024 - Exploring the Poop-ome, from the microbiome to metagenomics
Kat: Hello, and welcome to Genetics Unzipped - the Genetics Society podcast with me, Dr Kat Arney. In this episode we’re getting our hands dirty by delving into the poop-ome - the trillions of bacteria that live inside our guts and make up what’s known as the microbiome. Rather than simply being a bunch of bugs, the microbiome is now believed to play a role in virtually every aspect of health and disease. But what are they up to? How do we even know what species are in there? And can you blame your stinky farts on your gut bacteria?
Meet your microbiome
Kat: The microbiome is one of the hottest topics in scientific research right now. Our gut bacteria have been linked with everything from weight and mood to immune function and even response to cancer treatment.
Microbes populate our gut on the day we’re born, but exactly which ones we have depends on whether we pick them up from mum, the world around us to the foods we eat. But why is the microbiome so influential if it’s just bugs?
That’s the question I put to Tim Spector, professor of genetic epidemiology at King’s College London, author of The Diet Myth and scientific co-founder of the precision nutrition company ZOE, whose PREDICT research study is investigating how the microbiome and other factors affect personal responses to food.
Tim: It's just bugs but it's probably the most exciting discovery of the last decade because we have these hundreds of trillions of microbes inside us.
Most of them are in our large intestine and collectively, they form what is a virtual organ because they all work together.
They are essentially chemical factories that produce thousands of times more molecules than our own body can produce. They have several hundred times more genes in them to help make those chemicals.
It turns out these chemicals are incredibly important for keeping us healthy and on a stable footing in terms of our immune system, our mood, appetite and metabolism.
We're just scratching the service of all the things that these amazing chemicals produced by these bugs can actually do.
Kat: I'm used to the idea that for something like cows - or, I can accept that for humans as well - there's going to be things in our diets that our bodies, the enzymes in our stomachs and our intestines can't break down, so maybe we need bacteria to help with that.
So you're saying there's this whole other layer of molecules produced by these bacteria that could be really important - probably are really important and we don't really know?
Tim: Yes. So, digesting food is one part of -- the very simple part of what these bugs do. They can break down these complex carbohydrates that we otherwise can't do - your analogy about cows and other ruminants.
Once they've got this food inside them, they can convert that food into a whole range of other chemicals that they either use themselves or they pass onto the body or they allow us to then use in return, in a very symbiotic relationship.
Some of these are absolutely key to our wellbeing. Some of them are substances related to serotonin that is a major brain neurochemical that makes the difference between us being happy or sad.
Others are key components that keep our immune system healthy and stop us overreacting and getting autoimmune diseases or allergies. So we're just scratching the surface of these thousands of chemicals that we know very little about, trying to work our everything they do.
What's clear is that we can't afford to ignore them because in medicine, over half the drugs tested so far have a major interaction with these microbes, which explains why the same tablet isn't going to work for two people.
People will need different doses because depending on the state of your gut health, you're going to metabolise or break down that chemical very differently and it's going to have a very different effect. As with medicines, the same is also true of all our common foods.
It's requiring us to take a totally new look at everything that we put into our bodies because the way that's converted into other chemicals is proving to be highly complex, highly exciting and highly individual.
Kat: Let's dig into this a bit more. Tell me about some of the research that you've been doing to try and find out what is different in the microbes between people and the effects that it has on health.
Tim: What we did, starting about seven years ago, was to look within our twins cohort, Twins UK, which is the largest database of twins in this country - about 13,000 of them. We looked at two or three thousand twins who were both identical and non-identical.
For most of the traits we've looked at, we've generally seen a much stronger similarity between identical twins compared to non-identical twins, which tells you that genes play an important role. This is nearly every sort of thing we've looked at, whether it's fasting hormone levels or whether it's body size or intelligence or personality or major diseases.
Kat: That's good, this is a genetics podcast, good to know that genes are important for something!
Tim: Yes, I'm essentially a geneticist, so for 25 years I've been using the twin model to show sceptical people that actually, it's not just wear and tear and things like back pain and arthritis and depression have a strong genetic basis.
So I was well poised, really, to look at this question and say, well, how do the microbes fit into this vision of most things being highly genetic?
When we did this with the group at Cornell we got some surprising answers because overall, there's only a very weak genetic component to the microbiome. Most of it seems to be very individual, most of it is driven by the environment and our diet and is hard to predict.
