 So, thank you very much to the organisers of Galaxy and the Australian Biocomments in Galaxy Australia for inviting me to come and speak with you today. So, it's a oily, it's another small brown hoppy thing. So, the first thing you need to know if you come, don't come from Australia as most of our wildlife is, can be classified as a small brown hoppy thing. And here is one. This is a Indigenous painting of a bilby or a bilba in the Yellow Way language, and I'd like to take a moment to acknowledge the traditional owners of the lands in which I work, of which there are many, many nations of our work across Australia, and pay my respects to their elders past and present, and acknowledge their ongoing commitment to care of country. I love this photograph of anyone who's ever seen me speak before. This is the first photograph of the Earth from space taken by Bill Anders on the 24th of December 1968. And to me, it speaks of fragility and of hope and of technological marvel. Since we went and took this photograph, we've developed smartphones, genomic sequencing, artificial intelligence, machine learning, and the list goes on. It's amazing the technological advancements we've made in the last 50 odd years. In that same period of time, 950 species that we know of have already gone extinct on the planet. So unless you've been living under a rock, you know that we now find ourselves in the largest six-math extinction event, which is actually being driven by antrifugian impressions. And so what does that actually mean when we start to say we're losing biodiversity and what is the impact going to be on us as a species in the longer term? There's now over one million species on the threat of extinction on the planet, and there's only an estimated 13 million species, of which humanity is only one. And why is biodiversity important? Even if you don't care about nature and you don't care about animals, you do care about the food that you eat and you do care about the medicines that you take. And our health and our human health is very much linked to what is happening with biodiversity. Biodiversity contains a solution for climate change. It's also related to our food security, our human health and also its cultural and economic value to the many different countries of the world. Now, I don't know where you come from, but I hope that you played Gengar as a child that game where you stack up all the wooden blocks and you pull out one block at a time and the tower keeps standing. So people often say to me, what does it really matter if we lose a species? You know, who cares if we lose a butterfly or who cares if we lose the street? It doesn't really matter when you lose one species, but as you know, when you play Gengar, you start to pull out enough blocks, the tower starts to wobble. So I'm here to tell you the tower is now wobbling and humanity sits at the top of the tower. We have the hardest on the firm's place to fall. So when I talk about biodiversity, what actually is biodiversity? So biodiversity is all living things on this planet. And it's made up of three pillars. It's made up of a diversity of ecosystems. So the more variation we have in on the planet, the different ecosystems we have, the better. It's made up of a diversity of species. The more variation of species we have, we all know if you did year five science that monoculture is bad for productivity in agriculture, it's the same for the planet. If we have fewer species on the planet, we have less productivity. And of course, the third pillar is genetic diversity or diversity of genes. So for the first hundred years, ecologists have been really good at monitoring what's happening at an ecosystem level and at a species level. But the divegenic diversity pillar was always one of the hardest for us to crack and to understand in the biodiversity space. That was until the age of genomics. And what I like to find interesting is that when we started to put man into space in 1960s, we managed to sequence 76 base pairs of the yeast genome. We were super proud of ourselves. The genomic revolution however didn't really change until 2003 when we published the human genome. The publication of the human genome changed the face of genome sequencing forever. And along with that, it also changed the face of human medicine. You think about what human medicine was like in the 1980s and the 1990s and what it is today. We have precision medicine. We have greater understanding of genetic related diseases. We have a greater understanding of cancers. And that all came from us developing the human genome. In 2018, there was a bunch of us that came up with a wild and crazy idea of sequencing all eukaryotic life on Earth. Believe it or not, to sequence all eukaryotic life on Earth over a 10-year period will cost the same as sequencing the human genome in the late 1990s and the early 2000s. So when we say study a genome, I asked Gary who the audience was and he told me that you were mainly in the bioinformatics space so I thought I'd take us back to a little bit of biology. We don't actually use liver to sequence a genome. This comes from a human website. Liver is really bad in wildlife species. It tends to the enzymes and it tends to cause lots of problems with the sequencing technology. But essentially what we do is we extract a tissue DNA from either a tissue sample or blood or from fresh leaves if it's a plant. And we basically amplify that DNA and then we go ahead and sequence it. And then we end up with lots and lots of puzzle pieces. So all the pieces of a jigsaw puzzle. And when I first started working in this space and started working with Biocommons, I would say I need more. And the answer was how much more? And I'm like, I don't know how much more I need. I just know I need more. So what's the first thing you do when you make a jigsaw puzzle? You tip it out on the dining room table, don't you? And you spread it out. The bigger the table you've got, the much easier it is to find the pieces of the jigsaw puzzle. And I have learned that the bigger the compute I have, the much easier it is to spread out my pieces of my jigsaw puzzle. And why am I so interested in jigsaw puzzles? Except for the fact that my children keep buying new ones, which are really complicated to do at the moment, is because we're trying to make the puzzle box lid. So when we talk about a reference genome, a reference genome is essentially the puzzle box lid for the species that we're working on. And the reason for that is, is we don't sequence whole genomes to wildlife or other species, particularly if they threaten. We only do what we call reduced representation sequencing. So we only end up with maybe 5,000 pieces of a three billion piece jigsaw puzzle. And we need to know what each puzzle piece means and where it actually goes. So the use of genomics in conservation is not new. We've been using conservation genetics for a long period of time. And there's been a large number of reviews in about the last five years about why we should be using genomics in conservation. And this really interested me because I'm actually not a geneticist. I'm a conservation biologist. But I came to realise about 10, about 15 years ago now, I needed to use genomics and genetics for some of my threatened species management. And that's because genetics is an exceedingly powerful tool that we can answer a suite of questions with. And that's why it's