HADDOCK3 Antibody-Antigen Docking

Information-driven antibody-antigen docking with HADDOCK3: prepare gevokizumab and IL-1β, build ambiguous restraints from the paratope and epitope, then rigid-body docking, refinement, clustering and CAPRI evaluation.

by Temple Compute

Run HADDOCK3 Antibody-Antigen Docking on Horus

W-24 · HADDOCK3 Protein-Protein Docking

Domain: BioExcel Building Blocks

Overview

Antibody-antigen docking with HADDOCK3, following the official tutorial. By default it docks gevokizumab (PDB 4G6K) against interleukin-1 beta (4I1B) and scores against the known complex (4G6M); the PDB codes, chains, domain ranges and paratope/epitope residues are parameters. The heavy and light chains are cleaned, trimmed to the variable domains and merged with biobb_pdb_tools; paratope/epitope residues become ambiguous restraints; docking then runs rigid-body sampling, flexible refinement, energy minimization and FCC clustering, with CAPRI evaluation at each stage. The workflow has 80 tasks.

Pipeline

fetch_antibody / fetch_antigen / fetch_complex    Download antibody, antigen and reference complex
ab_H_* / ab_L_*  -> ab_zip_HL -> ab_merge -> ab_reres -> ab_chain -> ab_chainxseg -> ab_tidy_final   Antibody: heavy/light chains trimmed to VH/VL, merged as chain A
ag_*                                              Antigen, renamed to chain B
cx_*                                              Reference complex, same preparation
write_ab_actpass -> passive_from_active -> actpass_to_ambig    Paratope/epitope active and passive residues, ambiguous restraints
restrain_bodies                                   Restraints tying the antibody heavy and light chains
topology -> rigid_body -> capri_eval_1 -> sele_top -> flex_ref -> capri_eval_2
  -> em_ref -> capri_eval_3 -> clust_fcc -> sele_top_clusts -> capri_eval_4 -> contact_map

Inputs / Outputs

Inputs are the three PDB codes downloaded from the PDBe (antibody, antigen, reference complex). The chain IDs, domain ranges and paratope/epitope residues must match the chosen entries; residue numbers refer to the prepared structures (the antibody is renumbered from 1 after merging). Outputs (results/):

antibody_clean.pdb, antigen_clean.pdb, reference_clean.pdb: prepared antibody, antigen and reference complex
antibody_actpass.txt, antigen_actpass.txt, ambig-paratope-NMR-epitope.tbl, antibody-unambig.tbl: residue lists and restraint tables
2_docking.zip, 5_flexref.zip, 7_emref.zip: models after rigid-body, flexible refinement and minimization
3_caprieval.zip, 6_caprieval2.zip, 8_caprieval3.zip, 11_caprieval4.zip: CAPRI scores per stage
4_selected.zip, 9_clustfcc.zip, 10_seletopclusts.zip: selections and clusters
12_contact_map.zip: contact maps of the final cluster representatives

Parameters

| id | Default | Meaning | |---|---|---| | antibody_pdb_id | 4G6K | PDB code of the antibody. | | antigen_pdb_id | 4I1B | PDB code of the antigen. | | reference_pdb_id | 4G6M | PDB code of the known complex used as the CAPRI reference. | | heavy_chain | H | Heavy-chain ID, in the antibody and the reference complex. | | light_chain | L | Light-chain ID, in the antibody and the reference complex. | | reference_antigen_chain | A | Antigen chain ID in the reference complex. | | vh_range | 1:120 | Heavy-chain variable-domain residues kept (pdb_selres syntax). | | vl_range | 1:107 | Light-chain variable-domain residues kept (pdb_selres syntax). | | paratope_residues | 31 32 ... 214 216 (26 residues) | Antibody active residues, space-separated, in the merged and renumbered antibody. | | epitope_residues | 72,73,...,116,117 (17 residues) | Antigen active residues, comma-separated; passive residues are derived around them. | | rigid_body_models | 10 | Models sampled by rigid-body docking (10-10000; production runs use 1000). | | refine_top_models | 8 | Best rigid-body models passed to refinement (1-1000); keep at or below rigid_body_models. | | models_per_cluster | 4 | Top models kept per FCC cluster (1-50). |