Hi and welcome back. This is office hours for unit three of behavioral genetics. I'm your TA, Bridget Carey, and we have our professor, Matt McGraw here. So just a quick note before we get started. If you have any questions that you want us to address in office hours even if they're from past units, please put them in the most current unit's forum because those are the ones that we're going to read and address each week. So even if they're past topics, go ahead and put them in the current week. But we're going to start off with some questions about variants of components in the Falconer's model. [SOUND] Starting with a question from Edward Chen. And he wants to know so offspring inherit both additive and non-additive genes from their parents. Why do only additive genes play a part in the parent offspring resemblance? And in Galtonian inheritance, additive genetic effects outweigh the non-additive genetic effects, why is that? >> Okay. So there's two questions there. >> Mm-hm. >> So hopefully I'll remember both. But you'll catch me if I, I forget the second one. So this aspect of quantitative genetics is, is, is complex, it, it's a little bit subtle. The difference between additive genetic effects and non-additive genetic effects. Additive genetic effects are the effects of single alleles. >> Mm-hm. >> Added up over the multiple alleles and the multiple loci, in a quantitative system. Non-additive genetic effects really refer to the unique effect that is attributed to a particular combination of alleles that an individual inherits. The reason that parents and offspring only share additive effects is that if you look at a particular locus, any parent will only share one allele at that locus with his or her child. So parents really only share these single allele effects. They can't really share the particular combination of alleles. The unique effects, the non-additive effects that are due to the combination of alleles that you inherit together at a given locus. So non additive refers to combinations of alleles, additive refers to single allele effects. And it, it's parents, parents because they only transmit one allele at each locus or each low side to a child are sharing only those additive effects. One, one other thing about that, it's a little easy, I think, probably it's easiest to understand this distinction between additive and non additive when we're talking about the effects of a alleles at one locus. So the additive effect is just the effect of allele taken by itself regardless of what other allele it's pairing with at that locus. The non additive corresponds to the particular combination of alleles you've inherited at that locus. Parents are only going to give one allele so they can't share that unique effect. It gets a little bit more complicated and, and we don't need to go into it, as to these non-additive effects across loci. That's called epistasis rather than dominance. The same thing applies the mathematics are a little bit more complicated, but, but basically, the same principle applies. So, that's the first question. >> Yes. >> So what was the, I can't remember the second question now. >> He wants to know, why in Galtonian inheritance do additive genetic affects outweigh the non-additive, genetic affects. >> Okay, okay. Okay so that's a great question and it's actually a question that is currently debated in genetic epidemiology. And the reason that it's debated is, and we're going to touch on this a little bit later. I think when we get into Unit five and Unit six and we talk about genome-wide association studies. >> Mm-hm. >> And one of the things that will come out of that, and this is jumping a little bit ahead, but one of the things that will come out of those studies is that genome-wide association studies are unable to account for the heritability, the additive heritability of many diseases. And one possibility for the reason why they're not accounting for that heritability is that, well, they're only looking at the additive effects and they're not looking, maybe non-additive effects are there. So there's a lot of interest in this particular question now within genetic epidemiology. It doesn't have to necessarily be the case that additive genetic effects are dominant. Empirically, that appears to be true, that for most quantitative genetic systems, additive genetic effects dominate. The, maybe dominate is a bad term to use but they tend to be much larger than non-additive genetic effects. There are certainly some that are non-additive genetic effects are there and they're prominent, but as geneticists have looked quantitatively, at various phenotypes, it appears to be the case that the additive gene effects just empirically tend to be the, the most prominent ones. >> So another question from Edwin, or Edwin Chen. For DZ twins, the shared additive genetic variants is 50%, but why is their non-additive genetic variance 25% instead of 50%? >> Okay. So, yeah. So the reason that I'll, I'll get formal here and hopefully it makes sense. So the reason that it's 50% for the additive genetic effect is really the 50% is the probability, again, additive genetic effects are the genetic effects of alleles taking in isolation of the alleles their paired with, but added up over multiple low-side, multiple alleles. So the reason it's 50% is what is the it, it, it 50% is actually the probability the two dizygotic twins, or two full siblings, or actually, a parent and offspring. The