mlxmolkit is a GPU-accelerated molecular toolkit on Apple Silicon, it is a port of nvMolKit that uses Cuda.
There are now three pipelines
- Molecular Clustering — Morgan FP → Tanimoto similarity → Butina clustering
- 3D Conformer Generation — DG (4D) → ETK (3D) → MMFF94 optimization
- PM6_D semi-empirical SCF — full d-orbital NDDO (S/P/Cl/Br/I) with PM6-D3H4 corrections
Full details are on GitHub https://github.com/guillaume-osmo/mlxmolkit
What’s new
Semi-empirical SCF on Apple Silicon — 7 methods (RM1, AM1, PM3, PM6, PM6_SP, PM6_D, AM1*) plus PM6-D3H4 post-SCF corrections — bit-exact to PYSEQM for PM6_D, with no PYSEQM/PyTorch runtime dependency. Every entry point is covered by tests/test_{methods_api,pm6_d_native,pm6_d3h4,pyseqm_port,rm1_scf}.py (83 tests total).
Requires macOS with Apple Silicon (M1/M2/M3/M4). RDKit is needed for molecular input:
conda install -c conda-forge rdkit
pip install mlxmolkit-rdkit