
GenIce3 generates hydrogen-disordered ice and clathrate hydrate structures for molecular simulation. It is the current GenIce. You give a unit cell name, optional replication, guests, ions, and an exporter; GenIce3 writes coordinates that obey the ice rules (GROMACS, CIF, LAMMPS, and others). It does not minimize the energy.
pip install genice3
genice3 1h > ice.gro
Quick start: genice3 1h > ice.gro (1h is Ice Ih; 4 is Ice IV).
| Section | Description |
|---|---|
| Getting started | Installation, requirements, demo, what’s new in GenIce3 |
| CLI reference | Command-line usage and options |
| Basics | Generating ice, supercells, seed, density |
| Clathrate hydrates | Guest molecules, cage types, occupancy |
| Doping and defects | Ions (CLI), H₃O⁺/OH⁻ and Bjerrum defects (API) |
| Output formats | Exporters (GROMACS, CIF, LAMMPS, etc.) and generation stages |
| Unit cells | Table of unit cell symbols (1h, 4, CS2, CIF, …) |
| Water models | Built-in water models (exporter water_model suboption) |
| Guest molecules | Built-in guest molecules for clathrates |
| Plugins | Extra PyPI plugins and custom unit cell / exporter / molecule plugins |
| Changes from GenIce2 | Cage survey and other changes |
| Citation | How to cite GenIce |
| Contribute | How to contribute |
| License | MIT License |
The API examples show how to use GenIce3 from Python with embedded sample code (basic usage, CIF I/O, doping, guest occupancy, polarization, unit cell extension, topological defects).
A concise, structured overview of GenIce3 for AI/LLM systems that need to explain the project or to drive it themselves: For AI assistants. A machine-readable index of this site, in the llms.txt convention, is at /llms.txt.