Ep 131: Secrets in the structure (with Scott Edwards)
How are new technologies allowing biologists to look beyond simple genetic variation to whole genome structure? What is a pangenome?
In this episode, we talk with Scott Edwards, the Alexander Agassiz Professor of Zoology and Curator of Ornithology at Harvard University. On this episode, we talk to Scott about the new ways we can describe and understand large chunks of DNA that till recently we have not been able to characterize directly. We also talk with Scott about why engaging in natural history is important in genomics, and we discuss new tools that have allowed him to study avian “genetic structural variants” and pangenomes. With these tools, Scott’s lab has made some exciting discoveries about how large genetic changes have helped one bird species, the house finch, cope with a major disease outbreak.
Cover art: Keating Shahmehri.
-
Marty Martin 0:06
Greetings, Cam. Question for you, do you think the size of the bookstore influences the information you glean from the books you buy and read from them?
Cameron Ghalambor 0:13
Huh? That's a weird question. All right, let me think about that for a second. My first thought is no, because a book is a book, no matter where you buy or read it, same author, same cover, same content. But I know you well enough to realize this must be some kind of trick question or an attempt at wit.
Marty Martin 0:34
Oh how poorly you think of me assuming this is not some serious question with a straightforward answer? Though you're right. It's an attempt at wit and probably a bad one, but still, give it a go. Do you think the same info out of a book housed in a crappy bookstore versus a beautiful Barnes & Noble gets you the same information?
Cameron Ghalambor 0:51
Ah, okay, so that's a clue. You said "crappy bookstore", which automatically makes me think back to the old used bookstore I used to go to when I was an undergraduate in Los Angeles. It was a disorganized mess with classic literature mixed with sci-fi and everything else you have to like, sift through the chaos. And if you were lucky, you could find some real gems.
Marty Martin 1:15
Exactly! And so sure, once you find the book, all is okay, but the accessibility of that info could have a strong impact on your ability or willingness to read it.
Cameron Ghalambor 1:25
Okay, sure, but where are you going with this?
Marty Martin 1:28
Okay, let's say the bookstore is well organized, but it's just set up in such a way that some books are a bit hard to get.
Cameron Ghalambor 1:34
Example?
Marty Martin 1:35
Let's imagine that there are books on your favorite topics. You know, one floor is full of books on your precious LA Dodgers. Other floors are full of cookbooks and yet, other floors are full of those trashy romance novels you love.
Cameron Ghalambor 1:49
Wait, what romance novels? What are you talking about? Are you confusing me with Art?
Marty Martin 1:52
Ah, but here's the catch, the bookstore is five stories tall and it has no elevator. So how often are you gonna walk up five floors to get that copy of the Shohei Ohtani biography you've been so desperate to read when there's so much trashy, tasty fiction to read right there on the bottom floor waiting for you.
Cameron Ghalambor 2:12
Ah, okay, well, that would have been an easier question to answer if your example wasn't so crazy and ridiculous.
Marty Martin 2:21
Right, not that often are you going to bother to walk up all those stairs? At least less so than going for the easy access to the juicy romance novels. Who's gonna walk up so many stories when your other favorites are right there?
Cameron Ghalambor 2:32
I hope you're going somewhere with this story. By the way, have you seen my cookbook collection?
Marty Martin 2:34
Of course, our guest today, Scott Edwards, Professor of evolutionary biology and curator of Ornithology at Harvard University, has been studying information in genomes for years now. More specifically, Scott is becoming interested in how the structure of the genome, besides the variants we can easily describe, evolve and operate in a variety of organisms. In other words, in the context of my bookstore analogy, he's been trying to develop better ways to find certain books,
Cameron Ghalambor 2:35
But especially in birds.
Marty Martin 2:37
Exactly, especially in birds. Most work in genomics until recently, has been focused on small sequence variants like the Single Nucleotide Polymorphism or SNP, because these variants are much more accurately described by the technologies we now have.
Cameron Ghalambor 3:15
Even though we've known for decades that genomes have very complex structures-
Marty Martin 3:20
-they're all wrapped around histone proteins and coiled up into big groups of material we call chromosomes.
Cameron Ghalambor 3:25
And until very recently, we haven't been able to study structural variation very well.
Marty Martin 3:30
For instance, we've long known that in the evolutionary past of many lineages, big chunks of chromosomes somehow become inverted, literally flipped upside down. But we discovered many of these only because their impacts were so obvious.
Cameron Ghalambor 3:44
One of my favorite examples of this type of chromosomal inversion is from white throated sparrows. Males and females in this species either have a bright white crown or a dull tan crown, depending on whether they have or lack an inversion. We found the inversion because this species literally had two discrete color morphs and lots of correlated traits, suggesting we might look for groups of genes that were not being broken apart by recombination.
Marty Martin 4:11
Right and Scott's interest is to use new technologies to reliably sequence long stretches of DNA to find more of what he and others call structural variants, like copy number variants, deletions and insertion.
Cameron Ghalambor 4:24
Until just the last few years, we simply weren't able to accurately sequence longer DNA segments than just a few 100 base pairs. Instead, we had to chop up the genome with a bunch of enzymes, sequence those small chunks, then align them using computers and fancy math and check whether these simple variants led to any interesting insights. The good thing, this approach was effective. The bad thing, there's a lot more to genomes than simple variants.
Marty Martin 4:51
True, the typical vertebrate genome is billions of base pairs long without very large sequences, we can't see potential structure like inversions and other big changes in the genome. And to go back to my crappy bookstore analogy-
Cameron Ghalambor 5:06
Yes, it is pretty crappy.
Marty Martin 5:08
Haha. You know, that's not what I meant. Anyway, to go back to the bookstore analogy, now that we can sequence very long expanses of DNA, we can much more easily get to that top floor to find all sorts of new structures in genomes that we might want to know about.
Cameron Ghalambor 5:22
Ah, so kind of like installing an elevator so that I can easily get to the top floor to check out those Dodger books?
Marty Martin 5:28
Exactly, and please, no one hold Big Biology accountable for Cam's bad taste in baseball.
Cameron Ghalambor 5:33
Very funny. Let's see if you're still poking fun at my team when your precious Atlanta Braves are struggling again to even make the playoffs in October.
Marty Martin 5:41
Fair point. So time to let me let the baseball stuff go and talk science.
Cameron Ghalambor 5:45
But before we get to our chat with Scott, a quick reminder that we're now a subscription based podcast, so only the first 30 minutes of this show will be free.
