Binder name
Target chain IDs
Comma-separated chain IDs.
Hotspot residues (optional)
One "chain: residues" per line, e.g. "A: 54,56,58" or "A: 1-10". Leave blank to let AF2 pick a binding site.
Trim target around hotspots
Requires hotspot residues. Keeps local sequence segments around residues within the trim radius, reducing AlphaFold GPU-memory use without changing the uploaded PDB. Disconnected retained segments become separate chains and are recorded in a trim manifest.
Target trim radius, angstrom
A selected-chain residue is considered local context when any of its atoms is within this distance of a hotspot atom.
Target trim sequence padding
Additional residues retained before and after each spatially selected sequence segment to avoid cutting directly at the local structural context.
Minimum binder length
BindCraft's own default range is 65-150 for protein binders, 10-20 for peptides.
Maximum binder length
Accepted designs requested
Filter set
Default
Relaxed
No filters
Peptide
Peptide, relaxed
Omit amino acids
One-letter amino-acid codes ProteinMPNN redesign must not sample.
Use BigBang initialization
Helicity weight
Negative biases toward helical binders, positive toward non-helical, 0 is neutral.
Bias toward beta-sheet designs
Use ProteinMPNN-biased trajectory design
Allow a flexible target interface
Treat the target as a hard (rigid) target
Average pLDDT threshold
Average pTM threshold
Average interface pTM threshold
Average interface PAE threshold
Average surface hydrophobicity threshold
Minimum interface residues
Minimum interface H-bonds
Maximum hotspot RMSD, angstrom
Average binder pLDDT threshold
Maximum binder RMSD, angstrom
Maximum run time, minutes
Bounds this request's own subprocess; BindCraft itself has no built-in time limit.