9. Creating a digital version of you and all your genes, aka "digital twin", to better predict and treat diseases.

Creating a digital version of you and all your genes, aka "digital twin", to better predict and treat diseases. === [00:00:00] Sajung Yun: Five years from now, I predict there will be two types of patients. One who build their digital twin, another one who did not build digital twin. For the ones who did not build their digital twins physicians gonna say: this drug is known to cause this serious side effect. You might, have a very serious side effect. Basically they are testing the drug to themselves. But those who build their digital twins, we, we can simulate the drugs effect before they take the drugs. [00:00:48] Mizter Rad: How are you? [00:00:49] Sajung Yun: Good. Yeah, how are you? [00:00:52] Mizter Rad: Good, good, good. After some technical trouble. Always happens. We live on the 21st century and it's technology is so advanced and sometimes it's just, uh. It gives you trouble, let's say. [00:01:05] Sajung Yun: Yeah. No problem. [00:01:07] Mizter Rad: Cool. Where are you based right now actually? [00:01:09] Sajung Yun: Oh, actually, our company is based in The headquarter is in Silicon Valley. Mountain View, California. But I teach at Johns Hopkins University as a young professor for Data Department of Bioformatics. At the Center for Biotechnology Education at its graduate school for the last seven years. So I stay at Maryland Baltimore area. Actually Bethesda. And our company got the government funding in Arab Emirates. So I also go back and forth and Abu Dhabi. And currently I'm okay in Bangkok actually. [00:01:43] Mizter Rad: Oh, you're in Bangkok right now. So what time is it there? [00:01:46] Sajung Yun: It's 1:00 AM Yeah. [00:01:48] Mizter Rad: Whoa. Okay. Okay. Okay. You had a long day? [00:01:51] Sajung Yun: Yeah. It was a long day. Yeah. Good long day. [00:01:54] Mizter Rad: Okay. What are you doing in Bangkok? Are you working there or is it just holidays? [00:01:58] Sajung Yun: Oh yeah. I'm working there. Trying to make business partnerships. [00:02:01] Mizter Rad: Okay, cool. ,Sajung. Am I pronouncing right your name Sajung? [00:02:06] Sajung Yun: Yes. Yes, that's correct. [00:02:08] Mizter Rad: Cool. [00:02:08] Sajung Yun: How about your name? How do I pronounce your name? [00:02:11] Mizter Rad: Well, you can call me Rad. You can call me Rad. Or mr. Rad? Rad. [00:02:16] Sajung Yun: Oh, Mr. Rad. Okay. Rad. Yeah. . [00:02:19] Mizter Rad: Cool. Sajung, my man. Let's talk about predictive care, huh? Yeah, predictive care. That's the name of your company. How did you come up with your idea? [00:02:27] Sajung Yun: There was a beginning about seven years ago. My father got diagnosed with rare cancer, by the way, he's okay. But but because of this rare cancer he had to resect take out one third of his stomach cut out. And then, you know, my twin brother and I analyzed thousands of people's DNA before as an expert in this field, but not our parents. So we felt kind of bad. So we decided to analyze our parents, including my father DNA. After he got diagnosed with, uh, cancer going through all of this and and found out his cancer was actually predictable and preventable. So if, we were the expert even 10 years ago, and if he have done that, he didn't need to go through all this surgery and it could have been prevented. But that was not the case. So we felt so bad. So next thing we thought was, okay, why don't we analyze other serious disease so that we can prevent other serious disease from happening? And then we did it from the, you know, one disease and like Alzheimer's, diabetes, other colon cancer, et cetera. So we went all the way to 250 serious diseases. And one day my twin brother came to my place and pointing out to the windows and saying, you know, preventive medicine is like, uh, putting the screen door to keep the bugs out. But we install the screen door only for 250 cases out of tens of thousands of diseases, which is less than 10% of the window. And then we are hoping the bugs don't come in. So we basically install the 10% of the Windows screen door, and we are hoping the bug, which is disease, doesn't come in which doesn't make sense, right? So he's suggested, oh, let's do it all over again. And which is totally new way. And then it took after one year , and then it took another two and a half years for us. By the way, it is my twin brother Sijung is also faculty at John Hopkins University teaching Bioformatics and also he's a faculty at NIH, National Institute of Health FAES graduate school to teach the next generation sequencing. So he trained more than thousands of post thousand postdocs in this field. And so we decided... [00:04:49] Mizter Rad: I know you have a quite impressive team actually. Yeah. I was trying to look into teams or people in the world with the credentials and experience to talk about DNA based digital twins. And I stumbled upon you in Dubai and that's why I wanted to have you on this conversation cuz there's no one better in the world, I think. There's very few people that can talk with such expertise and knowledge about these topics. Your brother, your twin brother, is a bioinformatics expert at the NIH, and you both are John Hopkins University adjunt professors, right? [00:05:23] Sajung Yun: Correct, yes. [00:05:25] Mizter Rad: And you guys, combined with the rest of your team, have very diverse background in biomedical sciences, PhDs here, and there. And you yourself hold an MBA, so you also have this business touch to it. [00:05:38] Sajung Yun: Yes. [00:05:39] Mizter Rad: How has it been for you as a, because I believe you're former, you started being, you becoming a medical doctor scientist. Yeah. How does it, how is it for you to actually run a business now? [00:05:51] Sajung Yun: Yeah, I think this, all this training helped. So, you know, briefly, I went to the I got the PhD in biomedical sciences and then went to the MD program. I didn't finish that, but during the I completed 121 credit pretty much most of the curriculum, but didn't finish that. But on the third year of the MD program, which is, four year curriculum, right? out to establish a company and there was the AI physician company and software. And then from then it was my entrepreneurial journey has been started and we've been. I've been working on that for