The Blueprint for Your Body: Genetics and Big Data | Intel Business
(upbeat music) - [Narrator] Welcome to What That Means with Camille. In this series, Camille asks top technical experts to explain in plain English commonly used terms in their field. Here is Camille Morhardt. - Hi, and welcome to today's in Technology episode of What That Means: Genetics.
I'm talking with Michael Snyder, who is chairman of the Department of Genetics at Stanford University. He's also director of the Center for Genomics and Personalized Medicine at Stanford University. We are going to talk about genetics and its intersection with compute including artificial intelligence, big data. We're gonna talk about longevity and anti-aging. We're gonna talk about implants and micro samples and ethics as when it comes to genetics and what's going on in that space.
Stay tuned. Welcome to the podcast, Mike. - Oh, it's great to be here. - So Michael, can you just level set us as we kick off this enormously complex conversation by defining what is genetics? - Yeah, genetics is from your DNA and it's really your instruction manual for making you a human being. But how that instruction manual plays out counts on environmental factors and lifestyle. The food you eat, exercise, all that will impact your health.
So again, genetics is just the blueprint and then everything else actually does play into what makes you you. - Right, makes sense. There's the blueprint but then there's actually the building of the blueprint.
Can you tell us what is most exciting to you in this field right now? - For me it's big data. We can really change the way healthcare is done. We can now sequence people's DNA and predict your genetic risk for disease that was not possible even 10 years ago.
So we can can do that. We can also make lots and lots of measurements on you to see what's going on with your health at a level that's never been possible. So backing up a little bit, I think the healthcare system's broken. If you think about, it's very bizarre. It's mostly sick care. We treat people when they're sick rather than keep them healthy.
We actually take very few measurements when they go to a doctor's office and in fact the whole way we practice it is kind of crazy. People have to get in the car, drive there, show up in an office that looks pretty much the same as it's looked for the last 40 years. They draw a large amount of blood with a needle that hurts and then make very few measurements from that blood and then they'll make a decision about your health based on population based measurements. We think all of these things can change and that's what we've been trying to do. As examples,
we'll talk about this. I'm sure some of the remote monitoring and remote sampling we think will be very powerful. But I think on this whole idea about population average is really a, well to be honest, neanderthal idea. We should really be working on people based on individual measurements, meaning we can see exactly what your health profile looks like and detect changes presymptomatically before any symptoms appear. And that's an area we're very excited about.
We can bring in lots and lots of data to better manage your health. - I remember listening to one of the things that you said in a talk you gave that of the top 10 grossing drugs from pharmaceuticals in the United States. Only one in 10 or one in 24 actually work on a patient. - Yeah, meaning most of them don't work. That's absolutely right.
Crazy. So most people are taking drugs that have no chance of working and we don't really follow that. We do it all by trial and error and hope that give someone a drug come back two months later. Is it working? Is it not working? Well, I think so. I'm not so sure.
So we really need to get much more precise about it. That's just from the treatment side. But even catching disease, again, this is an area that we work hard on to try and keep people healthy catching disease before the symptoms appear. It's totally doable, but we don't do it. And how can we do it? Well, we use big data. So we're actually sequencing people's genomes to predict disease risk will make measurements from their blood.
There are new technologies out there that you can measure literally thousands and thousands of molecules from people's blood, meaning all their RNA and proteins and metabolites and lipids. These are important molecules in their blood and we measure them outta urine as well. We even measure something called the microbiome, which is really important for your health.
You may or may not know you have more cells in and on you that aren't you than are you, meaning a human being is 37 trillion cells. But we have about a hundred trillion microbes living in and on us, mostly in our gut that digests our food and makes essential vitamins, things like that. And we can measure that. And so we're in a world where you can now make measurements on people and catch disease before symptoms appear.
