 So these are the standard three slides that we have to show, and of course this is our introduction to metabolomics. You've already seen this outline before, but I'll use this again just to probably reiterate a couple of things. There's a lot of people from very different backgrounds. Some of you are very experienced, maybe even more experienced than I. Others are here learning about metabolomics for the first time. Some of you are mass spec specialists, others are specialists in statistics, others maybe specialists in other fields. So it means it's a diverse group and not everything that I'm going to give you is necessarily going to be new to you, and so if it isn't just sit patiently because it's likely going to be new to your neighbor. So it's a case that we're dealing with a very diverse audience with a wide range of backgrounds. Another thing that I'm doing here, probably for your first couple of lectures, is in order to do metabolomics, at least understand how to process the data, you should also understand the methods, the process. If you don't know where the data is coming from and how reliable it is, how it's been obtained, then even if you've got all the training and all the wisdom in the world as a statistician, it's not going to help you, and I can say that from experience. So it's really, really important to understand some basic aspects of how metabolomics is done, how the data is acquired, how things are measured, so at least you'll understand what to watch out for when it comes to the analysis and the other stuff that we'll deal with later on today and tomorrow. So this first part, these first two lectures are, how do we get the data? What's the process? What are the techniques? What are the things to watch out for? And that's, I think if people had known that about transcriptomics 10 years ago, we would have saved ourselves a lot of heartache and headaches. So that's just, I guess, framing things so that you guys will have a better idea of the perspective of why we're stepping through things the way we are, why sometimes recovering things that might seem really silly to you, but remember it might be really important for your neighbor. So this is the start. This is a slide I use very often, which is trying to explain the pyramid of life. And it's a pyramid because it's a pyramid in terms of numbers, with the largest number, arguably, of components, at least if we're looking at the human genome, about 25,000 genes, coding for many thousands of proteins, but of those proteins about 7,500 are enzymes. It's not that the other proteins aren't important, but these are the enzymes that do metabolism. These are the things that help transform and induce and transduce all kinds of things. And they're relevant to producing the metabolome. So the study of genes is genomics, the study of proteins is proteomics. The study of metabolites is metabolomics. And the reason why this is a pyramid is that what happens here profoundly affects this and even more profoundly affects this. And that's important to know. And so sometimes you see in the predator prey pyramid where there's a large bunch of prey and then there's a few predators at the top. What happens with the abundance or characteristics of the prey can significantly affect what happens at the top of the pyramid in the predator. It's also important to remember that there's a gradient as well in terms of environmental influence. What you just ate this morning, what you're breathing now, what you're drinking now isn't affecting your genome at all. And it probably won't affect it for many years to come unless you're eating the same stuff every day or inhaling the same toxic fumes every day. It may have and it currently is having a little bit of effect on your proteome, but only about three proteins are being changed. Insulin, glucagon, leptin, that's about it. And they're slowly rising and slowly falling. Right now, your metabolites are just going through the roof up and down and changing on a minute, even a second to second timescale because of what you've just been eating. And these things are being transformed and modified and being incorporated into proteins but also being modified by your gut microflore. So in effect, environment or the environmental influence is much greater on the metabolome than it is on the genome. And that's useful. It's not something that you should ignore or try to avoid. That's very useful because this is how we learn about what ultimately affects our genome, but also what affects our phenotype and our genotype and those environment gene interactions. So everyone here is probably familiar with at least the word metabolomics. This is just a quick definition. I think everyone understands genomics, so it's high throughput technologies, high throughput science, and we're trying to characterize all the genes in a cell, a tissue, and organism. It's exactly the same definition that substituted the word genes with metabolites or small molecules. And there's also been a confusion for a lot of people when you say metabolomics and say, oh, you study proteins, just small proteins. You know, we're studying small molecules. We're studying metabolites, chemicals. And we can look at cells and tissues and organisms, but we're studying the metabolome. So metabolites, this is a definition we'll use. People may use a slightly different one. Sometimes people use a molecular cutoff of 1,000 Dalton's. We'll use a molecular cutoff of 1,500 Dalton's. So molecules that are small. So that includes some peptides, a very small number actually. It includes some oligonucleotides, but not microRNA or anything like that. So really small things. It includes the other things like sugars and amino acids and organic acids and nucleosides and aldehydes and lipids and steroids and alkaloids. And then as you get into microbes and plants, you get into a lot of other things. But you can also get into the environment. Food that you eat, food additives, toxins that might be in the food, pollutants that you inhaled, walking into here, any drugs or drug metabolites. Those are all part of what we call metabolites. They can include, in the case of the human body, these would be the human metabolites in place of trapanosomes. It's the trapanosome metabolites. The human body and most organisms actually have microbes in them. And so you can also include the microbial products. So in the case of human metabolism, the gut microbiota, the mouth microbiota, extremely important. And then when we're defining metabolites, we're basically looking at the things that we can detect. There's a lot of theoretical metabolites. We can kind of guess what they might be, but they've never been detected. So in metabolomics, we're essentially looking at what can we detect, because that's what exists, to our eyes or to our instruments. So that's a metabolite. What's a metabolome? So that's the formal definition. That's all the small molecules. It includes endogenous and exogenous, meaning endogenous being produced from within, exogenous being taken in. It is defined by the detection technology. And because it's defined by detection technology, the metabolome is not perfectly defined. We know exactly how many genes are in E. coli. We know exactly how many genes are in Drosophila. We have a pretty good idea of how many genes are in the human genome, but we still have no idea how many metabolites are in the human metabolome. And that number is changing daily. And it will continue to change daily for the rest of your lives. We do have some idea in terms of what we can detect right now with today's technology. So in terms of mammals, metabolically, they're pretty simple. There's about 8,000 chemicals. There's 8,000 compounds we can detect in humans. There's about 8,000 compounds we can detect in cattle. There's about 8,000 compounds we can detect in chimps. It's about the same number. Give or take maybe 50, maybe 100. We've got a reptiles and birds. Maybe they add another 100. But it's still about the same. In terms of microbes, obviously very diverse organisms, they have to be able to produce everything on their own or most of the time. They're obviously occupied very different niches. It's not really known because there's many, many thousands, tens of thousands of species of microbes and parasites. But they're not quite as chemically diverse as plants, but more chemically diverse than mammals. The most diverse ones, metabolically, are plants. And we know this part because people have been doing a lot of natural product characterization for 100 years. And every day, they're finding something new in plants. Why are plants so metabolically diverse? The reason is they don't have legs or flagella. They can't move or they can't run away from predators. So what they do is they use chemical warfare to battle predators or pests or anything else. So it's critical for plants to have a huge arts. And of course, there's a wide range of plants that occupy very many niches. And again, to adapt to a stationary niche, they have to be able to have a gain, a wide range of metabolites that allow them to grow in tough conditions or survive in tough conditions