the last 10 years and all this training from a PhD in biomedical sciences and MD program study. And then I realized that I didn't have enough knowledge on business, so I went to Johns Hopkins to take the MBA courses and did receive the MBA in both concentration in healthcare management and entrepreneurship. And then I all these things quite actually helped for this entrepreneurial journey. [00:06:56] Mizter Rad: Absolutely. And so just to jump right away into the topic, what, how would you define DNA Digital twins? For someone in my audience that has never heard about the term DNA digital twins, how do you, how would you explain the, in, in simple words? [00:07:12] Sajung Yun: So I wanna explain the difference between DNA analytics, versus DNA based digital twins. DNA analytics is our old way. So let's say, we are interested in these disease, for example, Alzheimer or dementia. And we know that ApoE 4 gene, or ApoE 2 or other genes are associated with positively correlated with that. So we give the report of the that person, oh, you have a mutation in those genes and therefore you have a increased risk. But that's the DNA analytics. But if you the DNA based digital twin is different that we analyzed entire, your genome, which human beings are known to have 20, about 20,000 genes. And then we cross reference those genes, entire genetic mutations into the 22,000 diseases. And then we basically build a Clinical Decision Support Sytem, which is there are two major functions. One is uh, Clinical Digital Support System. So it provides relevant information in real time, such as ApoE 4 gene mutation, whether you have that mutation or not. [00:08:34] Mizter Rad: So what is that? Sorry. So let's, let's go, let's go slowly here. The digital twin, you have a, okay. So let's start from the beginning. I am a, a person that is interested in understanding about more about my genome. I go onto your website. It tells me to get a kit. I get a kit. A little package, a little box into my home. Yeah. I understand you ask people to take a nail off their of their hands and put it in that box and then send back the package to you guys. You guys then read the genome of the person. Yeah. Right. And then, yes. So that, that's, that's exactly what you're talking about now. So once you have it there. In your premises, in your labs, what do you do? How do you compare it? You compare it to what? [00:09:26] Sajung Yun: To the reference genome, which is the gold standard of the human genome. And then we tell that, oh, you have a mutation. Certain genes have a, you have a sequence, has a mutation in it. [00:09:38] Mizter Rad: is this gold standard of the human genome, a one single perfect superhuman, so to say, is that the perfect superhuman or how, how would you describe that? Or is it several or how, how. Explain me more a bit about this. [00:09:56] Sajung Yun: So that's the average of the, like you say just, it's a like a standard ruler for the genome. So this genome is known to be like just standard. So we, human being has a, this genome sequence that, that's the currently so center for library medicine NCBI. Uh, National Center for Biotechnology Information. And then they have a assembly of the genome reference from the Genome Reference Consortium that we use the GRCH38 of the spec particular the reference genome. But this is just in basic it just the gold standard. And this is it's like a superhuman that, you know, if you have a very typical human, it's gotta be the, this sequence. That's, that's what it is. [00:10:44] Mizter Rad: Okay. I understand. And who defi, who defined this superhuman genome? Who, what, when, when did it happen? Do you know about this? [00:10:54] Sajung Yun: Yeah. Genome Reference Consortium does that. And then GRch38 was published around 2013 December. [00:11:01] Mizter Rad: Okay, I see. So, okay, so now you have my nail. You have my DNA n yeah. You compare it to the genome reference consortium superhuman genome, let's call it like that. Yeah. And you discovered that I have certain things that you want me to know about. What are those things? Give me some examples. [00:11:24] Sajung Yun: For example, Colon cancer. If you have a colon cancer gene mutation associated with the gene associated colon cancer we can tell you you have a increased, much increased risk for the colon cancer or not. Then we can the, you know, healthcare providers can suggest instead of following the standard guideline of the, starting the colon cancer screening at the age of 50 and every five years of the colonoscopy the, your healthcare provider can suggest due to increased risk, it's better to start earlier for the colon cancer and then screening, and then more frequent than every five years. Something like that could be done. [00:12:05] Mizter Rad: Okay, I see. So, uh, we go back to the topic and maybe your name is very telling because it is about predicting and treating these possible diseases earlier than what we are doing now. So in that respect, why should we as humans have DNA digital twins? Cuz you were explaining there's a difference between DNA analytics and DNA digital twins. I still don't have clear that difference. So D DNA analytics has been around for a while, I guess. But now you come in with the d the DNA digital twin, and I want to know exactly why are you guys better? [00:12:47] Sajung Yun: So, DNA based digital twin has a two interesting features: which is realtime searchable, and then second one is realtime simulatable. So for the realtime, searchable means, instead of PDF report, other DNA analytics company gives. Which is like static one. But you can search for any disease of your interest or any genes of interest in real time. You are basically, you are Googling yourself, you know, Google, you can google pretty much everything, right? Except yourself, right? So you can, when you Google yourself in the digital twin, uh, platform, it will tell you the risk for any disease, whether you have an increased risk. So that's the first component. Realtime searchable function. Another thing is the realtime simulator. Five years from now, what I'm saying, I predict there will be two types of patients who, one who build their digital twin, another one who did not build digital twin. For the ones who did not build their digital twins, physicians gonna say: this drug is known to cause this serious side