 My lab is called proteomic big data. Normally we talk about big data, you are probably thinking about the text, a lot of web pages, a lot of pictures, a lot of video and the voice, music and so on. And now we are saying there's another type of big data which is proteomic big data, which is the key of life activity and the basis of our life. And they are changing and they're generating data. And this data is the key to diagnose disease and to treat diseases. Yeah. I was so happy about this Tianan, I'm very happy because I came to China in 2019 for about a month and I had some really fun, powerful times with the both the people, like the hospitality and the love, but also with the professors and with the institutions and what they were pushing the edge on. And so I did partnership interviews with Peking University in Beijing and also with the Peking University and also Westlake University in Hangzhou. And this was hugely inspired by also Kirill Piatkevich who helped make it happen, who I love you very much. Thank you. Thank you. It was a huge transformative life experience and it's a transformative life experience, so powerful. And now about, let's see, one, two years, a little over two years, let's say later, Kirill reached out again because we pretty much laid the foundations for unity, we could say, like bringing the world closer together, bringing the world's leading scientists and entrepreneurs and leaders together in a greater harmony and especially bridging together places like the US with China or Europe with China, Africa, Asia, just all together as one more and more. And so Kirill reached out again and with introducing me to Tianan Guo who is a principal investigator at Westlake University, he has two labs there and also he is the founder of Westlake OMIX which is using AI-powered big data to just try and understand what's happening at this micro level and increase our health and our well-being. So I'm really excited to spark this fire again with China and with Westlake University and welcome to the show, Tianan. Thank you. Thank you, Atlas. It's a great honor to be here and then explain our research to you. Yes. And we were bonding for a bit before we started the show and I was just more gaining a better understanding of what exactly is happening there because it's a lot easier when I'm there in person walking around Westlake University, going to your labs and actually seeing what's happening with the data and also with even all of the trials and the tests that are happening and the beautiful insights. But when we're remote, it's a little bit more difficult. So at least for now until we open up again and we're able to see each other in person. But why don't you go ahead and explain to our audience what exactly is going on at Westlake University in your two labs and let's start there and then we'll also bridge that to Westlake Omics as well. Okay, it's a great opportunity to explain here about our research in Westlake University and the Westlake Laboratory. Basically, we are studying proteins and other molecules which are a gears of our life activities and our body. So if you look at your finger, it looks very smooth but if you have a microscopy to look at your finger after amplified for 200 times, you will see a lot of cells. It's like an egg. And these are so many cells they can't defeat the cell reactions. But a cell is not the basic unit of life. If we have another technology to further amplify a cell by tens of thousands of times now but let's assume it's such a technology then we can see another universe. And in this universe, actually we have billions of molecules. These molecules are mainly proteins. They're also nucleic acids, lipids, small molecules. And they are the building spaces of cell and the cells, it's building spaces of our human body. And what we are studying is the movements, the dynamics of these proteins as the basis as the gear of life activities. You can imagine it's so challenging because they are so tiny, you cannot see it. You probably have heard another cool technology from Google recently. They have an alpha fold, alpha fold version two which use AI to predict a structure of a protein of a singular protein. So they develop a very smart algorithm. You can basically predict what protein looks like in the micro world. It's just a single, it's so exciting. The entire world is celebrating this a major progress. But if you go to the context of a cell and a context of a human body, that's just the one protein out of billions of proteins. So basically we are using some cutting urge technologies to study those are billions of protein as a whole. So we call this type of science called proteomics. You probably know genomics. Genomics is the study of thousands of genes and the proteomics is to study thousands of proteins. You know, there are billions of them, but there's no technology to measure every one of them. We can only measure thousands of them in our time, but the technology is advancing very fast. Then in West Lake University, we are collaborating with a lot of clinicians from different fields. Some of them are oncologists. And in the recent two years after you left West Lake University, you were stalked by the COVID-19. And we also have some research on COVID-19. We study why the virus can invade into human being and why invaded by infected by the same virus. Why some people, they just have nothing. They didn't even have a cough or fever, but some people have a cough and fever and they are long severe cases. They have mild symptoms, but some of them actually died from COVID-19. Why is that? Because they have different protein reaction to the virus infection. And then we have almost two years of study of this. We also are studying why after vaccination, some people have a high level antibody which can kill new virus infection, but some of them have a really low level of virus. Some of them need another second injection. Some of them, most of us need a third injection. And why? Omicron, they are different from the original stream. So all of these are related to proteomics. And also because of proteomics is so challenging. There are so many molecules involved. They are always constantly changing. So we cannot do this by our calculator. We have to use the AI to model the dynamics, the changing and predict the behavior of the micro weld molecules. So this is all our research happening in Westlake University. And also during the past two years, we've got a new mechanism, and also which leading to my second lab in Westlake University, which we call the iMarker lab, intelligent biomarker discovery laboratory. And so this lab belongs to what we call the Westlake laboratory, which is a provincial laboratory in Zhejiang province. And this laboratory, we focus not only on the proteomics, but also other omics. And with this other technology, we build up a cutting-edge facility, which can support research not only in my lab, but also in other laboratories in our campus and outside our campus. So this Westlake laboratory, we try to establish a new mechanism. We are not only going from zero to one, but also hoping we can go from one to 100 to 