Scientist Stories: Harris Lewin, Mammalian Genome Evolution

My name is Harris Lewin. I was born in Brooklyn, New York in 1957, and I'm a professor of evolution and ecology at the University of California, Davis, with a laboratory in the Genome Center and a joint appointment in the veterinary school in the Department of Reproduction and Population Health.

Who got you interested in science early in life?

I had a great chemistry teacher in middle school who was really— you know, he stimulated my fascination with chemistry early on. I have to say my parents were very nurturing with respect to getting me interested in the natural world, particularly my dad. And even though we were in Brooklyn, we were on the South Shore of Brooklyn, which is was still pretty wild in those days. And, and I spent a lot of time at the American Museum of Natural History as a kid and the Bronx Zoo. And those places were very, I would say, most influential in my development of my interest in natural history and biology.

Later, you became interested in studying dairy cows. How did you come to that?

So very early on, That was— those were formed. Always interested in animals. And then as I grew up, I started to have more experiences with agriculturally important animals, and I really thought veterinary medicine was my calling. And when I had the opportunity to go to Israel when I was in high school, my sister emigrated and moved to a kibbutz. And I said, well, this is a great opportunity to get experience in agriculture.

When I—

in the '70s, you know, mid-'70s. And so I— one summer I went and worked on a poultry farm. I liked that. I thought that was fascinating. I learned a lot about raising broilers. And then 2 years later, I decided I liked large animals a lot more, and I got to work with a veterinarian there, one of the most famous veterinarians in the country, large animal veterinarians, who really stimulated my interest in in dairy cattle and dairy cattle management reproduction. And I just became fascinated with the whole biology of the cow, you know, what they were eating and how they were converting their diet, very, very low-quality forage into high-quality product, if you will, you know, milk. And that was That has not gone away. I'm still absolutely fascinated by the biology of ruminants. Our recent papers attest to the fact that this is, to me, one of the most important adaptations on the planet, the development of the ruminant, this whole class of organisms, and how important they've been to human civilization and human evolution, even. Those were very formative experiences, actually working hands-on with the animals. And then when I got to Cornell as an undergraduate, I continued and started to work in the areas of animal genetics, learned about quantitative genetics and animal breeding, and decided that statistical genetics wasn't really what I was interested in. I was interested in genetics and diseases. And that led me to UC Davis as a graduate student where I began to work on, at the time, the major histocompatibility complex and genetic resistance to infectious diseases. That was a very important area from the late, you know, from the early, mid-'70s to the, really to the mid-'80s, as the biology of the major histocompatibility complex unfolded. And so, I started to work work on the Majestic compatibility complex of cattle. And I was very lucky to fall into a laboratory and onto a floor of people in the vet school who had made major discoveries on the bovine leukemia virus. And that was fortuitous because that turned out to be a model for the human T-cell leukemia viruses and even AIDS, when AIDS broke when I was a graduate student. And so I received many years of NIH funding to study genetic control of resistance subclinical progression of bovine leukemia virus in Holstein cattle.

Who were some of your key mentors in graduate school?

Mentoring was really critical. I was very lucky as a graduate student at Cornell when I shifted from undergraduate to graduate school to have a new assistant professor named Rod Deitert at Cornell who worked on, on oncodevelopmental genes in poultry. And so I learned a lot about immunogenetics. And he was terrific because he allowed me a lot of freedom in the laboratory to explore and to make mistakes and to learn on my own without being over— you know, without dominating every hour of every day. And I learned that that was the way you learned in the laboratory, was by experimenting and making your own mistakes and learning from them. Wow. He provided great context for what I was doing. Next was my PhD mentor, Domenico Brunocco, who was a very important principal figure in the elucidation of the HLA complex, discovered the HLA-C locus. He was one of the leading geneticists in Paul Terasaki's lab at UCLA. And I inherited him by accident, because I came to Davis to work for Clyde Stormont, who was the founder— a founder in the field of animal immunogenetics and blood groups and biochemical polymorphisms. And just about every person in the world who did animal genetics had come to Davis to learn from Stormont and his team. Horse blood groups, cattle blood groups, and all other domestic animals. But I was very lucky to get Bernoko, because Bernoko had been working on the MHC, whereas Stormont had been working on blood groups. Falling into the MHC was a real blessing because there was a locus of genes that were clearly involved in the immune system. And then to be working on the bovine leukemia virus, just luck, fell into it and really took that to completion, you know, mapped the genes, sequenced the genes, found the regions of the protein that were responsible for resistance and susceptibility, developed a genetic test, and all of the things that followed. So, it really was a beautiful story, and I was very lucky along the way to have great mentors who provided the context and great mentoring on how to approach such a complicated problem in a non-inbred species. I mean, I couldn't make crosses. I had to do it almost in the human context of using, you know, they were not wild populations, but they were still not inbred animals either. And so, it really was the first example of genetic control of resistance to infectious disease in a non-inbred species. And having people who understood, you know, how to approach that problem, like Bernoko, in a non-inbred species, because of all the HLA and disease association that was going on at that time, that was the most those were the most valuable lessons I could possibly learn at that time. So, again, lucky to be in the right place at the right time with the right people.

