Topology modules

[topoaa] module

The [topoaa] module is dedicated to the generation of CNS compatible parameters (.param) and topologies (.psf) for each of the input structures.

It will:

  • Detect missing atoms, including hydrogens
  • Re-build them when missing
  • Build and write out topologies (.psf) and coordinates (.pdb) files

This module is a prerequisite to run any downstream modules using CNS. Having access to parameters and topology is mandatory for any kind of EM/MD related tasks. Therefore this is the reason why the module [topoaa] is often used as first module in a workflow.

Note that for non-standard bio-molecules (apart from standard amino-acids, some modified ones, DNA, RNA, ions and carbohydrates ... see detailed list of supported molecules), such as small-molecules, parameters and topology must be obtained and provided by the user, as there is currently no built-in solution to generate them on the fly.

More information about [topoaa] parameters can be accessed here or retrieved by running:

haddock3-cfg -m topoaa

Here an example configuration file snapshot of a typical execution of the [topoaa] module in which a user specifies the protonation state of the histidine residues:

# Definition of 
run_dir = "example"
molecules = [
 "DNA_structure.pdb",
 "1abc.pdb",
]

[topoaa]
autohis = false
# Specify molecule 1 specific parameters
[topoaa.mol1]
5_phosphate = true
# Specify molecule 2 specific parameters
[topoaa.mol2]
nhisd = 1
hisd_1 = 76
nhise = 1
hise_1 = 15
charged_nter = true
charged_cter = false

# Workflow continues with other modules
# ...

The [topoaa.mol1] in square brackets is not as module, but allows to specify topoaa parameters for a given molecule. In this case (mol1), the parameters will be applied to the first molecule in the list of input molecules ("1abc.pdb").

Notable parameters

  • autohis: If set to false, you will need to specify the protonation states of histidines manually.
  • autotoppar: If set to true haddock will try to create automated topologies and paramters for unknown atoms using PRODRG.
  • ligand_param_fname and ligand_top_fname: Use these parameters to specify a custom ligand topology and parameters for your small molecules.
  • hydrogen_build: Use this parameter to control which hydrogen atoms will be (re)built (all (default) or only unknown). The unknown option might be useful for small ligands, or for very large systems in cases where the molecule already contains all hydrogens. For example, when it has already passed through HADDOCK.

Peptide cyclisation options

HADDOCK3 supports various cyclisation options:

  1. Through a peptide bond - defined at the molecule level (see below). Only checked if the cyclicpept parameter is set to true.
  2. Through an N-acetylated N-terminus to CYS bond detected using a minimum distance cutoff defined by the acecys_dist parameter (default 4Å). Only checked if the cyclicpept parameter is set to true
  3. Through disulphide bonds. These are automatically detected using a distance cutoff of 2.2Å, which can be changed with the disulphide_dist parameter in the [topoaa] module.

Parameters specific to each molecule

  • [topoaa.molX]: Allows the definition of specific topoaa parameters for molecule X.
    • nhisd, nhise allow to define the number of HISD, HISE in the molecule.
    • hisd_Y, hise_Y, allow to define which residue needs to be modified (e.g: hisd_1 = 3 means that we are defining the first HISD residue, and this residue has residue index of 3 in the file).
    • charged_nter, charged_cter allow to define the state of Nter and Cter residues. Note that chain breaks are not evaluated as termini residues.
    • 5_phosphate: Allows to define the state of the 5' end of nucleic acids sequences. If set to true, 5' end will be a phosphate group. Otherwise it will be an OH. (default false). Note that chain breaks are not evaluated as 5' ends.
    • cyclicpept: Allows to define that a molecule (peptide) is cyclized through a peptide bond. If set to true the module will check for the proximity of N- and C-ter and if within the cutoff distance defined by cyclicpept_dist it will create the peptide bond.

[topocg] module

The [topocg] module is dedicated to the mapping of the all-atom (AA) structures generated by the [topoaa] into their coarse-grained (CG) MARTINI2 representations and topologies. The associated CNS compatible topologies (.psf) are built for each of the input structures, as well as the necessary distance restraint tables for backmapping (*_cg_to_aa.tbl).

It will:

  • Build the CG structure from the AA, based on the MARTINI mapping (.pdb)
  • Build and write out topologies (.psf)
  • Build and write out distance restraint files (_cg_to_aa.tbl) necessary for mapping the CG structures back to AA.

The restraints created define atom-to-bead restraints, with one restraint per bead set to 0Å between the bead and the geometric center of its corresponding atoms (see the [cgtoaa] module for further details).

This module is a prerequisite when running CG docking. The topologies built are necessary for the subsequent sampling and refinement modules. Note that, with the current implementation, it is necessary to run this module right after the [topoaa] module.

Note that currently, apart from standard amino-acids, DNA, RNA, the module does not support non-standard biomolecules, such as small-molecules and carbohydrates. Those are ignored in CG mode, but will re-appear if present in the initial AA structure after the back-mapping stage ([cgtoaa] module) (provided parameters and topologies for the ligand have been provided or automatically generated).

More information about [topocg] parameters can be accessed here or retrieved by running:

haddock3-cfg -m topocg

Here an example configuration file snapshot of a typical execution of the [topocg] module:

# Definition of
run_dir = "example_CG"
molecules = [
 "1abc.pdb",
 "2abc.pdb",
]

[topoaa]

[topocg]
cgffversion = "martini2"

# Workflow continues with other modules
# ...

Notable parameters

The most important parameter for the [topocg] module is:

  • cgffversion: defines the force-field version for the CG mapping and parameters. Currently only MARTINI2 is supported.