Marty Martin 5:54
But if you'd like to hear the rest of the conversation, go to our substack page, www.bigbiology.substack.com and become a member.
Cameron Ghalambor 6:02
For just the cost of a cup of coffee a month, you can access the full show, as well as extras available only to members, such as Marty and my debrief after our conversations with guests, access to discussion boards to talk about the show with other listeners and a variety of other freebies.
Marty Martin 6:20
We've had to resort to a subscription model for most shows because of production costs that mostly entail funds for producers and interns. They're just too high to sustain through grants and listener donations.
Cameron Ghalambor 6:29
However, we never want your lack of funds to prevent you from listening. If you just can't afford a subscription, write to us and we'll give you one for free.
Marty Martin 6:38
But if you have the means, please consider supporting us, as we can then give access to students and other listeners that need help to be able to access all of the show.
Cameron Ghalambor 6:46
And just one more point before we bring in Scott, we want to emphasize what we're trying to do with Big Biology. We're trying to share with an international audience some of the most exciting and cutting edge research that's being done.
Marty Martin 6:59
Yes, scientific journals and other venues do this, but they're often beholden to publishers and other commercial interests, and we're not. And as we're producing this episode in early 2025 and are located in the US producing science podcasts these days have to become more difficult in the near future,
Cameron Ghalambor 7:17
We're not going to say a lot more about the future of federal funding for Science in the US, but we want these episodes to keep coming, and we need your help as subscribers to do this.
Marty Martin 7:35
We want to continue to bring you the ideas of the world's Big Biology thinkers without any obligations to businesses or other groups that might have agendas. We want to keep Big Biology about the science driven by curiosity, but we need your help.
Cameron Ghalambor 7:48
Your subscription can help make that possible. So please subscribe. Go to www.bigbiology.substack.com and become a subscriber.
Cameron Ghalambor 7:57
I'm Marty Martin.
Cameron Ghalambor 7:59
And I'm Cameron Ghalambor
Marty Martin 8:01
And this is Big Biology.
Cameron Ghalambor 8:16
Scott Edwards, thank you so much for joining us today on Big Biology.
Scott Edwards 8:20
Thank you.
Cameron Ghalambor 8:21
So we're really looking forward to talking to you today about your research. And kind of want to start off by saying, I recall the very first time I read one of your papers. I was just starting graduate school, and in our lab group, we read this paper that you and Shahid Naeem wrote in AmNat. This is like, maybe, like, 30 years ago, and it was about the importance of phylogeny, and the sort of phylogenetic component of cooperative breeding in birds. And at the time, you know, there was the Joe Felsenstein's phylogeny, you know, independent contrast, and a lot of studies looking at phylogeny. But I really remember and sort of appreciated that paper, because it it really made it very clear how important, not only phylogeny was when sort of comparing species, but also incorporating kind of the ecology and the natural history, which was often missing, I think, in a lot of those kinds of studies. And so, you know, that was back in the days where I guess maybe people were still doing AFLPs, or maybe microsatellites were starting to kind of-
Scott Edwards 9:34
I think that was before microsats
Cameron Ghalambor 9:39
But, you know, now we're doing long read sequencing, and, you know, I guess I was just kind of curious, like at that time till, till now, the present, like, Could you, could you have imagined sort of this arc in terms of, like, both your research and also in terms. Of, like, the boom in genomics, like, Was that something you could anticipate at the time when you were thinking about these ideas? Or was it like science fiction?
Scott Edwards 10:13
Yeah, those great questions. I mean, I'm glad you, you, you remember it and like that paper, because, yeah, it's widely unread, as I would imagine. But that was a fun paper, and I'm still in touch with Shahid. You should she'd interview him for your show, if you haven't already, yeah. And that was kind of inspired by some of my graduate work. I did my dissertation on sort of phylogeography of a group of Australian babblers, which are in the genus Pomatastomas. And you know, as you probably remember, the sort of heyday of cooperative breeding in the mid 80s, early 90s, there were a lot of different ideas about why birds were social, like that had helpers at the nest. And what I noticed about this babbler genus was just that they were all five of the species were cooperative to varying degrees, and yet lived in drastically different habitats. I mean, you have the chestnut crowned babbler in the deserts, semi-deserts of Australia. You have gray crowned babblers in the eastern woodlands, and then you have the New Guinea babbler in rainforest. And, you know, it was just unclear to me how ecology, in this case, was sort of the sole driver of cooperative breeding anyway. So yeah, we put this paper together. It wasn't really using very cutting edge techniques. Shahid was really helpful, just in terms of ideas and doing some simulations. And yeah, it was a fun paper.
Scott Edwards 11:43
If you want the real backstory, I'm happy to share it with you, because it was actually one of the more traumatic experiences in terms of publishing that I had as a grad student. But yeah, and you know, of course, many of the several the reviews of the paper like, oh yeah. Well, actually, I had said that already, you know, like, well, like Jerome Brown, who did a lot of really interesting work on cooperative breeding in New World jays, you know. And honestly, he did have phylogenetic ideas in some of his papers. No question. So in many ways, we were kind of just collating information together and synthesizing it to make a strong statement, and hopefully it just put the idea out there. I don't, I think today, folks wouldn't be so surprised by the conclusion. And I think now, you know, it'd be nice to there have been a couple papers that have sort of integrated phylogeny and environmental stuff, you know, very, very elegantly. And yeah, as you said about genomics, Cam, I mean, yeah, no concept these days that we'd be sequencing whole genomes. It's really remarkable the whole genomics revolution. And yeah, one of the really fun things about comparative biology today is just having these really well resolved phylogenies, but being able to map traits and features, behavioral, phenotypic traits, on these trees and integrate the two. That's really exciting.
Marty Martin 13:14
Yeah. Well, what you know the other dimension to continue to gush about the work that you've done, Scott, I mean, it's not just the sophistication of the genomics these days, but it's also, there's always a depth to the natural history. So you've always stayed true to the Berger roots. And, you know, we share an interest in sort of comparative immunology, and that's the kind of thing that's just so massively missing from these big areas of science. So, I mean what has it been about your history of being a natural historian? I mean, why have you kept it there? This is something we've talked to Harry Greene about a little while ago. I mean, he's made a giant point about how important this is. Is this something that you try to get your graduate students to pursue? And I mean, clearly it motivates you. But what's your what's your current thinking of natural history of genomics?
