repeats get reactivated – it’s basically hyper methylation of the areas of the genome where they are located. And these repeats actually get expressed; we have transcripts of these repeat elements or RNA. And what happens is that these transcripts, now, they can go and activate our innate immune system. So in a way, think about it as sort of our first sentinels, our first army to say to the rest of the organism, the cell is sick, let's kill the cell. Because it's sick. It's like having an endogenous infection. And basically, you kill this out and you establish homeostasis. So if everything is going well in your organism, you just have a healthy organism and the cells. But what if this mechanism goes rogue? In a way, you have to…it's very similar to what happens when you have a viral infection, you can have hyper activation that leads to a cytokine storm and damage to the whole organism. Or you can have hyper activation or latency that leads to diseases like herpes zoster, you know, you get chickenpox when you’re a child, and 50 years later you've developed herpes zoster. And the same thing happens here with the repeats. If it goes wrong, and you have hyper activation, you can get autoimmune diseases, you can get neurodegeneration, you can get aging, which is something very interesting. But also if you get hyper activation or certain types of hyper activation that can underlie some of the proliferation that we see in cancers. So what is happening here, and it could be both intrinsic factor in tumor cells as also an extrinsic factor in the tumor microenvironment. And that's what we're studying right now: the role that repeat plays in the pathogenesis of these different diseases. And where we're focusing right now is ultimately diseases and cancer, but also have a pretty healthy interest in neurodegeneration and aging.
In going back to kind of the taxonomy of the repeats, you have HERVs, and you have transposons, you have SINEs and Alus, which map those on to the different disease processes you talked about which ones, for example, trigger the innate immune system…
In terms of the transcripts that induce the innate immune system, what folks have shown is that LINE1, Alus and SINEs seem to be activators of the nucleic acid sensing pathways, like RIG-I, MDA5, and also cGAS in this tank pathway. And what is interesting is that repeats like LINE1 and HERVs encode for reverse transcriptase. So one of the things that fascinates me is that the repeating coded reverse transcriptase plays a big role in reverse transcribing, repeat RNAs into RNA DNA hybrids, and then double stranded DNA s. And now we can show that this hybrids and the double strand DNA activates cGAS-STING pathway. So, in a way, we're starting to interrogate whether those repeats are the genesis of sterile inflammation that has been implicated in a variety of autoimmune diseases. So again, it looks like an endogenous infection. Nobody knows where it comes from. But now you can sequence like if it takes cGAS, you can now sequence these DNA-RNA hybrids and double-stranded DNA and cGAS that are derived from repeat motifs.
Let's deep dive a little bit on cancer because I know that's one that you followed really closely. And so I remember you telling me an incredible story about the HERVs and actually treating cancer patients with anti retrovirals. Maybe you can kind of connect the dots a little bit on that.
Sure. So this is work actually from one of our scientific co-founders, David Ting, that started exploring that, where basically he saw the repeats and he saw the high expression of repeats in cancer and he hypothesized that if you actually would block the reverse transcriptase of those repeats, you would have an effect in cancer. The answer is that the jury's still out. I was mentioning before, whether there is an intrinsic effect or an extrinsic effect. And I think that it's going to be very much context dependent on whether inhibiting CAR-T is going to be beneficial or not for cancer, I think that for certain cancers, the answer is yes. For other cancer, the answer is no. And the reason for that is that if the cancer requires pro inflammatory signals to proliferate, then the answer is no. If the pro inflammatory signals actually blocks that particular cancer, like what you have with checkpoint inhibitors, the answer is yes. So one of the things that we're very curious about and researching right now, is whether you can use antiretroviral therapy together with checkpoint inhibition, and basically enhance checkpoint inhibition, because they should work hand in hand, because you're going to be activating the adaptive immune system and also regulating the innate immune system. I think with cancer, the story that's emerging right now, it's not so much about whether it activates innate immune system or not, but it's more about genomic instability- it’s the ability of repeats to integrate into the genome. And the other thing that I want to mention is that tumor suppressor genes, also regulate repeat expression. So when you have mutations in BRCA1 and p53, you have upregulation of repeat expression.
So, Rosana, the use of what you're learning in regulating or responding to changes to the Dark Genome and cancer seems really complex. And is that a reflection of just how complex cancer is as a constellation of diseases?
