11.8.22-Tokyo-Showcase-Modulus-Discovery

S. ROY KIMURA: All right, great. Thank you very much, Steven. And I want to thank, once again, the MIT Industry Liaison Program for this opportunity to present today. So my name is Roy Kimura, co-founder/CEO of Modulus. And I was actually a postdoc some time ago, actually.

And actually the research that I had been doing back then at MIT is directly related to what I'm doing today. And so I'd like to give you a little sneak peek on what we do. So the Modulus Discovery team members, prior to starting Modulus, recognized in the early 2010s that computational technology for drug discovery had finally reached commercial viability. That is, the accuracy of the predictions coming out of this technology became sufficiently high, such that you could drive drug discovery programs. And yet the cost of computation, as you know, continues to drop, even today, such that you could imagine using this at a very large scale to significantly impact and accelerate drug discovery.

However, through our experiences, we saw that this technology, especially the latest incarnation, the simulation technology, was not being leveraged to its fullest potential. This is because of legacy processes. And you can't blame the big companies. The processes are producing. It's not broken.

And it continues to produce new drugs. And so just because a new technology comes out, you don't want to change the whole process. However, in 2017, we saw this particular problem as an opportunity. And the Modulus Discovery team members came together with a shared vision in mind. And that vision was to build a brand-new organization from the ground up, with repeatable, cost-effective, and scalable processes, so that we can leverage this best simulation technology of today to discover the needed medicines of tomorrow and therefore dramatically improve patient lives.

So this is reflected in our mission. Modulus Discovery is dedicated to accelerating the discovery of new medicines for patients and their families, through the innovative synthesis of computational intelligence, biological inspiration, and global integration. Since our founding in 2017, we have built and established and validated a proprietary drug discovery platform, using state-of-the-art and internally-developed software.

And we created this repeatable, scalable internal process that can reduce R&D timelines by up to 50%. Using this platform, we developed three novel small molecule drug candidates that are currently in pre-clinical studies. We have established collaborations so far with PeptiDream, Astellas Pharma, Nissan Chemical Corporation, and Fujitsu, Limited. And a part of our success in Japan stems from the fact that we are headquartered in Tokyo with an office in Kendall Square. And we have a great cultural as well as logistical fit with companies here in Japan. We've so far raised approximately $50 million US from leading life science investors.

The way our simulation-driven drug discovery works is illustrated here. We start off with so-called seed assets. These are particular compounds that show some promise, a particular way to evaluate potential activity of compounds, or completely new targets in the body and validation data around that target. We take those ideas and fire up our platform, consisting of our large-scale simulation capabilities, project-specific machine learning algorithms, and synergistic experimental tools, to generate a very large number of compound ideas, which are evaluated in our simulation engine, our simulation so-called computational assays.

Every top-scoring compound that comes out of this is then experimentally synthesized and evaluated. We end up, every lead compound ends up being evaluated in industry standard-level drug discovery assay tiers. These are standard in vitro, cellular, in vivo, as well as ADME tox assays. And so we are quite certain that we have high quality molecules at the end of the day.

So our proprietary know-how suite consists of these six modules. We have ModSim and ModScale, which sits at the base of our platform and provides sort of the fundamental protocols and workflows that enable us to run very large simulations at scale. ModScreen is our ultra large-scale virtual screening protocols that enable us to run so-called in silico screens for a very large number, billions of compounds. ModOpt is our protocols that enable so-called lead optimization of molecules.

ModSolve is our technology to identify and characterize novel or allosteric or so-called cryptic binding pockets on targets or protein targets. And ModBind is our latest addition to our platform. It is a fast and accurate proprietary method to predict ligand efficacy using a unique approach. And in fact, back in August of this year, we just announced our research collaboration with Fujitsu on the implementation and optimization of ModBind.

This ModBind technology, as I said, is a proprietary simulation-based predictor of compound efficacy. One of the greatest strengths of this technology is that it is an absolute and not a relative predictor. That is, unlike the leading methods in the field, we do not require experimental measurements of existing compounds to be able to predict a related compound. We can predict from 0.

And so this actually opens up a very large area of workflow possibilities. The other advantage is that it's a lot simpler to set up, and also much faster than the existing state of the art. This is our internal benchmark showing about a 20-fold speedup of our technology. And we intend to take this to about 100-fold speedup with Fujitsu, while we also maintain the accuracy levels that are seen in state-of-the-art calculations.

This is how our pipeline looks like. We have 10 programs running mostly in oncology and chronic inflammation. And just a few snapshots of our programs, our most progressed lead program is called Mod-A. We have identified a potent and selective cathepsin C inhibitor, a clinical candidate.

And so far, we have shown in vivo proof of concept in multiple models, including this ANCA-associated vasculitis model in rats, showing significant improvement in lung and kidney histopathology. Mod-B is our highly selective cKIT inhibitor, showing selectivity against over 350 kinases, including PDGFRs. Just showing you, by comparison, imatinib is another KIT inhibitor already on the market. And you can see it hits a lot of different targets, whereas ours is very clean.

Again, we've shown in vivo POC of this compound. And Mod-C is our potential first in class PARG inhibitor showing tumor regression in a sensitive cancer cell line as monotherapy. We have multiple proprietary crystal structures, multiple lead compounds that are very potent, and compounds that are efficacious in the in vivo models.

So we are actively seeking additional strategic partnerships worldwide. These could include ModBind, discovery alliances, R&D collaborations, and licensing of our lead assets. So please contact us if you have any further questions. And thank you very much.