become so critical in the latest global biodiversity framework that was announced last year in December in Montreal. The goal, one of the key goals for the 2050 framework is that genetic diversity within populations of both wild and domestic species is maintained and safeguarded for their adaptive potential for a changing climate. And that's really critical for us because it's the first time that genetic diversity has really been front and centre in the Convention for Biological Diversity. And it's not that managers don't know that genetic diversity is important. Anyone who's done a biology degree knows that genetic diversity and inbreeding are inversely related. And that you need genetic diversity to maintain adaptive potential for the future changing climates that are coming ahead. And it's also how individuals maintain their fitness for survival. So it seems simple, right? We have a tool. We know it's important. So why aren't we using it? Why don't we use genomes in conservation? I asked this question back in 2018. For my many Sims in Australia, I sit on a suite of recovery teams. And it's always been traditionally considered as a nice to hack. So on a Sunday afternoon, obviously I had nothing better to do. I read 200 national vertebrate recovery plans here in Australia. And interesting enough, of those 200 plans, more than 80% of them had genetics listed as an action for the recovery of those species. Fantastic. We have the tools. We have the desire. We have the policy. Less than 15% are actually using genetic data. And my question was why? Why can't we actually get this to be used as an everyday tool in common management practice? We're sequencing all life on Earth. There's a range of different initiatives around the world that are sequencing life on Earth. So the resources are coming. And there's an increase in genomic resources. There's been a massive sharp increase in genetic data and reference genomes being published. However, it's an unequal increase in resources. It's much easier to go sample a common species. It is very, very hard to be able to get some high-quality tissue samples from threatened or endangered species. Of the 15,500 vertebrate species that are listed as threatened on this planet, less than 3% of them have genomic resources. And that became, never to the fore, was the greatest recognition for us in this area was when the bushfires hit. So in 2019, 2020, we had the megafires here in Australia. These fires burned hotter and faster and higher than any other fires on the planet at the time. Unfortunately, this great 12-meter height of our 2019 bushfires has now been superseded by the fires that have just occurred in California. The planet is burning and it will continue to burn. We have to come up with better solutions for how we're going to manage post-fire events. We lost about a billion animals in Australia. A billion individual animals died in those fires and we had 23 million acres burned. And then why is that so important really for Australia? Is that we have about 8,000 vertebrate species here and over 24,000 plants, but most of our species are endemic. Australia separated from the rest of the continent 53 million years ago. So most of our species are found here and nowhere else in the world. And when they did a study just after the fires to see what the impact was, all these species you can see on the screen here were already listed as vulnerable or threatened prior to the fires arriving and most of them lost almost 100% of their habitat. So everywhere they lived is now gone. And so how is the solution to that when you have this kind of catastrophic firestorm go through the natural environment? We can't simply sit back and wait for it to regenerate because the species that live there are already under pressure. So the solution was to go out in front of the firefronts and grab what we could. We were going out and grabbing frogs and other species after the fire scars had gone through. We had helicopters flying over, dropping food, carrots donated by a local supermarket out of the helicopters to provide some food source for the animals that had actually managed to survive the firefront. So we ran captive breeding programs. We have a massive problem in Australia with feral species eating our wildlife. So we tend to put them behind fences. We also go out and catch and clean the waterways. The ash has caused a significant runoff when the rains came or the ash went into the river systems and killed all the fish species. And then of course we have to start regenerating the plants. Every single one of those activities requires genetic data and easy, fast, rapid access to genetic data to be able to make smarter and more informed decisions when you start actually having to actively manage the landscape in such a way. So in the old days, as a conservation biologist, I'd use micro-satellites. There's no hard-tested sequencing, count my alleles, and off we go. Unfortunately, with the human genome project and the access to genome sequencing technology, genome biology has become this really specialist field, which now has created this massive gap between those of us who actually need access to the information and need to use it and those who are generating it. And everyone's been calling this the research implementation gap for quite a period of time, and I would argue that this is actually just a space because the space is flexible. It moves, sometimes you end up with people who know what they're doing most of the time we end up with this larger space between us. But there's now become a third space and this is where most of you guys sit and this is the cloud bioinformatic space. And as a conservation biologist, it's taken me about 10 years to get my head around the genome space. Please don't ask me to get my head around this space. I just literally do not have the mental capacity to figure out how to become a biome mathematician. And so this is now where you find yourself as a conservation manager, is you're just sitting there confused in the middle of it all, having no idea what you're supposed to do, how you're supposed to access and how you're supposed to use that information. And this is how the Threatened Species Initiative was formed. So this is the program that Nigel was talking about that I set up in 2018, we officially launched in May 2020, who decides to launch a massive program in the middle of a COVID pandemic. That would be me, but anyway. And what is the Threatened Species Initiative? It was a project developed by myself, Peter Latch of the Australian government, Kim Ottawell at the West Australian government and with support from Bioplatforms Australia, which is a increased facility here in Australia that supports omics infrastructure. And basically it's designed to set up genomic resources for Australia's threatened species, but we don't only just do a reference genome, we also do associated population genetic data, but more importantly, we're building an applied conservation genomics portal. So this is an online toolkit to improve genetic literacy and bioinformatic literacy in our conservation management community. And what does it actually mean to have a genome? So I've been famously known for saying around the world, a reference genome does nothing with conservation. You can make all the reference genomes in the world, if you don't annotate them, they're next to useless. It's like putting