Protein Design · Harvard / Broad Institute
Few-shot active learning for directed evolution — 24 mutants per round.
No backend is configured for this model.
30s–5min
Not runnable yet
This model isn't available to run right now.Run this model
About this model
EVOLVEpro sits on top of ESM2 and learns from your assay data as it arrives, proposing the next round of mutants. Validated across diverse enzymes with as few as 24 measurements per round, which turns directed evolution from a screening problem into a design problem.
Standardized I/O contract
Every model in the Hub speaks the same contract, which is what lets the Router and the agent call any of them without special-casing.
Inputs
Wild-type sequence
Assay data (CSV)
Next batch size
Outputs
Ranked next-round candidates with predicted fitness and diversity.
