def enumerateDihs(f): #openbabel
f_name, f_ext = os.path.splitext(f)
obconversion = ob.OBConversion()
obconversion.SetInFormat(f_ext)
obmol = ob.OBMol()
this = obconversion.ReadFile(obmol, f)
all_dihs = [i for i in ob.OBMolTorsionIter(obmol)]
return all_dihs
def _add_oop_angle_change_collection(self,
index,
threshold=0,
CURVE_TYPE=4,
zorder=50,
**kwargs):
"""Compute and draw torsion changes on the axes."""
CURVE_TYPE_3, CURVE_TYPE_4 = 3, 4
col = []
edgecolors = []
if CURVE_TYPE is CURVE_TYPE_3:
codes = [
Path.MOVETO,
Path.CURVE3,
Path.CURVE3,
]
elif CURVE_TYPE is CURVE_TYPE_4:
codes = [
Path.MOVETO,
Path.CURVE4,
Path.CURVE4,
Path.CURVE4,
]
for torsion in ob.OBMolTorsionIter(self.molecule):
atoms = [self.molecule.GetAtom(idx + 1) for idx in torsion]
atomsnc = [
self._to_normal_coordinates(atom, index) for atom in atoms
]
teq = self.molecule.GetTorsion(*atoms)
tnc = self.molecule.GetTorsion(*atomsnc)
amplitude = (tnc - teq + 360.0) % 360.0
if amplitude > 180.0:
amplitude -= 360.0
if abs(amplitude) <= threshold: continue
intensity = abs(amplitude / 40)
a, b, c, d = [self._2Dcoords(atom) for atom in atoms]
p2 = 0.5 * (b + c) # middle
p1 = intensity * (a - p2) + p2
p3 = intensity * (d - p2) + p2
color = self.oop_colors[0 if amplitude < 0.0 else 1]
if CURVE_TYPE is CURVE_TYPE_3:
verts = [p1, p2, p3]
elif CURVE_TYPE is CURVE_TYPE_4:
verts = [p1, b, c, p3]
curve = Path(verts, codes)
col.append(curve)
edgecolors.append(color)
kw = {'edgecolors': edgecolors, 'facecolors': 'none'}
kwargs.update(kw)
self._vib_oop = PathCollection(col, zorder=zorder, **kwargs)
self.axes.add_collection(self._vib_oop)
def get_topology(molecule):
"""
Return the molecular topology obtained via openabel
Args:
molecule (Molecule): pymatgen Molecule object
Returns:
atoms (list): list of atoms and their force field mapping
bonds (list): [[i,j, bond_type], ...] where bond_type is
a sorted tuple of the force field names of atoms i and j.
angles (list): [[i,j, k, angle_type], ...] where angle_type is
a sorted tuple of the force field names of atoms i, j and k.
dihedrals (list): [[i,j, k, l, dihedral_type], ...] where
dihedral_type is a sorted tuple of the force field names of
atoms i, j, k and l.
"""
bma = BabelMolAdaptor(molecule)
obmol = bma.openbabel_mol
#print obmol.NumAtoms(), obmol.NumBonds()
atoms = [x.GetIdx() - 1 for x in openbabel.OBMolAtomIter(obmol)]
bonds = [[x.GetBeginAtomIdx() - 1,
x.GetEndAtomIdx() - 1] for x in openbabel.OBMolBondIter(obmol)]
angles = [list(x) for x in openbabel.OBMolAngleIter(obmol)]
dihedrals = [list(x) for x in openbabel.OBMolTorsionIter(obmol)]
#print len(atoms), len(bonds), len(angles), len(dihedrals)
atoms = [
tuple([str(molecule[x].specie), molecule[x].ff_map]) for x in atoms
]
bonds = [
x + [tuple((molecule[x[0]].ff_map, molecule[x[1]].ff_map))]
for x in bonds
]
angles = [
x + [
tuple(((molecule[x[0]].ff_map, molecule[x[1]].ff_map,
molecule[x[2]].ff_map)))
] for x in angles
]
dihedrals = [
x + [
tuple((molecule[x[0]].ff_map, molecule[x[1]].ff_map,
molecule[x[2]].ff_map, molecule[x[3]].ff_map))
] for x in dihedrals
]
return atoms, bonds, angles, dihedrals
def read_parameters(dfrfile, txtfile):
# read the degrees of freedom from dfr into dictionaries
dfrBonds, dfrAngles, dfrDihedrals, dfrImpDih = read_dfr_dof(dfrfile)
# read the txt to a pybel mol object
mol, q, eps, sig = read_txt_to_mol(txtfile)
# check if all bonds are present in the .dfr
bondIterator = openbabel.OBMolBondIter(mol.OBMol)
for bond in bondIterator:
lbl1 = str(bond.GetBeginAtom().GetId() +
1) + " " + str(bond.GetEndAtom().GetId() + 1)
lbl2 = str(bond.GetEndAtom().GetId() +
1) + " " + str(bond.GetBeginAtom().GetId() + 1)
if (lbl1 not in dfrBonds) and (lbl2 not in dfrBonds):
print("The bond (%s) is not specified in your .dfr. Aborting..." %
lbl1)
sys.exit(0)
# check if all angles are present in the .dfr
angleIterator = openbabel.OBMolAngleIter(mol.OBMol)
for angle in angleIterator:
angidx = [str(x + 1) for x in angle]
lbl1 = angidx[1] + " " + angidx[0] + " " + angidx[2]
lbl2 = angidx[2] + " " + angidx[0] + " " + angidx[1]
if (lbl1 not in dfrAngles) and (lbl2 not in dfrAngles):
print("The angle (%s) is not specified in your .dfr. Aborting..." %
lbl1)
sys.exit(0)
# check if all dihedrals are present in the .dfr
torsionIterator = openbabel.OBMolTorsionIter(mol.OBMol)
for torsional in torsionIterator:
torsidx = [str(x + 1) for x in torsional]
lbl1 = torsidx[0] + " " + torsidx[1] + " " + torsidx[
2] + " " + torsidx[3]
lbl2 = torsidx[3] + " " + torsidx[2] + " " + torsidx[
1] + " " + torsidx[0]
if (lbl1 not in dfrDihedrals) and (lbl2 not in dfrDihedrals):
print(
"The dihedral (%s) is not specified in your .dfr. Aborting..."
