 You think from his house, maybe from California or something. OK. OK. Molly, you're on the screen here to say no. You ready? Let me kick things off for you. OK. Great. Well, good afternoon, everyone, and thank you for joining us for this special seminar. As her thesis advisor, I am excited to introduce Molly McDevitt, who is here to defend her thesis in partial fulfillment of the requirements for earning a PhD in biochemistry from the University of Wisconsin-Madison. Molly is from Iowa, Sioux City, Iowa, but went off to Omaha, Nebraska to Creighton for her undergraduate work, where she was quite a star student. She had a near perfect GPA and had an extensive research experience, including looking at heat shock responses in insect cells. Molly then decided to stay on at Creighton for a couple of years as a technician, where she further developed a broad range of research techniques studying, among other things, riboswitches, and then decided to go on to graduate school and we first came across Molly as a standout candidate for the IPIB program, not only from her great work at Creighton and her time as a technician, but really just across the board, great application, glowing layers of radical recommendation that gave her an opportunity to become a fellow on a competitive NIH training grant, the Molecular Biosciences Training Grant, which helped us recruit Molly here. So Molly, after her rotations, became very interested both in the technical side of mass spectrometry and in its biological application. And so she bravely took on a co-mentoring role with myself and with Professor Josh Kuhn, who's on her thesis committee here today. And after a couple of years of that and realizing it was hard enough to tolerate one of us, let alone two of us, Molly and Josh and I decided for her to join a single lab at the time that it made more sense for Molly to finish the biological aspects in my group under what Josh is still close advisement. So Molly's work has really taken our group in a different direction. As you'll see, it's really been applying these mass spectrometry skills to large genetic reference populations using systems genetics approaches to try to find new connections between genes and metabolites and lipids. It's really exciting work. It's led to two very nice co-lead author publications for Molly in a great journal, Cell Systems, and with another nice project in good shape for another one that somebody else would take on. So graduate school is not easy. It's hard, even when it's easy. And in my group, we like to talk about what are those traits, what are those characteristics that we value, that we think are important for our lab, for our culture, for keeping us ahead of our competition in a competitive field. And there are many of them, but we talk about how it'd be great if we could just be smarter than our competition. We could just figure things out that no one else figures out. Sometimes we do. It'd be great if we could be luckier and just serendipity came along. And sometimes that does as well. But one thing we talk about is that there's really no other way to ensure your success than to be willing to do things that your competitors aren't ready to do. To dive in, to be willing to struggle and wrestle with your projects, to have that perseverance and that grit to see through these hard projects, because they will be hard. And not all struggles look the same. But I could tell you that knowing Molly well and knowing her project, that I think there have been few if any that have demonstrated that kind of grit and that tenacity to persevere in grad school and to see through a very difficult project. She's really embodied that for our lab and that's not just great for Molly. That's not just great for a project. It's great for all of us and it's important for all of us. So I'm proud of her for demonstrating that for our group. Another thing I like to tell new faculty members is that when you're building a lab, when you're establishing your own organization, your own culture, that you of course need to get people that are talented like Molly. But what you really also need is you have to get people that you like because you're spending a lot of time with each other and you have to get people that you trust because you're putting a lot in people's hands. And with Molly it's been quite easy to check those two boxes. She is quite simply just a great person, a salt of the earth, just wonderful person. And she's a lot of fun. There's just a lot to like about Molly. She's famous for her love of cats and fantasy novels and Halloween costumes and many other things. And there's a lot we'll miss. I will miss, among other things, the seemingly annual event when a new member of our lab for the first time hears an absolutely blood curling scream coming from someplace in the cubicles of the lab and comes running to the rest of us only to be met with like a complete look of indifference. And a while like looking frightened and confused gets a shrug and says, that's just Molly. For those of you who have no idea what I'm talking about, Molly can be jumpy at times and I have to kind of like creep up in her cube like, does she see me? Because if I said the wrong thing, I'd be getting stuck back. But that's just one of our favorite traits about Molly. There are many things that we love and will miss. Molly, I am proud of you. I am proud of your science. And I'm excited that you now as our 10th PhD student get to tell your story, so take it away. All right, thank you, Dave, for that lovely introduction. Gonna make me cry right off the bat here. So today I'm gonna talk to