1000. And this Westlake laboratory serves as a bridge between zero to one and from one to many. And this is another big experiment originated in Westlake University. And we are trying to develop an optimized protocol for translation research in rapidly evolving Chinese science society, slightly different from other places. There's no recipe suitable for every weld, everything, but this is something I'm trying. And in addition to this two laboratory, we also now have a spin-off company. Since 2020 in June, so this also happened after you left Westlake University. And during the past one and a half year, we have finished two rounds of fundraising for fundraising, and we now have about 80 people in Westlake Omics. And Westlake Omics is trying to do from one to many. Basically, we are translating the technologies the biomarkers found from our research into clinical assets, which can help to improve diagnosis and improve cancer or cancer treatment or treatment of other diseases. And a major product we're developing is to diagnose thyroid nodules. We have a thyroid just in front of our throat and it's very small organ, but it's very important. It secretes essential hormones that maintains our life. It's also called CEO of the human body. We receive signals from the brain and they transduce the signal and controls our liver, our stomach, our bones and our muscle and also in motion. It's crucial. So thyroid also subject to diseases. So almost 50% of the adults have a thyroid nodules. You can easily detect it by ultrasound. You will see that at our age, many people have a thyroid nodules, but most of them are blind. You don't have to take additional actions except to sleep more and to be more relaxed. But for some of them, they are malignant. For malignant, you can also name it as thyroid cancer. If you had a cancer, then everyone is nervous. There's a gray area, it's about 30% of the thyroid nodules. Nobody can tell it is a cancer or lung cancer. It's a great challenge and also great pressure, emotional pressure to both the patients and also the doctors. So my collaborator in Singapore told me, so he's a senior in surgery. So he said, if he found such a case, he would tell the patient, be relaxed. So your nodule, probably 80% chance will be blind. You don't need to do more, you don't need to do a surgery. Just relax and adjust your lifestyle. But if he's a junior in surgery, so he will be very nervous. He will tell the patient, okay, now you have thyroid nodules. We don't know whether it's malignant. Probably 20% chance it's a malignant. Do you want to remove it? So the patient, talking to the younger surgeon, we're likely to remove his or her thyroid. But after removing the thyroid, he or she will have to take artificial hormone for the rest of his or her life. So it's so challenging. So we have developed a protein and AI-based test which can increase the precision to diagnose the malignancy of the thyroid nodules. And this is what we're trying to do in Westlake-Omins. We're also trying to develop AI pharmaceutical technologies to accelerate, improve the drug discovery. Because drugs treating cancer, treating diabetes, and so on, they are the most important, one of the most important industry in life science. So there are so many involving protein measurements. And this is what we are, this is what we're doing. We've been doing it for the past two years. Cool. Nice. Yeah, there's so many good things there. I love how you called proteins the gears of life. I like it a lot. Yeah, that's a good one. So there's several layers we could say. So at the level of the encoding of the double helix is the... So that's the genomics. And then you have the transcriptomics. And then we have the proteomics. Then we have the metabolomics. And so then there's the omics for all four of these levels, all the way up to the exposomics, the way that it expresses itself in the environment. And basically like the feedback loop that the environment has on the genetics itself. So the genetics are in a feedback loop with the environment feeding into each other. And okay, cool. So then we'll probably talk about that and which layer of omics you're most focused on proteomics. Is that right? Yes. Cool. And then another thing that you mentioned that I thought was great was that if we find the biomarkers, so we can use laboratories to find biomarkers like for example, the thyroiditis issue. And we can... If we identify at the level of the proteomics, the biomarker, right? The one that's showing there's a malignancy that's developing, right? And usually you can get that biomarker. Is that, can you get that without a biopsy? Can you get that through like saliva, blood? No, we have to get a biopsy and measure the proteins. Okay, you do. You have to get a biopsy itself. Interesting. So then there is like an invasive component to an extent of actually having to perform a surgery in the first place. Okay, cause then there's a level of prediction even before that that we can attend to even before needing to get a biopsy if we can find biomarkers about the malignancy developing even before that. That's even better. But I see, so then if you... Yeah, go ahead. Yeah, go ahead. It was a very important field because in most cases I have to explain this. Otherwise people may get confused like you. So for cancer diagnosis, there are two aspects. So first is for early diagnosis. So for early diagnosis, we need a sensitivity in a scientific term means if there's an early phase cancer, we need to detect, we cannot miss it. Because if we miss it, then you will, the cancer may grow to too late before you can treat it. So for early detection, also we call the screening, we need long invasive methods. Like in the case of thyroid cancer, this method is already there, which is ultrasound. Oh, ultrasound. Oh, cool. You have some lubricant and then you can detect it. You will see a lot of thyroid nodules and if you have a thyroid nodules, it's very likely it will be... It's likely the cancer and if it's a cancer, it's really malignant and they have some morphology and the expert can tell directly from the ultrasound. But the next phase is differential diagnosis. Okay, now you have a tumor. But this tumor can be beeline or can be malignant. This is crucial. If it's beeline, I don't need to cut it out. If it's malignant, we need to remove it. So in this case, we need a differential diagnosis. For differential diagnosis, the best way, the gold standard is to analyze the tumor itself. And this has to be evasive, unfortunately, because the next step will be surgery to cut it out. So for many other cancers, early diagnosis is a major issue like pancreatic cancer. It's very in our body. So you have to... You cannot easily take a vipersy for your pancreas. So it's very important we develop a long evasive or based on blood tests to tell whether it's a tumor or not. But for thyroid, the screening is not an issue. It's so easy to cut an ultrasound. The challenge, the clinical needs is to differential diagnosis. Yes, nice. Yeah, so the more that we can find it at the stage of the ultrasound, the better, basically. Because then it's as early