% lbl1)
sys.exit(0)
return mol, q, eps, sig, dfrBonds, dfrAngles, dfrDihedrals, dfrImpDih
def generate_fragfile(filename, outtype, ffparams=None, eqgeom=False):
# check outtype
if outtype not in ["flex", "header", "min"]:
sys.exit(
'Invalid argument indicating verbosity of .dfr (%s). Use "flex", "header" or "min".'
% outtype)
# get basename and file extension
base, ext = os.path.splitext(filename)
# set openbabel file format
obConversion = openbabel.OBConversion()
obConversion.SetInAndOutFormats(ext[1:], "xyz")
# read molecule to OBMol object
mol = openbabel.OBMol()
obConversion.ReadFile(mol, filename)
if ffparams:
# get atomic labels from pdb
idToAtomicLabel = {}
for res in openbabel.OBResidueIter(mol):
for atom in openbabel.OBResidueAtomIter(res):
idToAtomicLabel[atom.GetId()] = res.GetAtomID(atom).strip()
# read force field parameters and store into dictionaries
labelToSLabel = {}
charges = {}
epsilons = {}
sigmas = {}
bonds = {}
angles = {}
dihedrals = {}
impropers = {}
with open(ffparams, 'r') as f:
line = f.readline()
# read nb params
while "$bond" not in line:
if line.strip().startswith("#") or not line.strip():
line = f.readline()
continue
lbl = line.split()[0]
charges[lbl] = line.split()[1]
epsilons[lbl] = line.split()[2]
sigmas[lbl] = line.split()[3]
labelToSLabel[lbl] = line.split()[4]
line = f.readline()
# read bond params
line = f.readline()
while "$angle" not in line:
if line.strip().startswith(
"#") or "$end" in line or not line.strip():
line = f.readline()
continue
line = line.replace("–", "-")
# store the constants for the order of the input and the inverse order
consts = "\t".join(line.split()[1:])
bonds[line.split()[0]] = consts
bonds["-".join(line.split()[0].split("-")[::-1])] = consts
line = f.readline()
# read angle params
line = f.readline()
while "$dihedral" not in line:
if line.strip().startswith(
"#") or "$end" in line or not line.strip():
line = f.readline()
continue
line = line.replace("–", "-")
# store the constants for the order of the input and the inverse order
consts = "\t".join(line.split()[1:])
angles[line.split()[0]] = consts
angles["-".join(line.split()[0].split("-")[::-1])] = consts
line = f.readline()
# read dihedrals
line = f.readline()
while "$improper" not in line:
if line.strip().startswith(
"#") or "$end" in line or not line.strip():
line = f.readline()
continue
line = line.replace("–", "-")
# store the constants for the order of the input and the inverse order
consts = "\t".join(line.split()[1:])
dihedrals[line.split()[0]] = consts
dihedrals["-".join(line.split()[0].split("-")[::-1])] = consts
line = f.readline()
# read impropers
line = f.readline()
while line:
if line.strip().startswith(
"#") or "$end" in line or not line.strip():
line = f.readline()
continue
line = line.replace("–", "-")
# store the constants for the order of the input and the inverse order
consts = "\t".join(line.split()[1:])
impropers[line.split()[0]] = consts
impropers["-".join(line.split()[0].split("-")[::-1])] = consts
line = f.readline()
# check if there are unused labels
for lbl in charges.keys():
fnd = False
for i in idToAtomicLabel:
if lbl == idToAtomicLabel[i]:
fnd = True
break
if not fnd:
print(
"!!! WARNING: There are unused atoms in your parameter file (%s) !!!"
% lbl)
# split the molecule
fragments, fragConnection, dummyToAtom = split_mol_fragments_daylight(mol)
# dummy atoms ids
dummyAtoms = dummyToAtom.keys()
# write molecule to .txt file (passed as ljname to DICE)
with open(base + ".txt", "w") as f:
f.write("*\n1\n")
atomToPrint = []
for frag in fragments:
fragAtomIterator = openbabel.OBMolAtomIter(frag)
for atom in fragAtomIterator:
if atom.GetId() not in dummyAtoms:
atomToPrint.append(atom)
# print number of atoms
f.write(
str(len(atomToPrint)) +
" \t %s (generated with fragGen)\n" % os.path.basename(base))
# dictionary associating Atomic number with rdf label
rdfs = {}
rdf_label = 1
if ffparams:
# sort atoms by index and print (this prints the atoms, e.g., in the same order of the xyz input)
for atom in sorted(atomToPrint, key=lambda atom: atom.GetId()):
if atom.GetAtomicNum() not in rdfs.keys():
rdfs[atom.GetAtomicNum()] = str(rdf_label)
rdf_label += 1
f.write(rdfs[atom.GetAtomicNum()] + " " +
str(atom.GetAtomicNum()) + " \t" + str(atom.GetX()) +
" \t" + str(atom.GetY()) + " \t" +
str(atom.GetZ()) + " \t" +
charges[idToAtomicLabel[atom.GetId()]] + "\t" +
epsilons[idToAtomicLabel[atom.GetId()]] + "\t" +
sigmas[idToAtomicLabel[atom.GetId()]] + "\n")
f.write("$end\n")
else:
# sort atoms by index and print (this prints the atoms, e.g., in the same order of the xyz input)
for atom in sorted(atomToPrint, key=lambda atom: atom.GetId()):
if atom.GetAtomicNum() not in rdfs.keys():
rdfs[atom.GetAtomicNum()] = str(rdf_label)
rdf_label += 1
f.write(rdfs[atom.GetAtomicNum()] + " " +
str(atom.GetAtomicNum()) + " \t" + str(atom.GetX()) +
" \t" + str(atom.GetY()) + " \t" +
str(atom.GetZ()) + " \t" + "q" + "\t" +
"epsilon" + "\t" + "sigma\n")
f.write("$end\n")
# write info to dfr file
with open(base + ".dfr", "w") as f:
# fragments and fragments connections are printed to every outtype
f.write("$atoms fragments\n")
fragslst = []
for frag in fragments:
f.write(frag.GetTitle() + "\t[ ")
fragAtomIterator = openbabel.OBMolAtomIter(frag)
atomlst = [
str(dummyToAtom[x.GetId()] +
1) if x.GetId() in dummyAtoms else str(x.GetId() + 1)
for x in fragAtomIterator
]
fragslst.append(atomlst)
for atom in atomlst:
f.write(atom + "\t")
if outtype == "min" or outtype == "header":
f.write("] R\n")
else:
f.write("] F\n")
f.write("$end atoms fragments\n")
f.write("\n$fragment connection\n")
for frag1, frag2 in fragConnection:
f.write(frag1 + "\t" + frag2 + "\n")
f.write("$end fragment connection\n")
# bonds are printed to every outtype that is not header, since we need a connection matrix
if outtype != "header":
f.write("\n$bond\n")
bondIterator = openbabel.OBMolBondIter(mol)
if ffparams:
for bond in bondIterator:
try:
if eqgeom:
f.write(
str(bond.GetBeginAtom().GetId() + 1) + " " +
str(bond.GetEndAtom().GetId() + 1) + " \t" +
bonds[labelToSLabel[idToAtomicLabel[
bond.GetBeginAtom().GetId()]] + "-" +
labelToSLabel[idToAtomicLabel[
bond.GetEndAtom().GetId()]]].split()
[0] + "\t" + str("%.6f" % bond.GetLength()) +
"\n")
else:
f.write(
str(bond.GetBeginAtom().GetId() + 1) + " " +
str(bond.GetEndAtom().GetId() + 1) + " \t" +
bonds[labelToSLabel[idToAtomicLabel[
bond.GetBeginAtom().GetId()]] + "-" +
labelToSLabel[idToAtomicLabel[
bond.GetEndAtom().GetId()]]] + "\n")
except KeyError as e:
print(
"The parameters for atoms %d %d (%s) was not found in the bonds list\n"
% (bond.GetBeginAtom().GetId() + 1,
bond.GetEndAtom().GetId() + 1, e))
raise
else:
for bond in bondIterator:
f.write(
str(bond.GetBeginAtom().GetId() + 1) + " " +
str(bond.GetEndAtom().GetId() + 1) + " \t0.0\t" +
str("%.6f" % bond.GetLength()) + "\n")
f.write("$end bond\n")
# angles are only printed for outtype flex
if outtype == "flex":
f.write("\n$angle\n")
angleIterator = openbabel.OBMolAngleIter(mol)
if ffparams:
for angle in angleIterator:
try:
if eqgeom:
atom2 = mol.GetAtomById(angle[0])
atom1 = mol.GetAtomById(angle[1])
atom3 = mol.GetAtomById(angle[2])
aparams = angles[
labelToSLabel[idToAtomicLabel[angle[1]]] +
"-" +
labelToSLabel[idToAtomicLabel[angle[0]]] +
"-" + labelToSLabel[idToAtomicLabel[
angle[2]]]].split()
f.write(
str(angle[1] + 1) + " " + str(angle[0] + 1) +
" " + str(angle[2] + 1) + " \t" +
aparams[0] + "\t" + aparams[1] + "\t" +
str("%.6f" %
mol.GetAngle(atom1, atom2, atom3)) + "\n")
else:
f.write(
str(angle[1] + 1) + " " + str(angle[0] + 1) +
" " + str(angle[2] + 1) + " \t" + angles[
labelToSLabel[idToAtomicLabel[angle[1]]] +
"-" +
labelToSLabel[idToAtomicLabel[angle[0]]] +
"-" +
labelToSLabel[idToAtomicLabel[angle[2]]]] +
"\n")
except KeyError as e:
print(
"The parameters for atoms %d %d %d (%s) was not found in the angles list\n"
% (angle[1] + 1, angle[0] + 1, angle[2] + 1, e))
raise
else:
for angle in angleIterator:
# carefully select the atoms to find the angle
atom2 = mol.GetAtomById(angle[0])
atom1 = mol.GetAtomById(angle[1])
atom3 = mol.GetAtomById(angle[2])
f.write(
str(angle[1] + 1) + " " + str(angle[0] + 1) + " " +
str(angle[2] + 1) + " \tharmonic\tK\t" +
str("%.6f" % mol.GetAngle(atom1, atom2, atom3)) + "\n")
f.write("$end angle\n")
# all the dihedrals are printed to outtype flex, but only connection between fragments are printed if outtype is min
if outtype == "flex":
f.write("\n$dihedral\n")
torsionIterator = openbabel.OBMolTorsionIter(mol)
if ffparams:
for torsional in torsionIterator:
# Need to sum 1: http://forums.openbabel.org/Rotable-bonds-tp957795p957798.html
torsidx = [str(x + 1) for x in torsional]
try:
f.write(torsidx[0] + " " + torsidx[1] + " " +
torsidx[2] + " " + torsidx[3] + " \t" +
dihedrals["-".join([
labelToSLabel[idToAtomicLabel[x]]
for x in torsional
])] + "\n")
except KeyError as e:
print(
"The parameters for atoms %s %s %s %s (%s) was not found in the dihedrals list\n"
% (torsidx[0], torsidx[1], torsidx[2], torsidx[3],
e))
raise
else:
for torsional in torsionIterator:
# Need to sum 1: http://forums.openbabel.org/Rotable-bonds-tp957795p957798.html
torsional = [str(x + 1) for x in torsional]
f.write(torsional[0] + " " + torsional[1] + " " +
torsional[2] + " " + torsional[3] +
" \tTYPE\tV1\tV2\tV3\tf1\tf2\tf3\n")
f.write("$end dihedral\n")
# improper dihedral = carbon with only 3 atoms connected to it (SP2 hybridization)
# angle found following this definition --> http://cbio.bmt.tue.nl/pumma/index.php/Theory/Potentials
f.write("\n$improper dihedral\n")
atomIterator = openbabel.OBMolAtomIter(mol)
for atom in atomIterator:
# print(atom.GetHyb(), atom.GetAtomicNum(), atom.GetValence())
# if atom.GetAtomicNum() == 6 and atom.GetValence() == 3:
if atom.GetHyb() == 2 and atom.GetValence() == 3:
bondIterator = atom.BeginBonds()
nbrAtom = atom.BeginNbrAtom(bondIterator)
connectedAtoms = []
connectedAtoms.append(nbrAtom)
for i in range(2):
nbrAtom = atom.NextNbrAtom(bondIterator)
connectedAtoms.append(nbrAtom)
if ffparams:
torsional = [
atom.GetId(), connectedAtoms[0].GetId(),
connectedAtoms[1].GetId(),
connectedAtoms[2].GetId()
]
# create all the permutations to check if one is found
perms = list(itertools.permutations(torsional[1:]))
nfound = 0
for perm in perms:
try:
joined = "-".join([
labelToSLabel[idToAtomicLabel[
torsional[0]]]
] + [
labelToSLabel[idToAtomicLabel[x]]
for x in perm
])
f.write(
str(torsional[0] + 1) + " " +
str(torsional[1] + 1) + " " +
str(torsional[2] + 1) + " " +
str(torsional[3] + 1) + " \t" +
impropers[joined] + "\n")
except:
nfound += 1
if nfound == len(perms):
joined = "-".join(
[labelToSLabel[idToAtomicLabel[torsional[0]]]
] + [
labelToSLabel[idToAtomicLabel[x]]
for x in perms[0]
])
raise KeyError(
"The key %s (or its permutations) were not found in the improper dihedrals list\n"
% (joined))
else:
torsional = [
atom.GetId() + 1, connectedAtoms[0].GetId() + 1,
connectedAtoms[1].GetId() + 1,
connectedAtoms[2].GetId() + 1
]
torsionAngle = mol.GetTorsion(torsional[0],
torsional[1],
torsional[2],
torsional[3])
f.write(
str(torsional[0]) + " " + str(torsional[1]) + " " +
str(torsional[2]) + " " + str(torsional[3]) +
" \tV2\t" + str("%.6f" % torsionAngle) + "\n")
f.write("$end improper dihedral\n")
elif outtype == "min":
torsionIterator = openbabel.OBMolTorsionIter(mol)
# tjf = torsionals that join fragments
tjf = []
# find the tjfs by checking if all the atoms of a torsional belong to the same fragment
for tors in torsionIterator:
tors = [str(x + 1) for x in tors]
istjf = True
for atomlst in fragslst:
if (tors[0] in atomlst) and (tors[1] in atomlst) and (
tors[2] in atomlst) and (tors[3] in atomlst):
istjf = False
break
if istjf:
tjf.append(tors)
f.write("\n$dihedral\n")
if ffparams:
for torsidx in tjf:
torsional = [int(x) - 1 for x in torsidx]
try:
f.write(torsidx[0] + " " + torsidx[1] + " " +
torsidx[2] + " " + torsidx[3] + " \t" +
dihedrals["-".join([
labelToSLabel[idToAtomicLabel[x]]
for x in torsional
])] + "\n")
except KeyError as e:
print(
"The parameters for atoms %s %s %s %s (%s) was not found in the dihedrals list\n"
% (torsidx[0], torsidx[1], torsidx[2], torsidx[3],
e))
raise
else:
for torsional in tjf:
f.write(torsional[0] + " " + torsional[1] + " " +
torsional[2] + " " + torsional[3] +
" \tTYPE\tV1\tV2\tV3\tf1\tf2\tf3\n")
f.write("$end dihedral\n")
# create directory to store the fragments
if not os.path.exists(base + "_fragments"):
os.makedirs(base + "_fragments")
# write framents to the cml files
for frag in fragments:
obConversion.WriteFile(
frag,
os.path.join(
base + "_fragments",
os.path.basename(filename).split(".")[0] + "_fragment" +
frag.GetTitle() + ".xyz"))
def generate_patch_defs(H_atoms,
atom_dictionary,
squalane_mol2_file,
skeleton_file='squalane-patches_skeleton.txt'):
"""
Print necessary patch definitions for EVB parameters
:param H_atoms: List of H atoms that get within threshold distance of CN carbon
:return:
"""
with open(skeleton_file) as skeletonfile, open(
'squalane_patch_defs_all.txt', 'w') as outputfile:
outputfile.writelines(skeletonfile)
outputfile = open('squalane_patch_defs_all.txt', 'a')
obconversion = openbabel.OBConversion()
obconversion.SetInAndOutFormats("mol2", "mol2")
mol = openbabel.OBMol()
obconversion.ReadFile(mol, squalane_mol2_file)
atom_table = {}
for atom in openbabel.OBMolAtomBFSIter(mol):
atom_table[atom.GetIndex()] = atom.GetType()
for H_atom in H_atoms:
# To get numbering to start at 0
H_atom = H_atom - 1
if list(atom_table[H_atom])[0] != 'H':
print("ERROR: Atom %s is not an H atom." % H_atom)
break
else:
outputfile.write(
"PRES RH%s \t ! PATCH CN + squalane -> HCN + squalyl (missing H%s)"
% ((H_atom + 1), (H_atom + 1)))
for C_atom, connected_H_atoms in sorted(atom_dictionary.items()):
if type(connected_H_atoms
) == tuple and H_atom in connected_H_atoms:
outputfile.write("\nATOM 1C%s\t CSP2\t 0.00" %
(C_atom + 1))
elif type(connected_H_atoms
) == int and H_atom == connected_H_atoms:
outputfile.write("\nATOM 1C%s\t CSP2\t 0.00" %
(C_atom + 1))
else:
outputfile.write("\nATOM 1C%s\t CR\t 0.00" % (C_atom + 1))
for index, atom_type in atom_table.items():
if 'H' == atom_type:
outputfile.write("\nATOM 1H%s\t HC\t 0.00" %
(int(index) + 1))
outputfile.write("\nATOM 2C1\t CSP\t 0.00")
outputfile.write("\nATOM 2N2\t NSP\t 0.00")
for C_atom, connected_H_atoms in atom_dictionary.items():
if type(connected_H_atoms
) == tuple and H_atom in connected_H_atoms:
outputfile.write("\nDELETE BOND 1C%s 1H%s" %
((C_atom + 1), (H_atom + 1)))
outputfile.write("\nBOND 2C1 1H%s" % (H_atom + 1))
elif type(connected_H_atoms
) == int and H_atom == connected_H_atoms:
outputfile.write("\nDELETE BOND 1C%s 1H%s" %
((C_atom + 1), (H_atom + 1)))
outputfile.write("\nBOND 2C1 1H%s" % (H_atom + 1))
dihedrals_to_delete = []
for C_atom, connected_H_atoms in atom_dictionary.items():
for obtorsion in openbabel.OBMolTorsionIter(mol):
if H_atom in obtorsion and C_atom in obtorsion and obtorsion not in dihedrals_to_delete:
dihedrals_to_delete.append(obtorsion)
outputfile.write(