or protect themselves in tough conditions. So in the case of humans, and you'll notice it, a lot of my discussion will focus on humans partly because that's who's in this room. I think we all are kind of interested in ourselves. But it's a good reference frame for a lot of other things. So what I say about humans is also applies to rats and mice. And because humans also have microbes in them, it sometimes also pertains to microbial studies. So metabolites and humans can range from femtomolar, although we can't really detect them, up to almost molar levels. The most abundant metabolite that we measure in humans is urea. And it can get up to 100, 200 millimolar. So very high abundance. We know, or we can measure, as I say, about 8,000 metabolites, endogenous metabolites. But as I said, there are other things that we take in. So humans eat other metabolites. We eat plants. We eat animals. So there's lots of food products that also are part of our metabolome that are not produced endogenously. Some of them are as a food additives. Some of them are the plant or phytochemicals. We know that there are about 30,000 of those compounds. There's probably more, but that's what we've been able to measure. People also take drugs. No one individual takes all drugs. But in a population, we will find several hundred different types of drugs that can be detected by metabolomics. Drugs are broken down into drug metabolites. Likewise, food products are broken down into food metabolites. And then there are a variety of toxins and industrial chemicals that also come into contact with us and are absorbed and stay with us for most of our lives. As we climb this ladder, you notice that there's the abundance drops. Hopefully it should. You didn't want to see things like this up here. Otherwise, you'd be dead. Same sort of thing with drugs. This would be obviously seriously problematic. So typically, most drugs and food additives are in a micro molar or lower levels in blood or urine or tissues. But illustrated here are the numbers of these compounds that are known and then also some databases that track of these things. And these are databases that we've developed at the University of Alberta and maintain at the University of Alberta. And these are used by millions of people. So there are critical resources for metabolomics. Now, those are the known metabolomes. They're what we would call theoretical metabolomes. And this could apply for humans. It could apply to plants. It could apply to microbes. So there's another one we would call the secondary endogenous metabolites. These are metabolites that the body produces. And the body doesn't quite know sometimes what to do with them. And so it will treat them like xenobiotics and transform them. It'll hydroxylate them. It'll esterify them. It'll do something. And it wasn't supposed to do that to them, but it just happens. Mistakes happen. So these are things that are produced in the body. These are at very low levels. And these may account for a number of the mysterious peaks that people keep on seeing in mass spec based botanical. Food is another thing. When we eat food, it's broken down. And especially our gut microflora transform food products into a whole variety of different things, as does our liver. And it changes those food products into a variety of compounds, the same way that they transform drugs. There may be 100,000 different compounds. That drug metabolite set, there's 1,000 that we know, but there's probably 10,000 that we don't know. And so those are secondary drug metabolites. And then there's a thing called the lipidome. So the lipids that our body generates. And while we can measure about 6,000 lipids in the body, theoretically we know there are about 100,000 lipids that could be produced in the body. So we add all these up, and we're looking at well over 200, perhaps 250,000 compounds. So the complexity of the human metabolome is perhaps as great as what we currently know for the plant metabolome, the entire plant universe. And so that pyramid that was originally genome at the base metabolites at the top is actually starting to become inverted as we find out more and more about the metabolites that exist. So why is metabolomics important? Why are you here? So my background is as a protein chemist, a protein biologist, I come from the world of large molecules. I learned, I hated biochemistry. I hated metabolism. I hated all the stuff that had to do with small molecules because when I was a student, the most interesting things were large molecules. And for the most part, people still view them as being the most important. And they are important. But I think there's some statistics that might help put things in perspective. If you go to a doctor, tests that your doctor will perform today, probably 10 years from now, will still be measuring small molecules. Almost all clinical tests, the common clinical tests they're used today, still measure small molecules. And we'll explain this over the coming days, but there are good reasons for it. Even though biotech drugs are the ones of the future, everyone's aiming for, the proportion of small molecules to large molecules has remained the same for the last 20 years. And it will remain the same for the next 20 years. 90% of new drugs, 90% of old drugs are small molecules. What's more, those drugs that they're developing, half at least, are based on metabolites that we found in either microbes or plants or animals. So a great source of drug ideas and of drugs is from the natural world. If we even look at the genetic diseases that we've been studying for decades, about a third of them have to do with small molecules, either the control or signaling of what small molecules do. And then we also have to remember that small molecules, maybe they're not the engines of life, but they're the electrical wires of life. They're the things that act as the conduits, the signaling molecules, the cofactors that enable enzymes to function, that enable cells to communicate, that enable us to talk, to think, to communicate. One of the reasons why small molecules are used as clinical diagnostics, the reason why they're used in physicians' offices and why they'll continue to be used, is that they're basically the canaries of the genome, just like the canaries in a coal mine. It's ideally you could have canaries in your helmet when you go down in a coal mine and if a canary keeled over, it was time to get out of the coal mine because at least that would give you a minute or two before you, too, would keel over. What happens at the genome level is typically amplified manyfold at the protein level and even more at the metabolite level. And there are many examples where a single base change, the metabolite levels by a factor of 10,000 or more. So that makes it really easy to detect something that's going wrong. So even though it's getting cheaper to sequence, to find a single base change in one of us still means sequencing 3.1 billion base pairs and then checking against a reference genome and hoping the reference genome didn't have something else wrong with it, whereas sometimes just doing a simple dipstick test will tell you what's wrong and identify the gene and identify the probable mutation as well. This is the other thing I brought up earlier, which is the sensitivity to the environment and it's not only what you're eating, but also over the time. So what we just ate for breakfast, this is what's happening to your metabolism right now and it's gonna continue for the next hour or two at least. The response in terms of your two or three proteins, very slow and it's only marginal and then nothing's happening with your genome. The other analogy is, I've used this example, if you held your breath for five minutes, could we know what's happening? Most cases, if you could hold your breath for five minutes, you would die. If we were monitoring you over those five minutes metabolically, we would see something like this. Looking at your proteome or genome, we'd see no change. So by genomics and proteomics, we couldn't tell if you were dying or even if you were dead, but by metabolomics we would. Another advantage with metabolomics is that a lot of metabolism is understood. Again, biochemistry textbooks that you may have had to remember and pathways you had to look at. These were worked out and this is a diagram actually drawn in the 1960s. This is how much they knew 40 years ago, 50 years ago. It's really well understood. In fact, the people who were doing biochemistry 50 years ago probably know more about this than we know today. So unlike cases for signaling molecules for protein-protein interactions, gene-protein interactions, which we're still trying to figure out, this is understood. Not only are the pathways and locations for these things understood, it's understood at a quantitative level. We've got rate constants. We've got concentration data. It's quantitative in a way that nothing else in proteomics, genomics, or transcript ovens can match. Metabolome is also connected to the other ohms and it's pretty easy to see what genomes code for, proteins code for affects what happens