David Li - Meliora Therapeutics - Part 4

Intro - 00:00:01:

Welcome to The Biotech Startups Podcast by Excedr. Join us as we speak with first-time founders, serial entrepreneurs, and experienced investors about the challenges and triumphs of running a biotech startup from pre-seed to IPO with your host, Jon Chee. In our last episode, we spoke with David Li about his experience at Everest Detection, focusing on early-stage lung cancer detection and developing a non-invasive blood-based test. We also discussed the differences between life sciences and software innovation, product-market fit, de-risking core hypotheses, and distinguishing between one-way and two-way decisions. Additionally, we talked about fundraising, capital allocation, and the importance of knowing when to pivot or cut projects. If you missed it, be sure to go back and give Part 3 a listen. In Part 4, we talk with David about his transition from private equity to co-founding Meliora, focusing on their use of AI and ML in drug development. We'll also discuss mechanisms of action, the impact of Meliora's platform on precision medicine, and David's fundraising experiences. Additionally, David shares his advice for aspiring entrepreneurs and his reflections on thinking big and pursuing passions in the biotech industry.

David - 00:01:31:

So actually, this was contrary to the first couple of transitions where I knew I needed to leave private equity and KKR. At Benchling, I knew I wanted to dive deeper in life sciences. This time, in my transition to starting Meliora, being a co-founder and helping launch the company, I was not convinced at the very beginning that I was immediately going to jump in. I had run across Jason Sheltzer, my scientific co-founder's work, because he published some pretty controversial papers at the time, actually. He published a paper in Science Translational Medicine in December of 2019, and he found that there were quite a number of drug candidates, some preclinical, some clinical stage cancer drugs, that were mischaracterized in terms of their true mechanism of action. And the very simple experiment that he did was he did CRISPR knockout of the purported targets for these molecules in some cancer cell lines, and then treated those cancer cell lines. Now, the target's ablated via CRISPR using those same molecules. The expectation here is if the target's been ablated, then clearly you shouldn't see some of the strong anti-cancer effects you saw before. But for many of these molecules, Jason still observed, many of them had a very strong, potent anti-cancer effect. And so then he did some further wet lab follow-up assays and cross-screening to determine what was the true mechanism of those molecules, what was truly driving their anti-cancer profile. And then he published it. And so it was picked up in New York Times, Wired, a number of other consumer publications. And the reason why I think it was controversial is that you're basically saying biotech and biopharma, who spent a lot of time and money on some of these molecules, in some cases were treating humans in clinical trials, were not actually fully grasping what was the true mechanism and reason why these molecules were working. And I think that that obviously can ruffle some feathers. But as we've kind of gone deeper into this space and really bringing on really tenured, experienced drug discovery folks in our own team, this mechanism identification problem is one that folks have known about for a long time. There's always been dirty kinases that worked for a whole smorgasbord post of reasons, not because of a single target that it was really only hitting. So I think our story has evolved even beyond Jason and his other collaborators' initial scope. Because we realized that this is not just about kind of a mischaracterization problem in drug discovery. It's also fundamentally about, can we get mechanism right? Can we get the right chemistry with the right biology, with the right target biology, the right mix of target biology, such that we can start to rationally design in the truest form, a molecule that is connected to the right patient population. Because we know that it's sensitive to a particular mechanism, get the right biomarker and start developing a really fundamentally different probability success drug development funnel downstream when we get into the clinic. So that's kind of the ultimate vision of ultimately where we'd like to go. And the starting story then is I met Jason in, mid 2020, early 2021. It was just several months after Jason and his lab had been shut down for COVID and for the pandemic. And so Jason started asking the question. Really out of necessity, but also curiosity, if they could develop a computational path and algorithm to identify the true mechanism for molecules of interest, and then use that understanding mechanism to develop better cancer drugs. So in a very brief nutshell, then the thesis for the company was, and still is, can we use AI/ML to identify the true mechanism for molecules. And then leverage the understanding of molecules to connect it with the right mechanism, target biology, connect that with the right patient population, with the right biomarker, and improve our probability of success when we get downstream into clinical assets. We ran across Jason's work. Jason and I started talking, and the original founding team started talking quite frequently. Actually, everyone involved also made significant personal investments to get the company up and going and to generate the first data, which was fairly promising. And in mid-late October, really started thinking, well, I'm spending a lot of time here thinking through these foundational company formation questions. Jason was not going to leave his role at Yale, given that he really felt that he wanted to be an academic and that he was an academic at heart. And so we agreed that I would jump in as a founding CEO. And from there, I guess the story was we raised a small family and friends round, really built more of the tech, built more of the platform. In 2023, started expanding into becoming a full-stack therapeutics company. And in late 2022 help support that transition, we raised a seed round that was north of 10 million. So 11 million all in raised at this point. And we think we're only at the very, very beginning stages here of applying a lot of mechanistic insights that are coming out of our platform into developing real drugs.

