Hello and welcome to unit five of Office Hours. So this week, you guys covered schizophrenia and we got a lot of great questions. So, why don't we just jump right into them and start with the first questions. Which evolve around what is schizophrenia? And our first one comes from anonymous, and they ask, there are clinical reports that support the environmental stress related factors. Such as a relative's death or a period of financial aid, and also substance abuse can trigger schizophrenia. How can this be explained by genetics and epigenetics? Additionally, can we completely rule out the role of parenting as a trigger of schizophrenia? >> Okay, good question, good question. First of all, right, we talked in the, in the module that. Clearly environmental factors are important in the ideology, the development of schizophrenia. And the question is well what are those specific environmental factors, and I think maybe the first thing I should emphasize is, clearly we don't know what all the environmental factors are, in the same way we clearly don't know what all the genetic factors are. What I tried to emphasize in the lectures was that, where the strongest evidence for environmental influences arises are in early prenatal, or perinatal traumas. I think, personally, I think the evidence about post natal environmental influences is a little bit more uncertain. Not that they're, that they don't exist but I think that the evidence at this point is a little bit less certain. We talked about migration, and I think really one of the standard explanations for the migration effect is there's some cultural phenomena going on there. And, more specifically to the questions ask, that this person asked. Two things. One is, can we rule out parenting as an etiologically relevant risk factor for schizophrenia. I don't think we can rule out anything at this time. So, no, I would say we don't. Although I don't know of any evidence that parenting is a cause, or contributes to your risk for schizophrenia. I think there's stronger avenues, and maybe this is what, what the question asker is getting to. There's stronger evidence that your relationships with other people can effect the course of schizophrenia, given you have, whether or not you relapse, how severe this schizophrenia symptoms you have. But I think it's a little less certain whether or not those actually contribute to the initiation of schizophrenia, or the origins of schizophrenia. So that's the first thing I'll point out specifically. The second thing that maybe is, is, and I can't remember if I've mentioned this, Bridget, in the lecture. I apologize if I'm repeating myself. But stress per se, I don't think. There's a lot of evidence that stress, per se, is a trigger for schizophrenia, and to me, some of the more interesting data on that, because I remember this one, was it last Friday was the 70th anniversary of the Normandy invasion? Well, there's a study, it's extraordinary stress those guys were under, right? Being shot at as they invaded the beast, the beach. But they didn't develop Schizophrenia. They develop other psychological disorders, when they're under extreme stress. But, Schizophrenia is not what they typically develop. So, I think the evidence for stress is a little bit lesser. And, again, it probably does affect the clinical course. Ideology I think is much less certain. >> Okay, so our next question comes from Laura. And she says, could it be that schizophrenia is not actually a disease, but an effect of another underlying disease or genetic variation? I'm asking this because of the relationship between Velocardiofacial syndrome and schizophrenia. I was thinking, perhaps, schizophrenia is not a symptom, with, is not but a symptom, but derives symptoms on it's own for someone that is, for something that is, not yet discovered, or is another condition or disorder. >> Okay. That's a great question. Laura. Was it Laura? >> Laura. >> Yeah, Laura. That's a good question and, and and maybe there's two aspects to answering it. Most genetically influenced disorders are heterogeneous. That is, it, it's rare to have one genetic cause for a genetic disorder. There are hundreds, maybe thousands, I don't know, of heredity forms of deafness. Breast cancer the genetic risk factors for breast cancer, there's maybe eight or nine genes involved, where you could have a mutation. There's over a thousand different mutations. So I think she's right, in one way, although we can't tell today. It's probably undoubtedly true that there are different genetic etiologies, and that schizophrenia is a heterogeneous disorder, and that if we could resolve, and I think that this is one of the whole, of psychiatrists, if they could resolve the etiology, they have much sounder classification system and they would divide up schizophrenia, probably this broad category of schizophrenia. Into different disorders, and that might be important in terms of treatment. Right now I don't think it is, but it, it may be important. So you might treat somebody, as she points out, with velo-cardio-facial syndrome differently than someone who might have mutations in, in genes X, Y and Z. So I think that's the first part of her question. The other thing I'll