you about one of the things that I did in grad school, a systems biochemistry approach, investigating lipid metabolism. And the first thing that I'm gonna talk to you about today is the role of lipids in cells. They're not just a triglycerides and cholesterol that we think about. In addition, I'm gonna talk to you a little bit about the vast array and differences in lipid species, merely by mixing together a few different building blocks. And one thing that's really difficult about fully understanding lipid metabolism is that lipids are quantitative in nature, meaning that the abundance of the lipid arises from both genetic factors, as well as environmental factors. And one way to identify novel genes that are regulating lipid metabolism in some way is a method called quantitative trait loci or QTL mapping, which I will also be discussing. QTL mapping gives you hundreds or thousands of genes to look at. And so one of the very necessary things moving forward is prioritizing these genes that are under these QTLs and trying to identify the actual gene that is causing the phenotype in question. And even though I'm gonna spend a lot of time talking about QTL mapping, I'm gonna take a little segue and talk about one of the interesting stories that we found where lipids in the plasma of mice reflected their metabolic health. And furthermore, they reflected whether or not they had this non-alcoholic fatty liver disease or naffody. Fats have a very negative reputation in the media. I think about if I asked my parents what a fat was, they'd probably say it's bad for you. That's it. We all know the horror stories of high cholesterol and high triglycerides. And there's no denying that obesity is a worldwide epidemic at this point. However, that's not the whole picture. There's actually way more different lipids than cholesterol and triglycerides. And they serve a number of different functions. First, they're important for energy storage and transport. So here we ingest dietary triglycerides and cholesterol and they're uptaken by the small intestine and reformed, this is a very simplified model into these dense lipoproteins that triglycerides are extremely hydrophobic. And so the body tries to condense them all into a single spot. These lipoproteins filled with various proteins as well as triglycerides move through the bloodstream and will go back to the liver where they're repackaged and sent back out or they're sent on to other tissues to serve as energy, for example, for muscles that don't contain a lot of fat but might require more energy via beta oxidation. Secondly, they're vital to membrane structure. So sometimes we think of a cell as just this globular circle and in actuality, there's a bunch of different lipids making up these membranes, causing various different curvatures of these membranes. And the curvatures are important, one for fission and fusion of organelles. It also is sometimes used by specific proteins to actually bind to the organelle or cell membrane. And so there are three main types of lipids. The first is, the first two both form a cone shape. The head group in the first one is larger than the tails creating a normal, I guess, an ice cream cone shape. And this causes the membrane to pop out a little bit. Well, the opposite where the tails are bigger than the head group of the lipids, you get an inversion. And then there's also a number of cylindrical lipids which make up most of the membrane, which cause no membrane curvature at all. And finally, lipids are important in various signaling mechanisms. So various lipids can hear this is a very simple schematic a some messenger binds to a receptor on a cell membrane or an organelle membrane, which activates a protein, which may activate a second protein causing some cellular response. And lipids can actually be part of all different steps of this pathway. For example, lipids can bind to receptors and instigate this initial step. They can also act as second messengers causing this cascade that leads to an eventual cellular response. As I mentioned, lipids are extremely diverse. There is approximately 43,000 known lipids at this time and it's estimated that there's up to 100,000. Interestingly, they're all actually made up of very, very similar building blocks. So the hydrophobic tails of lipids are made up of three, mainly made up of three different groups. A fatty acyl group, an isoprene group or a ketide. And these can repeat a number of different times. So each tail can contain various numbers of carbons. They can also contain various numbers of points of unsaturation and those points of unsaturation can be found in different locations. And so you can see how the diversity of lipids increases exponentially. And then they're also made up of a head group which the tail is then connected to. Either one, two or three different tails depending on the species. I will talk a little bit more about these in the next slide. So there are eight different categories of lipids. Fatty acyls are these long hydrocarbon chains made up of any number of carbon molecules. Glycerolipids which are made up of this glycerol head group and three separate fatty acyl tails. Glycerophospholipids have these same fatty acid tails but also have a phosphate group as well as any number of different head groups. And these are often the lipids that we see causing the membrane curvature that I discussed earlier. And of note, glycerolipids include the triglycerides that I will be talking about later. Sphingolipids are similar and they have these fatty acid tails but they also contain a sphingocene head group shown here. Sacral