 So, my name is Larry Brody. I'm going to wear a couple of hats in the Genome Institute. One thing I do is I do human genetics research, mostly looking into birth defects and cancer susceptibility and metabolites. So that's my research hat. I also run what's known as the Division of Genomics in Society, which is an extramural program that is the one that houses the LC program. Has everyone heard of the LC program? So that is, and I'm going to cheat a little bit from the time on genetics and just tell you about it because it's pretty cool. That's the program that is money set aside from our budget to look into the ethical, legal, and social implications. That's what the ELSI is, of genetics and genomics. And it's been in our budget for 20 years. Part of it was a response to when the Genome Project was just getting started. Someone said, wow, you're going to do this genomics and genetic stuff. It's pretty scary. Aren't you worried about the ethical, legal, and social implications? And in fact, Jim Watson, who was heading the program at the time, said, we are concerned about this. And in fact, we're going to fund research into the area of these implications of the Genome Project and have been doing so for the last 20 years. So some of them has been bioethics. Some of them have been legal work. Has anyone not heard of GINA, the Genetic Non-Discrimination Act? So that's the act that says you can't be discriminated based on your genetics alone. That part came out of this program as well, helping to set up the legal framework for it. And so that's what I do with the other hat. Today, what we're going to do, though, is just talk about genetics and general genetics. I am going to focus this lecture. And this is a lecture that I also give to a summer institute for not sure what the formal title is, but the Nursing Institute has a summer institute on genetics. And I've been lecturing in that course for the last decade or so. And the students in that course tend to be more interested in complex disease than Mendelian disease, but we'll cover both. But most of the lecture will be on complex disease because I'm going to guess that most of you are more interested in complex disease than strongly inherited type issues. And I warn Gina, it was interactive, meaning there'll be questions and answers. This is informal. It's a small room. There's not that many of you. So you can feel free to ask questions. There'll be some quizzes during the talk. So how many points? A lot of points. A lot of points. So be prepared. No, you have to keep taking it again until you get it right. This is education, not, yeah. So let's just talk a little bit about complex disease. I think that, and I will preface this by, this is not a lecture about stuff that you don't know. In fact, this is a lecture about things that you will find you know it all, really. And I'm not going to make this up. You know it all. And the lecture is just to kind of put it in buckets and bins so that you can figure out the relationship between things. All the concepts that we're going to talk about, in fact, most of the details we're going to talk about, all of you already know. And, but you may not have thought about them in context. So that's what I'm really here to bring. That's why the quizzes should be pretty easy because you already know it. And there's not, there's no detail, minor details of this thing that you have to take home. It's just the relationship between the different genetic concepts. This is really the take-home message for the next hour. And matters of scale and that kind of thing, right? So, I think everyone knows what a complex disease is. It's a little bit of genetics. It's a little bit of environment. And the two work together in some way we don't totally understand. And the patient gets sick. Or the patient gets tall, as we'll talk about. Or the patient is short, right? Not that short as a disease. From my perspective anyway. And so, we used to have to say, are genes involved in this condition? And anyone who's going to look into the genetics of a specific condition used to have to bend over backwards to show that genes were involved. I think the field has gone so far to the other extreme is that you can look into the genetics of almost anything and everyone will believe you that there's a genetic basis. Probably because there is. And what really we want to figure out from today is how to decide whether there's a big genetic component or a small genetic component. So we don't have to sell genetics anymore. When I first started this, it was crazy to look into the genetics of breast cancer. Literally, I mean, Mary Claire King, one of the pioneers in the field, told she was crazy to look into the genetics of breast cancer. And I think everyone knows how that story turned out. There are genes involved in breast cancer, right? So that was a success. The other thing that we used to do a lot more was try to find genes. Now, we don't have to find the genes anymore because somewhere in this course we're probably going to talk about the genome being sequenced and the way to look at the information. So we don't have to find genes anymore. We just have to figure out which ones are important. So we'll talk a little bit about trying to figure out which ones are important. A little bit about how you do this. And I don't know whether this group is going to go back and wants to write grants or wants to do investigations or just wants to be more informed about your patients. So we'll talk a little bit about how you do it. That's because the how you do it tends to change really, really fast. And I said I'm not going to load the talk up with technical details as opposed to conceptual things. And then why should we care? We'll get into a little bit. But I think the fact that you're here and have decided to come to a place that is normally much more uncomfortable in the summer than it is today says you probably care because you think genetics is important. I hope by the end of the lecture, if you think maybe they right now think they're important, by the end of the lecture, I hope you think they will be. So complex disease. How many of people do some super highly specialized clinical or have done, I don't know exactly where you guys are but in your locations or careers or but how many people do really highly specialized that only takes care of patients that have strongly inherited disease? Anyone do medical genetics? So that means that you're going to be looking at complex at traits or diseases that are complex, right? It also means that in some cases the risk alleles. So the risk alleles