speaker-0: you Hey Momo, I have a question for you. All right. Do you know why DNA makes an absolutely terrible cook?

speaker-1: DNA cook. ⁓ I don't know, but I think you're going to tell me.

speaker-0: It's because it just sits in the nucleus, holding a giant book of recipes, but it never actually cooks anything. It just watches RNA do the actual work.

speaker-1: That's a good one. That's a good one. That's a PH dad joke with a twist of transcription.

speaker-0: Right. You know, more importantly, it's actually a biological fact. speaking of people who actually do the work in the kitchen, I think we are going to announce something really exciting today, which is the winner of our monthly challenge. Go for it, mama.

speaker-1: Yes, we have the winner of the March 2026 VIO monthly challenge. And her name is Zoe Miller-I6R. And she responded to our challenge by posting a recipe for the Mediterranean Chicken Chickpea Rice Pilaf with a salad. And she actually provided the receipt from the store and added it all up and divided it by the number of meals. And it turned out that this delicious and nutritious meal is only $3 per person, which is amazing because it looks so great. And it beats all fast food. It beats all processed food, all boxed food, all of that. Look at how delicious that is. So Zoe Miller won the Viom's FBI, or full body intelligence, which is our premium direct to consumer kit. And it was delivered to her last week. So congratulations, Zoe.

speaker-0: Congratulations, Zoe. And ⁓ this is also a great place for us to announce our next VIO Monthly Challenge, which is the April 26th VIO Monthly Challenge. And it's very, very simple. What you need to do is to post on any social network your best personalized nutrition insight. This can be your experience. It can be a publication. It can be a link to a resource. It can be whatever. And once you post that, Take the URL to that post and put it on one of our podcast episodes or send it to podcast.vyom.com. In your post, make sure that you include a hashtag, hashtag two PhDs on a pod. We will pick the winner based on the content and the engagement of your post. That's it. Very simple.

speaker-1: Amazing. Love it, love it, love it. OK. So now, speaking of personalized nutrition and the April challenge, we are going to discuss today about the differences between DNA and RNA. The world has been so focused on DNA over the last 25 years since the Human Genome Project was completed. And all the tests are focused on DNA. And we're going to show you many examples of how DNA compares to that cookbook. How many people do you know? that eat out all the time and you go to their kitchen and they don't cook anything in their kitchen, but guess what? There's a bunch of cookbooks there. Those cookbooks are basically the- But you cook a lot. So you use them. But there are people who have cookbooks, which is basically comparable to DNA. You have DNA, but it doesn't actually do anything. It just sits there and has recipes. It's the RNA that's doing the work. It's converting those recipes into actual meals. And so have to activate RNA in order to eat healthy, nutritious, delicious food. Otherwise, it's just a potential. Everyone can buy a cookbook and has a potential to cook delicious foods. But until you do, which is with RNA, you're not going to get to eat anything. So I'm Momo, a biochemist.

speaker-0: I am Guru, an AI expert. We're two PhDs on a pod. And today on two PhDs on a pod, we are breaking down why DNA is not your destiny. We're going to explain the difference between your genetic code and your RNA expression and show you how your meal, you know, that could be one of those meals like Zoe just gave us, that is literally telling your genes what to do. So Zoe is like the action-oriented RNA, not just the recipe that's sitting on your kitchen shelf. So, mama, let's start the episode. But before we start, as we always do, I want to ask you what's on your radar.

speaker-1: I'm glad you asked. I'm going to share what's on my radar. So this is another example. This is just another typical meal in our house. And ⁓ just another example of how you can so quickly make such a delicious and nutritious meal. It's incredible. So this meal is completely vegan, has no dairy, has no meats, has no animal products. And you can see here, all we've done is we have some whole wheat tortillas. We have beans in our house every single day. We have them all the time. And so we have two types of beans here. The bottom right is refried beans. The top is just cooked beans. We have some cilantro. And then the flavors are provided by Mexican spiced tofu and Mexican spiced grilled vegetables. Those are on the bottom left. And then you just assemble your burrito and you enjoy it. It's absolutely fabulous. So this is well under $4 per meal. And you just enjoy every bite. And then I also want to bring up a super exciting topic in that Guru and my teams have just published a peer-reviewed article. And this article shows a lot of data that we gathered over the last couple of years that examine the stability of the gut microbiome as quantified by RNA and not DNA. So we're focusing on the stability of both the composition, or who is there, and the function, or what are they doing, aspects of the microbiome. And I'm just going to share one takeaway slide, which is very important to answer a very commonly asked question. How often does human physiology, and specifically, how often does the gut microbiome change? Meaning, how often should I retest? And so we asked this question on a scale of something like 5,000 plus people. And you can see the data for those 5,000 plus people. And on the left side, we're showing the p-value for the difference between the gut microbiome at time point 0 versus the number of months that are plotted on the x-axis. And you can see that that p-value drops below the red line, which is the significance line, after about five to six months. In other words, what this is saying is that the composition of the gut microbiome significantly changes after five to six months. And then on the right side, we did the same thing, but here we're actually plotting the biochemical functions of the microbiome, not the composition. This is obviously specific to Viom's test because other tests cannot measure the functions. And here you can see a very similar trend that the p-value drops below the 0.05 line after about six months. So basically the takeaway here is on average, you should retest your gut microbiome after six months. That's it. Good-o. What's on your radar?