I think that you hit the nail on the head. Yes, I think that's part of it. And I think that's also because we're just learning about the role of repeats in cancer, and it's multi dimensional. So we have been talking about the role of the transcripts activating the innate immune system. Okay, so this is one aspect of it. The other aspect that we didn't talk about, I mentioned that LINE1 is an active retrotransposon transposable element in what several groups have shown now is that LINE1 is highly active in cancer. And that actually, you'll see more LINE1 insertions as the cancer progresses from the primary tumor into metastasis. The question for us, is that is that cause or consequence? Correct? Or people ask me this question all the time is, you know, the, the repeatome driving, or is it a passenger, and the way that I always like to refer to it, is that I believe it is a passenger with the foot in the gas. So there may be something else that initiates it, but without the repeat activity. For progression of the cancer, you're not going to get the type of proliferation metastasis that to get in the presence of an active LINE1 element and other elements. The other thing that I want to add is that LINE1, not only retrotransposons itself, is also responsible for retro transposing elements that don't have intrinsic retrotransposition capacity. So it can retrotransposons, SINEs and Alus, and so on, and also reverse transcribe them all. It seems that one of the day jobs for tumor suppressor genes like p53 and BRCA1. One is to basically keep repeat expression under control. So in cancers that have a p52 mutation or BRCA1 mutation, you see any increase in repeat expression. Again, is it cause or effect? And this is another piece that we're investigating right now, with our academic collaborators.
It seems like there's a lot of open fundamental research threads here. And I would like to return to that, like how you think about doing fundamental research in the context of a company. But you also mentioned something I want to double click on now, which is the relationship between aging and the repeatome. And it's an interesting link, because at least I think of aging is much more complicated than cancer and cancer is a disease state. There are many, many different types of cancer, but I don't even know how to name the types of aging or the ways in which we age or it's accelerated or decelerated. So, for me, it's strange that aging could be related to the repeatome. But is it more or less complex in that relationship than cancer?
I think it's more straightforward. I think that the seminal work from John Sedivy has pointed to a path in which there is a failure of the repeat surveillance system in aging, so that you get more active repeat expression as you age. And because of that, as you express more repeats, you activate inflammatory pathways that become chronic. And that's also part of aging. Also this: over active repeats seem to be involved in senescence. And what John has been able to show is that if you use antiretroviral therapy to be more specific nucleoside reverse transcriptase inhibitors that block LINE1 activity, you can also block senescence, at least in vitro. So this is an area of active interest in academia, in how antiretroviral therapy may play a role in basically delaying the progression of senescence and aging. This is still an emerging field, but I think it's going to be much more straightforward for cancer in that regard.
One of the things that I've noticed over my 20 years in life sciences is that there are some fields and some moments in time, where a community is able to organize itself and create a coherent, intellectual agenda. I remember last year on Theory & Practice, we had David Altshuler and it was just breathtaking, to watch how you had the Human Genome Project, and then the HapMap project, and then GWAS, and then sequencing. And you know, it was just this kind of amazing period of progress, because that community was able to organize itself. How do you see this playing out in the world of the Dark Genome? Do we have a set of organizing principles and efforts that are guiding the field? And if not, why not? And what would it take to get there?
You hit the nail on the head; I think this is going to be required for us to advance this field where it needs to go. Right now, there are different groups of people focusing on their little space, you know, the LINE1 folks organized through meetings and they talk about LINE1; and HERV folks organize meetings and talk about HERVs; the LINE1, folks don't talk to the HERV folks and vice versa. So there is no organization in terms of what are the overarching principles, as you say, and then how we can organize the different disciplines so they can collaborate and converge into a common arena. And that needs to happen now, so that we can actually get to an explosion of knowledge around repeats. And, you know, we call it the repeatome. Others call it the mobilome. I think that the technologies are here today that can help us investigate those repeats to help investigate their role. But if we don't get the community to really engage and create that very unified front, this is still going to continue to progress very slowly.
And you know, just to kind of dig into that for a second. How much of the lack of kind of coherent intellectual agenda is due to a lack of leadership? Is it that we're missing a small number of people that can really paint the vision and rally the community? Or is there something more structural or fundamental that's missing?