as possible in the detection of the malignancy forming. And that's our most ideal. And then when you do have the biomarkers, let's say when you do get into the later stages and you take a biopsy, you find the proper biomarker of analysis. I like how you describe that as like a zero to one. So Peter Teal's famous book. And so you have a zero to one is like the creation or the ideation or the realization of something like that where you find the biomarker. And then if you prove it empirically multiple times across different areas, different teams, let's say, then you can do one to many. So then you can basically take the realization that was found and then begin applying that across different hospitals around the world for when they're in that late stage of malignancy to then find a good method for transforming it back to its non-pathogenic state, to its fully healed state as much as possible. So thyroiditis is a pretty big one. That's, is that a million people a year across the planet or how many people a year? Yeah, huge numbers. I see many minions every year. So thyroid nodules occurs to 50% of the population. And for thyroid cancer, it's about five to seven percent of the population, so huge. Yeah. And then so, OK, so then there's a general understanding, let's say, and then it's good because we did a general understanding. And then we also gave a specific example for the thyroid, and then let's do another if we kind of zoom back out again and then we'll get to the pharma discovery part as well, which I find interesting. I think when we zoom out again, I like how you describe that. When, and this is a great way for also for us to just get more excited about life biology because when you do this like zooming in method and you see I like how you call it, you zoom 200 times in and then you see a bunch of little eggs that are cells. I think that's yeah, I think that's a great way to put it. And so you see these little clusters of there's like 30 trillion just in the body. And I like how that can act as an inspirational tool for young people is to say, well, what's happening at that cellular level and how to live a healthier, happier life and also help others live happier, healthier lives. And I like how you also mentioned alpha fold also. So like predicting what proteins will look like. That's such a huge part of this. So let's and also just the I marker lab and intelligent biomarker discovery and how different proteins respond to antigens. I thought that was also super interesting. So let's talk about if we zoom back out at the macro perspective and then we'll zoom back in. So if we if we kind of see proteins as these gears of life, are you looking at a specific like the specific proteins that like have a greater effect on most bodily functions. So is there a greater more, let's say more generally produced protein set by the genome that influences a greater amount of the bodily functions that is more important for you for your labs and company to focus on? Yeah, this is there are so many problems. There are nobody knows countless, really countless proteins, you know, internal micro world universe, you know, but and every protein has these functions, not only one function, but many functions, and it's different from the gear in that every protein is being synthesized and degraded. They are burst and then they are died and the degrading too. This is so interesting. So so the protein, what's the life cycle of the of the protein? So it's so it's the let me see if I can get this right. It's the ribosome that creates or assembles the protein. And then how long was the how long does the protein last in its sort of in its gear state, right with the multiple functionalities? And then how how long does until it dissipates, let's say, the topogen? Yeah, yeah, this is very another very good question. We know very little about it, but we even didn't know or even don't know how many proteins are there. And some proteins they are. They have a few hours and maybe a few days and some of them maybe a few months and nobody knows. There are several technologies to study this. And because this is very important in like Alzheimer disease, it's the because the degradation of the protein or some protein have a arrow have a bug and then it become severe disease for human being for brain to function. We know so little about it. So you know, we have a lot of rules in mathematics in chemistry physics. But it's very little rule in biology, because we even don't know how the elemental biology, which is the protein function, how the how long they be, and where are they, how they just don't keep inside the cell and how fast they move. No idea because we cannot simply cannot measure it directly. And currently we have to measure it by some indirect way. But one second. So the protein is also proteins are also moving from parts of the cell to other parts of the cell to potentially do a different functionality also. Exactly. So when your finger moves, when your mouth moves, actually, the proteins are moving. That's why your head can move. Your finger can move. So the proteins are enabling like the muscular skeletal system. Exactly. There are actions, the actions, they are, there are lots of actions, and they are moving. So then the cell are moving, the mass are moving, and then your finger are moving. Cool. Okay, yeah, you were, yeah, you were going to say you're continuing to to go on. Amazing. Well, we don't, we have a very little idea. So we try to interactively measure the protein movement, the dynamic, so that we have some idea about it. Then go back to your question. So which protein are there is their protein that we focus on? So my answer is, this depends on the question, the research question. So we define, let's say, the thyroid diagnosis, thyroid nodule diagnosis. This is a question. Then we try to find out what's the difference between the beeline nodules and the malignant nodules. There will be hundreds of proteins that difference. But if you know the key protein, which can be easily detected and easily used to fit into an model to distinguish this malignancy, that's the protein we want. And if we're talking about a treatment, let's say, you know, leukemia, it's very difficult to treat. But people found the chronic maloid leukemia, and the others have a key protein, which are different, which is a base rebel protein, which is a fusion protein. After we done by the fusion protein, we can develop a small molecule drug, targeting, interrupt, functioning of this so-called uncle protein, the protein causing cancer. Then the CML, the chronic maloid leukemia can be cured to 99%. So that means for every research question, every disease, there will be multiple proteins be involved. And our goal is to find out what are the proteins for each of these questions. Interesting. So another one of them, let's say, if we zoom in for a bit, is you mentioned the discovery process for things like treating diabetes. And so do we zoom in and we look at where, like, for example, we can find biomarkers at the proteomic level that show us where maliglances have developed in diabetes, and then which pharmaceuticals