How did the Human Genome Project impact your work?

Of course, those were very exciting years. It was really the mid-'80s, and I had, as I mentioned yesterday in my talk here, I heard a lecture from Jim Womack, and I started to hear about what Steve O'Brien was doing and what Jim was doing. And what a small group of others were doing with a mouse, which was really trying to develop comparative maps. You know, this principle of shared synteny between mammals was really pretty new, and there were very few genes. We didn't even have RFLPs in those days, or ESTs in those days. They were biochemical polymorphisms, primarily, that you could use somatic cell hybrid mapping to demonstrate that the genes were on the same chromosome of the different species. So, Jim was a— and Steve were very— their work was really important to me, because it opened my eyes to a whole— what a whole genome— what you could do with a whole genome rather than focusing on a particular locus. However important that gene was, there was so much more in the genome. And since I was interested in complex traits like lactation and reproduction, and how those could— how you could map genes for those kinds of traits that were important economically to either the dairy cattle or beef cattle industries, the large animals became a model, really, for how you approach mapping genes for complex traits. So, naturally, when they began to talk about the Human Genome Project and all of the techniques like RFLPs that started to come out and microsatellites, we quickly had to adopt those methods. And the human geneticists recognized the animal geneticists had unique models and things to contribute, populations to contribute, animals that had been under selection for individual traits for long periods of time. And so, this was a natural and very exciting opportunity to put people together to think about how we could use these new technologies that were developing out of human genetics to benefit the, you know, new genetics that was going on in domesticated animals. And so, when the Human Genome Project launched, that was a defining moment. For everybody. We realized that it was possible to sequence whole genomes. We never thought, when we started, that we'd ever be able to raise the money to sequence, say, a cow genome, or a pig genome, or a chicken genome. We were content with developing maps, because we knew that maps would be able to give us this marker-assisted breeding, these marker-assisted breeding strategies. But it quickly became apparent, as sequencing became more, you know, cost-efficient, that it would be possible at some point to do the sequencing of another animal genome. And the first one, after— first mammal, after the human, and the mouse, and the rat, was the cow. And we were right there at the right time with the right people thinking in the right way that the cow should be the next one.

Can you tell us how you began thinking of the feasibility of sequencing the cow genome?

Absolutely, because there are many of us that felt— of course, I was involved in the mapping side, but there were many of us that felt that sequencing would be inevitable once we knew that sequencing was possible, even though it cost $3 billion. We all began to say, "Well, what if we had the sequence?" You know, and it didn't take that much longer from 2003, the first, you know, draft of human genome published, until the time that we started. And George Weinstock and Richard Gibbs and the NIH got together, as well as the industry, and decided to put $50 million into sequencing the cattle genome. And, you know, that was only a period of a few years.

Can you tell us more about the process of sequencing the cow genome? Who were the key figures?

