 This one is like basically a quick introduction to metabolomics. It's not gonna be really quick. I guess it'll be about an hour and 20 minutes What we do with these Lectures or labs as we always define some learning objectives And hopefully at the end of the lecture at the end of the lab You'll have actually achieved those objectives One of the things at least for this is we'll be looking at the size of the metabolome This is something that's still a topic of plenty of discussion and people who were around a long time ago Probably also were part of the discussions about the size of the genome Which was debated for for many many years Still not fully resolved at least for the human genome But in the case of metabolomes, it's it's changing every day and we'll talk a little bit about that We'll talk about some of the applications of metabolomics You will see certainly many more applications and many of you've already described some of the other applications that you are using in your own work We're gonna dive into the principles of liquid chromatography gas chromatography MS and NMR and it's gonna be a 30,000 foot view of it, but certainly if you'd like we can go into it in more detail if you have questions And then we're going to distinguish between targeted and untargeted metabolomics because there's sort of a fundamental difference between the two A lot of our work will be more towards targeted, but we'll also give untargeted metabolomics a bit of a try here later on Schedule and is already given you and you have that in your notes and Hopefully We'll stick reasonably close to the time. We might be a little slow off the start here So let's start with a slide I often begin with in in discussions on metabolomics So use this pyramid picture here and at the base of the pyramid is the genome and what we study or how we study the genome It's using genomics and next-gen sequencing RNA-seq and transcriptomics and as we progress up the pyramid Genes code for proteins and the study of proteins is proteomics and there's at least two or three of you who are also specialists in proteomics At the top of the pyramid being the most important is metabolomics But it's really to indicate that that Things sort of flow up because subtle small changes in the genome are amplified Particularly up to the metabolome and this makes the metabolome a particularly sensitive Indicator of what's going on in terms of physiology Francis No, it's probably true It used to be that we draw the same pyramids It's to say the number of things we knew because we could go like 23,000 genes and then the number of proteins that are involved actually enzymatically is about five or six thousand and then the number of Metabolites that we knew about ten years ago is about 2,500 now the pyramids in inverted Estimates now there's up to two million metabolites in the human body And of course, there's all kinds of alternate splice Variants in proteome, so it's much larger than the genome so In terms of size it's inverted in terms of our knowledge. Yeah, probably There is a lot and so this is it's a it's a nice time to be involved in metabolomics, I think it's it's a big Sandbox sure I Yeah, yeah, what genes do you see much That's true. I think I mean No, it's I think in terms of pathways. We know a basic metabolism very well and that's been known probably since the 60s The I Think what what we'll try and learn about or I'll try and emphasize a lot is that Metabolomics isn't just about Catabolism and anabolism Metabolites actually play a huge role in signaling and in gene activation deactivation They also play really important roles at a physiological level in terms of Events in the body and those pathways Aren't in keg. They aren't anywhere some of them are in textbooks and in that regard it's a very poorly studied part of metabolomics and it's Probably 99% of the functions of metabolites are in these other areas That we just don't have the pathways for And it's unfortunate I think that because I mean we tend to pluck the low-hanging fruit and because the pathways in keg and other places are easy to find We tend to interpret metabolomic data purely from catabolism and anabolism and some of it's true but I think we're starting to realize that we're missing a huge amount By not including or encompassing those those other Elements of homeostasis pathway regulation everything else So we'll talk about that a bit more, but I think this is another component here. I've got two arrows showing which also go up in terms of the the effect But both environment and physiology have on the genome and the proteome and the metabolome Many of you have just had breakfast That is changing your metabolome Hopefully it's not changing your genome If it is I think we'd all be pretty pretty much Zombie mutants, I think but the point being is that The metabolome is at that interface between the genome and environment and The mix between genome and environment is is is the phenotype. So in that regard the metabolome is your best indicator of the phenotype There's also something that we tend to forget those of you involved in Biochemistry and even cell biology is that we tend to think of Organisms as single cells human is just one very large single cell and we're not we actually have Dozens of different organs that have fundamentally different metabolomes that interact with each other and much of the organ systems in our body we're involved to deal with and to produce different types of metabolomes so Physiology which is something we tend to neglect or forget in the biochemical world is is something that the metabolome is is very much a Sensitive to and reflects what's going on in terms of physiology We'll also talk about that a bit more Again, this is a almost a throwaway slide because I'm sure all of you know about genomics and metabolomics But the the genesis of the metabolomic name Which as I said started with the word metabolomics and Kenomics and chemical genomics is really just trying to capture the excitement from especially the 90s of genomics and proteomics So the idea of using high throughput technologies the idea of using computers To try and characterize as much as you can as quickly as you can and of course the focus in metabolomics is small molecules Whereas in genomics it's on genes There's also been a while evolving definitions of what our metabolites The one that we've used Is essentially organic molecules although you can probably include inorganic salts now But the typical thing is a molecular weight less than 1500 Dalton's Some people use a cut off a thousand some will use two thousand but we're sticking with the middle and it captures most things That people would consider metabolites So it can include short peptides and it can include short oligonucleotides That's okay, but what we mostly think about in terms of metabolites are The other kinds small molecules so the sugars and the organic acids and amino acids and steroids and lipids But the other part to remember is that the metabolome of many organisms also consists of what? Organisms eat We eat other metabolomes and so our bodies are composed of those metabolomes So there are plant metabolomes in our bodies. There are microbial metabolomes in our bodies But there are also xenobiotics There are thousands of synthetic molecules that are added to the yogurt you've been eating or the bread you've been eating We touch and inhale things that go through our skin or lungs that includes various pollutants Variety of toxic compounds that are found in the water and just about anywhere else Drugs that drug metabolites whether the prescription or over-the-counter or supplements those are also part of the metabolome And obviously they're part of ours, but they're also part of many other animals and plants and microbes So in terms of the humans we'll look at both, you know human products Synthetic as well as endogenous microbial products same with plants The concentration limits are variable depends on your instruments But as a cutoff will say anything greater than picomolar, which is typically below the sensitivity of most instruments So gain the metabolome has a definition has evolved So essentially people used to consider only the endogenous things only the things that our bodies or where enzyme pathways Directed, but we now realize that you know one of the largest organs in our body is the is the microbiome We also realize that there's a huge number of exogenous compounds that accumulate or Transiently exist in our bodies There's also a large collection of what we'll call theoretical molecules where there is essentially no authentic standard that you can buy or Synthesize, but we know the compound actually has to exist because of either Some work done in the 50s or 60s where they did actually isolate it briefly or where they know the intermediate has to exist based on chemistry So that means that in some regards the metabolome Evolves because it depends on the technology that we use to detect things and as technology improves the metabolome expands But also on our definitions, so people will distinguish between endogenous and exogenous metabolites The fact that it's dependent on technology dependent on some definitions, which are still evolving Dependent on different tissues