Predict biomolecular structure and binding affinity with Boltz-2. Submit a YAML config or use the form-based batch workflow below.
Fine-tune recycling steps, sampling, MSA behavior, and output options. Defaults work well for most submissions.
Define protein, RNA, DNA, and ligand chains for structure prediction. Add one or more sequence blocks to build a complex.
Apply geometric constraints such as bonds, binding pockets, or residue contacts to guide the prediction.
Provide structural templates to guide prediction. Assign templates to specific chains and set distance constraints.
Request predicted binding affinity between a binder and target chains defined in Sequences.