 I'll be telling you some of the work on how we started with an NSF Eager project on building a genome for a critter known as the 13-Line Ground Squirrel, and that how led to a new early-stage pharmaceutical company started by three of my postdocs. Much of the things I've been interested in are now being pursued in the private sector, including work in the direct-to-consumer space around how we begin to think about the use of genetic data in drug discovery. A lot of work in comparative genomics, so I'll tell you the story of FanaBio, which is the company that we spun out of Stanford. It turns out to be the second company from an Eager that we've started. There's a dog genetics company called Embark that was started in part based on some work we had done along with Lane Ostrander and others in dog genetics. And two brothers named the Boiko brothers that many of you may also know are now at the helm of that company, Edeland Diagnostics, which is a company in horse genetics, and my friend Sam Brooks is here in the audience. She's part of that adventure. And you'll also be hearing some of the work we've done at Dovetail, which is a company owned by Eden Rock, where I'm a director. So highly conflicted, but let's get all that out of the way. So why am I at this particular meeting? So for a long time, I've been very interested in new and emerging model systems for a while. We had a sort of cheeky term in the lab for this called gonzo genomic. So how quickly could we go into a new system, make a genome, make a variation map, and go out after interesting trait? Joanna Kelly, who was a postdoc in my lab, led a lot of this effort. And she'll be one of our speakers later on in the day. She was very interested in extreme biology. So we did the Antarctic Midge, the very smallest little only insect that lives in the Antarctic, as well as a bunch of killy fish species that will probably rear their heads in her presentation. But ultimately, our goal is really to use comparative genomics to inform our understanding of the human genome, ideally with an eye towards therapeutics. So I liken it to sort of GPS coordinates where you've got many different satellites and allow you to triangulate into interesting positions. And really one of the key challenges that I think we face as a field is that as we continue to develop these model systems, unfortunately the federal government can't be the only funder. It may be the funder of last resort, but it shouldn't be the funder of first resort, particularly when these things have economic value. So how do we incubate these ideas and allow them to take a life outside the university? And much of my thinking has been influenced by this book. If you haven't read it, I encourage you to take a look at it. This is Bill Janeway's book called Doing Capitalism in the Innovation Economy, and he sets it up as what he calls the three-player game. So you have investment coming in from early stage by the federal government that really is sort of push innovation. How do we delve deeply into basic science? You then have essentially speculation by financial markets, which is early stage investment. And then you have the true market that shakes this out. And so I'd like to take that as a lens into how we think about this. And ultimately, as I said, our goal really is to move into the world of precision health. And I'm a firm believer that in a world of aging demographic, it is going to be the prophylactic deployment of pharmaceuticals that will avert disease into the future. I'm going to repeat that. In a world of aging demographic, it is the prophylactic deployment of medication that is going to prevent disease in the future. And perhaps the best poster child that we have of this is now the deployment of statins and other medications, particularly in the context of averting cardiovascular disease. Perhaps the greatest success we have in going bench to bedside made use of human genetics. It was the discovery of PCSK-9 inhibitors. This came directly from the work of Brown and Goldstein in understanding the basic science of cholesterol metabolism. And then Helen Hobbs and others really sort of took this on through the study of individuals at the extremes of the distribution. So as we began to study a disease called familial hypercholesterolemia, that is very high levels of LDL, so-called bad cholesterol, we began to understand that over expression of PCSK-9 was one of the key components of this near monogenic form of high cholesterol. And then it turned out simply because of the beauty of biology that if you were born with a loss of function mutation in PCSK-9, you were naturally protected from LDL, not only from high LDL, but also from adverse cardiovascular disease. Regeneron and others led this in clinical trials, and it was sort of a record time or turnaround from the understanding of the biology into the time we had a monoclonal antibody therapy that was being deployed in many people. We can talk a lot about the health care economics of how they got wrong, the pricing, but this is kind of the right idea. Let's look at extremes of the human phenotype distribution and really try to identify medications that can avert disease into the future. Just yesterday, it was announced that the latest in this work has now reached stage three clinical trials. So for those of you who are following this at home, Vennacumab, which is another one of these protective loss of function mutations, just showed a massive reduction in LDL and homozygous FH carriers. These are folks who, unless you put them on this kind of medication very early on, will have heart attacks and strokes at an early age. How do we accelerate this discovery? Well, lots of bets have already been placed. It is now an absolute cliche that data is the new oil. I would argue from a market cap perspective, this is absolutely true. When I was in college, the largest companies in the world were all oil companies. Today, they're all the tech companies. Tech has no choice but to go into this pace. Largely, it is now the largest component of the US economy. So health care now makes up 20% of the US economy, which is why everybody's sort of gunning for this and the idea that you can use data analytics to identify human mutations that are then potentially drug-able is now just a very clear investment thesis being followed by both public-private partnerships as well as the things that we know about in terms of health care systems like Regeneron that have partnered with Geisinger to enable this to happen. So putting this into context, where does the comparative genomics side come in? And this is really sort of the thesis. This is what I believe we need to do far more of with early-stage investment by government and nonprofits. Some amount of bridge funding. So the story I'm telling