It turns out just from this proof of principle we've been doing as a research study, we, we actually have been following 109 people and in the first three and a half years we've been following for 10 years. But in the first three and a half years, nearly half, 49, had major health discoveries. Meaning we caught someone with early lymphoma, two people with pre-cancerous, two people with serious heart issues and we caught them all presymptomatically, meaning we would see shifts in their profiles that said something wasn't right. And then that indicator basically led the follow ups that caught the disease early. And then in some cases we think totally avoided these people from getting any serious complications like the lymphoma was caught so early, that person was treated and they've been disease free ever since we caught that.
It's like a jigsaw puzzle. So the way we do practice health now, they'll measure maybe five pieces and we are trying to take more like five or 600 pieces out of a thousand piece jigsaw puzzle so we get a much clearer picture of your health and we can catch these things presymptomatically and we've gone on to spinoff companies to try and scale it to get it out to the world. And then to your earlier point, we also think this is gonna be powerful for treating disease. We can see quickly how people are responding to treatments and not wait months and months.
We can see this high resolution measurement if you will, exactly how they're responding. So we think these technologies are gonna be very, very powerful. - So you're talking about a lot of different data collection methods. You're talking about genome sequencing, you're talking about microbiome sampling, you're talking about wearables. You probably have more than just the one on your wrist. - Yeah, I'm wearing four right now and believe it or not, these hearing aids, they're for hearing but they actually do measurements as well.
They'll measure my interactions throughout the day. They can also measure activity and things like that. Physiological measurements.
And I'm wearing, yeah, here's a glucose monitor. Whoops, can you see that? There we go yep. That's a continuous glucose monitor. So I use about eight of these devices every day. - So are you looking forward to a future where these kinds of devices can exist inside your body as opposed to something that you have to wear? Or does that kind of a future scare you? - I look for a future where there'll be one device instead of me wearing, you know, I'd use eight or nine of these things every day.
But I like the idea of them being inside of me. Quite frankly, that scares a lot of people 'cause of the privacy side. But I think for me, I like it because I think you get better measurements. I think that'll be valuable for monitoring health. - So obviously we already have some computers that exist inside of our bodies.
We have pacemakers and insulin pumps. But generally speaking, you're talking much smaller scale. So can you help us understand with the speed, like a chip that's implanted in a specific place in your body or would it be swimming through your bloodstream? Like how would that work? - Well, I guess what I would envision, they'd probably be chips in a fixed location for the most part. So you get very consistent measurement.
And what do we wanna measure? Well I think we wanna measure standard physiology, heart rate, heart rate variability, your temperature, blood oxygen. These are things you can measure right now with a smartwatch, but they would probably get more accurate inside of you. But I think measuring your glucose, your cortisol, cortisol by the way is involved in stress and things like that. So that would be useful. There's a lot of other markers that might be specifically useful for people who might have issues like some inflammation markers, they're called cytokines. Immune markers could be very, very valuable for certain kinds of inflammation.
And I also think, imagine people on drugs. We met talked before about taking drugs, you can measure people's levels of drugs and people metabolize drugs very differently. One person to the next and you could actually follow that with implantables that would be inside you. They'd have to be very specifically designed to do that. - So you're talking about combining a whole bunch of different pieces of information to try to identify which drug might match make with which person.
And I'm wondering what is the end goal here? Is it just that you would tell somebody in advance, okay this particular drug is not gonna work for you so don't bother taking it? Or how are we gonna now design drugs differently on a more personal scale? - Yeah, good question. I think ultimately we want to get to both taking the right drug for the right people. And so we don't know, most conditions that people have are complex and they're probably not one simple solution. So for example, depression, it's probably induced by many different things. That's true of most mental health. Diabetes is a good example.
There's at least five subtypes of type two diabetes. The different drugs work on the different subtypes. You don't really have one drug that works on all. So I think getting the right drug to the right people.