There were a few people that were very influential in getting the bovine genome sequencing initiative started. Jim Womack at Texas A&M University, who was the chair of the national, you know, sort of the cattle Genome Group within USDA. He was absolutely critical. There were— there was a group, Genome Canada, there was a group in Canada that was very important. In fact, the meeting to sequence, you know, one of the organizing meetings was held actually in Toronto. The USDA was very critical in this, the Agricultural Research Service and the group at Clay Center. Was also very important. Steve Kappes and the leaders at that time, Roger Garrett, Craig Caret, Rex Rhode, were really supporting this from the USDA side. And, you know, they went to NIH and said, you know, we can raise half the money to do this. And I think NIH saw that there was great interest, there was a community that could be involved in the annotation of the genome, and, you know, they were willing to go ahead and fund some of the substantial part of the sequencing to, of course, to have another mammal to compare at very high resolution, but also because of the impact that this could have on agriculture. It was a very important coalition that allowed the bovine genome sequencing to move forward. Those were pretty much the key individuals. I didn't play a tremendous role in that. My role was still, as a young— quite a young scientist, was keeping my nose to the grindstone and producing the map that eventually they used to hang the sequence on to really get chromosome-scale assemblies. And so, you know, I was involved in all of it from the beginning, but, you know, it was a very big move— political effort to get— to raise $50 million at that time. That was a lot of money. It cost $50 million to get that done. And, you know, what today we can do for maybe, you know, $20,000 was $50 million in those days. And that was a tough job, but we got it done. And the impact has been enormous.

You argued in a 2009 paper that good sequences require good maps. Can you elaborate?

You know, that paper, which we published in 2000, 29 is still reverberating 10 years later. And it was, in a large part, stimulated by the kinds of things they did with the 29, where, you know, mammals, where we had lots of sequences that were just reads that couldn't be really used unless you had a reference genome, a good genome to align them to. They were just a bunch of random reads. With no possibility to assemble them into chromosomes. And for me, for the work that I was doing, again, if you're going to study chromosome evolution, you need to have contiguous sequences in order to do that. But it turned out that the mapping, of course, was important, but it was the contiguity issues, because so many of the sequences, when we realized, and you went into the those genomes were missing parts, or they were assembled at those times incorrectly. You didn't have complete gene models if you wanted to do annotation. You didn't have— you couldn't determine whether there was a segmental duplication there. There were so many questions that came up because of the quality of the assemblies. They were good for SNPs, for single nucleotide polymorphisms, but not— and what you could do with SNPs, but not for much more than that. And so, you know, the maps were critical to us for studies of chromosome evolution, and to study how, you know, meiosis and mitosis works. And so, none of that was possible at that time. So, we made that the point around chromosome evolution, but it extended to many other issues that were related to contiguity and being absolutely sure gene— everything from gene regulation to epigenomics and such. And, you know, we're still having this debate. And the beauty is that the technology is finally catching up. Things like, you know, Hi-C and all of these scaffolding methods are now opening up— what we argued for is allowing the study of structural variation and the impact of structural variation on human genetic variation and any other species that you want to examine. You really need to have these highly contiguous assemblies so that you can find all the structural, you know, variants, the inversions, the translocations, the fusions and the fissions, and the small things, insertions and deletions and duplications. And they have been— we've known that they are important, but they've been very difficult to approach on a genome scale until now with these methods, these newer methods. And so, I think Everybody said, "Yeah, they're important, but they're beyond our reach at this time, so we're going to focus on these other things that we can answer with the more limited assemblies that we have." But people slowly began to gravitate toward, "Yep, not only is that important, but we think we can do it." And the Vertebrate Genomes Project, for example, is really leading the way on this now with the the kinds of assemblies that they're producing, and showing that it's possible to produce them at scale and at a reasonable cost. And that's opening up all these new areas for investigation that we knew were important back in 2009, but could only be approached 10 years later, now that we have these kinds of chromosome-scale assemblies.

What are some of the limits of the SNP studies?

It was a mixture of what was possible to do, and all the things that you could do with SNPs were really important to people, genetic variation. And that is a thing that, you know, that has led to the GWAS and the whole— the entire revolution in understanding population genetic histories and demographics and human migration patterns and human evolution and the co-evolution of domesticated animals with humans. These are things that were not possible without the SNPs. So they played an incredibly important role early on.

Do you think it's important to have a reference genome sequence for every species?

Yeah, I think the reference species— reference genome for at least for each— one representative of each taxonomic family will give us the backbone that we need to really understand the relationship between genotype and phenotype and how genomes evolve.

Your work on the cow genome led to an important discovery which greatly impacted agriculture. Tell us about it.