probability that they will share the same allele at a particular locus is 50%. So that's why the additive genetic effect is 50%. Because the probability they share single alleles is 50%. The reason that it, it's 25% for non-additive is the way I represented non-additive in the lecture was really as dominance, is this genetic dominance. And dominance corresponds to the unique effect owing to the combination of alleles you inherit together at a particular locus. The probability that dizygotic twins or full siblings share both alleles at a locus, the same, the same copies inherited from their parents turns out to be 25%. So the 25% is really one half they share the first allele times one half, they share the second allele. So 25% is for dominance. I didn't, I didn't, I really didn't get into epistasis, because it's just beyond what we need to do in an introductory course like this. For epistasis, it wouldn't necessarily be 25%. It would be some number less than 50%. It could be some number much less than 25%. It gets fairly complicated so it's not really worth going through it. >> So just for this, it was predominantly dominance? >> That's right, and, and, and the important point here is the conceptual point. I think, is that these effects in the case of the dominant effects here are really only shared by dizygotic twins and siblings because they're the only ones who can share both alleles and locus, inherit both the same alleles and a locus. >> And you actually kind of answered this question in the last one- >> Okay. >> But Sonya Durward asked, why do we assume 50% that there's a 50% genetic overlap for dizygotic twins? Couldn't it be more or less than that? So- >> Okay. >> You kind of just addressed that a minute ago, but-. >> It's, yeah, it's, so the fifth, one of the things as you know, one of the things that we say that on average, dizygotic twins share 50% of their additive genetic effects. And it is an average. It could vary, actually. Some dizygotic twins might share more than 50%, and some might share a little bit less than 50%. But on average it's 50%. The, the range at which they, they share is, doesn't actually, if it's over many loci, doesn't actually vary much from 50%. It might be as low as 45% and as high as 55%. But it can vary a little bit. There can be some dizygotic twins are genetically more similar than average, of course. And some that are less similar. And in fact, there's, I think, as you know there's, there's some ways of estimating heritability that just look at dizygotic twins or biological siblings. And ask the question, well, are the ones that are more genetically similar also more phenotypically similar- >> Mm-hm. >> Than the ones that are less genetically similar? >> And another question about, concerning Falconer's model. A question comes in asking what is assortative mating for traits, and why Falc, why does Falconer's model not allow it? >> Okay so assortative mating is just that literally that like mates with like. That spouses are similar on certain behavioral traits. And they are similar on some traits. They're, they're similar in height, more relevant to us in this course in terms of behaviors. Spouses correlate in their IQs. It's not a real high correlation but it's about 0.3. They have similar levels of education. They're not very similar in personality, spouses, for the most part. One of the strongest things that you see as sort assortative mating for interestingly enough. It's political ideology. Conservatives tend to mate with other conservatives. Liberals with liberals. So that's what assortative mating is. It, it actually, assortative mating varies quite a bit from one psychological trait to another. Being high for things like political ideology. Intermediate for things like ability and IQ and low for things like personality characteristics. How extroverted you are, or how neurotic you are. What assortative mating does is it creates if these, if, if, if mates are if a man and a woman who mate are similar on psychological traits that are heritable, that means that they're also genetically similar, on the underlying genetic factors. So, they're not two genetically random individuals. That has implications for the quantitative genetic models. And the major implication of that, without going through the mathematics, and I'm sure, I'm, I'm pretty sure she's not asking for the mathematics or. Is that if, if the parents are genetically similar, it raises the genetic overlap between DZ twins above that average of 50%. So it might go up to 55, and even, in some cases, for things that are, where there's very strong assortative mating it might go as high as 60%. If that's the case, then when you're taking the Falconer estimate of heritability and you're taking the difference between the MZ and DZ twin and doubling it. If DZ twins are not 50% genetically similar for additive genetic effects, but 60% because of assortative mating, if you go through the mathematics, you'll see that you will underestimate the heritability in that case. So, Falconer is assuming that DZ twins are 50% genetically similar that would only really be true if there's no genetic similarity for that trait in their parents. If there is some genetic similarity, that'll increase that above 50%, lead to an underestimation of heritability. >> Thank you. I think that was a great explanation. [SOUND] You presented a study on reading abilities using an Australian, Swedish, and US sample. I'm