So what we need to do first, and I think big data will be powerful for this, is that as we take data on people, we better understand what people will respond to what drugs we should be able to get predictive markers of that. Meaning we can tell for example from a continuous glucose monitor which follows people's glucose levels after something they eat. We believe you'll be able to tell what kind of problem of type two diabetes you have, what subtype you have and what drug you should take. And I think that's gonna be true for all these what are called complex diseases. Whether it's mental health like depression, bipolar, what have you, or whether it's metabolic diseases like type two diabetes or other things. And if we can get the right drugs.
So we should be able to make profiles and make predictions on them what drugs they'll respond to. And then we can monitor the response. Now what we're not very good at in the mental health space, we don't have good biomarkers for complex disease like for depression and things like that. It's generally run by surveys and how people feel.
And it's not a good way to measure people. - I feel like I just need to play devil's advocate and ask the cynical question of do we have a mismatch of incentives? Because if you know that, let's say you're the number 10 top grossing drug in the US and it's only working on one out of every 24 people, then if I've suddenly figured out that a 23 of those people are better off not taking the drug, have I now just dropped my income level on that? Is it better for me if people don't know whether or not it's gonna work? Or am I looking at this incorrectly? - Well, when you're first drug to market, what you wanna do is of course give it to everybody. And so from the pharmaceutical standpoint, it's sure it's in their best interest to give it to everyone and not know who will respond and not respond. The world's becoming more sophisticated and people do want to give the drugs.
Some of these drugs are really real expensive. So from the insurance company side, from the provider side, you actually do wanna know who's gonna respond. And I think we're gonna get better and better biomarkers, whether it be genetics, whether it be some of these molecular markers that we measure outta blood or digital markers from your smartwatch, what have you. I think we're gonna get better at predicting who will respond to what. And I'm a good example of this.
So I'm type two diabetic, you may or may not know. And it was predicted from my genome and I initially got it under control through a lifestyle change. But it came back and I couldn't really change my lifestyle. I did some things that improved it, but I never got it down to fully under control. So what I wound up doing was taking the obvious rontline drug Metformin and guess what? I'm a non-responder.
I also did some things like I lifted weights to improve muscle mass with the idea that would help my diabetes. That actually did not work either. And in hindsight I know why it didn't work. It turns out my cells respond to insulin, which is what choose, normally controlled diabetes and I make insulin, I just don't release it from the pancreas. We use big data to figure that out and it turns out then at the end there's a certain drug for that that helps release insulin from my pancreas. And had I known that from the get-go, it could have gone on that right from the start and been effective instead of spending a year and a half taking drugs that never had a chance.
So I think this is how the world's gonna move. If we can make the right measurements on people, we can give them the right drug so they get the right response at the right time. Yeah, be much more effective.
- So I wanna talk about longevity. I would be shocked if somebody hasn't heard about age reversing or longevity. You're the perfect person to ask. I wanna understand a lot of the work that you're doing seems to me like its goal is early detection or prevention by knowing what you're up against from a genetic perspective.
And I'm wondering your thoughts on actually reverse aging. Do you think that this is possible or will be possible? - I actually do believe you could reverse aging and I think the trickiest part's gonna be the brain, but I think the other parts of the body as we learn how to rejuvenate stem cells and replace organ parts and some of it may be mechanical, right? We do have pacemakers and hearts and things like that. So I do think most organs, we will be able to keep going and improve. The one area that's gonna be tricky is the brain. But I actually think even that we should be able to rejuvenate and keep going.
So I'm one of the few people who believes we should be able to reverse aging or at least prevent aging. And so I do think people right now a lot of people say, well 120 is the limit, that's about what people hit and then they'll die. But I actually believe it could go on forever. That creates all kinds of social implications as you might imagine. And we can talk about that if you like.
- I do wanna get into the social and ethical implications, but before that I wanna understand from you, how do you see the slowing down or reversal of aging actually happening? How does it work? - We can actually see how people are aging. People age very differently. It turns out some people are cardio agers, other people are kidney agers. I myself am a metabolic ager.
