Error encountered with DNA scoring function in lightdock3.py for DNA-DNA docking

[lightdock]

Error encountered with DNA scoring function in lightdock3.py for DNA-DNA docking

[alchemistcai]

Clear the tmpfiles generated by lightdock3_setup.py and try this conversion for AMBER94 atom type compatibility by conda and Ambertools:

install ambertools
convert atom types

conda install -c conda-forge ambertools
pdb4amber --reduce xxx.pdb > xxx.amber.pdb

Then use lightdock3_setup.py again.

See also #45 ,a similar problem.

@MarielEsquivel98

[brianjimenez]

Hi @MarielEsquivel98 , thanks for using LightDock.

First, we have never tested DNA-DNA docking, we are curious about this case scenario, but not sure if the dna scoring function will handle properly DNA-DNA since it was designed specifically for protein-DNA.

Concerning the issue with atom types, @alchemistcai pointed out a few different approaches. Another workaround would be to prepare structures as stated in the tutorial with reduce and then applying the reduce_to_amber.py script:

reduce -Trim dna.pdb > dna_noh.pdb
reduce -BUILD dna_noh.pdb > dna_h.pdb

reduce_to_amber.py:

#!/usr/bin/env python3

import os
import argparse
from lightdock.scoring.dna.data.amber import atoms_per_residue
from lightdock.pdbutil.PDBIO import read_atom_line


def _format_atom_name(atom_name):
    """Format ATOM name with correct padding"""
    if len(atom_name) == 4:
        return atom_name
    else:
        return " %s" % atom_name


def write_atom_line(atom, output):
    """Writes a PDB file format line to output."""
    if atom.__class__.__name__ == "HetAtom":
        atom_type = "HETATM"
    else:
        atom_type = "ATOM  "
    line = "%6s%5d %-4s%-1s%3s%2s%4d%1s   %8.3f%8.3f%8.3f%6.2f%6.2f%12s\n" % (
        atom_type,
        atom.number,
        _format_atom_name(atom.name),
        atom.alternative,
        atom.residue_name,
        atom.chain_id,
        atom.residue_number,
        atom.residue_insertion,
        atom.x,
        atom.y,
        atom.z,
        atom.occupancy,
        atom.b_factor,
        atom.element,
    )
    output.write(line)


def translate_rna(residue_name):
    if residue_name in ['U', 'G', 'A', 'C']:
        residue_name = f'R{residue_name}'
    return residue_name


translation = {"H5'":"H5'1", "H5''":"H5'2", "H2'":"H2'1", "H2''":"H2'2", "OP1":"O1P", "OP2":"O2P"}


if __name__ == "__main__":

    parser = argparse.ArgumentParser()
    parser.add_argument("input_pdb_file")
    parser.add_argument("output_pdb_file")
    args = parser.parse_args()

    with open(args.input_pdb_file) as ih:
        with open(args.output_pdb_file, 'w') as oh:
            for line in ih:
                line = line.rstrip(os.linesep)
                if line.startswith("ATOM  "):
                    atom = read_atom_line(line)
                    atom.residue_name = translate_rna(atom.residue_name)
                    if atom.residue_name not in atoms_per_residue:
                        print(f"[Warning] Not supported atom: {atom.residue_name}.{atom.name}")
                    else:
                        if atom.name not in atoms_per_residue[atom.residue_name]:
                            try:
                                atom.name = translation[atom.name]
                                write_atom_line(atom, oh)
                            except KeyError:
                                print(f"[Warning] Atom not found in mapping: {atom.residue_name}.{atom.name}")
                        else:
                            write_atom_line(atom, oh)

python3 reduce_to_amber.py dna_h.pdb > dna_fixed.pdb

Maybe it would be a good idea also to test the system on the LightDock server, many fixes will be performed automatically for you: https://server.lightdock.org/

[brianjimenez]

Closing issue since there is no feedback. Please feel free to reopen if needed.