There were some exceptions and there are some microbes that seem to particularly fit with certain people's genes.
We've explored in great detail things like Christensenella which is a microbe that if you have it, it protects you against getting overweight. That seems to have genetic basis, as did a few anti-inflammatory ones, but in general it was only very weak.
We followed this study up recently with the PREDICT study, which is the largest nutritional intervention study so far performed, where we took 1,000 mainly twins and gave them standardised meals and looked at various food responses and looked at the effect of the microbiome. When you looked at the metagenomes, we saw only a slight difference between the identical twins and the non-identical twins and unrelated people.
So, there was only a really tiny overall element of genetics making this up, which meant that most identical twins had really very different microbiomes. We believe this explains why even identical twins respond very differently to identical foods.
Kat: Obviously, identical twins and non-identical twins are born at the same time, one assumes mostly through the same route, whether that's a caesarean section or a vaginal birth.
Does that seem to have an impact on their microbiome? Because you might think that if they've come out the same way, they might get the same bacteria from mum, so it's strange that they're not more similar.
Tim: Yes, I think we were all surprised that even non-identical twins weren't much more similar than unrelateds. I guess it just shows that there's a huge luck element to what microbes you end up with.
Obviously, caesarean section is important and we know that microbes for the first period of life are different but it's hard to tell differences after the age of about five. So you have possibly these rather random changes in many kids early in life, that can have effects and the effects of medication and antibiotics etcetera.
So there seems to be a rather random element to people's early lives that we haven't really been able to explain yet. And also a flexibility in the early life - that you can quite rapidly change your microbiome from unhealthy to healthy or vice versa. So again, the microbiome continues to surprise us in ways we hadn't expected.
Kat: And for someone like me, as an adult thinking about the microbes in my tummy, what should I do to look after them, to improve them? What should I be looking for?
Tim: Well, so far, the one thing we do know is that the more diverse the number of species in your gut, the healthier you are going to be. This seems to be true for a whole range of diseases and disorders that we've looked at so far. So that's the number one thing that everyone should try and do.
The analogy is having a rich country garden full of different plants and herbs and flowers in there. If you've got that huge range, the soil is healthy. It means they're producing all kinds of chemicals that help each other and it stops other weeds coming in, other nasty diseases.
So, how do you do that? We did a large study comparing British guts and American guts of over 11,000 people. The one factor that increased the diversity of your gut microbes was the number of different plants you ate in a week.
Most of the food we eat are plants, we often don't realise it when it's in a packet, but it's the variety. So, if you can get to 30, that's the ideal number of different plants in a week, you can have the maximum diversity of your gut microbes.
It's not having the same kale smoothie every day, it's having a whole range of nuts and seeds and herbs and spices and different vegetables and fruits. If you can mix it up like that, that's how you get the greatest chance of getting a really diverse microbiome.
On top of that, you can help we think with fermented foods. Obviously people know about yogurt and cheese but the new ones, kefirs and kombuchas and kimchi are great as well, giving you extra probiotics and prebiotics combined.
Then looking for foods in your range of vegetables and fruits that are high in these particular chemicals that we now know are important called polyphenols.
Kat: Those are the colourful chemicals, aren't they?
Tim: They are generally very brightly coloured, the purples, the reds - and make fruit and vegetables particularly attractive. The stronger the colours, in general, you get more polyphenols. There's nine times more polyphenols for example, in Persian purple carrots, than there are in the orange type.
So, if you have these polyphenols, they are no good for you, but they are like rocket fuel for your gut microbes. Energise them and they can then make these amazing chemicals that get your immune system on track.
Kat: I'm going to go to the salad bar this lunch, I think.
Tim: Yes, and don't forget that as well as healthy salads, dark chocolate is also full of polyphenols.
Tim: As long as it's over about 70% and free of other sweetness and things, that's good. Olive oil, extra virgin olive oil is packed with polyphenols. Coffee, red wine, cider –
Kat: This is what I want to hear! This is the diet advice I need.
Tim: Yes, I think as long as you don't overdo it with one thing and you have a diverse range of these foods, everything in context. You can find a whole range of foods that are healthy for you, without having really too spartan a life.
Kat: Dark chocolate, red wine and coffee - that’s pretty much my perfect diet. Thank you to Tim Spector, for making me feel a bit better about my Cotes du Rhone consumption.