and how to design those pharmaceuticals to most optimally decrease those malignancies at the protein level. Okay. Diabetes is a very special case. Totally. There are different types of diabetes. There are different causes, and they are also affected by the environment, your lifestyle and your food and exercise and so on. It's not like cancer. Cancer, we call the malignancy. Malignancy means that it's cancer. Just because of one cell become, let's say, initially the normal liver cell, now it becomes liver cancer. They proliferate from one to two, from two to four. But diabetes is a metabolic disease. It's a systematic disease, very, very different. But what do you mean is some diabetes definitely occurred through the difference or disfunction of some proteins. And by correcting the function of these proteins, we can alleviate or even cure diabetes. This is a very interesting direction. And we have a project is not to cure diabetes, but to predict diabetes. We have monitors, about 2000 individuals over more than 10 years. Every two or three years, we collect their blood sample, their stool sample and urine sample. And we are going to predict, now you are healthy, but how about 10 years later, we'll get diabetes, we'll get hyper hypertension, and so on. So we build a protein-based AI model, also plus some of your basic phenotype, like your weight, your wisdom, lungs, your BMI, and so on. Using these clinical characteristics combined with protein, we can reach about 80 percent of accuracy to predict your chance to get a diabetes in 10 years time. Okay, so there's a data fusion of different components for prediction of diabetes, and then potentially the pharma discovery process for the healing of those dysfunctional proteomic components. Exactly. There are a lot of AI technologies using our world to do some prediction. You predict the pandemic, you predict the traffic, and all this prediction is based on data. And this data mostly from the micro welds you based on experience, and sometimes what we call intuition, we predict something by intuition. But in our research, we're trying to predict based on the molecules in the micro welds, proteins and other molecules. So this is entirely new new disciplines, new, well, because in the past we have very little idea how these proteins remove, what are their change in different physiological diseases. So now we have technologies to measure them, and we generate the data of the teams. And then we can apply AI. So that's why my lab is called proteomic big data. When you normally we talk about big data, you are probably thinking about the text, a lot of web pages, a lot of pictures, a lot of video and voice, music and so on. And now we are saying there's another type of big data, which is proteomic big data, which is the key of life activity and the basis of our life. And they are changing and they're generating data. And this data is the key to diagnose disease and to treat diseases. Yeah. Yeah. And even to get insights into diseases, even developing before they become cancerous or deadly, basically. So that's key. Yeah, I love this. So let's see if we can zoom out again, because I love how you described that the proteins are synthesized, they perform their functions, then they degrade. And then they can live for hours or days or even months. I love how they're just moving from part to part of the cell and potentially performing different functionalities. And just understanding all of that is in itself so fascinating. So there's the muscle cells. And in muscle cells specifically, there are the proteins that enable actins and myosins for the musculoskeletal movement. That's so cool. So muscle cells, so each cell group, so from the stem cell are all these different potentials for cells. And then inside of all of those cells are different proteins because so the DNA inside of all of those different cells is encoded slightly differently with small, small, small variations so that it can then make proteins for the specific functionality of that cell type. Can I make a correction here? Of course, yes, 100%. Yeah, let's say you took a muscle cell and a white blood cell or red tumor cell. The genome, the gene actually almost the same. So it stays the same, but the transcripted area is what's... Yeah, okay, perfect. Yeah, that means... Yeah, so it's like, it's like out of the all of the base pairs, there's like in each different cell, you could say there's like a little like flashlight on the area that's doing, you know, that you're only illuminating a specific portion of the sequence for this functionality that's needed of that cell. And in a different area, like neuronal would be a different portion of the... Exactly, and also protein protein are being synthesized at different speeds, different frequency, a little degraded through different mechanisms. And then the protein which we can detect is the synthesized minus the degraded. And this degradation system also controls the protein expression, independence of the genes. So it's very important, we want to really understand how life works that you need to measure proteins. If you measure DNA, you can hardly tell the difference of the muscle cell and the white blood cell and the cancer cell. Yeah. So there's also a like a cataloging of from the genomic level which regions are being transcribed into the creation of proteins in which cells, cell types. So there's like a big library of end of all stages. So there's the 10 year old or 20 year old that has the very healthy transcribing and protein development and no pathologies that have developed. And then there's let's say the 40 or 50 year old that's beginning to get some of the earliest signs. And so there's this big library of data, of big data and this AI powered analysis of that data to see like, oh, there's a little area in this specific cell type that is looking like that there's a dysfunctionality at the protein level that's starting. And that's going to cause a lot of other downstream bad health effects. And so so then from that, let's say from that library of let's say that you could have that we could have. Then there's also the like how to solve the dysfunctionality. So it's not only when the dysfunctionality arises, but also the solution to the dysfunctionality that also differs for different like proteins and different cell types in different regions of the body, etc. Is that also how you see it? Something like that? Yeah, but I can see the library that you mean, probably mean some location in the genome. Yes, you know, this location may be for protein A, as protein A may become dysfunctional and you're talking about, OK, now we located in this location of the genome. And then now we can find some cure for making protein A more healthy. This something actually I respectfully wants to correct. Yes, yeah, yeah, because most people actually thinking about life from the genome, because the genome is origin of the life. You start from the egg and the sperm from your parents. And this isn't a genome, but you are individual after 20, 30, 40 years of growth. And now you are individual and this is individually composed of mostly