CHERYL: So, what we-- I’m sorry, go ahead. JESSE: No, you're right. I was just gonna ask are there people now starting to sub-specialize in data science and going into it's almost like my significant other. She does some of this for a hospital like entering data and then also going through the data to look at certain things for doctors and nurses and all this kind of stuff. So, I am cursory aware and some of the processes that go on in this kind of field. But it seems like yeah, now you have all this data, now you need somebody or somebodies to sit down and say, okay, how do we make all of this data into like fields? So, it's not a bunch of disparate peer-reviewed papers, but it's data that it's like, this field appears in this way, paper and this paper and this paper so that you can start comparing it apples to apples instead of essentially a hoping that like your brain puts it all together in some useful information. So, are people coming out from Ph.D.s maybe similar to yours and saying, okay, I don't want to do the research, I want to figure out what the research is actually telling us? CHERYL: Absolutely. So, I work, I don't wanna say we're divided because we're not divided. But there's the sort of the wet lab aspect where people still involved in data generation, because we still need data to work with and ask questions. And science is mostly about asking questions and answering questions. And I think that some of the best scientists are individuals who know which questions to ask or can think of the best questions to ask and how to go about answering them. But then you also have what we call the dry lab people. So, the wet lab people, which is mostly me, I'm more of a wet lab person and you know, I spend a lot of my time moving small amounts of liquid from one tube to another. But you have the dry lab people who are mostly computational. And these are the people that are programmers, they know how to code. They're very computational, I think and also statisticians we need a lot of input. I work with some people who are much smarter than me who are experts in statistics. And when you know you're talking about very large datasets, you need those individuals. You need to know does this conclude-- is this conclusion valid? Do we have the power of the data to make these conclusions or predictions? And so I think, anyone who might be-- there's going to be-- there's no shortage of demand. I should say, ?? 02:49> it there's a lot of demand for computational people. So, if certainly, those individuals may be going into who are in college now or you know, thinking of going into that sort of field, really brushing up on your statistics and coding skills, invaluable, invaluable. And even my son now who's in high school, he's taking a coding class and I'm like, get a statistics class in there. You know, I just-- I did not have that when I was going through college, and even in the classes I took, excuse me, for my Ph.D. We just didn't have that big data then and there was not that same demand for high-level computing skills or statistics. And it's very much needed, there's a huge demand for people who, and it pays a higher salary a lot of times then people doing more wet lab work. So, absolutely, we really need a team effort nowadays to ask some of these big questions in science. JESSE: It definitely seems like-- it just, it kills me off the top of my head. I actually spoke to somebody who's a data scientist, and he's done several different things. But he is one of the only-- I’d have to like look up the episode right now, because I don't know why I'm blanking on his name. One of the only like 13 people to finish the Barkley Marathons. Are you familiar with that race? CHERYL: Yes, yes. Yes. Actually, I know someone who did that thing. He was a grad student at Penn State. I'm drawing a blank on JESSE: his name. JESSE: John Kelly. There we go. CHERYL: That wasn’t the person I was thinking of but I do personally know ?? 04:31> who did the Barkley Marathons. It's crazy. Did you see the documentary? JESSE: I did not. But John, I think John is in the documentary. And I had, I didn’t-- before I spoke to him, it’s Episode 31 for anybody that wants to listen to John's interview, but before I had-- I didn't have a chance to watch it before I had talked to him. But yeah, that's like that's what he's doing now and we talked about just when we were growing up, John and I are of a similar age, it wasn't even a thing. Then all a sudden, it’s become this very important thing because we have the computational power now to actually go through these giant data sets and things. I think he's working on like cybersecurity for like insurance firms or something, something very important. So-- CHERYL: ?? 05:17> actually funny enough, one of our, one of the people used to collaborate within the statistics department here at Penn State actually left to go work for a financial company in New York City. So, he was you know, if you have those skills, computational skills and the background, I mean, you can you know, really use that to your advantage in a lot of fields. JESSE: Yeah. But at the same time, I kind of give John credit for being in an important field, but I want to talk about the vision project because that seems like a very important field as well. And part of what, I know, you even said yourself a little bit self-deprecating, these I'll call them quant heads just to be loving are smarter than you. But it's not-- I don't believe that I just think that they have a different field of expertise. CHERYL: Oh, yeah. But that's certainly a nice way of looking at it anyway. Yeah. So, in this vision project, what we aim to do was to figure out ways to integrate