Building a new path to make medicines with AI

TL;DR

Isomorphic Labs declared "Our hypothesis has been proven" in its September 29 update. I have more questions about the business model. For a platform, I would want evidence that customers keep paying across successive programs. For a proprietary pipeline, I would look for named candidates, the evidence supporting them and a timetable for clinical testing.

Isomorphic Labs declared "Our hypothesis has been proven" in its September 29 update. Where I have more questions is about the business model.

The company reports designing molecules that open previously unknown pockets and validating their biological activity in the wet lab. It adds: "Our preclinical data is helping us to gear up for clinical developments."

That is meaningful progress. Given the ambition, I had expected to hear more by now about which medicines are approaching clinical testing.

I want to understand how much of the value Isomorphic can retain through discovery partnerships.

A platform business can work. I would want evidence that customers keep paying for it across successive programs. That would make a stronger case for a durable business than any single discovery result. After all, the best definition of product-market fit in my opinion is "how much will it hurt if I take it away again".

For a proprietary pipeline, I would look for named candidates, the evidence supporting them and a timetable for clinical testing. This update gives us little basis to judge that progress.

We have had the platform versus product debate in biotech for decades. I am still waiting to see whether AI changes the answer. Isomorphic is a particularly revealing test of it.