Scott Edwards 14:01
Yeah, no, that's a great question. And of course, you know, I think the challenge these days is to stay true to yourself, as you said, Marty. I mean, the genomics is so flashy, and honestly, the bioinformatics is so seductive and in some ways, so satisfying. I mean, you get like, a result on your desktop, just like in minutes. And honestly, I mean, I've seen sort of what I would call well-rounded biologists sort of go down the tunnel of bioinformatics and never look back. And so yeah, and for me, the questions really have been about the natural history. And yes, I definitely encourage my students to have that be their center as well. I mean, for any prospective students out there, you know, I'm not really interested. If you know how to sequence a genome, I want to know. I want to know, if you're just in the into the being of the field, are interested in diversity of birds, you don't have to know birds like, you know, in your backyard, but having this biodiversity be your center of gravity is really important because, yeah, I think that's, that gives meaning to all the genomic patterns we might find or and really is the inspiration for all the questions.
Cameron Ghalambor 15:16
Yeah, cool. Well, let's talk a little bit more about genomics, and specifically comparative genomics, and eventually make our way to talking about pangenomes, which is a new word that came has come up, I think, for some people. And I, well, I just want to say that, you know, like going back and thinking in the old days, like I remember when I when I first saw my first karyotype, and I thought how satisfying that was to just be able to see a picture of, like, you know, the genome, and especially bird karyotypes with their little micro chromosomes and everything. And we've come a long way, obviously, since, since then, but when we compare genomes across different groups and especially when we when we look at bird genomes, I think one of the things that I've been really kind of surprised by or interested in is how conserved the order of genes are on the chromosome. So this is what we call synteny and I guess one thing I haven't been able to kind of fully wrap my head around is whether this type of synteny like, what the ecological and evolutionary consequences of it are, and whether birds are are unique, like when we compare like, how different are birds in this regard, compared to, say, mammals or lizards or other groups?
Scott Edwards 16:50
Right, that's a great question. I mean, some of the hardest aspects of comparative genomics is kind of figuring out what, if anything, the synteny and the sort of higher order structure of a genome has, what sort of significance it has for ecology. You know, we're very good at looking at patterns of selection on individual genes, but the ways in which genes are ordered on a chromosome and sort of the higher organizational properties I think we're still a long way. I mean, either there isn't much consequence to the order of genes around the genome, or we just haven't figured out how to look at it. I think the number of examples, I mean, we have some great examples in birds, for example, of inversions, how inversions can flip around and suddenly capture a whole group of genes that might influence a phenotype, like in the example from ruffs, or there's some nice work done in redpolls and things like that. And that, I think is those are the best examples we have of how the order of genes on a chromosome can make a difference.
Scott Edwards 18:05
But there are a lot of other, you know, subtle things, like, you say, like, you know, gene copy numbers, and, you know, Marty, you had mentioned the MHC, and we find interesting differences between passerines and non-passerines, or even between different species. The significance of those kinds of changes are just, they're not obvious yet. Some of them, you know, probably correlate with increases in gene expression that help facilitate a certain trait. But yeah it's still, I would say, very much a work in progress. And you know, with these long read genomes that we're finally able to really fairly accurately capture the full structure of bird chromosomes. You know, we're really in a position to map things out, lay them out, see actually what differences in synteny we are seeing, and then try to figure out what they mean. My guess, honestly, is that birds probably do have less dynamic genomes than, say, mammals or reptiles, but they probably have more dynamic genomes than we've been led to believe by short read data. I think if we look closer, we'll see lots of evidence for gene duplications and micro shuffling events that we just haven't been able to see with the short read data.
Marty Martin 19:19
Right. Ooh, that's a big, big claim. That's an exciting one to dig into. So do you, I mean, what's really exciting about these long reads is that you can sort of describe synteny at the whole chromosome level. But then the next question becomes, and the technical challenge is, what are the other architectural pieces of the genome that we would want to describe. You use the word birds have probably more dynamic genomes, but I think you meant like in the evolutionary sense,
Scott Edwards 19:48
Yeah.
Marty Martin 19:48
But then there's the physiological sense, and all of the other things that we're just starting to measure, we've learned with the ENCODE project, and, you know, all of these other efforts that all of this stuff that was supposed to be. Junk is actually really important.
Scott Edwards 20:02
Yeah, right, that's at least what the you'd have. The human geneticist would have us believe that.
Marty Martin 20:09
Well, yeah, yeah, yeah. But so, so how are you thinking about that, if you had to put money into one of these bins about what part of this architecture, if you know, you're going to fund the group that's going to go develop the technology to describe this thing next. Where are you putting your money?
Scott Edwards 20:25
Wow. I mean, I probably, probably want people to look a little bit more at patterns of open chromatin around the genome, and that means, you know, which parts of the genome are kind of unwound at any given time during development or in different environments and sort of active for gene expression, versus being very closed and basically inert. So I'd want people to map those out at different time points during development, different phases of the life cycle, and then see how those regions interact with other regions around the genome. And so we know now that you know, the genome has this 3D structure, enhancers match up with their genes and help drive gene expression, and they can often be mega bases away, so sort of looping around. And so that's where I put my money. I think that's where some interesting higher order patterns might emerge. What genes are being regulated by a given piece of the genome, and how is the spatial organization of genes kind of important? And maybe birds as a group have a different architecture than mammals. Maybe their loops are smaller. I mean, we know their genomes are smaller, which means they have a lot less non-coding DNA than mammals and reptiles, but we probably don't know that much yet about how they're packing in more function per base pair than a mammal is, because surely, even though the bird genomes are smaller, they're still they can probably do just about the same range of things as a mammal genome can.
Marty Martin 21:58
What do we know about transposons in bird genomes, comparatively speaking, do we know much?
Scott Edwards 22:03
I mean, they're, yeah, they're, they're a lot less prevalent than in mammals, but they are gonna, the numbers will increase. I mean, you know, if you just look at genome sizes on NCBI, even for the same species, and you just, you just look, here's the genome that was sequenced in, you know, 2005 and here it is in 2015 and here it is in 2025 same species, like hmm that genome has been growing, what's going on? And it's because we're attacking it with better tools and so I think the repeats will increase, although they'll still be a little less muted than mammals. Honestly, I think the another real black box are the non-avian reptiles and lizards and turtles, and they're just gonna have a huge amount of diversity that I don't think has really been tapped yet, and that could give us a lot of context.
Cameron Ghalambor 22:57
So so the you know, when I think about the open versus closed regions of the genome and what can be read and not and also in the context of transposable elements, I start thinking more along the lines of epigenetics and kind of the dynamic, dynamism at that sort of short timescale. But are you thinking about it more over a kind of a, just like a longer time period, like where these are sort of set early on and then don't really change or..?
