From First Principles - New Rules For Heredity (Non-Mendelian Inheritance of Epigenetics) (EP 44)

Episode Date: June 26, 2026

Hosted by Lester Nare and Krishna Choudhary, this episode marks Krishna’s return to the studio after paternity leave — and the timing could not be more fitting. Today’s deep dive is about inheritance: not just the classic Mendelian rules most of us learned in biology class, but the stranger, more dynamic world of non-Mendelian epigenetic inheritance.Starting from Gregor Mendel and his pea plants, Lester and Krishna rebuild the foundations of genetics from first principles: dominant and recessive alleles, Punnett squares, chromosomes, fruit flies, DNA, and the physical mechanism behind inherited traits. Then they move into the “software layer” of biology: epigenetics, DNA methylation, chromatin packaging, RNA interference, and paramutation — cases where the genetic code is present, but the cell’s machinery silences or rewrites how that code is used.The episode centers on a new Nature Genetics paper, “Non-Mendelian inheritance of DNA methylation patterns in mice,” which suggests that non-Mendelian epigenetic inheritance may be more widespread in mammals than previously understood. The conversation also covers why Oxford Nanopore sequencing made this kind of analysis possible, why methylation patterns can be hard to trace across generations, and what all of this could mean for disease risk, drug response, sex differences, evolution, and the long-running nature-versus-nurture debate.Summary Mendel’s rules — how pea plants, true-breeding lines, dominant and recessive traits, and Punnett squares gave us the first mathematical laws of inheritance. The first cracks in Mendel — how chromosomes, fruit flies, sex-linked traits, and linked genes showed that inheritance is more complicated than independent assortment. DNA as hardware, epigenetics as software — why having a gene is not the same thing as expressing it, and how methylation and chromatin packaging can silence parts of the genome. Paramutation — how one allele can change the expression state of another allele across generations, creating inheritance patterns that do not follow standard Mendelian expectations. Oxford Nanopore and the technology shift — why long-read sequencing and direct methylation detection make it possible to trace epigenetic marks back to the parent they came from. The mouse methylation paper — how researchers used collaborative cross mice to show that most methylation inheritance looks Mendelian, but a meaningful fraction appears to follow stranger non-Mendelian rules. Why it matters — potential implications for clinical genetics, disease risk, drug efficacy, sex-specific biology, and the relationship between nature and nurture. Support the showDonate: FFPod.com/donateFollow: @FFPod on X / Instagram / TikTok / Facebook

This episode is brought to you by Accenture. When your advertising operations fall out of sync, everything else follows. Spotify and Accenture are working together to reinvent the rhythm of ad sales, using automation, analytics, and smarter workflows to simplify campaign delivery and access better data across the business. The result? Less time spent on operations, more time connecting brands with the moments and fandoms that matter most.

Learn more at Accenture.com slash Spotify. This spring, denim gets a softer, lighter update. Introducing Old Navy's drapey denim wide leg, a new fit that moves with you. It's everything you want denim to feel like for summer. Easy, breathable, and effortlessly cool. With a fit that creates natural movement and a wide leg that feels modern, not overwhelming. Plus, that signature, wait, for this price, moment. Old Navy's drapey denim wide leg.

And not everything is about genetics, meaning not everything. is about the hardware of the DNA. It's also about the sort of software, which is the epigenetics part of it. They've done a thorough analysis of the genome and found that up to 7% of the DNA obeys these kinds of different weird rules. 7% that's 1 in 14. So 1 in 14 genes, that's not a joke. Hello, internet. This is your captain speaking Lester Nare and breaking news. The candy man is home. Oh, ow. Oh, what is that?

Someone called a dentist. Is that a cavity? We are back here in studio by popular demand after months with just me and my back with our resident PhD, Krishna Chowdery. Welcome back, good sir. How are you? How's it going, man? I have barely had any sleep over the past two months.

But you've been doing a great job with all of your content. and now we are back in the studio. It's been crazy. It's been really crazy. And I remember, like, the baby came, you know, for our viewers, the baby came two weeks early, and the day that it happened, we were supposed to record two additional episodes

in preparation for my paternity break. And I texted you from the hospital like, hey, it's today, abort mission. Abort mission. We're going to just roll with it. But I'm glad to be back. Seems I've missed quite a bit. Quite a bit has happened in the world.

The most important thing that happened is probably Lewis Hamilton won his first race with Ferrari. Finally, the promise of Lewis Hamilton has paid off for Ferrari. So I'm wearing my Ferrari gear for that. We are very grateful to have Krishna back here in the Promise Land at FFP Nation. and in honor of you entering fatherhood, today's deep dive is going to be about non-Mendelian epigenetic inheritance. And so we're going to learn about the science from the ground up today because this is from first principles. So I was recently blessed with my first child and questions of inheritance are very former in my mind, right? questions that every person of color asks, like, how dark is the baby going to be?

You know, among other things, like what color are his eyes going to be? Is he going to be an athlete? But, you know, that's one of the big ones. If anyone out there who's brown, they know what I'm talking about. And so for most of the traits that are out there, you can use rules that were established 160 years ago by this guy called Gregor Mendel. And these traits follow Mendelian laws of. inheritance. This is the stuff that you learn in like AP Bio, if you ever took AP Bio in high school.

The rules are pretty simple and we're going to go over them. There's like these dominant genes, recessive genes, some, and by and large, those rules work. But as with all things in biology, there are always exceptions. And genetics is something that has a bunch of exceptions because you're, you know, down to the molecular level and there's always weird stuff happening. And not everything is about genetics, meaning not everything is about the hardware of the DNA. It's also about the sort of software, which is the epigenetics part of things. This is a very dynamic package. This is how the DNA is wrapped around inside the cell, how much the DNA has access to protein-making machinery, things like that. And those things can change much more quickly

than the actual genetic sequence itself. And we're only just starting to get into studying that higher level of genetic machinery, right? And it turns out that those traits can also be inherited. Like this software package on top of the hardware is something that is also passed down. And we're just starting to understand that. There's exceptions for how it gets passed down.

Most of the time, again, it follows Mendelian rules, but sometimes it doesn't. And that's what this story is about. There's a new paper out of Johns Hopkins University and Texas A&M that is documenting a new type of non-Mendellian inheritance of epigenetics in mammals. And it's showing that this genetics is far stranger and far more dynamic and far more interesting than I think Mendel could ever have imagined 160 years ago. And so, like, why is this a big deal, right?

Obviously, anything related to genetics is a big deal, because genetics is like the blueprint of life is the blueprint for who we are. But it touches on everything from like why identical twins can have different disease outcomes, right? Identical twins have the same exact DNA, but sometimes one twin is going to have one disease and the other twin is going to be just fine. That's because the software package on top is different, right? Diseases can skip generations. That's something that happens to do with epigenetics. and sometimes the same gene variant can be benign in one person.

If you do the genetic sequencing of two individuals, they'll have the same gene variant, but one individual will get the disease and the other one won't. Again, that has to do with the software on top. So this paper marks like a major shift in how we think about epigenetic inheritance, especially in mammals. And it shows that this is not actually an exception. That's the big part. Okay?

They've done a thorough analysis of the genome, and found that up to 7% of the DNA obeys these kinds of different weird rules. 7% that's 1 in 14. So 1 in 14 genes, that's not a joke. That's pretty statistically significant, as we like to say. Yes, exactly. Like, that's something that you do have to worry about

if we want to, like, get a really comprehensive idea of how genetics affects outcomes and all that stuff. So we're living through kind of a quiet revolution in the understanding of heredity. And in this episode, what I want to do is go through how it all started 200 years ago. We're going to start with Mendel. Then I'm going to take you through some of the history of the field. And then why this particular paper is happening now and not 10 years ago or 20 years ago.

There's some key technological advancements that have happened in recent years that is enabling this kind of analysis. And it's very, very cool. There's an interesting, there's a mutual friend that we have that had been doing work around some of this epigenetic stuff. you know, through grant funding through the NHS. And what is, I think this is very timely because there's been a lot of chatter and social media around, you know, this idea of generational trauma. Yeah. That's sort of a non-scientific term. Yeah.

And this, the idea that there is an impact on outcomes that are related to these concepts of like generational trauma. And I think this is going to start to brush up against some of these social conversations that have been happening. And I'm super curious to understand, you know, particularly your point about why now, are we able to start to get a better idea of what's going on on the epigenetic layer? Yes, that part is really, really cool. The technology there is really, really cool. And what you're alluding to is the fact that, you know, trauma is something that is not going to affect the genetic layer, but it is going to affect the epigenetic layer because the epigenetic layer is the one that is dynamic and changes throughout the life of an organism. we by and large have the DNA that we were born with,

unless, you know, if you look at tumors, there's a lot of mutations there, and tumor cells have very different DNA compared to normal benign cells, right? But epigenetically, the reason why this cell is a skin cell and this cell is an eye cell, and the cells inside my brain are neurons, that has to do with epigenetics, right?

Because the cell is, is making a choice what part of the gene to read out from and what part to ignore. That's the software level that I'm talking about. So it's extremely important for us to understand, right? So we're going to start, as always, on this podcast. Guys, I might be a little rusty. It's been two months of no, as I said, barely any sleep. So just bear with me if I'm a little slow.

But we're going to start as always with the history. Yes. We're going to go back. the story of genetics probably begins in the six um 160 years ago improbably in a monastery okay Gregor Johann Mendel joan he was an augustinian monk who spent eight years crossbreeding pea plants and the the green circle you see the guy in the green circle in the back row that's mendel along with all of his other monks yes in his monastery he was crossbreeding pea plants for like eight years he tracked 29,000 plants across multiple generations.