looking at their abilities in kindergarten and first grade, and there are a few questions on that. >> Mm-hm. >> so, Dinesh Banargi asks, if the demographics of different countries used in this heritability of reading study. Might explain the differences inheritability estimates. So for example, if the US is more ethnically or racially diverse than Sweden, should we observe a higher, heritability? >> Okay. I, I that study. Actually, were you involved with it at all? because you were at Boulder. >> I was at Boulder but I was involved. I know [CROSSTALK] You were involved? >> I was not involved. >> You were not involved. Ok. You knew the people >> I, yes. >> who were doing the study. I think. Okay, so she wa- she or he was not - I forget the name of the person here, but, they weren't asking about this. But I can't help but [LAUGH] say again why I like this study. I think, Yuri, it can be taken as an illustration of the limitation of heritability. >> Mm-hm. >> That heritability is not a constant. But I think it actually shows the strength. The utility of heritability. The finding that the heritability changes as you go from kindergarten to first grade or as you compare cross countries, really shows that the heritability is telling us something's going, interesting going on here. >> Mm-hm. >> So I think, although we might use it as an example. Well, heritability isn't some fixed biological constant. In fact, because it's not a fixed biological constant, I think it actually makes it a much more interesting statistic. And I, that's why I really like this study. I think it illustrates that. Okay, but I didn't answer the question. [LAUGH]. >> It's a, it's a good point. I don't think it's really affecting the heritability estimates here for a couple of reason. First of all, right, in the, the heritability of reading achievement is moderately to hi, to high in Australia and the U.S. in kindergarteners. Low in Swedish children. When they come back in first grade, the heritability in comparable across all countries. >> Mm-hm. >> So it, it, saying that Sweden, Swedish children are genetically more homogenous can't really explain that pattern because the heritability right recovers in a sense, in, at first grade. So that's not the explanation of the results but it's still a very good question. Could, do people in our business, do we worry about that possibility, that if we're comparing heritability estimates from Sweden, to the US, might we expect differences because Sweden is genetically more homogenous than the US. I don't know to what extent Sweden is more genetically homogenous than the US. It probably is somewhat more homogenous but it probably is not enough to really make a difference. We won't really talk about in this course I didn't think we really talk much about race or really what, what geneticists prefer, as you know the term ancestry, or ancestral differences. but, but what we do know from population genomic data is that most genetic variation is within populations, not between populations. Or another way of saying that is that the genetic variation that exists in Sweden probably accounts for most of the genetic variation in the world, maybe 85% of that so that we're not really restricting our variation that much when we take a particular population as large as Sweden. If we had very, very narrow groups, and then maybe it becomes a problem, but when we're talking about something as broad as Sweden, it's probably not an issue. >> [SOUND] And also in the same thread, an anonymous poster asked why isn't shared environment considered high for all three groups of first graders since each group attended schools where reading was part of the curriculum? >> Okay, okay, good. That's a good question, too. So so, the shared environment corresponds to the effects of environmental factors individuals growing up in the same home share that would lead to differences in their behavior. The important thing, I think, here in answering the question to recognize is what shared environmental effects correspond to are the consequences of environments that differentiate one family from the next and, therefore, might lead to differences in the children in those families. That's what the shared environment is. In this case the reason why, in first grade where the everybody is getting the same curriculum, is not a shared environmental effect, is because, that's not an environmental effects that differentiates family A from family B. Every family are get, is, the children in every family are getting the standardized curriculum. So, in some sense, it's a shared effect. But it's not a shared environmental effect here, because it doesn't differentiate the environments of children in one family from the environments of children in another family. >> [SOUND] So the next topic we're going to cover is gene by environment interaction. And we have a question from Diana M., and she says, in the last module of unit c you mention that g by e interactions are not the same as saying both genes in an environment are important. Maybe you could discuss that a little bit further. >> Okay. Well thanks for the question because this is really an important distinction and faculty in psychology departments have problems with this. So it's, it's a very important issue. The, when behavioral geneticist or actually genetic epidemiologist talk about gene environment interaction they're talking about and interactions in the statistical