| <!doctype html> |
| <html> |
| <head> |
| <meta charset="utf-8" /> |
| <meta name="viewport" content="width=device-width" /> |
| <title>MORAD</title> |
| <link rel="stylesheet" href="style.css" /> |
| </head> |
| <body> |
| <div class="card"> |
| <h1>MORAD</h1> |
| <p><b>Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design</b></p> |
| <p>MORAD post-trains a single backbone-conditioned RNA diffusion policy online with a six-objective reward covering tertiary geometry, base pairing, ensemble behavior and sequence composition. The trained policy designs sequences for a new target backbone directly, with no per-target optimization.</p> |
| <ul> |
| <li>🧬 <a href="https://huggingface.co/MORAD-RNA/MORAD-RIDE" target="_blank"><b>MORAD-RIDE</b></a>: the released model, post-trained from pretrained RIDE</li> |
| <li>📦 <a href="https://huggingface.co/datasets/MORAD-RNA/MORAD-Targets" target="_blank"><b>MORAD-Targets</b></a>: 527 training and 153 test RNA backbone targets</li> |
| <li>💻 <a href="https://github.com/Gabrile166/MORAD" target="_blank"><b>Code</b></a>: training, evaluation and reproduction scripts</li> |
| <li>🌐 <a href="https://gabrile166.github.io/MORAD/" target="_blank"><b>Project page</b></a>: overview, method and results</li> |
| </ul> |
| </div> |
| </body> |
| </html> |
|
|