is just the flavor of the different genes that are out there. So every gene comes in multiple flavors. There is no breast cancer gene. There are just a gene involved in breast cancer that comes in one flavor that gives you a lot of risk and another flavor that gives you not so much risk. It probably will true be true that a lot of these risk alleles are probably common in the population for common disease. And I'll show you and this is one of the examples we're going to tie scale and I'll show you in a few minutes how that works. This is just a list of some of the diseases that essentially would be considered complex. I didn't put the few things that we say don't have a genetic component because you always things like trauma and car accidents. You say, well, that doesn't have a genetic component. It probably obviously does if it's related to alcohol intoxication, which is a metabolic process that probably has a genetic component. It's related to risk and thrill seeking. That has a genetic component. So really you could say that the genetic component is present in everything. Again, we want to figure out which ones have a lot and which ones have a little bit. And so what we're really looking at is genetic variation. And so we're not talking about today are changes in your genome that make you have a tumor. Sorry, changes in your genome that lead to it becoming a tumor cell. Those are somatic changes. What we're going to talk about is inherited variations. We're talking about the constitution of the genetics that you inherited from your mom and your dad. And I just use this slide to illustrate variety in part because it was in a magazine called Variety. And if you look at these two guys and these two ladies, there's a lot of variety there. So you've got ethnic variety. You've got the environmental side of things. And that's the other theme is measuring the environment and the genetics. The fact that these guys have short hair is environmental. The fact that these two ladies have blonde hair may be environmental, but no one's saying. But we can clearly change a lot of things based in the environment. And for those of you that can't see, I always forget. That's maybe saying that Sharon Stone and Goldie Hawn. And that's the Dalai Lama who actually gave a talk in this building last year. And I don't know who the other guy is. The part of genetics that you probably are most familiar with or pops into mind is this, right? Everyone thinks that's genetics. So who am I channeling here? Mendel, right? And if it's not that, so here's Mendel, here's his piece. And this is part why I got into genetics because as a biologist, we have very few rules. And we have very few times when we can really predict the future. Mendel could predict the future. He could cross a plant and he could say, half of these peas are going to be yellow and half of these are going to be green. And he'd be right almost all of the time. And so many of you know that the people who worked in his garden kind of thought they knew what he wanted. So they started cheating a little bit because when you go back and reanalyze this data, it's too good. But we'll just say that Mendel was not involved in being that good. And so Mendel's laws tell us that we can predict what the outcome of a cross is going to be. In human genetics, there are some Mendelian traits, but most of them are disease traits. So for example, and this is the first quiz. You ready? You ready to sit up straight? How many Mendelian traits can you name that are not diseases in human beings? Eye color is actually a complex trait because the brown or blueish kind of are the closest to Mendelian. So you can either be in some huge spectrum of brown or a huge spectrum of blue-green. So that one I'll give you. I'll show you a slightly different height. Height is definitely not Mendelian. And I'll talk about height in a few minutes. So what was that? Tongue rolling. I never get a straight answer when I try to figure this out in that some people contend that you can learn how to do it. My kids can do it. I can not. And I can't teach myself. Maybe I don't have enough patience. So tongue rolling the jury's out on anything else? Hair color is complex trait. We now actually just relatively recently know that there's some genes that are responsible for blonde hair in Europeans and some other genes that are responsible for blonde hair in Polynesians. Not the same gene, but tilt. Skin color, definitely not. Skin color is a gradient. And we'll get to this in a second. So the quiz, the answer to the quiz is we've got 22,000 genes. Most of them don't behave like the peas. And that's your first kind of take home point to really think about. The genetics that we've been talking about in the past and we've been maybe get most of the public thinks with these peas is the really highly predictive genetics. And that's not the case for almost all human traits. There are thousands of human diseases. I don't remember the total number, but it's up to like three or four or 5,000 that are in the catalog, the Mendelian inheritance of man catalog that are mostly Mendelian. But those are specific diseases. Most of them rare. What type is a good one if you do ABO? Yeah, and obviously your minor antigens as well. So, but that's not a trait you can see. So I'll give it to you as a point. But it's if you think about what you can see because mental could see the peas, right? And the other one that people talk about is the free versus the detached earlobes tend probably is one. There's some dimpling things in a few others that are Mendelian, but by far most of what you see around you in variation in skin color and height and hair color and even eye color is not Mendelian. So when you're thinking genetics, think what kind of genetics? This is the other thing that you think of when you think genetics, right? This is the fruit fly experiments. And these got, these became a great tool to do genetics in part because peas you can only do, I don't know, once if you're in Austria, probably once a season. Fruit flies you can do every couple weeks. And they really were the workhorse at the turn of the century and started because there were flies with different color eyes that actually the offspring could produce white-eyed flies and red-eyed flies very much like the peas. And so that gave the first organism. Do you say most human genetics are not Mendelian? Most human traits, yeah. Most human traits are not, right? And Mendelian for those, and I'm going to stay away mostly from jargon, but Mendelian in this case means the mode of inheritance is pretty clear. It's