touch on here. And maybe it comes up later, I think you might have another question on this is that there, the boundaries between the disorders, the genetic research is indicating that the boundaries aren't as sharp as we once thought so that there's an overlap between schizophrenia and bipolar. so, what is that? What, what is it a unique disorder? Because we have these families where both are segregating. So, and, and we have the copy number variants where there are risk factors for both. >> Our next question comes from Peter, and he asks whether someone has been diagnosed with schizophrenia is a yes no question. But there are varying degrees of severity so is it possible to be quite schizophrenic but just fall short of the severity of symptoms needed for a positive diagnosis? At the other end of this hypothetical scale, is it possible to be just a little schizophrenic? >> So, strictly speaking, the, the Faulkner model conceptualizes the, threshold model, conceptualizes schizophrenia. Some, that there is some underlying continuum of severity. And that everybody is arrayed on it. So, so you're on there somewhere. And I'm on there somewhere. And somebody with schizophrenia is at the extreme tail. And they've developed schizophrenia. So, so Peter's right that, from this conceptualization, there are people who fall just short. And there are people above the threshold, various levels of severity. There is a movement within psychiatry. I'm not a diagnostician but there is a movement within psychiatry, at least among some diagnostician, to want to grade the severity of various psychiatric disorders. And I assume that schizophrenia would be a part of that. But certainly disorders like alcoholism and those things, where it's not, or ADHD, where it's not clearly an either/or, but there is some sort of gradation of severity. So, there's certainly evidence that for most psychiatric disorders that there is this gradation of severity. And that people who might not be diagnosed, may have some of the symptoms, and they may have some those symptoms because they're carrying some the relevant genetic risk factors. >> Our next question comes from Gwyneth and she says in unit five module one the graph variable age of onset indicated that 10% of women developed schizophrenia between the ages of 45 and 50. Is this showing that women have a higher risk than men of developing late onset schizophrenia? Also, I note the stages is within the range of the average perimenopausal age for women. Is this likely to have an impact on the risk for developing schizophrenia? >> So, she's right. So, there is, women have a higher likelihood than men to develop. Have a late onset form of schizophrenia. Most forms of schizophrenia are early onset. But, in, it. There are some, a certain percentage of cases that are late onset and women are much more likely than men to have those. Men are much more likely to have an early onset in late adolescence, early adulthood. In terms of the second part of her, Gwyneth, right? >> Right. >> In the second part of her question there is some speculation, I don't know all of the data on it. But there is some speculation, that estrogen, is a protective factor, and that you get this peak in women around the age of 45 to 50, because that's when women begin to undergo menopause and they're losing that protective factor of estrogen. So there is some evidence in supporting estrogen might be a protective factor, and I actually think that sometimes schizophrenics might actually be treated with estrogen. I don't know how effectively, but, I think, sometimes, it is used in their treatments. So, definitely, it's thought, at least by some researchers, to be a predictive factor. >> So, our next group of questions are about the familial risk of Schizophrenia. And, our first one comes from James. And it says in described labelled schizophrenia risk in Module C. It shows that the risk to a child is twice as great to the risk to a parent. Similarly, the risk to a niece is twice that to an aunt. So maybe you could explain that a little bit why we would see these differing risks for the same relationships. >> M-kay. [BLANK_AUDIO] There's a, there's a, a good explanation, and maybe there's kind of an uninteresting explanation. I'll, I'll begin with the uninteresting explanation. What you're seeing in that graph, is a compilation of a lot of different studies. And studies differ. In terms of the regular methodology and in particular, in how, how complete they actually ascertain the families. And, so sometimes when you're getting out to remote relatives part of the, part of the things, you kind of have to factor in is maybe the researchers weren't able to get the hard to get people and interview them. I don't know that that's the explanation, but that's certainly a possibility. One would, if you really wanted to make the distinction, one would probably want to look at the individual studies and see if they're comparable in terms of their methodology. However, it is the case, for sure. I, I'm, I'm less certain about the onset. Niece nephew difference for sure, if you look at the schizophrenia studies of, of comparable rigor the risk to a child is greater than the risk to the parent. That