Scott Edwards 23:34
No, no, I think I'm actually talking about epigenetics as well. I mean, I would consider chromatin profiles, you know, as measured by like a taqseek, to be a type, a type of epigenetics. But, yeah, you've also got the methylation and the histone modifications at time points. And so I think here's a great dissertation project for someone is just take a common species like black cap chickadees or goldfinches, and just sample them across the four seasons and just describe the patterns of gene expression, open chromatin in the different seasons. Because what's remarkable is that, you know this one genome is actually a different genome in every cell of body, and you know it's blinking on and off in terms of activity in different phases of development, different environments. And I just think we get a lot of insight into things like, you know, plumage change, molting behaviors, reproduction, if we could look at the genetic underpinnings of these as measured by epigenetics and chromatin, things like that.
Marty Martin 24:49
Well, this maybe is a good time to bring in this word, pangenome . Do you want to tell us what this is? I know that Cam looked for a definition on Wikipedia, and he was unsatisfied. And I think I'm just as confused. It's the entire set of genes from all strains within a clade. So that's a very inclusive definition. What would you what would you explain it to be?
Scott Edwards 25:12
Yeah, sounds like a microbial person wrote that. Yeah, no, pangenome, quite, quite the buzzword these days. And I should just go back a bit and just explain how I got really excited about these things. I was, this was, again, a consequence of the pandemic. And I was teaching my molecular ecology class, I think, in the fall of 2021, and I, you know, I it was online. And, you know, as any professor would, I'm looking for guest lectures, to like take a bit of the load off me. And you know, I realized that this very talented scientist named Heng Li, who you might remember, developed this technique called PSMC with Richard Durbin to look at population history using single genomes. I knew his name, but I hadn't really sunk into me that he was actually working at the Dana Farber Cancer Center, right in Boston, Massachusetts, and so, so I emailed him and said, "Hey, would you like to give a guest lecture?" And I sort of said, "Hey, maybe you want to talk about PSMC?" And he emailed back, said, I'd be happy to give a guest lecture, but I don't want to talk about psmc. It's 10 years old, and I've moved on. I want to talk about pangenomes. I want to talk about the Human Pangenome Project." And I'm like, okay, whatever, whatever. Anyway, he gave a guest lecture, and I was just, I was just riveted.
Scott Edwards 26:35
And so and so a pangenome is, is basically, I should say that this term has been sort of unabashedly stolen from the microbiologists who coined it. I think in around 2015 they coined this term, pangenome in microbial science. Yeah, it's microbes are mostly genes. It's refers to entire collection of genes, say in a population or species, across all the strain sort of summed up over all the strains. And of course, some strains will have some genes that are not in other strains. And so it's meant to be the sort of collection when with eukaryotes like birds, I mean we think of, we tend to think of a pangenome yeah, in part, defined by the genes, say, in a whole species. Although honestly, we're really not trained to think about different individuals of a species having different gene complements. I mean, it's, you know, most you carry on. People say, "Okay, you have a species has sort of a constant number of genes." But what we sort of broadened it to include, are all of the kinds of variation that could change the length or the order or the structure of a genome. And so what we call structural variation, and that includes things like insertions and deletions and inversions. And so the pangenome is basically, it's, it two things that are important. One, it's, it's a survey of all that variation that we can't access with short read data, so all the structural variation, in addition to the SNP variation, the single nucleotide variation. But more importantly, perhaps it's also this idea that in traditional genomics, we're usually designating an individual to be a "reference," to somehow represent the species, and then we'll map other the sequence reads of other individuals to that reference. A pangenome rejects that approach and instead says, Okay, we're not going to designate anyone as a reference, because after all, that reference probably doesn't have some DNA that other individuals in the population have. So instead, let's just go reference free. Let's just map everything to everything else, and we'll have, sort of, instead of a one way mapping, we'll have an all, all direction mapping. And then we'll be able to capture all the variation in every individual. We won't have to throw anything away. And that, I think, is the real promise of a pangenome approach, and yeah. And I think it's a really exciting new way to look at genetic variation. Not only can you look at the single nucleotide changes, but all these structural variants and whether or not they have adaptive significance and stuff like that.
Cameron Ghalambor 29:14
Yeah. So let's talk a little bit more about these, like structural variants and like the different sort of flavors that they can come in too. So when I think of a structural variant, I guess the the two things that I associate most strongly with that would be like a chromosomal inversion, you know, where the DNA has, like, kind of made a little loop on itself, or copy number variants for a given gene. But structural variation captures a lot more than just just those, although maybe those are the most common ones that are studied. I don't know.
Scott Edwards 29:51
Yeah, I would say yeah, inversions are definitely a big, a major kind of structural variation, you know, and honestly, we're still. At the very beginning of figuring out, well, how many inversions does a typical species have? The estimates are all over the map. I mean, even in humans, I think the current estimate is roughly around 800 or so.
Cameron Ghalambor 30:13
Wow
Scott Edwards 30:14
yeah, but there are other in some rodents, the claim has been made that there are several thousand within a given species. Of course, many of these are fairly small. Often a structural variant is defined as being involving greater than 50 base pairs, sort of an arbitrary cut off. So inversions of just a few base pairs would not be counted there. But surprisingly, inversions are not the most common kind of structural variant. Really. I think the most common ones will be simple insertions or deletions
Cameron Ghalambor 30:44
InDels?
Scott Edwards 30:45
InDels, that's right, and we are again, a lot will depend on how people are counting them. You know, some of our work now is with you, Cam, it's very exciting. We're basically trying to see how different methods result in different counts of these things, so at some point, the community has to come to some general agreement about standards. I think the standards will include, for example, at minimum, long read assemblies of every individual in the study, and then certain bioinformatic pipelines we're using one called the Pangenome Graph Builder, or pggb, developed by some very talented bioinformaticians, Erik Garrison and Andrea Guarracino. And you know, it's one of these all versus all unbiased mappings of to create what we call a pan genome graph. And that graph is sort of your bedrock tool for finding structural variants, finding inversions and deletions and insertions. I think in general, deletions and insertions will be the most common. Inversions will be, I would say substantially less common, but doesn't mean less impactful for fitness and stuff.
Marty Martin 31:59
So Scott, I mean, convince me I'm think I'm wrong about this, but we've always been interested in these kinds of possible variants, right? We just didn't, until recently, have the ability to find them. So I think you mentioned long read sequencing without getting into super gnarly details, can you tell us, like, why can we know about this variation now?
