A cross-attention-empowered SE(3)-equivariant graph neural network architecture for predicting enzyme substrate specificity. We are actively working on building the dedicated user interface (UI) for EZSpecificity. You can access the early version of our model (the frontend staging environment) via the following link: https://ezspecificity.frontend.staging.mmli1.ncsa.illinois.edu/home . Please note that while the full UI is being developed, this Jupyter Notebook will continue to serve as the temporary environment for performing EZSpecificity inference. If you have any questions, please do not hesitate to contact us. Thank you!