is, that if you have a parent with schizophrenia. The risk to his or her child is greater, I don't know exactly, I can't remember, it's about 10% I think, about 10% that a child of theirs would develop schizophrenia. However, if you flipped it around, if you took a child and asked what the risk to his or her parent was of schizophrenia, it's more like five or, it's about half, like, 5 or 6%, like James said. That's for sure, and the reason for, probably the major reason for that, is that and it relates back to the fertility in, in schizophrenia, is that many people with schizophrenia actually don't produce children, and so what you're doing is when, when you're looking at parents you actually have already filtered out. Maybe 50% of schizophrenic individuals, because they're not able to establish a relationship to produce a child, so that actually makes parents...it may sound odd to say, but becoming a parent is actually to some degree a selection for mental health. That is, individuals who successfully become parents, on average, are a little bit more mental, have better mental health than those who don't. So the lower risk that you see in the parents than in the children, and those are comparable genetic relationships, of course. Is because of this selection of parents for mental health. >> Our next question comes from Lisa. She says if a child is born if the child has an uncle on the mother's side with schizophrenia, and also an uncle on the father's side who has it, by how much does the risk increase for the child? Does it double? Or, or more than double, as compared to having an uncle from one side who suffers from it? >> so, the answer I think is it more than doubles. It's not quite additive. At the same time researchers, epidemiologists don't have familiar risks for all possible relationships that exist in families. So, I don't know if there's ever been a study that looks at second degree relatives on the father's side and second degree relatives on the mother's side. And then looked at the risk. The children might know such study might exist. However, there are studies of where both the mother and the father have schizophrenia. So it's, there we know it's, if it's genetic, and there is a genetic component to it, that so the genetic risk is coming through both lines of the family. What then, is the risk to the child? Well, we just mentioned. The risk to a child of someone with schizophrenia is about 10%. Your risk of developing schizophrenia if both parents have schizophrenia though, is about 50%. So it's five times higher, so it's not additive, it's not 10% per 10%, it really is if you, and so I would assume if there is, if someone could collect the data on. The aunt's and uncs, and, and uncles and what not from both sides. What'd you see is not an addition of risk. But at, rather some sort of multiplication of risk. Much like you'd see when you have both the affected mother and father. >> Our next question comes from Sally. And she's asked. In the slide about schizophrenia risk and the degree of kinship the risk of paternal twins is twice that of siblings. Is this a usual pattern in twin studies? One would expect the risk between paternal twins and siblings to be closer. Is this more indicative of an endogenetic and environmental influence? >> So, I saw this, I don't know the answer to this question, so, I saw this question coming. So I went to the expert. I went and asked Irve Gotissman, who, he's, he's the one that put together the research. So what was his explanation? It's, Sally's right. It, from a purely genetic perspective. The risk to siblings should be the same as the risk of dizygotic twins. Genetically they're, they're the same. They're 50% overlap. But it's, it's clearly higher for dizygotic twins. So why is that? Well he didn't, he didn't have an answer for me. So he, he didn't bail me out. the. It could be, so given that Herb doesn't have an answer, I'm assuming that researchers don't know since he's the world's expert on this. I assume it could be two things, one is what Sally's, Sally, was that right? >> Sally, mm-hm. >> Yeah. That she pointed out, and that is there could be some in uterine effect. The dizigon share but given that siblings are serially in the womb they wouldn't necessarily fear. So it could be an indication of that. And as we've said, there's certainly evidence for in utero effects on schizophrenia. So that's certainly a possibility. The other possibility again is this kind of uninteresting that, it might be some methodological thing. That the way twin studies have been done, are a little more rigorous than studies of siblings. And one way that you can see that is that with the twins they are the same age. So it's a little easier in assessing the risk and de, determining the risk period for schizophrenia. Maybe a little bit more difficult with siblings. So, it could be methological. It could be some sort of environmental trigger, like Sally is speculating, or it could be both. I th, th, think the safest thing to say at this time is we don't know. But it's an interesting anomaly in the risk figures. >> And then we have a question about heritability and liability. The question comes from Sophie, Sophie and she says, I'm