letters in a book on a shelf when you annotate it, that makes the words so people can start to read it. But a reference genome is the foundation on which your conservation house is built. It gives you a blueprint for what you're working with and it lets you create excessively massive, a huge amount of tools, downstream applications once you actually have the genome itself. And we can look at neutral diversity or the bits between the genes, which allows us to assess the differences between populations, population diversity, structure, parentage analysis, taxonomy. Believe it or not, a high proportion of reptiles and amphibians in Australia are actually taxonomically incorrect. So I'm not a taxonomist. I leave that to somebody else to deal with, but it is quite a life problem for us. It also allows us to look at functional diversity. So our research group specializes in immune gene diversity and disease resistance or recovery, but you can look at reproductive genes and fertility and heat tolerance and climate change. And the list goes on at the relationships you can look at between functional diversity and phenotype that's occurring in the environment. And more importantly, for cryptic species, we can actually use a reference genome to create species-specific DNA markers. So you can go collect a soil sample and air sample some water and actually detect whether your species is there. And also it allows us to develop some specific gene markers so we can run hundreds or even thousands of samples at a much cheaper rate than we currently are. So it really is an exceedingly powerful tool. And so when we look at TSI, at the moment we're doing about 87 species across 53 projects. I'm seeking funding to take that up to about 300 species in the next two years. Sorry, Steve, but there's gonna be a lot more data coming your way. But more importantly is this pie chart. More than half of the collaborators involved in the program actually are research scientists who work for conservation agencies or work for conservation NGOs and only about a third of them are academics. And so one of the key things we're trying to do in the program is create partnerships between the academic community and the conservation community to make sure we can get information going out of the academic sector into the hands of the people who need it in real time. The large proportion of plants is representative of the fact that of the 2000 threatened species we have in Australia more than 1400 of them are in fact plant species. And of course we've got a suite of critically endangered, endangered and vulnerable species. The gray DD is data deficient. These are species that we don't actually know whether they threatened or not because we don't even know anything about them. And the black ones are extinct in the wild. So we've actually de-sequenced the reference genomes for Australia's two extinct in the wild species which is a Christmas Island blue-tailed skink and the list is gecko. And so now we've got those working very well and working with the captive breeding program. And one of the greatest things we also try and do is where we can is work with the indigenous and local nations in Australia recognizing that some of our species like the bilby that I work on actually covers about 200 nations and has about 47, sorry, 74 different names, indigenous names just for that one species alone. So this is a bit of an example of some of the critters that we work on. Everything from crayfish and invertebrates through fish, reptiles, birds, mammals and plants. We don't do a lot of marsupials and the reason for that is because the Os mammal genome project actually occurred three years before we started TSI and they generated most of the marsupial genomes already in Australia. So this is me back in 2011 scratching my head and I thought I'd give you a quick synopsis of what we've actually built here for biodiversity genomics in Australia. So what is genetics and how do I use conservation action? That was my first question when I started running the devil program back in 2010. You know, what Australian species even have genetic data? How do I generate my genetic data? How do I analyze my genetic data? And then how do I actually take all that data that I produce and apply it to conservation decision making? So what we've done is we've developed we, this is in the final stage of developers hopefully the release at the end of this year is updating the TSI website which will have access to a community portal which will take you on to downstream analysis a suite of different tutorials and protocols and we'll be soft launch at the end of this year as an online genetic training course. Now these are 10 to 15 minute YouTube clips you can literally sit down we have any cup of tea if you're a conservation manager and you can watch some YouTube clips created by some of the best conservation geneticists Australia has to offer and it's nine modules there's actually 42 different recordings and we're building this at the moment to try and increase genetic literacy. How do you get things sequenced? How do you collect samples? How do you do the analysis? It's all gonna be in a one stop shop. So then the next question is what Australian species have genetic data? So when I made that little pie chart I said less than 3% of species that are threatened actually have genetic data. I had to go and manually search the world's data repository so that's NCBI ensemble and try and figure out where we find what species have genetic data. And this led us to develop ARGA or the Australian Reference Genome Atlas and this is a collaboration between the Australian Biocommons, the Atlas Living Australia Bioplatforms and it's funded by the ARDC which is the Australian Research Data Commons. And the idea is is you can go in type in your species name and it goes and uses API technology and tells you where all your data sits. This is publicly accessible data obviously. And then you can actually just go and click a button and send it into Galaxy for your analysis. So we're trying to make it idiot-proof for people like me. Biologists go in. I don't like anything that's a list. Point and click is my favorite option. And the idea is that I can actually say well, give me a list of all the species that are critically endangered in Australia that have genetic data. How easy will it be when we have bushfires and we need to know where information is for species in the area that we're looking at to simply type it in and know exactly where we can go pull that data from. So now whether you can go and check whether or not you've got the data for your species. If you don't have data for your species, how do you go and generate it? Well, this is the Australian Bioplatforms. As I said, they've got a number of different facilities that they support around Australia that does sequencing both Oxford Nanopore and Pac-Bio all the way down through Illumina and all the different types of both RNA and DNA sequencing available. And back in 2015 with the Koala Genome Project was their first Biodeversity Initiative. And these are a list of the different Biodeversity Initiative that they now fund across Australia and that helps support both reference genomes, phylogenomics and taxonomy. And the Threatened Species Initiative is really their conservation initiative which is about just generating resources to threaten species for conservation purposes. So