effect. You might, have a very serious side effect. Then come to me and I have to change it right away. So basically they are testing the drug to themselves. But those who build their digital twins, we, we can simulate the drugs effect before they take the drugs. So in that case, if we analyze CYP gene and some other things, then we can tell For gene to drug interaction, we can tell you have, you are expected to have side effects because of this gene mutation, or you are expected to have a, this drug not gonna work well. For example certain drugs, it has to be changed to the active form in liver. If you have a lack of those enzyme it's not gonna be changed to active active form. And even though you give that drug to that person, it's not gonna work because it's not active form. So by checking, assimilating your digital twin, you can choose the right drug. Yeah. [00:14:50] Mizter Rad: Okay. I get it. I think what I don't get yet, and it might be a very basic question, but I wanna get it, get it done with, because I think it's important to understand. The DNA of a human being doesn't change ever. Or does it? [00:15:07] Sajung Yun: There are two types of DNA. One is the germline, another one is somatic mutation. So germline, which you are born with it, and then you inherited it with your parents and it's everywhere in your, every single, each cell you have. That germline generally DNA don't mutate unless you have a you know, the strong, let's say you went to the the universe or something somewhere, and then you got the very strong UV or something. So you don't, [00:15:36] Mizter Rad: okay, so if I would be an astronaut... Sorry to interrupt. So if I would be an astronaut I would go to I don't know, outside. In the outer space, it might mu mutate my germline, DNA. [00:15:48] Sajung Yun: Yeah. Germline is the one, keep making the same thing for yourself. You inherited your parents. Yeah, usually. Yeah. It's regarded as you don't mutate. But somatic mutations like cancer, those cells get mutation and keep keep mutating. And those are the somatic mutation is keep happening. So germline you inherited, your parents. Don't mutate generally. And then somatic it does mutate, but in this case we are analyzing the germline so that you only need to do once in a lifetime this test. [00:16:18] Mizter Rad: Okay. So if you do it, so now, now I have multiple questions because if you do it once in a lifetime... yeah. And you say that you have real time searchable and real time simulatable services for me as a patient, for example, right? If I create my digital twin, now I can go and in real time search for my genes. But what is what is changing? Why is it important to be real time? Because if I have, on the other hand, a DNA analysis with a traditional, more traditional company and I have a pdf. I still get the information from my genes. if nothing is changing in that aspect, why is it important to have it real time searchable, for example? [00:17:04] Sajung Yun: Yeah. So the mutation typically for the whole exome people, humans get about 10,000 to 50,000 mutations, different mutations. [00:17:14] Mizter Rad: For the... Throughout the life, or what is this? , [00:17:17] Sajung Yun: So mutations yeah, it's actually single nucleotide polymorphism, but so basically people have a 50,000, I mean 10,000 diff different sequences from the reference genome, which is a gold standard human being. And if you worked, I mean, that's a whole exome case. The whole genome case, that's about a hundred times of that. So my point is if you analyze entire human genome and there are so many different mutations, and then if you cross reference to the so many diseases for the whole exome itself, we once try to make the PDF. And it was roughly about 3000 pages long. Okay. Report. And I understand, and clinically it's not possible to use that kind of information to in clinical practice. So other companies, what they do is they just have a make the list, okay, what's the most important diseases? And then, okay, there are 10 most important diseases or hundred most important diseases. And then they just keep the report. Mm-hmm. , that's the traditional way. And that's the, all other companies doing that. But our way is that, okay, let's make the Googling Google software for customized for that person. So it's gonna be real time searchable. So when they, for example, when they have certain sign and symptoms, you know, oh, I have a, a fever, or I have a cough, or I have a joint type of mobility. Then you type it and then it brings the relevant result to you. So, oh, with this cough, uh, usually when you think of the cough, you just think about maybe I got infected from the bacteria or virus. But there are also other reasons. Cardiomyopathy, for example, the muscle disease of the heart. Then it might itch the lung and then then people might cough. So that could be the only sign for the cardiomyopathy. So that way real time searches is very helpful. And also human also have a good mechanism to protect. So a lot of cases when the serious disease is happening and people have early these signs and symptoms, and a few months later people get better because. People adapt. Think about it, if you have a bone pain, it's an extreme bone pain. People cannot leave it. And then sometimes the human physiology is, uh, suppressing the dose pain. So people think that, oh, maybe I got better. I don't have anything. But internally they are actually growing those diseases even though pain might be reduced somehow. That, so this realtime searchable future is very important that for the prevention of the early, early phase of the serious disease. [00:20:06] Mizter Rad: I see. Okay. That I understand now. So again, I sent you my sample. You collect my data. Now I have access to a dashboard. A Google like search bar where I can search for. Whatever, if I have any symptom, I just go to your search bar and see what is related to my genome in terms of that specific symptom that I'm having at the moment. And so based on that, I can also then read the results or the Google search results myself. Or if I'm not mistaken, you also have a team that helps me understand what might be the, what, what's going wrong, and why am I having a symptom? If it's something to maybe really pay attention to or not. Is that correct? Because me, as a user that doesn't have much