and organisms. It means the metabolome size is always going to be ill-defined So we'll never have an answer an absolute answer like we will have or already have for the size of the genome of say microbes E. Coli or C. Eleganz and maybe in a couple years will have a firm number for the human genome In terms of the size of metabolomes and metabolites Right now we figure there's about 60,000 molecules that we can Definitely say are in humans and other mammals As I say that the probable number based on Variety of calculations it could be up upwards of two million But these are 60,000 that appear in the literature Microbes is a big debate about how how diverse the microbial community is I've chosen a number of a hundred thousand being slightly larger and more complex or in terms of chemistry than humans My own gut feeling actually is that microbial metabolome is actually smaller On the other hand the plant metabolome is the most complicated and the reason why the plant metabolome is so much Larger and more complicated is because plants can't run away from threats So they've evolved a chemical warfare approach and they've created a huge vast number of We'll call secondary metabolites that is Growing all the time and so from a chemists point of view or natural products chemists point of view Plants are a real goldmine Just an amazing diversity and amazing source also for drug drug leads and drug-inspired compounds Now I'm not including some of the other things that are you know between plants and microbes. So these are might be small micro Micro-organisms but multi-cellular Flagellates and other kinds of primitive organisms some of which also probably have incredible diversity in terms of their chemical repertoire So ironically we tend to think of mammals as being you know the most complicated and sophisticated But chemically they're the least This is sort of an enumeration of what we know about human metabolones and human metabolites It is about 20,000 endogenous metabolites that we've annotated in the human metabolome database So these are things that your enzymes are designed to produce transport transform Ones that we can sort of track to to pathways that are reasonably well known There are many other microbial metabolites, and then of course there are all the other exogenous Metabolones, so there are drugs There are huge numbers of food metabolites or food compounds in your body So these include many plants and 80 to 90 percent of your caloric intake Actually comes from plants and plant derivatives Drugs get broken down into drug metabolites, and then there are a variety of I guess we'll call them pollutants environmental chemicals toxins What I've indicated here is the range and concentrations that you'll see so the most Concentrated or abundant metabolite in your body is urea It can get up to concentrations of several hundred millimolar And then you can find metabolites all the way down to picomolar femtomolar levels So there's a broader range in terms of concentrations for endogenous metabolites Many many orders of magnitude Drugs are less abundant, and they are roughly at the same level that you'll find for foods and food additives And drug metabolites at lower levels, and then you hope if you're healthy that the toxins are even much much lower levels There are a number of databases that we've established over the years that try and capture this information So the human metabolome database is one drug bank is another food database or food DB is another and the toxic Exposome or T3 DB is another one We'll talk more about these later on but these are examples of resources that that help you identify But also understand the roles of these chemicals and they're not just intended to be only for humans the same Metabolites are found in mice and rats and cows And many of these things are also found in plants and microbes in fact the human metabolome includes Hundreds and hundreds if not thousands of microbial metabolites already Now I think it's also important to remember that there's what we know, and then there's what we don't know And I've mentioned this term theoretical human metabolites, and these are things that largely represent the Peaks or features you'll see in LCMS studies particularly Of which there are often tens of thousands of unidentified features and We know from variations in lipid structure and fatty acid structure that it's possibly easily generate several hundred thousand lipids Which are not in the databases. We know through Biotransformation studies that it's easily possible to generate tens of thousands of drug metabolites The transformations that go on in food additives which are treated as xenobiotics just like drugs also Hundreds of thousands and then even within our own metabolome the endogenous molecules Cytochrome p450s the glucuronidase is all these enzymes that are used for phase one and phase two metabolism also will Pick on these molecules and do some interesting chemistry on them So these are secondary endogenous metabolites So I put some almost random numbers beside them But as I say I think that the feeling now is it's perhaps up to two million theoretical compounds that would exist in your bodies Probably a very very low concentrations sort of the picomolar nanomolar levels But some of which we see as these unknown features in LCMS So why is metabolomics important most of you are here because you think metabolomics is important But I think here's some numbers that might help also justify why you're here Interestingly even today more than 95% of the common diagnostic assays look for small molecules Medicine revolves around the test of small molecules So maybe the headline grabbing stories are about the new gene test or a new protein test But the default is still to look to metabolites and the reason is is because we can absolutely Quantify metabolites It's still very hard to do absolute quantitation for proteins and it's still very hard to do absolute quantitation for gene transcripts So wherever you can get an absolute number that makes for a very powerful clinical medical test Just under 90% of all known drugs are small molecules So again, even though the headlines are mostly about new antibody drugs for every antibody drug that makes it to the market There's still another nine or ten small molecules that are also making it to the market So the drug industry still depends fundamentally on small molecules Most of the inspiration for drugs comes from natural products from small molecules Likewise, even in the realm of genetics almost a third of identified genetic disorders are also metabolic diseases and Then as I think we talked about before the small molecules play a key role not just in general Catabolism and anabolism, but they are really the cofactors and signaling molecules for for most of the proteins They play a vital role. It's really underappreciated I think both within the metabolomics community, but everywhere else Just one question just in terms of diagnostics for proteomics, there's so much false positives and difficulty in doing samples lab to library machines Is this are they developing diagnostics that will be much more reproducible Yeah, so Next year is the big year the next year is a big year, but it doesn't come No, it's The difficulty was that proteomics evolved initially as an identification field Rather than quantification the origins of metabolomics come from purely analytical chemistry and analytical chemists have always been obsessed about quantitation and and ironically both with transcriptomics RNA-seq and Proteomics the focus is let's this identify and we're not worried about really how much is there or let's worry about the relative quantitation So in metabolomics we have isotopic standards and we can use SRM MRM Whatever to try and actually get absolute values And in fact it is possible. It's routinely possible now to get highly reproducible Precise quantitative numbers that means what you measure here is the same as what you measure in Botswana or Mongolia And those same numbers can be transferred across. That's the ideal Clinical test So proteomics has just got that just figured that out about three or four years ago And so there's a big push in proteomics to do quantitative work It means synthesizing all kinds of C-13 labeled peptides and having libraries and spending tens of thousands of dollars to get those But those exist now and people are doing some pretty good quantitative proteomics work and that's completely changing the field There are now pretty reliable biomarkers that you can measure in in proteomics So I think we're going to see a resurgence in terms of biomarkers But the preference at least for proteomics is as soon as you've got a protein test convert it to an ELISA. It's cheaper With metabolomics because you can use a lot of these things on triple quads and triple quads are everywhere every clinical Testing lab in the world has one It's cheap to do it. And so it's very easy to convert Something you've got in the lab to a clinical test Okay So this just emphasizes this point about the canaries and the coal mine Metabolites are the canaries of the genome and this is this amplification effect that I talked about with the pyramid So this is again another reason why metabolomics