Scott Edwards 32:20
Yeah, well, there's been some really transformative technologies. And, you know, I full disclosure, I'm not paid by any sequencing company to promote their, but I will say, you know, for example, Pacific Biosciences HiFi long read sequencing, I think is a really great platform because it yields very long sequences on the order of 15 or 20 kilobases, thousands of base pairs. And, you know, they're very accurate. So getting this combination of high accuracy and long length is just a very winning combination. And that's only been with us for, you know, large scale, maybe only three or four years. And so that it's that sort of long read sequencing that facilitates very high quality assemblies of genomes, and which really kind of allows us to get this kind of landscape view as well as a very focused view individual at nucleotides, and so yeah, and I, having worked with PacBio data, I can just say it's just an absolute pleasure compared to the sort of the sort of honey beehive of chaos that is short read sequencing. There's so many short reads. There's so bazillions of them, they're, they're very short, like 150 base pairs. And it's just, it's just a lot more to work with. And so it is, I would say, a game changer. And I think the costs are coming down to make it really accessible to everyone.
Cameron Ghalambor 33:52
So Scott, like so I was talking to my wife about in anticipation of our conversation, and I was trying to explain to her what pan genomes were and so maybe this is helpful for some of our listeners who maybe don't think about this as much. But I used this analogy, and correct me where I'm really off with this. But if you imagine like the the genome as like a painting and as a let's say everybody in a population of a given species, the painting looks like the Mona Lisa, but, you know, you can't actually see it in its full form. So in the past, we had to, like, tear to pieces with into these little short reads and kind of stitch them together against some reference genome. And if you blurred your eyes, maybe you could kind of see that there was, like a picture of the Mona Lisa there. But what the pangenome does is, first of all, it with these longer reads, you see more of the picture intact, but also what we're like because. We're not like mapping to a single reference, and we're comparing to each other. What you see is that, like in this individual, the Mona Lisa's eye is like a different color, and like in this painting, this other individual, she's got a little mustache, and that's the level of variation that we just didn't know existed until relatively recently. Is that a is that an okay analogy?
Scott Edwards 35:28
I think that works very well. And to extend the analogy, with short read sequencing, basically we could only imagine seeing changes of individual pixels if we think of a digital Mona Lisa, whereas with structural variant, we can actually see that these larger features, like a mustache, or
Marty Martin 35:45
Of all the things you could have said, Cam, that's the trait you picked.
Cameron Ghalambor 35:53
It's just the image I had in my head,
Marty Martin 35:55
and everybody will have that image forever now. Good job.
Scott Edwards 35:59
I think the Mona Lisa actually does have a mustache. I mean,
Cameron Ghalambor 36:03
I was thinking, I was actually imagining like a Salvador Dali
Scott Edwards 36:07
Oh okay
Marty Martin 36:08
Yeah perfect
Cameron Ghalambor 36:10
Only on one side, asymmetrical
Scott Edwards 36:11
That's right. But, yeah, it's, I think it's a very good analogy. It works. Yeah, it's super, super exciting.
Cameron Ghalambor 36:19
Okay but then, but then, here's my follow up, and which is that- okay, so I knew about, you know, so again, like, the importance of, say, like an inversion, is that if you, if you have a whole bunch of genes that are inside of this inversion, it there's reduced recombination, and it keeps that cluster of genes together, and if those genes contribute to some adaptive function, you know, that's a good thing to kind of keep them together, but then I seem to also remember learning at some point that, you know, inversions were also barriers for reproduction and could cause reproductive incompatibilities. And so I was really surprised to find out that, like, there could be standing variation in a population for these kinds of structural variants. So does that mean that, like, some individuals within a population might be reproductively incompatible because their structural variation is such that it makes it harder for the gametes to form during reproduction?
Scott Edwards 37:29
Well, I you know, I'm pretty sure that, like in that ruff example, the ruff being the shore bird in Europe, there are some incompatible matings, right? I thought there was some combination that was sterile essentially resulted in no offspring. I'd have to check but, but you're absolutely right. It is a bit of a enigma. We do seem to see this. I mean, the whole definition of a structural variant means it's segregating. It's polymorphic in a population, and with inversions in particular, you're absolutely right. We would expect some sort we could expect some sort of incompatibility. And I guess the question is, there probably are ways during meiosis where chromosomes can resolve these inversions, but sometimes they do and sometimes they don't. And, of course, in some studies, like we know from studies in fruit flies that even within a normal, outbred population, there's actually lots and lots of sort of micro- incompatibilities segregating in the genome. And that's was just this really interesting result that Andy Clark up at Cornell and others reported a few years ago and so but you know, measuring that and detecting it takes really big sample sizes, and I'm not sure we're quite there yet in birds, but there might be some interesting ways to infer that from the allele frequency spectra, or something like that. I mean, I think your intuition is absolutely right, and it's just, we just have to look a little closer to see if we can actually detect it.
Marty Martin 39:00
My gosh, there's so many cool things to hit here, we haven't even jumped into one of the things we wanted to spend a lot of time on, the recent house finch pangenome effort that you made the paper in PNAS last year. You're finding sort of relationships between the number of these structural variants and how this species has dealt with a new pathogen. So maybe just sort of set the table about why you decided to work in that species, and the big picture about the utility of this pan genome approach to seeing how genetic variation has sort of shaped up across the landscape. This is a very broadly distributed passerine.
Scott Edwards 39:43
Yeah, thanks. That's a great question. I mean, so my entree into house finches is entirely due to Professor Jeff Hill at Auburn University you probably both know him.
Marty Martin 39:53
He dragged many people into the house world, for sure.
Scott Edwards 39:56
Yeah exactly
Cameron Ghalambor 39:57
Red bird in a brown bag.
Marty Martin 39:59
Yep, yep.
Scott Edwards 40:02
Well, I assume you, if you haven't had Jeff on your show recently,
Marty Martin 40:05
We haven't yet,
Scott Edwards 40:06
Yep, okay, because he's an obvious he's so prolific and so imaginative. Yeah, you know, he's one of these scientists where you can get from a flat out rejection from a journal, and Jeff will be like, "This is awesome. We're almost this is gonna get published." I'm like: "Jeff. It says reject." Somehow he doesn't, he doesn't see that like all, all cylinders ahead.