knowledge, unfortunately, of my biology or you know, medical knowledge, I don't have studies in it. Um, I maybe don't know the lingo as well as, Some professionals do. I would need help with that? Or is it, is this something now I'm talking from, from a user experience perspective, this could be a challenge for you guys or for the user itself, or how, how are you solving this? [00:21:26] Sajung Yun: Yeah. This is not the B2C software direct to consumer software. This is rather for the healthcare providers. So, alright. They so they delivered the result and interpret the result for our customers. And then one more thing I just wanna make sure is that yes, we do have a science and symptom feature developed, but we did not commercialize it. In our commercial version, it's not available yet. All right, so in our commercial version, it's only genes and then gene search, those kind of things available, but still... [00:21:55] Mizter Rad: it's a vision, let's say. [00:21:57] Sajung Yun: Yeah. [00:21:58] Mizter Rad: Okay. Alright. So you're b2b, so you're selling to uh, hospitals, um, to... what kind of customers do you have? [00:22:06] Sajung Yun: Yeah. So we originally started pilot phase and then we provided to 150 customers. But after that we are trying to scale up by B2B and B2G level. For example, we are in Abu Dhabi and we did receive the government funding from the UAE Arab Emirate government. And for example, UAE has the issue with the consent in your marriage, meaning that cousins marry each other and I mean, in Arab Emirates, I mean in Arab, middle East generally there is some tendency. And instead of discouraging cousins don't marry each other we can screen for those kind of things. When, you know, the fathers genome know and mothers genome, you can have a expected baby genome before you even start to think about making a baby. That way we can so we can have a customized newborn screening plan. Instead of constantly newborn screening is for is you, we check, uh, general 21 Tri, uh, trisomy, uh, which is the Down Syndrome and some other, uh, genetic diseases, like let's say 7,200 standard items. But what you are trying to change is that it's gotta be, yes, we do that dose two, and then in addition it's gotta be customized newborn screening from the parent. And so if mother has the homogenous mutation and the father has a homogenous mutation, the baby has a hundred percent of the chance for that loci homo mutation. So that kind of phenotypical scaling should be added. So that's so that we are working on the B2G side as well. And yeah. [00:23:51] Mizter Rad: That's interesting. I, I wanna, I wanna know more about this. So the customized newborn screening that means, yeah that you go there's a couple that is, that wants to have a baby and they come to you and you read their genome and you tell them. Hey, if you guys have a baby, there is a big chance that the baby, um, has this health problem. Is that correct? [00:24:14] Sajung Yun: Correct. [00:24:14] Mizter Rad: And based on that, then they would have the baby or not? [00:24:17] Sajung Yun: It's a, it's their choice. But, we just provide the chances. If we have a heterozygous mutation for each person, that particular location, then basically you have a 25 5% of the chance your baby has got that mutation, or if you have a homo 100%. So we simply tell the information, and then it's up to them to decide. [00:24:42] Mizter Rad: Right, of course. Why do you think you ended up in in the UAE and did not get, or why did you get funding there and not in the US? [00:24:52] Sajung Yun: Actually we did not apply for funding in US yet. Okay. But in our strategy our strategy is to expand overseas first and then come to US and that's our corporate strategy. [00:25:03] Mizter Rad: I understand. And from a government perspective, since government is also your customer, why do you think they're interested in what you're doing? [00:25:14] Sajung Yun: They have several advantages. One is they can buy, let's say you build a digital twin for the entire country, right? Then for example, if you type the vitamin vitamin in our digital twin, you can tell what vitamins genes you have defective. So that currently you are taking generalized vitamin with the same doses for the each item, like vitamin A, a hundred microgram. Vitamin B, 20 milligram or something like that. But, with digital twin built, we can tell, oh, you need a hundred milligram for vitamin A and 20 milligram for vitamin B, something like that. For the National Health standard it's gonna be the precision nutrition supplement is possible which will increase their health. So something like that. And in addition to the newborn screenings. So a lot of good things could be done for their people's health, so that it's gonna be, we can increase the, increase the quality healthy lifespan. So that's, yeah, I think that's the we can help to each country. [00:26:25] Mizter Rad: I understand. So from a country, from a government perspective, it makes sense because they improve not only the quality, Of the lives of their citizens, but also maybe even the longevity of their citizens. And then thing, of course, yeah. Yeah. Tell me... [00:26:43] Sajung Yun: one more big thing is a cost. So if you detect the serious disease only. For my father case he had to take out one third of the stomach, which is a huge cost for the surgery. In addition to the pain or these things. But if you could prevent that. Mean the medication for that and it's gonna be much cheaper. The Government wise if you could have a early intervention and prevent those kind of things, it would be much less financial strain. So that's also make merit for each government. [00:27:15] Mizter Rad: Yeah, absolutely. I, it always comes down to money as well, right? When I'm talking about government or talking about private companies, it money is just a way of measuring how efficient maybe a society is. And in this case, I can imagine that, of course, if you as a government, as a country are able to detect, to early detect diseases that your population might encounter in the next 10, 20, 50 years you can save money by doing some preventive medicine. And you guys are helping with that. And that's that sounds fantastic. When you talk about DNA screening and collecting the DNA and the genomes of people there is for whatever... so one of the things that come to my