And when you see what's going on, you can actually then design strategies to try and mitigate that. Certainly reduction of aging through antioxidants and things that we now know a lot of the hallmarks of aging and some of that I think we do know how to slow down, anti-inflammatories, antioxidants, things like that. I think if you really want to reverse aging, you've probably, they're these things called senolytics. These are drugs that kill your senescence cells, which are associated with aging. And I do think they're probably part of the equation as well as rejuvenating. So your stem cells can replace organs.
A good example is your liver. You may or may not know that your liver cells will divide once a year on average. And so they gradually regenerate your liver and you have actually some of these cells in all parts of your body, including your brain. So they have the potential to actually replace some of these things. They're just very few of them and they're kind of quiescent for the most part that may be one avenue for replacing them.
Other people are working on ways of trying to take cells that aren't stem cells that are already differentiated and reverse them. There are tricks out there that people are using genetic tricks and other means to try and actually reverse cells from their, what's called a more differentiated site. Say a cell that is a liver cell, turn it into one of these stem cells, actually let it keep dividing and replenish your organs. So I think there's some very fascinating technologies out there. It's obviously early days and this is, I wouldn't recommend any of the listeners go out and grab one of these experimental treatments 'cause they think safety has not been worked out at all. And the biggest concern when you're sort of rejuvenating stem cells and things like that is, are you gonna cause cancer? Are you gonna make cells that grow outta control and then that cause cancer? That's not good for you either.
So I feel like genetics is really exploding for understanding the genetic basis of disease. It's not all genetics. I think we do need to bring in lifestyle and environmental factors and get personalized predictive models. And so one area we didn't cover so much is AI. So I think artificial intelligence is gonna be very, very powerful for pulling in all these different kinds of data, your genetics, your environmental exposures, your other aspects, and then build predictive models about what's likely to cause disease. And we can even do this on a personalized fashion.
So for example, me, I follow myself a lot as you could tell with all my smart watches. So I can actually see what's going on with my various health measurements and I can make predictive models about when things are going to go off the rails when they start happening pretty early. And I think that's gonna be very mainstream. We'll be able to build personalized models to see, uh-oh, you're heading this way. Kind of like your car is, you know it's wearing out, you'll see different things happening and I think this is gonna be powerful.
So we'll need all kinds of tricks to make this to happen. But the field is moving very, very quickly. It may be too late for me, I may not be able to live forever and that's okay, but I don't think it's too late for people born today.
Everybody says everybody born today is gonna live to at least be a hundred. And I think many of us think that they may live to forever possibly. - I'm sure we can all brainstorm all kinds of social and ethical implications to living forever or living indefinitely. And I imagine some of them might be similar to the cloning conversations that people have had. But rather than just ask your opinion, I wonder if you can offer some insight into where should humanity begin to even address these questions or establish some sort of standards or regulations or frameworks or thinking around them? - Well that's a good question. I don't know that there are any laws or guidelines to stop people from doing that.
So I think right now it's totally open. I think anyone who can get themselves to live forever probably could. I don't think you would do it at the expense of anyone else.
On the social side, what do we do? The planet, you know, people will obviously keep reproducing. We will have more and more people. The planet only holds so much resources and I think we will need alternatives, meaning going to other planets, other things to be able to expand. And I mean we're capable of increasing the capacity of this planet up to a point.
But there is still is a question of how much can Earth hold for the number of people. And that number goes up over time. I think what may be space exploration and using other planets that you probably know, NASA's already talking about trying to put people on Mars by 2030 and some of the private programs wanna do it well in advance of that.
Obviously that's not the same as populating the planet, but that's a first step. So I could envision us moving out. I think we would have to.
I think people are gonna live extremely long lives. We're gonna have, there's all kinds of implications of how to deal with that. And again, it's not gonna be terribly useful.
People live long lives but still hit dementia at age 80. That's not going to be a very productive society. It's gonna create all kinds of issues that we'll have to deal with.
