CEOInterviews.AI
Start App
Get an email when George Church gives a new interview
One email per new interview. Unsubscribe in one click.
George Church
Professor of Genetics, Wyss Institute at Harvard University

From Launching the Human Genome Project to the Redesigning the Life Itself with Prof. George Church

📅 Oct 17, 2025 UM iGEM 73 MIN 202 VIEWS 71 SEGMENTS · 2 SPEAKERS
In this episode, we are joined by Professor George Church (Harvard Medical School, Wyss Institute, Colossal Biosciences) – a pioneering scientist who helped launch the fields of genomics, synthetic biology, and genome engineering. We explore his revolutionary work on the Human Genome Project, CRISPR technology, and the future of de-extinction, aging reversal, co-founding companies and engineering life. From sequencing the first human genome to designing new species, George Church shares his bold vision for the future of biology, and humanity itself. What you’ll learn: • How George Church’s p...

Questions asked in this interview

12
  1. 0:56So why and how did that happen, actually?
  2. 3:03So what did you actually learn from this setback at Duke?
  3. 6:25And if you had to say the most important thing you learned from all of these very good researchers at that time, what would it be? One thing?
  4. 10:46But would maybe doctors in the future be also looking at your genome when you just go for a checkup?
  5. 13:19So how can we better communicate this to a wider public so they're kind of more prone to it?
  6. 19:38This all happened within the university, within the lab, but do you think that this kind of DNA storage can become practical for real world use?
  7. 26:30And do you actually think that we are close to solving these issues now with the new technologies that we have, and with basically all of the stuff that we're doing now with AI, maybe?
  8. 32:57Maybe, for example, with a mosquito to wipe out mosquito-borne diseases?
  9. 38:31But in what ways do you think biotech can have the greatest impact on, yeah, on climate change and all of the pressing issues that are associated with it?
  10. 48:38So do you think that the technology development is underappreciated in academia relative to just the single discoveries?
  11. 54:42So what would you say is the ultimate goal of these kinds of de-extinction projects that Colossal does?
  12. 1:02:30Like, yeah, like the shift of the current, current ecological, like ecosystem or something like this, or do you think that it has more benefits?
Misha Bagirov 0:06 ↗
Hi everyone, I'm Misha and I'm a member of the 20125 team from Maastricht University CoreSpin. Within our project, we're introducing the IGM interview series where we want to bring the state-of-the-art synthetic biology closer to you. In each episode, we sit down with leading scientists, engineers, entrepreneurs to explore their groundbreaking research, novel applications, and the thoughts that are shaping the synthetic biology of tomorrow. Please, let's dive in.
Hello, George. It's truly an honor to have you as our guest today for our interview series. I'm Misha, an MSc student focusing on synthetic biology. And yeah, thank you very much for joining us.
George Church 0:55 ↗
My pleasure.
Misha Bagirov 0:56 ↗
So, first, let's go back to the very beginning of your scientific career. You entered your bachelor's at Duke University and completed a double major in zoology and chemistry in just two years. That's an unusually rapid start. So why and how did that happen, actually?
George Church 1:18 ↗
Yeah. So I had some advanced placement from high school and I didn't want to spend a lot of money getting an undergraduate education. I wanted to get into graduate school where they would pay me to do research, and I basically wanted to do research. And I figured I would learn as I went along. So it didn't feel like rushing. It just felt like getting to the next stage, you know.
Misha Bagirov 1:49 ↗
Okay, all right. So, and basically in these two years, I would imagine for us as undergraduates, you get the very beginning, the very first touch on your scientific career at that time. So did it shape your outlook in science in any way?
George Church 2:17 ↗
Well, I had done research in high school, independent research where I had keys to the chemistry lab, but I really wanted to get into... I also did research as an undergraduate around the courses. I took some courses that had a lab versus course option, and I would always take the optional lab course even though it didn't affect the grade. And also I would take graduate courses, and I worked in a lab for a salary as well. So I was getting research from a lot of different angles during those years, but it transformed another notch when I finally got to do full-time research in graduate school.
Misha Bagirov 3:03 ↗
Yeah, yeah. So now talking about graduate school at Duke, as I read, you dedicated a lot of time, like even you said up to 100 hours, to research so that you didn't really care, didn't really focus on the coursework, and you were withdrawn from the program. But you continued, you completed your PhD at Harvard with the Nobel laureate Wally Gilbert. So what did you actually learn from this setback at Duke?