That's huge volume. This was his hobby, and he took it very seriously. He's the first guy to really do large-scale statistical analysis of gardening, effectively. Okay? He was the first green thumb. Yeah. Yeah, yeah, yeah. Before Zach Gallifanakian had the Netflix show about gardening, which, by the way, a good friend of ours,

Trevor is the editor on that show, so go check it out on Netflix. It's a really good show, actually. So you can see his garden today, actually, the Mendel's Garden. You can see it in Berno, which is currently in the Czech Republic, or I think now it's called Chechia, if I'm watching the World Cup correctly. I've heard it, Chechia. Chequia, yeah. But it used to be called the Czech Republic.

That's what I'm comfortable with, but Chequia, Chechia, one of those. You can actually go see the garden and see the beds where. he was planting the pea plants, which I think is really cool. One day when FFP Pod goes on a field trip, that is definitely a place we're going to visit. It's the birthplace. That garden is the birthplace of genetics, right? I think that's so cool. So why peas, first of all? Okay. This is something that we talk about a lot on this podcast and is a staple for the biomedical research community, which is model organisms, right? You have to choose the organism that you want to, that you want to study as a system, and you have to make that choice very deliberately for very specific reasons.

Mendel is one of the first guys to do this. He chose peas because the bar, one, the garden pea is actually pretty easy to grow. You don't have to worry about it too much. It has easily observable contrasting traits. These are the traits that he was studying. For example, is the pea round or wrinkled? That's a binary. You can look at the thing and there's no middle ground. It's either round or it's wrinkled, right? When you cut up the pea, is it yellow or green inside? The flower, is it white or is it purple? Like, these are all binary things that there's no, you know, when you're trying to do

statistical analysis by hand, right? You don't have like computer vision nonsense, right? And you just got a notebook. You're just making tally marks. So you need really easily identifiable traits that are in binary. The pea plant has seven of these, which is really nice. Okay. So the idea of what you're saying is you want to have the fewest amount of variables to sort of control for. One, at least in this time period, from a management, it's like a management challenge.

Exactly. But two, there's less to deal with in terms of judging outcomes. Yes, exactly. There's no like subjectivity in what the offspring of this and this was, right? You can just be like, okay, I bred round and wrinkled and I got this many wrinkled, this many round, and I can just make the tally marks. Right? The other big thing about peat plants is they self-fertilize. So that means that Mendel can really easily control crosses by you can manually just transfer pollen. You can like take like a Q-tip or I think he used a paintbrush at the time. He would take the pollen from one paintbrush and put it in the stigma of the other one to like crossbreed manually. And I think we've got a we've got a photo of that in the next photo of photo five. This is this is how we did it. He would cut the stamen, which is the male part. or am I no no the pollen is the male part so he would cut the pollen part and then he would use a brush to take the the sperm effectively and put it in the female part of the of the flower so this was it was a controlled breeding yes because also again the the surface the surface area of how you would do so was was very discreet yeah exactly it's super easy to control right and the first thing he does is he develops true breeding lines what that means is he's going to self-pollinate a bunch of pea plants, which pea plants are totally chill with. They don't care about like, oh, lack of genetic diversity, all that. They're actually, that's one of the reasons why it's a good thing to use pea plants. And you keep self-pollinating until all of the offspring are something called true breeding.

Meaning they always produce the same offspring identical to the parent. It's like you're making clones over and over. So now what you do is you're starting out with a clean slate. Yes. Right? There's no genetic variability in this plant. and its gametes and its sperm and the egg, right?

It's like if I were to take the sperm and the egg of this and combine it, I would just get the same thing again. So it's like a feedback cycle and completely clean, right? For example, offspring of purple flour times purple flower, always going to give me purple flour at this point. That's why it's called true breeding. Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick.

Bet MGM and GameSense remind you to play responsibly. 19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact connects Ontario at 1-866-531-2,600 to speak to an advisor. Free of charge. BetMGEMGEM operates pursuant to an operating agreement with Eye Gaming Ontario.

If you want a $3,000 a month payday for life, what would you feel free to do? Maybe take a long weekend, every weekend? or try a bunch of new hobbies. Would you feel free to upgrade and listen ad-free? Don't worry, we get it. Every $20 ticket could win you $3,000 a month for life and supports life-saving cancer research at the Princess Margaret. Feel free to buy your payday for life ticket today.

Raffle number 155-2194. Please play responsibly. Okay. So the idea is we have everyone is effectively genetically the same. Yeah, and there's no nonsense in the, like, from one generation to the next, I'm not getting weird like effects. It's not just that this line of the generation is the same. It's that subsequent and prior to some number of generations forward and back are consistently

coming in as the same. So you're also removing the idea that, oh, this one was, but the next one might not. Yeah, yeah, exactly. So now when I do crosses in, I take the parent generation, the P, and I do a cross, let's say from a purple and a white, I'm going to get some weird effects, right? But I know that's not from the purple doing its own thing or the white doing its own thing. I know it's because of the interaction of these two. Does that make sense?

That's why you've got to start with that true breeding. This is a good, this is a really important point, which is we're creating the canvas from which we're going to test in a way that removes any prior muddying of the waters. Yes. So that we can look downstream and say we had a clean foundation with which we're building from. Exactly. Yeah. And now I can isolate the effect to an interaction between two individuals that have different genetics.

That makes sense. Okay. Here's what he observed. Okay. So he takes the parent generation. Let's just talk about the flower color right now. There's seven other traits that he could look at, but they all actually obeyed these rules. But it's nice to talk about flower color because it looks good on TV.

Okay. So I'm going to take a true breeding purple and a true breeding white plant. I'm going to breed those two. the first generation, that's the F1, the first filial generation is how he put it, that one has all purple. Immediately you notice something weird. You're not getting a mix, right?

You're not getting like lighter purple, like pink. It's not a blend. It's not a mix of half and half white purple. It's not even half and half. It's fully all purple. Okay? First thing that's kind of weird. Okay?

It's almost as if the first general, generation is also true breeding. Right? So you're like, hey, so is the first, is the white just like doing nothing? Right. Is everything coming from one? Is everything just coming from one? What's going on?

That's when you take, do the second cross. So you do a second filial F2 generation where you take two of the first generation ones and you cross them. And now you get something weird. You get the white back, but the ratio of purple to white is always, and I mean always three to one. in his notebooks. He's got 29,000 plants that he's doing this with,

and it's almost always three to one. Like statistically, that's the ratio that he's getting. Okay? I think I see it. It's very weird. It's very, and with no context and no foundation. Yeah, yeah, yeah. Now it's like trivial.

Right. But it's why did we, we had two different, the first generation, all only similar to one of the two parents, but the second generation that was bred from that F1 generation, where everything was the same as one parent is a three to one ratio of one parent to another. Yeah. Almost all of the time.

Almost all of the time. Okay? So let's try to figure out how this would happen. Okay? Mendel doesn't know. Right. Okay? Mendel doesn't know anything about genes.

He doesn't know anything about chromosomes. He just knows plants are doing this. Okay? So he takes a purely sort of empirical approach, which is I'm just going to describe the mathematics as I see it. Okay? I want to make up some rules. that are just math rules, okay, about what happens.

First thing he figures out, and what we've already seen, is the principle of dominance. There's two traits. One of them is going to dominate over the other. Okay? In this case that we saw, it was the purple. The purple trait, which now we call an allele. There's two different types of alleles, the purple and the white. And the purple allele is going to dominate over the white allele, which will call recessive.

So there's dominant and recepts. That's the first thing you established. Second is something that he establishes called the law of segregation. This is the segregation of traits. Okay. What he says is there's actually two copies of an allele that every individual has. And if I have a true breeding plant, let's say the trait that we're looking for is the color of the flower. And we're going to denote the type of allele by a capital A if it's dominant.

and a lowercase a if it's recessive. Okay? Now, if I have something that's true breeding, that means that I have two copies of the same dominant allele. Capital A, capital A, that's the one on the left up. And then the white true breeding is going to be lowercase A, lowercase A, or recessive, recessive. Okay. Now, when I cross them, there's going to be four different ways to choose the A and the

A, right? I can choose the first capital A and either of the two lower cases or the second capital A and all it. That's four total. But every single time I'm going to get a capital A and a lowercase A. So that first generation is going to have at least one dominant copy, capital A, all the time. Every single one is going to have a capital A and a lowercase A. And because the capital A is present, the dominant is present, I'm going to get the dominant phenotype. And the thing is going to be purple.

But now let's take the first generation that capital A lowercase A, which we're going to call heterozygous because there's hetero there's two different copies right one of each one of each yeah yeah compared to homozygous which is what we had in the parent generation that true breeding is now what we call homozygous so if we take two heterozygous parents and we cross them now commentatorically i'm going to get three offspring that have at least a capital a there's actually going to be one that's kind of like the true breeding parent in the beginning, the capital A, capital A, right? That's the one we have in this top left. Yeah, in the square. Those are called punnet squares now. But the top left, that's capital A, capital A, that's going to be purple. The two on the off-diagonal, that's going to have a capital A and a lower-case A.

But because there's at least one, I'm still going to get purple. And only one out of the four is going to have helmozygous recessive, lower-case, a, lower-case, and that's what's going to be white. This is interesting. And I think a key point here is we had the pure, the pure breed generation at first, two dominance from one, the purple, two recessives from the white. The first generation, because you always are going to have that dominant from the purple from the one side, you're always ever going to get this dominant recessive combo. Yeah. And so the reason why that first gen, first gen, always had a dominant.

Always has dominant. Always has dominant. And so that's why you don't ever see the white. Right. Because it's always being basically overrun by the recessive. white is being overrun by the dominant. The thing that the where you get the diversity of color begins to arise when you have two dominant recessive as the parents. Yeah, the heterozygous pair. The heterozygous pair. And that's finally when you start to see from the original

P1, the recessive begin to appear, but also from the the math you brought up earlier, it makes sense that only one. It's always going to be three to one. because of the way that it's this dynamic pair from each. Exactly. It's always going to be three to one. So this is an empirical rule that he comes up with. It makes sense. And the other nice thing about this kind of mathematics is it explains something else,

which is the law of independent assortment. What he did, and this is something that we're seeing actually in fruit flies, because the particular visual that I found was for fruit flies, but he also did this for pea plants as well. Okay? What you can do is you can take two true breeding, you can take a true breeding parent generation with different traits.