Right, so that, that's a very interesting story, and it was— it all happened quite by accident, just as many important discoveries do. We went back, I decided that we needed to resequence, do the first these were the first resequencing of the bovine genome, by picking two bulls, very important bulls. One of the bulls was born in 1963, and he had, you know, hundreds of thousands of offspring in dairy cattle. And his sons became the most important sons in terms of their ability to cause an increase in milk. Production in the female population of the cows. And so his lineage was absolutely critical to the development of the modern dairy cow. The name of that bull was Chief, Arland Farms Chief. And so we resequenced Chief because we wanted to look for signatures of selection over multiple generations since 1963, and see if we can identify those haplotypes, those parts of the region genome that were under selection. And it turned out that we sequenced Chief and his sons and a few others, resequenced them using 454 technology, which was the long-read technology of the day. And, you know, these were genomes that cost us about $150,000 each to do in those days. And, you know, we had a hugely successful project in doing this. We were even able to sort out haplotypes with what we did. And that project was done and published, that those results were done and published, and I got a call from USDA one day and said, "We have mapped a gene to cattle chromosome 5 that is clearly influencing fertility in dairy cattle, and the gene maps back to CHIEF." I had just moved to Davis, and Kurt Van Tassel called me with this problem. And he said, "Can you look on chromosome 5 to see if there's anything in this region of about 10 million— 10 million bases." There's a huge fishing expedition. "That could possibly be the cause of this, you know, loss of fertility." It was the gene with the biggest— it was the chromosomal region with the largest effect. That they found with, you know, millions of records in the National Dairy Cattle Breeding System. So I said, "Okay, sure." So I called the postdoc who had already moved on to Wisconsin, and I said, "Let's have a look at this region. Can you call up the region and let's use the, you know, very simple bioinformatics tools called ANOVAR at that time, and see if we can find something in this region. Let's first find— look at all the regions, the annotated genes, And so their comparative genomics comes in. Let's look at all the annotated genes, run this through, and see if we can find any mutations. And within 48 hours, it took less than 48 hours from that phone call till we found a stop codon in the 11th exon of the APOF1 gene. As I said, this is amazing. And I looked the gene up, I was immediately able to go to the the database and find that there was a parallel mutation in mice that caused this phenotype that was almost identical. It was basically embryonic lethal, influencing the entire development of the, you know, the notochord and the whole central nervous system in mice. It had beautiful pictures, and I said, you know, "This has to be it. We have a truncated— we clearly have a stop codon that was truncating this." this protein. And boy, this is the best candidate mutation, you know, that we've seen. And then, over the next, you know, few months, we were able to have genotyping on hundreds of thousands of individuals and all of the other types of confirmatory evidence, you know, no homozygotes at the population level, tracing it to another one of his sons who also you can map the loss of fertility to. And, you know, it turned out that this gene caused about 500,000 abortions, all worldwide. Because the spread, it turned out that 12.5% of all the genes in the current Holstein population could be traced to the single bull. So, eventually inbreeding will catch up with you, and inbreeding was starting to catch up with the industry at at that point. And, you know, these homozygotes, you never saw them, they were never born, but there was this clear decline in fertility along his line. So that led to a test. There was a commercial company, a company that picked up the mutation. We didn't patent it, we just gave it to the industry, said, "There's an opportunity to improve, you know, improve agriculture worldwide." And since then, they've been able to screen for the mutation, eliminate the mutation, and, you know, that mutation is no longer a problem for the industry, because fertility is, you know, probably the most important trait, because without having a calf, you don't have milk. So, it's a really important trait for animal agriculture.

Do you think genomics has led to better cattle breeding practices than the older traditional methods?

Well, I don't work in this field very much anymore. I've— my interests have completely shifted to genome evolution, but I do keep up with what's going on. And, you know, this is where genomics has had a huge impact, and it has happened faster and with greater— I would say greater overall impact than what has happened in the human genome with precision personalized medicine. The technologies were very quickly adopted into a process for genetic improvement, which we call genomic selection. And that would never have happened without the first having the sequence and then having all the SNPs identified and having very dense SNP maps, and then being able to relate quantitative traits to different parts of chromosomes. So this process of genomic selection is saving the industry, both the dairy— all the agriculturally important species of animals and plants now, millions of dollars, and accelerating genetic improvement at a rate, you know, very close to what would be predicted based on selection of individual, you know, genes affecting these traits over the entire genome. It's been transformative for the breeding industry because they can dramatically reduce their costs in generation. You can just make genetic improvement at a much faster rate than you could previously using the statistical animal breeding and plant breeding tools alone. So it was— they were the early adopters, and that industry has recognized recognized tremendous gains because of the genomic technologies.