I think it's because people came into it from different aspects, and people haven't connected the dots yet. And the other piece is that there hasn't been a real effort, maybe now there's going to be an effort, let's unlock the Dark Genome, and understand its functioning human health and disease and make this a mandate. You know, I mean, if you think about the Human Genome Project, that was a mandate; we got to all the greatest minds in the world and say, “We want to sequence the human genome, we want to understand what's in there.” And I was at Millennium at the time. And I have to say, it was amazing to watch how it was done. And I think that something like that is required right now. And I think the ENCODE project was a great start. I think that there are other programs that are happening right now. The telomere to telomere consortium now has mapped a lot of the Dark Genome with something that we needed, you know, we don't even have basic things. We can't even map the repeats into the genome, you know; this is basic, you know, we saw that you cannot even start discovery. And I think this is going to change a lot, but we need to organize the community, this is a call to arms, and let's break darkness and bring light into the Dark Genome.
So let's switch gears a little bit from talking about the Dark Genome and actually talking about you as a person a little bit. You know, tell our listeners a little bit about how you actually came to be an entrepreneur in residence at GV. In fact, you were our first EIR, so what brought you to us?
So, before joining GV, I was the founding CEO of Nimbus Therapeutics and all my career has been about bringing new medicines to patients. And do that using technology. I was always very interested in cutting edge technology. So Nimbus was all about computational chemistry, how can we use computational chemistry to drive the drug discovery process. And after eight years at Nimbus, and quite a bit of success, we have now three drugs in the clinic partnership with Gilead and Celgene and all kinds of good things happen. I was really interested in what's next. And I was interested either in doing something with new technologies, so - machine learning. And the other part I was very interested in is what is out there that nobody is working on, that can be, sort of, the new area of research that we can find novel targets, and really create new medicines to treat underserved diseases right now. And when I was leaving Nimbus, I called Krishna Yeshwant, who I had met, you know, when we were trying to finance Nimbus. And the first was a GV, I basically told that the whole GV team how Krishna did not decide to invest in Nimbus, you know, and how he missed the opportunity to have a 10x return. And I didn't realize how embarrassing this whole conversation was. And I didn't also realize why they hadn't invested on it, and which now makes complete sense to me. But over the years of trying to convince Krishna to invest in Nimbus, we developed a relationship. So when I was leaving Nimbus, I called him up and I said, “Look, I'm looking for my next opportunity, I could go and join another one of the traditional VC groups, but I would love to hear more about what GV is doing.” And we went for dinner, and we had a wonderful conversation at dinner. And I said to him, why not bring me in as an EIR. And Krishna turned to me and said, “We don't have an EIR program.” I said, “What a great opportunity to start one,” you know. And that conversation evolved for about nine months. And then at the end of it, I did join GV, and I never looked back. I mean, that was the best decision ever in my life. And I think that you guys have really built a tremendous team and are doing things in a in a quite different way. And I really appreciate that.
It’s speaking truth, you know, which, which certainly is something that you do extremely well. And I appreciate. So, some more about you. You know, I read that in 2012 viewer on the Boston Business Journal, Women to Watch list and you know, they were right, like, for you to be watched, because you've done incredible things. And I think, at least from where I sit, it's because you like taking on new challenges, right? I mean, you mentioned explicitly, what's something that literally nobody is doing, I mean, how much more audacious and ambitious and scary is that right now you're the CEO of a company that's, that's doing that with a mission to unlock, you know, even the word “unlock” - the Dark Genome implies, like level of ambition, and passion. And right now you're a CEO in a small group of just 10% of female CEOs of biopharma companies. Why do you think that this lack of gender diversity is persisting, even today?
Complex question. Let me take a step back here, because I think that we need to understand the evolution of gender diversity in our industry. 20 years ago, there were very few women, basically, you counted the Vicki Satos of the world and Deborah Dunsires of the world, basically, they would go into a room and have a lot of men around them. I think that with my generation, you started seeing more and more women rise to the top, you know, and nowadays, if you look around, there are more and more women in the C suite. I think that when it comes to the CEO level, it’s a question of, how are the investors going to trust women to basically return on their investment? And I think this goes back to the basics that, “like hires like,” so if you look at the investor community, most of the investors are men. You count basically in two hands, the female investors in R&D, it's actually improving. I have to say that. I think that the number of female CEOs will only increase when the number are female investors increase. I don't think there's any other way that this is going to play out. And I think that investors are making a lot of effort to go out there and recruit female CEOs and female CSOs and female board members. But I don't see the same level of interest in recruiting female investors. And if that doesn't change, the field is not going to change. I think that the idea of a glass ceiling that is not there anymore, but it's this idea of convincing people to trust you with their money. I think that's why you don't see a large number of women CEOs out there.