with the metabolites, but because metabolites also are cofactors, they affect the function of enzymes because metabolites also affect transcription and translation of genes. They also affect the genome. Genomes obviously affects what happens in the metabolome. All of these are connected. And to try and think of things as insular or isolated ohms is the wrong way. And so you're gonna hear mostly about the metabolome, but remember, proteome counts two, the genome counts two, the transcriptome counts two, but excluding the metabolome from your studies is essentially like covering one eye. You're not seeing the big picture. And so this just illustrates whether it's the small molecules that are the constituents of the genome and the transcriptome, AMP, CMP, GMP, and TMP, the 20 amino acids which makes up every protein in your body, the lipids that give your cells their structure, the small molecules which are the fuels to the body, the cofactors and signaling ones. One way of looking at it is in fact that the genome and proteome evolved to do chemistry, to make chemistry happen. So that life really was essentially a set of chemical reactions. And that's the way chemists typically look at it and that's what they're trying to do with early life. It's chemistry. But to make chemistry happen sufficiently fast so that life doesn't take thousands of years to give birth, you create proteins. You create enzymes and you create DNA to help generate those proteins. So that the proteome and genome arguably evolved just to be able to facilitate what goes on in the metabolome. Because metabolomics is uniquely quantitative in the sense that we understand metabolism so well, we have the rate constants down pat, it's actually been the source for most of the work in systems biology in terms of quantitative systems biology. A lot of the modeling is done with metabolism because we have the numbers. So by blending now that we get more quantitative data with proteomics, more quantitative data with genomics, people are combining these things and bringing it into systems biology. So as I say, try and keep this holistic picture of genomics, proteomics and metabolomics. There are lots of applications for metabolomics. We'll cover some of them. Some are fairly obvious in terms of clinical analyses. That's been around for decades, arguably. Tox testing, drug and pharma use metabolomics in clinical trials as well as preclinical trials. Food and nutraceutical industry uses metabolomics a lot. People are also using concepts behind metabolomics and water quality and petrochemical analysis. There's now imaging for metabolomics. It's a wide range of applications and they're growing. This isn't an exhaustive list. So I'm gonna switch gears and I'm gonna talk about the techniques of metabolomics because it is a multifaceted field that requires a lot of different methodologies. And this isn't bioinformatics, but as I say, if you're gonna do bioinformatics and statistics on your data, you need to understand where it came from. You need to understand how reliable or unreliable it is and what kind of data you're gonna be getting. So if you haven't done a metabolomics experiment, maybe I should just ask, how many have actually done a metabolomics experiment? One, two, three, four, five, six, seven, eight, nine. Just measuring metabolites. So when you're doing a metabolomics experiment, you have to start with a sample. And with a sample, you can typically arrest the metabolism by freezing it. And then typically after freezing it, you'll try and extract some kind of biofluid from it. Sometimes you can go straight from the tissue into the analysis, but most people like to try and extract material from it. So there's a quenching step from the live sample, it's an extraction step. And eventually with that extraction, you end up with some biofluid or extract. Sometimes with living systems, you can take the biofluid directly. So you can take urine or blood or saliva or cerebral spinal fluid from humans. That makes things a lot easier. So this is a sample prep step where there are lots of mistakes that are made. Once you've got to the sample prep, then you can go on to the analysis and you can use things like HPLC, or mass spec, or NMR, or GCMS, or a variety of other techniques to try and get an assessment of those molecules. From that, you're gonna get something that looks like this. This is an NMR spectrum, and it could be a mass spectrum, it could be an HPLC chromatogram. They're all a bunch of peaks. And what you have to do is try and identify those peaks or compare the spectra to each other so you can identify which peaks have changed. And then from there, there's a whole raft of other analysis. And that's what we're gonna focus on, mostly tomorrow and part of today. So metabolomics is a latecomer to the party. People have been doing genomics a lot longer, proteomics less long, but for a while as well. Metabolomics, as a field, is arguably about 10 or 12 years old. The name, the designation, the concept. And the reason why it's so young is because it's so tough. And this has a lot to do with chemical complexity or chemical diversity. So genomics is easy because chemically, you're just dealing with four different molecules. Chemistry was worked out decades ago. The enzymes, to do it, sangria sequencing. Really simple, just really one enzyme. So they've been able to refine it and refine it and refine it, and so now you have sequences that can sequence entire genomes in a day or less. Proteomics, also, is just 20 amino acids. Now there's a few other unnatural ones, but it's basically 20 amino acids. Chemically, very similar molecules allows you to do very similar kinds of processing. But with metabolomics, it's not four, it's not 20, it's 200,000 or more different chemicals with different properties and different features. And every one of them is different, so that means there's gonna be different analytical techniques you have to use. You can't, there isn't, and there never will be one method, one instrument, that can measure that much chemical diversity. So we use a bunch of techniques. We use ultra-high performance, liquid chromatography, high pressure, liquid chromatography, polyurelectrophoresis, and Imoidus, LCMS, GCMS, a variety of different types of mass spectrometers. We'll use NMR, we'll even crystallize compounds to try and figure out their structures. We'll use fluorescence and UV detection to detect or identify infrared spectroscopy has been used, is being used. Everything that you can possibly use in an analytical chemistry lab is brought to bear in trying to do metabolomics. So one of the first things that you have to deal with is that the fluids or the extracts you work with are incredibly complex. There are hundreds to thousands of molecules, and to deal with something that complex, you want to simplify it, and the best way of simplifying it is to separate things out. So one of the oldest analytical chemistry techniques is chromatography. And this is an example of what you can get with a separation technique to separate small molecules. This is integral to all aspects of sample preparation in metabolomics. And it's used in almost every method except for NMR. So chromatography, I think most people are familiar with it. It's trying to take something that's dissolved, mixed here in a mobile phase. So that could be blood or urine or a cell extract. And you're putting it through a stationary phase, which is the column, which is made up of usually some polymers. And that allows you to separate. There's, it comes to a differential partitioning. There's a whole range of separation techniques from thinlier chromatography, column chromatography, liquid phase, gas phase. In the liquid phase, there's affinity chromatography. There's separations by ion exchange, size exclusion, reverse phase, normal phase. You can work at gravity feed or high pressure, ultra high pressure. All of these things have been developed over the last century to help separate things. For small molecules, the thing that's preferred basically is high pressure or high performance liquid chromatography. That was largely developed in the 1970s in its revolutionized analytical chemistry and it's critical to all aspects of metabolomics. Typically use high pressures or ultra high pressures, 6,000 pounds per square inch. And you use very small particles, micron sized particles that can handle this pressure. And with HPLC people found soon that they were able to detect compounds at very low levels. And they're able to separate not just polar compounds but non-polar compounds as well. And so it opened a whole range of possibilities for seeing and identifying small molecules. So reverse phase is the one that's most commonly used and that separates non-polar molecules. And this is one of the reasons why most MS based metabolomics largely targets non-polar molecules because the separations you get typically with reverse phase are best. You also use a polar solvent to help get the separation happening. Another type of technique is normal phase chromatography. Most people don't know about this but it is a method for separating