Articulating the Long-Term Value of AI-Powered Approaches

In a conversation with Pharmaceutical Executive, Yerem Yeghiazarians founder and CEO of Soley Therapeutics, discusses the company's innovative drug discovery platform, which leverages stem cell research and AI/ML technologies. The platform, developed over 25 years, focuses on understanding cellular responses to stress and identifying novel molecules that drive desired cellular outcomes. The company aims to transform drug discovery by identifying drugs at the cellular level and then matching them to diseases and patients. Yeghiazarians emphasized the importance of a strong team, deep science, and adaptability in overcoming challenges in drug development.

A transcript of Yeghiazarians’ conversation with Pharmaceutical Executive can be found below.

Pharmaceutical Executive: How do you articulate the long-term value of AI-powered approaches to partners and investors who are increasingly overwhelmed with AI-driven discovery narratives?
Yerem Yeghiazarians: So, I think the key here, when you're trying to describe this to people, is to try to differentiate the approaches, and why we think our approach, has to be done clinical trials.

So traditionally, what's done is, again, as I said, people come up with a disease of interest, and then they study for years and decades the ideology of that disease to come up with a target. Then they come up with a drug that binds that target. They do a lot of MedChem, and there’s a lot of AI ml, involved with medicinal chemistry these days to bind the target better. Have predictive ways of binding target better, and then they have to validate it. They do all the downstream experiments. Well, that's one way of doing it. And you know at times it works, but the incorrect target selection surfaces, latent development, right? You think you know the target, but that target may be the wrong target, it may be insufficient target, and you won't know that until the clinical trial fails. And unfortunately, for complex diseases, as people and yourself are well aware of, there's over 90-95% failure rate for many reasons, but this is one of the main reasons, because we, unfortunately, despite advances in science on many fronts, we don't know the target that to go after for many of these diseases, so we decided not to do it that way.

So our approach initially identifies novel molecules that drive a desired cell fate. That's really important, because we can filter out broadly toxic molecules early, that's important. Number two, once we have a drug that’s identified that gives rise to those phenotypes, we can define key targets, networks and mechanism of action at the whole cell level, not just the one upstream target, and this allows us to uncover novel targets and mechanisms that otherwise would have been missed. So that's another important point that I still haven't told you about, disease and patient selection, we essentially have now a drug that we are very high on. It passes drug ability criteria. It's been filtered out for toxicity, and we have novel MOA. It's only then that we put a completely novel approach to figure out which disease and which patients with that disease who then enroll in the clinical trial. That's what I meant by flipping the whole thing 180 degrees. We don't start with disease, target drug, downstream experiment, we flipped it around. We discover the molecule at the cell level, do the MedChem to figure out how it works, then figure out which disease and which patients to go after. So this provides an unbiased view from the start, and the cell is helping us make the best selection.