"\nDELETE DIHE 1%s%s 1%s%s 1%s%s 1%s%s" %
(list(atom_table[obtorsion[0]])[0],
(obtorsion[0] + 1), list(
atom_table[obtorsion[1]])[0],
(obtorsion[1] + 1), list(
atom_table[obtorsion[2]])[0],
(obtorsion[2] + 1), list(
atom_table[obtorsion[3]])[0],
(obtorsion[3] + 1)))
outputfile.write("\n\n")
def parse_mol_info(fname, fcharges, axis, buffa, buffo, pbcbonds, printdih,
ignorebonds, ignoreimproper):
iaxis = {"x": 0, "y": 1, "z": 2}
if axis in iaxis:
repaxis = iaxis[axis]
else:
print("Error: invalid axis")
sys.exit(0)
if fcharges:
chargesLabel = {}
with open(fcharges, "r") as f:
for line in f:
chargesLabel[line.split()[0]] = float(line.split()[1])
# set openbabel file format
base, ext = os.path.splitext(fname)
obConversion = openbabel.OBConversion()
obConversion.SetInAndOutFormats(ext[1:], "xyz")
# trick to disable ring perception and make the ReadFile waaaay faster
# Source: https://sourceforge.net/p/openbabel/mailman/openbabel-discuss/thread/56e1812d-396a-db7c-096d-d378a077853f%40ipcms.unistra.fr/#msg36225392
obConversion.AddOption("b", openbabel.OBConversion.INOPTIONS)
# read molecule to OBMol object
mol = openbabel.OBMol()
obConversion.ReadFile(mol, fname)
mol.ConnectTheDots() # necessary because of the 'b' INOPTION
# split the molecules
molecules = mol.Separate()
# detect the molecules types
mTypes = {}
mapmTypes = {}
atomIdToMol = {}
nty = 0
for i, submol in enumerate(molecules, start=1):
atomiter = openbabel.OBMolAtomIter(submol)
atlist = []
for at in atomiter:
atlist.append(at.GetAtomicNum())
atomIdToMol[at.GetId()] = i
foundType = None
for ty in mTypes:
# check if there's already a molecule of this type
if atlist == mTypes[ty]:
foundType = ty
# if not, create a new type
if not foundType:
nty += 1
foundType = nty
mTypes[nty] = atlist
mapmTypes[i] = foundType
# get atomic labels from pdb
idToAtomicLabel = {}
if ext[1:] == "pdb":
for res in openbabel.OBResidueIter(mol):
for atom in openbabel.OBResidueAtomIter(res):
if (atomIdToMol[atom.GetId()] > 1) and (len(mTypes) > 1):
idToAtomicLabel[
atom.GetId()] = res.GetAtomID(atom).strip() + str(
mapmTypes[atomIdToMol[atom.GetId()]])
else:
idToAtomicLabel[atom.GetId()] = res.GetAtomID(atom).strip()
else:
if not ob3:
etab = openbabel.OBElementTable()
for atom in openbabel.OBMolAtomIter(mol):
if (atomIdToMol[atom.GetId()] > 1) and (len(mTypes) > 1):
if ob3:
idToAtomicLabel[atom.GetId()] = openbabel.GetSymbol(
atom.GetAtomicNum()) + str(
mapmTypes[atomIdToMol[atom.GetId()]])
else:
idToAtomicLabel[atom.GetId()] = etab.GetSymbol(
atom.GetAtomicNum()) + str(
mapmTypes[atomIdToMol[atom.GetId()]])
else:
if ob3:
idToAtomicLabel[atom.GetId()] = openbabel.GetSymbol(
atom.GetAtomicNum())
else:
idToAtomicLabel[atom.GetId()] = etab.GetSymbol(
atom.GetAtomicNum())
# print(idToAtomicLabel)
# identify atom types and get masses
outMasses = "Masses\n\n"
massTypes = {}
mapTypes = {}
nmassTypes = 0
atomIterator = openbabel.OBMolAtomIter(mol)
for atom in atomIterator:
i = atom.GetId()
if idToAtomicLabel[i] not in massTypes:
nmassTypes += 1
mapTypes[nmassTypes] = idToAtomicLabel[i]
massTypes[idToAtomicLabel[i]] = nmassTypes
outMasses += "\t%d\t%.3f\t# %s\n" % (
nmassTypes, atom.GetAtomicMass(), idToAtomicLabel[i])
# create atoms list
outAtoms = "Atoms # full\n\n"
xmin = float("inf")
xmax = float("-inf")
ymin = float("inf")
ymax = float("-inf")
zmin = float("inf")
zmax = float("-inf")
natoms = 0
acoords = []
for mnum, imol in enumerate(molecules, start=1):
atomIterator = openbabel.OBMolAtomIter(imol)
for atom in sorted(atomIterator, key=lambda x: x.GetId()):
natoms += 1
i = atom.GetId()
apos = (atom.GetX(), atom.GetY(), atom.GetZ())
acoords.append(Atom(atom.GetAtomicNum(), apos))
# look for the maximum and minimum x for the box (improve later with numpy and all coordinates)
if apos[0] > xmax:
xmax = apos[0]
if apos[0] < xmin:
xmin = apos[0]
if apos[1] > ymax:
ymax = apos[1]
if apos[1] < ymin:
ymin = apos[1]
if apos[2] > zmax:
zmax = apos[2]
if apos[2] < zmin:
zmin = apos[2]
if fcharges:
outAtoms += "\t%d\t%d\t%d\t%.6f\t%.4f\t%.4f\t%.4f\t# %s\n" % (
i + 1, mnum, massTypes[idToAtomicLabel[i]],
chargesLabel[idToAtomicLabel[i]], atom.GetX(), atom.GetY(),
atom.GetZ(), idToAtomicLabel[i])
else:
outAtoms += "\t%d\t%d\t%d\tX.XXXXXX\t%.4f\t%.4f\t%.4f\t# %s\n" % (
i + 1, mnum, massTypes[idToAtomicLabel[i]], atom.GetX(),
atom.GetY(), atom.GetZ(), idToAtomicLabel[i])
# define box shape and size
try:
fromBounds = False
rcell = mol.GetData(12)
cell = openbabel.toUnitCell(rcell)
v1 = [
cell.GetCellVectors()[0].GetX(),
cell.GetCellVectors()[0].GetY(),
cell.GetCellVectors()[0].GetZ()
]
v2 = [
cell.GetCellVectors()[1].GetX(),
cell.GetCellVectors()[1].GetY(),
cell.GetCellVectors()[1].GetZ()
]
v3 = [
cell.GetCellVectors()[2].GetX(),
cell.GetCellVectors()[2].GetY(),
cell.GetCellVectors()[2].GetZ()
]
boxinfo = [v1, v2, v3]
orthogonal = True
for i, array in enumerate(boxinfo):
for j in range(3):
if i == j:
continue
if not math.isclose(0., array[j], abs_tol=1e-6):
orthogonal = False
except:
fromBounds = True
v1 = [xmax - xmin, 0., 0.]