lipids contain a sugar, polyketides are just repeats of this subunit here. Sterils include cholesterol like we talked about and phenols which I won't talk about a lot but is very interesting to our group include a very important mitochondrial lipid coenzyme Q. So as I mentioned, lipids are quantitative traits meaning that they're regulated by environmental and genetic factors and I think it's really easy to see how diet or environmental components may affect lipid abundance. We know that if you eat a high fat diet you often gain weight, become obese, et cetera. So that's very easy to see. These are some examples here. So here looking at total triglycerides on the Y axis, this is a normal chow diet and this is a high fat sucrose diet which simulates the Western diet. And we can see that just altering diet drastically affects the total amount of triglycerides in these mice. Additionally, if things like age, for example, here we're looking at abundance and four to 11 weeks or old at 78 weeks and you can see that there are differences in these lipids based on the age of the patient. So it's easy to see how environmental components can affect lipid abundance but it's less clear how genetics affects lipid abundance. And one thing I want to talk about is the extremely complex nature of how all of these lipids are co-regulated. For example, this phosphatidic acid and diacylglycerol shown here are almost like hubs of lipid metabolism and give rise to a number of other lipids. And you can see how it's possible that altering the amount of PA, phosphatidic acid in a sample could affect diacylglycerol levels, triglyceride levels, even other phospholipids. And so each of these steps here requires one or more enzymes. And depending on how quickly these enzymes work, whether they're functional at all, that can greatly affect the abundance of each of these lipids. And because this is so intertwined, a defect going from a lysophosphatidic acid species to a phosphatidic acid species may actually affect a number of other lipids, which makes it really, really difficult to identify the exact genes that are regulating lipid metabolism. So what is a quantitative trait? A qualitative trait is like a yes-no trait. So you either have it or you don't. There is, whereas a quantitative trait, because it's affected by environment and genetic components, it actually looks more like a normal distribution shown here. And there might be some threshold at which individual gets a disease or a certain number of genes required for an individual to have red hair, for example. And one way to try to identify novel genes involved in regulating these quantitative traits is through quantitative trait loci mapping. In general terms, QTL mapping identifies genetic loci that correlate with various phenotypes. So you can do any phenotype of interest. The only thing that is required is you have to have a phenotype that is changing across the genetically diverse population. If the two samples are genetically identical, then you know that any differences that you see aren't coming from a genetic perspective, but rather coming from the environment. So if you have a variable phenotype and a genetically diverse population, you're able to identify genetic positions that in which a gene that is causing this phenotype is located. And I'll go into a little bit more details of how exactly we did that here. There are a number of ways to identify QTLs, but genetic reference populations, such as the BXD mouse cohort, are extremely useful. So two founder strains, the C57 Black 6, DBA2J, these are very common lab mouse strains. They were initially crossed and in the F1, all of these mice are genetically identical, as you can see. Therefore, that would be useless for QTL mapping. We wouldn't be able to identify any gene that caused any difference. These are brother-sister-mated and in the F2 population, we start to see some genetic diversity. And it's possible to do QTL mapping from an F2 generation. However, the BXD cohort and other cohorts like it, take it a step farther. You can see then F2 generation, these are heterozygous at various alleles. So it makes it, you have to account for heterozygosity and you have to genotype every new mouse that you make. So what they did is brother-sister mate these mice over 20 generations to create these completely inbred lines where they're homozygous at each allele, but each individual line is genetically unique from the other ones. And this is advantageous for a number of reasons. First, as I mentioned, you don't have to worry about heterozygosity. Second, you only have to genotype a single mouse. So any mouse from this line is assumed to be genetically identical. Third, here you can't do replicates because you can't control homologous recombination, but with the BXD cohort, you can actually do replicates and get statistical information. And finally, perhaps most importantly, the data that you get from these mice is completely collaborative, meaning the data that our collaborators in Switzerland, the Johan at Johan Orx's group, data that they got on these mice about their general phenotypes, their mRNA levels, their protein levels, et cetera, can all be applied to the lipid data that we got. This is a very powerful tool and this multi-omics approach is gonna be very important moving forward. In order to identify regions that may contain a gene that regulates a phenotype of interest, again, a simplified model, there are regions of the genome called single nucleotide polymorphisms or SNPs. And a SNP means that the B6 mouse and the D2 mouse have a different allele there that's