is important It's why it's used in clinical chemistry. It's why it's used in phenotyping in Plants and microbes and in mammals There's a temporal issue with metabolomics as well. I talked about because what you're eating is changing your metabolome In this case this person's metabolome is widely varying just because they're eating so quickly I think but it's it's a case that that many metabolites all of these things that are both foods and food products and digestion that's going on and it happens in seconds a moment the food touches your Tongue basically salivary enzymes are metabolizing it some of it's being absorbed very quickly It's going through the entire body. What's your breathing is also changing your metabolome, you know, hold your breath for 20 seconds and your metabolome changes hold your breath for 10 minutes Well, you're dead, but your metabolome has changed in that 10 minutes if we sequenced you We couldn't tell you that you're dead if we looked at your proteome. We probably couldn't tell you that you're dead So the proteome changes slower and in the case of eating food It's it's going to be a few enzymes insulin glucagon a couple other proteins will go up or down slowly Their effects are slower longer term and then as I said the genome is not supposed to change It's supposed to be incredibly stable if it wasn't we'd all be sort of these walking mutants. So This is this temporal sensitivity. So some think that's great others think that is a real problem and and I think you can look at it both ways correctly that way as We talked about metabolism at least as it relates to Catabolism and anabolism is pretty well understood We have more detailed pathways for metabolism than we have for gene signaling and protein signaling What we don't have though is we don't have the pathways for metabolite signaling and that's something that I think this is a real real issue If you do metabolomics, I think you'll find that because it's sort of at that top of the pyramid or the last part of the pipeline People will come to you and say, you know, how do I interpret my metabolomic data? You're gonna have to explain it both in terms of the proteome and the genome Um So that way metabolomics is uniquely connected many people doing Genomics will just be what is the silo on its own doesn't have to think about proteins and metabolites many people doing proteomics Also treated almost as a silo on its own But because metabolomics is one of the last emerging omics to appear But also because it's so close to the phenotype it ends up having to be much more connected to the other omics technologies and fields This connection is also mirrored in the fact that small molecules are used to make up the DNA in the RNA Small molecules are used to make up the proteins small molecules are used to make up the lipids which give the cells their shape and structure Obviously small molecules are the energy source the fuel for the body cofactors for signaling and you can kind of turn the whole field of biology a little bit on its head and thinking that really and I think it's quite justified the genome and the proteome really evolved to catalyze chemistry So That is the essence of living systems We're just trying to perform Chemistry and confined spaces by e-cells and we're trying to do it faster So you need proteins to make the proteins you need genes But this is this is how life evolved was just to make chemistry happen faster So the other part I think we're going to see and talk about is this how metabolomics because it's so connected to the other Omics really helps enable the concept of systems biology this integration of genome proteome metagenome metabolome So rather than seeing these things as layers Metabolomics through bioinformatics Tries to merge all of these things together and we'll see that tomorrow when we talk a little bit more about what's taboanalyst and other tools Lots of applications you guys have already mentioned many of the other things you're doing These are just some more examples some of which you're already doing but there's applications in clinical work in pharmaceutical work and toxicological environmental studies Food and beverage monitoring and testing Petrochemical analysis water quality assessments the the list just keeps on growing So I'm going to dive into the methods associated with metabolomics Again, this is sort of a 30,000 foot overview and it's partly to bring some people up to speed Some of you will probably start drifting off to sleep because you know a lot of these things already But some of you don't and I think as I said it's a diverse group and we feel we usually have to do this to keep everyone Sort of on par Does anyone have any questions about this first little bit? What I've mentioned Talked about Anyways, feel free to interrupt and chat if you'd like so in terms of workflow metabolomics works on many different substrates and sample types, so There's some people doing plant metabolomics. I'm doing bovine metabolomics some people doing microbial a lot of you doing human If the sample is solid We mush it up. We extract it if we can get something that is already fluidized That's our preference. So for plants we prefer sap or or or juices for humans We prefer blood or or or urine The reason is that most analytical instruments are designed for working with fluids Because this is what chemists tend to work with obviously, there's now work on tissue metabolomics and That's evolving and so again we can avoid the extraction procedure and do some imaging But even with tissues as a rule we just mush it up There's the chemical analysis step Which is LCGC, MS, NMR, IR whole bunch of different tools Most of those steps are pretty routine. I've been around for decades. The fundamental Revolution in metabolomics was going to the last step it was taking all that analytical stuff and translating it into compounds concentrations and Doing it with many cases a very complex mixture So that's what we're going to talk about most of today really So it's the software and databases that have emerged over the last 10 years that have enabled metabolomics It's not the extraction techniques. It's not the mass spec or NMR. Those have been around and largely Capable of doing these things for actually many decades and you look at the extraction techniques and many of them date from the 40s in the 50s So Other collaborators saying that when you are for example when you're studying Human physiology you're studying what we're studying is plasma for example, and that's essentially exchanging with the cells all the time so when you're mushing up everything when you're for an extraction you're Essentially mixing all the effects of all the different types of tissues Different types of cells both intercellular and extracellular and they're saying how can you be sure what you're seeing is an effect? Yeah, that's it's a good point. So I guess I should be repeating these questions, right? So the question is yeah, are we losing information by by extraction? Are we losing cellular information? And yes, we are the Ideally we'd love to be able to do single-cell metabolomics Because we are able to do single-cell sequencing single-cell transcriptomics But I think the other point to remind people is that that the metabolomics is designed and Metabolism and metabolites are really designed also to work at a physiological level so reducing Every individual to a whole bunch of trillions of cells Actually complicates the process rather than simplifies or really helps the process So we see and there are many hundreds of studies where they can look at plasma and see Significant effects in the metabolome for perturbations that are localized to just a single organ or a small number of cells in the tissue Cancer is often the best example But tumors, you know, a centimeter or two across can have very significant changes and you know over four or five liters of plasma It persists. What's often not appreciated is that many of those perturbations are concentrated in the urine So the urine is the repository for all things the body doesn't want and So by concentrating that you actually see an amplification now, it doesn't necessarily tell you exactly what's going on But because we know a fair bit about Catabolism and Abilism some of the sources for these molecules can be isolated or Localized or understood But yeah, if you want, you know a detailed tissue-specific Understanding this is I think they appeal for from metabolite imaging and you can see fascinating differences between Different tissues even within regions within a cell Polarization that happens and I think so there's there's different things that people expect or ask But I think from the perspective of people have been doing Genomics single cell proteomics They'll always complain that you know metabolomics just can't can't cut it You can't do single cell but I think you always have to counter that and metabolomics is about physiology Okay We're gonna talk a little bit about more about the coverage Again, we're using this pyramid. I like concept partly because we can always show these numbers In terms of what we can routinely measure obviously in a human or mammalian genomes It's about 20 22,000 genes proteomics routinely between five and 10,000 