non-polar molecules. It doesn't work as well as reverse but sometimes it works better depending on the type of stationary phase. A new type of chromatography which has only been around for maybe a decade is hydrophilic interaction liquid chromatography or hyalic. And this is becoming more and more popular in metabolomics and it allows you to actually get a hold of those polar molecules which have always been so tough in mass spec based metabolomics. In this case, the stationary phase is polar and the mobile phase is largely non-polar. So it's just the reverse phase. So HPLC columns, there's different types. Some are glass, some are made up of this sort of plastic thing called peak but most of them are made up of stainless steel. Some are very short, some are very long, some are very thin, some are wide depending on whether they're analytical or preparative and for different properties that people may be looking at. There's almost a column for every application. But the most common one as I say is the reverse phase chromatography and these are the particles that are put into the column. They're usually silica or sand if you want and they're derivatized with usually an organic compound and you'll work on the designation of the compound so if it's a C18 column it means that you have 18 carbon polymers, alkanes that are stuck to the surface of the polymer so they're greasy silicates. You can also have a C4 column so this is a very short set of carbon, an alkyl chain so it's much shorter not quite as greasy. You can also have aromatic compounds attached to them and this one has a cyanoglue so there's a variety of derivatizations you can do to change what's on top of the column or on top of the surface of the silica. This decoration is the key for getting the separations. And so that's done, some people still know how to do their own derivatization but this is done by manufacturers and they will pack the column specifically for you and that's also a real art but once that's done then the column can be used hundreds or even thousands of times. The quality of the separation you get depends a lot on the column and what's inside the column. A short column, which usually means a very short separation time, will only give you modest resolution. A longer column gives you enhanced separation. Small beads will give you modest separation. Ultra small beads, which you find in the UPLC columns, give you higher resolution even over the same length of the column. So that's the basis to UPLC or UHPLC is to go for very small particle sizes and higher pressures, shorter columns, therefore shorter separation times. So the way that HPLC systems can work is you can have a simple solution, solve it, the mobile phase, it's pumped in at high pressures into the column to eject your sample into the column just before the mobile phase passes through and then you might use it UV or mass spec or fluorescent detector to see what comes out of the column. This is the simplest form of HPLC, the most common form of HPLC is not just to use a single mobile phase but to actually use one, two, or even three solvents to create a gradient of solvents, a mixture of solvents that allows you to separate things, sometimes more precisely, targeting different types of molecules, some that move better with this solvent, some that move better with this solvent and some that move better with, say, the third solvent. So to do this, you typically have to have a mixer that combines these solvents, pumps that change programs, that change the mixing cycles, that adjust the concentrations, but again, all of it flows through, all of it comes out and if you play around with your solvents and your gradient schemes, you can get some amazing separations but it's a real art. So this is what you can get, particularly if you've done a gradient separation for a complex biological mixture. You can see in this one there's probably about 30 or 40 different peaks. These peaks may correspond to 300 to 400 different compounds. So under any given peak, there may be 10 other compounds but you've simplified your mixture and that's critical. So that's liquid chromatography and usually that's tied to other kinds of detection systems. There's another technique which is gas chromatography. You don't use gas chromatography for proteins. It's only used for small molecules and typically it's similar to what you have with liquid chromatography. Here's your source if you want, instead of liquid it's a gas. Here's your injector. Here is your column. Here is your detector. Same setup, instead of a column that's 50 millimeters long, this is a gas thin column that's about 10 meters long in a very tight little coil and it's very, very thin. And obviously you're not working with liquids, you're working with gases. The reason why gas chromatography works only for small molecules is that you have to vaporize your small molecules. They have to be converted into a gas and you can't readily do that with proteins or peptides or gases. So once it's vaporized, heat it up and you can move it through a column. A column is also similar to what you would see but it's hollow. It has a stationary phase which is stuck to the inside of the column. So it's not filled with beads like it is with HLC. It's essentially a surface on the inside of a very thin column that has polymers stuck outside. So columns in gas chromatography are not millimeters but meters in length. And they're not measured in centimeters in width. They are very, very thin, two, three millimeters thin. To make molecules vaporize, you have to derivatize them. And the trick they've learned over the last 50 years is to use tetramethylsilane, TMS, which allows you to make these compounds very, very volatile. So TMS derivatization is a step you have to do. You chemically modify things in gas chromatography. Sometimes you'll do a methodolysis, which is a, or some other modification which this is a sugar and it may introduce some methyl groups here. But the essential reaction that you're doing is that from this, you add tetramethyl or trimethylsilane. Some kind of halogen here. And it will react with hydroxyl groups, amino groups, acidic groups, all kinds of groups. And basically convert something like this, which is not volatile into something like this, which will just go poof into the air at very low temperatures. So once you've got your derivatized compounds, so this chemistry that's involved, once they're derivatized, then you inject them. You send, not liquid, but gas, usually an inert gas like helium through the column and you'll get your separation. With some compounds that have no affinity to that derivatized column, they'll just zip through. They'll be the first to come through. But others that have some affinity to the stationary phase will be held back. They'll be dragged along, so they'll come out later. And this is the type of separation you can get with gas chromatography. Very, very narrow, tight peaks, very reproducible, much more reproducible than HPLC. So to put a 10 meter column in, you don't stretch it through a room and all the way, you just kind of coil it because it's basically a little wire, a hollow wire. But inside that wire, as I say, you'll find there's this metal and there's some fused silica and then inside that orange ring is the stationary phase. These are the organic molecules that you've derivatized the surface with, which allows the small molecules to absorb. So these are examples of some of the things that people will stick in side, polysiloxane. So you can see that they're, in this case, there's benzoyl groups that can be stuck to it. So whether it's gas chromatography or liquid chromatography, we talk about things coming off at certain times because they'll be run for 10 minutes, 20 minutes, an hour. So how long a compound is retained in the column is essentially before it comes out, it's called a retention time or an RT. So that can be affected both in liquid and gas chromatography by the compound itself, by the column, big columns, narrow columns, long columns, flow rate, pressure, HPLC, gas chromatography, carrier gas, carrier liquid, even the temperature, both for gas and liquid chromatography. If everything is the same and manufacturers are consistent and quality control is great, you can actually measure and identify compounds because they'll come off exactly at the same time, same retention time over and over and over again, particularly with gas chromatography. So if you know what's supposed to be in your sample, you don't really need a whole lot of detection, you don't even need a whole lot of mass spec analysis. If you can tell when something's coming off and it comes off at 7.23 minutes and that's the compound and you've tested it, that's your compound. And that's how much of clinical chemistry is done today. So the retention time is a time that you can standardize it to your attention in depth. And that's done in gas chromatography more than in liquid chromatography. And it's a reference time that allows you to normalize things from slightly different columns, but you use essentially a standard mixture of alkanes. And they come off at very