Tutorials

These Jupyter notebooks run end to end on the built-in toy engine — no GROMACS and no GPU required. They are rendered here from their saved outputs; to run them yourself, launch them from examples/notebooks/ in a clone of the repository.

• AIMMD toy tutorial: 1D double-well path sampling
• AIMMD multi-system (multi-ligand) tutorial

• 1D toy system (AIMMD toy tutorial: 1D double-well path sampling) — a complete single-system AIMMD run on a 1-D double-well toy model, from configuration through launching, training, and analysis of the resulting path ensemble.

• Multi-system run (AIMMD multi-system (multi-ligand) tutorial) — two toy systems with different atom counts in one run, trained with a single shared committor network (contrasted with per-system networks), including per-system committor and rate analysis.