a lot of different data from different labs and different sources and find a way to model the data. So, then we can-- what insights can we get out of that? And we used a predictive modeling program called Ideas. I'm not going to go into the details of that. But it uses some modeling method and so, it’s a-- and I'll give you an example. So, it looks at like kind of output of the data and tries to figure out what information might have led to that. So, let's say you looked at your activities over the course of a week and it was, well, I guess I-- most runners would run regardless of whether it's sunny or raining. But let's say, though, that you'd like to garden, right. And you gardened on days Monday, Tuesday, Wednesday, and Friday, but on the other days, you spent a lot of time cleaning. And you might then be able to figure out based on looking at what activities you did, what was the weather like on those days? Because most likely on the days with gardening it was not raining, right. So, some of the statistics and integrated modeling kind of looks at the-- some of the output of the data and tries to predict what sort of states if you will, or in this case, the analogy of the weather what actually might have been. So, that's kind of just like a general way of looking at how some of these modeling programs work. And what we were trying to do is identify different regions in the mouse genome that play a role in regulation of blood cell development. So, we collected a lot of different resources, ran some of these modeling programs and attempted to pull out or identify these different regions, which might be important. And then we are now in the process of validating or testing those predictions. And when I say that this can be applied to other fields, it's because really the-- it's sort of a proof of concept. Because, in theory, you could integrate many different kinds of data depending on the field, in use this sort of modeling program, and then identify important regions in the genome. So, when I say, different fields, I mean, I'm still talking actually about more like genomics, but you maybe work on genes that might be involved in some other medical process or vision, like literal vision, like eye vision, not our vision of an acronym, or maybe in some other disease process, like Alzheimer’s or whatever. So, you can kind of feed various data in through what we call kind of our pipeline, our integrated pipeline, get your predictions and then test them. So, by kind of going through this process, we're hoping to show that this can be applied again in different areas to ask these questions. And we've come up with a number of different resources, mostly for clinicians who may work with patients who have blood cell disorders, as well as other researchers. And we've made a lot of these tools available you know, on different-- on a website where they can go in and kind of mine the data and ask different questions. JESSE: Obviously, you're in the weeds right now. But I mean, is the idea that we eventually have the kind of predictive ability to figure out-- I think your background on the project mentioned precision medicine, basically trying to figure out, as I understand it, trying to figure out what treatment is gonna be most effective for each patient. I mean, as part of the goal to say once we know your particular genome, we can figure out dial in, in this case, treatment for you for a particular disease. CHERYL: That would be the goal. I mean, of course, it's going to be more difficult with complex disorders. So, when you're talking about monogenic diseases or diseases that are due to a mutation in one particular gene, that sort of therapy is going to be much easier and more straightforward than something like diabetes or obesity, which have multiple factors and lifestyle components to it. But as one example that may be kind of a good illustration is sickle cell disease. So, sickle cell disease result is from a mutation in a gene that encodes beta subunit of hemoglobin. So, in normal, healthy individuals, humans consist of four subunits. So, we have two alpha subunits and two beta subunits. And they allow, that's in the red blood cells, and that carries oxygen. So, if it's going to carry oxygen to all your tissues, and then give the oxygen to the tissues, and the problem is with the people-- with sickle cell disease is they have a mutation in the beta subunit, which then does not-- the hemoglobin isn't formed properly, and it causes a ?? 12:31> phenotype or appearance to red blood cells whether that looks like a moon, like a quarter moon. And that can be, that first of all, impairs, oxygenation of tissues, it can be very painful. And many, many individuals, particularly African-Americans, in this country, are disproportionately affected by sickle cell disease. It's a very-- a lot of very poor quality of life. And you know, a lot of the treatments today are mostly in the form of transfusions and so forth. But one potential treatment that's very exciting in the field right now has to do with some gene editing experiments. And the reason is, is because when you're still developing as a fetus, you have, instead of expressing alpha-globin, and beta-globin, you express alpha-globin and gamma-globin. So, it's very similar to beta-globin, it's just slightly different. And excuse me, certainly, one reason that there is that difference is when the mother is carrying the baby, the gamma-globin has a higher affinity for-- the hemoglobin has a higher affinity for oxygen than the beta-globin. So, that allows the fetus to get that oxygen right from the mother's blood. But once you're born, your expression of levels of gamma-globin drops