shown by the B and the D here. So across the genome at every SNP, you would look at all of these different lines and indicate whether they had the D allele like shown here or the B allele as shown here. And then graph this against your phenotype of interest. Then you calculate a LOD or log rhythmic of odd score that looks at the probability that there's a QTL at this particular SNP over the probability that there isn't a QTL at all. And you graph this across genetic positions and you get these spikes here that indicate that there is likely a gene somewhere in this region that is regulating your phenotype of interest. One thing that's very cool about QTLs is that they can be mapped for any variable trait including mRNA and protein levels which are pretty new, especially mapping protein levels. And so as I got into this project, I actually originally was gonna look at protein levels and instead we started thinking about lipids. No one has looked at lipids before. And our hypothesis that individual lipid species were quantitative traits that could be mapped. And this was new because previous to our research, generally people had looked at total triglyceride levels, total cholesterol levels and not individual species. And as I'm gonna show you later on, not all lipid species of the same class are regulated in the same way. And so to lump all of them together like that can sometimes mask a lot of information. So in collaboration with Arnie Ulbrich in the Coon Lab, we identified 96 different hepatic lipids across 385 mouse strains. These, this was a lot of work so I'm gonna take a little bit of time on this. I first homogenized the livers and did a BCA to normalize for protein content and then extractions, extracted the lipids and ran these samples. And I would very much like to thank Adam Yocum for helping me with this without him. I probably would have lost my sanity as I homogenized 300 plus livers. From this data we identified 136 QTLs, 55 of them were in chow diet and 81 in high fat diet. And let me just orient you here. So here you're looking at the negative log 10 P value which is a significant threshold. And down here is every single chromosome in mice and each dot here represents a different lipid. Here this blue line represents a suggestive QTL meaning that it's possible there is a QTL there but it is not, it hasn't quite reached significance and the black line here shows significant QTLs. And as you can see, a lot of our QTLs that we identified were suggestive but we do have some that were significant. So here I just want to tell you tag means triglyceride, dag is diacylglycerol, these are different fatty acids. Here these two blue and the purple. PL is phospholipids, pink is a mitochondrial lipid called cardiolipin and oranges, coenzyme Q which is also a mitochondrial lipid. And what I think if we look directly at those significant QTLs I think it's easier to appreciate how large these regions actually are. Here it just looks like a bunch of dots all combined. And so here we're looking at the significant QTLs and the species are indicated here and color coded. And as you can see, these actually span large regions of the genome and these regions of the genome can include hundreds of genes. So the big question in QTL mapping is the whole point is to go from a variable trait to identifying a single gene that is regulating that trait. So how do we take this vast number of genes and try to narrow down the list? And so we used a pretty simple strategy. So we use this pipeline to prioritize genes for future follow up. First we looked at genes that contain a genetic variant meaning genes that in the B6 versus D2 mice are different because an amino acid has changed. For example, it may cause an early stop codon forcing the protein to not be fully processed or made. You can imagine that if the protein isn't there you're more likely to see variability in the phenotypes that it's causing. Next we limited our search to genes with EQTLs which is expression, so transcript QTLs and PQTLs are protein. And the cis here just means that when those traits were mapped they mapped back to the region of the genome in which the gene was actually located. Next we limited our search to genes that correlated with lipid species which I think makes sense given that if there's more of a given protein there's likely to be more of its product for example. And as I mentioned before variability is required for QTL mapping and so we look for genes that are variable in liver and mRNA levels. And from this we were able to narrow down the list of genes under each of these QTLs. In red font are genes fulfilling all four criteria. Orange is three criteria and gray is two. We were even able to identify some genes that have already been associated with liver lipid metabolism. Now future work is going to focus on fully verifying some of these genes but for now I'm gonna set that aside. We wanted to look at lipid levels in plasma because unlike liver where the lipid levels of liver are determined by the liver itself plasma is kind of a snapshot of the whole body. Lipids from adipose tissue, liver, et cetera all come together in the plasma and we wanted to know if there were any lipid species in plasma that reflected their abundance in liver. And so similar to the liver study took 280 mice and extracted lipids from plasma of 49 different strains on two different diets and we were able to identify 129 lipids per animal this time around which is a significant increase from what we were able to identify in the liver manuscript. We identified significantly more QTLs from the plasma data and