speaker-0: Well, this paper is on my radar too. And I want to emphasize one key point from that paper, which is that all of the data that we present in this paper, which is actually from about 6,150 customers of WIO, we present this data from a meta transcriptomic test. This data is very unique. You know, when we talk about 6,150 samples and we compare it to the total amount of meta transcriptomic data that's out there in the world right now, which is probably in the order of a few hundred. This is already a huge data set, but guess what? This is still a tiny fraction of the total number of samples we have collected and processed in Viome, which is more than a million samples from more than a hundred countries. So the kind of data that we have at our disposal to ask and answer all the difficult and interesting questions that we've always been asking. In science, that's all in our hands. And I think we'll be talking to you a lot about that data. Speaking of which, I also had one other thing on my radar, which is the phenotype data that we get from Viom's user base. And by phenotype, I mean the external characteristics of an individual. could be your demographics, your lifestyle, your symptoms at a given point in time. It could be the medications that you are using at the moment. It could be the surgeries you've had in the recent past. It could be anything that has to do with what you feel and what you can see in your external characteristics. We have a ton of this phenotype data and that is a compliment to the molecular data that we generate from Viom's customer samples. So when we are able to put this very large molecular data set on one hand and this very large phenotype data set on the other hand together, we are able to get insights that are unparalleled. And we're to be presenting a lot of that data here. Some of it we will be publishing, of course. And when we publish a paper, of course, we will be the first ones to tell you about it. So you can go read it yourself and you can ask us questions and so on. I was just doing some work with my colleagues and team members this week. I just realized how rich our phenotype data set is, how many different types of questions you can ask from it and answer from it. So that was a very good experience for me to just know that there's such rich data. So speaking of phenotype data, Momo, I want to get back to the main topic of our podcast episode, which is health. know, health A lot of people think is driven by of course, your genetic predisposition, which is what you were born with. Right. And there have been companies in the last 20, 25 years, which have told people that you take a cheek swab from your mouth and you send it in and they'll tell you your DNA. Most of the time, you know, all they tell you is your ancestry. They're not telling you much about your health. And even if they do, it's a tiny bit. Momo, the first question I want to ask you is, why is DNA that everyone has been so focused on for the last 20 years, why is it not very informative?

speaker-1: Yeah, that's a really good topic. And we're just going to briefly mention it here. But I would say as the highlight, it would be that we have now established as a scientific community, the DNA plays a minor role, and most of the time, very minor role, as a contributor to chronic diseases and cancers. And that has been established via many studies, including genetically identical twins. So large studies with genetically identical twins show that chronic diseases are actually environmental and not genetic. So that's one main point. The second main point is that there's still a lot of health care professionals that are propagating the myth of it's genetic. And so even recently, I had a family member who was just diagnosed with pre-diabetes. And she went to a doctor, and the doctor basically told her, take these pills. And when she asked, can I reverse this or can I slow this down with diet and lifestyle, the doctor said, no, it's genetic. And so you just have to take pills and live with it. But we know from so many studies that diabetes is a reversible disease and that diet and lifestyle play a major role in diabetes onset and progression. And so this is just one of many examples, but there are many others that we could, I'm sure we will have episodes just focused on this topic. So as we discussed earlier, DNA just represents your risk, average risk of a person who has that genetic profile at birth. And that's all it means. It's a risk. It doesn't mean you will develop that disease. And if you do develop that disease, the fact that you had a risk doesn't actually tell you what to do about it. So maybe its benefit is mostly in making people aware and maybe making people pay attention to it. Now there are rare genetic diseases and these affect about 5 % of the population where genetics does play a significant role. And so we're not discouraging people from doing the tests. But there are many other examples where there are currently genetic tests, but they provide very limited utility. Whereas RNA-based analysis, so gene expression-based analysis, are far more relevant to the human health. And we're going to hammer down that point throughout the next 30 to 45 minutes.

speaker-0: Yeah, Momo, I have understood in the last maybe 10 years or so after diving so deep into this kind of data that the vast majority, like 90 % of human disease is caused by lifestyle factors and environmental factors, which is basically RNA gene expression. the idea, and I've had family members who have come to me and said, ⁓ my God, know, my genetics are not very good. I might have heart disease, right? No, heart disease is not because of genetics. mean, there's a potential something might have happened because of genetics, but really what is going on is it's your lifestyle. It's your diet. It's your stress. It's the amount of sleep and exercise that you get. It's all of those things that have contributed to developing cardiovascular disease over years and years, sometimes even decades. And now that's why you have heart disease, right? So that's the key point. So that is happening because of gene expression or RNA. So Momo, I want to ask you, you're the biochemist. So I want to ask you about RNA. What is RNA? How does it work? And what effect does it have in the biochemistry of our body?