So it's structural in how these companies are financed in the first place. That is the problem.
I think so. And I think it's human. It's…look, everything we do. It's about our tribes. You know, we're talking a lot about inclusion. And we talk a lot about diversity. But let's be honest, look at your circle of friends, they all look like you. You know, your family looks like you, we still haven't done a good job, even in our personal lives, to diversify. When you get into a job, it's even more pronounced. So I think that once you have more female investors, more investors of color, you're going to see that diversity at the top change. Because “like hires like.” Look, look at my company ROME. 70% of my management team are women. We have to make an effort to hire man. It's serious. We actually made that concerted effort.
Yeah, that's so interesting. It reminds me of this quote from E.O. Wilson. The real problem with humanity is the following: “We have Paleolithic emotions, medieval institutions, and godlike technology.”
Yeah, agreed, agreed.
Rosana, I kind of want to just close out with a little bit about the future and where you're going with building ROME Therapeutics. And in fact, you know, how you're trying to take the theory of the Dark Genome, and then translate it into practice. So you know, when you actually think about trying to make a drug that goes after the Dark Genome, how do you approach this question? What are the modalities that are most promising? Is it gene editing? Is it small molecules? Is it antibodies? Is it something else?
That is a great question. So this is a new field, but I'm super pragmatic about how you get medicines to patients. And when we started ROME, we already started with that idea. And I have to say that we are currently focusing in small molecules and antibodies, because a lot of the targets that we have now are, in a way, traditional drug targets, even though it's the first time they're being drugged in humans. So for example, we are focusing right now in drugging LINE1 reverse transcriptase. And the reverse transcriptase activity is very similar to the reverse transcriptase activity that you see in viruses; we just have to generate compounds that are potent, specific and selective against the LINE1 RT versus the viral reverse transcriptase inhibitors. That's what we're doing. So being very pragmatic, we are going after antigen targets, because as I mentioned before, the HERVs make envelope proteins that are overexpressed in certain types of cancers, we are going after host targets, you know, that are responsible for splicing of this transcript. So there is a variety of the low hanging fruits that we're going after. This is going to give us the space to basically do a lot of work in finding novel targets that are repeatome based. And to do that, we're building a very strong data sciences team. So the way that I see that this is going to go hand in hand, is that ROME is the intersection of a therapeutics company with a data sciences company. That's the only way that we're going to be able to do that both in terms of identifying novel repeatome targets, but also identifying the right patients for which these medicines will have an impact. So that's how we are translating the theory to practice. Our first drug is going to be based or modelled in anti retroviral drugs that go after reverse transcriptase as a target. And the beauty of that is that we have 40/50 years of experience with reverse transcriptase inhibitors.
I think it's interesting to keep the risk focused on just the unknown part and be as pragmatic as possible about the rest of the therapeutics story. But if I could ask you to leave pragmatism behind a little bit and look forward like 10 or 20 years, do you think that we'll see interventions in the Dark Genome not for disease treatment, but for disease prevention?
Absolutely. I think that this is going to be the big future here is that we could use it to prevent diseases, like for example, David Ting and I, in the very beginning of our collaboration on the ROME idea and so on, he used to tell me that if you could modulate repeats that may be blocking reverse transcriptase - the repeatome called reverse transcriptase - that could be the aspirin for cancer, because if you can block insertion to the genome, you can block activation of the innate immune response or whatever it is that is involved in the pathogenesis, you could see that could be used prophylactically to prevent. And I think that this is going to be true, not only for cancer, but autoimmune diseases and neurodegeneration, and maybe possibly aging.
Rosana, thank you so much. This has been a fascinating conversation. You know, over the last few years at GV, you've been one of the people that I am most valued as a friend and as a thought partner. And you know, I think this is just a wonderful chance for our listeners to get to know you the way that I have. So thank you so much for doing it.