dominant, it's recessive or excellent. And you'll hear more about those in other parts of it. But most traits don't look like that. This is the other things that we also know, right? So we have, I'm just telling you genetics is maybe not as powerful as we think in a lot of cases, but this is a pretty powerful example of genetics, isn't it? These two girls are identical twins. We've not weaved Diane Arbus, dressed them, or they dressed as identical. And you know that identical twins share all of their DNA the same. We now know there's a few differences, but not many. And they have remarkable concordance in their phenotypes. Is anyone known identical twins? Are you, you're an identical twin? No, no. You are. Were you so identical that people couldn't tell you apart when you were younger? There's one person in the whole world that never got my sister, myself confused. And then how about later in life? Why is that? Because your genome didn't change. Stress. But environment. Yeah, I don't need to get personal about what. But right. So even identical twins, you can start to see this divergence. And we've been in a 35 year study with Marta, the middle-aged twin registry, Dr. Kenler. I know of this. Ken Kenler? Yeah, yeah. Yeah, there's actually a study of twins as a whole field of study. There's a journal on twin research. There's a group that's based in Australia that coordinates the entire world's twin research. Okay. What about these guys? Not twins, right? I'm going to guess that about half of you are more. Know who these guys are. Well, that should be both children and that is somewhat different. Right, right. I can tell you that when I show it to college students. Yeah. So these guys clearly show you the other range of variation. Right, so they're phenotypes. Their traits are very, very different. And yet you know that if we were to cross these guys like Mendel's piece, just to forget the biology in just a second, we would not get offspring that were tall short, tall short, tall short. Right? Or black, white, black, white, black, white. Right, so this is what I mean. You know this stuff already. You know that those things, the skin color and the height that we're talking about, is a complex trait because it does not segregate in a nice, easy fashion. We also know that environment has some influence but not over everything. I mean, Wilt Chamberlain went to high school with my dad. So they had the same environment. Do you think my dad was this tall? Did you have that? I don't think so. You play baseball actually, but not to this extreme. And you don't think so because you're looking at me and you're saying, yeah, that's not a tall guy. Student, younger students don't know who they are. Okay, if we finish early, I'm going to do that too. So here's height. Right, so here's how you think about complex traits. And everyone knows what this distribution is, right? Looks like a normal distribution, a bell curve. Height is mostly genetic. And so height is, your height is determined mostly by your genes. Yet in the population, we don't see short, tall, short, tall. We don't even see middle, short and tall, right? We see a nice, smooth distribution. This is only, I don't know how many kids are on here, but these were military Academy students. Looks like about 50 or 60 of them. Maybe more than that. Maybe 100 some. How many genes do you think control height in humans? 100, probably 100 strong ones. 100 strongest, let's say not 100 strong ones because we don't. And we didn't know that until recently until we started genome-wide association studies. If I take the strongest one and I say I could magically switch, say I have the weakest of the ones that have the influence and I take the one that has the strongest influence on height and I have the weak one now. And I gave myself two copies of the strong one. How much do you think I would shoot up? As far? No? No? Millimeter. Yeah. I mean, if I was born with it, yeah. Millimeter. So, am I moving a millimeter? I can't quite tell if I'm moving it. Little tiny effect, right? From all these genes. But together they all determine because everyone has their own constitution determines their height. The other, again, I'm going to hit these themes again and again. Look at the average here. It's about five, six, five, seven. And these are all guys, right? I'm five, six. I'm not average anymore. I'm short now, right? In my cohort, I'm short. And in the next cohort, I'll maybe still be short. And so why is it shifted to the right now? Environmental influence. Environmental influence. See, I don't need to be here because you guys do know this. So environment has shifted the entire curve to the right. So when you're thinking diseases, you want to think, especially if you're thinking from the patron's center point of view, you want to be thinking, what is it about the genetics and what about us in the environment that pushed this patient into this state that we see them at today? This is the eye color just to show you that there are a lot, there are multiple genes, and maybe there are genes that determine whether you're brown in the brown flavor or the non-brown flavor, but there are multiple genes. And we know most of them now. This is not realized. This is taken from a company that will let you design a doll that looks like you, which I find a little creepy. I don't know if it's a, no, maybe it's American girl. Could be. Yeah, I think it may have been that one. I just, I looked for eye color in the web. And so that, and you've, I mean, I don't know about in this room because it's dark, but we probably have almost this whole range even in this small sample within this room. Okay, so we talked about what that, sorry, I should have made this not step through. This just encapsulates what we're talking about in that you end up with a disease, you have a germline predisposition, you have environmental influence. There are times, especially in cancer, and you're probably going to hear more about it, where somatic changes, changes in the cells that make up the disease tissue are quite important. And I always underline this because we do a miserable job at accounting for and doing research on behavior, which is if you think of the top five causes of death, a lot of them are driven by behavior things. We talked a little bit about this already non-Mindelian, can be modified by environment. Height is an example of things where you get effects. It's not totally a truly epistasis and I won't, don't worry about that term if you haven't, you don't have it already. But height is where there's combinations of genes and sometimes