speaker-1: Woohoo. Let's dive in. This is super exciting for me because I started studying RNA in 1997. And that was basically, my PhD thesis was RNA. And it was RNA in many, many functions. And RNA is super, super exciting. And we should have a monthly book recommendation, Guru. And I think you had a book from Tom Cech that I thought was fantastic. I think- Yep. while I'm jib jabbering here, maybe you can look that book up and we can actually tell people about it. Because Tom Cech was one of my heroes. won a Nobel Prize in 1989 maybe-ish on RNA. And so he was one of my heroes during my PhD. anyways, let's dive in. What is RNA? Because most people are not familiar. And it's a very simple concept. So here I have a very simple graph showing that DNA, people are familiar with. It's ⁓ helix, know, double-stranded helix. two molecules are hybridized together or bonded together. And that is what you inherit from your mom and dad. And you inherit three billion of these nucleotides from your mom and 300 billion from your dad. ⁓ And so that's your genetic footprint. And that's a potential. And we're going to go over the next few slides to show you what that potential means, what kinds of possibilities there are. But if you actually, so here in this lighter blue color, we are showing ⁓ a gene. That gene is literally a physical object that exists in space. Like you said in your joke, it's in the nucleus, and it's physically there, but it doesn't do anything. It just sits there. And it can sit there forever without doing anything. But when we want to express the function of that gene, when we want to activate it, the first step is this transcription. So the gene is copied by the cellular machinery into RNA. And that means, aha, now we're going to activate that gene. We're going to make a functional molecule. And the number of more RNA molecules that are copied from that gene determine how much of that function will be activated. And that's a core concept to understand that when we talk about the Viome test, we're actually counting these RNA molecules, literally counting them so that we know that in person one, they have no RNA molecules from this gene. So it's completely silent and it's not doing anything in person. Two, there are a few of these molecules, and so the gene is active, but it has very low activity. And then in a person three, there are hundreds or thousands of copies of this gene, meaning hundreds or thousands of RNA molecules. And now that person's really activated that gene. So our method that we will discuss later is quantitative, and we know which genes are activated and to what extent, which is super important. Now what happens after this is very interesting and we can spend a whole episode on this, but I'm just going to highlight it. About half the genes in blood are translated. So that's this process translated into proteins. And these proteins are the ones that have enzymatic activity or structural activity. So they will perform chemical reactions or they will form the structures in our cells. They can also do some other things like bind and activate other proteins. or bind DNA and block it. So there's other functions. But what's not shown here, and Tom Cech's book definitely describes that and more, is that about half of the RNA molecules expressed in blood are non-coding. That means that they have a function as an RNA molecule. They do not need to be translated into proteins in order to have a function. And these are called non-coding RNAs. And there's a whole world of these. And they were discovered much more recently than the traditional messenger RNAs which code for protein. And so we can dive at some point in time into these RNAs because it's a whole lot of dark matter, but some of the ones that we've discovered play extremely important roles in our health. How awesome is that?

speaker-0: Yeah. ⁓ my God. This is such an awesome topic and I love that book by Tom Cech and I learned so much from it. I could just go on and on. I'll actually bring up a couple of points from that book. But before I go there, I want to ask you a couple of basic things. You mentioned nucleotides. Just tell us what are nucleotides? And second, you mentioned functions. And when you say function, what do mean by it? Just define those terms for us, please.

speaker-1: Okay, great. Yeah. So nucleotides are these little notches you see here in the RNA. So they're the building blocks of both DNA and RNA. So DNA has the oxyribonucleotides that build up DNA and people are very familiar with the bases called A, G, T, and C. And in RNA, they're very, very similar. They're not deoxy, they're just ribonucleotides. So they're missing that one oxygen. ⁓ And ⁓ they are, they're also A, G, and C, but T is replaced with U. So they're slightly different. That's in terms of the structure. So notice that DNA is double-stranded, whereas RNA is single-stranded, and that they have those little chemical changes. And then in terms of functions, so regulatory functions and structural functions and enzymatic functions that RNA can play are just what I just said. So RNAs that are non-coding, they can be structural, meaning they can form an actual structure. like they do in the ribosome. So I really encourage people to read about the ribosome. It's a nanomolecular machine that sits in every single living cell, and it produces protein. So it's performing this function. And so this translation is performed by a nanomolecular machinery called the ribosome. And about half of it is protein, but about half of it is RNA. I think it's half. It's a significant portion. So RNA can play structural and catalytic. It can actually perform chemical functions. And I'm sure you're to touch upon that.

speaker-0: that word, catalytic function is what I was actually waiting for because the book that we are talking about is called Catalyst. And Tom Cech is the guy who actually found a lot of properties of RNA that it can actually, one thing that blew my mind is that RNA can sometimes perform the same role as DNA, meaning it's actually encoding the functions. But at other times it can fold and it can actually create molecules that actually do the real work like proteins do. So it's actually the primordial molecule that can do literally all the different types of functions that we know of in biology. that's why they even have the hypothesis that it might've been the very first molecule that evolved from the full chemical cauldron of the planet billions of years ago, whatever it is.

speaker-1: I just love that you're excited about this. Just love it. And I hope the audience is excited as well. So yeah, so RNAs can also bind molecules. And like you said, there are many, many organisms even today, like SARS-CoV virus that causes COVID. That virus is an RNA virus. So RNA is the genome. That virus doesn't have any DNA, nor does it have any DNA intermediate. It's literally, it codes RNA as its genome. It then converts that RNA into another RNA that that codes for proteins and translates into proteins. So it's all RNA based. Yeah, RNA is a multifunctional molecule, and we're studying it in all its roles, not just coding. So it wouldn't be appropriate to say that the Vion tests are messenger RNA tests, because they're not only messenger RNA tests. We're seeing all RNAs and quantifying all RNAs, whether they're messenger, meaning they're going to get translated into proteins, but whether they're functional with any sort of function.

speaker-0: Yeah, Momo. So now that we know a little bit about RNA and there's a lot more to be talked about, let's go back to the relevance of RNA to human health, right? ⁓ What does RNA do?