In a pea plant, you could say it's always purple and it's always tall. And I'm going to cross that with always white, always short. Okay? So now I have homozygous for two traits in my parents. I'm going to get some offspring that are, Again, always purple, always tall, because the tall part is going to work the same way as the purple part. And so everyone's going to be the same.

But now I do a cross between those guys. Now, if you just do the mathematics, there's going to be 16 different choices instead of four because you're doing four by four across. And the mathematics is going to be such that I'm going to get a ratio of nine to three to three to one. Okay? nine being I'm going to get the dominant parts of both. Three being I'm going to get the dominant of one, the recessive of the other. The other three being I'm going to get the dominant of the second one, recessive of the first one. And only one out of 16, I'm going to recover that first generation of recessive recessive.

Okay? It scales. As you get to more traits here, the ratio of distribution effectively scales accordingly. Exactly. Accordingly. Yeah. And in this particular one, we're looking at fruit flies where we're seeing eye color and body color. Okay?

And eye color red is dominant and body color brown is dominant. Black for both is recessive. And so we're seeing nine to three to three to one. And you can go through the punnet square there and figure out how each of the alleles was matched. And you can convince yourself that this is the case. That makes full sense. Okay. This is called the independent assortment law.

meaning that the two different traits independently do combinatorics. They're not like related to one another. It's not like the eye color goes with body color. It's like correlated in some way. So that was his thing. Okay. So these are Mendel's laws. He presented these findings in 1865, published in 1866,

experiments on plant hybrids. The response, complete silence. Papers ignored. It's cited a handful of times. over the next 35 years. He died in 1884, unknown as a scientist. Terrible. And his discoveries were buried in some obscure journal.

Terrible. Okay? 35 years later, we've got three botanists who start doing the same thing. Okay? Hugo DeVries in the Netherlands, Karl Korrins in Germany, and Eric von Schermerk in Austria. And they independently figure all of this out.

And then as they're deep diving into the literature, because the first thing you do is, has someone else done it? You know, in the modern day, there's a joke in academia. There's always like a Zhang at all that's done it, that's done it before. So back then, they're doing the same thing. They're trying to find, like, has someone else done it? And all three of them actually find Mendel's old paper in the literature. And they realize, so these are the three.

And they rediscover Mendel's work in 1900. And they're like, well, somebody beat us to the punch. And this is now colloquially known as the rediscovery of Mendel. And it launched modern genetic. because now they're like, okay, so there is some kind of level of understanding that we can figure out, right? Now, the next focus is always how do we go from empirical math rules to some kind of mechanistic understanding? Right. Right.

What is actually happening in the organism that is doing this? Right. So we can predict, but we don't know how it works. Yeah, exactly. Okay. So first question, natural question, where are the alleles? Yeah. Right? You're saying there's these two types of alleles. Well, and it seems like there's like actual stuff, right? That's like moving. And like there's like physically two thingies that like get independently that come out and then they recombine and stuff like that. But if that's the case, there should be something I can find inside organisms that is an allele. Right. So Walter Sutton, he's a graduate student at Columbia University at the time. And he proposes in 1902, 1903, that these meant these.

alleles are located on chromosomes. Okay? He was actually working on grasshoppers at the time. And in grasshoppers, he could stain the chromosomes. And what he figured was that there's these threadlight structures called chromosomes. And they come in pairs. So that's interesting. That's interesting.

Because the alleles come in pairs. And they separate during cell division. And if you look at the gametes, which are the sperm and the egg, they each get half and half. So he's like, coincidence, I think not. Where there's smoke, there's fire. Yes, he's like, this could be the physical basis for Mendelian law of heredity. People are still doubting it because it's coincidence, okay?

You could be missing something that also goes half and half. Just because it has similar behavior does not mean it's not a causal link. Yeah, it's not a causal link. There's a correlation. But in order to establish causation, you got to dig deeper, right? And this is where a guy Thomas Hunt Morgan comes in, also at Columbia University. Early 1900s, Columbia was doing amazing with just like early biology, like even before DNA, before, you know, molecular biology. Just like this kind of early rudimentary work.

So Thomas Hunt Morgan at Columbia University, he starts using the Drosophila fruit fly. He's the one who actually establishes the fruit fly as the model organism. Nowadays, the fruit fly is used in neuroscience research, development, genetics, all sorts of things. He's the first guy to be like, let's do, let's go to town with the fruit fly. So Mendel was the pea plant. And then Morgan now moved us into a more complex biological organism. That is a better model organism. Yeah, that's an animal.

So it's like learn and stuff. Right. Like it sleeps. So you can do, like there's so many more things you can do with a drosophila, right? And Morgan's team discovers that there's some. traits that are sex linked like red eye color. Here's what I mean by that. So what he does is he finds a female with red eyes and then a male with white eyes.

And when they have offspring, all of the kids have red eyes. So you could just be like, oh, red eyes like this like, you know, homozygous type dominant trait or whatever. Now let's take the female again that has red eyes. So in this case, this is going to be a heterozygous female, right? Because the male had a white eye. So this female is going to have one red eye allele and one white eye allele. And now I'm going to cross that with a red-eyed male.

Okay? So the male, and we've done attributing with the male to make sure that the male is fully red. When we cross it, most of the offspring, in fact, every single one of the females, offstrings has red eyes. But the males are half and half. They're not even doing three to one. Right. Now they're doing half and half. Right. And he identifies that the red eye allele is actually on the sex chromosome. And the way sex chromosomes work is, we only get one if we're male, right? The white chromosome isn't really a chromosome, if I'm completely honest, right? We have X and Y. Like, if you look at the two side by side, the X chromosome is this big and has a bunch of extremely

important genes. The Y chromosome has absolutely like nothing, right? It's like, it's, it's really tiny. And it's, in fact, it's so tiny that there's, it's like, the way the chromosome is like tied up, there's a, there's a little centromere that like holds the two strands together. And the Y chromosome is so tiny that there's very little room for genetic recombination. And so that's why, like, the Y chromosome, like your white chromosome is identical to your dad's white chromosome, to his dads, to his dads for like 10 generations. Because it's so small that physically you can't like mix things. Anyways, so the point is he identified a particular chromosome with that trait.

Right. And the reason why males are coming up 50-50 is because they only get one X chromosome from the mom, right? So either you're going to get the red one or you're going to get the white one. Right? And that's it. The females, on the other hand, they're getting one from the mom and one from the dad. And the dad is guaranteed to be red. So even if I get the white one from the mom, the dad is guaranteed to be red. So all of the female offspring are going to have red eye color. That's fascinating.

Yeah, yeah. And so we've added another layer here. Yes. Because before we were saying the traits were not correlated. Yeah. You know, color, size. Yeah. But now with sex chromosomes.

Yeah. Now it's like if they're on that sex chromosome, then the math is a bit different. And actually the correlated part is the second thing that Thomas Hunt Morgan did, which is he showed that some of the traits happen to be inherited together. Okay. So, you know, before I was showing you about how body color and eye color under Drosophila, they were independently assorted. Yes. But there's some that if I have red and some other trait, those are going to go together. Okay.

Why would they go together? Well, now it's obvious because if they're on the chromosome, if they're on the same chromosome, then they're coming in a package, right? And so if they're coming in a package, you're going to have this correlation, and that violates Mendel's law of independent assortment. If I order something from Amazon, we thought it was going to come in two separate packages,

which could have two different delivery days. Yeah, yeah. But it actually came in a big box, two little boxes in it, and they were shipped at the same time. Yeah. Very bad, crude analogy. Yeah, but that's exactly what's happening, right?

And this is actually the first major empirical crack to Mendel's laws. Okay, this is the first time where it's like, no, Mendel was actually wrong, but also he did, it's fine. He had no idea. He did it in a cave with a box of scratch. Yeah, exactly.

Yeah, yeah, yeah. So it's fine. Morgan actually won the Nobel Prize in Physiology. He's one of the first Americans to win the Nobel Prize in Physiology in Medicine. And side note about Thomas Morgan, he's quite a character. So he's the nephew of Confederate General John Hunt Morgan,

who got his butt kicked. somewhere in the north. And then the great grandson of Francis Scott Key, who is the author of the Star-Spangled Banner. No way. Yeah, this guy's quite interesting. So he established something called the Fly Room in Columbia that became kind of this legendary place

where scientists from all over the world would come and hang out and learn about how to take care of fruit flies, how to make sure that the crossing is the same because for Mendel it's like pretty easy you just cut the thing you take a paint brush here you got to like make the flies mate and stuff right so it's a bit more of a process

but there's a lot more room to work with scientifically he became this legendary place established the fruit fly as a model organism and you know the lineage of that is seen in every single university

that does serious biomedical research they have a fly lab somewhere doing something is it an incredible something that people in there in their house are trying to look for ways to make these things go away. Not specifically the fruit fly, but just a fly in general. Yeah, fruit flies are a bit. And it is a cornerstone of us getting a greater understanding

not only of our own lives as humans, but the lives of all animals on the planet. And it's still going. Right, it's still happening. There's still so much we can learn from the fruit fly. It's pretty insane. Another cool thing about Thomas Hunt Morgan is he went on, to found the Department of Biology at Caltech when Caltech started in the 1920s.

Caltech was this insane place because like they, so they hired like the All-Stars. Like they just poached all of the big, because they just got a bunch of money. I believe from the Rockefeller Foundation, either them or Carnegie. Whatever, they're dead. But like if they were alive, they'd be pissed that I was mixing them up. But one of them gave a lot of money. And they just, so George Hale, who was kind of the founder at the time, he took this Pasadena community college type thing and was like, all right, we're just going to hire the biggest names. It's kind of like when you start like a new team or something, you just like hire the best coach.