you need to have a gene over here and a gene over here in order to get the trait. If you want a good take home project or some of you teach, it sounds like some of you teach already, look up Labrador Retrievers. If you want to use an example for epistasis, it's a great example for students to work through. Because you know they come in three flavors, one of which is really a flavor. And the genetics of them are, the genetics of those three flavors is actually more complicated because there's something called epistasis going on. It's a great teaching module. So instead of genotype, equaling phenotype, we already talked about genotype approximates phenotype to a certain extent. The equation is actually more complex. The phenotype is the genes plus the environment. It's actually more complex than that. And again, and so all these, can they, all the slides are going to be left here and you can have them all. I did pull a bunch of images off the web that are used with fair use but not for reproduction. You can use them but don't tell them you got them from me. So really the phenotype is the genes plus the environment but genes can interact with genes. So that's the gene-gene interaction and then genes can enact an environment in a multiplicative way. And so if we were really good and we wanted to be as good as Mendel predicting peas for any phenotype, let's say height, we would know all of these things and know how to put them in and then we would be able to make reasonably good predictions. We are nowhere close to doing that for most complex disease. In part because one of the things, if I look at my DNA now and say, what's the chance I'm going to get type 2 diabetes? I would know some of these things but I wouldn't know the ones in the future, right? And by the time I had my blood sugar being through the roof, then you'd probably be pretty good at predicting it but you wouldn't have needed the rest of the stuff anyway. So we're not there yet and that's a pretty important take-home message that we're just not able to make super strong predictions about complex disease. Can we use them? Can we use genetics in complex disease? Here's a trait, think of it as... Let's see, let's make the right to be bad. So let's think about blood pressure. We now know a lot of the genes that contribute to blood pressure has a strong environmental component, has a stress component but it's distributed like this in the population. What the geneticist learned, told us at the turn of the century was a little bit later in the turn of the century is that a distribution that looks like this really could be made up of something that looks like this. And if you were to sum those three distributions, they would add up to that one distribution, right? So this was the work of a geneticist called Pearson. Ever heard of him? Pearson correlation coefficient? That's the Pearson. He was a geneticist and he was using, he invented the Pearson correlation coefficient to help describe his quantitative genetic studies. So now let's imagine, and I'm oversimplifying it because I can't draw it if it's that complex. Let's imagine this is one genotype, this is one genotype, and this is another genotype. Seems to make logical sense that you'd have people who have a genotype that pulled the trait in one direction. You have people who have a certain flavor that pulled the trait in the other direction and yet there's lots of overlap in between the two. So the genetics itself doesn't project. If it was a Mendelian trait, we'd have three peaks separated by nothing, right, a gap. So can we use genetics? Take a look at, pretend this is where a disease level is. So whatever the trait is, you get up here you have disease. Now look at the proportion of the different genotypes in the disease pot, right? So whatever this, and I guess it is, colors didn't seem to bleed through the right way, but this distribution of this genotype, there's more people in there with the disease than the other genotypes. So if you were designing a public health intervention that could be benefit by stratification by something other than age, which is how we do most of our public health interventions, you might say, well, the people who have the yellow genotype, maybe they don't have to screen, let's say it's screening for something, screen as often or start as early, whereas the people in the green genotype or whatever this is, maybe they should start a little bit earlier, do it more often. That's the hope that we can have enough prediction, not to tell an individual what's going to happen to them, but that we could shift the populations and shift practice a little bit. And this could be the power of complex genetics because we can now add up all of those things. And I didn't bring this slide, but we're getting close to having enough data to start modeling this for breast cancer. For those of you that know, mammography is very controversial because of whether or not it has political components and there's people who say it doesn't really save any lives. There is undoubtedly, and we know in the population, breast cancer risk has got some different distributions. The question is, are they separated enough like this to actually revamp screening? Is it like the U.S. company that health service does? No. Because they cannot put the, you know, don't have enough mammography and don't have screening for prostate cancer, don't have screening. They do that for the entire population. They don't think about... They don't think about mass. Right, and this is where epidemiology, for some of you that study, and genetics is going to merge in a good way. So the epidemiologist has to believe that everyone in the population is the same. The geneticist believes that everyone's different and the environment doesn't matter. And they're both wrong, right? Everyone in the population is not the same. We know that. Geneticists know that like crazy, right? In fact, everyone in this room is different, except for you. You're not different from your district. But we also tend not to be good in the environment. So if those two things merge, then the U.S. preventative task force 20 years from now might be tailoring recommendations toward genotype. Yeah, Jean? Repeat the questions. Okay, thank you. I'll repeat the questions. So this is one of the other big take-homes, is this is what we think risk looks like. This happens to be for... I actually drew it in for breast cancer. But this is the one of the ones that you should reproduce and put in your book to think about, okay? So here is how common it is, right? So the more common is to the