Scott Edwards 40:32
Anyway, Jeff introduced me. We actually got a grant way back when to one of these Integrative Biology grants at NSF to look at the house finch mycoplasma interaction, which is this cool pathogen that jumped from poultry into house finches in the mid 1990s in the Eastern US, and sort of spread rapidly across the range of the house finches. As you probably know, the Western house finches are sort of, that's their native range. And they were introduced to the east about in the mid 40s, anyway. And so Jeff started us on this trajectory. And I've just been lucky to be able to continue it. And, you know, with, typically with grad students that come into the lab, or do post docs, I usually, you know, float "Hey, well, maybe you should work on house finches." And there usually is, like, a flat, "Sorry, no, you know they're too. They're too run of the mill." Actually, I think, I think house finches are great birds. They're just from many perspectives this logistically, they're small. You can get big sample sizes. They're easy to prepare in a for a museum. Specify, I know that's kind of gruesome, but
Marty Martin 41:42
Well, and I was gonna say they are almost as cool as house sparrows. I support you. They're almost as cool as house sparrow.
Scott Edwards 41:48
Ok. Important detail. And they are, they are different.
Marty Martin 41:54
Oh yes.
Scott Edwards 41:54
And so, yeah, they're, I love house finches, and they're just so variable, yeah, and so Bohao Fang, a very talented postdoc who actually got his PhD in Helsinki, with Juha Merilä, who, as you know, is a great evolutionary biologist. He sort of took to house finches, and just did this incredible job. We only looked at 16 birds, so it was kind of a small study, but, you know, we wanted to get the big picture demographics. You know, one of the principles I often tell my grad students, if they want to look at population variation or phylogeography, I sort of caution them like, "Okay, you might not have to look at hundreds of individuals depending on the study you're doing, because if for estimating demographic parameters, it's really more the number of loci, not so much the number of individuals that will influence things, but anyway. And so that's kind of the philosophy we took in the study. And yeah, it's a really exciting, I think it's a really exciting piece of work.
Scott Edwards 42:53
It's the first pangenome of a natural population of birds, and we saw the Bohao found this really interesting, eleven mega base inversion on chromosome one, and it turns out that it's, you know, the links with the pathogen, and sort of it's changed through time, I think, are compelling. It'd be nice to solidify them with even more evidence. One of the inversions, well, actually it's the heterozygous genotype of the inversion, which seems to have increased over time, and especially is high in populations that have been exposed to the pathogen for a long time, but at the same time, the same inversion seems to be found in the species that we use for an out group. This is this rose finch from Asia, which could suggest that this inversion has been around for a long time, and so, yeah, it's there's, as with any study, there's probably more questions than answers that are generated by it. But I think Bohao just did a fabulous job just laying out the broad picture of structural variation in that species, and also leveraging, actually, some short read data that was sort of around, but, and we were able to leverage to really increase our sample size to look at that particular version, which was really exciting, but
Cameron Ghalambor 44:09
So Scott, you know, for for our listeners who you know may not be completely familiar with the with the pangenome sort of pipeline, let's Say you were interested in the same question of like, okay, you've had this population that's been exposed to a pathogen, and you're interested in whether there's been selection on maybe, like, some MHC genes or other sort of immune genes. And you were to use kind of a, maybe the kind of more traditional population genomic approach, where you use short read sequencing. You've you have good coverage over the whole genome. You find lots of snips. Would you have missed the inversion if you had used the approach?
Scott Edwards 45:00
I think we would have because, in fact, a former grad student, Allison Schultz, also did a really nice, still unpublished, but looking at sort of SNP variation through time using mostly short read data. And honestly, we had no inkling that there is this big inversion there. So something about our mapping approach, maybe because we mapped to a reference genome, we couldn't see it at all. And so that's, yeah, it's just one of the really cool aspects of long read pangenome approaches. You start to see things on the landscape that you just couldn't see before. And so, so yeah, and I think, you know, it's the kind of thing where we really need to follow up with some experimental work. In fact, I've told Bohao you got to write a grant on this. And actually, I'll take this opportunity to keep others away from this particular idea, since we're going to submit a grant. But yeah, we want to take different inversion genotypes and experimentally infect them and see how they respond. You know, I think with the MHC, we still have not done a good job of looking at how that's varying in different populations or through time. And so the long read data shows that the MHC is a lot more complex than we imagined from short read data. So it's just it's set up some nice follow up studies.
Marty Martin 46:17
Wow. Al right, well, here's where we risk getting deep into the immunological weeds, and everybody turns off the show. So I'm not going to ask too many probing questions, but,
Scott Edwards 46:25
No, man, this is the good stuff.
Marty Martin 46:27
A lot of the genes that were in the inversion. I mean, well, ask you guys a question, how many genes that were in the inversion were obviously immunological? Was there much there? I can't remember.
Scott Edwards 46:36
There wasn't a huge amount. I mean, it's not like it was some different MHC haplotypes. I think there was a gene that facilitated proliferation of T cells upon infection, you know
Cameron Ghalambor 46:48
That's a big one.
Marty Martin 46:48
Yeah, that'll that matter, yeah, yeah.
Scott Edwards 46:50
I mean, you know how these gene ontology enrichment things go, it's like you're kind of grasping at straws. But, I mean, they really are. They're hypotheses to be tested at a later time. So I think the link with immune system is its potential, although certainly not, not not proven. But on the other hand, there's probably a lot of you know the MHC, the major histocompatibility complex. For those of you out there who are not initiated into this wonderfully complex world that in both the house finches and some other birds we're looking at, it's often still challenging to assemble, even with long reads. And so, you know, in some ways, the absence of MHC in Bohao's paper is a bit conspicuous, and that's because it is there and it is assembled, but sort of on a smaller frag, it's not a whole chromosome fragment. And yeah, the MHC is like, well, it's, it's one of these nagging, always one step ahead of us, you know?
Marty Martin 47:50
Well, it kind of evolved for that reason. If it wasn't one step ahead, none of us might be around so
Scott Edwards 47:56
Well, it's, it's proving that over and over again, because, like, even, like, for my postdoc, I and it's actually fun to come full circle studying house finches and scrub jays now. Because actually, for my postdoc, I was looking at MHC of these same birds using PCR methods and so, and it's just been just humbling, I would say, to revisit these systems with long read day and say, oh, okay, we actually weren't seeing very much of the MHC.
Marty Martin 48:22
Well one of the genes that you guys did see that didn't have anything to do with the MHC, but I found really intriguing, and I wanted to hear you know what your thinking is, without asking you to spill the beans, if this is part of the proposal that you're writing, it's about telomere length. So what do you think that's about? I mean, the upshot was that after the pathogen showed up, the survivors presumably had shorter telomeres than before the pathogen went through. So why did the telomeres shorten? I mean that that alone is sort of strange, because we think of short telomeres as aging vulnerability in a, you know, simple sense.