now we have an ability to generate the data. How do I analyse? And I never actually understood how complex it was when I first asked the question. If you surely, like, I've got my sequence, can I just put it in a computer and hit run? Yeah. Oh, well, I've been well trained by the Biocommons team here in Australia. So we have started a journey, I think it was probably in about 2020 when we started to really start to pull all of this together. So we now have the capacity to do genome assembly on the NCI, which is the National Supercomputer. The HiFi ASM workflow works really, really well in Galaxy Australia. And then we also use AWS for some of our bigger genomes like the frogs. We need a terabyte of RAM generally to assemble a frog genome. One of the other things, though, as I said, if you're not annotating your genome, it's not very valuable to us. And it's hard to use annotation programs that are built off protein sequences or transcriptomic sequences that only come from the Ethereum mammals from the Northern Hemisphere. When you live in Australia, they don't work. Well, they do work. We find lots of gene orthology but there's lots of gaps where we can see open reading frames and actually see the genes and we have no idea what they are. So we're really encouraging the Australian community to collect transcriptomic data along with their genomic data to be able to do reference guided genome annotations. And so at the moment, the transcriptome assembly pipeline is working really well on Causie and we use it on AWS. We've still got some tweaking to do with Galaxy. We've got a couple of dependency issues but we will get there in the end. And then we have our conservation genetics portal which I'll talk about in a minute with AWS and Ronan. And then we're supported by AWS in the open dataset program and we can actually got the Australasian genome site where we can load up our publicly available genomes. And so this actually turns out to be quite a lot of different places you have to log in to go and get your data, take it here to assemble it to go over there to do your transcriptome assembly. And thanks to the fantastic team at the Australian Access Federation and Steve Manos particularly from By Commons when I met with him a year and a half ago and said, come on, we've got all the pieces. Can't you just pull them all together in one place? Again, showing my biological ignorance into how hard it is to get the back ends of all this technology to talk to each other. But we now have a one place login that'll be launched at the end of this year which is a community portal you log in once and you'll have access to all the compute, the data storage and all the learning modules all in one place. So Kate Fierquiston who's a post-op with us as a biometrician has worked very closely with the By Commons and Galaxy team. So you can now go into the data portal with bioplatforms where the data is stored. And we used to have to copy the URL into Galaxy but I believe there's now a button that you can go send your data to Galaxy. Even better, Carolyn's idea of point and click. And you send it into the Galaxy workflow and it simply runs. The FastQC does a high file ASM assembly pipeline and it does a mercury downstream QC analysis as well. It's working so well. I gave it to my, I'm gonna give it to my 16 year old daughter to run. See if she'll assemble a genome but also we've got ecologists who don't know anything about genome assembly and nothing about bioinformatics who can now go and assemble their own genome. So we're getting ever closer to the place where we can start to generate data quite rapidly. So the last of course is once you've got all this is how to apply my genetic data. We've been building a conservation genetics portal. And again, this is something else that sounds very simple to do except sequencing companies to keep changing the read length of their read data. And when they do that, it starts to cause huge problems with your barcode. So what we, the plan with the genetics portal is as an end user, I can go collect my sample, get it sequence extracted and sequenced by a commercial provider. I receive the data. I can then come to our conservation analysis portal. It calls your variance against any reference genome that's publicly available. And then we'll run your VCF file through a series of downstream, globally recognized genetic metrics for conservation management. So heterozygosity in greeting FST values. And you end up, you get the data in a standardized report and this has actually been the feedback from the conservation community is that every time they get a report from a conservation geneticist, it comes in a different format. So it makes it very, very difficult for them to start to understand what the data is that they're looking at. And so what we've done is we've standardized the reporting format. And then as an end user, you can go and make your management decisions. So you can go and choose your decisions for translocation or captive breeding or habitat protection. And there's a large number of biological science students who are finishing their degrees who've done genetics or don't massive genetic projects during the undergraduate degree who then go work for government agencies and for conservation agencies. And this is actually the portal for them. It's a place for them to have easy access to be able to analyze their data and not actually have to have a university affiliation. This process already works. That's me in the field in January collecting the samples from a devil. And that's us translocating them six months later in the June. So we're down to about four months now. And actually two months of that is the sequencing time. So 2024, we've put all the sequencing companies on notice. We need to go and have a conversation with them about how we can start to shorten that sequencing time. So this is the car that we're currently driving for biodiversity genomics in Australia. It looks very clunky. That's because the time is, I liken it to a 1995 Ford Falcon. A little bit rusty, rattles a lot. Needs new shocks, probably some new tyres and definitely needs new brakes. But it drives. It's going down the road and it works. And so now, working with the Biocommons team and the rest of the people involved in the project, we're going to start polishing the car and making the rattles a little bit less and a little bit easier to drive. So for the second half of my talk, I kind of wanted to go through, like if genomics is a study of DNA and genes, how does actually studying in animals genome save them? So I thought I would walk you through what we do in our group. This is us, the Australian Wildlife Genomics Group. And yes, I'm very aware we have a large proportion of females in our group. I keep trying to recruit males, but I don't seem to be able to get them. A large proportion of women actually do molecular biology in Australia. And I co-lead the group here with Professor Kathy Bellop at the University of Sydney. And we've been responsible for some of the largest genome projects that have ever come out of Australia for Australian species. And we used to predominantly work on marsupials. We still work very heavily in marsupials, but we've been moving into some of the birds and amphibians lately as well. This is my favourite animal, much better than a koala. I work with koalas as well. The koalas bite and they claw and they