Computational design of protein binders targeting RBX1 using Proteina-Complexa and Germinal pipelines
Flow matching generative model produces backbone structures conditioned on the RBX1 target. 200 samples generated with AF2 reward scoring.
200 backbonesInverse folding model designs sequences for top backbones. 8 sequences sampled per backbone.
80 sequencesPredict complex structures with AlphaFold 3. Evaluate via ipTM, pLDDT, iPAE, and ranking score.
9 validatedAlphaFold-Multimer hallucination generates VHH nanobody backbones and sequences targeting RBX1 hotspot residues.
Hallucinated nanobodiesAntibody-specific ProteinMPNN redesigns CDR loops and framework residues for improved stability and binding.
5 variants per seedIndependent structure prediction with Chai-1 cofold. Evaluate binding via ipTM, pLDDT, iPAE, and hotspot contacts.
20 validatedClick any column header to sort. Binder_4 (highlighted) shows the best AF3 metrics overall.
| Rank | Binder Name | Length | Complexa Reward | Complexa ipTM | Complexa pLDDT | MPNN Score | AF3 ipTM | AF3 pTM | AF3 pLDDT | AF3 iPAE | AF3 Ranking | Actions |
|---|
Click any column header to sort. Top candidate (highlighted) shows the best Chai-1 cofold ipTM. All designs are VHH nanobodies (~130 aa).
| Rank | Name | Length | Halluc ipTM | Halluc pLDDT | Cofold ipTM | Cofold pLDDT | Cofold iPAE | Hotspot Contacts | Actions |
|---|