head right away is, how do you safely guard this information? Because it's very sensitive information. And someone could think, you know, if now insurance company get to know about my genome when I'm 10 years old or 15 years old, whatever, 20, 30 years old, and they know that I have a, a gene that might give me some issues with my stomach later on they're not gonna take me because, uh, whatever that, that becomes too expensive for them. Or something like that. How do you protect not only from insurance companies, but in general, how do you protect this sensitive data? [00:28:39] Sajung Yun: Sure. At first I wanna talk about the regulation. So the first thing I wanna mention is the Genetic Information Non-Discrimination Act, which was which is to protect people and then prohibit types of genetic discrimination. So there, there was, there, there is a act. It's called GINA; GINA. And then basically this protections Inc... intended to encourage Americans to take advantage of genetic testing for their medical care, by protecting them from the health insurance wrong usage or employment. So basically it protect individuals against discrimination based on personal understanding information. So that's the GINA. And then I think your question is how do you how do we comply with that GINA Gina? So, yes. So we at first gave the the right of their medical data or health data to our patient, customers. So the interesting thing nowadays is that, when you go to hospital or see a doctor, doctor write it down what you have signs. Oh I have a, a chief complaint of, fever or something. They write it down in the health record. But you know, who has the access to your own health record? I mean, probably not many, right? [00:29:59] Mizter Rad: I agree. I agree with you. I, this is something I always think about. I don't have my own records and that is not good . [00:30:07] Sajung Yun: Right. [00:30:07] Mizter Rad: I wish I had them. [00:30:09] Sajung Yun: It's your data then, right? You don't have access to your own record, which is, I think, totally wrong. So our solution is that, okay, so we give the ownership to the each person and each, it's just so natural thing. You have to have your own data and you have control, and then you are, we are letting your healthcare provider to use that data. So ownership is given to the each of our customers. However it's gotta be delivered through the healthcare provider. And then, whenever they authorize our customer authorize the usage to different doctors, they could use it. [00:30:45] Mizter Rad: Okay. I see. So I wanna go back to the process of how you as predictive as a company that creates digital twins of my DNA, or your DNA, anyone's DNA. How do you work? What's the process behind the scenes when it comes to you as a company? Because we explained what the user path is. What the user journey is. I send the box with my nail. Then you give me a dashboard where I can Google my genes and so on. But when you get that box, what do you guys do? Like specifically, you already told me that you cross-check it with the, what we called in this conversation, the superhuman DNA. But how does that exactly work? Is that on a lab? Do you have a team? Where's that team based? How do you do it? [00:31:32] Sajung Yun: Yeah. Our team is very internationally based. So when the data comes we analyze it. We made the automated pipeline, so in order to scale up we had to build that. It... So that's why it took so long time to develop. The technology has been developed over the last five years. And then so it's pretty much almost automatic. And then, so data twin has been... [00:31:56] Mizter Rad: so it's basically a database. Like you, you crosscheck one database with the other. [00:32:01] Sajung Yun: Right, but the, not just the, using the existing database, we had to write our own tool to to do that, yes. But generally, yes, it's we use the the database, which is the authoritative database like NCBI, which is commercially usable and also it's trustworthy database. But we have to write the many different tools to basically connect the dots. So yeah, so the people get the those digital twin and then... Basically you have your own Google for yourself. That's it. And then whenever you want Google yourself you Google it and then, uh, actually your healthcare providers Google it for you and then provide information more in the manner that you could, I mean right information could be given. So in order to prevent the misunderstanding. So that's the way yeah, it's delivered. [00:32:53] Mizter Rad: Okay. So you deliver the results, again through your customers, which are this medical professionals or the clinics, the hospital. You don't deliver directly to the end consumer. So that makes it easier for you. Maybe also to be more clear in the explanation of the results. Cuz as we were saying before it is complex to understand if you're not in the scene, if you're not a doctor, basically. [00:33:17] Sajung Yun: Yeah. And then I wanna clarify one thing. If you really need for the medical purpose, we, yeah, we are also, we like to try, we wanna help the community. So we'll consider and we'll do it and help out the, those people who really wanna do, I mean build a digital twin. But generally currently as a the small team we are not focusing on the b2c, but we are already focused on B2B and B2G. Because our aim is to provide a digital twin, like a country level, country by country level. So we want to provide like millions at a time, instead of one person at a time. That's why we didn't open we are not opening the B2C channel. So if you really want yours to be done, you wanna make a digital twin and we could contact well consider the case by case and then do it for you. But generally we don't because we could scale up. I mean city level, country level or the institution level. [00:34:13] Mizter Rad: I understand, but sign me in. I want one. Where do I have to live? If I wanted to have a digital twin? I talked to someone in my show some months ago. She is a security advisor for the Ministry of Defense in Estonia. And in Estonia they're doing this. They are collecting the DNA of all citizens and creating a database of, I don't know if they do it person by person or if basically or not; I'm not sure if they're advising or sharing with each individual their their digital twin. But they, as a collective, as