right. Here's how strong a risk is associated with a particular flavor of a gene. And once you start getting up to 10, 20, 30-fold risk, that starts to be Mendelian-like, which means you don't find it very often in the population and you find it quite often within the same family again and again, right? But you know those are rare, right? So this is 10% sub 10% of the population. So those kind of genetic effects are quite rare. This really is a continuum. Down here are the things that you find by genome-wide association studies. So for example, a 30% allele frequency means that most of us in the room would carry something. Actually, most of us carry lots of things that predispose us to specific diseases by just a little bit. One would be no risk at it, right? So they're in the 1 to 1.2, 1.5 risk range. And that's a little tiny amount of risk. And so there will be thousands and thousands of genes and flavors of genes that you hear about that fall down in here. Heights up to 100, diabetes is up to 100, breast cancer is up to 100, prostate cancer, I'm not sure where they're at. But any condition that we really push will have 100 things or more that do a little bit of risk. And that's hard to figure out what to do in the individual. Maybe okay in the population. And then here's like things where BRCA1 would stand, so in breast cancer. If a person carries a BRCA1 mutation, the risk of breast cancer goes up by about eight-fold. That's a lot given the occurrence, the population rate is so high. But very few, a few percent of the population actually carries those. And if you're not used to looking at relative risk, if you were to put lung cancer and smoking on this, the prevalence would be down around, actually it would be up here and be up there maybe, about 10-fold. It's about 15%, I think it's down to. It's going down, but it's still 30 million, 30 million Americans smoke. There are things that are very, very low risk and very infrequent, but we can't study them because you require probably samples of a million people to find them. So this boundary here is an artificial boundary because it's not tractable. There probably are not genetic things that are very, very common, high frequency here and very, very strong because we would know about them already from family studies. And this is really the landscape that you want to be thinking about if you want to translate genetics to the conditions or traits that you're interested in. This is a little bit of the why's in case you haven't heard this or internalized this yet. We really do think that there probably can be some predictive and preventative type benefit from some of these, understanding some of these genes. We still have a long way to go to do that. And the medical community has a long way to go to embracing prevention because right now, especially once you get into the hospital situation, if you're already in the hospital, it's too late to do prevention, right? And so most of the money is going into acute disease and chronic disease where prevention is too late. That's where the preventative task force can actually help. Disease classification you probably have and will have heard about using genetics to help understand better the disease. And then the other, and I'm sure there's a lecture in this course on pharmacogenetics and pharmacogenomics about understanding drug response. So these are a lot of the why's. Let me do a little bit on how you study it. We could study complex diseases in families if families looked like this, right? So if this was a Mendelian disease and the green box is someone who has the disease, so if this was a Mendelian disease and let's say it was dominant, how many of these kids would have the disease? Right, about half. So the fact that it's only three out of a gang tells you it's not Mendelian, but there's probably something going on in this family. Anyone come from a family that looks like this? One time someone did have 15 siblings when I used this slide. So we can't do this. We can't use family studies to study complex disease. We can use family studies to study Mendelian disease. So what we do is we do twin studies, as we already talked about, and we look for familial aggregation and those allow us to estimate heritability. Heritability is just a fancy term for how much of the phenotype is involved in a person's genome compared to a person's environmental exposures and other stuff we don't. I can tell you that we've been doing heritability studies for a long time. They probably overestimate the genetic component and geneticists haven't been as good about saying that because we like to have a lot of stuff in our column. But as the twins I showed earlier, they share a lot of environment too. And so the ideal way to do this is identical twins reared apart, and that's a hard thing to do. Reared apart at birth. And so that happens and you hear remarkable concordance things that happen, but we don't have tens of thousands of those to study. This is, again, back to the twins. Monozygotic twins share 100% of everything, their alleles, dizygotic 50%. The environment, you could decide whether it's similar in both cases or not. I think we need to get better at that. If you take a look at various diseases and look at the concordance in monozygotic twins, so take multiple sclerosis, 18% of cases where you have one identical twin with multiple sclerosis, the second twin has it. If you look at dizygotic, that concordance is 2%. So the delta here, the difference between this number hints at the genetic component. And multiple sclerosis is one of those things that even 10 years ago, 10, 15 years ago, if you said you were going to look into the genetics of it outside of HLA, the immune system, people would say that it's crazy. This data tells you that there is a genetic component. And so here's type 1 diabetes, pretty strong genetic component. This is, we know what actually one of the big players in this is the HLA, there's specific haplotypes in the HLA that strongly predisposed to type 1. Schizophrenia, arthritis, this got shifted. Rheumatoid arthritis, osteoarthritis, cleft lip and palate. So here's a range of conditions where twin studies have said if this number's a lot higher than this number, there's a genetic component. The other way to do it is epidemiologic studies. You can use families like this and do fancy segregation, but it requires getting everyone in the family and sometimes in multiple generations. And if you're looking at something that's very late onset, it's quite hard to get the two previous generations to do that. So often what we do is epidemiologic