Jong Lee Presents Day Zero Diagnostics at LSI USA ‘23

Transcription

Jong Lee  0:06  

Hi everybody, it's gonna be a little bit of a change of pace. I think most of you're used to seeing presentations on surgical devices, this a little bit of a life sciences turn rather than surgical therapies. So these are all diagnostics, we are infectious disease diagnostic company were founded in Boston by a group that includes people from Harvard, MIT and Mass General Hospital currently have about 15 employees. It's a pretty deep tech company. So most of the technical people in our organization. Company's raised 33 million in equity funding thus far. And we've supplemented that with about 20 million in non dilutive funding and we're currently raising a $20 million round. Importantly, one of the investors in our company is actually Becton Dickinson, which is one of the large strategics in the category. They invested in us after doing four months of due diligence to make sure that we weren't Theranos. Our mission is to redefine infectious disease diagnostics, using whole genome sequencing to enable rapid digital ID and AST. And for those of you who don't know what AST is that's antibiotic susceptibility testing. One of the primary missions we have as an organization is tackle the problem of growing antibiotic resistance, which you may have heard is a worldwide pandemic that is anticipated to create more deaths than cancer in the year 2015. In order to be able to do that, we've had to do a lot of deep tech development of unique capabilities and technologies that don't currently exist in the world. So we've developed proprietary methods for doing ultra high enrichment, which is gives us the ability to take a clinical sample like blood and make it sequential and find useful information. We've developed machine learning algorithms for highgrade species ID and antibiotic susceptibility profiling. And in order to be able to do that we've had to develop some world class databases that don't exist outside of our organization, to be able to support are able to do highgrade species ID, but most importantly, to be able to take a machine learning training approach to AST profiling. We've really taken our technology and targeted sepsis and bloodstream infections as our first application, the technologies, we're developing our platform, they're designed to work on different types of samples. But each sample requires its own optimization, and the clinical situation is different. So we've tried to start with the application that we think has the highest and greatest unmet clinical need. Sepsis is responsible for one in three hospital deaths. It has a 10 to 50% mortality rate, and it's responsible for 62 billion in health care costs just for CMS alone. So we're targeting the opportunity, because we think it's a great unmet need. But the other reason we're targeting bloodstream infections is because of all the different sample types I've put up there. Urine for UTI respiratory samples, CSF for meningitis, that's actually the most difficult sample to work with. And our strategy as a company is to say, if we can make it work in blood, we can make it work anywhere, whereas the reverse is not true for any of the other samples I mentioned. Just to give you a little flavor of the science that we work in, when I talk about ultra high enrichment and the need to develop a sample prep method that delivers ultra high enrichment. The reason for that is that in a sample of blood that I might take from you if you were sick, and in the ICU at the hospital, that sample of blood will typically contain about a billion times more human DNA, your DNA from your red blood cells, white blood cells, everything else compared to the pathogen, the bacterial cells that are in that sample of blood. So the problem we start with is how do you find something when there's a billion times more of a contaminating DNA, which is your DNA when what you're really interested in is the bacterial DNA. And the way we've tried to approach that problem is very different than the traditional infectious disease diagnostic molecular diagnostic approach, which is PCR probes. So PCR probes kind of take a targeted approach, they use binders to primers to bind to certain nucleic acids and extract that piece out of the haystack, it's using a magnet to pull the needle out of the haystack. Unfortunately, it's a very limited amount of information you're able to achieve. And frankly, it's not sensitive to work directly from blood. So people use blood cultures, which typically has about 10,000 to 100,000 bacterial cells in it, we're trying to work at the level of single digit bacterial cells in blood. So our approach has been different. Instead of using a magnet to pull the needle out of the haystack, we try to burn the entire haystack away. So we're agnostic ly left with whatever the needles are that happened to be there. Our technology blood back has been able to go through a multi step process, where the incoming sample starts with a billion times more human DNA than bacterial DNA. By the time we're sequencing we're typically generating 10 to 100 times more bacterial reads than sequencing the human DNA reads. So for those of you doing the math, that's about 10 to the 10th logs of enrichment, 10 logs of enrichment that we're getting to this the next closest thing we've seen in the space is about 10 to the third So literally seven orders of magnitude more enrichment through our process than what's currently available in the world. What that really allows us to do is we're able to recover the whole genome of the bacterial pathogen rather than just snippets of DNA, which is what everyone else retrieves. So people typically worked with maybe two 3% of the genome of the bacteria when they do their diagnostic. And they try to do species ID from that, it's impossible to do as t, what we've tried to do is develop a method to recover the entire genome. So we can do very high