v2 = [0., ymax - ymin, 0.]
v3 = [0., 0., zmax - zmin]
orthogonal = True
# add buffer
if orthogonal:
buf = []
boxinfo = [v1, v2, v3]
for i, val in enumerate(boxinfo[repaxis]):
if i == repaxis:
buf.append(val + buffa)
else:
buf.append(val)
boxinfo[repaxis] = buf
for i in range(3):
if i == repaxis:
continue
buf = []
for j, val in enumerate(boxinfo[i]):
if j == i:
buf.append(val + buffo)
else:
buf.append(val)
boxinfo[i] = buf
# print(boxinfo)
# Duplicate to get the bonds in the PBC. Taken from (method _crd2bond):
# https://github.com/tongzhugroup/mddatasetbuilder/blob/66eb0f15e972be0f5534dcda27af253cd8891ff2/mddatasetbuilder/detect.py#L213
if pbcbonds:
acoords = Atoms(acoords, cell=boxinfo, pbc=True)
repatoms = acoords.repeat(
2
)[natoms:] # repeat the unit cell in each direction (len(repatoms) = 7*natoms)
tree = cKDTree(acoords.get_positions())
d = tree.query(repatoms.get_positions(), k=1)[0]
nearest = d < 8.
ghost_atoms = repatoms[nearest]
realnumber = np.where(nearest)[0] % natoms
acoords += ghost_atoms
write("replicated.xyz",
acoords) # write the structure with the replicated atoms
# write new mol with new bonds
nmol = openbabel.OBMol()
nmol.BeginModify()
for idx, (num, position) in enumerate(
zip(acoords.get_atomic_numbers(), acoords.positions)):
a = nmol.NewAtom(idx)
a.SetAtomicNum(int(num))
a.SetVector(*position)
nmol.ConnectTheDots()
# nmol.PerceiveBondOrders() # super slow becauses it looks for rings
nmol.EndModify()
else:
acoords = Atoms(acoords, cell=boxinfo, pbc=False)
nmol = openbabel.OBMol()
nmol.BeginModify()
for idx, (num, position) in enumerate(
zip(acoords.get_atomic_numbers(), acoords.positions)):
a = nmol.NewAtom(idx)
a.SetAtomicNum(int(num))
a.SetVector(*position)
nmol.ConnectTheDots()
# nmol.PerceiveBondOrders() # super slow becauses it looks for rings
nmol.EndModify()
# identify bond types and create bond list
outBonds = "Bonds # harmonic\n\n"
bondTypes = {}
mapbTypes = {}
nbondTypes = 0
nbonds = 0
bondsToDelete = []
bondIterators = []
if ignorebonds:
sepmols = nmol.Separate()
for smol in sepmols[1:]:
bondIterators.append(openbabel.OBMolBondIter(smol))
else:
bondIterators.append(openbabel.OBMolBondIter(nmol))
lastidx = 1
for iterator in bondIterators:
for i, bond in enumerate(iterator, lastidx):
b1 = bond.GetBeginAtom().GetId()
b2 = bond.GetEndAtom().GetId()
# check if its a bond of the replica only
if (b1 >= natoms) and (b2 >= natoms):
bondsToDelete.append(bond)
continue
# remap to a real atom if needed
if b1 >= natoms:
b1 = realnumber[b1 - natoms]
if b2 >= natoms:
b2 = realnumber[b2 - natoms]
# identify bond type
btype1 = "%s - %s" % (idToAtomicLabel[b1], idToAtomicLabel[b2])
btype2 = "%s - %s" % (idToAtomicLabel[b2], idToAtomicLabel[b1])
if btype1 in bondTypes:
bondid = bondTypes[btype1]
bstring = btype1
elif btype2 in bondTypes:
bondid = bondTypes[btype2]
bstring = btype2
else:
nbondTypes += 1
mapbTypes[nbondTypes] = btype1
bondid = nbondTypes
bondTypes[btype1] = nbondTypes
bstring = btype1
nbonds += 1
outBonds += "\t%d\t%d\t%d\t%d\t# %s\n" % (nbonds, bondid, b1 + 1,
b2 + 1, bstring)
lastidx = i
# delete the bonds of atoms from other replicas
for bond in bondsToDelete:
nmol.DeleteBond(bond)
# identify angle types and create angle list
angleTypes = {}
mapaTypes = {}
nangleTypes = 0
nangles = 0
angleIterators = []
if ignorebonds:
sepmols = nmol.Separate()
for smol in sepmols[1:]:
smol.FindAngles()
angleIterators.append(openbabel.OBMolAngleIter(smol))
prevnumatoms = sepmols[0].NumAtoms()
else:
nmol.FindAngles()
angleIterators.append(openbabel.OBMolAngleIter(nmol))
outAngles = "Angles # harmonic\n\n"
lastidx = 1
for j, iterator in enumerate(angleIterators, 1):
for i, angle in enumerate(iterator, lastidx):
if ignorebonds:
a1 = angle[1] + prevnumatoms
a2 = angle[0] + prevnumatoms
a3 = angle[2] + prevnumatoms
else:
a1 = angle[1]
a2 = angle[0]
a3 = angle[2]
# remap to a real atom if needed
if a1 >= natoms:
a1 = realnumber[a1 - natoms]
if a2 >= natoms:
a2 = realnumber[a2 - natoms]
if a3 >= natoms:
a3 = realnumber[a3 - natoms]
atype1 = "%s - %s - %s" % (
idToAtomicLabel[a1], idToAtomicLabel[a2], idToAtomicLabel[a3])
atype2 = "%s - %s - %s" % (
idToAtomicLabel[a3], idToAtomicLabel[a2], idToAtomicLabel[a1])
if atype1 in angleTypes:
angleid = angleTypes[atype1]
astring = atype1
elif atype2 in angleTypes:
angleid = angleTypes[atype2]
astring = atype2
else:
nangleTypes += 1
mapaTypes[nangleTypes] = atype1
angleid = nangleTypes
angleTypes[atype1] = nangleTypes
astring = atype1
nangles += 1
outAngles += "\t%d\t%d\t%d\t%d\t%d\t# %s\n" % (
nangles, angleid, a1 + 1, a2 + 1, a3 + 1, astring)
lastidx = i
if ignorebonds:
prevnumatoms += sepmols[j].NumAtoms()
# identify dihedral types and create dihedral list
if printdih:
dihedralTypes = {}
mapdTypes = {}