studies. And those are great because they're simple. Well, something happened in this one too. Let me just give you some examples of how we can, before you go and watch in, how you can look at it. So here's a condition that is present in one individual and then present in another family member. But this, not that present in the population. And so you can look at the, the formatting all got thrown off on this one. You can look at the frequency in the population and the frequency in siblings and see what the ratio is. And these, this one, some parts have fallen out of the slide. So I'm going to skip this because the date on the slide is not right right now. So this is kind of what I was talking about, linkage analysis in families, Sib pairs in individuals, and then these things where we're counting. And so we can think about genetics moving into the counting realm, which I don't have the slide yet, where you can take thousands of individuals with a disease and a thousand, thousands of individuals, and it really does take thousands without a disease. And you can look at their genome and figure out what is more present in one group than the other. And it's, other than lots of multiplication, lots of statistics, it's no more complicated than that. And that's what we're doing a lot for complex diseases. I'll just use an example. This is, everyone know what this is? It's a mammography, anyone read mammography? This will help. That's not good. So that's a speculated calcium rich tumor. If you wanted to study the genetics of breast cancer in a population, this shows you the data on the rates per hundred thousand for the different ethnic groups. And what you'll see is that very few women have cancer down here. Most of the cases are up here. If you were to study breast cancer and you wanted to get to the genetics, you'd be better off enriching for these younger onset cases because that's getting them earlier before the environmental exposure has also contributed to it. And that's obviously works for breast cancer because we have breast cancer genes. This is what breast cancer might look like in a family. Note, what about this guy? Can someone tell me what's going on with that person? Yeah, so that's incomplete penetrance. And in this case, it's what's known as sex-limited incomplete penetrance because he's a he. And so he, presumably from this information, you can tell that he inherited whatever predisposition was in his sister and his mom and passed it on to his daughter but did not manifest the disease because he was a guy. Now, he could because one percent of breast cancer actually occurs in men, but he didn't. And the person who is, these women have breast cancer, I can tell you now, and I'm not going to talk about it, but identification of these genes has now led to treatments based on understanding BRCA1 and 2 biology to treat tumors in people who are BRCA1 and 2 positive. And drugs that have already gotten to, they're in phase 2 trials right now, and they directly came out of understanding the genes. What we also can do is this woman wants to know what her risk is. And so who thinks her risk is the same as the population risk, which is about 10% lifetime. You already, again, another thing that you already know, intuitively, her risk must be higher. Empirically, we know that if you don't have any other information other than an affected mom, her risk is doubled. The risk goes up for having more relatives. If I told you that this was not breast cancer and ovarian cancer, does anyone work in this area? Does that make the risk go up or down that her risk of breast cancer go up or down? Because ovarian cancer is a feature of having a BRCA1 mutation, and it makes it more likely that something is going on in this family. So this is where the BRCA1 and P53, which is leaf-ralmeni syndrome, kind of fall on this risk spectrum. There's a bunch of things that are starting to fill out the middle in breast cancer. And this, there's already 30 to 40 of these in here. It's the P53, it's the formal name of the P53 gene, which is the gene when you inherit a mutation in it, you have leaf-ralmeni syndrome, which is an early susceptibility to lots of different types of cancers. It's a very rare syndrome. It also, P53 happens to be the gene that is mutated in lots of tumor cells. In fact, about half of tumors have a mutation in it. So it's a very, very, very highly studied gene. And you can have germline and somatic changes in it. This alphabet soup will expand going forward and already has. Genetic variation can kind of produce protein variants. You can ignore this because we don't do this anymore. A lot of us spent way too much time on RFLPs. This is where a lot of the action is and DNA sequencing is how you find variants. This is just what a SNP is if you haven't had it already. If I look at my chromosome, I might have an A there and everyone else in the room might have G's. You might have a T there and everyone else might have C's. That's all it is. It's just minor changes. In this room, we have 18 people. We've got easily 30 million differences between all of us. Most of these differences don't really mean anything. Oops, sorry, I did that. Why do we have that? DNA is amazing in that it can copy itself. So if I were to ask a scribe, the best scribe in the world to copy a document, word for word, the best they can do is making one mistake per document. So if it's got a thousand elements, the best they can do is one in a thousand. And this is actually information theory because you have to copy one thing to the other. DNA actually almost is as perfect when it replicates. The reason that I say luckily for us it has made mistakes because if you didn't make mistakes in copying DNA, where would we be? We'd still be in the primordial soup because we couldn't have evolved. So there would still be nothing without these errors. Most of these errors that are in DNA and all this variation really just doesn't mean anything. It's great for geneticists because we can follow it around in populations in the genome, but they're really not of consequences. So finding a genetic difference does not mean it causes anything. That's another thing to keep in mind. It also gets, this is the frequency of genetic variation. They get injected into the population at some rate and they drift out of the population at some rate. So there's always new genetic variation. Most and all of us in this room have new variants because when the two cells that came together that made us, each of those had some new variants, but most of them will go to the grave with us because they won't be