Jon - 00:06:49:

Awesome. And did you just bump into him at a conference? How did this serendipitous meeting even take place?

David - 00:06:56:

On the side at Everest, I had started doing some more angel investing and looking at really interesting science and new data from scientists that perhaps wasn't even a company yet, or perhaps it was an early-stage company. And I really liked that. I thought, one, I had done some of that at Everest in the sense that we were looking at different technologies just to try to figure out if there was a commercial path to doing early-stage cancer detection in lungs specifically. And a lot of the scientists I felt benefited from having a good mix of kind of bio and business experience. For Jason specifically, yeah, in the very beginning, I actually invested before coming full on full time. And it was clear that we needed money to kind of de-risk a bit further around the data set. But yeah, as I was mentioning, we kind of started talking very frequently. We were needing more money. And as much as I would have loved to just keep self-financing it, I realized that we would need to go out and raise more money. The minute you need to go out and raise more money, you need to kind of really seriously think about what is the team that's going to be around to take that money and turn it into real value. And things started to fall into place of, well, if you're spending a lot of time with this team and the data is interesting and you need someone as a full-time CEO to step into it, perhaps this is a next step that we could take. And in the very early days of how we even met Jason, we met actually not even in person at first. Because we were still in COVID, we were at a number of scientific angel groups on Twitter. And I think that that was kind of the initial contact of meeting Jason because they were trying to do some experiments and put together the initial data package for the company.

Jon - 00:08:33:

The internet is a beautiful thing to be able to make those connections that otherwise may not have crossed paths, to be quite honest, and very cool that that happened for you. And so, you know, your co-founder is like, okay, look, I'm going to stay at Yale. And so like, you said, you're going to raise some institutional capital. Can you talk about the early founding team, building that team? And how did you approach that?

David - 00:08:54:

Yeah, definitely. So we had a technical co-founder, her name is Joan Smith, and she built a lot of the original algorithms, kind of original scaffolding for the platform. We also then hired on as our early ML scientist. We met through, again, the internet. There's like a Y Combinator, like co-founder dating site that we were both on. And so we met through there. And that, as good teams sometimes are likely to do, that you got to start bringing in folks that you have worked with in the past and know of. And so the first couple of folks were like that. And then over time, we really started thinking, at first we didn't have a full-time chemist, didn't have full-time drug discovery. We kind of outsourced that and got consultants in. But over time, we realized we really need that. And we especially wanted to have drug discovery feed right back into the platform. So in early 2023, we've been looking, actually a longer period of time, but early 2023, we were successful in recruiting our head of drug discovery. Now our CSO, his name is Claudio Chuaqui. And he has a 25-year run in the biotech industry as a chemist. And so he's worked on a bunch of drugs that have gone into the clinic, including co-leading a project that turned into osimertinib, which is Tagrisso, massive drug for AstraZeneca. I think it was last year it grossed over 7 billion. So he's seen a lot of different targets across his time at AstraZeneca, Celgene, BioGene, and most recently before joining us at Syros, which was an oncology therapeutics company. And so, I think that is really the core of the foundational team. In our view, having drug discovery side by side with the platform team very early is really critical. So even though we didn't immediately jump to that, because honestly we probably couldn't have afforded it, but the minute we started bringing in real capital, institutional capital at scale, so in the millions of dollars, we realized that we wanted to have drug discovery very early to help set the DNA of the company. I think one of the things that we're very clear about is even though we're a computational platform, we're here to make drugs. We're here to develop real assets that are clinically differentiated. We're not here just to kind of demonstrate that a computational platform can work and can have some impact of drug discovery. We're here to say, with our platform, we think we have a fundamentally better chance at creating drugs that will change patients' lives. So we're very critically focused on that. And you really need to have drug discovery people in order to have drugs.

Jon - 00:11:29:

Yeah, that makes a lot of sense. And so as you guys were scoping the market, obviously your platform has a very differentiated approach. Can you talk a little bit about the status quo and how your guys' platform is disrupting the status quo?