proteins now it's climbing It's getting better. The average metabolomic study Will identify less than 200 chemicals So experimentally Metabolomics is lagging way way behind genomics and proteomics Now in metabolomic studies people can claim that they see 10,000 or even 20,000 features But the actual number of identified compounds is is tiny So this is a fundamental weakness of metabolomics and figuring out ways of Improving or increasing that number I think is the major goal over the next few years in metabolomics research One of the reasons why it's so hard and why we cover so few compounds or components in metabolomics is because of the diversity So in DNA we just have to worry about the chemistry of four bases Proteomics we just have to worry about the chemistry of 20 protein amino acids the number of chemical classes in very large terms is something on the order of four or five thousand the number of chemicals is Hundreds of thousands to millions each of them requires either a separate Purification protocol or separate isolation or deterrent characterization So the fact that you're having to deal with so many different chemicals is the fundamental reason why Metabolomics is more difficult than proteomics and genomics And so if you're you know genomic and proteomic Colleagues or pointing fingers and laughing at you just tell them you're dealing with a much more complicated problem So it also means that to do metabolomics You can't get away with just a high-seq DNA sequencer or an aphi-metrics gene chip or a single Maldi mass spec for for proteomics You have to use a lot of different tools Chromatography HPLC uplc Capillatrophoresis general chromatography SP and then a whole bunch of different types of mass spectrometers Summit low-resolution summit high NMR compliments GC another critical one people using FTI are Solving the structures of some of these things actually requires crystallography So the technologies required by metabolomics are much broader because of the chemical complexity that you deal with So it makes it a more challenging field overall So I'm going to talk about some of these technologies We'll start off with chromatography separation So we're looking at Separating things on a from a mixture and we're using both a mobile phase and a stationary phase Chromatography comes from the word of separating by color and it's essentially how The very first chromatographic studies were done where they sort of separate food dyes or other equivalents There are many types of chromatography TLC thin layer chromatography something that some of you maybe many of you've done most people probably run columns But you can run it both in the liquid or gas so that mobile phase can be a gas phase There's certain types of affinity and ion exchange and science exclusion Primarily more for larger molecules in the case of small molecules. We will use reverse phase normal phase Helic Sometimes we'll just let things pass through with gravity other cases will use higher pressure and in the case of higher pressure We'll use HPLC high pressure or ultra high pressure or performance liquid chromatography So this has been fundamental to the advances in small molecule research for the last three or four decades Higher pressures are used Instead of 20 psi sort of the common Atmospheric pressure, it's 30 times greater. We use very small particles It's a very sensitive technique if you've got a good Detector and you can do separations of all kinds of different small molecules So reverse phase we typically separate non-polar molecules hydrophobic molecules in that case you use a non-polar C8 C18 Stationary phase and then you usually use some kind of polar solvent Methanol water or cedanitrile Normal phase which is actually the first modality for HPLC hardly used anymore Partly because the separations aren't so great But in this case you use a polar stationary phase instead of non-polar and a non-polar organic Mobile phase Helic which is picking up in popularity is To separate polar molecules and as it turns out most of the molecules in the body are polar And so this is a chromatography that's you know desperately needed and probably is becoming increasingly more widely used HPLC columns can be made up of different types of Material obviously they have to be able to withstand high pressures So there are Not too many but there are glass columns or peak columns of plastic and then most are stainless steel They are both preparative and analytical columns so the preparative ones are big thick tubes Generally relatively short and the analytical ones are somewhat narrower longer tubes if you Pull apart an HPLC column. Has anyone ever pulled apart or packed an HPLC column? Everyone just buys them now, but if you pull them apart for fun When your boss isn't looking you'll find that they're filled with beads and the beads are actually decorated with Chains and the chains are Referred when they talk about a C 18 column basically means an aliphatic chain of 18 carbons a C4 column will be an aliphatic chain of C4. So this is the hydrophobic almost lipid like Surface that's put on these beads that are a few microns across but you can also play around with other types of substrates you can put on Bifenyl groups you can put on Cetanitrile groups and other things that can change the the polarity of the column So the the the trick really is is what you put on the surface of these beads You can also improve the separation By changing the length of your column so a longer column and longer column run improves your separation You can get the same effect by working with smaller beads And so this is the basis to Uplc So the five micron beads are what you find in HPLC columns the one one and a half micron beads So what you find in the Uplc columns? But to push things through the tiny tiny beads means higher pressures So the simplest HPLC Setup and it's also same sort of thing you could do with gravity feed columns You'll have a solvent. That's your mobile phase you pump it through for high pressure. You you have a more complex solvent delivery system Samples injected in front of the pump between pump in the column and things are pushed out and you detect And so you can detect with UV you can detect with the vacuum light scattering You can detect with fluorescence or you can detect with mass back or a combination of all of those So the detector is obviously key it allows you to identify what's coming off You can also do gradients You can have a and b and this is the most common configuration for HPLC For lipids people use three and sometimes even four solvents to get some incredibly complex ternary quaternary gradients and The very very skilled HPLC people can get amazing separations by playing around with the solvents So it's not much more difficult than just sort of the running the one but Defining your your delivery and mixing processes often very challenging so HPLC runs typically take between 20 and 40 minutes Uplc runs can be shorter 10 minutes or less and this is the type of separation you can get So in this case we're seeing dozens of peaks But what under each of those peaks? There may be several compounds or hundreds of compounds in some cases And what we're signaling seeing is typically absorbent. So these are things that have UV Absorptive moieties Many other compounds are not and so there's lots of other hidden or unknown peaks that you aren't seeing with this particular separation So HPLC is great, but the best separation tool is gas chromatography and Unfortunately, not too many people appreciate it probably because it's an older technology But in terms of plate count, which is how people measure separation gas chromatography wins hands down So this is the simple setup it uses a much much longer column and instead of using a liquid Mobile phases it uses a gas mobile phase It also means having to make your Sample volatile it has to be turned into a gas so not everything likes to be turned into a gas There is still an inert phase that is responsible for the separation Usually it's a polymer that's attached to the column to say the columns are really long typically 10 meters or more and the Column widths are very tiny millimeters For some things in order to fly in a GC column you have to derivatize them with trimethylsilane So this makes them very volatile And there's other modifications you may have to do Ethoxymation other things may also be used to make things volatile But this is this TMS addition that here we're taking again a sugar and an example of how it's being volatilized with a trimethylsilane Derivatization Is a pain and it's one of the reasons why GCMS isn't universally used in It's it takes time its efficiencies are debatable and typically not everything is going to be modified equally so for this compound which is trimethylsilanated on 1 2 3 4 You'll end up with probably four or five different variants of that even though it's the same molecule Because it will have two TMS's or three TMS's or Only one TMS, but not the other anyways all of those things will lead to different separation properties Which makes it complicated On the other hand there are a lot of compounds that we don't appreciate most of the things that are responsible for the taste of food our volatile So our