Scott Edwards 48:59
Yeah, I think, I think there's gonna be... the telomeres was a really interesting angle. And I think boho can take full credit for looking at that. And I think that, well, one is, I'm actually really excited about this new way of, sort of measuring telomeres. I don't think the field will switch entirely this, like long read, sequencing everything, because that's it's just too expensive, but I will say that it does give us a much finer look at what the telomeres are actually doing, and, for example, which telomeres are actually shortening faster than others. We didn't get into that at all in the house finch study, but it's something we want to do. My guess is that telomere biology is going to be relevant at the level of individual development and aging, as you say, but it will also be relevant at the population level. For example, there's a really cool paper published in Genome Biology and Evolution, just recently, with theory suggesting that telomere length, sort of average telomere length, of a species, could be influenced by their population size, wherein species with small populations would tend to fix telomere variants, such as short telomeres, faster than species with large population sizes. So the finding we had in house finches could reflect some sort of selection, or it could represent some sort of overall stress in the population, which could result in shorter telomeres. I think again, it's for one of these observations is sort of crying for a follow up study.
Marty Martin 50:32
Yeah, do you, sorry to keep probing this, but you said something that, you know, when I read that result,
Scott Edwards 50:37
Were you one of the reviewers of our paper?
Marty Martin 50:38
No, no, no. I never saw the paper until I read it, I promise, Scott, I never saw it before. Do you think it might be a cellularity change? Do you know anything? I mean, are you getting shorter telomeres, because the dominant circulating cell type and the blood samples that you had is different than it used to be? I don't know how much the cells vary in the average length of their telomeres?
Scott Edwards 50:59
That's a really good question. And honestly, despite our use of long range sequencing, we were obviously very coarse with regard to cell types. And yeah, and this has gotten me excited to look at some of your papers, Marty, because you probably talked about this. And you know, people probably in humans, for example, they're probably looking at single cell telomere length distributions. And, absolutely, I think there very well could be differences in the cell composition of blood or tissue samples. Yes, all of those are really great details. And I think this whole this new wave of 'omics technologies is just beckoning us to look at the cellular level, we're going to find a lot more granularity than than we imagined.
Cameron Ghalambor 51:44
Yeah,so, you know, in the house finch, example, you know, you have this exposure to this pathogen, and, you know, pre- and post-, and so, you know, you you could clearly see, for example, that this frequency of this inversion had, like increased in the population. But one of the other things that you were able to infer from some of these pangenomic analysis were the distribution of fitness effects for these structural variants. And so I'm thinking like for the ecologists who are like listening to the show. How are you able to infer? Can you explain how you can infer whether changes in the genome are beneficial or deleterious, just by looking at this kind of sequence variation without like measuring something like experimentally like survival or fecundity, or something like that?
Scott Edwards 52:44
Yeah, great question. I think you're sort of exposing the fact that, you know, at their core, population geneticists are just very lazy. And you've also exposed a little bit of the smoke and mirrors that is population genetics. That's not entirely fair, but, you know, yeah, the distribution of fitness effects is a really interesting development that was really kind of pushed by, yeah, folks like Adam Eyre-Walker in the UK. And really interesting. You know, the way, the fundamental observation is you have this thing called the site frequency spectrum. It's a way to describe genetic variation in populations, and basically it's just the count of the number of variants, either SNPs or structural variants, that have reached a certain frequency in your population. So say you have 10 individuals, some of your variants will be present, say in nine, nine versus one individual, eight versus two individuals, seven versus three, six versus four, etc. And we can, of course, if we use an out group, we can determine what the derived allele frequency is, because we know the ancestral one. So estimating with distribution fitness effects basically builds on the fundamental assumption that derived variants that are rare on average, are more likely to be deleterious than advantageous. If they're advantageous, they would have swept to fixation already, and we probably wouldn't even see them as polymorphisms. If they're neutral, then, well, every mutation has to start rare, but the likelihood that a neutral one will be drift to mid-frequency is probably greater than if it were deleterious.
Scott Edwards 54:37
And so you can, you can imagine how if you turn to take that logic to its extreme, you can sort of imagine that, on average, rare variance will be deleterious, whereas those that have reached higher frequency will be neutral or nearly neutral, and that's essentially the logic of estimating distribution of fitness effects. There's also some cool ways to play off the number of fixed differences between populations versus the number of polymorphic differences, to come up with ways of estimating sort of the average proportion of variants, for example, that have gone to fixation by selection. Again, this is really foundational work by Adam Eyre-Walker in the UK. And so the question is, we see a bunch of differences between, say, two species. How many of those differences were driven to fixation by natural selection versus coming to fixation just by genetic drift? And it turns out that, you know, again, through this really clever kind of looking at ratios of fixed differences versus polymorphic differences, we can actually tease that out. And so that's,
Cameron Ghalambor 55:51
That's super cool.
Scott Edwards 55:52
Yeah, it's super cool. I know it's like, it's like, black magic. And now you're absolutely right. Experimentalists, I think, would be just inwardly chuckling, saying, "Oh, how naive there." And you probably remember from the days of allozymes and protein variants, there were lots of really cool studies, people taking this variant and looking at how flies are faring in this environment versus so there's a lot of experimental confirmation we, you know, that's the species that we study are hard to do that. And I also wonder if the effects of individual genes, you'd be able to actually see them in the context of a whole complex you carry out, you know, with thousands of genes. And that, that is, I think, I mean, what we've been finding it was reported in Bohao's papers that, on average, these structural variants are slightly deleterious, that they're and that's kind of interesting, because I think the publication bias has been such that we've kind of these inversions are all adaptive, capturing all this amazing variation. So our result is not contradicting that. It's just saying, like, Well, wait a second, like most things are deleterious. Of course, as usual, we're like, dumping water on the party, on the fire and stuff.
Cameron Ghalambor 57:05
Yeah, but I that seems very, very profound, actually, and probably very important, because exactly for the reason you said, like, I'd say every paper I've ever read that's mentioned copy number variants, it's in some sort of adaptive context, and we're not thinking about all the other examples where these kinds of structural variants actually might not be. I mean, it's just like any other kind of, you know, genetic variation in the population.
Scott Edwards 57:38
Yeah, that's right. And actually, that to me that's a really beautiful thing about population genetics and sort of the neutral theory, in general. Of course, there's a neutral theory in ecology as well. And I just it's, we're very lucky in sort of population genetics and evolutionary biology to have good null hypotheses, like what do we expect if something is neutral? I think it's harder to have those null hypotheses in some areas of biology, you know, maybe like molecular biology or even animal behavior, it's sometimes hard to know, well, what's the neutral expectation? And so it's a nice aspect of population genetics that we can we're comfortable saying, well, some of this is neutral or even not so good for you.
