AutoDock CrankPep (ADCP)

Peptide docking that combines folding and docking

AutoDock CrankPep (ADCP) is a docking engine specialized for peptides. It combines a protein folding technique with AutoDock’s efficient representation of a rigid receptor as affinity grids. A Monte Carlo search folds the peptide while it optimizes the peptide’s interaction with the receptor, which yields docked peptides. ADCP takes a peptide as a sequence string, or as a 3D structure in a PDB file (-i), and has docked peptides of up to 20 amino acids successfully.

Two superimposed helical peptide poses in a receptor pocket

ADCP example, PDB 5GRD

ADCP example, PDB 3AV9

ADCP example, PDB 1N7F

ADCP example, PDB 4AK4

Crankshaft move rotating a cyclic peptide about the axis between two C-alpha atoms

Versions

Version Status Platforms Install
1.1 current Linux (macOS through an Ubuntu virtual machine) micromamba or conda, Download
1.0 deprecated Linux, macOS (Windows through WSL) ADFRsuite 1.0 installer, Download

What changed in version 1.1

Version 1.1 adds support for non-standard amino acids and lets you minimize docked poses with OpenMM. In detail:

  1. Streamlined output. Docked poses are saved in a single multi-model PDB file, xxx_out.pdb, where xxx is the job name (or a default name), and a summary file xxx_summary.dlg is written.
  2. Full side chains. Docked poses contain all-atom side chains instead of a rotamer index.
  3. Number of poses. Set how many docked poses to write with -m / --nmodes.
  4. Command line changes. The target file is given with -T and can be a .trg file or a folder containing one. The new option -k keeps calculation details.
  5. Rotamer libraries. Side-chain rotamers are loaded from .lib text files; system rotamer libraries are selected with -L.
  6. Non-standard amino acids. Rotamers for modified side chains can be added without recompiling. The syntax <FOO> names the side chain used for the full-atom representation, preceded by the letter of the coarse potential.
  7. D-amino acids. Write & before the coarse potential letter.
  8. Histidine protonation states. HIE and HID are supported.
  9. Automatic affinity maps. ADCP loads affinity maps for every atom type detected in the peptide.
  10. MD refinement. Docked poses can be minimized with OpenMM, with options to control it.

Source code

The source code is on GitHub: sannerlab/ADCP. The source of ADCP as published in 2019 is archived at sannerlab/ADCP-2019.

License

ADCP is distributed under the Sannerlab MIT-like Source-Available License. See License.

How to cite

If you use ADCP in your work, please cite:

Back to top