Structure Prediction
Use State of The Art Biomolecular Interaction Prediction Tools on Tamarind Bio
Webserver: https://app.tamarind.bio/boltz
API: https://app.tamarind.bio/api-docs
With DeepMind releasing AlphaFold3, the field gained a new state of the art model for bimolecular interaction prediction. AF3 leads every other tool in antibodies, protein(-ligand) complexes, covalent modifications, and nucleic acids. However, even though DeepMind recently released code for the model, the weights required to use it for any practical drug discovery or protein engineering work are behind an application, only available for academic use.
With this, teams from MIT, and Chai Discovery have launched alternatives to the original, reaching or sometimes beating the performance of AlphaFold3, with no restrictions on commercial use.
Today, we launch our web interface, API, and high throughput interface for AlphaFold3. Try out the web interfaces today: Boltz-1, Chai-1.
Get in touch to learn more on our secure, confidential cloud offering. We also support a programmatic interface, Boltz-1 finetuning, and running tools at very high throughput (have had 1M+ sequences before!). You own all of your data, inputs and derivatives!
We already work with dozens of biotech and large pharmaceutical companies, supporting diverse work in antibody therapeutics, enzyme engineering, protein design and many others.
Chai-1 Residue Distance Restrictions & Epitope Specification
Chai-1's capabilities extend beyond its base sequence modeling. Users can specify epitopes by identifying maximum distances between two residues in separate chains, for example those retrieved from experimental methods. The authors find that this can improve docking prediction by double digit percentages. Most notably, they find that this epitope limitation doubles antibody-antigen structure prediction accuracy.
Performance
