 nice to be able to present to the CLSA participants. I want to start off by expressing my gratitude for all of the work that has been done by CLSA participants to participate in this cohort. And I have been working as a researcher interested in the genetic determinants of disease and I have been working with CLSA data for at least about the last eight years. And I'm just going to share with you some of the insights that we've been able to make and try to provide this into a context which may be relevant to most of the participants on this call. So I'm going to start off by maybe bringing some of you back to 1978 when in June 30 boys went on a canoe trip in Lake Temiskaming and 12 of them died. And one of the adults on the trip also died. And this was a tragedy at the time in Canada and in fact was brought to a provincial commission. And within this context a thorough examination as to the causes of this tragedy was brought to fore. And when people think about tragedies they often think well this person has served and left into my lane in front of me or this critical decision was made that led to the death of these 12 boys. But what the commission actually found when they examined the causes of this tragedy is that there was no actual single cause of this tragedy. Rather, it was a confluence of risk factors. And that was that these were inexperienced paddlers. They were in a moderate wind. There were moderate waves. Notice that these were not severe waves. These were moderate waves. They were paddling for four hours so not a long period of time. They didn't sleep very well. There was cold water. There was no swim test and there was only one adult. And this confluence of risk factors led to a tragedy which actually changed the shape of outdoor education in Ontario and across Canada and changed how people think about participating in outdoor activities. Because it became apparent that this was not just unique to this tragedy but actually was a common theme in many tragedies across Canada. And I'm going to share with you how a similar analogy can be made to actually risk of disease. And I'm going to start off by talking about breast cancer. I'm going to talk about it because it's common. Almost everybody on the call will have known somebody in their life who has been impacted by breast cancer or maybe been a breast cancer survivor themselves. And when we think about what causes breast cancer, a lot of people tend to think about a specific cause. And in fact, there's a very well-known example of a woman who had a single genetic cause which led to her having a double mastectomy to prevent occurrence of breast cancer. But when we think about most people who have breast cancer, we find that they actually do not have a single genetic cause. But actually they have a confluence of many causes. And so I'm going to take you through a little bit of a story today to tell you a little bit about how we have begun to think more and more about the major causes of diseases in our population and how these are actually a confluence and mix of causes. Because most aging related diseases caused by hundreds of little nudges. And some of those little nudges are the things that we eat, the number of times we go for a run, the number of cigarettes that we smoke. But a lot of that risk, and this varies by disease, but a lot of that risk is actually inherited as little tiny nudges from our parents in the form of our genome. And what I'm going to do is take you through a graphical demonstration of how we've been able to try to discern these little nudges over time and how we peer into the genome to be able to do so and what this means to you and your families. So this is a map of the human genome. Each of these vertical lines is actually a chromosome. And this is the state of knowledge in 2005 where each one of these little pins represents a region of the human genome that was reproducibly associated with the risk of disease. And you can risk of common disease. And you can see that back in 2005, which was around the time CLSA was getting started, we knew very, very little. And then if we fast forward to 2020, excuse me, 2010, this map becomes much more dense. We start to see many more pins show up, where we're reproducibly associating different regions of the genome with risk of common diseases. And if we move to 2020, this becomes very, very, very full. And in fact, I'm actually only able to show you some of the chromosomes, because if I had to show you all of them, they wouldn't fit into one slide. And so how has this happened? Well, it's happened via the work that people like you have done, which is sharing your DNA as well as medical information so that we can map the regions of the genome that cause common diseases. And we have been able to understand that just like in the case of the late Tomiskaming tragedy, as well as most causes of breast cancer, that our risk of disease is predominantly conferred from a genetic perspective by tiny little nudges from many thousands of genetic variants that impact our risk. And now, with that as an understanding, I'm going to move forward to try to explain to you how we can use this information to help our patients. And I'm going to frame this in three different categories. The first is how we can use this information for disease prediction, how we can use it for improving diagnoses. And the third, how we can use this information to identify causes of disease. So let's get started and talk about disease prediction. Okay, so we'll focus on breast cancer. And this is a very interesting slide that was published a few years ago now, and things have actually improved since then in our ability to be able to discern those at risk of breast cancer. And what the horizontal axis shows here is the age of a woman, and what on the y-axis shows here is their absolute risk of getting breast cancer. Most people do not contract breast cancer over their lifetime. But what these different colored lines do is they summarize the information across the genome and what's called a polygenic rescor to be able to put people on different tracks for different absolute risks of having breast cancer. And you can see that if you're in the 99th percentile of risk, so it'd be a very, very high risk and your risk would be higher than 99% of the population, that your risk of breast cancer starts to become appreciably high even in your late 30s and becomes very high by the age of 80. Whereas if you're in the bottom one percentile of risk, this action, your risk actually never becomes appreciably high, where in most jurisdictions would suggest a mammogram until quite later in your life. And most people, of course, are binded to different groups depending upon their risk. But what we can actually now do is quantify this risk through something called a polygenic risk store, which is just a fancy way of counting all of those little nudges that we have in their genome and seeing who is lucky enough to have very few and who is unlucky enough to have many. And so you can easily think about this as a shared cumulative risk that that person has