not sure to have understood the concept of liability with regards to heritability. I think I'm confused with the two notions. Could you explain them a little bit? >> Okay, so it, it's a little bit of a complex thing and I don't want to, of course, get into the mathematics of it, right. I'll spare people that. But maybe a little bit of history, history here helps. It used to be thought that if you had a, a, an either or disorder, something where you had a diagnosis, that you had to have a specific genetic factor underlying that. So, it used to be that people thought that schizophrenia, if it was a genetically influenced trait, that there was a single locus in the genome, that was the schizophrenia locus, because otherwise, how do you have this kind of discontinuous trait? And it really wasn't until Faulkner Faulkner thought about this, and thought about what he called the threshold model. That people recognize that you can have either or traits, even though the underlying genetics is quantitative. And it, it get's mathematically a little bit, and maybe conceptually a little bit complex. But the notion is, is that underlying discontinuities like whether or not you have a diagnosis of schizophrenia or bipolar disorder or even things potentially like hypertension, although blood pressure is continuous, or a heart attack, there is something continuous, some quantitative trait. And you develop the disorder when your standing on that trade, it's called the liability distribution, your standing on that is above the threshold. So the, in, from this model or conceptualization, the inheritance of schizophrenia. Is really the inheritance of the liability of this underlying, unobserved quantitative trait. The heritability of schizophrenia, strictly speaking, the heritability statistics that are recorded, strictly speaking, aren't for schizophrenia, they are the heritability of this underlying liability. So, in the case of schizophrenia, people say heritability is 80%. And then, I actually, I don't think anyone asked, maybe someone did. If it's 80%, how come monozygotic twins aren't concordant 80% of the time, they're only concordant 50% of the time, right? Well, it's really not the heritability of schizophrenia, I hope I'm not confusing people more. But it's really the heritability that's underlying liability that's 80%. So, the underlying liability was a way of taking categorical traits like schizophrenia and other diagnoses, and acknowledging that they might have a quantitative inheritance underlying them. And actually, in a sense, Faulkner has been validated, because we now know, for sure, the liability for schizophrenia, there are multiple genetic factors contributing, hundreds maybe over thousands. >> Okay, our next group of questions are about the genetics of schizophrenia, and the first one comes from Gregory, and he says, I don't understand what's the profit from using tag SNPs for gene loss. They actually do not add any new information. It's just imputed from real snips. They are already there. So if there is an association with tag snips, the same association should be there with the original snip. We are just increasing the number of points to analyze. Diluting the original data set, and maybe shadowing other possible points of interest. Have I missed something? What information did the tags add? >> Okay, well that's a good observation, from Gregory, yeah? Because I didn't really get into this in depth. And he's right, if what's meant by. So I, I, I don't remember off the top of my head, but I think in the, in the, let's take the Saunders paper. I think in the Saunders paper they genotyped around 600 snips, and then they tagged 400 snips. Or, if you do a GWAS, you might genotype, today you might genotype a million or 2 million snips, but then, you'll then estimate. Another 6, 7, 8 million snips. So if it, if it's the case that you predict one, a snip you didn't genotype by one you did genotype, then he's absolutely correct. There's, you're not gaining any information at all by doing that. You might as well just analyze the snips you genotype. So, if you're predicting what you didn't genotype from one snip, there's no added information. Where you're getting added information, and this gets a little bit beyond, I think what we can do in this course, is that what you're predicting in those tags snips, or the ungenotized snips that you're analyzing. You're predicting them from combination of snips, of snips that you have genotype. And this gets into things called linkages, equilibrium, and haplotypes that, that we really have, I haven't really laid the quantitative genetic foundation for in this course. So, it's a very good observation and an important one. That if you're, if what you don't genotype is really just being predicted from one step, there's no added information, but really what's happening is the what you don't genotype is being predicted from combinations of snips that you did genotype. And it's those combinations that are adding information in the analysis. So our next question comes from Erica. And she asks, do any of the risks alleles found in GWAS, in schizophrenia coincide with the snips used in the positional cloning strategy? So that, in the secondary