selected for, they won't be in our future generations. Occasionally you get things that either is selected for or just drifts up to being a high frequency and these are things that can be selected for. If you look into any population, you see that these are variants at all different frequencies and so if I were to look at everyone in the world, I'd find all kinds of frequencies of variation. What I won't find is too many variants that are present in one ethnic group compared to another. You might think that there are variants present that define an ethnic group and don't define another. In fact, because most of these variants were in the human population when the human population lived in Africa and then diverged out, they're all shared across. What you'll find is different frequencies in some populations to others. Some groups will have high counts of a certain variant compared to others, but almost all of those don't have any biological function at all. We used to worry about how you could find... If you were going to go back and do a study, how could you find genetic variation? You used to have to roll up your sleeves in the lab and find it in the people you wanted to study. Now you don't have to do it because it's all online. Thanks in part to our institute. There are whole genome sequences online. There's catalogs that have millions and millions of SNPs online. If you want to look for genetic variation to do studies, it's all been found for you, or almost all been found for you. The other trend in genetics, which you're going to hear more about, and I think everyone has to show this slide, is that DNA sequencing has gotten incredibly cheap compared to what we're used to. Not as cheap as I think I would like. To give you an idea of what this means, it's really come down 100,000 fold in a decade. A little bit more than a decade ago, I bought a Honda Civic. If I were to buy it again at this time, and the price of my car had come down in that rate, how cheap do you think my car would be? Just to tie it home. A dime. That would be a good deal. Everyone thinks they've got a great deal on their car. A dime would be a really good dive room. Technology is constantly evolving. This is why we're not going to spend much time on it, but there are DNA chips that you can do, and there's DNA sequencing machines that can sequence an entire genome in a few days now. There are maps of genome-wide association studies. This is an online map. I haven't updated it in a long time, but it's too crowded now. This is the different chromosomes, and every dot shows you where there's a trait that's been tied to a genome-wide association study result. If we were just doing height, you could imagine there would be 100 dots just for height. What you don't want to do is say, look for the new genes. I'm going to find the new genes that are reasons that my patients have kidney disease, and I need to take their blood. The first thing you need to do is look on the existing maps and see whether or not someone else has already found the genes that predispose to kidney disease. Some of them are harder to find in the literature compared to these online resources. It's easier to use these online resources. This is just an example of what sequencing machines look. Are there going to go to NISC? I'll make you not feel bad about missing the tour. You'll see a room with some machines in it that look like this. You used to need all these machines, and now you need just smaller machines. It is actually a good tour, but nothing blows up with bubbles or anything like that. It's just small things. This is a lab that would have been a DNA sequencing production center using the sequencing technology of five years ago. There's this and there's a backside. Today, if you were to go up there, this would take care of all that. It's gone from that, which are automated but still need someone to do some care and feeding to this. Our sequencing center has not hired new people. In fact, we shed some people because we don't need as many people because these machines, these will do this entire row in a day. This entire row would probably take a week to do what one of these machines do in a day, one of these new looking machines. If you were going to NISC for the tour, you'd see some boxes that are gray. Let's talk a little bit and wind down with what you can do with this. We talked a little bit before about personalized medicine, really having some impact in diagnostics, pharmacogenetics, risk assessment. Hopefully, if you can do risk assessment, you can do risk modification, as well as understanding the new biology and new drugs. The example I gave there was the BRCA-1. Understanding what BRCA-1 and 2 did opened up a whole new avenue of biology and the understanding of DNA repair, and that allowed some smart people to say, well, if these people have faulty DNA repair in their tumors, maybe I can get a drug that will just hit their tumor. It really does. It hits the tumor, there's some nausea involved, but compared to conventional chemotherapy, which is toxic to all cells, it's really quite a breakthrough. The problem with these, I'll just tell you in advance, when we target toward the tumor, often the tumor figures out how to outsmart it and comes back. We're not there with the breast cancer ones yet, but the lung cancer experience is such that you can give someone more months of life, but you can't cure them yet. But it's really helped to actually open up new avenues, because otherwise, what we're doing, most of you have heard what oncologists do is, you know, slash poison or burn, and having a tool that's a little more fancy than that would be great. So these are the realms where we think genetics and clinical practice really can have an impact. I wanted to leave at least five minutes for questions. We have seven, so I'm ahead of schedule. And I wanted to end with that kind of, again, it was a broad brush of things that mostly you know, but now, especially that one diagram of the landscape is something you should really kind of think about. So when you hear gene for type 2 diabetes, is it a common allele and is it high risk? Or is it a commonly on low risk? Think about the parameters of frequency and potency, for lack of a better term when you're talking about genes. Or conditions, right? You had a question? Yeah, so the question was about wheels coming and going in the population. And don't they stay in Hardy-Weinberg, whereas the next generation almost has the same frequency as the previous generation. And you're absolutely right. What was not put on this slide was the time scale. And so