protein. Most proteins are not well, because you are also well understood after the protein is translated from RNA and the protein has lots of other additional genome independent modifications and dynamics. This completely independent of a genome because genome, we have plenty, three pairs of chromosomes and they are mostly sleeping in the nucleus. Our brain cells is said that they never or they seldom divide. You have this brain cells and they are the genome never been reactivated, they're open, they're sleeping in the nucleus, unless they're tumor. But then how can you get a function of your brain by analyzing the genome which is sleeping in the nucleus? It's not the correct way to do this. The correct way is to analyze the protein, which are the gears of the other actions. And let's say protein A, although it's from the same location of the genome, it can be in the nucleus of the cell. It can also be in the memory of the cell. It can just locate in the floating, it can also secret out into the outside of the cell and go to the circulation and take the functions. And let's say if you want to make protein A more healthier, then we need to have a drug or whatever treatment. And these are actions through binding to the protein A or proteins related to protein A and so that it can affect the function of this protein A. And protein A can also change if you move, if you have a high temperature, if you are cold, all this can change the function of the proteins. So actually we are trying to tell the community, not only the scientific community, but also the social community, that actually we should study, understand and treat the disease from the perspective of protein. So we are now launching a big project we called Proteomics Driven Precision Medicine. Something like correcting the inherent or the older way of thinking about the diseases. You know, the diseases are classified according to like a WHO guideline or other guidelines based on the patient's symptoms. If you get a fever, then this type of disease, if you get overweight, it's another type of diseases. And also based on pathology. Pathology means the morphology under microscopy amplified by 200 times. And more and more diseases are classified based on the genome. Like what you mentioned just now, some location of this genome, which is eventually translating to a protein A, maybe it's the cause of the disease. And based on that, we can predict the disease progress prognosis. We can also decide which treatments used to treat this disease. But frequently people found based on the symptoms, based on the pathology, based on the genome changes, we still cannot precisely predict the behavior of many diseases. And now because you have not measured directly the proteins which are the gears of disease and line. So but why why that? Because to measure protein need a very sophisticated technology. And the kind of occasional protein is very challenging, technically challenging. Okay, I love this topic. I love this topic. So how do we Yeah, how do we actually look at proteins? Because we have to go even more than 200 X zoom. And we have to find some way of like, illuminating them, but still being able to like capture the data in them, and also how they function. So how they like express themselves. So and how do we, how do we see that? And like, do we, do we have that like a molecular level? And then how do we like also see the, how do we like color it to see the different, yeah, the coloration of it, all of that. Yeah, to differentiate it basically. Yeah. So conventionally, we use antibody to measure each protein. So antibody means if you know, you know, you have your vaccination, you need antibody against the dead virus or the virus protein. And this antibody will bind to the virus of let's say, SARS-CoV-2, not to another coronavirus, but only SARS-CoV-2, because this antibody has the specificity to bind to a protein of the virus. And this is called affinity binding. And this immuno affinity, this is the basis of using antibody to measure any protein you want to measure. So there's a lot of technologies like what a carer is doing is to have antibody and they bind to a particular protein. And this protein may be in the brain, maybe in the tumor, and we can use a microscopy, give it a color, a pseudo color, that you can see it. You can see this protein is a nucleus, not a protein membrane of the cell. But you know, these are the limitations. Every time you can measure only one protein. And to measure one protein, you need to consume one slide of the sample, one piece of sample. You know, they are, I already told you there are millions of proteins in each cell. So Tiananis, is it still the most common for us to take in antibodies that then have the like the genetic sequence for us to be able to like when they bind to the target protein. So then they have to go to the target protein. So they have to be coded to go to the target proteins delivered there. And then when they bind the antibody to the target proteins, let's say, then is there like an optogenetic component or something like that for us to be able to have an illumination of the protein to be able to read it? Is that still one of the most common ways or what's Yeah, that is the most common ways has been used for many dozens of years. People develop tens of thousands of antibodies to study almost every protein that they know in a human body or almost. But there are due to these limitations, it's very difficult to quantify it. You know, you can see it. But how much of it is another question. What are they? It's first question. What? And second question how many but to to get how many you have to get how many because there are many proteins that express the both in the belay and the malignant cells. The difference is the amount of the protein. So to measure how many the best way actually is using mass spectrometry. And this is the technology that we're using here. So mass spectrometry like for example, I can recognize you atlas by seeing by your nose, face, eye and the year and so on. But I can also identify you by your weight. If I can measure your weight with 20 decimals or 30 decimals. So with mass spectrometry, we can measure molecules with 0.0030 zero plus one kilogram the site position at this level that we can simply by measure the weight, we can get the identity of the protein. And this is all this is a theoretical basis for mass spectrometry based the protein measurement. And with a tiny amount of tissue, which you can hardly see, we can measure tens of thousands of proteins. And this cannot be achieved by any other technologies right now. And with this technology, then we can actually measure many hundreds, thousands of samples, which are relying on another batch of sample, which are malignant of different severity that we can use AI to find out what are the different proteins and how much are they and find the biomarkers. And to build models to distinguish them. So building a models to distinguish at the level of 30 decimal places using mass spectrometry, so spectrometry. So so so then so the different molecular weights, basically, will tell you which molecules