ndihedralTypes = 0
ndihedrals = 0
dihedralIterators = []
if ignorebonds:
sepmols = nmol.Separate()
for smol in sepmols[1:]:
smol.FindTorsions()
dihedralIterators.append(openbabel.OBMolTorsionIter(smol))
else:
nmol.FindTorsions()
dihedralIterators.append(openbabel.OBMolTorsionIter(nmol))
outDihedrals = "Dihedrals # charmmfsw\n\n"
lastidx = 1
for iterator in dihedralIterators:
for i, dihedral in enumerate(iterator, lastidx):
a1 = dihedral[0]
a2 = dihedral[1]
a3 = dihedral[2]
a4 = dihedral[3]
# remap to a real atom if needed
if a1 >= natoms:
a1 = realnumber[a1 - natoms]
if a2 >= natoms:
a2 = realnumber[a2 - natoms]
if a3 >= natoms:
a3 = realnumber[a3 - natoms]
if a4 >= natoms:
a4 = realnumber[a4 - natoms]
dtype1 = "%s - %s - %s - %s" % (
idToAtomicLabel[a1], idToAtomicLabel[a2],
idToAtomicLabel[a3], idToAtomicLabel[a4])
dtype2 = "%s - %s - %s - %s" % (
idToAtomicLabel[a4], idToAtomicLabel[a3],
idToAtomicLabel[a2], idToAtomicLabel[a1])
if dtype1 in dihedralTypes:
dihedralid = dihedralTypes[dtype1]
dstring = dtype1
elif dtype2 in dihedralTypes:
dihedralid = dihedralTypes[dtype2]
dstring = dtype2
else:
ndihedralTypes += 1
mapdTypes[ndihedralTypes] = dtype1
dihedralid = ndihedralTypes
dihedralTypes[dtype1] = ndihedralTypes
dstring = dtype1
ndihedrals += 1
outDihedrals += "\t%d\t%d\t%d\t%d\t%d\t%d\t# %s\n" % (
ndihedrals, dihedralid, a1 + 1, a2 + 1, a3 + 1, a4 + 1,
dstring)
lastidx = i
if not ignoreimproper:
# look for the improper dihedrals
improperDihedralTypes = {}
mapiDTypes = {}
niDihedralTypes = 0
niDihedrals = 0
mollist = []
if ignorebonds:
sepmols = nmol.Separate()
for smol in sepmols[1:]:
smol.PerceiveBondOrders()
mollist.append(smol)
else:
nmol.PerceiveBondOrders()
mollist.append(nmol)
outImpropers = "Impropers # harmonic\n\n"
for imol in mollist:
atomIterator = openbabel.OBMolAtomIter(imol)
for atom in atomIterator:
try:
# print(atom.GetHyb(), atom.GetAtomicNum(), atom.GetValence())
expDegree = atom.GetValence()
except:
# print(atom.GetHyb(), atom.GetAtomicNum(), atom.GetExplicitDegree())
expDegree = atom.GetExplicitDegree()
# returns impropers for atoms with connected to other 3 atoms and SP2 hybridization
if atom.GetHyb() == 2 and expDegree == 3:
connectedAtoms = []
for atom2, depth in openbabel.OBMolAtomBFSIter(
imol,
atom.GetId() + 1):
if depth == 2:
connectedAtoms.append(atom2)
torsional = [
atom.GetId() + 1, connectedAtoms[0].GetId() + 1,
connectedAtoms[1].GetId() + 1,
connectedAtoms[2].GetId() + 1
]
a1 = torsional[0] - 1
a2 = torsional[1] - 1
a3 = torsional[2] - 1
a4 = torsional[3] - 1
# remap to a real atom if needed
if a1 >= natoms:
a1 = realnumber[a1 - natoms]
if a2 >= natoms:
a2 = realnumber[a2 - natoms]
if a3 >= natoms:
a3 = realnumber[a3 - natoms]
if a4 >= natoms:
a4 = realnumber[a4 - natoms]
dtype1 = "%s - %s - %s - %s" % (
idToAtomicLabel[a1], idToAtomicLabel[a2],
idToAtomicLabel[a3], idToAtomicLabel[a4])
dtype2 = "%s - %s - %s - %s" % (
idToAtomicLabel[a4], idToAtomicLabel[a3],
idToAtomicLabel[a2], idToAtomicLabel[a1])
if dtype1 in improperDihedralTypes:
idihedralid = improperDihedralTypes[dtype1]
dstring = dtype1
elif dtype2 in improperDihedralTypes:
idihedralid = improperDihedralTypes[dtype2]
dstring = dtype2
else:
niDihedralTypes += 1
mapiDTypes[niDihedralTypes] = dtype1
idihedralid = niDihedralTypes
improperDihedralTypes[dtype1] = niDihedralTypes
dstring = dtype1
niDihedrals += 1
outImpropers += "\t%d\t%d\t%d\t%d\t%d\t%d\t# %s\n" % (
niDihedrals, idihedralid, a1 + 1, a2 + 1, a3 + 1,
a4 + 1, dstring)
# print header
if printdih and (ndihedrals > 0):
if ignoreimproper or (niDihedrals == 0):
header = "LAMMPS topology created from %s using pdb2lmp.py - By Henrique Musseli Cezar, 2020\n\n\t%d atoms\n\t%d bonds\n\t%d angles\n\t%d dihedrals\n\n\t%d atom types\n\t%d bond types\n\t%d angle types\n\t%d dihedral types\n\n" % (
fname, natoms, nbonds, nangles, ndihedrals, nmassTypes,
nbondTypes, nangleTypes, ndihedralTypes)
else:
header = "LAMMPS topology created from %s using pdb2lmp.py - By Henrique Musseli Cezar, 2020\n\n\t%d atoms\n\t%d bonds\n\t%d angles\n\t%d dihedrals\n\t%d impropers\n\n\t%d atom types\n\t%d bond types\n\t%d angle types\n\t%d dihedral types\n\t%d improper types\n\n" % (
fname, natoms, nbonds, nangles, ndihedrals, niDihedrals,
nmassTypes, nbondTypes, nangleTypes, ndihedralTypes,
niDihedralTypes)
else:
header = "LAMMPS topology created from %s using pdb2lmp.py - By Henrique Musseli Cezar, 2020\n\n\t%d atoms\n\t%d bonds\n\t%d angles\n\n\t%d atom types\n\t%d bond types\n\t%d angle types\n\n" % (
fname, natoms, nbonds, nangles, nmassTypes, nbondTypes,
nangleTypes)
# add box info
if fromBounds:
boxsize = [(xmin, xmax), (ymin, ymax), (zmin, zmax)]
boxsize[repaxis] = (boxsize[repaxis][0] - buffa / 2.,
boxsize[repaxis][1] + buffa / 2.)
