def convert_topology(src_filename, set_backbone=True, in_place=False, split_dir=None):
""" Converts the D.E.Shaw topology to Charm-usable force field with
TIP4P water molecules. THis re-sorts the atoms in the water solvent chains
(C2,C3) to group all 4 atoms as one contiguous residue. It also aliases the
ion charges 'pseu' atom to 'OM' for use with the standard tip4p.par parameter
file. The set_backbone option is to set constraint during initial minimzation"""
# Grab unit cell description (should be on first few lines:
cryst = None
with open(src_filename) as src:
for line in src.readlines():
if line.startswith('CRYST1'):
cryst = line
break
# Read in source PDB (DEShaw original format)
src_pdb = PdbStructure(open(src_filename))
atoms = list(src_pdb.iter_atoms())
topo = md.load(src_filename).top
# Break into 4 segments
segment_list = ['C1', 'C2', 'C3', 'C4']
segment = {l:[] for l in segment_list}
for i in atoms:
segment[i.segment_id].append(i)
# Set temperature factor (for gradual heating)
if set_backbone:
backbone = topo.select("backbone")
for i in range(0, len(segment['C1'])):
if i in backbone:
segment['C1'][i].location.temperature_factor = 1.0
# Resort water segements and alias "pseu" to OM (tip4p forcefield)
for wat in ['C2', 'C3']:
segment[wat] = sorted(segment[wat], key = lambda i: i.residue_number)
start_serial_num = min(segment[wat], key= lambda i: i.serial_number)
for i in range(0, len(segment[wat])):
newsn = i + start_serial_num.serial_number
segment[wat][i].serial_number = newsn
if segment[wat][i].get_name == 'pseu':
segment[wat][i].set_name_with_spaces(' OM ')
# FOR RE-RUNNING THE PSFGEN
if split_dir is not None:
for s in segment_list:
with open(split_dir + '/%s.pdb' % s, 'w') as dest:
for atom in segment[s]:
_=dest.write(str(atom) + '\n')
# Writeout new file
if in_place:
dest = open(src_filename, 'w')
if cryst is not None:
dest.write(cryst)
for s in segment_list:
for atom in segment[s]:
_=dest.write(str(atom) + '\n')
_=dest.write('END')
dest.close()
def reset_pdb(src_filename):
""" Updates PDB file with crytalline unit cell data and backbone temp
control to enable gradual heating"""
# Standard unit cell description (first lines)
cryst= 'CRYST1 51.263 51.263 51.263 90.00 90.00 90.00 P 1 1\n'
# Read in source PDB (DEShaw original format)
src_pdb = PdbStructure(open(src_filename))
atoms = list(src_pdb.iter_atoms())
topo = md.load(src_filename).top
# Set temperature factor (for gradual heating)
backbone = topo.select("backbone")
for i in range(0, len(atoms)):
if i in backbone:
atoms[i].location.temperature_factor = 1.0
# Write out to file
dest = open(src_filename, 'w')
if cryst is not None:
dest.write(cryst)
dest.write('REMARK Dynamically Generated from data driven controller (DDC)\n')
for atom in atoms:
_=dest.write(str(atom) + '\n')
_=dest.write('END')
dest.close()
def reset_pdb(src_filename):
""" Updates PDB file with crytalline unit cell data and backbone temp
control to enable gradual heating"""
# Standard unit cell description (first lines)
cryst = 'CRYST1 51.263 51.263 51.263 90.00 90.00 90.00 P 1 1\n'
# Read in source PDB (DEShaw original format)
src_pdb = PdbStructure(open(src_filename))
atoms = list(src_pdb.iter_atoms())
topo = md.load(src_filename).top
# Set temperature factor (for gradual heating)
backbone = topo.select("backbone")
for i in range(0, len(atoms)):
if i in backbone:
atoms[i].location.temperature_factor = 1.0