are present and and then and then how much of them also in the protein. And then if so, then if there is like a 17 year old, let's say they're supposed to be like a normal level of all molecules present in the protein. But if there's a 45 year old that is beginning to have one amount of molecules higher than the other, that could be an indicator for pathology dysfunction. Yeah, yeah, we also have a project of aging. We measure the proteins in different organs along the lifetime of a mouse. And then we can see differences. How different organs getting old and when they get old, we will never get old. What are the proteins which are changed? We will stomach getting old. We are brain getting old. What are the difference? Yeah, that's so cool. And then so then so I love that. So the stomach, the brain, the muscles, just the different areas when they get old in this library, you could say is the data that shows that in those different regions, there's usually this a molecule will get too much of in the protein or whatnot. And so there'll be like a pattern analysis that will have in this library to predict dysfunctionality. So would you say that when we when we predict the dysfunctionality? Let's say I have my levels, you have your levels. There's young kid levels. There's older people levels. And they're all interestingly different, even in the same, let's say, area of the stomach cell and that the stomach cell has a specific composition of of molecules that is supposed to be there at a healthy state. And then when there's a a dysfunctionality in the at the at the level of molecules, again, it's so cool. So mass spectrometry will tell us that there's that there's that dysfunctionality developing. And then so then what would be your solution then? So is this where the the AI big pharma component comes in to to do this like testing on those regions? That kind of thing to see if it can bring it back to homeostasis or what? Yeah, so, you know, these or agent actually evolves from the molecular basis, basically the protein level and to the cellular basis. So that which means when you're getting old, your protein change first and then you will at the time you don't see any other difference by looking at your face. You don't see any difference by looking at your finger after under microscopy the cell, they're the same or proteins getting old. But then after a certain threshold after reaching some threshold, you begin to see the difference at a cell level. And then after a certain time, then you start to see from the macro level your face or your hair become seems different. So if we have the ability to if we have a glass, OK, we can see if a camera can see you from the other end of the world. But if we have the proteomic technology, we can see changes, so subtle changes which will precede the macro field change. And this is the all the basis for us to manage diseases. For example, if we are cancer patients, they are already diagnosed with cancer and we want to know whether which drug should be used to treat this cancer. The ordinary practice is treated with drug A because drug A is effective in 80 percent of people, patients like him, like this patient. Then we treat it that after three months we do a CT or six months do a CT to see whether tumor shrink initially is so big, now shrink. If we shrink, OK, successful. We should continue to use this drug. If it doesn't shrink or even grow, OK, we waste three months. So the idea is if we have a cutting edge proteomic technology, an AM model, then we can maybe tell at one week or two weeks before the tumor shrink or grow. We only can see protein level change because every drug they're actually through some proteins or multiple proteins. And the key is we don't know which protein it will, the drug will bind to because we know some of them. We know maybe every drug developed have some what we call MOA mode of action. We know something about it, but we don't know everything. We don't know how much change this drug will induce to affect the protein. And this is what we are doing. And this change can be quantitative change can be used to build an AM model. And this model can predict whether it shrink three months. And this is the principle of our vision. This will work in clinic. And this is also what we are trying to do with a few clinical trials with the ovarian cancer with the trigger negative breast cancer and thyroid cancer and so on. Interesting. So in the later stages, like you're describing now is the like is developing us as best of a pharmaceutical remedy as possible, deploying that and then analyzing the cancer and seeing if it's decreased. So that's later stage. And then earlier stage is potentially the like as best as we can, creating like computer simulations of the of what it would look like to to bring in some sort of like, you know, everyone's been, you know, feeling the whole thing of like, what can I actually like take that can like boost the biology of my vehicle? Like what like what could actually boost like the entirety of it? Not only cognitively, metabolically, etc. All of it. And so is there something that we could be in a sense even before it gets to cancer or tumor or malignancy? Could we go even like running these simulations? Let's say where even at the beginning level of like a molecular increase or decrease from the normal, let's say from what it's supposed to be that then we could have in that simulation. Is this pharmaceutical or is this specific remedy that we're applying? Is it actually able to bring us back to homeostasis to like to full health or even better than that? And so I love that. And that that's also that's also really exciting. And it feels like it feels like it feels like anticipating the future more than waiting until the later stages of the disease kick in like basically tackling aging at the level of like you said, protein dysfunctionality before it even like really as it's starting basically as it imbalances. Yeah. Yeah. Yeah. So the human being is a huge machine composed of trillions of cells and each cell has a countless proteins. If we can measure the protein, we know the secret of how the machine runs. Now we can predict the microbial behavior of this huge machine. Yes. But this is not easy. So this is our vision. We still need to accumulate sufficient data, big data before we can have a reliable model to predict the central sophisticated system. So, you know, we have more than seven billion individuals in the Earth on Earth. And we are more than eight now. Eight now. OK. But there's still a small number compared to the proteins in one single cell. You think so? Are there more than eight billion proteins in a single cell? Sure. Yeah. Really? I heard last time that it was like thousands of different proteins. But so so there could be like you're saying there could be like millions of each one of those thousands or something like that. Yeah, there are a huge number. I didn't know that we were talking over billions in each cell. Billions of proteins. Yeah, there are other. It's estimated and nobody knows how many, but you know, there are millions type of proteins. And for each protein, there