you today really was an eager grant. One of these things that NSF wants to turn around rapidly. SBIRs, I think, are the lifeblood for many early-stage companies. And that really then leads into placing the right bets. And we will let the market shake it out, right? Eventually, many of the things that are being developed internally then get sold up the biopharma chain. And you just need to look at that structure to understand there can't be too many $100 billion pharma companies. So they have to go through this sort of procession of gobbling up innovation. They themselves cannot innovate internally. So they innovate by allowing the financial markets to place the early bets and then stepping in. So in this context, this is Katie Graybeck, who's in the audience. She was a postdoc in my lab. She came from Sandy Martin's lab, where she did her PhD. And she brought this incredibly cool project, right? So she said, Carlos, I want to study hibernation and 13-line ground squirrels. And as I said, well, that sounds like a really interesting idea. Tell me about it. And turns out that these are critters that can get down to four degrees Celsius, which sounds pretty cool to me, right? So here's what the sort of lifecycle of this little critter looks like. This is the torpor that they go into during hibernation. So here is sort of the early fall. These guys are finishing their bouts of feeding. Then they go down into this almost pre-programmed circum-annual clock near freezing. Every few weeks, they've got to shake off the icicles. And we'll talk about why that's really, really interesting. And then the whole process repeats itself. So not only can we study this through the means of comparative and population genomics, because they are an emerging model system, what made this absolutely possible was NIH investment in being able to do experiments, right? So this isn't just an observational project. You can actually go in and carefully record all matter of life history for these critters. They can go into hibernaculum that you can do experiments, time, and begin to study how the variation actually plays out. And we'll talk some about why this was just really critical to getting this project right. So again, to sort of set up the biology, these guys are feeding, feeding, feeding all up into the summer. By this point, when they begin hibernation, they switch from glucose to lipid metabolism. There's absolutely no feeding, obviously, while they're in hibernation. And again, the sort of cycle repeats itself. So this really is a very, very interesting cardiometabolic phenotype to study as you're switching. And the question becomes, what could we learn from this? The other thing that happens, which we'll return to, is that hibernating animals are some of the only animals that we know of that can have neurodegeneration. So when they go into hibernation, they actually prune the dendritic tree, lose many of the neuronal connections during these bouts of interarousal. They actually redevelop these in a massive effort. And so the goal is, can we also potentially study this for some of the phenotypes of greatest interest, which include really the pressing issues with neurodegeneration. And as many of you know, billions and billions of dollars have been thrown at Alzheimer's therapies, none of which have really made it to market in any successful way. So the first thing we did was try to improve upon the existing genome assembly. So working in collaboration with Dovetail Genomics, which has one of these high C and high C-like products allowed us to dramatically improve the quality of the assembly. So we've now got hundreds of megabase scaffolding on which to hang much of this work. We then develop the kind of genotype bi-sequencing strategy one needs to generate markers all over the genome. I don't want to tell you about that, because it's now sort of a trend of true technology that many of us can develop. But there's just a ton under the hood here that I don't have the time to tell you. The goal, of course, was to take and ask questions about the inherent biology. So here is some variation from critter to critter in when they go into onset of hibernation. Here's a structure plot for the critters one has access to. And of course, there's tons of population structure, which is both good and bad when you're trying to map as many of you know, but you want to make the best use of the scenario. And in our case, given our expertise in high level models that can take into account both local structure and global structure, we were able to really use this to get at some very interesting biology as part of this NSF grant, including the top hits that are driving this sort of seasonal onset of hibernation. Many of these are, of course, in the pathways that you'd like, including feeding and insulin response and many of the traits that you'd, of course, want to try to go after. So from the point of view of the NSF, we had promised that we'd build a better assembly. We'd build a genome map and begin to get at the genetics of the trait. We were able to demonstrate timing of hibernation is heritable, much of the variance explained by a few loci. Many of these loci cluster in the genes that you'd be interested in. And the hypothesis that a multi-omics approach would allow us to get into potential biological pathways that would be potentially druggable. So at this point is where the story got very interesting, having the privilege of mentoring just extraordinary folks in the lab. I had three colleagues who decided they wanted to go on an adventure. And so three of the postdocs got together and said, there's something potentially really interesting here, put together a business plan, worked with a local accelerator, and got half a million dollars in funding to take this to the next stage. So the company's now led by Ashley Zender, who is a DVM PhD with expertise in cancer biology. She's founder and CEO. Katie, who is our expert in hibernation biology, is the chief science officer. And then Linda, who had worked with Eleanor and had been a postdoc at the Broad, leading many of the efforts in the Malian genome project, spent some time in my lab as a postdoc and then joined the company as CTO. And I think it's just been an incredible founding team to get this off the ground. The sort of investment thesis behind the company is that the huge array of interesting biology that gets explored by mammals could very well be what we need to sort of improve the bets that get placed in pharma. As I mentioned, the PCSK9 story told us that the genes that predict LDL cholesterol also predict early heart attack. So it begins to tell us how we might be able to make these investments. The contrast is that the genes that predict HDL don't predict MI. Five different drugs went to market because