speaker-1: Yes. Okay. So these are now a couple of slides that I think are the main takeaways from this entire thing that I think our audience can go and at dinner table or at a party, I mean, a geeky party, but still a party, you can talk about this and you can blow people's minds because people don't necessarily know this information. It's kind of mind blowing. And so we talked about DNA being the cookbook or the recipe book, right? So when you inherit your DNA, All of your cells in your body have the exact same DNA from your parents, for the most part. All of your cells. And what that means is that all of your cells have the same cookbook, but all different cells and different organs make different recipes out of that cookbook. And that's what makes those organs different. So if we take a biopsy from one person and sequence their kidney cells, their liver cells, and brain cells, and we do DNA sequencing, What the results are going to be is that, this is the same person, and this is their DNA. That's all I know. I have no idea where you extracted that DNA from. It could have been extracted from skin or cheek, swab, or any part of the body. It's just their DNA. That's it. And it could be any tissue, any organ. But you sequence RNA from kidneys, liver, and brain from the same person, and the data are black and white different. And the reason is that kidney cells must perform certain functions that are very unique to the kidneys. So they express only a fraction of those 22,000 genes that are encoded by DNA, and they express only the genes that are needed for kidney functions. Kidneys do not need to express the genes that create teeth or that create neurons or that create other tissues, only to filter blood and to filter toxins out of the blood, right? Liver cells are going to express a different set of genes. And that's what makes liver liver. It's perform it's detoxifying blood. And then brain obviously is expressing completely different genes because it has to make neurons and the different connections and so on. So RNA can tell now very, with very high accuracy, the difference between different tissues. And that's just a start. The next slide is really, really, really important. This is probably to me, the single most important slide that depicts the importance of RNA over DNA. So what we have here in this slide is we have the intestine from a Crohn's disease patient and on the left side and the right side are the same intestine from about a month apart. So if you sequence DNA of the intestine a month apart, what do you think the results are going to be good?

speaker-0: If you sequence DNA, there's no difference.

speaker-1: There's absolutely no difference. It's the exact same person, a month apart. Their DNA didn't change. So why is it that on the left side, this person is healthy? They can go out. They can eat whatever they want. They can cook whatever they want. They are healthy. They have no pain. A month later, they are literally in pain 24 hours a day. They have 20 diarhias per day. Their life has turned upside down. And they have the exact same DNA. How can anyone think that DNA is the determinant of whether you're sick or healthy. I mean, that's crazy visual here and crazy experience in life, right? So if you sequence RNA, so if you biopsy this tissue when it's healthy and biopsy this tissue when the person is sick just a month later, and what I'm showing on the right is kind of a technical graph. It's called a heat map, but it's basically showing the level of gene expression. This is human gene expression, and each row is a different human gene. And the blue and the red color show the level of expression. You can see that during the healthy state and during the sick state, the gene expression is completely different for many human genes, not all of them, but many. So there's this transcriptomic or gene expression signature where these cells are now expressing different genes when they're healthy and when they're sick. And this is what we're going to spend many episodes where Guru is going to be explaining. how we are basically doing the same thing for many, different diseases and how we're identifying gene expression patterns that are associated with health and the disease. And then we're going to talk about how we modulate gene expression. And that's another major, major advantage of RNA is that gene expression is modulated. That when we say diet and lifestyle influences our health, that's not some magic. It's literally changing the expression of genes and keeping a healthy gene expression versus wrong diet, wrong lifestyle, lack of sleep, express wrong genes, and they make us not healthy and sick. And this is a nice visual, but there are many other ways to do that. How cool is that, Guru?

speaker-0: Man, it always blows my mind when I see this picture. So let me ask you another fundamental question before I go to the next topic, okay? Which is how many genes or what proportion of genes, generally speaking, are expressed in, let's say, the human body at any given point in time? Out of the 100 % DNA genes that we have, what proportion of those are likely expressed at any given point in time, just in broad terms.

speaker-1: Okay, I'm not an expert in this, so I'm going to make some estimates, but there are genes that are required only in certain stages of life, meaning that when you're a baby, those are expressed, or when you're in utero, those are expressed, or for example, some genes are expressed in a breast tissue only during breastfeeding and so on. So there are these genes that are temporally regulated during your lifetime. I would say that those are probably 2,000 to 4,000 out of the 22,000 genes. This fact should be checked. So the vast majority of genes are expressed during our lifetime in some tissue. And so for example, what we see is that in blood, well, you're going to go over those data later, so let's go there. But yeah, I would say that about 80 to 90 % of the genes are expressed in some tissue at some level at any given, like in our adult life.

speaker-0: in our entire life. But at any given point in time, think what you're saying is approximately 10 to 20 percent of the genes could be expressed at any given point in time.

speaker-1: No, I'm sorry. I didn't say that properly. I was saying 10 to 20 % of the genes, like 2 to 4,000, are only expressed during very narrow parts of life, but the rest of the genes are expressed throughout life. most of the genes are expressed most of the time. It's just that there's like 10 to 20 % that are expressed in very specific times of life, like in utero and during childhood development and maybe during like breastfeeding and things like that. Special, special genes.

speaker-0: Let me also ask you, just to be clear for our audience over here, when we're talking about these RNA molecules, the gene expression that's happening, it's not only in the human genome, but it's also happening in the microbial genome, right? So if a bacteria, for example, we've talked about having, you 3,000 genes or whatever, there are some bacterial genes that are expressed for certain functions and other bacterial genes that are expressed for certain other functions. So there's a fraction of the bacteria genome that's expressed at any given point in time as well. And I actually remember some data from one of our experiments where we looked at the potential number of genes that are in a given person's microbiome, which was in the millions and maybe like two and a half, three, four million genes that are available. But at any given point in time, there's probably a small fraction, maybe two, three, less than 5 % of the microbial genes were expressed in the sample set that we were looking at. Is that correct?

