Getting Started#
Enerzyme trains neural network potentials (NNPs) for enzymatic and molecular systems, then uses them for prediction, simulation, and active learning. The tutorials below follow a typical workflow:
install → prepare data → train → predict / simulate
↓
active learning ← QM annotate ← extract
Optional Enerzymette tools assist with PLUMED scans, NEB path building, ORCA/TeraChem bridges, and workflow launchers.
Contents:
- Installation
- Preparing a Neural Network Potential Dataset
- Training a Neural Network Potential
- Evaluating a Trained Model
- Running Simulations with a Trained Model
- Enhanced Sampling and Hybrid Potentials
- Active Learning for Neural Network Potentials
- QM Data Annotation
- Extracting Fragments by Local Uncertainty
- Prediction Server and Enerzymette Integration
- Bond Assignment and Utility Tools