smell and devils are just lovely. So this, for anyone who doesn't come from Australia, is a Tasmanian devil. And they are in fact, the world's largest mycicle carnivore. They got this name after we managed to hunt the Tasmanian tiger to extinction. And unfortunately, the last Tasmanian tiger died in the Hobart Zoo in 1936. And they're not very big. So they're about the size of a small dog. If anyone has copperspaniel or one of those brown, small, furry, yucky things, that's pretty much the size of a devil. So males are about 10 kilos. Females are about 8 kilos. And like all good marsupials, they give birth to underdeveloped young. They're only pregnant for a period of 18 days. And interestingly enough, Tasmanian devils actually give birth to about 30 embryos. But because they've only got four teats in their pouch, they can only ever have a maximum of four offspring. And this photo down here used to make sense when we used money, but nobody knows what a 20 cent piece is anymore. Yeah, but as you can see, they're highly underdeveloped and they attach to their mother's teat and they develop in the pouch. And then around about July, August, they start to get dropped in the den. Mum goes back and forth to feed them. And then they start to disperse in December of February every year. And they live for about five or six years when they're in the wild in the old days. And I'll talk about what that means shortly. So why are they called a Tasmanian devil? Probably the people in line from here that have grown up here. This is my colleague, Phil, and he's just trying to scan to see whether or not the devil in the bag has a microchip. And normally they're not like this, but this is a deep nose. He's very upset with us because he was asleep in the trap and we rudely woke him up and we don't come through the bag. As soon as Phil puts his hand on his head and stops, it's amazing. They'll be absolutely losing it and you just hold them in your lap and they just, oh, okay, I feel safe. So the reason they're called Tasmanian devils is because that actually is the sound they make underneath your house in Tasmania. They like to set up dens under your house and when the Europeans first arrived, they thought the devil was screaming underneath their house in the middle of the night and that's literally how they got their name. And people from overseas who used to see the Looney Tunes character called Taz just thought it was an imaginary sound. They had Tasmaic. It's actually a real sound of a Tasmanian devil. Oops, there we go. So we dump them in bags and then we weigh them and then we also check inside their mouths. I'll show you a bit more of that in a minute. We also take a little bit of a bit of a biopsy from the ear for genetic sampling and then of course we also check their pouch to see whether or not they're reproducing. And so people are always fascinated when we work with devils because they're actually really calm. So this is an awake devil. Devils have the second largest bite force of any land animal. If you basically force their jaw open by blowing in their face. We use a very technical piece of scientific equipment to check under their tongue. It's called stick. And then you just let them go once you've looked in their mouth and they're just quite happily sitting in your lap. We can also bleed them from the jugular vein when they're awake. You can climb your lap, hold their head up and you can take blood samples from them. And this is our little friend who is screaming his head off. 10 minutes. It only takes about 10 minutes to process a devil. And then you try and get them out of the bag and then this happens. They don't want to leave. Still doesn't want to leave. But they're kind of cute. His name's Yosemite Sam because that was a national park here. We were naming devils off the national park and after screaming so loudly I think he sounded like the Yosemite Sam character. They also move pretty quick. It's not very sexy sometimes working in the field. I've cleaned, I calculated actually the other day I cleaned over 3,000 devil traps in the last 10 years of working on the devil program. And we also get to work in really exciting conditions like shipping containers when it's only one degree. Doing analysis on my PhD students from a few years ago Rowena Chong doing some fickle accounts. And then sometimes we get to do really exciting things like setting up a lab that's sterilized on a piece of aluminium foil while you're putting up tubes on the table in the Airbnb. They did know I was doing that. I did ask. But Tasmanian devils actually went extinct in the Australian mainland about 3,000 years ago and they now only restricted to the little island state of Tasmania which is always that little blip at the bottom of Australia if you're not from here. And they've been through a number of unexplained population crashes over the last 20,000 years and that's really important to know because that's actually left them with very low inherent genetic diversity. And they're now actually classified as endangered. So I hope no one had bacon for breakfast this morning. Sorry about the gruesome photos but this is why they're endangered. This is devil facial tumour disease. It's an infectious, clonal cancer which is wiping out the population. It starts off as a little pimple on the bottom of a devil's jaw or around on the cheekbone and then it simply grows. And it grows to the point where it either suffocates the animal or starts to death. That's the only way to die. And several facial tumour disease in fact started up here in the northeast of Tasmania. It was first photographed by Christo Bars who was a National Geographic photographer and it has sped slowly southwards and westwards over the last 20 years. And we've seen local declines of about 95% on the east coast and as we've lost about 80% of the species now across its entire range. This map is a little bit old. The disease front is now right up to here at Wulmoth. We saw a devil with devil facial tumour disease three kilometres from the Wulmoth gate. So it will actually arrive to the tip here in the next year or so. It'll take a bit longer. There's a very large mountain range down here through Tasmania to get down into the southwest. And it's only restricted to devils. It's only found in Tasmanian devils. And as I said, it's an infectious cancer. Infectious cancer, you say. There's only three infectious cancers in vertebrates. One is it occurs in dogs. They've had it for about 20,000 years and it doesn't kill their host. And the other two in fact occur in Tasmanian devils. And even though we've been studying the disease for 20 years, we still don't know what the incubation period is. We don't technically have a diagnostic test how there is a PhD student at the Mendys Medical Research Institute who is very close to having one working. And once the lesions appear, we say they are nearly always fatal. And the reason I say that is because there are a few incidences where we have tumour regression with the cancer. And a second cancer was actually found in 2012. And how do we know that it's a clonal cancer? The DNA. So in 2012, this medicine actually released one of the first Tasmanian devils genomes. We know that the cancer arose from a Schwann cell. It's known as the Schwann cell carcinoma. It's a nerve cell in the face of a female devil. How do we know it's from a female devil? It has an X chromosome. The