sense. They're not talking about simply that both genes and the environment influence a trait in question, or individual differences in trait in question. Gene, for every trait we're going to study in this course, for any behavioural trait, it's a truism, but both, that both the genes and environment influence the trait. Therefore it, it would almost say nothing, to say, if we mean gene environment interaction in that sense it would be kind of a vacuous statement. Because it doesn't say anything, it's true of everything. What we mean by gene environment interaction is, and I'll elaborate on this in a second. It's really a statistical, what statisticians mean by interaction, is that the effects of the genes and the effects of the environment statistically interact in influencing individual differences in the trait. Well what does that mean? It means that for interaction to occur, it eans that the genetic effect differs at different levels of the environment. So in phenocateneria it's a gene environment interaction. It's not just that genes in our environment have an effect, but it's a gene environment interaction, because the effect of inheriting the phenylketonuria genotype depends upon the environment you're in. If you're in a phenylalanine-rich environment, then the effect of inheriting or not inheriting that genotype is profound. Because it will lead to intellectual disability if you inherit the genotype but not if you do not inherit the genotype. So the effect of that genotype is very profound in that environment. However in a completely different environment, in a phenylalanine ,in a low phenylalanine die diet environment the effect of the genotype is very, very small. And maybe vanishingly small in, in terms of something like IQ. So the effect of the genotype depends on the environment, that's a statistical interaction, that is what we mean by gene environment interaction. Not just that both are important, but that they're both important in a particular way. [SOUND]. >> And moving on. [SOUND] And a way from gen-environment interactions. Ross Buckingham wants to know where choice fits into the equation. So he asks, do behavioral geneticists recognize the ability and freedom to express personal choice, in any situation that requires behavior response? >> Okay. >> So just because gene-environment play a role do we still belive in free will. >> Free will? Okay, okay so this is a hard question for me, I'm not a philosopher and we actually we'll we'll get into to the the this issue here, this often comes up in behavioral genetics what are the implications of behavioral genetics for the notion of free will? And we're going to touch on some of these things in, actually, the last week of this course. >> Mm-hm. >> We'll talk about the legal implications of behavioral genetic research. actually, I'm going to try to change the question a little bit. >> Okay. >> Bridget, because I, I think it's, it's a little bit difficult to talk about, would, I think is really essential, in essence a philosophical concept of free will in a scientific discourse. Philosophers debate this, I think most behavioral geneticists are determinists, and so if they, if, if, if, pushed maybe they would say that free will really might be located in the nonshared environmental component in a variance advice book, but that's not a very satisfying answer. I think what I'd rather do in trying to address what is I think a useful question is to say, and I think this is really what Ross is getting at perhaps, is does behavioral genetics research absolve the individual from responsibility. Regardless of whether or not there's a notion of free will, does genetic information absolve us of the responsibility of our choices? We do have choices. A determinist, I guess, would say that those choices might have some ultimate genetic or environmental determinants of those, but we do have choices. If our behavior is genetically influenced, are we less responsible for the way we behave? I think most researchers in this area would say no, it doesn't. And maybe maybe to take a pa, particular example of something that we won't talk about much in this course, if genetic factors influence your risk of alcoholism, which, they certainly do. Alcoholism is maybe 50% heritable, surprise, surprise. But, if there are genetic factors that make it more likely that if you drink, you become alcoholic, does it, the existence of those genetic factors reduce your responsibility? Or I, I, I think you could equally argue that it increases your responsibility. That is, knowing that, and knowing that you've inherited those factors, might actually increase your responsibility to drink responsibly or not drink at all. Similarly, and here the debate rarely comes up. If we know there are genetic factors for your risk for developing late onset or insolent independent diabetes, there's certainly are. >> Mm-hm. >> It's heritable. Does the existence of those factors absolve the individual that inherits them from acting responsibly on knowledge? I, I think no, I think it actually increases their responsibility, because now they know they've inherited the risk factor. It's now up to them to make the choices to mitigate that risk. I think the same thing applies here. Regardless of whether or not we have free will, we do have choices and those choices I think genetic research ultimately tells us, I think increases our risk responsibility to act in a responsible way. >> [SOUND] Okay. And