defined times and you reference the time that your analyte comes off to this first bunch of alkanes that you've run a standardization curve with or a standardization test with. With gas chromatography, with liquid chromatography, you can identify and quantify, as I said, by looking at the retention time. So this particular compound, the Krilomide, always comes off at 2.85 minutes. So you just look in that interval, 2.85 plus or minus 0.01. And if you see your peak there, that's a Krilomide. How much a Krilomide is there? You measure the area under the curve, and you can calibrate that. So if you got a standard solution of a Krilomide, ran it through, and you ran your mixture through, you could actually measure how much a Krilomide really is there. You get a quantitative result. So again, seeing it for liquid chromatography, HPLC, this is an example of a GC run, great separation. GC, as a separation technique, always beats out UPLC and HPLC. But it's not something that a lot of people like or do, partly because it's, quote, old technology. Now whether it's HPLC or GC, it's still kind of useless unless you can detect it. So all of the chromatograms I was showing you still had something detecting the peaks. And so some cases it could be UV, some cases it could be fluorescence. In most cases, that was mass spectrometry. That was the thing that was allowing you to detect what was coming off. So mass spectrometry has been around for more than 100 years. It's a standard method to measure atomic weights. I think many people have seen mass spectrometers. This is a time of flight mass spec. Some are horizontal, some are vertical. They've been around for decades. Principles of mass spec are that you can identify compounds by their mass. Just like everyone in this room, we could potentially identify you by your weight. Because I don't think anyone weighs exactly the same. And so if we had a list of everyone's weight and name and we asked you to stand on blindfolded that we asked you to stand on a weight scale, we could figure out who's who. Same thing with small molecules. Many compounds can be uniquely identified by their mass. With the best instruments like time of flight or FTMS, you can determine the molecular weight down to one part per million. Or an accuracy of 0.0001 Dalton's. For big proteins, the accuracy isn't quite that good. But that still corresponds to maybe one Dalton with a 40 kilodalton protein. So with mass spectrometry, you typically attach a mass spec detector to a chromatography tool. So gas chromatography attach a mass spec becomes GCMS. Take a liquid chromatography, HPLC or UPLC attached it to a mass spec. It's now LCMS. You can also do separations on separations by mass spec. You can separate molecules by mass spectrometry. That's possible. And then you can smash those molecules into a gas and you can now do tandem mass spectrometry. And so you can do LCMS and so you can just do MSMS and you can do direct injection MS. So a variety of combinations. So mass spectrometry fortunately has been getting better and better every year. In the old days, the resolution of most instruments only allowed you to measure the average mass because you would see all of the different isotopomers just sort of merged into one giant peak. Now today it's routinely possible to measure the mono isotopic mass of most molecules. So here is a lipid and here is an example if you had a very high resolution mass spectrometer, you would see four or even five different peaks corresponding to different isotope variations of that same molecule. This is the parent ion mass. This is the mono isotopic mass. This is the natural abundant mass of the most common isotopes. So then there's a step of one, a step of one, a step of one, which is essentially the addition of one proton. That might be a carbon 13 variation or a deuterium variation of that molecule. So most organic molecules in living systems are just hydrogen carbon and nitrogen and those isotope abundances are very low. Deuterium 0.02%, carbon is 1.1%, nitrogen is about 0.1%. But occasionally when you get things like chlorine, you have a real mix of isotopes. You have the C-35, chlorine 35 and chlorine 37, almost an equal abundance. And so if you just had a regular molecule, so this is a chlorobenzene, you would see or could calculate based on the abundance of these isotopes what you would see. And so in the case of chlorine 35, you could have this one and chlorine 37, you could have a mix of chlorine 37 and chlorine 35. You have carbon 13, you have deuterium. All of these combinations would create a fairly abundant isotopamer peak that would have about 30% of the abundance of the main monoisotopic mass. And you can calculate the game with different combinations of different isotopes and actually calculate the intensity and the molecular weight of these isotopamers or isotopic variants. So a high resolution FTMS mass factor of this would show pattern like this. So that's actually fairly informative. And if you saw a pattern like this, it would actually not only tell you from the molecular weight, you could actually determine what the molecular formula is and to some extent even the structure. M over Z is the mass over charge. And so in mass spectrometry to be able to measure things, you actually have to charge them. Usually you'll convert them to a positive or a negative charge. And then they're sent down, it's typically a tube through high voltage plates that allows things to move around. If things aren't charged, you won't get mass spectrometry. So something is charged, something's don't. And so that's sometimes a problem. Sugars don't easily charge in mass spectrometry. So sugars are harder to detect in mass spec. So this is this sort of explaining it here is that typically we'll take a sample, whether it's from gas or liquid, and we'll put it into a little cone, this is the electric spray, with a very strong charge. And that charge will actually cause the liquid to sort of spray out like an aerosol can. And then it's sent into essentially an electric, set of electric fields or magnetic fields that will move the charges around. And then it'll be sent in or these particles will send into essentially a particle detector, which allows us to detect the abundance of those ions. So this is a mass spectrum. Essentially it could be an MS or an electron impact mass spectrum of aspirin. And what you'll see is a mass high up at the far end, which is the parent ion mass, usually mass plus one positive charge. The under-dead gulfins. And then what you're seeing are fragments of aspirin that come up with essentially breaking up of the molecule, especially if it's gone through an electron impact or a tandem mass spectrometer. And this is a pattern, a characteristic that actually is like a fingerprint for the molecule. So mass spec, unlike the HPLC or GC spectra that we saw, have very, very sharp and very narrow peaks. So this kind of looks like an HPLC picture, but we're not measuring time, we're measuring mass to charge. And this isn't of a mixture of molecules really. It's one molecule that's just being smashed up into different fragments. But those fragments allow us to identify what that molecule is. So that x-axis horizontal is the mass to charge ratio. That's an ion. And the height is the relative abundance of that ion. However, unlike in gas chromatography or liquid chromatography, the peak intensity doesn't tell you about the abundance of that molecule. That doesn't give you your concentration. It's very unreliable. Peak intensity is only a measure of the ability of that ion to fly or to desorb or become volatilized in a gas. So mass spectrometry's been getting better. Resolution's been getting better. And the way we measure resolution, just like we measure a resolution in HPLC, is by looking at the width of the peak relative to its mass. So it's sort of this delta M over M. So the measured mass of a molecule and then the width of those two peaks that you can resolve with your mass spectrometer. So here are two peaks. Here's another two peaks. And typically we use about the half height, 50% half height. And these two peaks are typically resolvable. And so the width at half height is that delta M, if you want. So here's a lower resolution mass spec. Looking at this, it's actually a peptide, but I'm like the mass of about $2,847. Here's M over Z, here's your intensity. And you can see it's one big peak. If you use a high resolution, the time of flight instead of an ion trap, you see six or seven peaks. This corresponds to the different isotopes or the isotopemers of this particular peptide. So different mass specs, and depending on how much money you're willing to pay, will give you different resolutions. So the blue curve is a low resolution mass spec, which would have unit mass, or not even unit mass resolution. So it's cheap as you can get very old instruments. The red peak gives you a resolution of about 3,000, and that's probably close to sort of what that ion trap was. The green peaks gives you a resolution of 10,000, and that might be a lower quality kind of flight. And then this black peak, a resolution of 