David - 00:11:41:

You know, the usual format for stages of drug discovery is you find some targets, you try to find some molecules that can screen and kind of hit those targets and bind against those targets. And then if you do find one that binds, then you put it into a mouse model or some other preclinical model, see if you have some anti-cancer effect. In our case, we're looking at cancer drugs. And then you put it into the clinic and hope it works in a human. So there's a lot of ways where we are fundamentally different in the way we think about drug discovery. And the main way I would say is trying to pull forward that mechanistic understanding of what the molecule is doing to the very early stages of preclinical development. In the process that I just painted in terms of the kind of traditional status quo, we oftentimes are equating or at least ascribing the true mechanism based upon incomplete set of assay results or data sets to try to predict or make the identification or link for mechanism. And this is the part where I think if you ask any drug discovery person that's been in the industry for a long time, they'll tell you it's not something that is very well sorted out. It's not easy to do, mainly because we can't run a wet lab assay for every single type of target, for every single type of hit. It's just not physically feasible and cost-efficiently at all. And so we have a molecule, for example, that's in our pipeline that originally started off its life as a kinase inhibitor for a certain kinase target. And then it was brought forward into a phase one, generated by a different company, generated good safety, but not enough efficacy to warrant progression into a phase two. And then we looked at that molecule. So it was parked. The molecule was then parked. We looked at that molecule and said, okay, looked at the computational signature of that molecule and what it was doing to cancer biology in our cancer models. And we used our ML system to really understand those analyses and find patterns. And we found that what it was doing looked a lot like CDK11 inhibition rather than the purported target that it originally was being developed against. And so then we tested against CDK11. And in biochemistry assays, found it to be a very potent sub-59 molar IC50 value. You. And this gave us more confidence that we had identified a different mechanism for that molecule that actually was driving the anti-cancer effect. So maybe a quick pause here. How could that have happened in general from a traditional method? How would that have been missed? There's a lot of blind spots here. One way is if you're developing, for this example, a kinase inhibitor, you will do a kinome scan usually. Do it from reaction biology or urofins or someone. And generally speaking, those panels only have the first top most popular several hundred kinases on there. If it's not on the panel, clearly it's not going to come back. And this is one of the big category of failures is kind of essentially a false negative because you need to have a priority brought into an experiment, your hypothesis for what you're testing for. Otherwise, you will never get a positive result. And that's one of the challenges. If you don't know what to look for, you're not going to find it. Versus in our computational approach, we're looking holistically at what this molecule is doing through a multimodal set of bioassay outputs and using the ML to try to guide us to, well, maybe you should look in these specific sets of targets that we can then ascertain using some wet lab. And that's a big failure mode is that you don't know the mechanism and then you kind of ascribe that mechanism to that molecule. And then similar to this particular molecule that I'm talking about for the CK11 inhibitor, maybe you go downstream, you actually put it into the clinic, but if you haven't found the right patient population where there is especially sensitivity, to that particular mechanism, then it's highly likely that either one, you're not going to get enough efficacy signal as was the case in this particular instance, or two, maybe you do see some, but it's associated with a tox and therefore your therapeutic index is not wide enough that you can't successfully make a drug. And so there's just a lot of ways where you should really be knowing what the mechanism is way early, as early as possible, so that then you can know what to optimize that molecule against and ultimately, again, kind of have stacked the deck in your favor, to ultimately have a successful clinical asset.

Jon - 00:15:55:

I love that kind of overview. And also, there are a bunch of listeners of the podcast, who may not come from the scientific background, but just kind of see these drugs get to market. And there's like, it was all figured out. We have a complete 100% understanding of the mechanism. He's like, no, like, despite stuff getting on the market and having these blockbusters, there are still very much large gaps in our understanding. As much as we wish it wasn't the case. There's huge gaps that we need to fill and really close that because it is a tiny miracle, if anything gets all the way through every round of the clinic. But the fact that you can go all the way through and still not know the mechanism is like mind-blowing.

David - 00:16:34:

And there's no requirement, right? It's not like the FDA is saying you have to know the mechanism. And we're not making that argument either. We're not trying to say, hey, you need to have the mechanism known to approve the drug. We're just saying you're going to have a higher chance of having something approved if you do know the mechanism. And being able to rationally design each of the experimental steps in getting that.