Microbiome plus genomics equals metagenomics
The kind of analysis that Tim and his team are doing to study the microbiome in twins involves collecting a stool sample - yup, you’ve gotta scoop the poop - extracting the DNA from all the bacteria and running it through a DNA sequencer, then sifting through all the data using clever bioinformatics to try and figure out what bacterial species were in there in the first place.
This approach is known as metagenomics, and researchers are now using it to study bacterial communities in all sorts of other places from every nook and cranny of the body to food and drink or the soil and the sea.
Deconvoluting all this genetic data isn’t an easy task, as I discovered when I spoke to Rob Finn, a team leader at the European Bioinformatics Institute just outside Cambridge.
Rob: So, when I first started out in the field of molecular biology and bioinformatics, the traditional way was you used to go and have to find one of these microbes and then you had to grow it in the lab.
But these things, these bacteria for example, are quite fastidious. They like to eat certain things and to get that all working in the lab can be quite hard. Sometimes they need their friends to come along with them that also help them to grow, so it can be quite difficult.
It's estimated that we've only actually isolated about 1% of all organisms, which means there's 99% out there still to discover.
Over the last few years, with sequencing getting much cheaper, what we've been able to do is actually go and sample directly. Therefore you might have a whole collection of different microbes from viruses, bacteria and small fungi, for example. Then with sequencing getting cheaper, you just sequence all the DNA you can collect.
Then the problem is; I've got all of these fragments, how do I get to a genome? People have likened this to a catastrophe in a jigsaw puzzle shop, where someone has just pulled all the pieces off the shelf and mixed them all up, taken some of those pieces away and the pictures.
Then what we have to use are computational algorithms, to try and guess which pieces should go back together again. So, you might start off with - going back to the analogy of jigsaw puzzles, you might start off with blue as being the sky. You use some sort of knowledge about what we've seen before.
We've been doing this more and more and the algorithms have got better and better, so now it is possible to just about get draft genomes, or a good idea of these genomes back out from these.
There are certain limitations, in the fact we can only recover the most common things. It's the bacteria, typically, that are more common in a sample, not the very few.
Kat: What sort of samples are we talking about? Where could you just grab a whole bunch of bacteria and throw them through your machines?
Rob: Microbes are very good at finding all different places to live. There are not very many environments where you don't find microbes. So in theory, any sample.
Traditionally, the common areas in which people have had interest are the human gut, ocean samples and soil. That's really where the field of metagenomics started but it has broadened out. People like to understand the metagenomes of cheese and wine for example, now.
Kat: I can see that, yes.
Rob: Yes, it's very important that we understand that. But there are other things, where people are going into, say, compost heaps and trying to find microbes or enzymes that could be important for generating biofuel. It's really gone from those three major biomes, to lots and lots of different environments.
Ultimately it comes down to, can you get enough sample of the microbes to then extract the DNA?
Kat: The one I've always wanted to do is go around a load of music venues and rehearsal studios and do the microphones because I bet there's some weird bugs in there.
Rob: Yes, there'll be all sorts. I mean whether they are growing and able to sustain… but I'm sure there's all sorts of horrible things. People have done built up environments and let's just say, microbes are all around us, all of the time.
Kat: And when it comes to taking the sample, you know, you take a cup full of water out of the ocean, extract all the DNA from it, put it through a DNA sequencer and you get all these fragments, As, Ts, Cs and Gs, these little strings.
How do you then go about putting that puzzle together? Can you explain some of the ways that you would start analysing it?
Rob: The way that we go through these sorts of puzzle problems is, some of it is just brute force. Essentially what you're looking for is little pieces that overlap. We've got lots and lots of fragments and then we start looking for pieces where maybe a part of the string is in common.
Then you can start building it up. You want a certain amount of depth of that, so usually when you pile all of those little fragments up, we're typically looking at about five times or more coverage before we can start piecing these together and reliably say, yes, we think this is a piece of DNA from a particular genome.
Kat: So, to go back to your jigsaw analogy it would be like, okay, we definitely want to find the same piece from five copies of this jigsaw before we know that's probably a real piece?
Rob: Yes, essentially that's what we want to do. What we've done in my research team is we've looked across many, many different samples. You get more confidence when if you treat each sample independently, then you recover the same genome over and over again. Which is always quite reassuring, that actually, these algorithms are not just producing something arbitrary.
