Biocentral Paper released!

Our paper Biocentral: Embedding-based Protein Predictions was released and published in the Journal of Molecular Biology (JMB)! It is accompanied by the biocentral v1.0.0 release and the public availability of our service.

The paper highlights the internal of the biocentral ecosystem and services. Additionally, we provide two showcases for its abilities:

• Using the TMbed model for trans-membrane predictions for all virus sequences in the BFVD: The model is available as part of biocentral and our service was used to create predictions for the roundabout 350k sequences. In the paper, we further discuss the applicability of TMbed to the viral domain based on these predictions. We'd love to see domain experts apply the provided prediction models to their field of science! This allows for interesting insights, further investigation of their actual abilities and limitations, and improving model development in the future.
• Reproducing the ExoTox predictor: We show how the biocentral platform can be used to create powerful prediction models in a short period of time. Take a look at the reproduction code to learn how to use it yourself!

The biocentral ecosystem now consists of the following elements:

• biocentral IRE - The biocentral IntegratedResearchEnvironment (frontend application)
• biocentral API - The biocentral API for programmatic access to our service (Python/Dart)
• biocentral Service - The biocentral service, providing compute for the IRE and API.
• biocentral GitHub - The biocentral GitHub repositories, containing the free and open code for all applications.

Checkout the biocentral tour to find your perfect starting point, or dive right in with our quickstart guide.

Finally, we'd like to thank all our contributors and co-authors. We warmly say thank-you to the iteratec GmbH as well for supporting parts of the development. You can read more about our collaboration in their blogpost (german).