So we really have to extend all aspects of people's health span, including their mental health. - What kinds of things do you think regular people should know about genetics and some of the advancements in the field right now? I'm not talking about medical students or people in biology departments, but what should everybody be aware of? - Well, I think people should be more attuned to their health. I think the idea that people get more attuned to their health when problems arise, especially when they hit the fifties or later. I think we can start that process much, much earlier.
Get people attuned to their health right from the get-go. You may or may not know, we can tell when people are getting a viral infection in advance of symptoms. So we can tell for example, if you're getting COVID on average three days before a symptom's onset from a simple smartwatch. So one of these watches we have a alerting system, your heart rate jumps up, we can follow that.
So we think health monitoring will take that kind of stage if you will. And the analogy I like to use is a car. So our car has lots of sensors, race cars have 400 sensors on them and they relay all this information back to a dashboard that you see and you use that to manage the health of your car. If something goes off check engine light goes on. And I think that's what's gonna happen with human health monitoring. I think we can have these devices that are passively monitoring our health or we'll do checkup, so to speak from home.
We can talk about micro sampling if you want, where you do a little prick of blood, mail it in kind of like Amazon for health. You would mail it in, you'd get back detailed report of your metabolic and other health and then that would say, whoa, something might be off. And we don't do that today. I think this is what's called longitudinal monitoring. Following people over time is very, very powerful for following shifts in people's health.
And that's because we're all different. We all have different baselines. Like you may not realize this, but we all have different body temperatures, meaning we're not all at 98.6, some people are 94.6. And for them to actually get, see if they're going to get ill, they can see that just if they go up two degrees, they'll be able to see it with a smartwatch. But a physician today will never spot that because they'll measure them at 98.6.
So we need, I think where I'm going with this is we're trying to understand people's baselines with this frequent monitoring and then see these shifts and I think that's gonna be a paradigm change in keeping people healthy and living long, healthy lives. - Can you tell us a little bit more about micro sampling and how that works in personalized medicine? I had read a paper that you were part of about weight loss and taking microbiome samples to understand how people are digesting either carbohydrates or fats. Give us a little bit more insight into that.
- What you're referring to is a study we had out recently that showed everybody's different. That is to say, some people who go on low carb diet respond better than those who go on low fat. They're both healthy diets, one's healthy, low carb, one's healthy low fat. And it's very personal.
We can actually predict, it goes back to what we talked about earlier. We can predict from some early markers who will respond and who won't respond to these different kinds of diets. And again, it's because we're all different. We have a very cool study we just did with this micro sampling that you referred to where again, you take small drops of blood, you mail it in and we do a detailed analysis. We can measure 2,200 molecules in that blood, and some of them are very, very important health molecules.
The one things we discovered is we had people drink a very simple shake, it's called an ensure shake. You may have seen it in CVS or your grocery store. And what'll happen is we had 32 people drink this and everybody responded differently.
Some people their carbs plummeted, others went up, some people's inflammatory markers went way down, others went up to the exact same shake. So we all respond very differently. And that makes sense. People when they eat different foods, some people get indigestion, other people don't. And we have different biologies. It's probably a lot of it's our immune system and our microbiome we referred to earlier that actually probably play into this.
And so I think we can actually figure this out by, first of all making predictions, knowing what molecular markers will predict these things. And the other is by measuring by this kind of micro sampling and other means, we can tell how you're responding to that piece of bread. You just ate at a level that's never been possible. And so I think this is really gonna be great for managing people's diet and lifestyles.
- I don't wanna let you go without actually having you define and describe how CRISPR works. I think we've all heard of it, but I'd love to have a true geneticist give us that explanation. - So CRISPR is a new method, if you will, for editing DNA and it's obviously being used a lot in the research setting. It's being used, you know, to change DNA to see how genes work and this sort of thing. And it's just starting to play into human. And a good example where there's some trials is something called sickle cell anemia where folks have a mutation in one of their blood genes called hemoglobin.
