def reprepare(self):
"""Repeat the system preparation, after the user edited the .data table.
You should only modify the value of the .data.forced_protonation column on the basis of the values
in the .data.resid, .data.insertion, .data.chain attributes. Any other change will be ignored.
Returns
-------
mol_out : Molecule
the molecule titrated and optimized. The molecule object contains an additional attribute,
resData : ResidueData
a table of residues with the corresponding protonation states, pKas, and other information
Examples
--------
mol, prepData = proteinPrepare(Molecule("3PTB"), returnDetails=True)
d = prepData.data
d.loc[d.resid == 40, 'forced_protonation'] = 'HIP'
mHIP40, pHIP40 = prepData.reprepare()
"""
from pdb2pqr.src.hydrogens import hydrogenRoutines
from pdb2pqr.src.forcefield import Forcefield
from pdb2pqr.src.definitions import Definition
from htmd.builder.preparation import _buildResAndMol
d = self.data
routines = self.pdb2pqr_routines
p = routines.protein
keep_pka_columns = ('forced_protonation', 'buried', 'z', 'membraneExposed',
'pKa', 'pka_group_id', 'pka_residue_type', 'pka_type',
'pka_charge', 'pka_atom_name', 'pka_atom_sybyl_type')
copy_of_resname = d['resname']
copy_of_protonation = d['protonation']
copy_of_default_protonation = d['default_protonation']
list_of_forced_protonations = ~ pd.isnull(d['forced_protonation'])
neutraln = neutralc = False
assign_only = clean = False
debump = opt = True
# Some water molecules seem to have fixed=1 at this point. This makes initializeFullOptimization() skip them
# when it builds the resmap dict, which in turns breaks later stages. So, let's reset it.
for res in p.getResidues():
try:
if res.fixed == 1:
logger.debug("Resetting fixed flag on {:s}".format(str(res)))
res.fixed = 0
except:
logger.debug("Residue {:s} has no fixed attribute".format(str(res)))
pass
# Code lifted from resinter.py
routines.removeHydrogens()
for index, oldResidue in enumerate(p.getResidues()):
chain = p.chainmap[oldResidue.chainID]
chainIndex = chain.residues.index(oldResidue)
d_idx = d.pdb2pqr_idx == index
if sum(d_idx) != 1:
logger.warning("Residue {:s} appears {:d} times in data table".format(str(oldResidue), sum(d_idx)))
continue
newResidueName = d.forced_protonation[d_idx].iloc[0]
if pd.isnull(newResidueName):
# newResidueName = d.protonation[d_idx].iloc[0]
continue
logger.debug("Replacing {} with {}".format(oldResidue, newResidueName))
# Create the replacement residue
residueAtoms = oldResidue.atoms
newResidue = routines.protein.createResidue(residueAtoms, newResidueName)
# Make sure our names are cleaned up for output.
newResidue.renameResidue(newResidueName)
# Drop it in
p.residues[index] = newResidue
chain.residues[chainIndex] = newResidue
# Run the meaty bits of PDB2PQR
routines.setTermini(neutraln, neutralc)
routines.updateBonds()
if not clean and not assign_only:
routines.updateSSbridges()
if debump:
routines.debumpProtein()
routines.addHydrogens()
hydRoutines = hydrogenRoutines(routines)
if debump:
routines.debumpProtein()
if opt:
hydRoutines.setOptimizeableHydrogens()
hydRoutines.initializeFullOptimization()
hydRoutines.optimizeHydrogens()
else:
hydRoutines.initializeWaterOptimization()
hydRoutines.optimizeHydrogens()
# Special for GLH/ASH, since both conformations were added
hydRoutines.cleanup()
ff = "parse"
ffout = "amber"
usernames = userff = None
routines.setStates() # ?
mydef = Definition() # ?
myForcefield = Forcefield(ff, mydef, userff, usernames)
hitlist, misslist = routines.applyForcefield(myForcefield)
# reslist, charge = routines.getCharge() # <--- ?
# Copied from runPDB2PQR = ?
if not ffout is None:
if ffout != ff:
myNameScheme = Forcefield(ffout, mydef, userff)
else:
myNameScheme = myForcefield
routines.applyNameScheme(myNameScheme)
p.reSerialize()
newMol, newResData = _buildResAndMol(p)
# Assume that the number and order of residues does not change
# Carry over old pka and other useful info
newResData.data['resname'] = copy_of_resname
newResData.data['protonation'] = copy_of_protonation
newResData.data['default_protonation'] = copy_of_default_protonation
newResData.data.loc[list_of_forced_protonations, 'protonation'] = \
d.loc[list_of_forced_protonations, 'forced_protonation']
for cn in keep_pka_columns:
newResData.data[cn] = d[cn]
newResData.pdb2pqr_routines = routines
newResData.pdb2pqr_protein = routines.protein
newResData.missedLigands = self.missedLigands
return newMol, newResData
def reprepare(self):
"""Repeat the system preparation, after the user edited the .data table.