second cancer, how do we know it's so different from the first cancer? Genome sequences are very different. But more importantly, the second cancer has a Y chromosome. So it actually rose in the mouth. So it's a separate event. And what happens is that one female, Liz Moterson has managed to prove, fits seven other devils. And those seven devils got the cancer from that one female that first arose in. And that's basically how the epidemic started. And it's just been transmitted through the population, through biting. And devils bite themselves, bite each other, predominantly during the breeding season. So that's usually in April. And we see a massive increase in the tumours usually in September or October. It takes about six months for the lesions to actually start to appear. So Kathy and I started working together on this program back in 2010. Kathy actually started working on devil immunogenetics in 2007. And as I said, we've got the DNA, we've got the reference genome, first reference genome in 2012. We've just published, Liz Moterson's group just published a new one. We have a huge amount of RNA data for this species. And we've really been investigating the interplay between the disease itself and the devil's immune system. We've done massive studies into microbiome. We've done genome-wide association studies to try and understand why the tumours have progressed. We've studied peptides that are produced in the mother's milk because a female devil can have the cancer, but none of the offspring actually get the cancer. And we think that some of these peptides that are produced in the mother's pouch should protect the offspring when they're born because they have no immune system when they're born because they're so underdeveloped are actually helping protect the offspring. And all that information is fed into making management decisions. So when Kathy and I first started working together, I actually worked in the Zoo and Aquarium Association which ran the captive insurance population here in Australia and Kathy was at academia. And what normally happens in the traditional conservation research approach is you have your research idea, you do your research, you get published. Someone may or may not who works in management should read your publication if it's not behind a paywall. And they're implemented into management and you may have a conservation outcome. From start to finish, that's a linear approach. It's about 10 to 15 years before you actually start to see outcomes. We wanted to do what we called the tools and tech approach. Could we actually accelerate how we were gonna do things? Could we take management questions into research ideas, leverage management funding, work back and forth between the management teams and the research teams at the university and allow students to go on to publish their data and their research papers. As managers, we didn't have to wait for that to happen. We were just receiving data in real time allowing us to adapt our management decisions. And it works very well. It works so well with Tasmanian devils. We then rolled it out with bilbies and with boilies and this is pretty much the precursor for the threatened species initiative. This is what we've built it on because we know it works. We know if we can get data into the hands of managers in real time, we can actually start to make a difference on the ground. So for Tasmanian devils, we've now sequenced over, I think it's 6,000 devils over the last 15 years at about 1,300 SNPs, which doesn't sound like a lot, but that is a lot of SNPs for a devil and about 500 functional genes. So we have a target capture method that allows us to look at both the immune genes, the genes associated with their 2D and reproductive genes. And we use these decisions to inform captive breeding, to run the annual monetary program in collaboration with the Tasmanian government to make translocation decisions or the translocation decisions for the species that are made based on genetic data and also to use it for post-release monitoring. We've had some pretty significant challenges. When I first met Kathy, it was simply because all the devils came in the insurance population came from the northwest corner. And my simple question to Kathy in 2011 was, can you use genetics to help me tell whether or not the devils, patterns are related to the animal that came into the insurance population, whether related to one another? Turns out it took us seven years to answer that question because of the low genetic diversity in devils and not having sequencing technology and compute available to us at the time to be able to process the information. We also look a lot at the MHC region. There's a very strong link between MHC genes, which are immune genes and the cancer itself. Unfortunately, the genes have a 98.6% similarity. So when you short read sequence them on an Illumina platform, you don't know which order to put them in when you get them back. So the only way we can actually look at the MHC region is through long read sequencing. And even if we've got a 45X coverage home genome for Tasmanian devils, we still can't call haphal types. And that's because they've got such low diversity. So one of the things that we did is the Tasmanian government asked us, if we're going to translate devils and move animals around, which population should we start with? Where should we get devils from? What's the survival rate upon release and do they breed them as they release? All those questions can be answered with genetics. And so what we did is we took animals from Mariah Island where we had already released different devils from the insurance population and we translocated them to a place up here at Stony Head. And we wanted to see if we were going to make a more genetically robust devils. So if you can take devils that are genetically differentiated, if you put them together, can we get a boost in genetic diversity? But does that boost actually maintain for multiple generations? And what does that mean particularly at an immune gene level? So this works being led by two of our postdocs with the planin and Wanwan Chen. And just to give you an idea here, the pink dots is the genetic signature of the animals at Stony Head when we did the release and the darker blue dots are the animals from Mariah Island that we released. And the turquoise dots you can see between the two are the offspring that they produced. And so what we found using the genetic data is the minimum number of individuals that we released that went on to breed because we can see them in the parentage analysis. So there you go. You don't have to go spend 1,000 hours trapping animals. You just go the year after the breeding season and catch their offspring. We do know that Stony Head devils bred with other Stony Head devils. So they wanted to mate with the ones they knew. Same with Mariah animals. This is actually Nutella and Boomer's kids. Nutella and Boomer just mate it all the time together on Mariah Island. And then we put them at Stony Head and they continued to mate. They obviously didn't ever want to get the force. So they just kept producing the same offspring. But actually there's a suite of animals here between the dark blue dots and the pink dots which are actually the hybrid offspring. So those animals, when we do their parentage analysis we can tell where their parents came from. When we went on to then