Cameron Ghalambor 58:23
Yeah.
Marty Martin 58:25
So one more sort of population genetics question, Scott, we briefly mentioned that this species is sort of native to the Western US and then showed up in the 1940s in the East. So some fraction of the individuals are sort of invaders or introduced. You showed some statistics in the paper of a signature that you would expect with regard to, you know, genetic diversity and these kinds of things. And those were sort of validations based on other, you know, data that you and other people had done before, but now that you have these long reads, were there any surprises beyond what we were just talking about with respect to fitness? Was there anything surprising? Did you have maybe more genetic diversity than you expected? Or was there anything new when you use this technology as opposed to old approaches in these introduced populations?
Scott Edwards 59:11
Yeah, great question. I mean, I think you know, genetic diversity is fairly easy to measure. I think what was surprising with some of the measurements that Bohao took was that we could detect, for example, lower diversity in structural variance, sort of non-SNP variance, and so things like insertions and deletions, we actually were seeing lower diversity in the East than in the West. And we have to remember that you know, the magnitude of the reduction of diversity was not great. I mean, you know, anecdotally, these birds went through a bottleneck of about 200 individuals, which is some might consider actually a mild bottleneck. I think a strong bottleneck would be something more like 20 or 10 individuals. And so, you know, the fact that we could see reductions in diversity even in these structural variants, I thought was really, yeah, and in some ways, in sort of very indirect way, it shows that these variants are responding to genetic drift in the same way as SNPs are in a lot of cases. And so, yeah, it's, I think that's, it'll be interesting to look at individual cases in more depth and see how they were affected some some structural variants were probably affected a lot more by the bottleneck than others, and they're probably complex interactions between selection and drift as well.
Cameron Ghalambor 1:00:32
Yeah. So Scott, I have one last question for you, sort of a big picture kind of question. So one of the real benefits of being able to look at snips and looking at, especially like in the context of GWAS, looking at like different regions of the genome that contribute to quantitative traits. So kind of this connection between population and quantitative genetics, you know, many of the traits that we're interested in have a polygenic sort of basis to them, genetic architecture underlying them. So looking to the future, do you see these kinds of pangenome approaches being able to sort of replace the kind of more traditional, you know, looking at SNPs. And is there an opportunity here for more of, like, identifying these structural variants to underlie these more, you know, complex quantitative traits?
Scott Edwards 1:01:38
Yeah, absolutely. I definitely think there is. And I think the folks working on plants, especially sort of agriculturally important plants, are probably already doing this in terms of associating structural variants with certain growth patterns or certain adult sizes or whatever. Actually, I was just, it's nice that you kind of brought quantitative genetics into the picture, because, you know, it's an example of, like, a really essential component of evolutionary biology that a lot of folks, unfortunately tend to dismiss these days, is, but you're absolutely right that is continuous traits and polygenic traits that is very much still in play. And I think that structural variants will have a big role to play there. There probably are some cool papers and plants. I don't know of any, even in humans, that I can think about, although they're probably out there. But I think with the long read data and the pangenome approaches we have, we'll have enough markers around the genome, structural variants around the genome, to be able to start to get at these questions,
Cameron Ghalambor 1:02:45
Yeah, because I'm thinking about, like human height, we've got like tens of thousands of SNPs.
Marty Martin 1:02:45
I wondered if you were going to bring up that example,
Cameron Ghalambor 1:02:54
Yeah, and you know, we're still only explaining a small fraction, well, maybe not a small but still, there's a lot of unaccounted for variation and and perhaps that variation is tied up in in these structural variants, and-
Scott Edwards 1:03:08
Absolutely, and I actually, as soon as we finish this call, I'm going to Google and see if I can find this, because there probably are some cool studies looking at structural variation in human height. It's just sort of an obvious question.
Marty Martin 1:03:20
Well, good. So, so you can get to that Google search, Scott, we always end with sort of, you know, blue sky. Is there anything that you want to say? Advice to graduate students, appeals to funders, inspirational stories to collaborate. I mean, what kinds of things? What else do you like to say?
Scott Edwards 1:03:34
Well, I'll keep politics out of it, for now. But yeah, I feel very privileged to be an evolutionary biologist today, it's a super exciting time. I would encourage younger scientists, folks thinking about a career in science to just go for it and having a good background in evolutionary biology. It's very much a good background in biology generally. And I would say, you know, gosh, being a scientist, super cool. It's just a really fun opportunity you gotta get to teach the next generation. Then, honestly, you use, maybe it's because I'm getting a little advanced in my career, I start thinking these sentimental thoughts, but you're really, you're really passing a torch, the torch of knowledge and a particular discipline, to the next generation. So yeah, just, just go for it. Follow your dreams, and hopefully, you know, you'll get the support along the way to make an impact. Yeah, and don't forget to sprinkle it a little bird watching in there. That's great antidote for your paper rejection or grant rejection or whatever. And it's a great, really good place to turn to when everything, everything else seems bleak,
Cameron Ghalambor 1:04:45
Nice. Well, thanks so much for taking time.
Scott Edwards 1:04:48
Thank you. This was great. You guys make a great team. I love it's like click and clack almost. You know,
Marty Martin 1:04:57
Laurel and Hardy
Scott Edwards 1:04:58
Exactly, exactly. That's true. That's right, exactly. I'm really excited. Thank you so much for the opportunity. It's super fun.
Marty Martin 1:05:07
Yeah thanks for doing this.
Cameron Ghalambor 1:05:08
Thank you.
Cameron Ghalambor 1:05:20
Thanks for listening to this episode. If you like what you hear, let us know via Twitter, blue sky, Facebook, Instagram, or leave a review wherever you get your podcast, and if you don't like what you hear, well, we'd love to know that too. All feedback is good feedback.
Marty Martin 1:05:35
Thank you to Steve Lane, who manages the website, and Molly Magid for producing the episode.
Cameron Ghalambor 1:05:39
Thanks also to interns, Dayna de la Cruz, Caroline Merriman and Brady Quinn for helping with this episode. Keating Shahmehri produces our awesome cover art.
Marty Martin 1:05:49
Thanks the College of Public Health at the University of South Florida, the National Science Foundation, and our Patreon and Substack subscribers for their support.
Cameron Ghalambor 1:05:56
Music on the episode is from Podington, bear and Tieren Costello.