tongue can only taste Four or five things you know sweet sour bitter But our noses Effectively taste Thousands of things so the flavor of an apple is not what you taste on your tongue It's it's basically the aromas that are picked up to there. So the volatile components of Foods and beverages same thing with wines and beers again It's the taste comes from what's in the volatile those are things that are ideally characterized by GCMS So here's this typically the inert gas usually helium Separation some things that have higher affinity stick to the walls of the column those migrate more slowly those that come off more quickly Are less have a lower avidity or affinity? And you can see the level of separation Which you get in GCMS, which is just amazing It's it's if all chromatography could be this way then Our lives would be very very simple the Instead of say a C18 column as we use an HPLC There's typically a compound called polysiloxane that's attached or Adhered to the interior of the column. So remember this is very thin very thin too and this polymer Which is a mix of both aliphatic and aromatic compounds Is what's responsible for the affinity of different? Volatiles or volatilized compounds Now the time it takes for something to move through either HPLC column or GC column is called the retention time So it's how long it takes that anolite to pass through the cone So as I say, it's used the term is used both for gas chromatography and liquid chromatography It's affected by a lot of different things. It's affected by column dimensions by the material that's inside the column flow rates Temperatures a huge difference In the case of gas, it's the pressure as well as the carrier in the case of HPLC and UPLC. It's also a pressure So retention time calculations for HPLC are basically hopeless because there's so many different columns however in the case of gas chromatography and It is possible to predict quite accurately and to calibrate using a thing called the retention index and This is by using the retention time and normalizing to Set of LK. It's a fairly standard set. So GC is remarkably standardized and standardizable as Is GC MS? Which we'll talk about a bit more Whereas LC MS is just about anything under the Sun So this makes it really tough and it's kind of unfortunate because in GC MS It was the only game in town for many years the analytical chemists quickly moved in and said this is how you must do it The evolution of LC MS was largely I think driven by people in mass spec in Proteomics who were trying to come up with every possible way of doing it and so no one stepped in and said this is how you have to standardize it So from the metabolomics perspective we need to go towards more standardized approaches This is sexy how you can do some compound identification and quantification when it comes to GC or LC as well And this is measuring both retention time and The area under the curve. So this might be attached to some kind of UV or Flaminization or a fluorescent or ELSD We're not attaching to a mass spec, but this is still a valid way for doing identification and quantification By knowing the retention time and then by measuring or having other another physical property you can characterize So the area under the curve area under the peak is is still often one of the best routes for Quantifying via either HPLC or GC game this is just a picture and if you compare or flip back and forth between the HPLC and the GC you can see this remarkable difference in terms of precision and plate count and the narrow narrow peaks So those are quick perspectives on chromatography There's also How to detect things and in this case LC and GC are usually attached to mass specs So in mass spectrometry, we're just trying to measure the weight of molecules to some extent or the mass to charge ratio Here's an older style Qtoff instrument will explain a little bit more, but the concept is that you can distinguish molecules by their molecular weight and Different molecules have different masses Just like if we weight everyone in this room, we could probably uniquely identify you by your weight I doubt if there's anyone who's weighing exactly the same and so this is the concept with mass spectrometry For for chemicals in particular with mass spec we can measure with incredible precision now less than a PPM and With that kind of precision it's possible to determine the atomic or molecular formula For large molecules, it's also become very very powerful and means even for 40 and 50 kilo Dalton proteins We can measure the atomic or molecular weight to to one Dalton So with mass spectrometry we couple it either to a GC instrument or an LC instrument or to another mass spec So you're gonna have GC MS LC MS or MS MS So you can do separations by a mass spectrometry and then also fragmentation followed by a more characterization by mass spec And then you could have LC MS MS You have GC LC MS MS, but anyways it gets a little crazy With mass spectrometry we have Peaks typically with the defined mass to charge ratios and with the better instruments these days you can get easily identifying the individual Isotopic components C13 or deuterated or chlorinated compounds and so you will see a collection of Typically sharp narrow peaks if you have a lower resolution instrument you will see Sort of a single peak Which represents an average of all of those masses in this case is the average mass Which in this case for this compound would be 11 5 6 point 3? There'd be a big mound So what we typically see in higher resolution mass specs Q toff's Orbitraps FTMS we is this collection of these isotopic peaks and this has to do as I said with the abundance of certain isotopes deuterium carbon in this case chlorobenzene chlorine which has very high abundance of chlorine 37 almost 32 percent and So if you think about the formula and combining deuterated and C13 and chlorine compounds You can see that there's different combinations which will include different abundance of C13 or deuterium to produce this array of Molecular weights from 112 213 214 915 and so on so calculating either in your head or With an isotopic calculator you can come up with how many? Different conformations with different combinations of isotopes and you can estimate the intensity of those peaks as well as their positions That information is incredibly valuable for determining a molecular formula And so this is what the chlorobenzene Spectrum would look like in a higher resolution mass spec and you can see that there's it's not just the steady drop In peak intensities as you would have with something without chlorine you see this second peak showing up It's two dolpins away. That's it's more intense So isotopic distributions are important for molecular formula determination, but they're also important to understand as these sort of extra Peaks that seem to always show up beside your primary peak So at mass spec The general principles are the same for just about all mass spec instruments. There's an ionization step There's a mass analyzer, which actually performs separation through Either magnets or or quadruples or traps and then there's a detector which allows you to see your signal So a mass spectrum looks like a gas chromatogram actually very sharp peaks Instead of having time on the axis you have m over zed, but there's an intensity which relates to the detector and the abundance of those ions and This is the I think this is an EIMS of aspirin So like gas chromatography very sharp narrow peaks The mass to charge ratio is on the x-axis the intensity is basically the abundance of the ion But unlike in gas or liquid chromatography where the intensity of the peak is basically related to the abundance It's not directly Quantifiable in mass spectrometry You can't look at your peak and simply say based on the area under this peak. This is my concentration You have to have another known reference that you spiked in usually an isotopic reference or a calibration curve to actually get a Quantitation so mass spectrometry on its own without these isotopic pairs or Spikens cannot be quantifiable Whereas with chromatography you can quantify and this has to do with the fact that ions have a different ability to fly Which is still not theoretically understood so different mass spectrometers have different resolution and resolving power and that's measured by the mass that you're measuring and the mass difference between two masses that you can potentially separate So this is equivalent of resolving power in a microscope So how many nanometers or microns can you go down to distinguish? So you can see on the right there those two peaks But the top are easily resolved you can see that there are two peaks if we sum the areas Down and that second one it might be hard for some of you to distinguish between those So we're just at the point where we can resolve those two peaks So this is sort of the theoretical one and it's the same thing that's used in NMR spectroscopy Everywhere, it's just what is your ability to resolve between two peaks? So the narrower the peaks are the better your resolution So here's a low resolution linear ion trap instrument mass spec It has a resolution of about one Dalton and you can see just barely that there's sort of