You're welcome. Can I leave you with one additional thought? In terms of therapeutic modalities, which you guys did not ask me. So I said that I would start with small molecules and antibodies. But the one area that I'm super interested in is RNA editing; you're not going to be able to do editing at the genome, but your RNA editing could be a big new therapeutic modality to interfere with repeat function. Just leaving it there.
Wow. We'll be back in season four with a successful demonstration of RNA editing by Rosana at ROME Therapeutics. Exactly.
I'm just saying, if you're saying about future blue sky, what I say out there: RNA editing.
Thanks so much for your time.
Guys. This was fun.
So, you know, Alex, that was such a great interview with Rosana, I'm really excited to kick off this new season of Theory & Practice. And I thought it might be nice to use this time at the end, to talk a little bit about how we're thinking about the third series, and topics that we'll be exploring and why we're choosing to explore them.
That sounds great. So I guess for this one, we'll set aside the normal Hammer & Nail discussion. Maybe we'll talk about toolboxes or exactly different fields of carpentry or something like that.
Yeah, let's go meta for a second. And well, actually, sorry, I shouldn't say “go meta” - a different use of that phrase at our current time. Let's go high level. Yeah. 30,000 foot view. And talk about how we're thinking about this series. You know, I think one of the things that you and I heard first, I mean, it's been great to get a lot of positive feedback and a lot of input on the first two series. And one of the things that we heard from people was, although they really liked sometimes getting to know the guests on a personal level, what was often even more exciting, was not just their personal trajectory in life, but actually diving deep into an intellectual topic. And so I think this series, we're going to be focusing more on intellectual topics than we are on just the biographies of the people that we’re interviewing, which I think is a great change.
Yeah, absolutely. And I think one thing that we heard a lot of last season that we're definitely going to hear this season, are people who were there for the births of the fields, and the discoveries that they're known for. It's kind of a weird concept to think about, like a field of science not existing. And then it comes into existence, that birth process can be quick, or sometimes it can take decades. But this season, I think we're going to be hearing a lot more about that and seeing people straddling fields that are coming into existence or actively helping to bring them into existence. And I think our conversation with Rosana is exactly one of those things. She's working in this field of dark genomics, which is kind of a field but it's not quite established as much as at least she would like it to be established.
I mean, totally. And actually, again, I thought that was really interesting when she was touching upon it at the end, when she was talking about challenges, not just for her company, or for her own intellectual research interests, but actually kind of for the field as a whole. And I think there are a few things that are worth diving into today. So you know, the first is how a new field gets born. And then also how the way it's set up at birth, can often follow it through into its adolescence and old age. To choose one example, you think about the birth of molecular biology. And you know, it's kind of amazing when I started med school, there were still a few people around that actually got to witness the birth of molecular biology. And they kind of would recount the time where the medical schools focus was one much more on anatomy and physiology. And these things that almost aren't even academic departments. And then suddenly, the molecular biology revolution hit, and there was this huge pendulum swing towards creating that field. And what you saw was a whole new type of scientist that was born in terms of people who are working on true biochemistry. And you also saw certain cultural traits that were different than their academic ancestors. You know, Watson and Crick were very competitive and kind of view that they were in a race. And a lot of the initial generation of molecular biologists carried that ethos forward, and were very intellectually aggressive and competitive. And that was kind of part of the way that molecular biology worked as a culture. Does that make sense?
Yeah, absolutely. To me, it's strange that like, you can think of molecular biology like not existing at a point of time, because when I was kind of coming up in, in my training, that was kind of like asking a fish about water, right, it pervaded not just the kind of core set of practitioners, but its ethos, and its tooling. And its perspective on science touches fields that were much farther than just kind of the set of core researchers. So it's interesting to think about the birth of these fields, there's a really good book on this topic for those that that care called, “Operators and promoters,” which kind of details some of the early work that led to the establishment of this field. It was kind of, at first, this unholy union of chemistry and biology; it's weird to think but those things didn't fit together, at least in the minds of people doing science at the time.