30,000, and that might correspond to an orbitrap or an FTMS. What is an orbitrap? It's a type of mass spec. It's kind of an ion trap design, and it's specially designed for high resolution. So mass specs are made up of different components, and we've seen about the, essentially the HPLC allows you to send things in, but then how do you get things to convert into ions? There's different methods to convert that solution or that mixture into ions so that it can then go into the mass analyzer and then can be detected. And we're gonna talk about the ion source a little bit and we'll talk about the mass analyzers. So the oldest ionizer, the one that's most commonly used historically is called electron impact or electron ionization. Still probably the most useful method gives you detailed information about structures. There's another approach which is called chemical ionization. It doesn't smash up molecules. It produces fairly simple parent ion spectra. So this is called a hard ionization, a semi-hard ionization, and then these two are called soft ionization, and these have historically been used for proteins and for nucleosides, but they also work well for small molecules. And these have been done for very large proteins, particularly moldy. So electrospray ionization is ESI, and then matrix-assisted laser-dissorption ionization that's moldy. So these are different techniques, all of them usable, all of them viable, all used in metabolomics. There's no one that's particularly any better or any worse. They're used for certain applications. If you have the instrumentation to do all four, great because you'll see more compounds. Yeah, so ionization, hard ionization breaks the molecule up, it blows it up. Soft ionization keeps the molecule intact. So soft ionization means that you're gonna measure the parent ion. So take a car, you measure cars weight, it's 2,000 pounds, blow it up. Now you measure the tire and the wheels and the steering wheel. All of those pieces are what you measure with electron impact ionization. So for electron impact ionization, the process is to essentially send in your ions. They go through here, but instead of to make things tougher on those ions is you send in essentially gas molecules which are set-ups or flying in through here I guess with the sort of electron, 70 volt or 70 electron volt potential. And what happens is that those molecules that we're in in this chamber collide with those neutral molecules and they break up. And that smashing, so this is introduced gas-based bombardithy electrons from that filament, so it's a 70 electron volts. 70 electron volts is about 10, 50 times stronger than the electron volts or the kilocalories that hold bonds together. So essentially the electrons smash, your molecules up, they become ionized and they go zipping through into the mass analyzer. So electron impact is the standard mass spec technique used in gas chromatography. And it gives you detailed structural information. And older mass spectroscopists can look at molecules like methanol and draw out these equations by heart knowing exactly how molecules will fragment. So methanol can fragment into getting positive ion so that's your parent ion. But you can also lose a hydrogen, create this kind of unstable molecule, you can lose hydroxyl, you can create this. So you can take this molecule and use, let another hydrogen and create that. These are all types of fragments that you can see. And as they say, some people, not very many anymore, can look at molecules and actually predict a lot of the fragmentation patterns and therefore predict a lot of what's in the mass as they would see in an EI mass spec. So here's an EI mass spec of methanol. And so there is the parent or molecular ion of $32 and here's one where we've lost a hydrogen and then here's the other fragments here. And so these are the fragments that we see. This is a fingerprint. So if you see this fingerprint on a mass spec, an EI mass spec, you have no doubt that that is methanol. And so you can create fingerprints for all kinds of molecules. And because it's the same 70 volts, it's the same collision energy, everything is standardized, you can create a set of standard reference fingerprints that allows you to compare and identify hundreds, thousands of molecules. Yes. You're gonna have to speak up. It's very difficult. You can get some idea of what might be there so you can identify certain, you can say there's a Benzel group here, there's maybe some hydroxyl groups, there's a sulfate group. Those are things that you can identify but to try and piece them together, whether it's para or meta or whether there's two of them and not just one, you can't. So you have to know the compound before and to do novel compound identification. And I guess I'll be up front of it. The very best chemists, PhDs, do totally novel compound identification. Usually need about 10 to 20 milligrams of the compound. Can take them up to three years. There hasn't been any really new compound identified in metabolomics ever. Every case I'm aware of is people have ultimately found the compound. It was previously identified. 30, 50 years ago, whatever. It's non-trivial to determine novel compounds. So talking about the ionization techniques, these are the ways of getting things to fly to convert M to M over Z, create an ion. So the electrospray ionization typically has to be a fluid, great for biofluids. So this is the preferred technique. There's also methods where you can put a fluid on top of a surface, dry it off, and then use lasers and moldy to try and measure metabolites. This is how they're doing metabolite imaging. Electrospray, whether it's done for proteomics or metabolites, the same thing. Solvent is piped through a very small capillary. There's a sheath gas that passes through it. It's a high voltage. Gas and the high voltage cause essentially an aerosol to form. So you get a spray that comes out and then you go into progressively stronger and stronger vacuums that allow these ions to fly through your chamber or your analysis. So here's the droplets as they come out from the sheath flow and they start evaporating because the vacuum is so strong and then eventually they start to propel and they actually explode into single ions as they travel down the tube. That's the process of electrospray and electrospray ionization, where you're converting what may have been neutral molecules to charged molecules that can be positively or negatively charged. To do it, you actually have to work with very small capillaries at very low flow rates. Tens of microliters per minute. Strong voltage is applied at the tip to get your aerosol and you have to make sure that the carrier fluid is free of salt. That's really problematic because urine is full of salt. Blood is full of salt. Any extract you get is usually full of salt. So again, this is the reason why so much of mass spectrometry is with largely non-polar molecules because then you can dissolve them in organic solvents. Anyways, as I say, the ionization creates tiny little droplets. You can tune things depending on your solvents and your voltage to get little sprays or giant spray cones. And that's something that, again, that is worked out in their protocols that every lab has. You can go from micro-spray to nano-spray. Very small amounts of material. That's one of the great benefits of mass spectrometry. Salt and detergents kill the whole thing. You can work in either positive mode or negative mode depending on the molecules you're wanting to look at. So these are the ionization methods. Then there's the mass analysis. And I've already hinted at that before because there are different methods for separating your ions. The very first mass spectrometers use magnets. So magnetic sector analysis. Some of them are still made. They're huge, expensive, very accurate. But now the cheapest ones are the quadrical analyzers. They only have very low resolution. They're cheap. Time of flight analyzers, which measure how long it takes for an ion to pass through. So these instead of being magnets, this is century electric fields that you're working with. So very high resolution. And then the highest resolution of all is this ion cyclotron resonance, or FTICR type mass spectrometer. Very expensive instruments. Sometimes that hurt your words, Q-tough. That's right. The same on this one? So Q-tough is a quadruple time of flight. So it's a combination of these things. You can also make a quadruple, plus a quadruple, plus a quadruple, and that's called a triple quad. So you can get combinations. And people can even connect the magnetic sector with the ion cyclotron if they want. They will pair these things up to do, in some cases, different types of separation and different types of ionization, different types of measurements. So they're kind of almost like chromatographic instruments. So here's this measure of resolution or mass accuracy, I guess, in terms of how precisely you can measure masses. So you would call that very low resolution ion trap mass spec. So low resolution, low mass accuracy. Here's these three quadruples. Here's that Q-tough, if you're mentioned. So