case, they fail to find a relationship because of sm, a smaller, smaller sample. >> Yeah. Yeah, that's a, that's a good question actually, Erica And I think that there may be some, she used the term positional cloning, I'll just call it gene studies. There may be some cases where somebody studied the candid gene, and it has emerged in GWAS. Of schizophrenia, I'm not familiar with any specific one, I think in general what's happened with GWAS is kind of interesting, and that is, it's not just true of schizophrenia it's true of many other disorders. Most of the things that are found in GWAS. Were not observed in any, they, they weren't even studied, in earlier studies. So they couldn't have been identified. So, in general, I think, the conclusion wa, is that we did candidate gene studies for maybe, almost 20 years. And now that we're doing this very large scale GWAS'. The things that were studied in the candidate gene studies aren't emerging. I mean every once in a while, I'm sure they do, but they don't often emerge. Which means that unless they're really small effects and you need even larger samples, the detective. Maybe we spent 20 years not accomplishing it. Maybe it was a cynical thing to say, I'm sorry. But maybe we didn't accomplish a lot, but I think, in general, it's one of the arguments that people make for GWAS is that you know, how else are you going to find these things? If you look again, you can look in the hypertension or other diseases. You get away from psychiatric phenotypes. And they're finding things that people had never thought were relevant. And they, they that was thought to be a major justification, for spending all this money for GWAS. Because it's opening up new biological insights into these disorders. [SOUND]. >> Our next one comes from Mary. And she say's I notice that you talk a lot about effects of individual alleles. Have researchers looked at the effects of combinations of genes? For example, where neither may appear to be risk factors, but jointly they are. Similarly, presumably, some genes are only risk factors if certain environmental conditions coexist. >> Okay, so Mary is getting an interaction, I think is what, and that's an important issue. And I'll, and, and one of it is gene, gene interaction. Right? So epistasis would be the term for that. And the other is gene environment interaction. So I'll take those in turn. Gene, gene interaction. As people look at the success of GWAS. And I think it, there's some debate about this, I'm sure. Some people say, well, you know, you need these massive sample sizes, you get these things with smaller packages. I personally think it has been worth it. I think it's really providing some significant insights. So let's take that as a reasonable premise, that g watts has been somewhat successful. Then the question is that in a GWAS, you're taking one snip at a time. And so, you analyze snip one, two three and on up to 2 million. And, what Mary was asking about, well you might snip 1 and snip 13, might by themselves have no effect. But if you took 1 and 13 together, they might have an effect. And geneticists would agree with Mary that that's a very reasonable conjecture. So let's, we need to develop ways to do that. The problem is, and, I, I can't do the common [UNKNOWN] in my head, but if you had a million snips, if you start taking all possible pairs of a million. You know, you have trillions and trillions and trillions of snips, so you need, you need to talk to some computer scientists to, to, to develop data-mining techniques. So, actually a very active area of research in genetic epidemiology now. Is to try to develop data mining techniques to do what Mary's suggesting and that is to try to look at combinations rather than just one snip at a time because maybe in combinations things, we'll find things. Similarly, there are people that argue, well maybe this snip doesn't have an effect in this environment but if we put it in another environment it does. the, I think there's more speculation about that then actual research, because, it, it, it becomes difficult to do the studies. I think there's some discussion, maybe what we need to do to find the snip's disorder act, let's make it generic. We need to find what the right relevant environment factor is and only study people in that environment because that, only then will the ge, the genes express themselves. The, the difficultly is two fold. What's the relevant environment? And can we get enough people? Because even in that environment, even if the genetic effect is enhanced. It may not be enhanced enough. That, that you need less than a, a large sample. So again, it's, I think it's a good point by Mary. People are trying to develop research strategies to address these points at this time. >> Our next question comes from James, and he asks, are methods for identifying schizophrenia alleles, such as GWAS strategies. Actually contributing in any degree to prevention. And, early intervention, at this moment in time. If not, can you see this happening in the near future? >> Well. They aren't. To my, to my knowledge they're not, they're not contributing at all, Do I think, in the future? Oh, I think absolutely, in the future. And, I