I think this is for two reasons. First, there was a lot less manipulation of the sample prior to extraction. We instead of having to normalize for protein content, homogenize, et cetera we were just able to extract immediately. And I think that is indicated by the increased reproducibility that we see. One thing we noticed as I mentioned before was that lipids of the same class can be differentially regulated and we notice this in both the liver and the plasma. Here I'm showing liver lipid species. Down here is their correlation with liver weight so red indicates it correlates positively with liver weight and blue correlates negatively. Along this axis are a number of lipid species and I want you to take notice of the fact that there are triglycerides here that if we go along this axis over to here they correlate positively with liver weight. However, there are also a number of triglycerides that correlate negatively to liver weight. And this is something that we saw over and over and over again. And we continually saw that triglyceride species with lower points of unsaturation or less double bonds were more likely to correlate positively with liver weight while those with increased numbers of double bonds correlated negatively with liver weight. And as I mentioned, we saw this in plasma as well. So here on this axis we're looking at a number of different lipid species. I'll highlight these triglycerides again here. And on this axis we're looking at a number of phenotypes measured from these mice. So for example, there is heart weight, how big the heart is, lean mass, fasting insulin levels. These are all traits that tend to be reflective of metabolic health. And there's supposed to be some brackets here. But basically these healthy traits in this blue square here correlated, the healthy traits correlated negatively with unhealthy lipid species while healthy species correlated positively with these healthy traits. And so next we wanted to look at, well, these species can be reflective of metabolic health. And perhaps maybe they can tell us a little more about disease. So we looked into non-alcoholic fatty liver disease or naffody. Hopefully most of us here have a healthy liver. Naffody is caused by an increased level of triglycerides in the liver. And it can either stay at naffody or continue on to steatosis. Non-alcoholic steatosis means parts of the liver starts to die off. And cirrhosis or cancer and eventually death. And so one of the things that is interesting about naffody right now is that there isn't necessarily, in order to diagnose someone with naffody, you usually have to take a biopsy, which is invasive and painful. And we wondered if maybe some of the plasma species reflected the abundance in liver and correlated with liver weight. And then instead of having to do a biopsy, perhaps in the future, an individual would just have to do a blood test in order to test whether they had this disease or not. And so the first thing that we looked for was whether or not the abundance in plasma reflected that abundance in liver, so whether they correlated together. And here's an example of one species. This is in chow diet. We're looking at the abundance in plasma versus liver. And you can see that this is a relatively straight line indicating that as abundance increases in the liver, it also increases in the plasma. And even more important than this was the fact that there were nine different lipid species that reflected the liver abundance in plasma irrespective of the diet. So this is that same lipid species. And in black now, we can see that high fat diet species also correlated in plasma and liver. And this is important because controlling the environmental factors of human patients is extremely difficult. You can tell someone to eat something. The chances that they actually follow that diet are not that high. And so we really wanted to look at lipid species that reflected their abundance in liver no matter what diet they were on because we knew that we couldn't control that in a human population. As I mentioned, there were nine species that followed these criteria. As you can see here, most of them are triglyceride species. And there's one diacylglyceride species that actually correlated negatively in liver and plasma. And so armed with this data, we next thought, okay, well, these seem to be indicative of liver weight. How do these different lipids correlate with naffody traits? So on the y-axis here, there are a number of different naffody readouts such as body weight, liver mass, fasting glucose, fasting insulin that are indicative of liver health. And again, blue is negatively correlated and red is positively correlated. And when we look at, we see that these red species here correlate positively with these naffody readouts. And we call these pro-naffody signatures. Well, as the blue correlated negatively with these naffody readouts, and we call them anti-naffody signatures. Now, I wanna be clear that the use of the word signature is we haven't identified a true biomarker at this point. And I just wanna make that exceedingly clear. However, we did test whether these same lipids reflected naffody in an independent mouse model. In this mouse model, mice were fed chow diet at age zero for seven weeks. And then they were switched to, this is high-fat, high-sucrose diet, which is a Western-type diet that induces naffody, induces fatty liver. And then these mice were treated with this NR, nicotinamide riboside which enhances myocardial function and decreases triglyceride levels in the liver. And so we wanted to see if this independent mouse model