Kat: And in terms of the number of species that you might find, say there's a cup of water on the desk here, the standard 200ml of water - what number of species of bacteria and viruses might we find in there if that was a cup of seawater, or even maybe Cambridgeshire tap water?
Rob: All biomes have a different level of complexity in terms of the different numbers of microbes. Even within an ocean environment it varies. For an ocean sample we roughly say it's in the thousands range, so 1,000 to 10,000. But that varies depending on how far you are from the coast and the depth that you are at sea.
Kat: What about in a soil? What if that was a cup full of soil, how many different species would we find?
Rob: In a soil it's even harder. The easiest microbiomes to work on is typically the human gut. That's about 400 species, typically, in a human gut sample. We can recover those fairly easily. Then water is the next step up. It's an order of magnitude.
Then soil is even worse, it's another order of magnitude. You're talking tens of thousands. Our ability to recover genomes, we can do that in the gut, we can just about do it in the marine samples, soil is still one of those areas where we need improvements of techniques to really access those microbes reliably.
Kat: Actually, thinking about that our gut is not the only place in the body where bacteria live. I'm thinking about your armpits, your bladder, your lungs. What do we know about those kinds of populations and can we take the same approach there?
Rob: So yes, the same approach can be applied there. I've also worked on the human scalp. It's actually a simpler microbiome. But bacteria everywhere, the bacteria you find say, in your scalp or in your armpit or in the vagina, they are all very different.
So there are different communities living on us in different areas and that's understandable. Your gut doesn't have a lot of oxygen in it whereas on your head, it's exposed to oxygen all the time, so the bacteria are adapted to that different environment.
Kat: Is there anything that you've found that has been just very strange, or a bacterium turning up where you're like, what are you doing there?
Rob: That's a good question. I wouldn't say there's so much of that. I think one of the things that we're always amazed about is that we see logical links - so when we did this big survey of gut microbes, we found some cases where there were things that are commonly found in the mouth. But that's not really that surprising, the fact that your oral microbiome is linked to your gut.
Kat: Yes, seems legit, it checks out.
Rob: Yes, it checks out. Because we're in this voyage of discovery, a lot of the bacteria we recover we've just never seen before, so to know it's an anomaly is actually quite hard to say.
Kat: It could be perfectly common it's just we never knew it was there?
Rob: Correct. That's the sort of thing we're thinking about.
Kat: Finally, what's the point of doing this? Once we know that all of these different populations of bacteria are here or there or in the sea or in the soil or in our guts, what can we do with that information?
Rob: Okay, I'll go from the immediate that's happening now, to the longer-term aspirations. We're already seeing people starting to mine these metagenomic datasets for novel enzymes, for examples.
There are lots and lots of applications. You can just think about some simple cases. So if you're after an enzyme that works at a low temperature, say for washing powder, if you go to an environment that is at a low temperature, you've got a good chance of finding an enzyme that works at a low temperature. It's fairly logical.
We're beginning to see the metagenomics datasets being tapped in for novel enzymes and biotech discovery. So, that's where we are now.
The next stage is starting to understand how we can gather more interesting things like novel antibiotics. That usually involves many different enzymes and cassettes. But if we can start discovering that or understand antibiotic resistance mechanisms, that's interesting.
So that's still from a single enzyme to a single genome. The longer term aspiration is; can we understand a microbiome sufficiently enough that if it's in a disease state or if there's an environmental impact - whether that's climate change or an oil spill - can we understand how to reverse the impact of that and bring it back to a so-called normal state?
Kat: This may be a bit of a strange question, but do you have a favourite bacteria?
Rob: I wouldn't say I want to be too prejudiced about my favourite bacteria. I like them all, I want to understand them all but there are certain groups that metagenomics is really giving us access to. There's a group called the candidate phyla radiation. They are quite new to us in the scientific world.
There's lots and lots of examples that have come from the metagenomics world that we have never seen before. They have some really interesting biology in the fact that they have lost a lot of the things that you would consider normal. Some of them have lost complete amino acid biosynthesis pathways - they can't make their own proteins themselves, in isolation.
Some of them are what they call epibionts. They live on a yeast and then they seem to just absorb the amino acids that are just given off by that yeast and so can then thrive. So understanding how those two organisms have co-evolved is very interesting.
Kat: Rob Finn from the European Bioinformatics Institute, giving us a glimpse into the mysterious microbial world.