George Church 3:44 ↗
Well, I never really felt it was a setback. I just kind of kept doing stuff. The reason I had gone to graduate school is so I could be a full-time researcher. And then they told me I had to take the courses I had already taken, more or equivalent courses, and I just said no, I already took those as an undergraduate, same school, professors. And so I said I want to do full-time research, and I did it. And then when they dropped me out of the program, I just did full-time research as a technician. And then my adviser, wonderful adviser Sono Kim, said, you know, you really should try a PhD again. And so then I fortunately got into Harvard. It was the only place I applied. But anyway, then I had to take courses again for the third time. But this time I said, okay, I will spend some time doing the courses.
Misha Bagirov 4:39 ↗
Yeah, that's fair. And also talking a bit more about the Harvard time, Gilbert's lab is just an amazing place, I would imagine, right? With the co-invention of the first DNA sequencing with Sanger and everything, the novel techniques. So what was it like to be a young researcher in this era of the breakthrough technologies when we got the first sequencing methods and basically modern biochemistry and molecular biology was developing?
George Church 5:28 ↗
Yeah. Well, molecular biology was being turned into a technology. It was not just reading DNA but also writing it. So the whole recombinant DNA was happening all around me. It really laid the groundwork for next-gen sequencing and synthetic biology, two things I was going to really focus on in the future. I mean, it wasn't obvious then, but looking backwards you can see how it could have helped you think about it. And it wasn't just Wally. I mean, we were surrounded by very talented students and professors, you know, Tomatis and Mark Ptashne, Nancy Kleckner, Matt Meselson. These were some of the labs that I rotated in.
Misha Bagirov 6:25 ↗
And if you had to say the most important thing you learned from all of these very good researchers at that time, what would it be? One thing?
George Church 6:37 ↗
Well, from Wally, you know, I got reinforcement of two of my natural tendencies, which is towards technology development and being able to work in multiple fields at once. Sort of the uber multidisciplinarity showed that that could be made to work. I'm not sure that he or I did it completely consciously, but it was very reinforcing. Boldness, you know, don't do something small. It's going to take almost exactly the same amount of time to do something big. So yeah, those are some of the lessons.
Misha Bagirov 7:20 ↗
Nice, nice, nice. And jumping forward to 1984, at a meeting with the Department of Energy, you and some scientists proposed sequencing an entire genome, like basically the idea of the Human Genome Project was first established. So how did you and your colleagues manage to launch the Human Genome Project?
George Church 7:52 ↗
Well, it kind of fell in our lap in a certain sense. It was kind of an accident. The meeting was supposed to be on estimating mutation rates in populations, and we felt we couldn't do that without at least a reference genome to say how they were different. And then we had all kind of been independently thinking a little bit on the edges of that, but we kind of all put it together and said, well, we could do a genome. And then we described how we would do it. And then that went back to Department of Energy headquarters and they thought it was great. And you know, there were only about 10 scientists in the room, and a couple months later they were writing checks to get the genome project going. So that was 1985-ish, and it really was going quite well under the Department of Energy with technology development, computer robots, all kinds of things. And then NIH got in the game. They didn't want DOE to be responsible for health care or responsible for the greatest breakthrough in health care. And then I think it went downhill from there for a while, innovation was downplayed and production was up. And then it wasn't really until 2004 that victory was declared before it was actually done. And that allowed us to get back to doing technology development. And almost overnight, as soon as they opened the door, we got to NextGen Sequencing, which changed it from a $3 billion project to a $1,000 or $100 project.
Misha Bagirov 9:43 ↗
Yeah, I just wanted to ask you about this actually. So how did that happen? Actually, if we think about it, this is just insane, right? Something first costs like $3 billion as you mentioned, and then it goes down to a couple of hundred dollars.
George Church 10:05 ↗
And the quality went up too. So it wasn't just bringing down the price. It went from three billion base pairs, which isn't even a practical genome in a clinical sense. You need to have both your parents' contributions, which is three billion, which is six billion. So it should have been twice as big. And it's actually harder to separate two genomes that are that closely related and tell which one came from the mother and father. All that's important. So it's a very low-quality genome in every way, and it had lots of gaps in it. Anyway, now we can make nearly perfect separation of maternal and paternal chromosomes.