investigate the immune genes of these individuals what we found is that we had introduced new genetic variants. We didn't swamp out any of the local variants which is really key in a translocation not to swamp out any local diversity. But we introduced new genetic diversity and that actually confers a fitness benefit for the offspring in the fact that they have lower parasite loads and they have a greater capacity to start to mount an immune response to cancer. So is this the devil's solution? Is the solution to go and move devils around the landscape because they can't move themselves anymore because the density is so low and they live in a highly fragmented landscape? We don't actually know the answer to that question at the moment but that's pretty much what we're working towards because I live for the day that we get to see this happen where you've got a disease-free devil getting to run free in the landscape. It is the largest carnivore in the landscape and we know how the story ends from the Australian mainland if we don't protect our carnivores. We will lose the rest of the species that live in Tasmania. So really the age of genomics is here. It's here and we are using it. We can use it to enhance our understanding of global biodiversity. It has a wide range of applications for management and policy, ecology, transportation, evolutionary biology but there's still a number of challenges to solve. But the challenges that we need to solve should not preclude us from driving the dodgy old car down the road at the moment and actually just using it already in management applications. We really need to be able to develop methods where we can use long-read sequencing with low-quality DNA. That is the panacea for conservation biology. We can't get high-quality genetic samples. We need more efficient and effective annotation pipelines, preferably ones that use transcriptomic data from the country in which the species comes from. Oh my God, whole genome resequence data. I've got 70 terabytes of data for koalas sitting in the cloud at the moment, transferring that data 438 kilobytes per second to NCBI in the US is going to take months. But also just being able to do analysis pipelines for most of the population genetics tools that we use at the moment to investigate whole genomes comes from micro-satellite data. So now you go from having a few micro-satellites to 55 million SNPs. We need to be able to multi-thread those analyses. If we can't, they just take us too many months to run. We also need to start thinking about methods to interrogate large disparate datasets. So how do you take genomic data, species distribution data, habitat data, climate data like ultimate goal as a conservation manager, which is a challenge for all of you, is I want to be able to draw a polygon on a map and I want AI and ML to go find all that information for me from anywhere in the world, clean it up, bring it to me and give it to me in a really fast time point and stop laughing. Come on, you only ever like to get a challenge if I push you to it. But can you imagine after the bushfires, if we were able to do that, just draw a polygon on a map and know what was there and potentially what we had lost. And my favorite one of all is, can we please make really easy to use compute interfaces? Point and click is my desire. And the reason for that is, is because Galaxy is a really powerful platform. But if you are not a specialist, it is actually very complicated to use. And if you don't know what type of analysis you need to do, it's even harder. So if we want to make these tools more accessible to the people like myself who really need to use them, then we need much easier interfaces for us to work with. And I once asked one of my colleagues in the devil program really for him, what has been the difference with us providing all the genetic data to him? And he basically said to me that having genetic data available for the management of a threatened species is the difference between flying blind and flying with a navigation system. And so TSI is not just me, there are hundreds of people who are involved in this project from the conservation agencies who provide us with all the samples, the sequencing companies who sequence all the cloud compute and all the tech guys who really have come to the party about developing a suite of different tools for us to the funding agencies from the Australian Research Council. It really is a massive joint effort. And not only that, it's also been a massive effort from our research group and all the students and postdocs past and present. This is definitely not me. I don't even know how to run R, believe it or not. But I have a suite of amazing people who do wonderful work for us and really everything I've presented today is really standing on their shoulders and also for all the field teams who provide us and will work with over the years. So I'd like to finish today for where I started. So you look at what we've achieved in the 50 years since we took this photo. And my challenge to all of you here is how what you do in your everyday job is going to technologically advance us to allow people like me to make a difference on the ground to helping save the remaining divide diversity that we have left. Thank you very much. Inspiring. Thank you very much, Caroline. We can take questions from the audience and from online. If you've got a question online just put it in the Zoom chat and Melissa will advocate on your behalf. Any questions or comments for Caroline? I've traumatised everyone. I was wondering what caused the devils to die out 3,000 years ago and has there been any thought about like trying to reintroduce them onto the mainland? And would that be useful? I always get asked that question when I'm on the radio. So 3,000 years ago we think devils might have disappeared from the mainland for when Dingo's first arrived to Australia. Should we reintroduce them back to the mainland? Well, at the moment we're having trouble reintroducing them back to Tasmania. So I might wish to get that one done first. The biggest problem with the Australian mainland is the Australian mainland now suffers from significant predatory pressure. So not only do you have the native predators but you've also got the introduced predators. And my question to you is, how are you going to solve the problem if you put another predator into that scenario? Because all you're doing is adding another problem to a system that's already under severe stress. And there is a belief that, you know, if you introduce devils then you'll get rid of cats and foxes. Devils, when they have a run in with a dog, which is, you know, smaller than, like, same size as a fox, lose every time. So I don't think devils and foxes will ever, devils will never win in the fight with the fox. We don't actually know that because there's no foxes in Tasmania. With cats in Tasmania, devils don't get rid of cats. They just change the foraging behavior of cats. So cats will forage earlier in the evening if the devils are around because devils generally come out at 10 o'clock at night. So even if you introduce devils back to the mainland to try and get rid of the cat problem, you're not really just going to shift the problem. Hi, Caroline. Fantastic presentation. Lovely to see all the massive work that you guys have done. I was wondering about how far down the track are we to potentially consider