If you got the money, right, you hire the best coach. You hire like Erling Holland or whatever. The perfect analogy for this in soccer and football is PSG a couple of years ago. Oh, really? The French League has been viewed for a while as being sort of a Farmers League. They got Namar. They got Messi. They got Mbapé, Donoram, a bunch of other players.

They just got like literally all the best players in the world. Yeah, at the same time. At the same time. Yeah. That's what Caltech did. That's what they did. In the 1920s. They got Milliken for physics.

They got Thomas Hunt and Morgan for biology. It was just an insane place to like start out like that. I will note it's sometimes. It doesn't always work because it did not work for PSG. They had that. They couldn't win the Champions League. Oh, they couldn't? And then Messy and Mbapé left and Namor left, and then they won the Champions League.

Oh, that's hilarious. So it can work. It can work. Yeah. It worked in Caltech's case. Yeah, certainly. Because they became like a top institution immediately, which usually it takes years of work to do that, right? And his students have won the Nobel Prize, too.

Like John Howard Northrop, he's a Berkeley professor in 1946 Chemistry Nobel for isolation and crystallization of enzymes. and Hermann Miller won the 1946 Nobel Prize in Medicine for x-rays to make mutations. So a big lineage from Thomas Hunt Morgan. Anyways, after Morgan, then you have the modern synthesis of genetics. So this is, we've covered this in a previous episode with James Watson in his discovery of DNA. But Avery, McLeod, and McCarty, they showed in 1944 that genes are from DNA. and then Watson and Crick, together crucially with evidence from Rosalind Franklin,

discovered the helical structure of DNA, and how the sequence of base pairs is really how genes work. We had a huge deep dive on this, particularly the work of Rosalind Franklin, which in part was not included in a lot of the accolades that came afterwards, has since been rectified at least slightly. At least slightly in the modern era. It's a great deep dive if you haven't watched that. already because that also has a great, the science is great, but the palace intrigue in how

the university ecosystem worked and was being established not only in the UK but in the US at the time was really, really fascinating. Yeah. Yeah. And just how many shenanigans went into, so much so much shenanigans. So now we've got kind of a more complete picture of how genes work and how Mendel's laws come from genes. So I just want to go over all of that. Okay. So first we know now from, you know, all of the work is DNA stores the data of these alleles and genes. And what an allele really is is one version of the ATGC sequence. Okay. The different allele will be a different version of that ATGC sequence, right? And the two alleles are different in only a few base pairs, because that's going to code for a different protein.

And that protein is going to make the different trait. For folks who might not know when you say ATCG sequence, can you clarify? Yeah, yeah, yeah, yeah, good point. The ATCGs are the nucleotides that make up DNA. So DNA is a molecule that's kind of a twisted ladder, and the rungs of the ladder come in four different types, A, T, G, and C. The sequence of those letters in that ladder

tell you what kind of protein to make. Okay? So it's like their rungs on the ladder, they can be in different order, and so you can do different things based on the order that they're in, and each chromosome comes in with the different rungs on the ladder order from each of the parents in this kind of. context. Yes, exactly. And when we say that there are two alleles, what we really mean is there's

two distinct sequences of ATGCs, right, that will code for the protein. And that DNA is stored in the chromosome. So when we see the chromosome, like the two strands, each strand is a giant molecule, a single molecule of DNA that if you were to unravel would be like this big, literally, like that fits inside every single one of ourselves. But it's like, it's like several inches long. That's so horrible. It's a single molecule of just ATGs. species strung together, right? And it's wrapped around a bunch of stuff to create the chromosomes. So that's what a chromosome is. All right, now we know what a chromosome is. Now each chromosome, as we said, has two chromatids. Those are the two different sequences of DNA. And each allele

is located on the different chromatids. So now, when we look at the laws of independent assortment and things like that and how to create gametes, well, during meiosis, the two chromatids, they replicate and then they physically separate into different cells. So the spur-s, cells all get half one chromatid each, and that's now making sense, right? With the whole punnet squares and all that stuff, how's that working? Well, when the egg and the sperm come together, you get half and half to create a hole, right? Right. So now that's making sense. Now the independent assortment works. The whole correlated genes also works, because if you're on the same chromatid, then you're going to get the same package. And then finally, how does the dominant and

recessive work? Right. Well, a lot of times the dominant version just codes for the protein that works. And the recessive version codes for the protein that doesn't work. So if you get two dominance, you just get maybe a little bit more of the protein. There can be some gene regulation stuff to like downplay one of the chromatids so that I don't like have too much of the same protein. There's actually stuff that goes on there, which is very interesting. But if I have like the heterozygous, meaning one functional type and one dysfunctional type, I still have a working copy so I can create that gene. But if I have two dysfunctional types, then I've got no way to create the gene.

Eye color is a very good example of this. The OCA2 allele combinations. So there's a gene for melanosome, melanosome. That's the protein that gives brown eye color. Okay? And if you've got two versions of that protein, then you're going to create melanosome, and you're going to have brown eyes.

If you've got single version of the protein, body can still create melanosome. But if you've got no functional melanosome, but if you've got no functional melanosome, then it's kind of like the naked retina, or not retina, iris, is the color that you're going to see. It's a bit more complicated,

obviously because you have like green and hazel, and so there's different versions of even melanosome and there's other little things that go in. But by and large, this is how dominant versus recessive works. Hey, y'all's Kelly Clarkson with Wayfair. Ever order furniture online and wonder, what if? Like, what if it doesn't hold up? That sofa was four days old.

You should have ordered from Wayfair. With Wayfair, there's no what if. The style you love and quality you can trust. Visit wayfair.ca. Wayfair, every style, every home. And this is why people have to wear sunglasses if they have light colored eyes because they don't have that melanchome. That provides that additional protection for the eyes as one example.

Exactly. So, and so far what we've been covering is effectively it's still been Mendelian. Even like Thomas Hunt Morgan talking about the packages coming together, that's still Mendelian in some sense, right? I mean, you're kind of proving his point. that they come in these packages. It's just that some of them are correlated. There are now a growing, there is now a growing body of work that there are a lot of exceptions. And so that's where we're going to get to next. But before we do that, we've got some housekeeping. Some housekeeping here.

So we are super excited, everybody, to have our resident PhD, Krishna Chowdhury, back in the studio. I'm so excited for our upcoming episodes. We are working through a studio redesign. You might notice that our dock behind us is in a different location. We have some neon signs coming in shortly. And we are adding the FFP bookshelf that will be behind us. And we are going to have a collection of some of our favorite books. We would love to get some suggestions for what you think should go into the FFP Library. So over the next couple of episodes, you'll start to see things begin to take place

and a little bit of changing of scenery. if you enjoy the show, if you are happy to see us back doing these deep dives not only on the frontier of science research, but the context with how we got there. You can always support the show by sharing clips or your favorite episodes with your friends, family, and colleagues. You can watch the show. We always encourage you to watch the show on YouTube and Spotify. The full video podcast is available there. You can listen wherever you get your podcast in. for clips or across our socials, TikTok, Instagram, X, some of the best key moments from the show.

Easy to send that DM to slide into the DMs. Getting more eyes on the show is super helpful for us. If you can support the show directly, we have a donation portal on the website, ffppod.com backslash donate. And if you haven't been to the website, you can watch full episodes. We have chaptering. We link out to all of our research papers that we cover. connected to each episode. And I will note there have been some of you that have been asking for an

RSS feed for the website. We are working on that as well. So keep an eye out for that over the next couple of weeks. So be sure to comment what books you want to see on the bookshelf. I hope we didn't order neon signs that are too big. They're going to be quite large. So hopefully it works out here. But we are super grateful for all of you in FFP Nation and the audience who have been super supportive of the show. We are super excited to get back on the hopper here. We have some great, great episodes lined up. But before we get to those great episodes, we need to finish up because I'm very curious about how we now talk about epigenetics as it relates to this background. We've just talked about on Mendelian genetics. Right. So what we've talked about so far is just the

genetic part of things. And even then, we've only talked about Mendel's rules. And from our understanding now, it's been 160 years of study, it is a gross oversimplification of what is actually happening. Now, the goal here is to predict from genetic input what the offspring's genotype is going to be, what the offspring's genetic makeup is going to be, and then how the offspring is going to look on the outside, right? By look, I mean, like, what are the traits that it has, okay, even if it's on the inside, right? But that's what we call a phenotype compared to the genotype, which is just the actual genetic makeup. Now, as people worked on this over many, many years, anomalies started piling up. Okay, there are several.

We are going to focus on just one, okay? The anomaly is called paramutation. It starts with maze. It was first documented in the 1950s by geneticist Alexander Brink when he was studying the red one locus in maze, location of the red one. gene in maize. Maze is just corn, but biologists always say maize for some reason. Okay, maybe someone can tell me why in the comments, but I've never seen someone just say corn. But then I look at the photos and it's just corn. Okay, anyways. So the red one location in this organism controls for the anthocyanin pigment. Okay. And that's the pigment that is responsible

for the dark coloring of corn kernels. You've seen like those like exotic corn that like it's like purple. That's what this is. Okay. So that's the trait that we're going to look at. In Zim, we have this dish called Sazza, which is cornmeal. And you can have sort of, you know, different color styles. You usually don't use when you have that red pigment style. You'll usually just cook it straight. You won't make that into cornmeal. It'll usually be either yellow or white. But that's, it's a, we call it Mays back home. Yeah. Which is interesting. Yeah, but like, but these are Americans that are working on. That's what I just want to know what, what's going on. Okay. Like, even like, um, um, Dorothy, um, no, it's Barbara McClintock with the jumping jeans.