are multiple copies. Yeah, interesting. So millions of millions of types. And then there are multiple copies. Cool. That's crazy. Oh, in each cell. Each cell. Yeah, that's nice. OK, so I have a couple more questions. So one of them is. Do you want to share anything else about the way that like Westlake is like about your lab that has about I think you said 30 or so people right now or also about Westlake omics and how it has like 80 or so people right now and how you're providing these conventional applications, translating the scientific insights. Do you want to talk about anything else about that funnel or that process or about how you're like recruiting for it or creating like international relationships for it? Do you want to share anything else about that? Yeah, very good question. Actually, we collaborate a lot with international collaborators. We have actually very nice facility and also exciting research topics. We hope more people could join us in either in Westlake University, Westlake laboratory, also the Westlake omics for this exciting research and application. So I'm also the general secretary of Chinese Society for Human Proteal Organization. And we are gathering a bunch of exciting passionate young scientists, their graduate students, their post-docs. They are willing to communicate with other colleagues, peers outside China. And we are initiating a big science project with Professor Fuchu He in Beijing, which is funded by most Minister of Science and Technology in China. We are really hoping to collaborate with clinicians, healthcare practitioners, and AI scientists all over the world to move forward in this exciting proteomic data research. You know, this cannot be achieved by a single group, by a single country and the benefit will not only belong to Hanzhou or China for the sake of the entire human being. After we solve any disease like salary cancer or leukemia or viral cancer to be helpful for anyone in the world. However, there are indeed a lot of challenges. You know, most people, they understand the genomics. And very few people really know what is a protein, how many proteins are there, and how soft-based they are there, how dynamic are there, and how to measure them, and how can we use that for disease diagnosis and circuitry. And these are all open questions and we need an education so that more people can understand. And this is particularly important for application of this technology in a clinic, because you know, a clinic diagnosis is not something like an iPhone. You invent an iPhone and then you can sell it in the market. So we develop a biotechnology, we develop a drug, we need to get a proof from the authorities like FDA in the US or in the Chinese NPA and so on. And we need to let them understand what's the gears of life and what's the molecular basis of the therapeutics. And this will need some time. It will not take a long time, but you know, we still need some explanation why why protein is the molecule that we should measure. And AI also, some people have concern. So AI, most people think, okay, it's powerful, but it looks like a black box. What's inside AI? Can you explain to me how do you use this protein and then you get a score, for the risk score? But this, you know, this need a different logic. And many people think mutation of a gene is the cause of a disease. Then if you measure protein, what's mutation inter-corresponding to? And you know, for this type of question, and we have to tell them, okay, just like what I did to you, the past one hour, the protein have a different logic. Proteins, there are many copies that changing, they have been set aside and they've been integrated. And all this, we need help from you and many other colleagues from different fields. First, you should maybe understand this science and understand now we can do a lot of cool protein measurements. So I started to protein research more than 10 years ago, studying Singapore. At that time, the protein measurements is very expensive, much, much more expensive than what we have now. And also not very precise. So now everything changed. But many people, they are still some of the scientists who has done proteomics in the past. They have this older impression, okay, proteomics is expensive and not very precise. And they didn't know noise changing. So we are talking a lot to scientists in China. We have been interviewed for by many times to explain this after normally after one hour or so, they will understand now his difference is a different field. And they'll have a lot of cool technologies. But to the international community, we also need such education. So I'm also involved in HEOPOL, which is the International Organization for Human Proteomics. And they are scientific peers from different countries. If you look at 20, about 20 years ago, proteomics get a lot of support from academia, from industrial. And they are having very good research outcome. And they have a high expectation from many fields, saying that proteomics going to change the world. But during the past 15 years, there are a lot of frustrations because of the immature technology and overclaims and so on. So in many countries, actually, people are not very supportive of proteomic research. So this makes many of our peers in other countries relatively tough. But recently, you probably heard President Joe Biden, he relaunched the Cancer Moonshot Project on the second of February. And one major project in this initiative is called CPTEC, the Clinical Proteomics Analysis of Cancer Tissue Consulting. And this project actually is mainly proteomics. And the director, Henry Rodgers, is also our long-term collaborators. When I was in Zurich and Sydney, actually we joined Cancer Moonshots via Switzerland, UK, and Australia. So President Biden also joined our Hupo meeting I think in 2019, I don't know, sorry, it's 2016 also. So he spoke in the proteome meeting, annual meeting, and also held a great promise for the field. So with your kind interview, I hope to also advocate the more understanding of the proteomics field, which has changed dramatically during the past couple of years. And now we really have a practical technology to use for clinic. And we have a lot of companies like our company, Westlakeomics, are equipped with cutting-edge proteomic technology and AI models, which can provide practical access to treat some diseases such as thyroid nodular diagnosis. So we also hope to expand our... So before this test can be a proven clinic as IVDE individual diagnostic tools, it will also be LDT, lab development technology, LDT. So we hope also to test our LDT as a research phase in different places in China and also maybe overseas. Our friends in California and Stanford and we are talking to them because they have a quite mature industry for LDT. So in contrast to China, it's just starting to develop LDT. So regulation is not really mature. So hopefully through the collaboration, scientific collaboration between different countries, China and the US, they are not competitors but they are actually, we can, our scientists can work together to promote the healthcare, like President Joe