Misha Bagirov 10:46 ↗
And looking ahead a little bit, how would you see maybe these technologies integrating in routine healthcare? For example, we now take blood tests or all kinds of other tests. But would maybe doctors in the future be also looking at your genome when you just go for a checkup?
George Church 11:14 ↗
Well, certainly in the future, and to some extent already in the present, a large fraction of cancers should and are — you do genomes both preventatively and also of the cancer itself, if you've already got cancer.
Misha Bagirov 11:34 ↗
And also as a next step in the Human Genome Project, in 2005 you launched the Personal Genome Project to sequence the genomes of different people and make the data actually publicly available. And at that time, and I feel like we still are worried a bit about genetic privacy. So what actually motivated you to push for open access personal genetics?
George Church 12:05 ↗
Well, it wasn't that I wanted everybody in the world to forget about privacy. It was more a matter of if we were going to be sharing the data, which everybody wanted to do, we should properly consent people. We should allow them to look at their own data, which was not possible before my project. We should allow them to share their own data if they want to. But most importantly, we should communicate to them not just the possibility but the likelihood of it escaping. While most consenting at that time said we can't tell you what we found, you know, and we will keep it secret from everybody else, when in fact what happened is because of hacking and lost disc drives and whatnot, it could be that lots of people know your data and you don't. So it was really my reaction to human subjects consenting. Now, in terms of privacy, I think now we can make — with things like homomorphic encryption queries, we can ask questions about genomes without decrypting them. So if they're encrypted, you can make it so that nobody can read it. They can ask a very limited set of questions about it.
Misha Bagirov 13:19 ↗
Yeah, that's a fairly good point. But still, I guess it's a problem of communication of these technologies a little bit, because when I prepared for the interview, I asked my parents what they think about the Personal Genome Project, would they like to potentially share their information with the scientists, and they were also kind of skeptical. So how can we better communicate this to a wider public so they're kind of more prone to it?
George Church 14:10 ↗
Again, the objective was not to get 8 billion people to share their data. It was to test new technologies, including multiomics technologies, on a cohort that knew what they were getting into and could prove that they did by getting 100% on an exam that showed that they knew what they're getting into. It's going to be a tiny subset of society. But as far as I can tell, there's a diverse — you know, there are people from almost every ancestry that are willing to do this. You don't need everybody. But furthermore, the sharing question is not whether you're going to share it with scientists, it's whether you're going to share it with hospital staff. Scientists for the most part couldn't care less who your name is. But hospital staff, if a celebrity comes in, they're going to poke around and find out what's going on, which they shouldn't be, but they do. So I think people do share all kinds of medical data, not just sequences. And the medical data is actually in a way more personal, more private than your genome, because your genome might say that you're at risk for Alzheimer's — everybody is if they live long enough — but your medical records will say, oh, you have STDs and you have early onset dementia and etc. So anyway, I think it's a personal decision, and we do need to communicate better. It's interesting that people aren't getting their own — it's not a matter of sharing the data, they don't even want to share it with themselves. They don't want to get their own genome sequenced. It could save trillions of dollars. It could save personal family anxiety. And if you have someone with a very severe genetic disease, then somebody in the family often has to quit their job, sometimes both parents anyway. There's going to be a lot of education in the future. And there are many technologies that take a while for people to understand what they're good for. For example, computers. When I was young, I thought computers were great. And I asked people if they wanted them and they said no way. And now you can't get it loose from them, you know. So it's just a matter of seeing the benefits and just evaluating for yourself what's good and what's not, and maybe for the bigger good.
Misha Bagirov 16:55 ↗
Yeah, yeah. And also talking more about applications of DNA, you've also used DNA for data storage. And famously you included your full entire book, Regenesis, with more than 300 pages with images, into synthetic DNA. And it's just insane — just a drop of this DNA liquid, as I read it, contains 70 billion copies of the text. So what actually inspired you to do this experiment, and how did you actually do it?
George Church 17:37 ↗
Yeah. So I had been working on reading and writing DNA basically since I was a graduate student that we talked about earlier, and it just got exponentially better every year. And then one year I realized it was good enough that we could store a book in it. And I wrote a review for an article that I wrote in DNA, and the editors let it go through. So I had the software already done just for that stunt of reviewing an article on synthetic biology and DNA. And so it was fairly easy. I mean, we used two of the techniques from the lab. One was to write it using a chip-based synthesis which we had published in 2004, and then amplify it. So actually one of the advantages, you could make billions of copies almost for free, for pennies. And then to read it with fluorescent next-gen sequencing, which was one of the two main methods that I had helped develop, the other one being nanopore, which wasn't quite ready at the time. This is around 2014. Now it is totally ready. And then compare the input with the output and saw that they were basically the same. We had error checking capabilities which made it so you really didn't have to have very accurate reading or writing because of error correction. But it was really on the order of 99-plus percent for both reading and writing raw data. That's just for one read, and then the redundancy went up, got exponentially more accurate.