She discovered those in maze. It's maze. It's maze. Okay, whatever. Anyways. So, let's, let's talk about this red one location, okay? The anthocyanin pigment pathway. Here is what he observed, okay? What I'm going to do is do kind of the same thing that, um, Mendel used to do, which is I'm going to get two true breeding plants, one that has the purple pigment thingy, and then one that I'm going to cross those. Weirdly, I get all recessive. Even though the purple is the pigment, right? So that should be dominant because that's the working pigment.

According to what I told you about what is dominant and recessive. It's not just arbitrary. It's just like one is the protein is actually working. The other one, the protein's not working. Here, if I have a true breeding purple and a true breeding green, I'm actually getting no working pigment. I'm getting all green. It's like the opposite of the F1 generation.

Right? This isn't even getting to F2. This is like in the F1 generation, it's as if Mendel did purple cross white and got all white. That's, okay, that's fast. Which would not make sense given that the- Given that it was all purple. It had the dominance across. It should have the dominance in all four of the offspring. Yeah, okay. So, again, what the hell?

What's going on there? What's going on there? So now, next thing to do is now, I mean, usually what you do is you take that and then you crossbreed with itself, so you do a self-fertilization. But now things are weird. So now let's cross it with a true breeding purple again, right? Because let's say something weird happened and I got like recessive, recessive here. And then the purple.

Or maybe I got like one bit of purple, but like it's not showing up. In the in the F2 generation, I should get some purple. I cross it. I get all green. No purple. It's almost as if the green, whatever it is, has like completely nixed the advantage. of the purple gene and silenced it. This is called a paramutation.

Fascinating. And it's also, it's as if there's the newly altered silent allele that's there, right? Because I know I have the purple allele. But that purple allele has somehow been silenced. And now that purple allele, you would assume if it made it with another purple allele, I would get purple. But this purple allele has turned to the dark side somehow. and started nixing all of its other compatriots

with the purple allele. We'll call it the an anoreal. Yeah, yeah, yeah, yeah, something like that, right? It's like you've got like a weird corrupted computer file that now goes and infects every other computer. And so what this brings up is that there's not this 100% all of the time sort of viewpoint. That's the first thing, yeah. Because this should be dominant.

Yeah, first of all, it's not dominant. But okay, fine. But now it's making everything else recessive somehow. It has an ability. Right. Yeah. Cross generation downstream. Yeah, yeah. It's not just a one-off outlier. Exactly. Yeah. This is very strange. Okay. So it was in 1950s. It was a very big deal when it was discovered. Okay. And so now to understand, now we've got a better understanding of what's going on. So in order to understand this, we've got to understand two things. Okay. First thing is gene regulation.

That is the, that is some of the software on top of the genetics. Okay. So we've got two types of gene regulation pathways. You've got enhancers and promoters. We've talked about this a lot on the podcast, right? When you've got a gene, the gene is a recipe for creating a protein, right? But there is an on button that is telling the machinery to come make the protein. And then there is a volume control button that is saying how much protein to make. That is the promoter and the enhancer respectively. The promoter is the on button.

The enhancer is the volume control. I loved that episode where we talked about this. Yeah. We've talked about this a lot, right? And frequently what happens is the promoter is right next to the gene, because basically what happens is you've got the gene on this part, this is the part that has the recipe. The promoter is right next to it.

So what it's going to do is take something called RNA polymerase, which is the thing that makes it into RNA that then goes into making proteins. The promoter is going to be like, yo, RNA polymerase, come over here, attach to me and then go this way. And it's going to be like, da, da, da, da, and it goes that way. The enhancer can be way far apart from where the, from where the gene is. And the genetic code can literally fold such that the enhancer gets close to the promoter.

The enhancer recruits the RNA polymerase to come. And it's very essential for creating that two-step sort of regulation. Yes. Right. So this is genetic regulation. That's the first thing we got to understand. Second thing we got to understand is epigenetics. Okay.

Epigenetics is the software and the modifying of the structure of DNA and chromatin. Chromotin is the thread that the DNA is, right? There's two levels to doing this that are, by and large, these are the two levels of epigenetics. The first one is something called methylation. Effectively, as I told you, DNA is made up of four distinct nucleotides, AT, G, and C. Now, the Cidazine, can sometimes be modified by attaching a methyl group. A methyl group is, I think, just CH3. You just take a CH3 part of a molecule and you attach it to the cytosine.

When you do that, if you've got a bunch of cytosine and a segment of DNA that has these methyl groups, the RNA polymerase can't really get to it because you've altered the look and the RNA polymerase comes in and it's like, I don't know what I'm looking at. I'm going to go somewhere else. I'm going to go somewhere else. Right? So this is one way to silence parts of the gene, parts of the genetic sequence by just like adding a little chemical tag. And those are the little tiny, like, red pins that you're seeing.

Okay? And they only attach crucially to the C, the cytosine. Okay? So the A, T's, and Gs are always left alone. Okay? The part of what we're sort of speaking to here is there's, in the practice of your body generating the proteins that your genetics sort of has the instructions for, that's a whole functional process. in and of itself. And much like all functional processes,

and there are opportunities for it to break down in multiple different places. And that's where you begin to get the impacts on outcomes because you might have the instructions. But when you go to the factory to build it, the power goes off. The doors are locked. You offshore it to China. Whatever.

Yeah, exactly. Yeah, exactly. And so that's the first part, the methylation. Okay. The second step is in order to actually create the chromatin, which is the stuff that makes chromosomes, it's like a thicker sort of wool compared to stringy. The DNA is like a string,

and then the chromatin is like wool. You have to like sort of bunch up a bunch of string together to create the thicker part, right? And in order to do that, if you bring that photo back, you wrap it around these nucleosomes, which are made up of histones. Okay?

So they're like little proteins that you wrap. it around, and then those sort of get packaged together, right? Now, if you wrap it around, that part of DNA that's wrapped around the nucleosome, the RNA polymerase can't get to. Right. Right. So this is another level of silencing certain parts of the DNA. If you don't ever want this part of DNA to be expressed in a cell, like, let's say

you're an eye cell. I don't want the proteins that neurons want. Package it up. and, you know, put it in a warehouse, put it in your garage effectively, right? Only read the part that tells me how to be an eye. And this is how it's able to control for if all cells all have the same instruction list, how it only focuses on a certain portion of it in order to build out the complex different parts of us where it's not just reading everything and not really having any idea.

We don't have like a bunch of mini like Krishnas, right? It's like, no, there's like Skin Krishna, Hair Krishna and all that, all that other kind of stuff. So that is epigenetics, okay, these two sort of layers. That's effectively what we want to focus on. How does this explain the maze paramutation that we observed? Well, it's a combination of the genetic regulation with the epigenetic silencing. Okay, right, right. Here's what's happening.

There's something called RNA interference. what that is is tiny little bits of RNA that act like mobile silencers. A part of the DNA makes this RNA interference thingy. It's a little piece of RNA that goes and like attaches to DNA elsewhere. And because this thing is small enough, it can just move around the nucleus.

Okay? And what it does is recruit DNA methylation machinery to come in and put those methyl groups in the Cs. Okay? So now what's going to happen? What's going to happen is I've got my purple, I've got my purple trait,

and I've got the green trait. They both actually have the gene for purple. What's different is that the green one has a mechanism to create this small tiny thingy of RNA to go and silence that purple gene. Right. Okay? Right.

The purple one doesn't have that. But now when I cross it, I'm going to get one, one of the chromatins is going to have that thing that makes the little tiny RNA, the RNA interference. But that RNA interference is going to go do it for both. Right. Right? Because it's not, it's not limited by, it is free, is able to free flow through the nucleus and do what it did previously to get the green.

Yeah. And so regardless of the fact that both have this dominant purple, this sort of secondary functional process is free to move despite the genetic code and instruction. Exactly, yeah. And because now the RNA interference is on both the chromatins, now when I take that and I subsequently breed it with another true breeding purple, again, all the offspring are erected. Right.

Because the RNA interference is kind of like this little mini molecular virus type thing, right? Just because you have the instructions doesn't mean your body is going to execute on them. Yeah, yeah. If you've got an agent that's just like, like, erasing it. Right. Or like, you know, putting like a paper on top of the instructions or whatever. That's a really interesting note. Again, going back to this idea that there is sort of a few layers between the instruction and the completed production of the protein.

Yeah. And there are mechanisms that can go and silence component parts in that production process. Yeah. And in the case we just brought up, that was happening, even though it was a purple dominant on the green, that's what was happening. And so if you just looked purely at the genetic code. At the genetic thing, you'd be like, this should be, the green should also be purple. It's got the purple gene on it. And now this epigenetic layer around the silencing on the sea.

Yeah, getting methylated. Getting methylated is what is causing the different phenotypic outcome. Yes, very good. Exactly. And this is non-Mendelian because it's doing all this weird shenanigans. And it's the epigenetics that's doing it. Okay. So this is already an example of non-Mendellian epigenetic inheritance.

But for the longest time, this has been ubiquitously reported in plants, insects like Drosophila, because we're so good at Rosophila genome that like, it's like easy to find, okay? But proving their prevalence in higher order mammals was next to impossible. And we're going to get into why. That's where the technological revolution comes in that enables this paper. Okay? And just a side note, in 2006, there was a paper that showed that mice can have RNA-mediated non-Mandallian inheritance. But there was a lot of dismissal of the findings because they were like, oh, this is just something that happens in strange transgenic mice.

Because the mouse model that was used here was transgenic. And I think it's totally fair. I just want to, this is the funding issue. Oh, there's the funding issue thing. We don't have to get into it. We don't have to get into it. The transgenic mice, yes. Transgenic mice, right?

Please do not cut our funding. We're not making mice trans. But in any case, it was an interesting little finding. I mean, it got published in nature, so it was a big deal. But again, it's in transgenic mice. It's easier to do analysis in transgenic mice because you've got a very nice blueprint of what the genetics is. It's not a naturally like sort of. super diverse genetic population, right?