Growing gut bugs
Once researchers have identified interesting species of bacteria with peculiar biological properties, the next step is to try and grow them in the lab to take a closer look. But growing gut bugs is a lot harder than you might expect, according to Hilary Browne, a researcher at the Wellcome Sanger Institute in Cambridge.
Hilary: So, it's quite easy to get hold of these samples. We're quite lucky in that way. What's challenging though is that your gut is anaerobic, so this means there is very little to no oxygen in there.
So the majority of the bacteria in this environment cannot survive in the presence of oxygen. Some of them will die within seconds of being exposed to oxygen.
This is a challenging part of our work. We have to work with these samples in anaerobic conditions, so we have these special hoods where we can stick our arms in and the inside of them is completely anaerobic, it's pumped with anaerobic gas. This is how we work with these bacteria.
Kat: It seems like the weirdest job in the world. You're getting samples of poo from people, you're trying to work with them in an environment with no oxygen at all, and then what are you doing with them? Presumably, there's a lot of different bacteria in there?
Hilary: Yes. What we've been trying to do is to grow as many of them as possible - and this is really basic microbiology - so if you can grow the bacteria then you can study them in the lab and you can learn more about their biology.
The field of microbiome research has exploded in the past three years because of DNA sequencing technologies. This is because sequencing has become much cheaper and we've also become much better at it.
By doing this approach you get an idea of who is there in the community but if you can work with the bacteria and you can work with them in the lab, you can actually understand what they are doing.
Kat: It seems a bit strange to me, as someone who has worked in a lab, you get some bacteria, E.coli is the most common one, that's originally a gut bacteria, you stick them in some broth or you smear them on a plate and bugs just grow.
Why is it so hard for these gut bacteria to grow? What makes it so challenging to try and grow them in the lab? You don't just smear some poo on a plate and hope for the best?
Hilary: Well actually, E.coli is not very common in the gut at all. It is there, but it's at very low numbers. E.coli can also grow in the presence of oxygen, so again this makes it very easy to work with.
The reason the majority of the gut bacteria differ from E.coli is that some of them have very strict nutritional requirements. Some of the carbon sources that are in media that we commonly use, these gut bacteria do not like them. That's why we have to use different media and different culturing techniques to try and isolate them.
Kat: Take me through the process. You get a sample of poo from someone. First of all, who from? Is it you, is it your lab mates? Where do you get the poop from?
Hilary: We would have studies with collaborators in hospitals. So for example, we have a study looking at patients with IBS. In these scenarios, the patients would produce the samples, stick it in their freezer, then we send a courier around to pick up the samples on dry ice.
The reason we do this is because we want to keep the bacteria alive and to get them here as soon as possible. A good way to do that is to freeze them. If the patient freezes them in their freezer and then the courier brings them here on dry ice, then we can stick them into the freezer here and this preserves the bacteria.
Kat: It's like, woohoo, Christmas! Oh…
Hilary: Yes. And obviously the patient has to be very careful when they go reaching for stuff in the freezer, as well.
Kat: Oh my god. Yes, I can imagine that. So what's the next step in trying to get these bacteria out of this sample? You've got the poop, you've got your wonderful hood with no oxygen in it, what happens next?
Hilary: So, first we have our frozen poo, then we defrost it in the cabinet. Then what we do is we dilute it maybe a million times. If you just plated a piece of poo on an agar plate that had nutrients for bacteria, you would be amazed at the amount of growth that would be there. This really gives you an indication of how much bacteria there are in your gut.
What we have to do is we have to dilute that faecal sample down thousands of times, until we can see distinct bacterial colonies. So when these are growing on the plate, then we can pick them.
Then what we do is we sequence a part of the genome to identify them, to give them a name. Then we would freeze that particular isolate in a freezer at minus 80 degrees. Now we have a frozen isolate in the freezer that we can do experiments within the lab and we would also sequence its genome. This is very informative as well.
Kat: Tell me about some of the bacteria that you've found. Have there been any that have been completely new, surprising? Any that are like, woah, what are you doing in there?
Hilary: Well, we were amazed at the amount of novel bacteria that we actually recovered. The more culturing we do, the more novel bacteria we find. We're not just finding novel strains or novel species; we're finding bacteria from novel families as well - this is kind of a higher taxonomic level.