dramatically, and instead of expressing gamma-globin, you express beta-globin. So, that’s the adult form, right. So, one of the treatments that's being explored right now for sickle cell patients is to find ways to turn on that gamma-globin gene. So, it's still there, it's still in your genome, it's just not being expressed. And one of the key players in keeping that gamma-globin repressed is a protein called BCL11A, and what that does is it’s called a repressor. And what I mean by that is it prevents expression of gamma-globin. And so one of the experiments if you're trying to edit the genes so that you can allow gamma-globin to be expressed. And proof of concept, I mean, there's been-- they've shown that this works in cell culture. And I think there may be some clinical trials with various gene editing in tech trials to try and increase gamma-globin expression. And you can get an increase in normal blood cells, in normal looking blood cells ?? 15:16> so better oxygenation and so forth. And, excuse me, one thing about blood cells is you have access to stem cells. You could, in theory, take out stem cells from a person's bone marrow, and do some gene editing and put them back in and perhaps have that population expand enough to produce enough normal cells with gamma-globin and alpha-globin to allow the person better oxygenation, better quality of life there. I mean, there are some problems with gene editing right now, and there are some off-target effects meaning it might not just edit where you want it to edit, but there might be other problems. And so I mean, there's still a lot that needs to be done, but I think we're getting much closer. And that's like a really, I think, pretty straightforward-- pretty straightforward disease to try and tackle using gene editing. So, we work with some collaborators at Children's Hospital, Philadelphia and St. Jude's Children's Hospital that are working on these sorts of projects. And we do a lot of DNA sequencing for them and some data analysis and it's pretty exciting, it’s a pretty exciting field. JESSE: This is-- I'm jumping forward in the research because this is just a pop culture head that's going on there. I'm like, okay, where are we going? So, it makes you wonder, so if the beta-globin is the issue and we turn on the gamma-globin a little bit to essentially resolve the oxygenation issue with sickle cell is the goal basically to repress beta-globin and switch it back to alpha and gamma entirely? Or is it simply having, like, a portion of your blood cells as alpha gamma, and then the other portion is the normal like adult expression of alpha and beta? CHERYL: Yeah, that's a good question. I don't think that the goal or would be to completely replace that. I think the idea though, is it's with this sort of treatment, you're hoping to provide enough modified or edited cells to overcome the clinical manifestations, right. So, I don't think you're ever going to be completely normal, like normal levels of hemoglobin, but can you meet a threshold. Now, I mean, where things go down the road ?? 17:54> happen, but that's really kind of, I think, where the current thoughts are. JESSE: So, it’s essentially just like a, the way I guess I think about it would be, if you could successfully do this in a person, I know from at least from the sound of it, we're ways off from like any kind of human clinical trial. But if you could successfully do this in a person, it's almost like a self-replicating therapy instead of like having to get transfusions to fix the issues with the blood. Am I on ?? 18:27> there? CHERYL: ?? 18:26>. I think there are some clinical trials working on...editing. I'm not involved with those, so I don't want to like comment on any particular details. I'm definitely more on the basic research side, but are our collaborators are working with-- some of our collaborators are working with some sickle cell patients or blood from various patients and human blood. So-- I’m sorry, I forgot the second part of your question. JESSE: No, no, it's okay. I think that pretty much got it. And we're running short on time. So, we could probably keep going because you've done a lot. But before we run out of time, so in the interviews you listen to I don't know, if you've got to the point where I had a question I asked everybody in season one, we're now in season two, because this is the second year I've been doing the show. So, this year, I'm asking everybody an opinion question. What do you think the purpose of sport is? CHERYL: That's a good question. I think for me, I think it's been key to help build confidence and resilience. And I think it's also helped me learn to manage my time and priorities. I think everyone has their own view or purpose of why they do it. But of course, there's obvious health benefits, right, from exercise and so forth. I ?? 20:04> too, I think it also, for me, meets a bit of a psychological need to because I like to push myself, I like to be competitive. And I'm 48 but I still aim for those overall places. I'm not ready to settle for too many age groups, but I've been forced to face my age a bit. But yeah, I think I'm gonna go with-- Yeah, I really like the, I think confidence and resilience and time and managing your time and your life. JESSE: It's just a question-- It's something I think about sometimes. I'm like, why don't I ask people I'm talking to, to see what they think. Just because it's--
Smart Athlete Podcast Ep. 39 - Dr. Cheryl Keller Capone - GENOME SEQUENCING - Part 3 of 3
So, what we-- I’m sorry, go ahead
No, you're right. I was just gonna ask are there people now starting to sub-specialize in data science and going into it's almost like my significant other.