a nation, they already have collected a bunch of or a pretty big sample, I believe. So it's interesting that you said that you're focused on nation, on a nation's level or in a country level because I believe there might be more countries like Estonia interested in this. Which countries are pioneering on this? [00:35:13] Sajung Yun: Yeah, I think that's a great question. So Estonia is one of them. So actually I was in Estonia a couple of years ago and gave gave a, a presentation there at the particular Genome Institute in Tartu, Estonia. And they have a, they want to sequence the 70% of the entire population, which is 1.4 million. So yes, there's, they do that, but, oh, by the way, I need to clarify this. They are doing the genotyping, not the sequencing. The difference between genotyping and sequencing is that sequencing you get the entire sequence by next generation sequencing technique. But so basically from DNA chromosome one, two, chromozone 2, 3, 4, et cetera, you just basically get all the sequence. ACGT sequence. But genotyping is the one that they they do the it's called SNP chips which is you get the um, you get the difference between the, you basically have a bunch of the sequence, like 30 base pair, et cetera, and then you tell that, oh, there is a mutation there. Something like that. So that's the SNP chips. So for example, Estonia, they do the most of them is a SNP chips, and then they impute sequencing. And then yes, they do have a very precious sequencing data too. Quite big number too. But but Estonia genome project is mostly about SNP chips, and... [00:36:37] Mizter Rad: sorry, did you say stitching? [00:36:40] Sajung Yun: SNP , single-nucleotide polymorphism chip, chip. yeah. So there's a different technique compared to the next generation sequencing. Okay. I'd be happy to explain the difference, technical differences, but I'm not sure you wanna listen to this. [00:36:57] Mizter Rad: I don't think we need to go down. I mean, un unless you can explain in, in very simple words for my audience. [00:37:02] Sajung Yun: Sure. Next generation sequencing you use the laser and then it is, it has the color. You, and then each sequence have a different color. There are four types of DNA, uh, other ones inside enzyme, into, uh, four different colors there. And then you take the ccd. That picture, and then you just get the you read that sequence. That's the next generation sequencing technology. Versus SNP chip technology is that you basically have a little hole or well little space, and then you put the standard sequence. And then you attach those 30 base pair sequence. And if you have a mutation, it's not gonna stick there that much. If you have a perfect match, it's gonna stick there um well, so that when you see the grade scale, how much it got attached if you have a mutation it's gonna be like light. Depending on how you set the grade scale. But anyway, it's not gonna attach that much versus attach very well. So you can tell the difference of the, I mean the mutations, that's the genotyping of the SNP chip. [00:38:08] Mizter Rad: And so basically the end result or the objective with either of those two, genotyping or sequencing is to do what? [00:38:18] Sajung Yun: Find out the mutations of the... [00:38:20] Mizter Rad: find out the mutation... [00:38:21] Sajung Yun: your DNA. Yeah. [00:38:22] Mizter Rad: Of your DNA. Okay. So, but in the end, both of them give you a result that is optimal, or which one do you think is more optimal? [00:38:31] Sajung Yun: Yeah. Next generation sequencing is the optimal. Well, good. Okay. Because it gives the specific sequence. Versus SNP chip is giving the mutation information for the currently there is a Illumina chip. There was a 1.4 million possibilities. Instead of 3 billion, so SNP chip is, you can tell, you can regard SNP chip is a more simplified version of the sequencing. Yeah. [00:38:56] Mizter Rad: Okay. I understand. Going back to the countries that are pioneering on this. Cause I think this is important. I like to look at the world as a as a complex interdependent organism. And this organism has different bodies. Different parts. And some parts are moving faster than others. So when you look at the world's from a macroeconomic lens, of course you see there are some superpowers, super economies moving faster than others. But when you look at health I'm curious, who is moving faster? What's your opinion on this? [00:39:31] Sajung Yun: Yeah. In terms of the sequencing, for example, Arab Emirates, UAE, has a project Emirati Genome project, which is plan to sequence entire citizen of the country. Entire citizen is about 10% of the entire population because there are a lot of expats. But anyway, the Emirati Genome Project is gearing toward to sequence entire citizen of the country. And then likewise, there is a UK Genomic England aiming to sequence 5 million people. And and the Scandinavian countries. And like the Estonia and other countries, countries. So they do have those country by country. Also France. They have a big genome sequencing project. Including US. Yeah. [00:40:14] Mizter Rad: And how expensive is this for a country? Do you charge this per person, per citizen, or how do you come about budgeting this? , [00:40:22] Sajung Yun: I believe those things they do the sequencing for free, I think most of them. But let me tell you how much it costs in-house from their side of those two sequence. So currently roughly you can think of $1,000 to sequence one person, whole genome. But there was a new announcement. It could go down as low as 100 to $200 within one year from now. [00:40:45] Mizter Rad: And why is that? Why is that change? Why is that coming down so much? [00:40:48] Sajung Yun: Because they the Illumina, is the one company leading this sequencing industry. And then they announced that they make the NovaSeq X which will be providing the the reagent in the very follower price which was announced I think one or two months ago by CEO. So that's the, and then yeah, so there are some other competitors like BGI MGI. And so there are other competitors providing services more affordable. And there are some other sequencing technology available like Oxford Nanopore and some other companies as well. [00:41:24] Mizter Rad: Okay. I mean, it's like with every big technology at first people are afraid of it. People don't understand