Demos
From every protein to one design problem.
Each one starts with every protein in an organism, narrows to a single target, puts it where it actually sits in a cell, and then asks what a drug against it would do to everything else. Nothing on screen is an illustration.
Every source and version travels with the sequence that uses it.
Design sequences
01
PD-1
The receptor behind Keytruda and Opdivo, traced from a map of every human protein into a model of how the whole cell runs.
Watch
02
EGFR
A cancer target with six approved drugs, where the side effect that caps the dose comes from hitting the target itself and shows up in the skin. An earlier design of ours bound it at 630 pM.
Watch
03
GLP-1R
The receptor behind the GLP-1 drugs, whose molecules were aimed at what happens to the whole body rather than at the receptor itself.
Watch
04
Acetylcholinesterase
The Alzheimer’s target where thirty years of chemistry produced excellent binders, four approved drugs, and no change to the course of the disease.
Watch
05
Calvin-Benson cycle
How plants capture carbon: thirteen steps, a solved structure for each, and every one of them improved at the same time.
Watch
06
Andes hantavirus
Three viral proteins against every drug the FDA has approved, in 33 seconds, and the structures behind the two candidates it puts at the top.
Watch
Interactive views
The same data, with the controls exposed.
Which proteins touch which, for seven organisms. What binds them. How a bacterium reaches into a human cell. Where everything sits. Hover, filter and read the predictions yourself.
Open the interactive views