Misha Bagirov 19:38 ↗
That's insane. Insane. And do you think for the real world use — like we talked about this, right? This all happened within the university, within the lab, but do you think that this kind of DNA storage can become practical for real world use?
George Church 20:02 ↗
So outside of academia, it's mostly outside of academia after that paper, where there's an international consortium that's trying to do this. I think the first applications are not going to be the top priority of that consortium, which was essentially writing digital data into DNA. I think it's going to be writing biological data into DNA because it's an easier route, it's a much cheaper route. You don't have to go through the interface between digital and biological. So for example, we and others have demonstrated you can record all kinds of developmental lineage data and physiological data into DNA or RNA and even proteins, and then you can read it out in a way where you go to a particular part of the body and you know that data was from that body part, and you can get a time record and so forth. So I think that's the lowest hanging fruit. It's not that low hanging, but it is lower hanging, I think, than trying to compete with disc drives. I think for a while, you can get potentially exabytes of data in a living system, while writing exabytes with the digital-to-biological route that we described in my first paper is still out of reach. But doing it entirely biologically, you can record that level, and then you can do selective retrieval and sequencing.
Misha Bagirov 21:48 ↗
Yeah. And also over the past few decades, you've been working on an astonishing range of biomedical advances. Everything from the Human Genome Project that we talked about, to CRISPR gene editing, and also personal genomics, right, the age reversal and the de-extinction of animals. And my question is, how do you decide — so this is the next bit, the next biggest thing that I want to work on, especially when these areas might not have immediate payoffs like that?
George Church 22:35 ↗
Yeah, well, it's been my experience that things tend to pay off faster than you expect rather than slower in this particular set of fields. These are exponential fields, and so people look at the short term and it seems very slow, but just even a moderate term, the exponential is surprising. And that's computing and AI and reading and writing DNA and cellular — anything that comes from reading and writing DNA, all gone exponential by 10 million fold or more. So the way you pick then is, you can either go forward where you say, you know, what combination of new methods can be put into an even newer method. So like we just went through taking reading and writing DNA and integrating them, or you can take cell biology and barcoding, the multiplexing that made both reading and writing go exponential. You multiplex viral delivery vector development. You can multiplex cells. So you can do millions or billions or even in some cases trillions of assays all in one tube. And then you just pick the winners. You can't necessarily sequence trillions in every case, but you can pick the winners and sequence those. So you don't need to see the immediate payoffs. Although I think in general some things were faster than others. You know, fluorescent next-gen sequencing was picked up much faster than nanopore. CRISPR was much faster than de-extinction. Both required editing, and we would kind of collect ideas for why you would need multiplexing. You know, multiplexing was driving the radical reduction in cost and improvement of quality, and we'd say, well, we need multiplex editing for recoding genomes, for humanizing, for the extinction of animals and so forth. And then we would just go through and try to make them happen.
Misha Bagirov 25:09 ↗
So do you plan on revisiting maybe the original genome project? And you're still continuing, right?
George Church 25:19 ↗
Oh yeah. So I mean, with most of these exponential technologies, as soon as it becomes available, you can redo everything that's been done in a short period of time inexpensively. So for example, in 2009 we published a paper where the genome project was just finished in 2004. In 2009 we did five genomes, diploid genomes, so they were twice as big, so really 10 times as much data with higher quality. And then today we are aiming to do all 8 billion, or at least the subset of the 8 billion that want to have better health care. And you can see it happening in places like the UAE, the United Kingdom, and Iceland. The UAE is getting close to getting its whole population sequenced. And the United Kingdom is close to having the same number of people, but it's a smaller fraction of their total population, you know, on the order of a couple million people. But it's going to be billions if, you know, assuming people see the health and financial benefits.