# Write out to file
dest = open(src_filename, 'w')
if cryst is not None:
dest.write(cryst)
dest.write(
'REMARK Dynamically Generated from data driven controller (DDC)\n')
for atom in atoms:
_ = dest.write(str(atom) + '\n')
_ = dest.write('END')
dest.close()
def test_load_multiframe():
with open(get_fn('multiframe.pdb')) as f:
pdb = PdbStructure(f)
yield lambda: eq(len(pdb.models), 2)
yield lambda: eq(len(pdb.models[0].chains), 1)
yield lambda: eq(len(pdb.models[0].chains[0].residues), 3)
yield lambda: eq(ilen(pdb.models[0].iter_atoms()), 22)
yield lambda: eq(len(pdb.models[1].chains), 1)
yield lambda: eq(len(pdb.models[1].chains[0].residues), 3)
yield lambda: eq(ilen(pdb.models[1].iter_atoms()), 22)
t = load(get_fn('multiframe.pdb'))
yield lambda: eq(t.n_frames, 2)
yield lambda: eq(t.n_atoms, 22)
yield lambda: eq(t.xyz[0], t.xyz[1])
def test_load_multiframe(get_fn):
with open(get_fn('multiframe.pdb')) as f:
pdb = PdbStructure(f)
assert eq(len(pdb.models), 2)
assert eq(len(pdb.models[0].chains), 1)
assert eq(len(pdb.models[0].chains[0].residues), 3)
assert eq(len(list(pdb.models[0].iter_atoms())), 22)
assert eq(len(pdb.models[1].chains), 1)
assert eq(len(pdb.models[1].chains[0].residues), 3)
assert eq(len(list(pdb.models[1].iter_atoms())), 22)
t = load(get_fn('multiframe.pdb'))
assert eq(t.n_frames, 2)
assert eq(t.n_atoms, 22)
assert eq(t.xyz[0], t.xyz[1])
def _read_models(self):
if not self._mode == 'r':
raise ValueError('file not opened for reading')
self._topology = Topology()
pdb = PdbStructure(self._file, load_all_models=True)
atomByNumber = {}
for chain in pdb.iter_chains():
c = self._topology.add_chain()
for residue in chain.iter_residues():
resName = residue.get_name()
if resName in PDBTrajectoryFile._residueNameReplacements:
resName = PDBTrajectoryFile._residueNameReplacements[resName]
r = self._topology.add_residue(resName, c, residue.number)
if resName in PDBTrajectoryFile._atomNameReplacements:
atomReplacements = PDBTrajectoryFile._atomNameReplacements[resName]
else:
atomReplacements = {}
for atom in residue.atoms:
atomName = atom.get_name()
if atomName in atomReplacements:
atomName = atomReplacements[atomName]
atomName = atomName.strip()
element = atom.element
if element is None:
element = self._guess_element(atomName, residue)
newAtom = self._topology.add_atom(atomName, element, r, serial=atom.serial_number)
atomByNumber[atom.serial_number] = newAtom
# load all of the positions (from every model)
_positions = []
for model in pdb.iter_models(use_all_models=True):
coords = []
for chain in model.iter_chains():
for residue in chain.iter_residues():
for atom in residue.atoms:
coords.append(atom.get_position())
_positions.append(coords)
if not all(len(f) == len(_positions[0]) for f in _positions):
raise ValueError('PDB Error: All MODELs must contain the same number of ATOMs')
self._positions = np.array(_positions)
## The atom positions read from the PDB file
self._unitcell_lengths = pdb.get_unit_cell_lengths()
self._unitcell_angles = pdb.get_unit_cell_angles()
self._topology.create_standard_bonds()
self._topology.create_disulfide_bonds(self.positions[0])
# Add bonds based on CONECT records.
connectBonds = []
for connect in pdb.models[0].connects:
i = connect[0]
for j in connect[1:]:
if i in atomByNumber and j in atomByNumber:
connectBonds.append((atomByNumber[i], atomByNumber[j]))
if len(connectBonds) > 0:
# Only add bonds that don't already exist.
existingBonds = set(self._topology.bonds)
for bond in connectBonds:
if bond not in existingBonds and (bond[1], bond[0]) not in existingBonds:
self._topology.add_bond(bond[0], bond[1])
existingBonds.add(bond)
def _read_models(self):
if not self._mode == 'r':
raise ValueError('file not opened for reading')
self._topology = Topology()
pdb = PdbStructure(self._file, load_all_models=True)
atomByNumber = {}
for chain in pdb.iter_chains():
c = self._topology.add_chain()
for residue in chain.iter_residues():
resName = residue.get_name()
if resName in PDBTrajectoryFile._residueNameReplacements:
resName = PDBTrajectoryFile._residueNameReplacements[resName]
r = self._topology.add_residue(resName, c, residue.number)
r.segment_id = residue.segment_id
if resName in PDBTrajectoryFile._atomNameReplacements:
atomReplacements = PDBTrajectoryFile._atomNameReplacements[resName]
else:
atomReplacements = {}
for atom in residue.atoms:
atomName = atom.get_name()
if atomName in atomReplacements:
atomName = atomReplacements[atomName]
atomName = atomName.strip()
element = atom.element
if element is None:
element = self._guess_element(atomName, residue)
newAtom = self._topology.add_atom(atomName, element, r, serial=atom.serial_number)
atomByNumber[atom.serial_number] = newAtom
# load all of the positions (from every model)
_positions = []
for model in pdb.iter_models(use_all_models=True):
coords = []
for chain in model.iter_chains():
for residue in chain.iter_residues():
for atom in residue.atoms:
coords.append(atom.get_position())
_positions.append(coords)
if not all(len(f) == len(_positions[0]) for f in _positions):
raise ValueError('PDB Error: All MODELs must contain the same number of ATOMs')
self._positions = np.array(_positions)
## The atom positions read from the PDB file
self._unitcell_lengths = pdb.get_unit_cell_lengths()
self._unitcell_angles = pdb.get_unit_cell_angles()
self._topology.create_standard_bonds()
self._topology.create_disulfide_bonds(self.positions[0])
# Add bonds based on CONECT records.
connectBonds = []
for connect in pdb.models[-1].connects:
i = connect[0]
for j in connect[1:]:
if i in atomByNumber and j in atomByNumber:
connectBonds.append((atomByNumber[i], atomByNumber[j]))
if len(connectBonds) > 0:
# Only add bonds that don't already exist.
existingBonds = set(self._topology.bonds)
for bond in connectBonds:
if bond not in existingBonds and (bond[1], bond[0]) not in existingBonds:
self._topology.add_bond(bond[0], bond[1])
existingBonds.add(bond)
def convert_topology(src_filename,
set_backbone=True,
in_place=False,
split_dir=None):
""" Converts the D.E.Shaw topology to Charm-usable force field with
TIP4P water molecules. THis re-sorts the atoms in the water solvent chains
(C2,C3) to group all 4 atoms as one contiguous residue. It also aliases the
ion charges 'pseu' atom to 'OM' for use with the standard tip4p.par parameter
file. The set_backbone option is to set constraint during initial minimzation"""
# Grab unit cell description (should be on first few lines:
cryst = None
with open(src_filename) as src:
for line in src.readlines():
if line.startswith('CRYST1'):
cryst = line
break
# Read in source PDB (DEShaw original format)
src_pdb = PdbStructure(open(src_filename))
atoms = list(src_pdb.iter_atoms())
topo = md.load(src_filename).top
# Break into 4 segments
segment_list = ['C1', 'C2', 'C3', 'C4']
segment = {l: [] for l in segment_list}
for i in atoms:
segment[i.segment_id].append(i)
# Set temperature factor (for gradual heating)
if set_backbone:
backbone = topo.select("backbone")
for i in range(0, len(segment['C1'])):
if i in backbone:
segment['C1'][i].location.temperature_factor = 1.0
# Resort water segements and alias "pseu" to OM (tip4p forcefield)
for wat in ['C2', 'C3']:
segment[wat] = sorted(segment[wat], key=lambda i: i.residue_number)
start_serial_num = min(segment[wat], key=lambda i: i.serial_number)
for i in range(0, len(segment[wat])):
newsn = i + start_serial_num.serial_number
segment[wat][i].serial_number = newsn
if segment[wat][i].get_name == 'pseu':
segment[wat][i].set_name_with_spaces(' OM ')
# FOR RE-RUNNING THE PSFGEN
if split_dir is not None:
for s in segment_list:
with open(split_dir + '/%s.pdb' % s, 'w') as dest:
for atom in segment[s]:
_ = dest.write(str(atom) + '\n')
# Writeout new file
if in_place:
dest = open(src_filename, 'w')
if cryst is not None:
dest.write(cryst)
for s in segment_list:
for atom in segment[s]:
_ = dest.write(str(atom) + '\n')
_ = dest.write('END')
dest.close()
