tmol
tmol is a protein structure & design tool whose code is published by RosettaCommons on GitHub, first public in 2026, described in arXiv preprint (2026). GPU-native PyTorch implementation of the Rosetta energy function (RosettaGenFF): batched, differentiable all-atom scoring, packing, minimization and FastRelax for proteins, nucleic acids and ligands, usable as a reward in generative design.
| Category | Protein structure & design |
|---|---|
| Developer (GitHub) | RosettaCommons |
| Task | All-atom scoring (beta2016 / RosettaGenFF), side-chain packing, Cartesian/torsion minimization, FastRelax, gradients through coordinates |
| Interface | Python package (pip install tmol; tmol[cu12] / tmol[cu13] for CUDA), Colab tutorial notebooks |
| Website | rosettacommons.github.io/tmol |
| GitHub | RosettaCommons/tmol |
| Released | 2026-11-11 (source) |
| Paper | https://arxiv.org/abs/2611.21857 · arXiv preprint (2026) |
| Code license | BSD-3-Clause |
| Weights license | no separate learned model weights (parameterized Rosetta energy function) |
| What this allows | Permissive (commercial OK) Use, modify and ship it inside closed-source commercial products; keep the license text and copyright notices. What do these mean? |
| License source | api.github.com/repos/RosettaCommons/tmol · checked 2026-11-19 |
| Pricing | Open source |
| Tags | #energy-function#rosetta#differentiable#pytorch#protein-design |
| Last checked | · links verified |
Can I use tmol commercially?
Yes. The recorded license is permissive, so closed-source commercial use is allowed if you keep the notices.
- Code: BSD-3-Clause
- Weights: no separate learned model weights (parameterized Rosetta energy function)
- Source: api.github.com/repos/RosettaCommons/tmol · license checked 2026-11-19
What the license types mean · not legal advice.
Specs
| Task | All-atom scoring (beta2016 / RosettaGenFF), side-chain packing, Cartesian/torsion minimization, FastRelax, gradients through coordinates |
|---|---|
| Modalities | — |
| MSA required | — |
| Open weights | — |
| GPU need | Optional: CPU (incl. Apple silicon) or NVIDIA CUDA 12/13; Python 3.12+, PyTorch 2.11+ |
| GitHub stars | 166 |
| Interface | Python package (pip install tmol; tmol[cu12] / tmol[cu13] for CUDA), Colab tutorial notebooks |
Compare with…
Pick 1–3 other Protein structure & design tools to compare side by side with tmol.
Related
DeepMind's open-source protein structure prediction model (monomer and multimer).
Structure prediction of proteins, nucleic acids, ligands and their complexes; hosted web server plus open code, weights on request for non-commercial use.
Open biomolecular interaction models (Boltz-1, Boltz-2) for complex structure and binding-affinity prediction.