1Mby1M Virtual Accelerator AI Investor Forum: With Mo Islam, Partner at Threshold Ventures (Part 2)

Sramana Mitra: So let’s double click down on drug discovery. It’s something that I’ve been spending time on as well. What have you invested in and how are you thinking about AI-enabled drug discovery?

What’s the criteria of what you want to invest in and what have you invested in?

Mo Islam: We have a couple of companies in our portfolio, like Atomwise, which was one of the earlier companies leveraging deep learning for matching small molecules to their various targets. It’s a deep learning-based approach to drug discovery. That was an early company we’d invested in. Verge Genomics is another investment, which is focused on neurodegenerative diseases. Both of these companies are AI platforms.

Many drug discovery companies have a single asset—a small molecule or biologic—that they are betting on. We tend to focus on AI-enabled platform companies that could have multiple assets to potentially take to market. The probability of success to get a drug with any one predicted asset is very low, and the probability of failure is high. So, having an AI platform that can help predict and take multiple shots on goals is phenomenal.

Sramana Mitra: Let me ask a question on that one. So, Atomwise has a go-to-market strategy as a platform as a service on which other people are going to do drug discovery, or is this company doing drug discovery itself?

Mo Islam: They’re essentially partnering with a number of different biotech and pharma companies where they would provide the drug discovery engine and then those companies would take that technology platform and advance assets. For example, if a small molecule shows early signs of efficacy, they would move it into animal models and eventually in vivo studies in humans.

Verge is doing a lot of this themselves, keeping some programs in-house and taking them all the way through. Different companies in our portfolio have different go-to-market strategies, but our thesis largely revolves around backing AI platforms, especially in areas with unmet needs.

Atomwise has several programs they are working on. Verge focuses on neurodegenerative diseases like Alzheimer’s and ALS, which are very tough areas with a lot of unmet need. They have a unique approach using human data, which we found interesting and that’s why we backed them.

Platforms within drug discovery have been a big part of our thesis, but different go-to-market strategies can work to get those drugs to market. Ultimately, the long-term goal is to discover something within the platform and either partner with a biotech or pharma company. Or if you’re going to take the risk yourself, raise enough capital and bring a particular program to market. It is a 10-15 year journey, so it is a lot.

Sramana Mitra: I was about to ask you where in the cycle of drug discovery are you. Let’s take Atomwise. How far along are the customers of Atomwise who are using that platform in getting close to maybe Stage 1 clinical trial? Where are we in this evolution?

Mo Islam: Verge is in Phase 1 now with one of their programs, so they’ve actually made it quite far. With Atomwise, it’s somewhat dependent on their partners and how far they’ve taken it. They’re largely in the preclinical stage today, but we see promise in where they can go. Drug discovery is a marathon process, and it can take quite a while, but that’s what we look for in these companies—eventually getting to market.

The nice thing with drug discovery is that you can get upfront payments along the way. This has been used to capitalize some of these businesses. Often, the model includes upfront payments, milestone-based payments, and eventually royalties when the drug is actually brought to market. Even before that, there are ways to capitalize the business by making progress in the overall process.

Sramana Mitra: Who is paying those upfront payments? Are the pharmaceutical companies paying to Atomwise? Is that what you’re talking about?

Mo Islam: Correct. It’s the pharma and biotech companies that they’re working with. Pharma has been outsourcing R&D to smaller biotech companies for a long time.

Sramana Mitra: Yes, for a long time. They’re buying technology; so obviously, they have to pay upfront. They don’t have to wait for the drug to be successful for that to be payment worthy.

Mo Islam: Exactly. That is why they’re willing to do it. If someone has a more novel approach that could make that marathon process faster, that’s worth risking tens of millions of dollars because the prize at the end could be worth billions.

Sramana Mitra: Let’s say you have a startup that works on the Atomwise platform and is trying to do their own drug. That’s more on that 10-15 year timescale, right?

Mo Islam: Exactly. Even for these AI platform companies like Atomwise and Verge, ultimately the goal is to get a drug in market that’s been discovered from their platform, right? Ultimately, that’s what long-term success looks like.

Both these businesses are doing pretty well just based on the progress that they’ve already made, but long term success’s about getting a drug to market.

This segment is part 2 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Mo Islam, Partner at Threshold Ventures
1 2 3 4 5