think. In two ways they're likely to be helpful in the future. But how far away the future is, you know, maybe I'm not around and you can't hold me to this, but I think one is, and there's some controversy about this and I'll touch on it just briefly here, but one is to identify individuals at risk. Early in their lives to try to construct an environment that would prevent them from developing a disorder. And of course the controversy there is labeling somebody at risk for a psychiatric disorder and will that carry stigma. I don't know. I mean that's something that clinicians and ethicists are debating. And should debate, and maybe it's less of an issue with something like schizophrenia than things like alcohols or ADHD or what not. But I think, we, we're not there today, but I think maybe in ten years, I think, it, it, within the psychiatric disorders, I think it's certainly conceivable. That people could be identified as being at high-risk. And, there might be preventive steps that they might be able to take. In, including pharmacological innovations, that might modulate or reduce that risk. So, that's one way, I think, it's likely to help in the future. The other way, is, certainly, in treatment. And here I, I, I tend to think more in terms of pharmacological treatment and it gets back I forget the women's name who, who asked the early on question about header geneous nature of schizophrenia. If there are different types of schizophrenia that we could, associate with specific ideologies, then those might be most effectively treated in different ways. Certainly you see that with cancer now. Genetics is not my area but genetics has certainly transferred, transformed the treatments of cancer because now many cancers are genotyped and your treatment is based on the gene type. I think that's in the future of psychiatry. So I, today, no. In the future, I'm extraordinarily optimistic that this will have a big payoff for the, for the field of psychi, psychiatric. >> So our next question comes from Sabrina and she is looking for a little, more information about the missing heritability problem. So would you mind just explaining that just a little bit? >> Okay, missing heritability? >> Missing heritability. >> Actually, this is a question that, somebody answered I thought excellently [LAUGH] and I'll just give him a shout out as it were. Thomas, who's done a great job of answering question and making my life. Maybe your life too, a little bit easier. >> Oh yeah. >> But I thought, it's, it's maybe worth talking about here because it's a, it's a big question now in genetic computing now. So what is missing heritability? Heritability is, for most disorders, schizophrenia, hypertension, obesity, bipolar, depression, so on and so forth. Heritability is estimated, using biometric methods. From twin studies. From family studies. And so we have those heritability estimates. And then now people have done GWAS, where their systematically surveying the whole genome. Identifying genetic variants. And then looking at the effects of those genetic variance in aggregate, to see how much of the risk for the disorders are accounting for. On the one hand, the heritability estimates are high, but what we can account for by existing GWAS is not so high. I'll take a non-psychiatric disorder. Height, the heritability of height is on the order of 80%. What we can account for today in terms of [UNKNOWN] findings is maybe 10 to 15%, probably about 15%. 15%, 80% the difference between the two is the missing heritability. So a question that really is heavily debated in genetic epidemiology is. Well where, what's accounting for that difference between 15% and 80% for height? Is it ever smaller effects of these snips? Is it types of genetic variance other than snips like copy number variance or tandem repeats? Is it, what Mary was talking about? The interactions that aren't being accounted for in the GWAS analyses. So, what geneticists would really like to do is account for all that heritability, and so they're trying to explore all these different possibilities to see what's, what really is, is, is, can account for the missing heritage. [BLANK_AUDIO] >> So our next question comes from Edwin. And he says that you mentioned CMV's as a nonspecific risk factor for neurode, developmental disorders. And he says, why is the quantifier nonspecific added? Is it, is a snip specific? In what ways is it more specific than a copy number variant? >> So, I was using non-specific as, and I think its really one of the most fascinating things that has emerged in psychiatric genetic research of the last couple years, is that you have copy number variants, and, to tell you the truth I can't remember all of them off the top of my head. But, you have copy number variants. Such that, if you inherited deletion in this region, there, they recur. So the one that's most famous is 16 on 16 P. So if you inherited deletion in this region, and I might, this is where I might mess up, is what's associated with the deletion versus a duplication. But a deletion in the region and, and someone can just go back I think I have a slide on this in the, in the records. But deletion let's say is associated with an increased risk for autism, an