Superbio’s binder design pipeline streamlines de novo protein binder creation by integrating three state-of-the-art AI tools:

• RFDiffusion: Generates diverse protein binder backbones using diffusion models, based on your target structure (provided as PDB) and defined binding hotspots.
• ProteinMPNN: Designs amino acid sequences that are predicted to fold into the backbones generated by RFDiffusion, optimizing for stability and binding.
• Boltz-1: Co-folds both binder and target protein chains, assigning pLDDT scores and helping to select the most promising candidates.

This end-to-end pipeline lets you go from a target protein to a list of potential binder candidates in a single step.

This pipeline is complemented by the following tools:

PeSTo-PPIs: Identifying potential epitope regions on target proteins. Especially useful for poorly characterized targets.

AlphaFold2: Predicting PDB structure for potential target proteins / antigens for which a good crystal structure is not readily available.

Example use case:

Designing novel protein binders targeting specific protein structures for therapeutic development or biochemical research.

Technology:

Boltz-1, RFDiffusion, ProteinMPNN