speaker-1: That is exactly right. And that's a part of a publication that we need to submit for review. yeah, it's basically exactly what you're saying. The DNA of a typical gut microbiome, average microbiome, encodes about 4 million genes, somewhere in that order, whereas only about 100,000 genes are actually expressed in any appreciable level. And we're talking about a level of one part per million, meaning if a gene is expressed and that RNA... that from the expression of that gene contributes less than one part per million to the whole community, we're going to ignore it for this use case. Only a few percent, like under 5 % of microbially encoded genes are expressed. And that's a very important point that when we publish that paper, we should really review that and point out why sequencing DNA gives you a huge amount of data and a huge amount of noise that doesn't actually contribute to our body and our health at all.

speaker-0: Exactly. That's a very important data point that I think we should keep in mind, but I want to step back now and ask once you know that certain RNA molecules are expressed in certain ways in the body, what can we do with that data?

speaker-1: Great point. Yep. Okay. So I'm just going to show you one slide and we're going to touch upon this, but this is really an entire podcast that I will be querying guru on how we use these data. So let me show you just an example. Okay. So I think that a lot of people know that in this particular case, cruciferous vegetables encode sulfur containing compounds. are glucosinolates and this particular molecule is called sulforaphane. So Sulforaphane is a substrate for specific microbial genes, meaning when they're expressed, they can chemically convert sulforaphane into hydrogen sulfide via a cascade of reactions, which we showed a pathway on one of our previous episodes. And hydrogen sulfide is a key metabolite in our health. So while over the course of human evolution, we have consumed sulfur containing compounds. The microbiome that has colonized us and has become our best friend has been converting sulforaphane to hydrogen sulfide. And our own physiology has then adapted to that and said, well, wait a minute. We can use this hydrogen sulfide as both food source or energy source and as a signal for gut motility. And so hydrogen sulfide is one of five or more molecules that are produced by the gut microbiome that regulate our gut motility, whether we're going to be constipated, whether we're going to be irregular, or have diarrhea. Down below what I'm showing is apples have pectin, which is a polysaccharide. It's a fiber. And specific microbial genes are able to convert that pectin into butyrate, just like the example I gave last time where brewer's yeast convert fiber into alcohol or beer. In this case, these specific microbial genes convert pectin into butyrate. And while DNA methods can tell you that you have microbes that are capable of converting sulforaphane into hydrogen sulfide, they cannot tell you at all whether that process is actually taking place or to what extent. That's it. Whereas our RNA test specifically quantifies how active is this biochemical pathway and how active is this biochemical pathway. And we're going to obviously dive into this a lot deeper with Eric and you later on where I'll be querying you. But basically, if a person has too little hydrogen sulfide, we want to push this reaction to go forward and produce more. And if another person has too much production, we want to slow this down by depriving that person of the substrate, or not depriving the person, but actually depriving their gut microbiome of these substrates. And so these are substrates that are found in foods, and these are metabolites produced by the gut microbiome that affect our health.

speaker-0: Yeah, Momo, let's get into some specific examples of how the pathways that are active in your microbiome, the RNA expression, can help us interpret whether somebody's gut microbiome is healthy or not.

speaker-1: All right, that's a great question, Guru. And we have some examples of that. So I'm really excited to share these. So there's something that I call the E. coli paradox. And E. coli is this very controversial microorganism in that most people think of it as a killer. And it comes in the news that so many people died of E. coli poisoning, right? And then if you read the scientific literature, E. coli turns out can be very beneficial, and it's associated with many health conditions. And so I'm just going to give you here an example where a typical or one of the E. coli strains is able to do five different biochemical functions. can produce a genotoxin called colibactin. And this genotoxin has been strongly associated with colorectal cancer. It can also produce, all E. coli's can produce endotoxin called lipopolysaccharides. So these are two very harmful functions for our body, but the same exact genome. also codes for genes that produce indole. And indole is a crazy metabolite in that indole produced in the gut microbiome actually travels to the brain and stimulates neurogeneration. It can actually stimulate neurons to divide and create new neurons, which is amazing. Just recently discovered. coli is one of the best producers of vitamins K2 and several vitamins B naturally. And then E. coli is great at producing acetate, which is a very beneficial short chain fatty acid. So now, if someone did a DNA test that says E. coli, someone can say, ⁓ my god, I have E. coli. How do I kill it? Some people may read the literature and say, wow, E. coli produces this genotoxin. This is dangerous. How do I kill it? Other people may read the paper that say, wow, this is great. E. coli is doing all these beneficial things for me. But based on the DNA test, you have no data at all as to which one of those functions, if any, your E. coli is performing. The Viome test will actually quantify the level of gene activation or gene expression for all the genes making those biochemical reactions happen. And then we can say whether your E. coli is doing good things for you or bad things for you or both or neither based on your biology. And so that's pretty exciting about, not pretty exciting, that's hyper exciting that we have those data natively generated. So that's one example. And the other example I will provide is this molecule called TMAO. I think that. A lot of people are now familiar with trimethylamine oxide. It's a molecule made in the liver, but the necessary substrate for TMAO is a molecule called TMA, or trimethylamine. And that molecule can only be made by microbes, gut microbes. And it's made from carnitine and choline that we consume. So if you remember, substrates for TMA are carnitine and choline. And so with a DNA test, someone can say, ⁓ no, you have microbes that are making TMA. When in fact, all they can say is, have microbes that have genes with a potential to create TMA. I have zero information, whether your microbiome is making any TMA or not, zero information. DNA cannot provide that information. It's just a cookbook as we discussed. But with our RNA test, we can actually quantify the exact level of expression of those genes that are actually performing that function. And then we can tell someone you have high TMA production or low or moderate. And by the way, TMA is a very typical kind of a metabolite, just like hydrogen sulfide and butyrate that we discussed just a minute ago, where small amounts are very beneficial for the host, but large amounts are not very beneficial for the host. They're in fact very, very bad in that they can cause atherosclerosis or cardiovascular disease. So these are just tiny little slivers. of what our technology can provide. And we're going to obviously dive much deeper into all the scores and all the disease correlations and everything. But now, let's go into an exciting thing that is exciting to me, which is, Guru, can you please talk about the data? You talked about a million samples earlier. Can you please share with the audience what kind of the highlights of what have we seen in those data?