they were able to target LDL lowering and improve the clinical endpoints. Five different drugs failed in clinical trials. They were able to raise HDL, but they weren't able to make any movement on the clinical endpoints. So if we could add on top of this the comparative mammalian approach, this could rapidly accelerate. The argument is that human genetics maybe gives you a 2 to 3x improvement on placing the bets. How much could comparative mammalian genomics get you? Could you get to a 10x improvement and dramatically reduce the cost of bringing these pharma pseudocals to market? It's about a $2 billion in 10-year process. So anything we could do to improve that efficiency translates into better drugs for all of us. Hibernators turn out, because of their fascinating biology, to have lots of potential use in understanding diabetes, osteoporosis, ischemic injury, Alzheimer's disease, and many others. And so we began there. But of course, the idea is that it could go in many additional directions. Of course, what we want to take is ultimately a multi-omics approach with great data analytics. And so I would argue if you look at, say, the failure of IBM Watson to bring AI to cancer, it wasn't the AI. It was probably the data. There's only so much you can do from just ingesting a ton of articles. You actually need to do some damn experiments. So maybe if we did this experiment at scale in the way that I would say NHGRI has led the field, the wonderful thing NHGRI has done is continue time and again to invest in these technologies and enable us to do things at scale. We all remember the candidate Gene Winter in the early 2000s, where every week there'd be a different paper with an irreproducible result. Say what you will about GWAS for the love of God. At least it was reproducible. You could go into population after population and get odds ratio within a standard error, which is exactly what you want to do, and that has led to an acceleration in our understanding of the biology. The same thing is now happening with GTEC and other experiments that do functional genomics at a reproducible scale. So if we think about the opportunity to bring this into the context of industry, as this is getting cheaper and cheaper and cheaper, you're now beginning to build those layers of evidence that may be necessary to break this through. So the idea here is quite straightforward. Let's begin with the ground squirrel as a model system, understand the core genes that are involved in different aspects of the life cycle, and see whether this could be useful in our understanding of the biology we want to punch. Here is part of the tissue bank that was licensed from Colorado. We were able to take tissues at different stages during the life cycle. I'll focus on things that have to do with the late arousal. So this is right when the critters are in hibernation but have sped up their metabolism to shake the icicles. This is also when a lot of that neuroregeneration is happening, so potentially quite useful in understanding potential points to impact for neurogeneration. So, of course, what you want to do is take a kind of approach where you're using clustered gene expression analysis to get at pathways that might be useful, and that's exactly what I'm showing you here. So this is a slide from Fauna, where you're showing over expression in red and under expression in blue of different tissues, different genes in the hypothalamus during the late arousal period, the inter-period, inter-about period, and so on. And then you put this through a tool similar. This is the Brode's connectivity map, which has about one and a half million experiments for 5,000 small molecules and 3,000 genetic reagents in what their expression profile is. So you kind of play the inverse game, let me look for expression profiles that might target the small molecule. And in our case, when we looked at the hypothalamus, it turned out that 11 of the top 13 hits were already implicated in either Alzheimer's or similar in sort of beta-amyloid biology and many of the things that we knew go wrong when beta-amyloid begins to accumulate. In fact, this included the only approved treatment for Alzheimer's, memantine, which we'll talk about in a moment, as well as two new compounds, Fauna 2001-2002, that we've now begun to do some very early stage functional experiments. And I'm very grateful to the team for allowing me to share some of the data. So this is what typical assay looks like. You're putting this into a well-understood cell model for neuron. These are cells derived from a four-year-old patient in the 70s. They were then chosen and differentiated into these sort of neuronal pathways. When you add beta-amyloid, you begin to get cell death, and that's measured by this MTT assay, and then you can begin to rescue the phenotype. In this case, this is with the memantine, which is the approved therapy. And then here's what's happening with the new compounds. In fact, they seem to do a little bit better than memantine, and this is a decreasing sort of dose of the compound. Again, very, very early stage, but it begins to say, boy, there's some interesting biology here. There's ways that we could begin to add onto existing therapies. So this is sort of the logic, begin with whole genome data, add interesting layers of data that then improve our dark target identification, begin to license some of these assets up the pharma chain, and on and on we go as we improve our understanding of the biology and how to punch. And so this really kind of brings us back to the model I set out earlier, the sort of three-player game with the state investing early technology, some bridge funding, until some adventurous venture capitals are willing to put some money in. In this case, it was the age one incubator. In the case of fauna, they then were able to rapidly follow up with a seed round of investment. And I'll leave you with this last slide. It was a $300,000 initial investment in late 2016. It was actually December of 2016 when this was awarded. Less than two years later, age one picks this up with about a 20x increase in the nominal value based on the capitalization of the deal. And then a few months later, a second round of funding with the seed coming in led by TrueVenture gives it another sort of 2x on that. So we've basically gone in three years and increased the value of this idea by about 30 to 40 fold. And in my mind, I think this is exactly what we need to do to begin to take these and let others place the bets that allow the market to then shake out and see what's valuable. With that, I'm happy to take any questions you might have as part of the panel that we're all going on to. Thank you very much. So yeah, so could the speakers come down and sit at this lovely table here in front of everybody?