Jon - 00:16:52:

Absolutely. That makes a lot of sense. And you're talking about developing your own pipeline. You mentioned partnering as well. From a purely business perspective, how are you approaching that? Is it just a pure pipeline development, internal pipeline development? Or are you also partnering with external parties with your platform?

David - 00:17:10:

We are doing both, but we do believe that the vast majority of the value will be driven by our internal pipeline. So we think that platform validation for partnerships and doing some co-development to leverage the mechanistic insights that can come from our platform, that is all great and kind of part of the financing strategy for the business. Having BD dollars help support the overall financing needs, I think, is always a good thing. With a caveat that BD partnerships do take quite a bit of time and overhead, and you have to make sure you get that, we'll call it cash positive effect rather than having it really take you far afield. But yeah, we want to do several partnerships that hopefully can really help accelerate our internal pipeline. And the view is that the sooner and more effectively we can drive assets into the clinic and get the clinical proof of concept, the sooner we can actually establish this mechanistic way of thinking and AI-powered and ML-powered computational approach of identifying mechanism as a gold standard for what you should really be doing in drug discovery, at first in small molecule oncology drugs, but then over time in more medications and more modalities in ways that are broadly applicable through a lot of the drug development industry.

Jon - 00:18:26:

Awesome. Yeah. And it makes a lot of sense. And also the idea of supplementing cash flow with these partnerships as a leader is incredibly important to have this kind of diversity in financing. Again, as the old adage goes, don't put all your eggs into one basket. And then, you know, it's like your mom will tell you, yeah, whatever mom, like, yeah, but it's like, it's true. Not to say that you have to be splitting your attention 50-50, but it's always like this athlete where they can coexist. You can do two things at once. You can walk and chew gum simultaneously. So for anyone who's embarking on this, like, you know, we're in a new era in terms of fundraising. It's not 2020. It's not 2021 anymore. You got to start thinking creatively here and really figure out how can you leverage your know-how, expertise, your technology to the best of your ability to set your company up for success. And on that note, with regards to your fundraising journey for Meliora, can you talk a little bit about what that was like? Because I know 2020, 2021, just like to the moon, 22, 23 was a big hangover. We're now in 24. Can you talk a little bit about that journey and how it was for you?

David - 00:19:33:

So the fundraising environment has changed pretty dramatically. When we first started, it was already probably past the peak. We were kind of in late 2021, early 2022, things had already peaked in biotech. And so I always joke that we started fundraising. Every successive round has probably been in a worse environment. What I will say though, is I think it has become a lot harder to fundraise. But if you can really crisply tell a story that draws a line between where the world currently is and where the world needs to be in order to create value and start showing traction against that. I think there's still a market for that. It just takes longer and takes more data to convince folks that you do have that story to tell. And that's not necessarily, I would say, a bad thing for the industry. That gives us credibility over time that whatever we bring forward is actually going to create value. I think it really instills a good amount of discipline. Definitely not easy though, definitely not easy. And I think speaks to, as a leader, one of my biggest learnings so far in helping start the company and lead the company has been, that psychological resilience around, you know, it's never going to feel up and to the right. In fact, it's a lot going to feel like you're going sideways and there's real risk and this might not work. And as a leader, you have to acknowledge it, but you have to absolutely be able to own it and move forward effectively and lead the team in a way that is extremely effective and into a future that you can inspire confidence about it. Because if you don't have it, then clearly it's not going to pass along to the potential investors. And you do need in order to cross the next bridge.

Jon - 00:21:11:

Absolutely. And I think there's nothing wrong with being one transparent and vulnerable about it. It's like, look, things are hard, but that inspires confidence in itself that you can be self-aware. A leader that is not self-aware does not inspire confidence either where you're just like, it's all good. It's like, no, it's everything is perfect right now. And so I've always thought for Excedr's perspective is that kind of transparency. When things aren't going right, we'll talk about it. When things are going right, we're also going to talk about that. And hopefully on the net, it is up into the right. You talked about it's probably not a straight line. There's like volatility within it and embrace and acknowledging the volatility, but having the confidence that it will net out positive. So as you're looking forward the next one, two years, what's in store for you and Meliora?