speaker-0: ⁓ yeah, Momo, we have so much data in Viome that sometimes it's a head spinning. Let me explain to you the amount of data that we deal with, right? Every sample that a Viome customer sends us goes through Momo's lab process and it ends up in the sequencer, which converts all of the molecules that's in the sample into bits and bytes, into data. Now, that data happens to be in... tiny bits, it's really about 150 nucleotides at a time. There's tons of these reads, as we call them. Now those reads have to be reassembled and we have to identify which organism or which function is a particular read corresponding to. And that process, the algorithms for doing that conversion from the individual reads to the organism or the function is what our bioinformatics algorithms do in Viya. So I'll give you an example of some of these translations that happen in the bioinformatics. One is we have a catalog of tens of thousands of different organisms. Like on the picture you see here, we have something like 25,000 genomes from bacteria. We have a ton of genomes from viruses from archaea also fungi fungi i'm particularly proud of because we had to work quite hard to clean up the genomes and to get them into our catalog but now that we have them we can take every one of these reads and this is you know we're talking about millions tens of millions of reads and we are trying to figure out which genome do this or these set of reads correspond to and there's an algorithm called an alignment algorithm and there's many different varieties of that algorithm that we have tried and we have landed on a specific type of algorithm, which gives us a very high, reliable data on which genomes are likely active. That's on the first side. But as we've been talking about, knowing about genomes is not going to be very helpful because it's only telling us who the organisms are. What we really want to do is to learn about their functions, meaning that we need to actually figure out what the genes are that are expressed from those genomes. for that purpose, we have a different catalog of all the microbial genes and in fact, all the genes that are possible to be found in the kinds of samples that we normally process, like the stool samples, saliva samples and blood samples and so forth. You can find many, many different types of genes that are expressed in those samples. So we have collected a very large catalog of genes from all the different sources of gene curation that exists out in the world. And that has resulted in a catalog of more than 98 million. So let's say, you know, rounded up to a hundred million genes in that catalog. Of course, just knowing the genes is not enough because if there's information about what the function of a gene is, that's called an annotation or a functional annotation, there are databases which give us these functional annotations. One of the very well-known functional annotation databases called the Kyoto Encyclopedia of Genes and Genomes, also known as KEGG. So we have something like 16 plus thousand annotations of different types of functions that we overlay on top of all the genes that we detect from a given sample. And we can then determine how these functions are relevant to somebody's disease state or health state. Now, I also want to point out one more super cool thing over here. You see at the bottom of the middle panel, you see there's approximately 60,000 transcripts that we are able to see. Now, these transcripts are 60,000. We talk about the human genome having more like 20, 22,000. So there is a very fancy process that happens with RNA, which, by the way, is in the Tom Cech book, the Catalyst book, which you know, again, I was so excited to read about, which is the whole process of splicing, right? So you have different components of your genes that are actually cut and connected back in multiple different ways. So in the same gene, you may have multiple of these introns and exons as they're called, and they get chopped and they get spliced back together in slightly different ways to get actually multiple functions from the same G. Isn't that super cool? It's just a way that nature has figured out that you can in fact, multiplex, meaning you can have multiple functions from a single G. And this splicing mechanism helps us do that. So we have something like 60,000 of these transcriptional elements that we can detect. And on the right-hand side in the slide, I also show the number of various kinds of genes and genomes that we can detect. So for example, in this particular instance, we see here that we can detect approximately eight and a half million unique microbial genes across all the stool samples that we've ever seen. And for any individual stool sample, we see approximately 80,000 genes. That's a huge number of, you can call them biomarkers if you want, right? But from those biomarkers, from those genes, we can annotate approximately 11,400 or so of these functions or the KOs that we talked about, the keg orthologs. And for a given stool sample, we normally see approximately 3,000 of these functions from microbial population of your body, which is your gut microbiome. Let's say in this case, it's a stool sample, so it's a gut microbiome. So approximately 3,000 functions. in a single person's gut microbiome. But if you want to look at the species, and sometimes it's useful to kind of just look at both the functions and the species and correlate them. And we have a lot of things to say about why it's not enough to just look at the species alone, you have to look at the functions. But you can see here that we've seen approximately 7,780 species across all stool samples. And for given stool samples, we have something like 1,100 specific species. Now, the same method can also be used to process blood samples and figure out what the human genes that are expressed in your blood samples are. So in our case, have seen on the whole across all our blood samples, we've seen approximately 19 and a half thousand human protein coding genes. And in an average sample, we see approximately 11,500 genes that are expressed. So that is the extent of data that we get from every single sample, stool sample, blood sample. I also have saliva samples that we can talk about in a future episode, but that's a huge amount of data. So we get all this data. Now we have to convert that into what are the higher level biochemical functions that are happening in the person's body that correlates with health and disease, right? But in order for us to do that analysis, which we'll get to in future episodes, we need to be sure that the quality of the data that's coming from the lab is very good and very high. So I want to ask you, Momo, what are the kinds of analysis and validation that we have done in the lab to make sure that we're getting high quality data? And in particular, There's a lot of people who think that RNA is super dynamic. It changes. mean, it's unstable. It just degrades very fast. tell us a little bit about how do we preserve the RNA molecule from the point that it's collected all the way to the lab. And then how do we keep the data quality high so that all the analysis that I do and our team does is still reliable.