our last question comes from Gwyneth James. And she says, my question relates to the graph titled results. What are the effects of carrying the s-allele on depression risk in module E? The graph cle, clearly shows that environment starts to impact on the risk of depression after two stressful life events occur. And it appears to be straightforward correlation. However, I recently read two bits of research that appear incompatible with this graph. So in first it was found that grandchildren of people who were exposed to significant emotional trauma were more susceptible to depression and anxiety due to altered cortisol levels, and there's increasingly ev, increasing evidence that this occurs due to a genetic change caused by the trauma suffered by the grandparent. In the second, it has been found that although levels of serotonin are implicated in some forms of depression, it is not merely, nearly as previously believed, indicating that there are other causes. Both of these studies seem to support the suggestion that the cause of depression may be the result of the impact of several alleles. Can you comment on this? >> Okay again, two questions so I have to try to sort the, make sure I, I, I, try to address both. They're both very interesting questions and they're kind of challenging questions, so thanks Gwyneth. So the, the first is really about grandparent effects. Something that we haven't touched on yet at this course, and we really won't touch on. But, well, well we, well we, we'll kind of skirt up to the border of grandparent effects in this next week. Actually the lectures for the fourth week will be posted by the time this, this video is posted. There's increasing interest about grandparent effects. So maybe I'll just comment on it, on that. within, within the field, certainly within behavioral science and as well as genetics more broadly, and the reason for that, I'm not familiar with the particular study Gwyneth refers to which is it sounds like it's the descendants, the grandchildren of individuals who suffered the Holocaust. >> Mm-hm. >> Survived but suffered through the Holocaust. The interest in, in grandparent effects is really, comes out of an epigenetics which we will talk about. It's actually the last module in Unit Four, talk about epigenetics. And there's two ways that people think about epigenetics. One, which there's a phenomenal amount of empirical data supporting its existence. Epigenetics talks about, again, you'll have to look at the last module and Unit Four to get at least an introduction to epigenetics. But epigenetics refers to factors, exogenous factors, factors not in the DNA code itself that actually affect the expression of the DNA, things in our environments that actually lead to stable changes in how genes are expressed. And that's really the essence of epigenetics. There's a whole set of studies now that show that there are things in our environment, including stress, that might affect how our genes are expressed. By that's not what's Gwyneth is asking about in grandparent effects. In grandparent effects, you're talking about effects that not only affect our DNA but the DNA in our germ cells, our sperm, our egg cells that then or passed on from one generation to the next. >> Mm-hm. >> Epigenetic inheritance. There are some s, studies that suggest that that may be going on, but whether, there's no doubt about epigenetics in the first sense. >> Mm-hm. >> That's important, we talk about it in Unifor. The second sense, it's a very interesting phenomenon, epigenetic inheritance, whether or not you get these epigenetic factors being transmitted from one generation to the next, so that you get what happened to your grandparents affecting how your genes are expressed. That, I think, is much more hotly debated, at least in mammalian species. >> Mm-hm. >> There are some suggestive studies certainly out there. I think it's still a very open debate as to how important that is at this time. But it's certainly things that are, people are actively researching. So it's a great question. Okay, but that was only the first one, right? Number two, what was, oh, number two was the s-elleles, right? >> Yes. >> Okay. >> Correct. >> And I think it was yeah, so this one, I guess, is a little bit more simple. If the s-allele at the serotonin transporter gene locus, if it's a risk factor for depression, and it appears to be one especially if you're exposed to stress, if it's a risk factor. Right, yeah, it's not the only genetic risk factor. There are many many other genetic risk factors that will be contributing to depression. >> Mm-hm. >> And we'll begin to see this when we talk about, we won't talk a lot about depression, but we'll talk about schizophrenia. >> Mm-hm. And in the case of schizophrenia, there certainly 100's, probably maybe over 1,000 specific genetic factors contributing to risk. So you're absolutely right. That's an example of one genetic variant that might affect your risk for depression, but there are no doubt many, many others that are not being accounted for in this particular study or in any of the studies really at this time on depression. Okay? >> Great. Well that's about it for Unit Three. We'll be back next week with Unit Four, and again, just a reminder if you have questions from any of these past units, just post them on the most current forum and we'll address them when we get them. All right, thank you. >> Thank you. [SOUND] [BLANK_AUDIO]