I love the phrase asking a fish about water, we grew up with it, at least the way in which I can understand it is that in my own lifetime, I was lucky enough to get to watch the birth of genomics as a field. The human genome was sequenced in the year that I started med school. And I still remember one of my professors asking me about research interests. And I said I was really interested in in genomics. And I remember that person kind of being snide and saying, genomics is not a discipline, it's a tool, you should find a discipline. And you know, to be fair, at that time, it was true. It wasn't an intellectual discipline unto itself, it was a tool to do genetics. But you know, sure enough, over the coming years, genomics actually matured into a field unto itself. And actually, since then, you've started to see lots of subfields start to emerge, whether it be single cell genomics, like when we talked about Aviv Regev last year, or the field of human genetics and human genomics, which we talked about with David Altshuler. And so it's interesting to watch a field give birth to subfields. And going back to this idea that the way a field is founded, can often permeate into its future. You know, one of the rallying cries in the early days of genomics was that the data should be made available to all and you know, you had Celera on the private side versus the Human Genome Project on the public side. And the people on the public side, that data sharing was almost the reason why they should exist. And what you see is that as Genomics move forward as a field, data sharing, ended up happening much more there than in other branches of medical research. You know, the initial generation of people doing GWASs got together and agreed to share data and do meta analyses relatively early on. And you know, conversely, you look at fields like clinical trials, which we've explored on other episodes in the past, that's a field that still, you know, it's really reluctant to let anybody see their trial datasets. And very rarely do you see a meta analysis happen across different trials. So it's interesting to me how the way a field is formed, can actually kind of carry forward generations later.
Absolutely, in, I think that there's parallels between the field that you saw birthed when you were training and one that I saw, it wasn't birth, but it came into prominence, which is the use of machine learning in biology. So I remember reading the first ImageNet paper, I remember reading the first TSNE papers. And it was just a shock to me that these, you know, programs could do these things, like, detect objects inside of an image. I mean, at the time, I didn't know that there was a long history of computer vision coming before it, but it was really taken with the idea of what these new models could do. And everybody that I talked to about it said, you know, that's just some fancy algorithm, like stick to what you know, like, stick to means and averages and standard deviations and, and things like that, obviously, machine learning and deep learning have really taken science writ large by storm, and they've been incredibly useful. In the birth of that field is also interesting. And I think we can draw a parallel to genomics. Everybody in machine learning, publishes everything in the open. Yeah, all the code is available. Generally, people put their papers up on archive for free view. All of the conferences are open, and they're staffed by volunteers. The entire field of machine learning owns its own publishing stack, which is actually quite unique among scientific disciplines. It's just permeated through the field in the discipline of machine learning. And so part of that is kind of blending into biology actually, and you know, there wasn't bioRxiv before which is this preprint service where you can put papers up online before they're peer reviewed, that would have been shocking to propose something like that to your advisor 10 years ago, but now people routinely share their research before it's actually peer reviewed and published. And I think that's explicitly, directly because the ethos of computer scientists and machine learners have begun to influence biologists in small and subtle ways.
Like I said, I think this is going to be a recurrent motif, leitmotif if you will, throughout this season. Later on, we're going to talk to Karl Deisseroth. And I have to say it was just a thrill to hear him talk about the birth of optogenetics, which is a field that you are kind of had a front row seat in during your PhD.
Absolutely. And it was exactly one of those things. You mentioned that at the time, you know, it was considered a tool or an interesting tool. People weren't even sure it was working. And then it was a tool that absolutely took neuroscience by storm. And if you're doing Systems Neuroscience, you have to have that tool in your toolbox. And that transition from obscurity to proof of concept and then to this just firestorm where it's taking over the field was really wonderful to watch. And I think folks will enjoy the conversation with Karl as well.
Excellent. I think this is going to be a great series. Alex, really looking forward to doing this with you for our third series.
Third series, back strong.
Exactly, exactly. Alright, my friend. Until next time,
Until next time, Anthony that was fun. Next episode, we will be speaking to Professor David Liu from Harvard University about gene editing. Then later in the series, we'll be discussing diverse topics that we think will be impactful in the next 10 to 20 years, such as the exciting developments in cancer medicine, our new understanding of how we think and feel protein folding, and the molecular and cellular basis of aging. If you've got any questions for us, or our guests email email@example.com, or tweet the GV team @gvteam, we'd love to hear from you.
This is the GV podcast and a Blanchard House production. Our science producer was Hilary Guite. Executive producer Rosie Pye with music by Dalo. I'm Anthony Philippakis. I'm Alex Wiltschko. And this is Theory & Practice.