How is genomics currently helping to address the variety of other pressing issues facing the agricultural industry?

Yeah, this is a really new and very cutting-edge area that you're suggesting. Of course, all of the selection has been on traits of economic, direct economic relevance, such as milk production or components of milk or or parts of the beef cattle industry that are important. But now people are starting to look at things like, you know, methane production, other areas of sustainability that are extremely, you know, important in the, in the new environment that we're in, in terms of how these animals and how the industries will adapt to changes that are being caused by, for example, by climate change. And so I think, you know, this is where the emphasis needs to be. There needs to be much more, a much greater understanding about how to use genetic variation as it relates to extreme environments, which we're seeing all over the planet. There are places where it's not just temperature, but it's the insects. It's everything that comes along with the changes in climate that affect the ability to conduct— or, you know, it affects the entire industry and in turn affects the food supply, the stability of the food supply, safety of the food supply through, you know, the pathogens that can enter. And so this is a very active area of research. It's going on in Australia, going on in the United States, and going on in other areas of the world. And it's very complicated. It's not as simple as direct selection, because there's been no selection on these traits, so it's not that easy to find the genes that you're actually looking for. So it's an area of active investigation and one that's really important to the future.

How have recent developments in sequencing technology aided your work? What's on the horizon?

You know, these long-read sequencing technologies and the cost of doing them has been an absolute game changer for the goal of producing chromosome-scale assemblies. It's not just the long reads, but it's also the scaffolding methods that have made a huge difference in the transition from the short reads to the long reads. And then the combination of these methodologies is really what's enabling us to get to almost telomere-to-telomere sequences. And so long read technology, I think, is the key. We know it's the key to contiguity. And it's, you know, if the costs— the costs are already at a point with the new technology, say, by PacBio, the $8 million chip that they have. You know, a human genome at 60x can now be sequenced. You know, for basically on 2 chips, that's $2,400. And so, that's a 3-gigabase genome. And as we're thinking about sequencing all eukaryotes still, the— that's a big genome, 3 gigabase. I mean, the average is probably about 2. And so, the cost of producing high-quality genomes across the entire you know, eukaryotic tree of life is now, you know, it's now really in our grasp. It's reasonable. I think we can, you know, we can do the kinds of assemblies that are even better than the, you know, 2003, the primary human genome assembly. We can do that now for, you know, a couple of thousand dollars for if we want to go with that quality. For each new species that we want to sequence. So having a reference quality genome is, as we've heard in the last two days of this meeting, is what every scientist really wants. But, you know, people just did not have a way. We just did not have a reasonable strategy to that, for doing that, until, you know, it just became cost-effective when we have these new longer read technologies.

How do we make sure the next generation of researchers are equipped to keep pushing genomics forward?

This is where, of course, training programs and these— looking at this from different perspectives, from NSF's perspective, from NIH's perspective. I mean, we need stronger training programs in technology. In population genomics and evolution and in bioinformatics. We barely have the tools to scale beyond where we are today. You know, going from a couple hundred reference quality genomes that we have now to thousands of them, to tens of thousands of them, will require expertise that we don't have. If we don't have the good fundamental training programs and be able to attract people into those programs, we're not going to have the tools. We'll have the genomes before we have the tools. You know, we'll have assembled genomes, but we won't have the tools to visualize them or interrogate them in a meaningful way. So the training programs are inadequate in my view. I think we need to broaden those. I think we need to think about interagency. Approaches to training graduate students and postdocs. And we need to be able to do it at scale and not in these little onesie-twosie kinds of things that some of the funding agencies have, but in larger programs where this— it's only possible to train, I think, the number of people that are going to be required if you do it— the only way to do it efficiently is to do it at scale. And, you know, we just don't have those programs right now.

