Compute functionality for biocentral. Provided via a dockerized FastAPI server.
Learn how to best use the biocentral_server functionality for your use case by taking the tour!
Or check out the API or Frontend directly.
These are the currently provided modules:
| Module | Features | External Dependencies |
|---|---|---|
| embeddings | Embed protein sequences | biotrainer |
| ppi | Work with protein-protein interactions | hvi_toolkit |
| predict | Create predictions from pre-trained models | TMbed, VespaG |
| custom_models | Train and evaluate models trained on embeddings | biotrainer |
| proteins | Work with protein data | taxoniq |
See the docs on how to get started with biocentral_server.
For development, see the development guide.
Please cite our paper if you are using biocentral in your work:
@Article{Franz2026,
author = {Franz, Sebastian and Olenyi, Tobias and Schloetermann, Paula and Smaoui, Amine and Jimenez-Soto, Luisa F. and Rost, Burkhard},
journal = {Journal of Molecular Biology},
title = {biocentral: embedding-based protein predictions},
year = {2026},
issn = {0022-2836},
month = jan,
pages = {169673},
doi = {10.1016/j.jmb.2026.169673},
publisher = {Elsevier BV},
}