these other peaks in there But they're too broad they all merge and you can't see the isotopic distribution Run the same thing on a high resolution time of flight instrument and you can see all the peaks So this is a case where the peaks are narrower. It's still the same Window in terms of mass Variations, but the narrower peaks are now resolvable So this is an illustration looking at if you want the old technology of 15 or 20 years ago To the latest probably in terms of orbit trap looking at the resolving power Where in this case we're looking at a molecule that had a mon isotopic weight of 3400 so it's a peptide But with a resolving power of a thousand that's a big broad blue peak Versus a resolving power of 30,000 which are those narrow easily seen black lines so there's a huge difference and Obviously better the resolution more likely you can actually characterize the molecule in terms of determining its its Molecular formula so This is a Schematic of a mass spec. There's an inlet There is Which is typically your HPLC or GC then there's an iron source, which is the thing that converts the molecules into ions Then there's the analyzer which essentially is responsible for separating the ions With their mass to charge ratio and then there's the detector To keep all of this working you need lots of vacuum pumps And these are the things that cause no end of pain to mass spectroscopists Along with actually the iron source if it's not Regularly clean So the ionization Approaches differ so mass spectrometry requires ions you have to convert something to either a positively charged ion or a negatively charged ion So the electron ionization is probably the original ionization method And it's called EI and it's also Designated as the hard ionization method because it it fragments things into tiny tiny components It's a very useful technique because of the fragmentation Actually allows you to determine the structure of molecules and there are still a few people around who can look at a GCMS spectrum and actually figure out the structure of a molecule now there's a Chemical ionization which showed up a little later It's a game designed for small molecules, and it's not going to break things up into tiny fragments, but it's it's a gentler approach It's useful for certain kinds of molecules And has emerged is actually increasingly being used Then there's ESI and Maldi. So these are the ones that have been developed for proteomics primarily and Couple of Nobel prizes have been awarded to people that developed the concepts They're very very soft ionization methods So they're just able to get molecules sort of a single or a few charges But you're not fragmenting the molecule very much if at all so the soft ionizations were great for peptides and proteins They're normally now used for small molecules to get the parent ion mass But we typically want to then fragment the small molecules again to get some structural information using a triple quad or other collision cell So EIMS because it was developed by analytical chemists became very very standardized almost immediately small molecules are let in Electrons are emitted from a filament typically tungsten or rhenium and they're sent off at a fine voltage 70 electron volts never varies or never supposed to vary And then the ions that are generated are fed into the mass spec or the analyzer if you want You need the high energies 70 electron volts to break up the bonds in your chemical Because the chemical bonds are only held together with the strength of about five or six electron volts So this bombardment this impact with electrons is the thing that's designed to shatter Small molecules Those fragments are then things that are analyzed so EIMS is standard with GCMS So standardization again is retention indices for GCMS standard voltage for for EIMS Makes for actually a really useful and easy way for characterizing small molecules So this is an example of what happens with the EIM packed with a simple molecule like methanol First you'll ionize it so it can fly so it gets a positive charge, but then you can knock off Other hydrogens and produce a various ions can fragment and removes a hydroxyl group So you've just got a positively charged methyl group And then you can create other variations So an EIMS spectrum of something as simple as methanol or produce a mass spectrum which would have four or five different peaks and the positions of those peaks are uniquely characteristic of this molecule and represent its fingerprint or its signature So if you're familiar with NMR, this is almost like an NMR spectrum. It's sufficient to determine the structure of this molecule Now the soft ionization methods EI ESI and Maldi are different Maldi uses the laser to Excite or cause little explosions and send off ions ESI uses a high voltage Most metabolomics methods use ESI With imaging people are starting to do that, but there's also efforts now to develop small molecule Maldi systems For things that are matrix free I'll talk more about electric spray because as it says the most common in this case things are fed in through a fluid usually your HPLC and What you do is you have a sharp tip with a gas sheath around the fluid and a high voltage and For reasons that are still not completely known. This causes an aerosol To form so things start spraying out in a tiny tiny Mist basically that is now charged So if you could look at it in detail this little capillary that's being fed into this vacuum change chamber is spitting out fluid the droplets are Charged because of this high voltage and as they pass into the vacuum. They start rapidly evaporating and as they evaporate the Droplet becomes increasingly concentrated that the analyte Concentration increases and increases, but it's now filled with these charged molecules and then they basically Explode because of the charge repulsion that goes on and so they Explode and evaporate even further So this is the concept of how ESI largely works, but it in the end you're left with just You hope single ions Some of which may have one or two or three charges But just these ions that are representing your analyte of interest So you have to use polar buffers Let it reasonably volatile. You can't work with salts because that messes ESI up Stainless steel is needed to prevent things from corroding strong voltages of three or four thousand volts and you aerosol things and Depending on the polarity of your solvent. You can produce different types of aerosol events at different voltages So this just illustrates, you know high water versus high acetonitrile at which point you'll get this spraying phenomenon You can have high volumes or low volumes the nano spray techniques are now much preferred over the Micro spray where you work with very low flow It's very sensitive You can get away with tiny amounts, but you have to get rid of detergents. You have to get rid of salts in ESI you can work both in a positive mode and a negative mode and that depends on the type of solvent or Carry your your adding so some people add a bit of formic acid to get a positive ion a bit of ammonia to get negative ions So the charging and the spraying or the ion source depends on the instrument We could go for several hours just talking about that. We don't have time We're just sort of yes. They're giving you a quick overview. There's the mass analyzer This is the thing that takes those ions that have been generated and Separates them based on their mass to charge ratio So there's time of flight Quadrupole iron traps orbit traps magnetic sector Fourier transform ion cyclotron resonance instruments There's a whole bunch and so when you start mixing and matching the ion sources with the type of mass analyzers you can end up with dozens of instruments and This is of course great for manufacturers, but it's not great for standardization And and this is a problem for for metabolomics But anyways the original mass analyzer was a magnetic sector analyzer big giant magnets Very high resolution. They were very popular in the 70s, especially for doing environmental monitoring Because you needed to do that for Things like the EPA and you can get very high high resolution exact mass quadruples Single quads or triple quads very low resolution But as I mentioned earlier, these are ubiquitous in just about every clinical chemistry lab in the world Time of flight is sort of the poor man's high resolution mass spec. They're getting to be very very good actually They support very high throughput Very very high resolution now The highest resolution mass spec is the ion cyclotron resonance mass spectrometer They're not that popular They're Very expensive and they're very low throughput instruments actually. I don't know has anyone ever worked with an FTICR Just one survivor Now, you know Victoria has one of the the I think it's the biggest one in Canada you get some amazingly high resolution data with with the FTMS and for certain applications that are the only instrument you can use so in terms of Accuracy or in some respects resolution. This is what you typically see for these different instruments So 0.1 to 1 ppm for an FTMS the orbitrap Below 1 ppm Time of flight instruments are also approaching that now so they can get to 1 to 2 ppm though older ones