Biden said, he wants cancer, mortality decreased by 50% or 25% in 25 years, with 50% by 25 years. So no country can do this by their own and the benefit will benefit the entire human being, not only the United States, but also Europe and China. So I hope through the collaboration between scientists and technologists, I have so many collaborators in the United States and I really hope you can, when they come to here and see what we are doing and also have some of our voice heard by not only the people in our field but also by the ordinary people, by the patients to have them aware of this new technologies. Yes, yeah, beautiful. And I feel like that's one of the main goals is to make it so that we don't have data silos around the planet, but rather that there's data interoperability across labs and across institutions and then that we can leverage that data interoperability to be even more like a single unified planetary force and do things like decrease the amount of people dying by cancer because we've increased our data interoperability, very similar with like cellular communication as well. So Lino, yeah, on the micro levels, the same thing on the macro level as above so below. And the other thing that you mentioned that I really respect a lot is the patients, right? So the people and the average like population who would love to hear and learn more about this because it's going to be so applicable when let's say they have a family member going through a disease of some sort for them to know, oh, hey, I remembered that there's actually these important labs doing this work and maybe I can get involved in what they're doing so that my family member can live longer so they can be healed or whatnot, whatever it is. So I love that also. So yeah, hugely excited about the international collaborations and data interoperability and success moving forward as a unified planet as well as getting this out to the masses via even content like this and beyond just making it more and more friendly and simple for us to understand about the advances that are happening and how they can be applicable for our lives. So we're going to host a meeting in collaboration with New England Journal of Medicine and also Cancer Hospital in China. We are focusing on molecular diagnosis of thyroid cancers and we have invited all the opinion leaders in China and also in the United States, in Harvard, in Montmarie, in Snow-Catering and also people from California, professor in Japan who is a major contributor to the WHO guideline for thyroid nodule diagnosis. So we will meet together some of them virtually because of the pandemic. Hopefully next year we can gather physically. So we're going to talk about counter status of thyroid nodule diagnosis and what are the existing technology without the limitations of any and what are the emerging technologies to have a better management of this disease. And we're also trying to interview some of the patients. There are so many patients in China and actually if you're interested, you probably can join our events and do some things, some interview some people in the States or I believe a lot of people will have thyroid nodules. You just ask them what they think about it. Have they cut it? Do they have concern about it? And what are their concerns? And we're also trying to make some cartoon in both Chinese and English to explain what's the thyroid and what's the biology and what are the disease of this thyroid organ? And I think that would be very, very interesting. Every year there are millions of people have their thyroid cut out, removed. And after removal, they found that actually they didn't have to. This is simply because there's no precise diagnosis method. Yeah, and this sounds great. This sounds like an awesome upcoming event in March for doing exactly what we just said, bringing us closer together, creating more interoperability between data, healing more people around the world. So that's great. And I do feel like there's gonna be a lot more opportunities for us to collaborate and create together and just to, you know, re-spark our relationship with West Lake as well as China at large. And it's exciting for me to bring a greater sense of unity together and meeting basic needs and actualizing everyone's full potential. And it just, you know, that's what drives my heart and that's what drives our heart. That's why we do what we do. So, thanks so much for coming on the show, Tianan. Thank you, brother. Oh, thanks so much for your time. I'm really glad to talk into you. Haven't talked into foreign friends for a long time. You know, such a casual way. Yes. You actually invited me to talk a lot more than what I have thought about. Nice. Yeah, that's always also exciting is when you feel like we had a good conversation for your further ideas and creativity. That's always very important. Yeah. Yeah, you have a very nice summary every time when I talk about something and this inspired me to talk about more. Well, yeah. That's great. And also the nuances where you like helped correct and make it more clear that I was actually knowledgeable about the central dogma of biology. So just, I appreciated your corrections also. Those were very important. Yeah, that's my job. Yeah. Do you have any website of your company that I can learn more about what you are doing? Yeah, of course. Yeah, so one of the main ones that I'll send you is the No Limit Society. So that's the main thing that we're building. And then of course, Simulation, which is the show, the show that I host. So yeah, so I'll send you those links, of course, and I'll have them in the bio as well. I'll also put the links to the gulmix.com as well as the westlakeomix.com links in the bio for people to check out. And also if you guys would like to, you can leave us a comment below with your thoughts on the episode. We would love to hear from you. You can also like the video to help the algorithm, subscribe to the channel if you haven't yet. Share the video. If you think this conversation about proteomics, about big data, if you think this has been helpful, go ahead and share this with other people that you feel like this would resonate with. And check out the links in the bio below to Tianan's work. And if you wanna get involved in their international collaborations, if you wanna just reach out to create together with them, feel free to dive deeper into what they have available online. And yeah, I think that's it. Do you have anything else you'd like to add? Yeah, this is very nice. And I have a team, now they are not here. Later after you send me this information, we will, I will work with my team to see how to collaborate with you in the future, maybe in more extent in the March meeting or more events that we're going to do. So first of all, read your information. And then let's think about if we have more opportunity to work together, we can schedule another meeting to discuss part of it. Yeah, sweet, sweet. Okay, thanks. Yeah, I'm gonna end our, just the recording. So just stay in the room for one more minute with me, okay? Yeah. Okay, thanks everyone. Thanks for tuning in, we love you. See you soon.