increased risk for maybe epilepsy, increased risk for schizophrenia, and a duplication might be again, I may not have the disorders right here. But a duplication would be associated with increased risk, maybe also for autism, but maybe not for bipolar disorder. What's nonspecific is you have this one genetic factor that's associated with multiple psychiatric disorders, that may be have been thought to be somehow related in the past, but not this. And it, and it's both when the region is deleted as well as duplicated. So there's something pretty interesting and important going on there. So, I use non specific in the sense that the CND's any of them seem to be. A risk factor, when they're deleted. Also a risk factor, may be when they're duplicated from multiple psychiatric disorders, suggesting there's something that ties at least some, some aspects of these disorders together. [UNKNOWN] >> And how would you, how do researchers go about, finding C and Ds in the genome? I mean, you don't [UNKNOWN], >> Yeah, oh, so how did they, yeah, oh, how were they originally detected? I'm not 100% sure. How they're detected now is like if somebody, was concerned that they had a child with what I'll call Velocardiofacial syndrome, so a deletion on chromosome 22. They use a technique called FISH. F, it's an acronym, F-I-S-H. And I can't remember all the names that go into the acronym, but it's fluorescent, in situ, yeah, in situ hybridization. I guess I do remember. So that's what FISH is. So what, what they have is they have a probe. To the region which is essentially the complimentary DNA. And then they radioactively label that probe. And then they, they co-join with, with the, the DNA from that individual. And, if they have both, both chromosome 22's with that region then they, the probe should anneal to that, to that site and you should see two signals in their DNA. If, they only have one chromosome 22, then only one of them will light up, and the other, chromosome 22, it won't, because it doesn't have that region. The complimentary DNA will bind to it. That's the hybridization. And so you only get one. So FISH is used now, and it's actually used quite a bit in perinatal and prenatal diagnosis to look at children, especially children with developmental delays now. It's very standard for children with would appear to be anomalous neurological developmental delays that they would maybe get a panel of CNDs to try, with these probes to try to identify if they have either an extra copy of the region of the chromosome, or deleted regions that might be in a database somehow related to neurological development. >> Okay, so we have one last question, and this is comes from some earlier material. >> Okay. >> And this is from Isabelle and she says, I was wondering or I was under the impression that since females have two x chromosomes, they are less likely to express recessive genes that are carried only on the x. IE they would have to inherit a recessive gene on each of their two X chromosomes. How can this hold true if one of their X chromosomes is deactivated preemptively in every cell? >> Okay, this, that's a very good question by Isabelle. And actually it's, even though it's from an earlier lecture. It's something that we come back to again this week because we talk about X linked intellectual disability. Two things to point out. First of all, there is a period of time, it doesn't last very long, where both X's are active in a female. So there is that period of time. It, it's a matter of days, but, it they are apparently active early on. Maybe more specifically to her question, and, and I'll try to give an example, a little bit of an arm waving example, Suppose you had a gene that's expressed in a neuron that's on the X chromosome, and what she's really getting at is the heterozygote women. It may be and it probably is the case that okay, this neuron has, so let's say that what she's carrying the women she's a heterozygote, and what she has on one of her X chromosomes. The allele that would lead to an intellectual disability in the recessive state, so, if a male inherited that he would have an intellectual disability, but on the others she has the normal allele. So, randomly one or the other of those allele X chromosomes is going to turn off. But as you look at her neurons, this neuron will have the good allele, I'll call the good allele expressed. The next one won't and this one will. It probably is the case when you're a heterozygote, you have enough neurons expressing the good allele. That you compensate for the ones who, that are expressing the bad allele, so even in the heterozygote state you're protected. So that's why you, you get these lower risk in females because, because they really have to have two hits. They have to be homozygote because in the heterozygote state. Given the random turning off pattern, they have enough of the function of the good allele to compensate for the, the ones where the bad allele is the one that's the, the x exactly. >> Okay. So that concludes our questions for schizophrenia. This week you guys are talking about General Cognitive Ability. So get those questions in by Friday so we can address them next week for you. Thank You. [BLANK_AUDIO]