Can you tell us about the Earth Biogenome Project? What gave you the idea for this ambitious project?

Well, it goes back to your earlier question. What motivated it was that I reached a dead end. I had no more genomes of chromosome scale that I could use to go deeper into the questions of chromosome evolution. The ancestral organization of the mammalian genome. There were just— there were 8 genomes, and then there were 10 genomes, and okay, you know, you could just go so far with 10 genomes. And so, that was— that's where I realized that unless there was something we were going to be able to do at scale, you know, we were not going to be able to understand things at the level that we were interested in exploring. And so, you know, on a flight back from the UK, I started to sort of pencil this out of what it would take to produce large numbers of reference genomes. All of this was inspired by the G10K, you know, the thing that Oliver Ryder and Steve O'Brien and David Haussler started. I was a founding member of that community. I was the only one advocating, and would Steve was the only one supporting me for chromosome-scale assemblies. Everybody wanted to do these draft assemblies, and I said, you know, we have to build it in. If you look at the founding G10K document, you'll see that we argued in there for chromosome-scale assemblies. And it was the first place that we did so. But they weren't being— even though it— that was 2009, that paper and that whole thing started about that time, but here was 4, 5, 6 years later, and we just weren't getting any traction. They really— it was still too expensive. We didn't have the technology. And then things started to change. PacBio came along. And so this was led to this little pencil calculations. You know, we had Illumina was there with their technology. The short-read technologies were there, the long-read technologies were coming along. So, you know, just penciling this out, I said, "Well, what would it take to do, you know, 1,000 genomes?" I looked at the number, I said, "This is, you know, if this is what it takes, why aren't we doing this?" And I started to go 1,000 genomes, 10,000 genomes, 100,000. Well, why not everything? And then, you know, I said that number came out to like 2.7 billion. This was 2000, about 2005. 2014, 2015, you know, just based on sequencing tech costs and technology at that time. I said, "This is crazy. This is doable, you know, this is actually possible to do." And so I started to talk to a few colleagues about it who were thinking about the same thing. Gene Robinson at the University of Illinois, and John Kress at the Smithsonian, and we decided to put a workshop together at the Smithsonian in 2015. 2015. And that's— Eric Green was there and Richard Gibbs and a number of others, representatives from all the agencies that are here today at this meeting. And they all said, you know, you've got to do a lot of work to figure out how you would do such a thing, but it's not crazy. And it really would be transformative if we could do it. And then we spent the next 2 years trying to figure it out and eventually came up with a a reasonable strategy that everyone believed, you know, could work. And we published it in 2018. Here we are another year and a half later, and it's turning into a global movement of very committed people who believe that this is the best thing we could do for biology and medicine and agriculture and environmental science at this point in history.

How do you see genomic sequencing affecting wildlife conservation.

Having the sequence is the first thing that everyone is going to want to do. We just heard from Beth Shapiro and her talk. I mean, you could sequence a single individual and you could look at runs of homozygosity, long stretches where you find no variation, and there's already been demonstrated a correlation between stretches of homozygosity in the genome and their risk for extinction. So I think as a first step, this is what— this is what will be done for every species, because if there's lots of existing genetic variation, you have some very reasonable tools and strategies already in hand for how you would manage that by, you know, moving breeding individuals and that sort of thing. But if you have a population that seems to be be a great risk because of its almost absence of variation. You know, you might— you either might say it's functionally extinct already, or you might have to take very different measures and— to try and conserve that species. And so there are decisions that will be made based on what we will find out. About genomes, you know, what variation still exists in individuals and at the population level. And those— that information will be absolutely critical to the decisions on which species you're going to try and conserve in their natural habitats, which you might move to confined, you know, breeding for perhaps reintroduction at a future date. And, you know, which ones you might move individuals from different isolated populations to increase variation. Those are the ways that the information will ultimately be used.

Can you give us some examples of how sequencing is already impacting conservation?