it, and it's also super expensive. Mostly for the richer individuals. But as time goes by, people start understanding it more, they start finding the new technology useful for themselves. They're not afraid anymore and on top of that it gets cheaper. So what would you tell those individuals that at this point, cuz this is such a new technology, that are still afraid of sharing their DNA with a private company like yours, or a government like the UAE for example. [00:42:07] Sajung Yun: Before that, let me tell you about the DNA first. You know, DNA is the sequence that, think of a DNA is a blueprint or copy machine. So DNA is the copier that has the entire your protein library. We extract the DNA, for example, in our company from from the fingernails. So basically break the disulfide bonds of the keratin, which is the major component of the fingernails. And then extract the DNA from it. And then, which just simply breed what the DNA sequence is. So DNA sequence is nothing but continuous, like ACGT, ACGT, something like that from number one to the entire chromosome. And then once you read that, nothing people might worry about is the cloning. You know? Oh, what if, you know, you just clone the DNA and make something like my copy of me without my permission, et cetera. The synthetic biology which kind of clone the DNA, et cetera. Um, I mean, at first it's ethically not right. So you legally you cannot do it. Secondly, by to synthesize entire human being, it's not possible with our current technology. And of course you probably have heard of CRISPR CAS 9, gene editing. You could edit the... [00:43:32] Mizter Rad: the gene editing technique. Yeah. [00:43:35] Sajung Yun: Yeah. Gene editing is something like, let me explain about, briefly about gene editing, about CRISPR CAS 9. Because I've been working on that at NIH for the five, five years about that. [00:43:45] Mizter Rad: Please do that. [00:43:46] Sajung Yun: Yeah. So CRISPR CAS 9 that is, there is a binding region and then cutting region. And then another one is the template. So the binding region is the RNA, so RNA guide that we call guide RNA, so that search for the specific DNA sequence and bind at the target. So once you bind the guided guide RNA binds to the specific target, then CAS 9 protein is the one cuts. It cuts the DNA. Usually it's a double standard cut. But anyway, so it CAS 9 protein. There are different proteins like CAS 12 and the other things, but CAS 9 is the most popular. The CAS, C A S protein cut. And then once they cut, then DNA there is a repair process that reconnect. And then in order to insert the right sequence you edit the gene, we are providing the template. So it binds the left side of the cut DNA and then right side of the cut DNA. And then there is there is a specific, if you wanna put it in something like that. So that's actually called Homology Directed Repair template. HDR template. So in summary, CRISPR CAS 9 is nothing but combination of guide R N A that binds to the target and then cut with a CAS 9 protein. And then when you repair, it has a template attaching there and then inserting the right nucleotide. The right things. So that's called gene editing. [00:45:24] Mizter Rad: And this is all, this is all, this is all possible now in a lab. [00:45:28] Sajung Yun: Yes. Yes, it is possible. [00:45:29] Mizter Rad: the, And the cons, the, let's say the consequences of this is that you in theory can now when a baby is born or maybe even before, , you can determine if this baby, if maybe there's a precondition of this baby genetically. A precondition for having stomach cancer, now you can change that and edit that gene so that he doesn't, or she doesn't have stomach cancer anymore. That's pretty much how it works? [00:45:59] Sajung Yun: Yes. Stomach cancer, g gene anymore. So we added the germline for the new baby. I mean... [00:46:05] Mizter Rad: but is this legal? Is this legal? Is this legal? Sorry Sajung. [00:46:09] Sajung Yun: No. [00:46:09] Mizter Rad: Okay. [00:46:10] Sajung Yun: So since this is ethically questionable. So let me give the example of the mice. So let's think let's say, I wanna make the green-eyed mouth instead of the, red eye mouth, right? Then we find the specific protein that is responsible for the color of the eye, right? And then we design specific mutation. And then to guide DNA to code for the, like from red eye to the green eye. Then we put it into, back into the cell. And then we can make the and then model mouse. And then it can make the, instead of the red eye eye newborn mouse, we can make the green eye newborn mouse. So technically, if you give me three months, or I mean, let's say three to four months, and I could make the green eyed mouse. But that's the general theories. [00:47:09] Mizter Rad: Yeah. Okay. I understand. But this is not legal yet. And for humans, there's for humans, yeah. And there, there's a big concern, ethical concern around it, obviously. .. [00:47:20] Sajung Yun: Two big concerns. One is the off-target effect, which is, you know, when you bind something, it could bind to wrong place and then make the wrong cut. Because you know... [00:47:30] Mizter Rad: So could be creating creatures that are not to our knowledge, basically. [00:47:36] Sajung Yun: Yeah. Not you intended to. And think about it, DNA is not straightforward. It could fold. So when DNA got folded the guide RNA could bind to the wrong sequence rate. It will just recognize the straight line to unfolded form, et cetera. So folding will change. I mean, give the possibility, potential for the wrong binding. So it's called off-target effect. And then and then DNA repair enzyme, Homology Directive Repair, which is not always a hundred percent right. It could, DNA has the other other repairing possibility, which is called non NHEJ which is the, so you can basically sequence the wrong sequence could be inserted. It's called Non-Homologous End Joining Pathway instead of the Homology Directed Repair. Okay. So that case wrong sequence could be put in. So there are two important technological issues we need, which is off-target effect versus the the wrong sequence bringing yeah. [00:48:40] Mizter Rad: I understand. And so all these mistakes, let's call them. All these errors. Yeah. All these downsides. Have you seen them throughout time, since you're in the scene and you're in the industry. Have