the average time for something to go from here to here is it's a statistical model, is four times the effective population size times the number of generations. And so in the founding human population they think it was about 10,000 individuals. Four times 10,000 is 40,000 times 20 would be that number of years to take to have the shift. So you're absolutely right that these things take tremendous amount of time to drift. One example, and this is why you cannot make the population better by doing the kind of breeding that people tried to do in the eugenics movement. It just doesn't work. So anyone take care of cystic fibrosis patients? So up until recently no one with cystic fibrosis has ever reproduced. Because the males are infertile and the women have often died before reproduction. Has the frequency of cystic fibrosis in the European population changed at all in the last 100, 200, 500 years? Probably not. And that's with having everyone who has two bad alleles never reproduce. So that shows you how stable the Hardy-Weinberg kind of thing is because there's a lot of heterozygos. So if people tell you we're going to use genetics to evolve the population it's totally crazy. You just can't do it. Because of this big pool. Yes? Sure. So the question is these new compounds that have been they're now in phase two trials for treating tumors in individuals who are BRC1 or 2 positive. They take advantage of the fact and I'm ashamed I have a picture of it and it's easier in a picture but I'll just tell you is that DNA has two strands, right? And think of these two strands being really long in a chromosome. If you have a double-stranded break in those two strands then the chromosomes can fall apart, right? Because they're no longer attached to each other. So if you have a big chromosome. So double-stranded breaks are really bad. If you have enough of them the cell dies. If you have a few of them the cell might become cancerous. A single-stranded break on the other hand if I have one strand up here and I have a break here in the other strand everything stays together. And it's easier to fix that. So there are two different components to fixing DNA. There's a whole bunch of proteins. There's actually a lot of proteins. It's almost 20 or 30 of them. That specialize in sensing and fixing double-stranded breaks. There's another set of proteins to fix single-stranded breaks. In a BRCA1 cell the cell that's deficient of BRCA1 it has some extra double-stranded breaks. But it apparently can survive because it doesn't have that many. If I take a chromosome now that has a single-stranded break and I'm ready to copy it so the strands come apart I now have one I can copy and then I have another one that falls apart because it's already cut. And so what the folks in the UK said is what if I inhibit the single-stranded break machinery that would flood a cell with double-stranded breaks if it tried to copy itself and a cell that a normal cell might be able to handle those double-stranded breaks but a BRCA deficient cell might not and it would commit, commit, sell your suicide apoptosis. And in fact they did that in a dish. They did it in a mouse. They had some change in disease-free progression in phase one trials and in phase two trials there's significant disease-free progression and they're doing mostly in ovarian cancer patients who have failed all therapy. The study that came out actually just a couple weeks ago in Lancet Oncology showed a five or six month shift in disease-free progression. It didn't show a decrease in mortality in patients that were studied. They're still at the midpoint. So that's that particular mechanism derived from two things. Understanding this complexity of DNA repair and understanding where BRCA plays a role in it. Happy to answer any questions about, yeah? So the question is, where do you see the role of nurses in dealing with this personalization of medicine? And I will tell you that it's been shaped by a student that I had in my lab who was a nurse who did a PhD with me and by Jean, too. We're totally under-utilizing nurses in general. They're really trying to make sure that they're able to do that. There are a lot of nurses in general. There really should be a specialty there's some technical term for it, but a thing that you could certify in so that you could do risk assessment and you could do behavior modification if needed. And you could be the local expert in genetics because we know the medical community is not ready to handle this. There's not enough genetic counselors to handle this. And there are a lot of nurses and a lot of them are very anxious to that's why you're here, right? You want to learn more. And so Jean has written about this. It's a ready-made workforce to help deal with all aspects of this and act as a facilitator, translator and a knowledge expert. You just need to do these courses and convince whoever it is in the hospitals in the medical community this is a good role. The amount of people that are available is tremendous and the amount of yearning for new and different experiences is also tremendous as well. So it's the biggest untapped resources we've taught. We don't need to train zillion internists about this. We need to train people who are available in the medical system. So I think you could have a role in both and there needs to be some official buy-in from the entire system and as you know that's always complicated. The comment was there's a certification for genetics and nursing. I'm going to get that it's mostly involved in medical genetics as traditionally which is the Mendelian disorders now genetics is drifting into everything and that's really where you need it. So the cardiology practice should have someone who is a nurse who's part of the practice who is up on what the latest low-risk penetrance were because that practice is really going to see a familial hypochrestremia patient which is what the Mendelian people would see. They're going to see people who are drift onto one side and certainly the oncologist need this as well because they're not only as the I think that the advanced treatment based on what the tumor genome is like that's going to happen quite quickly but the susceptibilities is not being covered very well or even understood very well. We still hear cases of oncologists who've told women took a family history so that's a plus they actually took a family history and then found that all the breast cancer was on their father's side and said don't worry about it because it's all on your father's side so that's not that shouldn't be happening today and it still does.