speaker-1: tell you that in a minute, but I want to also comment on this slide. So super impressive large numbers that we already have. What I'd like to point out is that these numbers are growing in terms of the functional space for microbial functions. So we're continuously learning new functions of bacteria. So for example, the example that I gave earlier about TMA production, the genes responsible for making TMA were not known until just a couple of years ago. And so until then, we knew of TMA. as a metabolite, but we didn't know which genes were responsible for making it. And now we do, and Viome can just integrate that knowledge into our Viome recommendation engine. And so we can power our nutritional recommendations not just from the knowledge we gain, but also from the rest of the world. So that's very important. But I also want to point out that the fungi part here. You mentioned, that we made a large effort to curate fungal genomes. And I want to make sure that people understand the complexity of this, that Fungi are likely a very hidden root cause of many chronic diseases today. And some people who have suffered from mold illness are very familiar with that or who have chronic fungal infections. We know that fungi are responsible for many chronic diseases, but they're kind of a dark matter. And one of the reasons for that is because the world has focused on environmental fungi and not medically relevant fungi. And second is that... fungi really act via mycotoxins and there are so many of them and their genes that the genes that are actually encoding mycotoxins are not very well understood. And so that whole space is really kind of a dark matter understudied. And what we're trying to do at Viome is first step was to curate these genomes so that we have a high quality. Now we really need to make an effort to annotate those genes that are making mycotoxins. because I've glanced at one of our data analysis projects and found out that the biggest determinant of whether someone has brain fog or not is actually the fungal community, not the bacterial community. So we have one very serious and very prevalent condition, brain fog, that affects people under 65 that is predominantly determined by fungi. And we really need to nail that down. So now I will answer your question. So first of all, yes, most people know that RNA is very labile molecule. And one of the features of RNA is that it actually self cleaves. So while DNA, you can deposit it in the environment and unless something actively choose on it, it'll be there for thousands of years. Whereas RNA, it will self cleave. It doesn't need anything to degrade it. And so it's naturally unstable molecule. And so people always wonder or criticize Viome saying, well, RNA is unstable. You cannot quantify it. Well, it turns out before Viome was even formed, we already created a proprietary preservative that we're using at Viome. And I just want to highlight here that we have published three papers that revealed the clinical validation of the stability of RNA in stool blood and saliva samples. And those are going to be posted in the show notes. And these three papers not only perform clinical validation of the stability, but also on all of our measurements. So when you ask, how do we ensure data quality, we have a clinically validated method that is a part of a CLIA license. And the US government, as well as the state of California and New York, oversee the whole process. And they come and audit everything and review everything and so on. And we can talk about CLIA labs later on. But I think for now, this is good enough to highlight how this is a clinically validated process.

speaker-0: Yeah, Momo, I think it's super important that everybody knows that this is a very reliable platform that you've built. It's also very highly scalable because we could do as many samples as we can possibly get from anywhere in the world. And number three, it's also not expensive. It's pretty low cost. That's why it's possible for a consumer to be able to buy a test from Vibe. Right? So all of these three things have happened in a, very scalable way. And of course the data analysis and all the clinical applications, such as the risk determination, the recommendations, all of those things are also happening in conjunction with the lab. And all of that is happening on the, so to speak, the dry lab and the dry lab is also highly scalable. And it's possible to take this data and look at it in many different ways and make it useful for many different people. Speaking of making it useful. Let's just touch on this one last point before we close the episode, which is how does all of this RNA analysis, both on the human side and the microbial side, translate to clinical impact ultimately?