reflected the things that we saw in the BXD mice. We're excited to see that for a number of these, we were able to verify what we saw in the BXD cohort. So let's start with this initial triglyceride species. That is a pro-naffody signature, meaning it positively correlates with naffody. On a chow diet, you can see that its levels are low and these levels increase upon induction of a fatty liver. Unfortunately, for this specific triglyceride species, we didn't see a decrease in the species when treated with the NR, but we did see decreases in three of the other lipid species which suggests that these species are indicative of liver weight and liver health. And the same is true for these anti-naffody signatures. They are more abundant in the chow diet or the healthy mice and then decrease upon induction of naffody. However, it should be noted again that treatment here did not completely rescue the levels of these anti-naffody signatures. However, when we looked at naffody readouts again, body weight, fat mass, lean mass, liver mass, et cetera, for these lipid species that I outlined previously, we can see that for the pro-naffody lipids, they correlate positively with these naffody readouts and negatively with the concentration of NAD, which is the treatment of the nicotinamide riboside increases NAD levels and it's thought that this increase in NAD levels increases mitochondrial function and that causes the decrease in fat accumulation in the liver. Whereas the anti-naffody shown in blue, here in general correlate negatively with these traits. And finally, we wanted to see how this translated to human patients. So mice are similar to humans, but if the same trends are not seen in human patients, it's never going to be able to be used as a biomarker or a signature of naffody in the human population. And so we looked at naffody readouts in patients that were either healthy or had naffody. And although you can see that this is quite a bit muted and I apologize, I don't have the y-axis here, but it's the same as the previous one. The naffody, again, these pro-naffody species correlate positively with naffody readouts and negatively with the anti-naffody species, although to a much lesser extent. And so although I don't think that we have identified a biomolecule in any way, I do think we've made significant progress in even identifying the fact that it's possible that lipids in the plasma can reflect their abundance in the liver, and perhaps in the future, these can be used to diagnose various diseases without having to do biopsies. And so I'm just going to discuss my future directions here. So as I mentioned, verifying the candidate genes is very, very difficult, but it's important. It is what the end goal of QTL mapping is, and therefore we should spend more time trying to verify these genes. And although our lab particularly might not do this, I think that we've provided a great foundation and resource for future studies. And another thing that I think will be important moving forward is expanding lipid QTL or LQTL mapping to include more strains. For example, it's already being done on the collaborative cross, which instead of starting from two founder strains, comes from eight, which are incredibly genetically diverse and gives us more resolution and limits the space that we're looking for, candidate genes. And I think that's really important next step forward. I think looking at these lipids in different tissues, not just liver, is also important. And perhaps the plasma levels of different lipid species are reflective of other tissues and may indicate other diseases besides naffody. And I think that what we did here, we looked at a whole cell. So we just took a snapshot of the whole cell, but we know that different organelles are separated from each other by lipid membranes, and their lipids are regulated in different ways. They have different lipogenic needs. And so I think the next step is, instead of looking at whole tissue, whole cells, is looking at individual organelles and seeing if we can't unmask some of these novel candidate genes. And finally, we, as I mentioned, I identified approximately 100 lipids. Since I've been here, we can now identify thousands. And so that is an automatically. Originally I was doing by hand. I was identifying lipids by hand in the beginning. And this has opened up whole new doors. And I think that using all of these different lipids and trying to continue to map these traits and perhaps identify even novel lipids from these searches is extremely important moving forward. And with that, I would like to thank my lab. Sorry, I'm going to get emotional. I would especially like to thank Dave for standing by me throughout this process. As many of you know, as he said, it was not easy for me. And I couldn't have asked for a better mentor, a better scientist, or a better person to work under. I would like to thank all the Pagler Nila members, new and old. This is our current group. We've tripled in size since I started. Thank you for making every day worth coming into work. Thank you for being, pushing me scientifically, making me the best person that I can be. I am truly grateful for every single person that has come through the lab. It is impressive how well we all get along. And I say that's a statement to Dave and is choosing the people that join our lab. I would like to thank Adam, like I said, for homogenizing tissues with me for days. And also two former members, John Steffley and Brendan Floyd. John, who really pioneered the lipidomics in our lab. And Brendan, who was instrumental in a different project. I'd like