just by collecting all of those little nudges that moves them along this risk distribution. And what we can start to think about doing for these people is instead of just identifying people like Angelina Jolie, who are at risk for a single mutation, which increases importantly the risk of breast cancer, but identifying the very, very large proportion of the population in compared to those with a single gene mutation, who actually have a similar or higher risk of disease, but is much more common in the population because they happen to be in the top 5% or the top 1% of the risk accumulation. And so I'll next talk about how we can use this information to try to improve diagnoses. And I'm going to shift to using genetic insights in the case of diabetes. I'm going to show you an example of how we can actually use the same type of polygenic risk score to differentiate between type 1 and type 2 diabetes. And why is this a problem? Well, many of you may know someone who has diabetes and you probably know more people who have type 2 diabetes. Type 2 diabetes is a disease that is often later onset in age and is often more driven by obesity. But about half of the people who develop type 1 diabetes actually develop it after the age of 18. And so I'm an endocrinologist. I treat people with diabetes all the time. And very early in their disease, it's very difficult to try to understand which type of diabetes they have. But interestingly, these two different diseases, type 1 and type 2 diabetes, have very different genetic risk factors. And so we can actually quantify a person's risk of the disease, both type 1 and type 2 diabetes, and see if they are someone who more clearly has genetic predisposition to one type of diabetes over the other. And doing so can really help us to rationalize when the patient should begin on insulin therapy. And so here is a slide where researchers have done this. They have done this by looking at two genetic risk profiles, one for people with type 1 diabetes, and the other for type 2 diabetes. And you can see that you can spread out these two populations quite nicely where some people overlap, but the majority you can actually paint them into more probably a diagnosis of type 1 diabetes or more probably a diagnosis of type 2 diabetes. And this information I'm actually rolling out into our clinic right now at McGill University in Montreal to help clinicians and their patients decide which type of diabetes they have so that they can receive the appropriate therapies faster. The third thing that I promised to tell you about was identifying causes of disease, and this is a way that we can use human genetic information to be able to disentangle a lot of the causes and consequences of disease using some simple concepts that I'll describe to you right now. So oftentimes we're trying to understand whether or not a risk factor causes a disease, and here I'm going to switch gears to talk about osteoporosis. I'm actually going to use data from you to be able to show you these insights. We can measure a risk factor for a disease and we can measure these in the bloods. And one thing that we have recently measured in approximately 10,000 people in CLSA is 1,000 different metabolites. And so some people on this call, your blood has been surveyed for 1,000 circulating metabolites. And we're interested if any of these cause aging-related diseases, and I'm just going to show you the example of osteoporosis. So when we do that, when we measure these metabolites and test if they're associated with osteoporosis, we should be really worried about confounders. And confounders could be things like body mass index, so having a high body mass index could influence on the tablite and could increase or decrease your risk of osteoporosis. And if we don't understand this relationship and this potential confounding factor, then we can draw spurious conclusions and doing so actually plagues a lot of our understanding of cause and consequences of disease. And so using the human genetic data that we've generated from your DNA, we can actually disentangle this problem by identifying genetic variation. So some people have one flavor of a piece of DNA and other people have a different flavor of a piece of DNA that strongly associates with metabolites. And what we can then do is using the biological fact that these genetic variants are randomized in the population at conception. So whether or not you got a specific genetic variant or your brother or sister got a specific genetic variant is essentially a random process. And that random process breaks association with confounding factors. And doing so allows us to be able to test via these genetic variants their effect upon disease freeing ourselves of these confounding factors. And so we've done this recently using data generously contributed by participants on this call. And we were able to estimate the causal effects of 1000 metabolites on bone density, which many of you will know is the most important risk factor for osteoporosis. And we found that some of the metabolites had clear effects upon bone marrow density, such as this metabolite here, which I'll name, which is called orotate, which is found that individuals who had higher levels of orotate had much lower bone marrow density. And importantly, we can make a statement that orotate influences bone marrow density causally, and is unlikely to be biased by confounding. And this information can be very, very helpful for therapeutic development, because we can actually identify targets that are causal in humans, which is a difficult thing to do across the entire biomedical enterprise of information, because it's very rare that we can actually make causal insights in humans. And therefore, being able to target something that actually causes the disease is much more effective than something that is actually just associated with the disease, or something that is actually caused by the disease. And so this is an example of how we've used your data to be able to gain causal insights in disease, providing us a platform for being able to undertake therapeutic interventions that intervene upon these metabolites and decrease risk of disease or its consequences. So I'll just summarize here. The first point that I'd like to make is that rapid advances in measuring millions of small influence on disease risk has allowed us to better identify individuals at risk of disease, clarify diagnoses, and help to identify causes of disease, which can be used as targets for therapeutic interventions. This is only possible through large-scale collections of data from humans such as you, the participants of CLSA, and for which we are very grateful. I'll close by just reminding you that, in fact, most tragedies that affect our lives are due to a confluence of causes. And this is something that we're beginning to be able to quantify quite clearly using large-scale human collections, such as those afforded by DNA, and by the DNA in CLSA. And I look forward to answering any questions that you have in discussing in the Q&A session. So thank you for your time.