ProSSTHub: ProSST models in the SaprotHub and OPMC ecosystem

Welcome to ProSSTHub, a collaborative community for sharing and reusing ProSST models. ProSSTHub extends the SaprotHub and OPMC ecosystem with community-trained ProSST models.

SaprotHub, introduced in our Nature Biotechnology paper, makes protein language model training, prediction, storage, and sharing accessible to biologists without requiring advanced machine learning or coding expertise.

ColabProSST brings ProSST into this ecosystem, enabling users to train models, make predictions, and share their models directly with ProSSTHub.

ProSST is a structure-aware protein language model that combines amino-acid sequences with quantized structure tokens. See the ProSST paper and the official base models released by AI4Protein.

When using shared models, please cite the original publications where available. If you use ColabProSST, please cite the SaprotHub and OPMC paper and the ProSST paper.