to thank the Coon Lab, especially Professor Coon, for allowing me to be part of your group. For the few years that I was, I learned an incredible amount. And I look forward to applying that in my future endeavors. I'd like to thank a past member, Arnie, who with John really pushed the whole lipidomics methods forward. And I would like to thank Paul for all of his continued work and the automation of the identification of lipid species. It was truly life-saving. I'd like to thank my committee members, Carl Broman, Alan Addie, Mike Susman, in addition to Professor Coon. Thank you for listening to me all these years. Thank you for your advice. I very much appreciate it. And I would like to thank my collaborators, Johann Orx. His group is one of the best in mitochondrial research, the BXD cohort. And I was lucky enough to work with them. Arnie doesn't go here. But Pooja, who put up with me through the whole process, and Evan, who started it off. And I tried to put pictures in, but it got too hard. And so I would just like to thank my friends that I've made here. It's truly been a blessing. I want to take a special shout out to a former lab mate, Tim Rhodes, who during the past few weeks, as I've stressed out, has managed to remain calm and keep me calm in the process. It's been incredibly helpful. I would also like to especially thank my best friend, Tucker Croci. If you don't know him, you should. We're kind of like two peas in a pod. And there's no one I would rather grocery shop or pay bills with. And finally, I would like to thank my family. There's a number of them in the audience today. Thank you for supporting me always. No matter what. Believing in me when I didn't believe in myself. Nicole and David, thank you for joining this crazy family. Marty and Maggie, thank you for making me want to be a role model for you and pushing me to be the best I can be. And to my parents, Michelle and Dave, there are no words. Thank you. Thank you from the bottom of my heart. And with that, I will take any questions. It's more of like a histology, like looking at it, rather than specifically measuring, like, total triglycerides or something. It's like it looks fatty is more of the, with this one. Or I'm not sure I understand your question. They, I don't know exactly how they are, but they likely are in some way just because of how the whole network is so interconnected. But exactly what regulates the number of double bonds and triglycerides, I'm not sure. Yeah, like if all of these come from, yeah, it's very possible. So for the mouse model is in a B6 background. So in the B6 background, they do not. You have to induce it. Though some of our mice from the BXD cohort, though not specifically diagnosed with naffoldy, clearly have significantly larger livers, fattier livers than others. So it's, it's, the black six doesn't without some instigation. And I would have to go back and look at the data and see how often those fatty livers are associated with mice on the high fat diet. I presume that it's much more likely when they're on the high fat diet. We, we didn't do QTL mapping in the high fat, high sucrose mice, because they, they lack the genetic diversity that is required. They all have the same background. But that would be something in the future as an, you know, you can change the environmental components and that would be something that could be tried in the future. So that would be, those specific, I don't know about the specific lipid species. I, I'm not 100% sure, but it probably hasn't been really deeply looked into about the individual triglycerides that make up, for example, butter or any of the other processed foods that we eat. But it's possible that is where it comes from. Natalie. One of these species, I can't remember off the top of my head exactly, but we looked at relative abundance. And if I remember correctly, these species are middle to low. None of them are the, the most abundant species, so they weren't kind of taking over the total amount of triglyceride that we saw. Let's say these blue lips there, really something to do with these, with these complex phenotypes. I know you don't know the answer that nobody does, but like what, how might that work? Do you have something else? And your correlation is a causative thing or these things? What do you think? I, it would surprise me if it was actually causative. But I think there is something in the fact that the, the anti-nathalie signatures or the, the healthy lipid signatures tend to have an increased number of double bonds. And we know that if you have increased double bonds, you create more space. It's less densely packed. And so I think that might kind of help prevent some of the diseases that are caused by buildup of triglycerides if they can't quite pack as well. I don't know if I want to answer this. That was a great question. I don't know about, I would assume it had a higher caloric intake. I do know that it has unsurprisingly significantly more fat, but the exact percentages of them are written down somewhere, but I don't know them off the top of my head. Yes. Yes. Yes. So clearly the environmental part, the diet has a very clear effect, but it doesn't tell the whole story. And there is definitely a genetic component and for, even though most of the lipid species were affected by diet, there are still some genetic component there that remains to be discovered. Oh, God. Research. What does it feel like here in the lab? Can you hand it to me around? Yeah. I don't know. Yeah. I might have to say go bears on that one. I'm not sure. I think we'll end it there. Thank you again for agreeing.