So you know exactly what to look for. But perhaps it's not so general, right? And if it's an RNA-mediated thing, maybe it's just special, and it's not as general as, like, DNA methylation. This RNA-mediated mechanism was not through DNA methylation. It was through something else. There's a bunch of other stuff that can happen, right?

So the point was they were looking at that this mice population is not necessarily representative of the natural world in a way that we can extrapolate the results from this more generally. And it's like, at the end of the day, I want to see, like, is this something I can apply to humans? Right. Right. So still some scientists suspected that paramut mutation is probably happening in mammals. I mean, it's happening in plants so ubiquitously. It should probably, there shouldn't be, there's no like first principles reason why it shouldn't be happening in mammals, right? But you need to actually get the evidence. And for a long time, it was largely at the level of theory. Now, why is that the case? Why is it so hard to prove? So it has to do with measuring DNA methylation. Okay. Okay. How does one, if I get a strand of DNA, it's actually very easy to sequence DNA.

Okay, there's ever since the days of the 1960s and 70s, I mean, we've sequenced the entire human genome. Now we've sequenced tens of thousands of human genomes. I think it's up to the even might be millions now. So, you know, sequencing DNA is not a big deal because it's distinct. A, T, G, C, right? There's ways that I can make the A, like, combine with a T that has a little, like, you know, a little GFP tag that, like, lights up a certain color. So then, like, every time it's red, I know it's an A. Every time it's a G, I know it's green, it's a G. Right.

And, like, there's so many cool sequencing technologies. But those sequencing technologies aren't good with methylation, okay? One way to do it that a lot of people use is bisulfide sequencing, okay? Here's how you do it. You take your DNA. Okay? You take your DNA. Now your DNA is going to have ATGCs all over.

Some of the Cs, some of the cytosines is going to have that methyl group. Those little red pointies. Yeah, those little red pointies. Here as well, they're little like red tags on the Cs, right? There's some Cs that are like blue. And then there's other Cs you can tell that are not. And those are like normal cytosines that aren't methylated. Right?

What you're going to do is you're going to treat it with a detergent effectively, okay, a chemical treatment that is going to convert unmethalated cytosines into uracils, use. U is the letter that it's used in RNA. Okay. So it's going to leave the cytosines with the methyl alone, but it's going to change the season to use, okay? And then I just sequence like normal, like I used to. Right. Okay? Right.

Because I'm good at sequencing. I'm not good at the, like, finding where the methyl is. But this is one way, right? Because I've just, I've taken all the ones that don't have the methyl groups. I've turned them into another letter. And then now I just sequence like normal. And I'm like, oh, this is a U. The DNA shouldn't have a U.

That's probably where the C was, right? Or I can, like, I can sequence the original gene, and then I can sequence the gene after treating it with this detergent. And wherever the difference is, that's where the Cs are. wherever the C remains, that's where the methyl groups are. Right, and now we know where the methylated cytosine is. Yeah, because any cytosine that survived had to have had a methyl group on it. That makes sense.

Okay. Which now gives us a different map to analyze. That's not just indistinguishable Cs, methylated versus non-methalated. Now, you can imagine treating a molecule like DNA with chemical is not a good idea. Okay? And of course not. Because what ends up happening is something called catastrophic DNA degradation. Here's how sequencing works.

You've got a giant thing of DNA. When I treat it with detergent, it's going to break up into short parts. And I'm going to get a bunch of different short parts. And now I have a combinatoric puzzle to stitch together all of these short parts into a big part, right? Yes. The problem is these short reads that are about 150 base pairs long, max. It's not a lot. It's not a lot.

Okay? Because a lot of times nothing happens. happens. Okay. And so if I've got repetitive sequences of DNA, I could just fool myself into thinking, oh, this is all part of the same sequence rather than this is the first repeat. This is the second repeat. It's kind of like, have you seen that like crazy puzzle of the Beatle white album? Right? Where it's just, it's a, the Beatles white album, and it just says the Beatles with a white. And like, it's incredibly difficult because everything is white. So imagine here, like if I've got repeated sections of DNA, I don't know whether this thing came from this part or this part or this part or this part, right?

100%. And some puzzle pieces go together when they're not supposed to be next to each other. Yeah, yeah. Yeah, because just, you know. Just kind of the couple. Yeah, it's like this side of the puzzle piece goes together. But when you put it together with the rest, the other three parts don't make any sense. Yeah, like it's a really incredibly hard puzzle.

Right. So that's one reason why you don't want that. The next is loss of allelic phasing, right? If I want to study non-Mandallian inheritance, what I really like to know is which parent the epigenetic mark came from. That's why the maze thing was such a big deal, right? Because I could tell what was happening. But in order to do that,

what that requires is each of these short reads to have a mutation from one parent and not from the other. It's like, oh, this part is a T. I know that the mom had the T here and not the dad had a C here. So I know that this part comes from the mom. But if I have short enough reads

that are 150, the probability of getting one of these single nucleotide permutations in my short read is very low. Right? Mm-hmm. Because the shorter the thing is the probability of something going wrong is not that high. Right. So the methylation could be different between two identical, but I don't know which one came from the mom, which one came from the dad. That's actually a really good point.

And that's fundamentally... Yeah. And fundamentally, the whole point of what you're trying to do. Right. Right. You need to be able to make that connection. Yeah, I'd like to know which paternal chromosome it came from from the entire whole gene. from the entire whole genome.

It's like you have two white puzzles. One is the Beatles' white album. One is whatever, the white stripes white album. Yeah. And it's like the white piece is good. Yeah. It's like they're all white.

Yeah. But is this the white outfit? Yeah. That's exactly right. That's a good analogy. Yeah. Okay. Okay.

So that's been the problem. I don't know how to do it. Yeah. It's like we've been relying on the sequencing technology that gives us the sequence of DNA, but it's not designed for methylation reads. Right, which is key for us to be able to identify, you know, where we're going to see these mutations that cause the dominant to get suppressed. Yeah, yeah. And all of this weird shenanigans. All the weird stuff.

All the weird stuff. It kind of derives, at least in part, from this. From understanding which chromosome is the parent, right? Which chromatid came from which parent. Okay. So now we get to why it's happening now, the Oxford Nanopore Revolution. Okay. This is an insanely cool technology.

Here's how it works. Okay? It goes to biophysical detection. What you do is you take your native DNA strand and you pass it through a nanopore. It's a little hole in a membrane, kind of like ion channels. Like there's a membrane that separates in and out, and then you've got a little hole that's made out of protein that one strand of DNA goes through. There's a thing above it that unzips the DNA so that one strand goes this way, one strand goes through. Okay?

And what you're going to do is measure the current across the membrane, okay? The current that's going through from this side to this side. Now, A, T, G, and C are different molecules. They have different sizes. They have different little residues. They have different shapes. And so every time the four goes through. there's going to be a tiny different current for each of the four.

Are you one of those media strategy people clicking through slides, scrolling spreadsheets? Yes? Good. This is for you. Because on Spotify, there's an audience that's different. Locked in. Loyal, invested. They're called fans. Fans don't just listen to music. They feel seen by it like it belongs to them. So when your brand shows up on Spotify, that's who you're talking to. And you're right next to artists like me, Lizzo. So, are you ready to talk to fans?

Spotify advertising. You're among fans. Isn't that sick? I'm so... We're so smart, dude. Isn't that so sick? That's very clever. The visual analogy I sort of think about is like,

if you unzip the zipper of your jacket, it's like the unzipping of the DNA. Yeah. Each of these are then going into something that's going to read each segment of four, which is basically just going to have a different fingerprint. Exactly. Like, imagine your zipper. Let's go with that analogy.

I like that. Imagine your zipper, but each of those tiny little zip thing, is a different composition of plastic. Okay? And so when I go through this little hole, it's going to sound different, right? The red plastic is going to sound different from the white plastic.

And so if I've got a little sound recorder that just record the tur-d-d-d-d-d-d-d-d-d-d-d-d-right, I can then have a computer program, decode what type of plastic was the four zipper parts. And that's what's happening here. And so on the right-hand side, you see the current versus the time Yeah, yeah, yeah, yeah. Completely different signatures.

And then you've got like, you know, it's probably not even machine learning. This is a very simple, like, computational task to do. This is really good. Okay. Now, this was originally, this was originally invented for just genome sequencing, because A, T, G's and C's, right? But now a C with a methyl group is going to sound even more different. Bingo.

You see? Bingo. Yeah. And so it's just going to be a fifth signature that I have. to look for. Yeah. That it makes, I mean,

it makes total sense. It makes total sense. Why is this advantageous? One, you can do ultra long reads. You're not treating the thing with detergent. Right.

Right. You just, you just, you just, you just, you just stick this thing through and you can get hundreds of thousands

of base pairs. We're going from 150 to now several orders of magnitude of like how long you can make the read, right? So it's like, it's like,

you know when, when sometimes you, you like go to the Airbnb and like you open the puzzle box and half of it's done. they're like, yo. Right?

Because then you can post it on Instagram like two hours later. I did it. It's a really interesting note though that through this biophysical detection method, which again was used for sequencing, the point is because we understood the methylation process in and of itself,

you could then extrapolate that the C in this biophysical detection is going to just have Again, that slightly different fingerprint. And that's all you need to look for now. And again, just this narrow, now we can, and again, now being able to sequence longer. Yeah, it's huge. Gets back to this chromatid.

Yeah, because now you've got a longer sequence. In 100,000 base pairs, you're going to be able to identify which one came from the mom and which one came from the dad. 100%. Wow. Isn't that sick? And the thing is, this Oxford Nanopore evolution was not necessarily for methylation. No. No.

Detection. No. It's just someone was like, wait a minute. Yeah. The physics works out here, right? It's called Oxford Nanopore detection because it was a technology licensed by Oxford Nanopore Technologies of Oxford, UK. They licensed the patent, but the patent actually is a University of California patent. Okay.