reliability ID but also be able to do a as T prediction. Once we generate the sequencing data, which takes us about an hour or two, which is unprecedented, we apply two algorithms to it kinome, Id kinome jst to be able to do organism ID, and then ASP profiling. Both of those are really based off of databases that we've curated and built over time over the last five years, including micro ref dB, which is our database for doing highgrade organism ID. But more importantly, the more unique acid is micro dB, which is our database for doing resistance prediction. So micro dB, sorry, I'm gonna go back. microbe DB is one of the world's largest data sets of its type in the world. There are lots of datasets in the world that have sequencing data for pathogens, there's only two data sets of large scale in the world that combines the pathogen sequencing data with ASTM formation, the antibiotic susceptibility profiling information, and our company happens to have one of those. Unfortunately, I've gone ahead and now I can't figure out how to go backwards on the slides. I don't know if you guys can help me there and put me back about five or six slides. Keep going, keep going, keep going, keep going. Keep going. Great. But we'll start there. So going back to organism ID, we can show you a lot of verification data, it'll look like a lot of plots and things like that. So I'm going to avoid that. Instead, I'll just focus on something we recently did for cardbacks, which is one of our funders. They asked her to do a blinded challenge to test whether our capabilities are what we said they are. And the blinded challenge was they sent us 40 isolates of bacterial pathogens without telling us what they were, we had to run it through our process at single digit Cfu per ml Spikings into blood. And we reported our results to them. And then they told us whether we were right or not. So out of 40 samples that they sent us, we got 39 out of 40 Correct 12 Different organisms identified that were unique. One thing we're really proud of is that we had no false positives in these results. And that is the endemic problem of molecular diagnostics is that most nucleic acid based diagnostics are actually quite sensitive, but they have poor sensitivity specificity. One of the things we're very proud of is the fact that we have very reliable organism ID results. So with our organism ID and AST capabilities, we're going after a $10 billion market. But really, in principle, the way to think about what we're trying to do is we're trying to create an entirely new pathway for doing diagnostic testing in the hospital, Microbiology and Molecular labs. Today, all of bacteriology is based in culture, culture takes days to grow. And then it takes additional days to do testing after that. So by the time diagnostic results are available from culture, they're almost irrelevant to the treatment decisions that have already been made. Ours two days in advance of the diagnostic results actually being returned. It's a terrible standard of care in infectious disease diagnostics, because it's a carpet bombing approach. And the reason it's a carpet bombing approach is because diagnostic information is not available to a clinician in the time that it needs to be to do something more precise. We're trying to use sequencing to be able to do that. The other aspect of our business is not just that we do infectious disease testing. But if you think about how our diagnostics is intended to work, every time somebody runs our diagnostic, they generate a whole genome sequence that didn't exist before. And because of that, and our ability to aggregate that into a map, large dataset that's been generated every day and attitude, we get the opportunity to opportunity opportunity to analyze this data at scale, and be able to do things like string tracking, so we can tell somebody whether something the sequence today in New York is the first time that pathogen has been seen in the New York metro area. We can tell somebody whether something the sequence today is identical to something sequenced yesterday from a different patient, ie a hospital outbreak that's occurring on the spot that they can diagnose early and get in front of, we can also do epidemiological prediction. Just I'll finish with one thought here in terms of we're currently on the path to developing this diagnostic and creating the hardware device. We're in the Alpha prototype stage for that we have a few years left to go to complete our cycle of clinical testing and FDA Michigan. So from an investor standpoint, I think one question to ask is Why is now a good time to think about investing in the kind of stuff we do. And I would argue that this is a very unique period of time. In the last five years strategics in the space have been taking a wait and watch approach because people like Becton, Dickinson, Bo Murrow Thermo Fisher don't really have an incentive necessarily to push sequencing, it really disrupts their current franchise business in the microbiology space. What's happened just in the last six months is there has been a major shift in the technology profile of sequencing with some of the advancements that have recently come to market. There's been a shift in the way the FDA and major government agencies are approaching sequencing. And there's been a shift in the way the industry ecosystem is talking about how they're thinking about sequencing. The thesis we have as a company is that in the next three years sequencing based diagnostics are going to hit the microbiology lab in the hospital lab incredibly hard. And Day Zero diagnostic is very unique and having a safe set of capabilities that allows not just testing, but actually entire pathway, a workflow capability in a microbiology lab that we think will be a really unique asset that helps one of the strategic players in the industry, become a winner in the next generation of technology that impacts microbiology. Thank you everyone for your time.