--- title: M3Site emoji: 📉 colorFrom: indigo colorTo: indigo sdk: docker pinned: true license: cc-by-4.0 short_description: 'An interactive demo for M3Site.' --- # M3Site: Leveraging Multi-Class Multi-Modal Learning for Accurate Protein Active Site Identification and Classification This an interactive demo for the inference, and the source code of the paper can be found in [Github](). We provide some example `.pdb` files in `case_study` folder. You can use these files to use this demo. Take `A0A384E143.pdb` as an example, the predicted protein active site results using M3Site are shown below: ![image1](img/image1.png) Besides, you can visualize and analyze the predicted results interactively like below: ![image2](img/image2.png)