It was developed in UC Davis and UC Santa Cruz. The fundamental research comes from David Deemer, who he started at UC Davis and then Santa Cruz poached. him. He's like, yeah, I'll go, I'll go live in Santa Cruz. Santa Cruz is kind of nice. Yeah. It's kind of nice. And I really do think that this is, um, this is something that is Nobel Prizeworthy because it's, it is now the foremost way of not just doing methylation, but also just sequencing DNA, right? In general. In general, like this is now, it's super fast because you go, right? And then, and then have a computer just read the whole thing.

Um, and, and these ultra long reads are just good regardless. Right. Right. For any number of context. We have a specific focus for this episode, but the applications are endless. And Oxford Nanopore Technologies, they created something called a mini on sequencing device. It's like a tabletop sequencing device. So you can just like have it. It's not like the giant machines that you used to be. It's just a tabletop. Stick your pipette in and then it'll just do the thingy. Is it analogous maybe to the sort of the phase shift from vacuum tubes to like desktop personal computer or desktop to mobile in the Yeah, and I would say like, like, yeah, desktop to like laptop. Okay, desktop to laptop.

You know what I mean? Because laptop I can now like go anywhere. I don't have to be plugged into a wall somewhere. Yeah, exactly, and things like that. Okay, okay. And so I think it's a very big deal. And they actually took one to the international space station because they do like weird genetics experiments up there

with like plants and like how they deal with microgravity. Makes sense. It makes sense, right? Like how does bacteria deal with microgravity? Because ultimately, if you want to extend life to several years in space, you need to understand genetic. genetic how like does the genetics at the molecular biology level do they care probably they shouldn't because the gravitational potential difference between like one end of the cell and the other shouldn't

be that much but you know you never know with biology right right you got a yeah like the fact we're gonna shut down yeah like just because i took out like a tiny bit of wind effectively right right that's like going in one direction so the i ss is using a mini ion sequencing device on it that's really cool yeah so now Now finally, with this, we can get to the paper at hand. An hour in, an hour into our episode, this is the paper that we were talking about, okay? Non-Mendelian inheritance of DNA methylation patterns in mice. It was out in nature genetics.

It was during my paternity break, and I was like, oh, this is really cool. But now I can read that title and understand why this is a big deal. That's right. You miss me? Yeah, like, this is, ladies and gentlemen, the greatest. This is why it's the greatest. But like and it's it's important because these words have meaning. Yeah.

A lot of times people see these titles and they're just like, ah, it's just big words. Yeah. I don't mean anything. There's actually real even in the short title. Even in the short title, you get it. And it's in mice, not transgenic mice, not like special mice. Just mice. Just plain old mice.

That's why it's very important. He's doing it in mammals, right? Because this is from the 2006 one. This is basically the next stage, the next development there. Yeah. The one of the key author is Andrew Feinberg. He's from Johns Hopkins. He has a career in disease.

genetics and epigenomics. He's like one of the epigenomics guys at the conferences, right? Like that's like giving the presidential lecture or whatever. And then David Threadgill at Texas A&M, he's been studying human-like complex genetics in mice. So they decided that they wanted to collaborate about 10 years ago trying to study epigenomics in the mouse model. And they're trying to compare methylation patterns in response to diet in these inbred strains. And they realize that in order to do that, what they really want to observe is allel-specific epigenomics, meaning this allele came from the mom, this allel came from the dad, how is the methylation changing, all of that kind of stuff. Which this methodology we just talked about before is an enabling layer

to that specific research question. Exactly, yeah. And so now we can start doing that. The model that they chose was something called the collaborative cross-model. This is key here, because what this collaborative cross mouse model is, is it's derived from eight distinct founder strains. So the founder strains are kind of like your P generation in the Mendels, like the true breeding. But because now you're combining these eight different models, you get a lot of genetic diversity. And so it's much more closer, it's much closer to like, you know, normal organisms. But they're still lab in the sense that I still have very nice sequencing data about them. So I can still do very nice, nitty-gritty comparisons.

Okay? So that's the strains they use. The specifically one, the strains that they use was the CC-019 and the CC-037. Those are the two specific like sort of F-2 or, in this case, like FN generations, right? There's so many that you can make. And the experimental design is actually quite simple. It's kind of like Mendelian experiments. Okay.

What you do is you take your, you take your CC-019 and your C-C-019 and your C. CC 037, those lines, very well established what the epigenetic markers are there and what the genetics are there. So even though they're not like true breeding in a sense, you have a good understanding of like what the genetics are. Now you cross them and then you cross those again, right? Just like in Mendel's experiments. But in this case, what we're looking at is the methylation pattern on the chromatin instead of the actual genetic sequence. Okay? So the genetic sequences could be the same, but depending on whether this chromatin came from the mom, this one came from the dad, all of these things are now going to start changing. Okay? This is very, very, and it makes sense now

why it would be changing. One, because we understand the sex chromosomes, so you get one from each. Two, because the methylation is going to change what proteins are actually produced or what's suppressed. And so now, with all of that understanding and background and context, and the ability to actually detect and sequence the methylated C. Nice. Like that, that is, they can now do real statistical analysis. Exactly. That's meaningful in this, because they've done, they have the correct,

they've done the mice work. They've done all the other background work. Anyway, so that I, it's very, very cool. It's all kind of just. And one final thing that I want to mention before we get into the results. Yes. Is again, about methods because we do want to focus on methods in this. So we've got the.

you've got the mouse models of CC19 and CC37. Now, when you sequence the offspring's DNA, right? Usually when you want to put together, even when you've got these long reads, you've got to have a reference genome to put stuff together to. But we don't really have a reference genome for the offspring of these two lines, right? And it's not great to just use one or the other because you really, you're going to have to have like some middle ground. And so one thing that I do want to talk about is like in traditional bioinformatics pipelines, right, you've got these.

And I think photo 30 will show that in a traditional bioinformatics pipeline, right? You've got fragmented DNA. Even if they're long, you sequence the DNA, then you get all these unaligned sequences, and then you try to align those sequences to a reference genome. Okay. And the reference genome is something that you get, like if you've got a human genetics, right, you've got the human genome project. And you align your own DNA to that because there's not going to be that many. differences, right? But here, those tiny differences are going to matter in this study, right? And so what they did was they actually solved it by constructing a pseudo-hybrid intermediate genome, which is, I think, very cool that they were

careful enough, you know, to create this kind of custom software that constructs a synthetic reference genome for every single cross, right? And it includes all of the strain-specific insertions of the parent genomes, but it's like unbiased analysis of what the, what the kids are going to be showing in terms of methylation. And this, this is important because, again, that reference genome is otherwise you're kind of, you have no, you have literally no reference. Yeah, exactly. And then it's impossible, but you can't use one or the other.

You've got to use this like hybrid. And so they were careful enough to do that. I thought that was quite cool. That's an important. Again, shows the, they're really trying to do it in a robust. Yeah, very meticulously. fashion, especially also from a replication perspective. Exactly, yeah.

So now we can finally get to the results figures, okay? So let's talk about the control, okay? So Mendelian baseline, 93% of the time, the epigenetics is doing what Mendel's laws say. Here's how we're going to read these figures, okay? All of the subsequent figures are going to be the same way. So the pink is coming from the mom. The green is coming from the dad.

What we're looking at on the y-axis is the percent methylation, the amount of methyl-methylation, meaning how much of the Cs have a methyl attached to it. 100% would mean every single C has a methyl group attached. Zero means none of the Cs have a methyl attached, right? And what we're looking at on the X-axis is just along the DNA. Okay? So along the DNA, you can see that there's a certain region where the mom has almost no methyl groups attached to the seas,

and the dad has a bunch. Okay? The solid green lines and the solid purple lines are the parent generation, and then there's dotted green lines and dotted purple lines for the kids. The key thing to see here is that the dotted green is on top of the solid green, and the dotted purple is on top of the solid purple, Meaning the methyl pattern of the kids is the same as the moms or the dad. Nothing's changed.

Right. Like I got the chromatin. I got the methyl. Nothing weird happened in between. So this is the Mendelian baseline. Yes. Okay. And they actually use two different tissues from the liver and from the muscle to see how like

if there's any organ dependent effects that are happening. That's actually good. Yeah. Yeah. So this is like normal. Okay. This is something that we thought should happen every single time. Yes.

All right. Now let's look at our first anomaly. And because what we're saying is 93% of the time Mendelian genetics holds up. Yeah. But there's the 7% outlier where it doesn't. And that's weird. Yeah. And that's where the weirdness is.

Okay. Now let's look at non-dominant transacting methylated. What is it? Something, QTLs. I forget what the full form. is, but the ME is the methylated part. Yeah, yeah, yeah. Okay? Here, what we're looking at is, let's look at the liver, which is the first figure,

the first sort of middle figure, okay, at the top. Over there, what you're seeing is the top line is the pink solid. The bottom line is the green solid. So that's the mom methylation and the dad methylation. You can see the mom has a lot of methylation in this region. The dad has almost no methylation in this region. Which is an invert of what we just like that. Yeah, yeah.

But that's not what's important because this part of the gene might be different. What's important is the dotted lines are all in the middle. They're not coinciding to one or the other. Meaning that I got the DNA from them, but then the methyl parts are half and half. They're blended. They're blended. That's fascinating. That's non-Mendellian.

That's what we're looking at, okay? The offspring, some of the methylation is there. Some of the methylation is not. That's interesting. Some of the female chromosome is getting more methylion. Or sorry, no, some of the female chromosome is getting less methylated and some of the male chromosome is getting more methylated. Yeah, because based on where they are, they're in, they're in, again, the male is the one that's almost has none, but the, the, the, the offspring are blended and so. Yeah, that's why the dotted lines are all in the middle.