moving the Tasmania devils back to mainland or whether that may be a consideration given that you have this cancer, infectious cancer that potentially can reach those areas where at the moment those populations are free of it and wondering in whether that is your plan B to potentially consider. No, and it wouldn't be my plan B because another cancer arrived roast spontaneously by itself in 2012. So Desi Euras, which is the family of my super family that devils come from, tend to get a lot of cancer anyway. They get lots of carcinomas in the climate. The fact that we've had two infectious cancers arise in the same species doesn't give me confidence that we should put them anywhere else where there's not the disease are really occurring in the environment. I think there's too many unknowns there. Not at this time. Come to you in the back after. So thank you for the wonderful talk. So you have now this pipeline for the assembly then you add the transcriptomes to figure out where the genes are. How important does it, or how much would it help you to have this functional analysis of the genes because I guess that's also a huge bottleneck. That would be hugely important. And actually we published a paper out of our research group this year that shows that automated annotation pipelines are only pulling out about 30% of the immune genes. We still have to manually go and annotate them. For every single species we work on, we manually annotate the immune genes. So if anyone can come up with a solution for that one, we'd be ever so grateful because it is very time consuming but we just can't, because immune genes particularly co-evolve with the pathogens that the species are exposed to, they differ in every species. So you can find the conserved portion of the gene using current pipelines and annotation methods but it's the variable regions which are actually the sexy parts really that we're interested in are the hardest parts to annotate. And from the kind of galaxy perspective, I have another question. We had a yesterday a meeting discussing kind of where we want to go with Galaxy and one of the questions is what do our users need? And I think you have a good idea. So it would be interesting to see kind of what does it specifically mean, the points that you've mentioned and kind of simplifying the interface because we have a room full of people here that no Galaxy inside out. And so I think it would be useful to figure out. One click would be my first request. And I think sometimes I think Gareth like Dredd saying my name come up on his, I know Steve Manos is like, oh God, here she goes again. I am not a shy wolf now when it comes to saying it doesn't work and I don't like it and I want something easier to use please. But then it helps, I think it helps you because otherwise, you know, you don't want to be making something that no one's finding useful. Thank you for the really awesome talk. I'm gonna be so excited to see that comes out of TSI and actually having like a dialogue between conservation managers and the people generating the genetic data. I was just curious about, I know you talked about using the genetic data for translocation of the devils. Has there been any like population modeling of possible like other conservation interventions? How that might affect the populations if you just let it play out over time or not yet? Yeah, so all the modeling, well, originally the modeling said that the species is gonna go extinct from the disease in 20 to 25 years. That hasn't happened. But what we found is that the disease has pushed the populations down so much. So some populations only have three breeding females in them. And we have a massive problem with road kill in Tasmania. So, you know, black devil at night eating something that's been hit on the road gets hit by a car. So if we lose one or two breeding females out of the population, the population just goes. So the problem now for devils, the disease is an issue but another problem now are small population pressures. So increase in breeding, low genetic diversity, road kill and fragmented habitat. So it sounds bad to say, but it's actually the conservation gift that just keeps on giving because it's got every single problem known for a species. And so because it's also such an iconic species here in Australia, it's been very well funded. And because of that, we've been able to develop all these different tools that we can now apply to other threatened species that have similar situations and kind of have more informed decision like how to do it. But yeah, we do use it for modeling parts significantly. Thank you. Thank you for a great talk. What opportunities do you see for cities and municipal government, especially those interfacing directly with the indigenous community? So I guess I'm thinking more in populated area versus remote area. Do you see any opportunities for city governments to take leadership using similar approaches? What, to using genetics? Yeah. I would like to see genetics in every government and conservation agency and local government areas and indigenous communities are managing their own land just the same way we use camera traps and satellite tracking devices. So I used to do satellite tracking back in the 90s when it first came out and used to have to know how to tell net and you know, dial up internet because shout out all day. You can tell net in and you get these long masses sequences of information out and you have to cut up the code to know what the data was actually saying. And it was really quite complex to use. These days, you put a tracking device on an animal and you just download an Excel spreadsheet from the internet. So, you know, I want to get to that point with genetic data that it just becomes yet another tool in the conservation toolbox that becomes readily available and easy to use. And the problem at the moment is because it's such a specialist skill for the sequencing and it's such a special skill for bioinformatics. People in the conservation sector are scared of trying to use it because they just think, I don't have time to try and learn something new. And that's how I feel with bioinformatics. I just don't have time to try and get my brain to expand anymore. So I think there's lots of capacity. We just need to build the tools to make it more accessible for people who need it. Thank you. Just a quick follow-up question on the buzzes. So, approximately, is there sort of ballpark? Would it be too large for a small city to take on? What the sequencing for populations? Yeah, to run a project like this, led by a city or a municipal government, is there sort of ballpark in terms of the overall budget to run a project like this? If you live in North America, it's about 300% cheaper than what I pay, to be honest. Sequencing in Australia is very expensive. But we can sequence a reference genome and do population genetics of an individual, of up to 100 individuals for eight particular species for about 13,000 Australian dollars. Thanks very much, everyone. I can see there are more questions. Caroline, you're here for a little bit. I'm leaving after lunch. Leaving after lunch. So if you've got more questions or want to have a conversation with Caroline, please grab her in a break. And join me again in thanking Caroline for a really inspiring talk. Thank you. Unreliably informed, it's now time for posters and a coffee break out in the queue.