There's a huge amount of diversity in the gut and we're still really scratching at the surface in terms of capturing the diversity of it all. It's hard to pinpoint one specific example really, because there's so much there that we still have to discover.
Kat: And presumably there are still, in your gut and in my gut, bacteria that are currently unknown to science?
Hilary: Yes, loads, yes. And if you consider us here in the Western World, eating a particular diet, we will have a particular type of microbiome. This is where much of the research is focused to date in Europe and the US, but if you looked into different populations then you're going to find a completely different set of bacteria as well, that haven't been discovered.
Kat: Who knows what else is out there? Hilary Browne from the Wellcome Sanger Institute.
Can you blame your microbiome for stinky farts?
And of course, I couldn’t make a programme about gut bacteria without tackling the one question that my inner ten-year-old has been dying to ask - what’s the connection between the microbiome and farts? Tim Spector humoured me with the answer.
Tim: A very good question. We don’t really know, is the answer. There's very little research goes on about farts. We do know that the average person does about 14 farts a day.
Kat: Good to know.
Tim: Some admit it, some are proud of it and others are not. So it is quite normal to expel wind. Some foods, when people are changing onto certain foods or changing fibre from a junk food diet to a healthy diet, they might experience more wind as their guts are fermenting some of these products.
Some people do have some medical problems with this and can end up with irritable bowel type symptoms from over-fermenting. It could be due to the length of a bowel, which differs about three-fold between people. It could be due to the speed of eating and it could be just due to certain resident microbes that produce their own methane and live off methane.
Kat: You’ve got the farty bugs, basically?
Tim: Yes. So next time you have an embarrassing moment, rather than just saying sorry, say, "Please excuse my microbes".
Kat: Great, blame the microbes! It's like blaming the dog, isn't it?
Tim: Yes, they are there for all purposes, to say they are great and also to blame all our troubles on. The nice thing is, we can change them. Next time you eat anything, realise that with a hundred trillion microbes inside you, you are never going to eat alone again.
Kat: Not so much feeding the five thousand as feeding the five trillion. That’s Tim Spector from King’s College London.
And finally, a couple of episodes ago we looked at the evolutionary origins of sex. But what about the evolution of asexual reproduction? While doing-it-yourself is common in invertebrate species, there are around 90 species of vertebrate that have evolved to reproduce asexually, including Darevskia lizards, where mums make babies without any input from males.
In this clip from the latest podcast from Heredity, the Genetics Society’s journal, James Burgon talks to Susana Freitas from the University of Lausanne in Switzerland about the origins of these unusual animals.
James: So yes, maybe you could just tell us a bit about this group of lizards and why they were interesting for your study?
Susana: So, historically they are actually important because they were the first asexual vertebrates to be found. In the 1960s there were some researchers that were studying these populations of these lizards. In a specific population, they only found females.
Every time they went back there, they only found the females, the never found males. Those were the first observations of vertebrate asexuality.
It's an amazing group. Like I told you in the beginning, there's about 90 species in total of vertebrate asexuals, but in this group alone there are seven.
Kat: You can hear the full interview in the latest Heredity podcast - just search for Heredity in your favourite podcast app, or follow this link.
That’s all for now. For more information about this podcast including show notes, transcripts, links, references and everything else head over to geneticsunzipped.com You can find us on Twitter @geneticsunzip and please do take a moment to rate and review us on Apple podcasts - it really makes a difference and helps more people discover the show.
Genetics Unzipped is presented by me, Kat Arney, and produced by First Create the Media for the Genetics Society - one of the oldest learned societies in the world dedicated to supporting and promoting the research, teaching and application of genetics. You can find out more and apply to join at genetics.org.uk
Our theme music was composed by Dan Pollard, and the logo was designed by James Mayall, transcription is by Viv Andrews and production was by Hannah Varrall. Thanks for listening, and until next time, goodbye.
References and further reading:
The role of hybridisation in the origin and evolutionary persistence of vertebrate parthenogens: a case study of Darevskia lizards https://www.nature.com/articles/s41437-019-0256-5
Predicting Personal Metabolic Responses to Food Using Multi-omics Machine Learning in over 1000 Twins and Singletons from the UK and US: The PREDICT I Study (OR31-01-19). Sarah Berry et al. Current Developments in Nutrition, Volume 3, Issue Supplement_1, June 2019, nzz037.OR31-01-19, https://doi.org/10.1093/cdn/nzz037.OR31-01-19