In the middle. In the middle, so you can be higher and the opposite of her. Very interesting. Okay. Okay. Now let's go to the next one. Yes. This one's weird. This is the cap N11 gene. What we're looking at here is is effectively paramutation. Okay. Okay. So in this one, in this particular region, the mom has, let's say, intermediate methylation,

the solid pink line is like 50%. The dad is all the way down to like 10%. Okay, those are the solid green line that's way down there in that highlighted region. And the mom is down, there's a dip, but it's only 50%. The dotted, pink line and the dotted green line are all on top of the green.

None of the mom's effects of methylation have come through in the subsequent offspring. Okay, so that's that top part. Now let's look at the F2 generation, okay, which is in the bottom. Every single offspring has the methylation pattern of the dad. It's like the mom's methylate, like obviously the genes are there. Right. But the mom's methylation pattern is completely gone. The mom's epigenetics have been completely signed.

In both F1 and F2? Yes. And in F2, subsequently, yes. F2 is there to show that like whatever happened to F1, it didn't recover. Right. Right. Whatever happened in that first generation, the grandkids now just get no influence from the mom's epigenetics. All of the epigenetic influence is coming from the dad.

That's, that's, in this particular allele. Right, right, right. Right. This is very weird. This is very weird. Because what is... This is something that we saw in the maze, right? Right.

Where it's like, even when I crossed the purple with the green, it was only green. Even the grandkids all were all green. This was the suppression thing we talked about there. This is now the the allele is completely suppressed. And so now we're seeing this in animals as well. We'd seen it in the maze before, but now we're seeing it in more complex biology, which is fascinating. And there are some very cool like consequences of this that we'll talk about at the very end, okay? The last few figures that I want to talk about

This next one, figure number 34 This is, I think, figure six and seven This shows some emergent epigenetic patterns Again, the main thing to see is that The dotted pink and green lines Are different from the solid ones. The one on the left, figure six, That shows something very weird,

which is that the offspring sometimes just have more methylation than either of the parents. So that, I don't even know how that got there. And neither of your parents It's not even a blend Or one of the parent is dominant over the other Which would it make almost more sense? Yeah, yeah

This one's just like This doesn't Yeah, that's why they're calling it Emergent Epigenetic You know, it's just like Put it in that bin Yeah, we'll come back to that way Yeah, yeah

It's like this happens And we report it So now you got to size And then on the right hand side We've got genomic imprinting This is, again, this is similar to the mom and dad thing. But here, so in the mom and dad one, it didn't actually matter which chromosome,

which, if the allele came from the mom or if the allele came from the dad. Because both sets of alleles, if they were from the mom or the dad, it was allele specific. Here, it's actually mom and dad specific. It's a process where certain genes are expressed in a parent of origin specific manner. In this case, the epigenetics is expressed in a parent of origin matter. Okay? So if the methylation is with the mom, only then is there going to be methylation in the kid. Okay? Doesn't matter what type of allele it is. Which is actually an important distinction. Yeah. So it's a parent of origin. So there can both be allele-specific divergence in methylation. in, you know, the FN generations.

And that can be true on top of the fact that there's this genomic imprint. Yeah, yeah. Like, they're not necessarily independently true. Yeah. So it's super, super, super complex. Yeah, because the combinatorics becomes. Right? The common, yeah, exactly. It becomes quite difficult.

Yeah. And the last one I thought was actually really, really cool. I mean, I think all of this stuff is really, really cool. But here what we're seeing is a difference in one organ-dependent methylation. So the same region in the liver versus the muscle. The muscle has about 80% methylation in both males and females. And in the liver, the males have 80% methylation, but the females have something like 50% methylation. This is crazy.

Highly significant difference. This is crazy. Okay. So you can also have divergence across organs. And on top of sex. Now, why is this important? This was found, so it's hyper-methalated in females and then not that much in males, right? Why is this important?

Well, it's found exclusively in the liver. What does the liver do? One of the big things that the liver does is dictate tissue-specific, like it dictates hormonal environments. It dictates how drugs are detoxified. So this could be one of the main reasons. why certain drugs work one way in one sex and then not in another sex. Yeah, yeah. Right. Right. Because that's a known effect. Right. We know we have. That's kind of mysterious. Right. Right. This could be it. Like the epigenetics is different in the liver,

but not in other other places. That's a really important potential insight among many in this is that there is this organ-specific epigenetic difference in methylation in the liver, which would affect drug delivery. Yeah, and drug like efficacy. Efficacy, right, right. Kind of crazy. I'm thinking about so many things, but I want to be not,

we will talk for the next three hours. Obviously, this has very deep implications. That's, that's the main headline. And what it's really saying is, look, future clinical genetics is now going to need to focus on genetics and epigenetic risk. Okay, if we want to start charting, you know, like how an individual's genome is going to affect their life and so on and so forth, right? There's also some very cool,

just like fundamental science things that are happening here that have to do with, like, for example, evolution, right? The foundational science of evolution. There used to be this theory called the Waddington's theory of genetic assimilation. Basically, it's saying that environmentally induced traits can become hard-coded into the DNA, right? This is not very Darwinian. Darwinian is just natural selection, survival of the fittest. It's like the offspring get a random assortment, and then they live their lives, and then if they survive, they procreate, right?

What Wattington's theory is saying, it's more Lamarckian in the sense that your life is going to dictate your genetic makeup. Now, for a long time, this was discredited, and for good reason, because genetics is, like, as I said, the DNA part is like very hard-coded, right? But now there's a link between the software and perhaps the hardware. Yeah. Because methylated cytosines,

which are the Cs with the methyl groups, those are inherently chemically unstable. Okay? So what that means is they frequently mutate into thymine t's. So if I have more methylation, I'm going to have more mutations, and I'm going to have more T. where there should be Cs at three times a higher rate.

So what's the conclusion? That means that if among the inherited epigenetic traits, right, if my life is changing the methylation pattern, that methylation pattern could change the hardware itself because of this 3x difference in mutation rate. It's a recursive self-improvement or not improvement. Or not improve. Yeah, it could go either way.

It could go either way. Kind of crazy. It's always been the conversation about nature versus nurture. It's more like nature and nurture. And there's this combination between the two. And as this sort of continues to permeate as now posing some very difficult questions, there is, and obviously the sort of epigenomics and epigenetics conceptually have now been around for some time. These are the kind of big results that I think put us into a new frame, you know, a new, what do they call it?

The Overton window on this is moving a little bit to now really think about, okay, because I think your point is right. Over the course of your life, if your life itself is impacting the methylation on these cytosines, that then become go from C. to T and your offspring fundamentally get different genetics as a result of that, that's pretty meaningful. Yeah. Yeah. That's pretty huge, right? I mean, even in terms of disease risk, right?

This is a big deal, right? Right. Because you can explain now how a mutated allele could epigenetically silence a healthy allele. Yeah. Right? So it's not just all about genetics, even if you're a heterozygo and you've got the correct allele to, like, survive the disease. If you've got some shenanigans going on like this, and the healthy allele just doesn't even get expressed, you're biologically null for that gene, right?

This makes me think about things like, um, um, like anemia being more prominent in, like, Mediterranean or, uh, African populations as opposed to, and it's like, is there? Well, that one, that one is actually a single nucleotide polymorphism. It is. Yeah, that's actually, there's like a straight up hardwired genetic difference. That's not an epigenetic thing. That's very much. Yeah, and that one's well documented because the mutation is, I believe, in the, in the heme protein or something like that, that, like, affects the shape of red blood cells. Yeah.

Where you cannot store the hemoglobin as well. Yeah. Hemoglobin is, it's like in a weird shape. Some of them are a weird shape. So then that makes the cell a weird shape. So, but there's so many. other things, right? And this is, like, there are traits, I forget which, but like, there's some traits where, like, there's, like, there's, like, weak evidence that shows that the grandmother's

exposure to stuff alters the disease risk of grandchildren. Okay? And maybe that's another episode that we could do. But, like, this is, this is now getting to that question, right? And we can now ask, we can actually probe at the, probe at it, um, a much more, um, in a much more robust fashion. Yeah. Not just create a framework or have things that we can't quite then follow up with empirical data.

Yeah. And I mean, the next step, the next clear step, and this is something that Feinberg and his colleagues are doing next, is to do this long read sequencing stuff for humans. Yeah.

Right? And seek paramutation in humans. Which makes total sense. So, because if it's like 7% in mice, it could even be higher in humans because humans are a relatively recent evolutionary, organism, right? And our mutation rate is hella high. And that mutation rate, we're really talking at

the hardware level. If the hardware level is already high, imagine what the mutation rate at the software level is, you know? This is fascinating. A big, big return episode for our resident PhD, Krishna Chowdhury, new rules for this idea of non-Mendellian inheritance of epigenetics. I mean, there's so much here that we again we can't cover everything in these pods there's a lot of detail that we can't quite dive as much into but hopefully we gave a nice overview for folks who are interested in this conversation and again like I mentioned there's so many you know social cultural conversations around this idea of quote unquote generational trauma and its impacts on your your actual physical outcome and

we're starting to be able to actually poke at some of these areas in interesting ways. Do we want to leave, because this is a long one. So if you've made it this far, congratulations. You are a lifelong learner like the rest of us. Do we want to have any questions for folks to leave in the comments for this episode? Oh, yeah. Well, I was going to say, what you guys should do is tell us what books to put on the, as you said, in the thing. Yes.

But if you made it this far, and do you have any advice for a new father? That's a good one. Even if you're single and you don't have kids, just give me some really terrible advice. I want to see the worst advice you can think of. My name is Lester Nare, joined as always by my co-host and our resident PhD. Welcome back to Zadhi, Krishna Chowdery. It's great to have you back. This is always so fun.

It's good to be back. We got some fun ones coming up for you guys here in the next couple of weeks as we get back on our regular scheduling cycle. We appreciate you all for staying with us and we will see you all next week.