they're about equal in terms of their resolution. Time of flight equal to a Q-tough. There's that magnetic sector. No one uses these many much anymore. But here's the orbitrap, and here's the FTMS. So the orbitrap is a little worse than FTMS, but not too bad. And these are very, very high resolutions. So one part per million means that you're reading at point zero, zero, zero, zero, one Dalton's. Very accurate. So which one is the most popular? This is the most popular actually for historic reasons. It's a very stable instrument, very robust instrument, relatively cheap. If people have the money, they typically like to go to this. Because once you get up to here, with this kind of mass accuracy, you're able to take the mass and get a pretty good guess of what its molecular formula is. And we'll talk about that later. These ones, it's not quite possible, or you'll get too many formulas generated. So as I said, mass spectrometry is a little bit like chromatography because what you end up with is kind of like a retention time information because if something's going through an electric spray ionization, you're getting electric chromatography, you're getting compounds coming in over time, but then you're gonna be measuring information. So the stuff that you're measuring is actually the ion intensity, or the signal intensity is as these things go into the mass spectrometer and hit the detector. So there's things like that. Total ion current, or Tic chromatogram, that's the sumbed intensity. And there's the base peak chromatogram, B, B, C. And then there's the extracted ion chromatogram, E, I, C. So this is sort of like a total ion chromatogram. This is the base peak chromatogram. And this is the extracted ion chromatogram that you can get from mass chromatography. And different labs will display things in different ways. Some people will also sort of normalize, but this is probably a more understood type of chromatogram that you can generate from mass spec. So here's an example of essentially a chromatogram where we've taken HPLC with mass spec. So what we're showing here is essentially things like retention time, but then we're also showing some of the masses that were measured before those particular retention times. So one was a tomato extract, one is a rabidopsis, and you're seeing different abundances for different molecules or for the same molecules. So you can see these are clearly plants, but they have different levels of certain metabolites. So we're running out of time, but I'm gonna try and see if I can squeeze in the last part, which was another approach to doing metabolomics. So we've talked about chromatography, we've talked about mass spectrometry very quickly. This is another approach, and we're gonna segue into this into our next lab. So most of the people here do not do NMR, but because we could get the software for free, that's what we're gonna use. And in fact, I'm gonna try and persuade you that NMR is a useful technique for metabolomics. So how many have ever seen an NMR spectrum or worked on an NMR spectrometer? Anyways, I think people probably, if they've ever taken organic chemistry, at least seen an NMR spectrum, the idea with NMR is to use a magnet and to put a solution under a very strong magnetic field and then to shine or blast radio waves onto this magnetically excited fluid. And when those radio waves are absorbed, some of them are absorbed selectively. And so you'll get just like a UV absorption spectrum, and have some certain peaks that are absorbed very strongly and others where there's no absorption. And that absorption phenomenon is a function of what's inside the tube. If you don't have the magnet, then nothing happens. No RF or radio frequency absorption happens, so this is just a blank spectrum. So there's parallels to magnetic NMR and UV, where UV, we're looking at what's absorbed and we're measuring peaks. Same thing, it's just we're looking at what's absorbed with radio frequency radiation. In NMR, we're looking at how a nuclear magnetism, it's not like an electromagnet and it's not as if it's radioactive. It's just every compound has a nuclear magnetic moment. But we're just measuring how that magnetism changes, that's how we detect it. And that's essentially the absorption phenomenon. So we're measuring white, but we're not measuring UV light, we're measuring radio frequency light. And we can only see this phenomenon when something's put in a really strong magnetic field. So if you've ever had an MRI, we have to put you into a strong magnetic field in order to see things. If we didn't put you into a strong magnetic field, you're transparent to the radio waves. We also know that different nuclei absorb at different frequencies. And there's carbon nuclei and there's hydrogen nitrogen and fluorine. They all have different absorption frequencies. We also know from physics that the different protons, which are in the nucleus, have different types of spins. Some spin up when they rotate clockwise. Others seem to rotate counterclockwise. And so they have a down spin. So these are quantum numbers. Electrons have spin as well. And that explains the electron orbitals. Butters 2 and 8 and 16. So protons also have a spin and they have an up spin or a down spin. Because protons have a charge on them, if you've got a charge that's spinning, that creates a magnetic field. Same thing if you take a copper wire and wind it around a nail and put a little battery to it. It makes a current. Current goes around, spinning around. That's the spinning phenomenon. So anytime current goes around in a circle, it creates a little magnet. There's this little north and south pole. So if they're spinning up, the north pole is up. If they're spinning down, the north pole is down. So they like little miniature magnets. And so if you have a solution filled with molecules, which have protons, some of these will have spins that are up and others that are spins that are down. Spins that are down are lower energy. Spins that are up are high energy. If we send in a radio wave, some of these spins are gonna absorb. If it's the right radio frequency, they start resonating, nuclear magnetic resonance. They start oscillating and then they suddenly flip. And so you can see that this spin flipped up and went up. That's a high energy state. They absorbed the radio wave, absorption. That's what spectroscopy's about. And so we measure the absorption. And because these are electrical changes and magnetic changes, it's actually pretty easy to measure those things. So we can measure where that absorption has happened and what frequency or wavelength that happened. We can also change the strength of the magnetic field. And that also allows us to increase the frequency at which we see certain absorption phenomena. So big magnets are good. Huge magnets are even better. And so with a modern NMR instrument, you'll take fluids, you'll inject them into refrigerator-sized magnets. These can pick up a city bus. Extended radio waves go in, radio waves are absorbed. What's absorbed is then measured electronically. And we can measure the absorption spectrum, which is the NMR spectrum. So as I say, these are very powerful, superconducting magnets. They are big tanks filled with liquid helium, liquid nitrogen, layers and layers of insulation on the inside. But what the magnet is, is really just a special metal wrapped and wrapped miles of wire that allows you to keep that superconducting magnet going. And then inside the magnet is a tiny, tiny little coil, which is the radio frequency antenna and receiver. And that's the thing that sends in the radio waves and also measures the absorption of the radio waves. And so that's the tiny little antenna, and then you put a tiny little test tube inside that antenna. So if you've ever played with an NMR, you'll use these NMR and you'll put them in a spinner and you'll drop the tube all the way down into the magnet. And that's what you're measuring is this little saddle coil, which sends in the radio frequency. But if you didn't have it in the high magnetic field, nothing would happen. So we pulse it, we collect the absorption and these are the wavelengths where we see the absorption. They're very sharp, they're very well-defined, they look a lot like a mass spectrum to some extent, or an HPLC chromatogram. But they have chemical shifts, they have splitting patterns and they have different intensities. So this information tells us a lot. It tells us in terms of hydrogens, it tells us the type of compounds or groups that they're involved, and then how close things are. So NMR is actually the only way you can determine the structure of an unknown. Mass spectrometry can give you some hints, but NMR is the only way you can determine the structure of an unknown molecule. So the chemical shifts tell you an awful lot because different hydrogen shifts are defined by patterns and fingerprints, just like we talked about the fingerprint pattern for mass spectra. And they're affected by neighboring atoms. They're affected by bonds, they're affected by groups. And there are tables where people have worked out how an acidic group of chemical