|
Download README.md from InstaDeepAI/Ibis: direct link, hf CLI and curl.
- Browser
- Download file 921 Bytes
-
https://huggingface.co/InstaDeepAI/Ibis/resolve/main/README.md
- Command line
-
hf download hf://InstaDeepAI/Ibis/README.md
-
curl -L -o README.md https://huggingface.co/InstaDeepAI/Ibis/resolve/main/README.md
921 Bytes
| license: cc-by-nc-4.0 | |
| # Ibis | |
| Ibis is a flow-matching model that generates conformational ensembles of proteins at Cα resolution from their amino acid sequence. It is the model studied in [Ablating Design Choices for Protein Ensemble Generators Across Structural Regimes](https://arxiv.org/abs/2611.05055) (Quinn & Bent, 2026). | |
| ## Checkpoints | |
| | File | In the paper | Training data for disordered proteins | | |
| |---|---|---| | |
| | `base/model.ckpt` | Base | CALVADOS coarse-grained simulations | | |
| | `mix/model.ckpt` | Fragment 50% | half CALVADOS simulations, half IDRome-o fragment-assembled ensembles | | |
| Both models are also trained on AlphaFold DB structures of folded proteins and on CALVADOS simulations of multi-domain proteins. They take sequences of up to 510 residues. In the paper's evaluation against experimental data, `mix` agreed better with NMR chemical shifts and slightly worse with SAXS than `base`. | |