ml-jku/clamp

Code for the paper Enhancing Activity Prediction Models in Drug Discovery with the Ability to Understand Human Language

47
/ 100
Emerging

Implements contrastive pre-training on molecule-bioassay pairs using CLIP-based assay encodings and Morgan fingerprints, enabling zero-shot activity prediction from natural language bioassay descriptions. Provides pretrained models (including Frequent Hitter baseline) and supports downstream evaluation via linear probing on MoleculeNet benchmarks. Includes utilities for dataset preprocessing (FS-Mol, PubChem with 500k+ assays) and assay description augmentation.

109 stars.

No Package No Dependents
Maintenance 10 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 12 / 25

How are scores calculated?

Stars

109

Forks

10

Language

Python

License

Last pushed

Feb 26, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/ml-jku/clamp"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.