shifts that you typically see, if you're at tetramethylsilane, remember that combo? It almost always has a chemical shift near zero. Alkenes all have chemical shifts around one. Aromatic groups around seven or eight. These are all things that people have measured over and over, and they tell us a tremendous amount. And a skilled NMR spectroscopist can actually look at a spectrum and tell you very quickly which components are in an unknown molecule. So here's a Bromo ethane, and here's an example of where we can see three hydrogens from the methyl group and then two hydrogens from the methyl methyl group here. And so depending on the fact that you're very close to a bromine, very electronegative, it shifts the positions of these CH2's down here. This is further away from the bromine, so the chemical shift is closer to one part per million. The coupling patterns are all related to how close, whether there's two hydrogens nearby or three hydrogens nearby. Don't have time to give you everything about NMR, but this is just sort of concepts. Same sort of thing. Here's another molecule, another spectrum. We can see the aromatic, which is around seven PPM, the methyl group, the methyl group. What you're gonna learn in the next session is how to work a little bit with NMR spectra. These NMR spectra aren't quite as nice as we're showing here. Typically when you collect NMR spectra from urine or blood, they look like this. In this case, there's a huge urine peak, a urea peak, there's a huge water peak, things are out of phase, they're pretty messy. What you need to do in NMR spectroscopy is you need to fix the spectra so that this, which is too high, is brought down, this, which is out of phase, is brought up. The water, which is a mess, is eliminated. So these are things called shimming, water removal, phasing, get rid of this negative peak, and referencing, make sure that you're marking things at the right position, at zero parts per million, and making sure that the top is even with the lower one, so this is called baseline correction. And this is actually done for mass spectra as well. So this isn't just unique to NMR. There are things where you're doing baseline correction, there are things where you are eliminating large peaks, occasionally even with FTMS, there are things like phasing. So these are things like what you will see in other types of chromatographic work, but it's sort of unique to NMR. This just sort of explains these phenomena and you can look back at these in your book, so I'm not gonna talk about them now. But the thing about NMR is that unlike chromatography is we're not separating, so if we're looking at a bio fluid like blood or urine, this is not gone through a separation, but we're seeing what looks like a chromatogram because all the compounds in there have different chemical shifts, and they naturally disperse over the range of the detected. So this is standing from 5.5 to 9, so I could bring this up and put it here. So I'm not measuring time or retention time, but every one of these peaks corresponds to one or more compounds. So that's an essential advantage of NMR, but the other thing is that it's a quantitative technique. I can measure and tell you exactly how much what the concentration is for each of these compounds. Yes. Parts per million is a frequency, and the frequencies are 500,267,728.262, and if you had to say that every time, it's just really hard, so you might as well divide by 500 million and take the difference in that. It's something that we can all memorize. And because different spectrometers have different frequencies, it would be tough. This normalizes it so everyone can speak the same language and the same frequency. So it's like the retention index for gas chromatography. So this is what we're gonna try and do in the next round, and I'm just gonna wrap up. I'm not sure how much more I'll give on this, but different techniques, different sensitivity. Yes? Yeah. Does it know you get what each program has to say? No, it's a little more complicated than that, and we'll learn more about this later. So mass stack or chromatography, we know that this is one compound. Well in NMR, number 25 might actually correspond to, here's one peak of 25, and maybe here's 25 here, here's 25 again there. So three peaks are corresponding to 25 because 25 has three different protons in three very different locations in its chemical structure. So that actually gives you a much clearer picture of what's there. So if you can match all three peaks in exactly the same position with the right intensity, you have 100% confidence that that's the compound. Whereas with mass spec, you only see one peak, typically after you've normalized and done the isotope, and then say, okay, that's the mass, sort of the retention time, I'm not really sure. So in the case of NMR, if there's only one peak, which is sort of what you have for then with mass spec, we have to spike it in to be sure of what it is, which is what you should do for mass spec too, because you want to have an isotope variant to be sure that it's the real compound. So NMR, as appealing as it is in the sense of being able to confidently identify and quantify, is the least sensitive of all the techniques I've talked about. The most sensitive one is mass spectrometry, which is great, but there are differences. I mean, mass spectrometry is great for non-polar molecules, NMR is great for polar molecules. In the middle is gas chromatography, which is a nice compromise between what LC-MS and atom work does. This is the most quantitative method, this is the next most quantitative one, and this one is still struggling with good quantitation. The other thing is that as you get more and more sensitive, you start seeing the things that no one has any idea what they are, so there's more and more unknowns. Some of those unknowns are real, a lot of them aren't, and that's another challenge as you get to something that's too sensitive. So this is just sort of a comparison between the different techniques, and I'll just let you guys read through that because we're short on time, and this just sort of continues that comparison. So if you're given a sample and someone says go do metabolomics, different platforms give you different sensitivity and numbers of compounds. So NMR, 50 to 100 compounds with micro molar sensitivity. GCMS, 70 to 100 compounds with micro molar sensitivity. Direct injection in LC-MS, anywhere from 150 to 350 compounds with nano molar sensitivity. Lipidomics can measure up to 3,000. And there's certain types of other compounds where you need very specialized assays to detect them because their compounds don't work well or are too unstable with these other techniques. We're gonna be focusing, like there are two ways of doing metabolomics and they're somewhat similar or somewhat different. We're gonna be focusing here on quantitative or targeted methods. Historically, most metabolomics went through this way, which is a profiling approach, chemometric way, which is where you collect many, many spectra, compare them and see where they're different, identify which peaks cause the difference, and then you go back and say, what's the compound? Problem with that approach has been the differentiating compound is an unknown. I just told you how long it takes to identify an unknown and the fact that no unknown in metabolomics history has been identified. So it's actually better to stick with what you know. So this approach is to say, knowing what we know, let's identify all the compounds first and then try and differentiate. So concept with profiling is that it's the very first approach to metabolomics, take many, many samples, collect many, many spectra, do standard principle component analysis, find where the differences are, and then try and identify what the differentiating metabolite is. Sometimes you're lucky and that unknown is identifiable in the sense that it's already a known molecule. Many times you aren't. The other approach is to say, okay, take the sample, let's take the time to identify all that we know is in there, and from the list of known metabolites, then do the statistics, and then since we now know what the differentiating metabolites are, then we can do the biological interpretation. So I do have a bias for that, but I've done metabolomics for 15 years. I think I can tell you what works so it doesn't. So what you try and do, whether it's with targeted or untargeted metabolomics, is you, in the end, you wanna take your data and generate lists, compounds. It could be known, it could be some unknown compounds with some information about their relative or absolute quantity. From there, you wanna go from these lists, things like pathways, that's where we get the biological interpretation, so we're gonna talk about that later on. And then from those pathways and lists, we wanna go to things like biological models, systems, biology models, biomarkers, the stuff that we really care about. And so that's what we're gonna focus on for the rest of the day, the rest of the two days, and these just sort of highlight what we're gonna be looking at over the next couple lectures.