are 3 to 5 The magnetic sector which is almost no one has anymore, but it's almost matched what the orbitrap could do the triple quad If you do some tricks you can get higher resolution, but typically It's generally more on the order of 50 to 100 ppm, which is a similar to the ion trap so Different instruments different resolution different precision or accuracy So the output from a mass spec usually coupled to LC Will be a collection which you'll have over time Either a total ion current chromatogram tick a base peak chromatogram BPC Or what's most commonly used is the extracted ion chromatogram Which is basically one of the analytes extracted from the base peak chromatogram or the total ion care So the base bottom there is just illustrating the different things you will see or the types of data that are collected and or stored from a mass spec LCMS type run So one which is the red one an unpleasant looking one The base peak chromatogram, which is generally more appealing and used to impress your friends And then the extracted ion chromatogram, which is where you do most of your work So if you're working with LCMS, you typically will have an LC chromatogram and then associated with that You'll be able to extract your ions and then identify Hopefully these individual peaks with individual masses. So what's marked here is both the LC run But also the mass associated mass Therefore potentially the compound under those peaks So Again quick overview of mass spec We're going to jump to NMR spectroscopy now So how many people do LC or GCMS for their metabolomics? So about half and how many people do NMR for their metabolomics? Zero so maybe I'll just race through this one, but in fact So Anyways NMR Was actually probably the original method for metabolomics. This is what launched the field And it was a group in London Jeremy Nicholson based at Imperial College which approached or used NMR spectroscopy to look at mixtures and NMR spectra look a lot like GC spectra or high-releases so very very narrow peaks distributed over a chemical shift range The concept with NMR is Totally different from mass spec. You're not weighing molecules What you are doing is you're putting a solution under a high magnetic field and Then you send radio waves into this magnetized fluid and you measure the absorption So you basically have a giant Radio receiver transceiver transmitter and you're measuring absorbance. And so just like with a color absorbance with UV or You will see bands that are absorbed Not necessarily on a color But different bands at different chemical shifts So nuclear magnetic resonance measures nuclear magnetism Doesn't work with radioactive compounds Which is sometimes Confusing things for people it measures radio waves or light at radio frequencies typically on the order of several hundred Megahertz so a little higher frequencies than your FM bands on radio When you're measuring NMR, you're looking at changes in the nucleus and the nuclear spin You can only have something that are something it's NMR active if it's under a strong magnetic field and Only certain types of nuclei will absorb at different energies just like with UV only aromatic molecules absorb Sort of in the UV range only certain types of nuclei can absorb Radio frequency energy Fortunately hydrogen is one of the strongest absorbers and hydrogen is just about it in every single organic molecule Which is made NMR incredibly useful So in the nucleus you have protons and neutrons Every atom every molecule has protons and these protons spin and Because they have a charge Anytime you have a charge that's spinning or rotating you produce a magnetic field so something is spinning proverbially up just like a clockwise versus counterclockwise spin or spinning down you will have a Sort of the right-hand rule with something pointing up or Counterclockwise a spin pointing down when you've got protons spinning they produce little miniature magnetic fields So these little mini magnets which are in every molecule under Well all the time Can be detected so a sample which has trillions of hydrogen molecules or atoms Can be oriented so when you have a very very strong magnet measured in tens of Tesla Tesla Which is enough to pick up a city bus basically You have a material or a sample that's basically primed So if you send in a radio frequency wave at the appropriate frequency you will cause Spins to flip which is equivalent of causing a magnet to flip So we excite the sample with the electromagnetic radiation of the appropriate frequency It causes what used to be so the blue spins to spin up or flip up And become red spins or red atoms these are high energy forms And then we turn off the radio frequency and things will relax and so they will then sort of flip down or Oscillate as they flip back down. We measure that oscillation over time After that excitation or that pulse of radio frequency and that allows us to measure the frequency of absorption of those particular Atoms In NMR you need really big magnets and the stronger the magnet the higher the frequency That you're able to measure the greater the dispersion the higher the resolution So the biggest magnets in the world Measure frequencies at about 1.2 gigahertz Or billion Hertz Most NMR instruments have around five or six hundred megahertz and those are measured in terms of 10 to 15 Tesla So in NMR you're not working with Gases You're working with liquids You pull the liquids up and you'll put them in some cases either into tubes or into flow cells and And the magnet which is shown as that silver Canister which is about the size of a refrigerator is connected to a radio frequency transmitter and receiver Which is also about the size of a refrigerator and then the signals that are collected Come out of a probe and are transferred to a computer Magnets are about a million dollars. They can weigh several tons They are superconducting magnets And they are heavily insulated with Layers both of liquid helium on the inside and liquid nitrogen They are not electro magnets. They're actually permanent magnets. They're charged once and as long as you keep them cool They will stay going forever or almost forever They unique thing about NMR is That they're incredibly reliable So an NMR instrument never goes down if you take care of it Whereas I think if anyone's worked with a GCMS or an LCMS they go down every two weeks So this is a central advantage of NMR This is just yes It is yeah Yeah, the liquid nitrogen is actually used to cool off the liquid helium which is in the interior So you're trying to keep the magnet at about four degrees Kelvin or minus 265 Celcius so this is you know very very cold It's kept that cold by the liquid helium So you can see the green layer in that cross section That's where the liquid helium is surrounding the magnet coils which are a niobium tin alloy Which are wrapped around and then you have layers of liquid nitrogen to help keep things even cooler or cold all the time You drop the sample in through the top in the bottom you have a probe which is made up of this Which has a little saddle coil. It's a a little set of wires and Essentially a tube which contains your sample, which is not much wider than a pen or a pencil sits inside that saddle coil and that's where all of the magnetic and radio frequency Activity happens so remember radio waves have both an electric and a magnetic Component and so it's that magnetic component of the radio waves that are responsible for pulsing The sample in the NMR tube So there's an example of an NMR tube as I say it's about the size of a pencil it sits in that saddle coil And that's where all the magic happens, but it has to be surrounded by this very very strong magnetic field So an NMR spectrum as I said it looks a lot like a GC MS type or GC spectrum or an MS spectrum narrow narrow peaks You're not separating by mass to charge. You're not separating by time. You're separating by chemical shift So the chemical shifts are reported in frequencies or in parts per million You'll have some interesting patterns these are splitting patterns due to spin coupling and Then you'll have different intensities So unlike in mass spec where the intensity is a matter of how well the ions fly in NMR The intensity is directly proportional to the number of hydrogens NMR allows you to quantify very precisely So not only NMR is it very reliable. It's also quantifiable It also does the separations automatically you don't have to run anything through a chromatogram Because things are separated on the basis of chemical shifts So this is why NMR in fact was the first technology used for metabolomics it did everything all at once it quantified It separated it allowed you to identify the compounds because you can look at both the chemical shifts and the splitting patterns To unambiguously identify the compounds the chemical shifts are the reason why NMR works Different hydrogen atoms in different molecules will absorb at different frequencies So there's a unique pattern of chemical shifts just like there's a unique pattern for EIMS for molecules So these are fingerprints And that chemical shift is defined by the electronegativity and this is a standard chart So again, there are people who can look at an NMR spectrum and Actually