You know, again, Oliver Ryder's efforts with the northern white rhino, which is basically functionally extinct, and there's only a few individuals left in So radical decisions are being made about how to conserve that species. But there's everything in between. You know, we heard about Florida panther today, and we've seen the impacts of introduction of new genetics and new genetic variation into that population to stabilize it, and now that population is thriving. And there's a whole host of species. Species that are sort of at that point, and many of which we know about, and many of which we don't know about. And so that's what I think one of the beauties is of the Earth BioGenome Project is we're immediately going to be able to access that, even through the single individual. If we can get, you know, haplotype-phased assemblies, we're going to be able to see stretches of homozygosity, you know, and just understand, you know, how much variation actually exists, even if it's from a single individual, it should be a mirror into what is happening in the general population. And if we do what we propose in the Earth BioGenome Project, which is to get at the genetic variation by population sampling of at least 10 individuals, and with a good reference genome, we're really going to have a pretty good handle on where those species are with respect to their risk for extinction.

How have you built a research community around the Earth Biogenome Project?

The project in itself, the vision of the project itself, is highly motivating. And it's amazing that young people who write to me almost every day from around the world are totally captivated, you know, by the vision of sequencing everything. So they're not constrained, you know, they don't ask the same questions that you'll hear say, at this meeting. They just say, "How can I be involved? I think this is terrific. I want to be involved in this project. Who can I go study with? Do you have funds to do this kind of thing?" We have to engage scientists and young people and the public. This is a project that is not just about trained PhDs and people in science. It's about engaging the entire planet, really, in this initiative, because we won't be able to collect the samples, we won't be able to do many things unless the public actively engages in what we're trying to do. To answer your question directly, the project is building its community through— it's a network, global network of networks. So it's tying together all of these groups that are working on specific taxonomic groups of organisms, you know, whether it be insects or microbial eukaryotes or vertebrates, and putting everybody under a common framework. And so the community engagement comes in just the way the governance of the project is organized. We have representatives from all of those communities, and those communities are tied to everything from, you know, groups of people who are interested in specific taxonomic groups to individuals who were interested in a specific species. And so this thing just kind of spreads and grows organically in a way that just, you know, draws people in, because anytime you have an organism that has an interesting biology, today everybody wants to have its genome sequenced. And so this is It's not very difficult to build communities when you have that kind of pull across the entire, you know, entire living world and across biology where there are so many organisms being looked at. Like you said, I think before, you know, it's no longer is this a model, you know, or every organism in itself is a model in itself and for everything else because everything is connected. So this is a very— I think a very motivating vision for many people, and I think this is only going to grow as the project, you know, really gets off the ground.

What for you was the most surprising legacy of the Human Genome Project?

Well, the thing that's most surprised me is the speed in which the technology has enabled us to create resources and ask questions that we never thought we'd be able to ask. And if the end product is these high-quality genomes, reference-quality genomes, if we can produce them, they are— that is— it turns the question around. It's the infrastructure for the future of biology, because any questions can be answered by having reference quality genomes. You know, and if you do it and you include the Hi-C and you include the epigenomics components, epigenetic marks, and so on, you really have a resource that will last for a long time into the future. And I think that's the, you know, that's something that we all, I think, want to see happen. It's a great legacy because these species are disappearing at a rate in which we also would not have predicted even 5 years ago, the acceleration in the extinction rates. We're in the 6th mass extinction. We could lose 50% of the species on the planet by the end of the next century. You know, what the consequences are for us and the rest of life on Earth, nobody knows, but we're starting to you know, in these accelerated extinction patterns. And so I think, you know, really looking forward from there to answer your question, you know, most surprising thing is the speed we've been able to be able to do this. Now we need to figure out how to do it, and then we need to focus our efforts on what we can do with the information that once we get it. And this is where, you know, Beth's ideas and people— things that people don't really want to talk about because it's sort of at the fringe, like the extinction. People don't want to really talk about it, but it's, you know, if we have the code, it's almost Michael Crichton, you know, it's, it's almost Jurassic Park. If we, if we have the code, we just believe, and, you know, if we believe in our ability to change the future, we know If we have the code of everything, the true book of life, the digital sequence, the library of everything that exists on the planet, we don't know what the future holds, and we don't know what technologies are going to be there, but we know if we have that digital information, that's— we have a record of that species, and what we can do with that in the future, you know, could be very important. To the survival of our descendants. And so, you know, I urge people not to dismiss it as science fiction because what we're doing today would have been viewed as science fiction when the Human Genome Project started.