Misha Bagirov 26:30 ↗
And now going to another area of your research, CRISPR-Cas9 systems. CRISPR is definitely a revolutionary technology, but the technical complications currently, like off-target effects, are still being mitigated, right? And the delivery methods of the CRISPR therapies still haven't been worked out fully yet. And do you actually think that we are close to solving these issues now with the new technologies that we have, and with basically all of the stuff that we're doing now with AI, maybe?
George Church 27:15 ↗
Yeah. So we were working on other editing methods before CRISPR, and we're working on other ones afterwards as well. I think that we were the first one to raise the issue of off-target, I think. But it hasn't been that big a problem really. Just briefly there was an off-target problem with C-base editors, but that was quickly solved by David Liu's group by adapting the A-base editors to Cs. But the vanilla original CRISPR, you know, in 2013 that we published, pretty much had such low — if you choose your sequences carefully, could have very low off-targets. But I think multiplexing is hard, that's doing lots of edits. I think the record now for CAR-T multiplexing is on the order of six edits at once, that came from our group at Khan. But if you really want to do higher levels, then germline is a good way to go, and we do that in pigs where we can edit up to 69 edits and probably more in the near future. And you can make them almost perfect because you're doing clonal. So you have one cell that makes the embryo. And so that means that you can make that cell essentially perfect, so that you can sequence the whole genome and make sure that what you have is correct, you haven't missed anything you wanted, and you didn't get anything you didn't want. So I think that's — and even though we can't do germline in humans yet, you can definitely do germline in pigs, and then that makes its way into humans.
Misha Bagirov 29:00 ↗
Yeah, that's true. And also sticking to this topic, you have introduced, as you said, 69 or more gene edits in pig cells, and the goal was to inactivate the retroviruses in the pig genome for basically making the organs from pigs transplantable to humans. So specifically on that project for organ transplantation, how close are you now to normalizing? Because I've heard that there was a story in China in 2024, the first liver from a pig, I believe with your technology, was transplanted into a human. How close are we to normalizing transplanting gene-edited pig organs into humans on a normal basis, basically in each hospital, to end the organ shortage?
George Church 30:11 ↗
Right, well, it's more than just the organ shortage. We can make organs that are better than organs in principle, things that have better immune profiles that are not rejected the way human-to-human organs are occasionally rejected. We can make them so that they're pathogen resistant. In fact, pig organs are intrinsically resistant to all human-specific pathogens, things like the syphilis bacterium and HIV virus and so forth. But anyway, the way that it gets — we both have better organs and solve the crisis of shortage. It's already happening. We've transplanted two kidneys now, and both people are still setting records for longest survival. So one of them is now 7 months out and is pretty healthy, I mean, considering that they had, in addition to the kidney problem, other health problems, because that's how they were prioritized to be involved in the study. But as we get to healthier and healthier patients where their only problem is kidney failure, I'm sure it'll go better and better. But already these two are doing well. And kidney is probably the biggest unmet need right now worldwide, hundreds of thousands of people. So normalized — I mean, I think this, unlike getting your genome read, which seems to require some explanation as to why it's useful, this is obvious to all the patients. Almost everybody on dialysis wants to get off dialysis. They really — well, I think the difference is one is 100% guaranteed to be better. Dialysis is — you have brain fog, you have this really inconvenient, constantly going to the clinic kind of limits you geographically. While with genetics it's like, well, there's a 97% chance your kids are going to be fine, or your first child will be fine. And so in one case it's a 100% chance you're going to have a better outcome if you do the transplant, and in the other case it's like maybe a 3% chance that you'll benefit from it. And I think that's what people can't wrap their head around. But from a public health standpoint, it should be obvious, and it will be.

35 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Church, G. (2025, October 17). From Launching the Human Genome Project to the Redesigning the Life Itself with Prof. George Church [Interview transcript]. UM iGEM. CEOInterviews.AI. https://ceointerviews.ai/interview/3241792/

MLA

George Church. "From Launching the Human Genome Project to the Redesigning the Life Itself with Prof. George Church." UM iGEM, 17 Oct. 2025. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/3241792/.

BibTeX
@misc{church2025_3241792,
  author       = {George Church},
  title        = {From Launching the Human Genome Project to the Redesigning the Life Itself with Prof. George Church},
  howpublished = {Interview transcript, UM iGEM. CEOInterviews.AI},
  year         = {2025},
  month        = {oct},
  url          = {https://ceointerviews.ai/interview/3241792/},
  note         = {Speaker-attributed transcript with timestamps}
}