figure out what the molecule is Because they've memorized this little chart and they have a good understanding of chemistry So this is an example of a really simple spectrum. So this is a bromo ethane But you can see how the influence of the bromine atom leads to Pushing or shifting the a atom so the a protons further down Downfield and then the B atoms less downfield So this is the electronegativity of bromine shifting things and then you can see also the coupling patterns where there's a Quartet and a triplet That's caused by the spin couplings by the proximity to the CH3 or CH2 atoms So again somebody skilled in NMR could just look at that spectrum not knowing the molecule and probably guess Structure in the composition pretty accurately. It's another example and again a fairly simple spectra Indicating this is the influence of aromatic rings. So you'll see chemical shifts around seven parts per million and then Substituents now NMR spectra are when you initially collect them quite ugly They have to be phased. They have to be referenced. They have to be Samples have to be shimmed. There's has to be some baseline correction to give things Some some flattening so you can see the top one, which is what you get immediately after Your pulses and then the fixed spectrum which generally looks a lot nicer and is far more useful And it's the same sort of thing with the mass spec versus the total ion chromatogram versus the extracted ion Sort of the same Extraction or fixing you do and these are the the steps that people will do to reference tetramethylsilane or DSS such in a sulfonated derivative TMS Shimming to make the lines look nice phasing to make things look like they're all Pointing up in the same direction You normally collect in water from a tabulomac. So you have to get rid of the water signal and you try and get rid of the wobbly lines So this is an NMR spectrum of mixture So it looks a lot like something like a GCMS and LCMS and an LC they all have a Great deal of similarity, but the thing that distinguishes them typically is the x-axis. Is it time? Is it m over z? Is it ppm? Now why isn't everyone using NMR? Because there's lots of pluses for it a lot of it has to do with sensitivity So this graph just illustrates the differences between sensitivity and And the instruments here the Mass spec LCMS you're able to detect in picomolar levels up to 10,000 features NMR it's typically down to a hundred features or a hundred compounds and Typically micromolar sensitivity GCMS is sort of in the middle in the case of NMR because it's so insensitive. You're usually measuring things that are known So that's not terribly exciting to people In the world of LCMS you're measuring or hopefully measuring things that have never been seen before Problem is that as I say we only identify about 200 compounds even with LCMS, which is roughly the same you measure by NMR So currently there's no difference What we're hoping over time is that you know with LCMS will measure thousandths, but that's still not happening Can I ask a question on that? So If you move to more and more sensitivity It comes at the expense of throughput like you're gonna you know, there's a there's a trade-off between at least Samples So people like sensitivity, but we're already more sensitive than what we can understand So, you know right now NMR is That's nice even stable so you can characterize and understand everything in NMR So From one perspective, that's probably all we should be working with Until we can figure out what all these other compounds are that we can't identify It is yeah Yeah So this is just a quick comparison between the different types of techniques and given their time because they're running out of time I'll just have people sort of sort of look at that but typically in a more low sensitivity So you need lots of volume Mass spec LCMS you don't need very much at all. You can get away with 10 microliters GCMS again sort of intermediate There's the differences in throughput Some I think with with automation they can almost all be about the same in terms of throughput But again the limit of detection is quite different so micro molar for NMR nano even Pico molar for LCMS Interestingly there's not a lot of overlap So what you measure when NMR is often not what you'll be able to measure in GCMS and is not what you'll be able to measure in LCMS So if you're doing a comprehensive metabolomic study you should use all three techniques This just underlines it says a typical results that people can get so NMR Actually holds a couple records for most compounds identified even though it's the least sensitive technique GCMS some people can get a lot of compounds identified not very many quantified LCMS methods it's people are able to get up to several hundred now in certain certain situations lipid omics People can identify many many classes of metabolites lipids, but not the unique lipids typically But again Significant advances are happening The other thing to remember and this is sort of the last point was to talk about the distinction between targeted and Untargeted metabolomics. How many people do untargeted metabolomics? About half how many people do targeted metabolomics About a third So The concept between untargeted metabolomics is just the original version or view of metabolomics Which is to just get lots and lots of samples From cohorts or collections Treat the the spectra the resulting spectra as you know without assigning them Cluster the data so convert the lists of peaks or or bin the peaks whatever you want to do And and perform some data reduction, which allows to identify what features are Significant The things are at this stage still unassigned Then once you've identified which features are significant or important then to go back and identify Or try your best to identify them tablets Targeted methods are different in the sense that they will Right away either use specific standards or spend a great deal of time doing compound identification and quantification at the start Then take all the lists of the known compounds known concentrations and do the data reduction PCA PST and then do the data interpretation after that now There is a very strong trend now to actually go more and more towards targeted or quantitative metabolomics And this is because in large part people have found that the untargeted approach is Leading you to this sort of tantalizing set of Compounds, but you can't identify Of course, you can't publish And so you've done all this work and you're left sort of you know hanging high and trying The targeted approach has actually allowed you to work with identified metabolites But if you're doing the targeting well, you're quantifying precisely And so you can still report the same compounds that everyone's always seeing but because you know the concentrations You can actually distinguish the differences quite objectively and consistently So that means you can publish And it also allows to come up with things like biomarkers which then are reproducible and usable in other other labs question I guess you're you're using targeted in terms of both when you have a platform of Metabolites you've already described and characterized as well as the approach of first trying to identify In some literature I've been reading it's that the targeted refers to a specific platform or a specific Yeah, yeah, and this is I don't like the term targeted Which is why usually put in word quantitative Because yeah, some people will have sets and it depends on the platform So if you do mass spec you have to have a sort of collection of a library that you put in if you do NMR, you don't have a library of compounds. You just have a reference Set of spectra that you look up GCS is also a little different too, and then there's other techniques where you can do chemically selective targeted but in the end if you can quantify and if you identify and So if that bundle I'll call targeted Then you're way ahead of the game But if you treat things simply as here's a collection of features I'm just going to process the features and this is the chemometric approach Which was the original way of doing it? you still end up with a lot of data, but not much to publish on and And It's it's been very problematic, and it's I think one of the things that has held metabolomics back so Typically what you try to do in the end is you go from your spectra to lists So whether it's the untargeted or target you still want to come up with some lists of compounds Now ideally the lists aren't just the compounds. You might have some relative or absolute quantitation From the list you want to go to pathways Pathways allow you to do some biological interpretation and relate to physiology or environment or genes and proteins And this is this integration and connection From pathways you can also go to models. This is where the systems biology comes in But also to biomarkers, which is where the clinical or veterinary or botanical or farming applications come in So we're going to talk about those things in the next session about going from spectra lists And we're also going to be talking about going from lists to pathways and biomarkers, and that's really the subject for the next two days really so We're behind but I guess we'll