⬡ App ⬇ Download 📗 Books ⚗ QSAR 🔬 Research 👤 About

Enter a SMILES string

Select which activity models to run, then calculate.

2D molecule structure
A note on accuracy — descriptors (MW, LogP, TPSA, etc.) are exact RDKit calculations. Activity predictions are estimates from a machine learning model trained on public ChEMBL data — treat them as a rough potency ranking, not a lab measurement. Each prediction shows a confidence level based on how similar the input molecule is to the training data.

Upload a CSV of SMILES

CSV should have a column named smiles (or SMILES will be read from the first column). Up to 100 rows per upload. Structure images are skipped in batch mode for speed.

Supported activity predictions

Descriptors and drug-likeness checks work for any valid molecule. Activity predictions are available for the targets below — more will be added as new models are trained.

HDAC11 (ChEMBL CHEMBL3310)