| --- |
| license: apache-2.0 |
| tags: |
| - chemistry |
| - biology |
| pipeline_tag: other |
| --- |
| |
| <p align="center"> |
| <img src="assets/disco.png" alt="DISCO: Diffusion for Sequence-Structure Co-design" width="900"/> |
| </p> |
|
|
| <p align="center"> |
| <img src="assets/carbene.gif" width="700"/> |
| </p> |
| |
| <p align="center"> |
| <a href="https://arxiv.org/abs/2604.05181"><img src="https://img.shields.io/badge/arXiv-94133F?style=for-the-badge&logo=arxiv" alt="arXiv"/></a> |
| <a href="https://disco-design.github.io/"><img src="https://img.shields.io/badge/📝%20Blog-007A87?style=for-the-badge&logoColor=white" alt="Blog"/></a> |
| <a href="https://github.com/DISCO-design/DISCO"><img src="https://img.shields.io/badge/GitHub-747474.svg?style=for-the-badge&logo=GitHub&logoColor=white" alt="HF"/></a> |
| </p> |
|
|
| DISCO (DIffusion for Sequence-structure CO-design) is a multimodal generative model that simultaneously co-designs protein sequences and 3D structures, conditioned on and co-folded with arbitrary biomolecules — including small-molecule ligands, DNA, and RNA. Unlike sequential pipelines that first generate a backbone and then apply inverse folding, DISCO generates both modalities jointly, enabling sequence-based objectives to inform structure generation and vice versa. |
|
|
| The model was introduced in the paper [General Multimodal Protein Design Enables DNA-Encoding of Chemistry](https://huggingface.co/papers/2604.05181). |
|
|
| ## Sample Usage |
|
|
| To run inference, first follow the installation instructions in the [official GitHub repository](https://github.com/DISCO-design/DISCO). You can then run generation using the provided runner: |
|
|
| ```bash |
| python runner/inference.py \ |
| experiment=designable \ |
| input_json_path=input_jsons/unconditional_config.json \ |
| seeds=\[0,1,2,3,4\] |
| ``` |
|
|
| ### Key Parameters: |
| - `experiment`: Use `designable` (steers toward samples more likely to refold correctly) or `diverse` (produces greater structural variety). |
| - `input_json_path`: Path to the JSON file describing the generation target (masked sequences, ligands, etc.). |
| - `effort`: Use `max` for full quality (200 diffusion steps, 4 recycling cycles) or `fast` for prototyping. |
|
|
| ## Abstract |
|
|
| Evolution is an extraordinary engine for enzymatic diversity, yet the chemistry it has explored remains a narrow slice of what DNA can encode. Deep generative models can design new proteins that bind ligands, but none have created enzymes without pre-specifying catalytic residues. We introduce DISCO (DIffusion for Sequence-structure CO-design), a multimodal model that co-designs protein sequence and 3D structure around arbitrary biomolecules. Conditioned solely on reactive intermediates, DISCO designs diverse heme enzymes with novel active-site geometries that catalyze new-to-nature carbene-transfer reactions with high activities exceeding those of engineered enzymes. |
|
|
| ## Citation |
|
|
| ```bibtex |
| @Article{disco2026, |
| title={General Multimodal Protein Design Enables DNA-Encoding of Chemistry}, |
| author={Jarrid Rector-Brooks and Théophile Lambert and Marta Skreta and Daniel Roth and Yueming Long and Zi-Qi Li and Xi Zhang and Miruna Cretu and Francesca-Zhoufan Li and Tanvi Ganapathy and Emily Jin and Avishek Joey Bose and Jason Yang and Kirill Neklyudov and Yoshua Bengio and Alexander Tong and Frances H. Arnold and Cheng-Hao Liu}, |
| year={2026}, |
| eprint={2604.05181}, |
| archivePrefix={arXiv}, |
| primaryClass={cs.LG}, |
| url={https://arxiv.org/abs/2604.05181}, |
| } |
| ``` |