David - 00:21:58:

For the next couple of years, we really are excited about how our pipeline is coming together and would love to nominate multiple programs that we feel have deep conviction around the target, have deep conviction around our clinical product profile differentiation for where the ultimately, is going to move the needle clinically. And then capitalize on all those insights coming out of our platform such that we can grow and make this transition from seed to a series A. And I think there's a lot of discipline right now around series A's. They're not easy to raise. But, we're excited and confident that with a lot of the progress we're making on the pipeline. And that's really where life science species in particular want to see that you have something differentiated that's coming out of your product engine. So, yeah, we're very excited about making a lot of great progress on all those fronts and looking forward to being a part of this next wave of AI for drug discovery. That's thinking about it. And drug discovery in a different way and really starting to move the needle clinically and how we get to some proof of concepts that this is a different way of developing drugs in the pharma industry at large.

Jon - 00:23:07:

That's awesome. And I'm really excited to just watch your journey because like in my lab at Berkeley, mechanism of action was just drilled into me. And I never myself never developed a pipeline. But I was like, oh, yeah, like, isn't that what you figured out out the gates and like hearing it from music? No, that's not the case. You're just like, oh, my God, this would be incredibly helpful to know up front, truly. But, as we're kind of rounding out our conversation here and thank you again for being so generous with your time. We have two traditional closing questions. So the first question is, would you like to give any shout outs to anyone that supported you along the way?

David - 00:23:39:

Absolutely. The first one, this one I haven't really talked about in public, but thinking through all of my experiences professionally over the last dozen plus 15 years has had a through line to my time back at Penn and the dual degree program and life science management program. And not a lot of people know this, but I actually was not originally slated to be admitted. So I was on the wait list for this program all the way up until I think a week before the first day of school and freshman year. But the director of the program, who I'm pretty sure made a unilateral decision to bring me off the wait list, his name was Andrew Coopersmith. It brought me into the fold. And I think from there really changed my trajectory and experiences in many, many different ways. The program has really fueled in concrete ways, but also in setting my interpersonal vision and aspirations for my career. So I would love to give him a shout out. I don't think I've ever talked about in public, but what I'd love to be able to thank him if the chance arises, but that's definitely one where I think it really made a really significant difference. And I think a very traditional one that I'd love to include is my family, both my parents, but also my grandparents who raised me and just gave me a lot of the foundational character traits and outlook on life that I think really resonate to me to today.

Jon - 00:24:59:

That's amazing. And the last question is, if you can give any advice to your 21-year-old self, what would it be?

David - 00:25:06:

Think bigger. Don't worry about what others are thinking about, think about exactly what you want and then work like hell to try to make it happen. That's actually the best reward is to just be working towards something that you're really excited about rather than somebody else's path. So I really strongly encourage everyone, my younger self, especially to really think hard about that question and build a life that they're really excited about.

Jon - 00:25:29:

That's amazing. And honestly, I don't think there's any other place that I could end this conversation that I feel the exact same way. It is those like little things, like getting off the waitlist that can have massive inflection points for a one's personal journey. So that's a beautiful thing. David, thanks again. You've been so generous with your time.

David - 00:25:46:

Thanks so much, Jon. It's really an awesome time to be here. And thanks so much for doing this series. I enjoyed it a lot and it was an honor to be here.

Jon - 00:25:53:

Thank you.

Outro - 00:25:55:

That's all for this episode of The Biotech Startups Podcast. We hope you enjoyed our four-part series with David Li. Be sure to tune into our next series where we chat with Nathan Clark, founder and CEO of Ganymede. Ganymede is the modern cloud data platform for the life sciences and manufacturing. Their lab-as-code technology allows you to quickly integrate and harmonize lab instruments and app data, automate analyses, visualize all your data in dashboards built over a powerful data lake, and ultimately speed up your operations to accelerate science or production. Prior to founding Ganymede, Nathan was product manager for several of Benchling's data products, including the Insights BI tool and machine learning team. Additionally, Nathan worked at a firm as a senior product manager and was also a trader at Goldman Sachs. Nathan's extensive background in machine learning and data systems across financial technology and laboratory technology and their applications in life sciences offers unique insights for founders to benefit from. The Biotech Startups Podcast is produced by Excedr. Don't want to miss an episode? Search for The Biotech Startups Podcast wherever you get your podcasts and click subscribe. Excedr provides research labs with equipment leases on founder-friendly terms to support paths to exceptional outcomes. To learn more, visit our website, www.excedr.com. On behalf of the team here at Excedr, thanks for listening. The Biotech Startups podcast provides general insights into the life science sector through the experiences of its guests. The use of information on this podcast or materials linked from the podcast is at the user's own risk. The views expressed by the participants are their own and are not the views of Excedr or sponsors. No reference to any product, service or company in the podcast is an endorsement by Excedr or its guests.