You should only modify the value of the .data.forced_protonation column on the basis of the values
in the .data.resid, .data.insertion, .data.chain attributes. Any other change will be ignored.
Returns
-------
mol_out : Molecule
the molecule titrated and optimized. The molecule object contains an additional attribute,
resData : ResidueData
a table of residues with the corresponding protonation states, pKas, and other information
Examples
--------
mol, prepData = proteinPrepare(Molecule("3PTB"), returnDetails=True)
d = prepData.data
d.loc[d.resid == 40, 'forced_protonation'] = 'HIP'
mHIP40, pHIP40 = prepData.reprepare()
"""
from pdb2pqr.src.hydrogens import hydrogenRoutines
from pdb2pqr.src.forcefield import Forcefield
from pdb2pqr.src.definitions import Definition
from htmd.builder.preparation import _buildResAndMol
d = self.data
routines = self.pdb2pqr_routines
p = routines.protein
keep_pka_columns = ('forced_protonation', 'buried', 'z', 'membraneExposed',
'pKa', 'pka_group_id', 'pka_residue_type', 'pka_type',
'pka_charge', 'pka_atom_name', 'pka_atom_sybyl_type')
copy_of_resname = d['resname']
copy_of_protonation = d['protonation']
copy_of_default_protonation = d['default_protonation']
list_of_forced_protonations = ~ pd.isnull(d['forced_protonation'])
neutraln = neutralc = False
assign_only = clean = False
debump = opt = True
# Some water molecules seem to have fixed=1 at this point. This makes initializeFullOptimization() skip them
# when it builds the resmap dict, which in turns breaks later stages. So, let's reset it.
for res in p.getResidues():
try:
if res.fixed == 1:
logger.debug("Resetting fixed flag on {:s}".format(str(res)))
res.fixed = 0
except:
logger.debug("Residue {:s} has no fixed attribute".format(str(res)))
pass
# Code lifted from resinter.py
routines.removeHydrogens()
for index, oldResidue in enumerate(p.getResidues()):
chain = p.chainmap[oldResidue.chainID]
chainIndex = chain.residues.index(oldResidue)
d_idx = d.pdb2pqr_idx == index
if sum(d_idx) != 1:
logger.warning("Residue {:s} appears {:d} times in data table".format(str(oldResidue), sum(d_idx)))
continue
newResidueName = d.forced_protonation[d_idx].iloc[0]
if pd.isnull(newResidueName):
# newResidueName = d.protonation[d_idx].iloc[0]
continue
logger.debug("Replacing {} with {}".format(oldResidue, newResidueName))
# Create the replacement residue
residueAtoms = oldResidue.atoms
newResidue = routines.protein.createResidue(residueAtoms, newResidueName)
# Make sure our names are cleaned up for output.
newResidue.renameResidue(newResidueName)
# Drop it in
p.residues[index] = newResidue
chain.residues[chainIndex] = newResidue
# Run the meaty bits of PDB2PQR
routines.setTermini(neutraln, neutralc)
routines.updateBonds()
if not clean and not assign_only:
routines.updateSSbridges()
if debump:
routines.debumpProtein()
routines.addHydrogens()
hydRoutines = hydrogenRoutines(routines)
if debump:
routines.debumpProtein()
if opt:
hydRoutines.setOptimizeableHydrogens()
hydRoutines.initializeFullOptimization()
hydRoutines.optimizeHydrogens()
else:
hydRoutines.initializeWaterOptimization()
hydRoutines.optimizeHydrogens()
# Special for GLH/ASH, since both conformations were added
hydRoutines.cleanup()
ff = "parse"
ffout = "amber"
usernames = userff = None
routines.setStates() # ?
mydef = Definition() # ?
myForcefield = Forcefield(ff, mydef, userff, usernames)
hitlist, misslist = routines.applyForcefield(myForcefield)
# reslist, charge = routines.getCharge() # <--- ?
# Copied from runPDB2PQR = ?
if not ffout is None:
if ffout != ff:
myNameScheme = Forcefield(ffout, mydef, userff)
else:
myNameScheme = myForcefield
routines.applyNameScheme(myNameScheme)
p.reSerialize()
newMol, newResData = _buildResAndMol(p)
# Assume that the number and order of residues does not change
# Carry over old pka and other useful info
newResData.data['resname'] = copy_of_resname
newResData.data['protonation'] = copy_of_protonation
newResData.data['default_protonation'] = copy_of_default_protonation
newResData.data.loc[list_of_forced_protonations, 'protonation'] = \
d.loc[list_of_forced_protonations, 'forced_protonation']
for cn in keep_pka_columns:
newResData.data[cn] = d[cn]
newResData.pdb2pqr_routines = routines
newResData.pdb2pqr_protein = routines.protein
newResData.missedLigands = self.missedLigands
return newMol, newResData