Tessel retweeted
When a drug meets a target, they both tend to change shape
Here's a sneak peek of Sesame-2: the first generative model to design around induced fit
@arvindtessel and @kayatsenko literally making moves with molecules
Tessel retweeted
Had a blast repping @TesselBio on the @PsymedVentures podcast!
Come for the hot takes on AI x Bio; stay for the nuanced discussion of what it means for a model to be good.
New Business Trip epi on virtual models in biology w/ @NTallapragada of Tessel. The past couple of episodes helped us shape our view:
- You need proprietary data, and right data beats big data
- Focus on functional data. A drug can change a bunch of genes and cure nothing.
- The cell is the wrong scale for startups . Tessel focuses on tissue, Powerhouse on mitochondria.
- Perturbation atlases are a tool, not a company
Listen to the full episode: businesstrip.fm/1173518/epis…
Tessel retweeted
*This* is The Bitter Lesson of Biology
(and why we focus on tissue *function* @TesselBio)
“A few bytes representing the result of one clean experiment resting on few assumptions can easily outweigh the causal information content of terabytes of associational data.”
Perhaps one day the field will stop chasing the red herring of dataset size - data that has the quality to reliably answer mechanistic questions is the scarce resource for understanding disease, data quantity by itself without any qualifiers is irrelevant.
A few bytes representing the result of one clean experiment resting on few assumptions can easily outweigh the causal information content of terabytes of associational data.
Tessel retweeted
1/ Open, Sesame ⚗️
This week, we released Sesame: the first generative chemistry model that fills any protein pocket with drug-like molecules – de novo or from a scaffold – with no constraints on size or class.
Tessel retweeted
Big day for @TesselBio: We’ve been awarded up to $4.2M by @ARPA_H to help make the “animal-free IND” a reality!