And you can actually make a change there that improves, it keeps their hemoglobin up if you will, keeps it functional. And so there are now trials to do this where you can make these changes and some of the results look promising. And so the hope is that you can change these severe genetic diseases with this and it's gonna be easiest to do on blood diseases initially because we have ways of getting your blood cells out changing it and putting 'em back in. They'll go into your bone marrow and they'll repopulate your blood system, if you will, your immune system and and your blood cells.
So I think the blood's with this CRISPR technology for correcting these kinds of changes, if you will, they're generally errors or so it's thought in your DNA will be powerful. It'll be harder to do it in non-blood cells, but people are working on that. There's some tricks out there. There are ways of adding genes to people that don't involve CRISPR. People are working on for things like muscular dystrophy and some eye diseases, things like that. There's now ways of adding genes to cells in your eye to help correct vision.
It's pretty powerful. What scares a lot of people is, well what about behavioral genes and are we gonna try and make people smarter and all these sorts of things? And that's still a ways off. We don't first of all understand what makes people smart and all that stuff at any reasonable level. Still a long ways from all that.
But I think for correcting these severe diseases that's starting to happen now, especially on the blood diseases. - So once you may make a change in the DNA via something like CRISPR, how long does it take to roll out the benefits of that change? Like for example, if it takes your liver an entire year to replicate itself, would it take an entire year to see that benefit or do some things happen faster? - No, for the blood one, it could all probably be done in a few weeks to a few months, depends on the particular situation. So that can move very, very quickly because again, you can pull these blood cells out, they work pretty quickly and they divide reasonably fast. Now most cells in your body don't divide that quickly and replace very fast.
So you can't really do that kind of manipulation. Liver is one of the few actually where you do get cells dividing pretty slowly about once a year. So that may be possible too. It's not gonna be easy to fix cells and muscle and things like that, but it's easy to add genes to cells or easier to add genes than it is to try and make those precise changes.
Does that make sense? Is getting a little technical. - Yeah, that makes sense. I can see that it would be easier to add something if you're missing it, rather than to go in and actually alter something within your DNA.
- Yeah, unless you can pull cells out and put 'em back. It's hard for making these precise changes. But we will be in a world where you can fix a lot of genetic mutations. You know the ones that are gonna be particularly trickier brain, for example, like fixing Parkinson's. How do you do that? That's not gonna be easy. And some of the genetic changes there, they're kind of dominant changes, meaning you can't just add a new gene and get the desired result.
There's a problem there that's kind of dominant and it's hard to get rid of. So there in those situations you probably need drugs, you will need CRISPR but that's gonna be very, very hard to do in the brain. - What are we missing? What topic have we failed to cover in this conversation? I don't wanna miss something gigantic in genetics. - Well, I think we're getting better at understanding complex disease and I hope we'll get better at predicting who's at risk for them. There are some very debilitating diseases, things like ALS, muscular dystrophy, things like that als, it's a complex disease, meaning probably many genes are involved and we can actually go in now and and have a good idea about many of those genes that are involved. And I think on the verge of getting predicted models, who's likely to get the disease and who's not, which is not well done now.
And I think if we can start there and then start understanding then how to prevent that. So first we have to understand the basis of disease and then I think we can set up preventative strategies. I think that could be very, very powerful.
- Michael Snyder, thank you so much for joining today. Again, Chair of the Department of Genetics at Stanford University and Director of the Center for Genomics and Personalized Medicine. Really appreciate your time. - Oh, it's been my pleasure. - [Narrator 1] Never miss an episode of "What That Means" with Camille by following us here on YouTube or search for InTechnology, wherever you get your podcasts. - [Narrator 2] The views and opinions expressed are those of the guests and author and do not necessarily reflect the official policy or position of Intel Corporation.]
(upbeat music)
2023-05-29 06:12