for i in range(3):
if i == repaxis:
continue
boxsize[i] = (boxsize[i][0] - buffo / 2.,
boxsize[i][1] + buffo / 2.)
header += "\t%.8f\t%.8f\t xlo xhi\n\t%.8f\t%.8f\t ylo yhi\n\t%.8f\t%.8f\t zlo zhi\n" % (
boxsize[0][0], boxsize[0][1], boxsize[1][0], boxsize[1][1],
boxsize[2][0], boxsize[2][1])
else:
if orthogonal:
header += "\t%.8f\t%.8f\t xlo xhi\n\t%.8f\t%.8f\t ylo yhi\n\t%.8f\t%.8f\t zlo zhi\n" % (
0., boxinfo[0][0], 0., boxinfo[1][1], 0., boxinfo[2][2])
else:
header += "\t%.8f\t%.8f\t xlo xhi\n\t%.8f\t%.8f\t ylo yhi\n\t%.8f\t%.8f\t zlo zhi\n\t%.8f\t%.8f\t%.8f\t xy xz yz\n" % (
0., boxinfo[0][0], 0., boxinfo[1][1], 0., boxinfo[2][2],
boxinfo[1][0], boxinfo[2][0], boxinfo[2][1])
# print Coeffs
outCoeffs = "Pair Coeffs\n\n"
for i in range(1, nmassTypes + 1):
outCoeffs += "\t%d\teps\tsig\t# %s\n" % (i, mapTypes[i])
outCoeffs += "\nBond Coeffs\n\n"
for i in range(1, nbondTypes + 1):
outCoeffs += "\t%d\tK\tr_0\t# %s\n" % (i, mapbTypes[i])
outCoeffs += "\nAngle Coeffs\n\n"
for i in range(1, nangleTypes + 1):
outCoeffs += "\t%d\tK\ttetha_0 (deg)\t# %s\n" % (i, mapaTypes[i])
if printdih and (ndihedrals > 0):
outCoeffs += "\nDihedral Coeffs\n\n"
for i in range(1, ndihedralTypes + 1):
outCoeffs += "\t%d\tK\tn\tphi_0 (deg)\tw\t# %s\n" % (i,
mapdTypes[i])
if not ignoreimproper and (niDihedralTypes > 0):
outCoeffs += "\nImproper Coeffs\n\n"
for i in range(1, niDihedralTypes + 1):
outCoeffs += "\t%d\tK\txi_0 (deg)\t# %s\n" % (i, mapiDTypes[i])
if printdih and (ndihedrals > 0):
if ignoreimproper or (niDihedralTypes == 0):
return header + "\n" + outMasses + "\n" + outCoeffs + "\n" + outAtoms + "\n" + outBonds + "\n" + outAngles + "\n" + outDihedrals
else:
return header + "\n" + outMasses + "\n" + outCoeffs + "\n" + outAtoms + "\n" + outBonds + "\n" + outAngles + "\n" + outDihedrals + "\n" + outImpropers
else:
return header + "\n" + outMasses + "\n" + outCoeffs + "\n" + outAtoms + "\n" + outBonds + "\n" + outAngles
def binarify_molecules(inputfile, outputfile):
ofile = open(outputfile, 'wb')
obconv = ob.OBConversion()
obconv.SetInFormat('xyz')
for molidx, xyzfile in enumerate(open(inputfile).readlines()):
mol = ob.OBMol()
obconv.ReadFile(mol, xyzfile.strip())
molecule_id = molidx
number_of_atoms = mol.NumAtoms()
# atom properties
nuclear_coordinates, element_numbers = [], []
first = None
for atom in ob.OBMolAtomIter(mol):
element_numbers.append(atom.GetAtomicNum())
# Shift molecule to implicitly place the first atom at the origin
if first is None:
first = [atom.GetX(), atom.GetY(), atom.GetZ()]
else:
nuclear_coordinates.append([
atom.GetX() - first[0],
atom.GetY() - first[1],
atom.GetZ() - first[2]
])
# bond properties
bond_orders = []
bonds = []
for bond in ob.OBMolBondIter(mol):
bond_orders.append(bond.GetBO())
bonds.append(
[bond.GetBeginAtomIdx() - 1,
bond.GetEndAtomIdx() - 1])
number_of_bonds = len(bonds)
# dihedral properties
dihedrals = []
for dihedral in ob.OBMolTorsionIter(mol):
heavy = [element_numbers[_] > 1 for _ in dihedral]
# assert middle atoms to be heavy
if not (heavy[1] and heavy[2]):
continue
# skip if two H are involved
if not heavy[0] and not heavy[3]:
continue
dihedrals.append(list(dihedral))
number_of_dihedrals = len(dihedrals)
# binary packing
formatstring = '=IB%dB%ddB%dB%dBB%dB' % (
number_of_atoms, # element numbers
3 * (number_of_atoms - 1), # atomic coordinates
2 * number_of_bonds, # atom indices in bonds
number_of_bonds, # bond orders
4 * number_of_dihedrals) # atom indices in dihedrals
nuclear_coordinates = sum(nuclear_coordinates, [])
bonds = sum(bonds, [])
dihedrals = sum(dihedrals, [])
ofile.write(
struct.pack(formatstring, molecule_id, number_of_atoms,
*element_numbers, *nuclear_coordinates,
number_of_bonds, *bonds, *bond_orders,
number_of_dihedrals, *dihedrals))
ofile.close()