speaker-1: All right, great. So back to the future a little bit. Guru, you will lead our next episode, which will focus on personalized nutrition towards minimizing the glycemic response. So the glucose spikes after meals. And we have leveraged that machine-learned model that we're going to dive deeper into next episode to create a viral recommendation engine that can basically move certain foods to either eat more or less of for each person such that their blood sugar spikes are minimized, not eliminated, just minimized. And we have leveraged that platform to run a randomized clinical trial. This is a blinded, randomized placebo-controlled trial where we enrolled people with predominantly pre-diabetes. And we published a pre-print, which is now under peer review, that shows that we can reduce HbA1c, which is the major biomarker of diabetes and pre-diabetes, we can reduce it by 0.42 % relative to placebo. And this is clinically and statistically significant value. So we're very proud of that. But what I'm also super proud of is these preliminary data that we obtained from our interim analyses. These are not statistically significant yet. This is just a first look. But the effect size here is very large. You can see here that we're measuring fasting glucose and fasting insulin. And what you can see here is that we reduced the fasting glucose by 25 points. in people in just three months. That's a very significant amount. For example, normal blood glucose is between, let's say 80 and 100 milligrams per deciliter. And so anything above 100 would be pre-diabetes. And so you can see the 25 units is a very large number relative to that. And then when we go over to fasting insulin, fasting insulin is normal under 10. It's pre-diabetic from 10 to 15, and then above 15, it's diabetic. And what you can see here is that In three months, on average, we reduced the fasting insulin by five milli international units per liter in the interventional group. That's this VPNP, whereas the placebo, actually went up slightly. So these have not reached statistical significance. You can see the P values are 0.11 and 0.117, but this is from a preliminary analysis. So we're hoping by the end of the trial that these will be improved, but we'll have to see. So these are very, very exciting data showing that our personalized nutrition works well with these very important biomarkers. And then I'll just flash some preliminary results we got from two other placebo-controlled trials. And this one has reached statistical significance. It's a GAD7 endpoint. This is generalized anxiety disorder seven. So it's a seven question clinically validated questionnaire. And you can see here that 50 % of the people in the violent personalized nutrition program were transferred from anxiety to some level of anxiety to healthy, whereas only 28 % were in the placebo arm. That has a statistically significant value, so that's very important. And then in a completely randomized controlled trial where we enrolled people with IBS, all levels of IBS, we have even more dramatic effects in that 67 % of the people who had constipation at the beginning of the trial no longer had constipation at the end after three months, but only 13 % of the people in the placebo arm experience the same transition and the p-value is very low for that. So a few preliminary trials, these take a lot of time, a lot of money, and they involve a lot of people as you know how complex these trials are. But we wanted to give the audience a glimpse of what we're working on to translate all of these data into actions and then to measure how those actions may influence human health.

speaker-0: Yeah. So what I take away from that Momo is that once you understand the dynamics of the biology for any type of function, let's say for glucose response as one of the functions, right? Once you understand the biological functions, then you can modulate it because you can understand what is driving the biological function high or what is suppressing it and keeping it low. And you can modulate it by providing the right kinds of things to either boost it or to dampen it so that you can then get the health benefits that you're seeking. And glycemic response or sugar response is a great example of that, which we're going to get into the next episode. But before I talk about the next episode, let me step back and summarize the key takeaways. I'm learning all the biology and biochemistry from you. So the RNA molecule is one of the most amazing, if not the most amazing molecule in biology after especially after I read Tom Cech's book I am convinced that that is it and so you are very

speaker-1: Let's just call it the most amazing molecule.

speaker-0: It is the most amazing molecule. And I think you were very wise and definitely very lucky that you got into the field of RNA science so early in your career and you've learned all these things about RNA. So I'm like super amazed. And one of the things that I've learned is it's not your DNA. No, DNA is sort of like the background of what is possible, but RNA is the foreground. It's actually what is happening at any given point in time. And that determines health. It determines disease. It determines your phenotype. It determines which organ does what function. It's basically the chef in the kitchen that is cooking the meal. It's not the recipe book that's sitting in your shelf. I have several of those recipe books in my shelf which do nothing. It's there. There's knowledge in there. There's a lot of very good ideas in there, but none of those ideas gets realized in the real world until a chef comes along and picks up one of the recipes and starts to cook with the recipe. That's what RNA does. So that's my main takeaway from this entire episode. And we can talk about all the downstream impacts of that, which is once you understand the activity of RNA, you can see what you can do. You know, the great thing about RNA is that you can modulate it, right? On both on the microbial side and on the human side, you can modulate it. Therefore you can take an unhealthy state into a more healthy state, right? And if you are not careful, you can take a healthy state. out of a healthy state, make it unhealthy if you're not paying attention to your diet, you're not paying attention to your exercise and sleep and stress and everything else, right? So that can also...

speaker-1: Or if you follow a generic advice. If you follow generic advice, you can easily fall into... mean, there are many people who have IBD and they're in remission, and so they have no symptoms, and they follow ⁓ just a generic healthy advice, and yet they fall into a relapse because they just don't know what triggers that relapse. And different people will trigger a relapse for different reasons. So it's really personalization that's key here. But like you said, The first step is understanding what's driving health versus disease. And DNA is not the molecule to tell you that. It's the RNA.

speaker-0: DNA is not your destiny, as we said in the title of this episode. That brings us pretty much to the end of this episode. I just want to mention again, the next episode is likely going to be about blood sugar response and all the studies we've done and others have done and how we use that environment and what impact it has. One of the studies that Momo showed is actually one of the direct impacts of that. So we'll talk about all of those things and we'll have some guests on the show as well. And the other thing I want to mention is the April challenge, just a reminder. the April monthly Viome challenge simply asks you to post something that you find meaningful and insightful about personalized nutrition. It can be anything. It can be your experience. It can be a paper. It can be some resource. It can be anything. You post it on a social media site and you put the link to that social media site in our episode comments or send it to podcast.viome.com. And a lucky winner will be picked based on the content and the engagement of that post. And Momo and I will announce it next month, just like we announced for the month of March earlier in this episode. We had a lucky winner, Zoe Miller. And I think she's going to get a volume full body intelligence. And similarly for April, somebody is going to get a very powerful and I would call it mind blowing educational resource that Viome full body intelligence testers. with that, I'll just remind everybody to subscribe and engage with this episode. And I thank you all for joining us. I'm Guru, an AI expert.

speaker-1: I'm Momo, a biochemist.

speaker-0: We're two PhDs on a phone. Yep, time.

speaker-1: Woo!

speaker-0: Two PhDs on a pod is a Biome Health Science podcast production written and hosted by Biome CTO Guru Banavar and Biome CSO Momo Vujicic. Produced and edited by Lisa Shomo. Podcast art by Kevin Wu. If you enjoyed today's episode, like and subscribe for new episodes every other week. We'll see you next time.