you seen that these errors have been diminishing? Are we getting better at this? Do you think at some point will be so good, we'll have such good technology and knowledge and confidence to actually apply CRISPR on humans? [00:49:07] Sajung Yun: Oh. Yes. There are several techniques developed since the original CRISPR Cas 9. For example Dr. David Liu lab in Harvard. They developed the prime editing. So instead of double stranded cut, they can, DNA has a double strand, right? Instead, they can just cut the single strand and then put in, into the just one right, uh, sequence there. Insert something like that. So yes, there were a lot of advances, technological advances. And efficacy of the gene editing is increasing. However, it's not like a hundred percent chance effective. You know, roughly, you can think of the Homology Directed Repair efficacy from 1% to the 20% range. Yeah. For okay. As of today. [00:49:50] Mizter Rad: So it's still very low, [00:49:51] Sajung Yun: yeah. Yeah. [00:49:52] Mizter Rad: What's your opinion in general of optimizing longevity? Now talking about CRISPR. Talking about what you're doing with digital twins. Helping people focus on preventive medicine rather than the type of, uh, medicine that we have now that is basically going to the doctor when you feel sick. And, you know, other trends that we see. Better nutrition. Countries again, being more predictive with their citizens and trying to understand their, from a, from a genetical predisposition, how can they help them to have a better life and maybe a longer life. So I guess my question is, again, what's your opinion on optimizing longevity and do you think at some point we'll be immortal? Will we have the capacity to say, okay, I'm not, you know, I can live a hundred years longer or 200 years longer, but I just don't want it to I just don't wanna do it. I'm gonna switch off my system, so to say, but I'll have that power to decide when to die, not as now that we pretty much don't have that power. What are your thoughts on this? [00:50:59] Sajung Yun: Sure. So let me just start with my experience. During my PhD time, training time, I was working at the Asia Aging Study. Which is the longevity study. So they tracked the 3000 World of Veterans for 30 years or more. And they took prospective cohort study, so for the longevity. From that experience. I'm in data analytics. That was my job. And I realized that in a basically balanced life. It's a very common thing, you know, the balanced life with the nutrition. And then behavioral modification, for example, for chronic disease patient like hypertension, diabetes, and like exercise and reduce the you know, reduce the, those junk food, et cetera. And then reduce the stress. Those kind of things actually do contribute the longevity and the immortality. So let's think about, how people die. So for example cancer cells has cancer cells, has the uncontrolled cell division. They, therefore it grows a mass and inhibit the normal function of whatever the organ the cancers have, for example lung that the, those mass and then mass also influence the, those chemical reaction and oxygen exchange with what the lung does. And then those lung cancer patient have a difficulty in breathing, et cetera. And then, for example, kidney cancer. Kidney is known to purify you know, the spleen, for example, purify the blood and then those kind of things. And p d E is also purify the blood and make the urine. And when those functions are not working, the blood is becoming toxic, and then people cannot handle those kind of things. That's the way, eventual the organ failure and then death. So in the future, eventually, you know, they, people are working on the, those synthetic organs. And then and I, I don't know that it will lead to the immortality fast enough in our lifetime. But so when you have a certain organ failure maybe synthetic organs might be available and then and it keep the functions going. So I think there are good technology getting developed and then we have a more and more chance to live longer. And then on, but the, I wanna just stress out one thing aspect, which is the preventive medicine because and a predictive aspect in a, you know, when have a longer and a healthy lifespan. That's actually my goal for the society. That's my mission, life mission. I wanna make our society, our community live longer with a healthy lifespan. And then the, I think the one, one of the best way is to analyze the people's propensity for the future serious diseases and try prevention as much as we can so that we delay the onset of those conditions or the even prevent that. So that's the, I think the best way. And then that's why I'm devoting myself with our team for the preventive health. [00:54:14] Mizter Rad: Absolutely. I am fascinated with what you're doing. I would love to get my digital twin copy myself. It was great to have you here, Sajung. It was fantastic. I'm very grateful about this opportunity and very happy that we managed to talk. I learned a lot from this conversation. I hope you had a good time. And let's see if Predictiv grows as you intend it to grow, and maybe we see you here in Germany as well at some point. Or in Europe, a bit more spread out. Any last words for those listening. Those interested in knowing more. Those interested in your career? Anything you wanna share? [00:54:55] Sajung Yun: Sure. I think, for the people who really want the digital twin, we'd be happy to consider case by cases and do it. Even though we are focusing on the B2B and B2G scale. But still, every single person is a precious life. Very, very precious. I think most precious thing is the life. so one person's digital twin is as important as a hundred thousand people, digital twin. That way so if you wanna be wanna contact us we can contact by predictivcare.com. Actually pre PREDICTIV. No letter E. predictivcare.com. And the email address is [email protected]. Then our customer representative will escalate this case to the, our committee. And then so basically we would like to contribute and help The people to have the increased quality, I mean, healthy life span. So that's, that's what we are aiming for. Yeah. Thanks. [00:55:53] Mizter Rad: That's fantastic. Sajung, thank you so much. I wish you a happy evening in Bangkok and hope to see you soon. [00:56:01] Sajung Yun: Yeah, yeah. Thank you.