def __iter__(self):
if self.properties is None:
i = 0
while True:
mol = Chem.BasicMolecule()
try:
if self.r.read(mol):
yield sanitize_mol(mol)
else:
break
except IOError:
yield None
i += 1
if i == self.nr_mols:
break
else:
i = 0
while True:
mol = Chem.BasicMolecule()
try:
if self.r.read(mol):
read_properties = self._extract_properties_from_mol(
mol)
yield sanitize_mol(mol), read_properties
else:
break
except IOError:
yield None
i += 1
if i == self.nr_mols:
break
def cdfMol_pdb(pdb, output, name):
initial_time = time.time()
cdf_mol = Chem.BasicMolecule()
pdb_mol = Chem.BasicMolecule()
pdb_str = open(pdb, 'r').read().replace('WAT', 'HOH').replace('HIE', 'HIS')
pdb_reader = Biomol.PDBMoleculeReader(Base.StringIOStream(pdb_str))
Biomol.setPDBApplyDictAtomBondingToNonStdResiduesParameter(
pdb_reader, True)
if not pdb_reader.read(pdb_mol):
return None
Chem.calcImplicitHydrogenCounts(pdb_mol, False)
Chem.perceiveHybridizationStates(pdb_mol, False)
Chem.setAtomSymbolsFromTypes(pdb_mol, False)
Chem.perceiveSSSR(pdb_mol, False)
Chem.setRingFlags(pdb_mol, False)
Chem.setAromaticityFlags(pdb_mol, False)
cdf_mol.assign(pdb_mol)
for atom in cdf_mol.atoms:
Chem.set3DCoordinatesArray(atom, Math.Vector3DArray())
i = 0
while i < cdf_mol.numAtoms:
Chem.get3DCoordinatesArray(cdf_mol.getAtom(i)).addElement(
Chem.get3DCoordinates(pdb_mol.getAtom(i)))
i += 1
tmp_output = output + name + ".cdf"
try:
Chem.FileCDFMolecularGraphWriter(tmp_output).write(cdf_mol)
except:
print('> Cdf_mol writing failure.')
raise
residues = Biomol.ResidueList(cdf_mol)
tmp_output = output + name + "_residue_info.txt"
with open(tmp_output, 'w') as txt_writer:
txt_writer.write('residue name_resid_chain\n')
for res in residues:
res_id = getResidueID(res)
txt_writer.write('{}: \n'.format(res_id))
calc_time = time.time() - initial_time
print('> Cdf and amino acid residue number list files generated in {}s'.
format(int(calc_time)))
def getMoleculeFromAtom(atom: Chem.BasicAtom, protein: Chem.BasicMolecule) -> (Chem.BasicMolecule, list):
"""
Given an atom and a protein structure, find the ligand the atom corresponds to.
Traverses the molecule by its bonds until no longer any atoms are attached. All atoms and bonds are assigned to a
new molecule, which is being returned.
:param atom:
:param protein:
:return: The found ligand as well as the atom indices of the ligand in the parent molecule.
"""
ligand = Chem.Fragment()
neighbors = set() # atoms not being added already
neighborsAdded = set() # keep track of added atoms to not process twice
atomsToRemove = []
neighbors.add(atom)
while len(neighbors) > 0:
n = neighbors.pop()
neighborsAdded.add(n)
ligand.addAtom(n)
atomsToRemove.append(protein.getAtomIndex(n))
# get all the neighbor atoms
for i, b in enumerate(n.bonds):
for a in b.atoms:
if a != n:
if a not in neighbors and a not in neighborsAdded: # new atom
neighbors.add(a)
ligand.addBond(b) # ignored if already exists
Chem.perceiveComponents(ligand, True)
mol = Chem.BasicMolecule()
mol.assign(ligand)
return mol, atomsToRemove
def getSurfaceAtoms(mol):
surfaceATomExtractor = Chem.SurfaceAtomExtractor()
f = Chem.Fragment()
surfaceATomExtractor.extract(mol, mol, f)
surfaceAtoms = Chem.BasicMolecule()
surfaceAtoms.assign(f)
return surfaceAtoms
def readPDBFromStream(stream: Base.IOStream):
from Protein import Protein
from MoleculeTools import sanitize_mol
r = Biomol.PDBMoleculeReader(stream)
mol = Chem.BasicMolecule()
r.read(mol)
sanitize_mol(mol, makeHydrogenComplete=True)
return Protein(mol)
def process():
if len(sys.argv) < 4:
print('Usage:',
sys.argv[0],
'[input.sdf] [exclude-molecules.sdf] [output.sdf]',
file=sys.stderr)
sys.exit(2)
ifs = Base.FileIOStream(sys.argv[1], 'r')
xifs = Base.FileIOStream(sys.argv[2], 'r')
ofs = Base.FileIOStream(sys.argv[3], 'w')
reader = Chem.SDFMoleculeReader(ifs)
xreader = Chem.SDFMoleculeReader(xifs)
writer = Chem.SDFMolecularGraphWriter(ofs)
mol = Chem.BasicMolecule()
Chem.setMultiConfImportParameter(reader, False)
Chem.setMultiConfImportParameter(xreader, False)
Chem.setMultiConfExportParameter(writer, False)
stats = Stats()
stats.read = 0
stats.dropped = 0
xhashes = set()
while xreader.read(mol):
setupMolecule(mol)
hashcode = Chem.calcHashCode(mol)
xhashes.add(hashcode)
while reader.read(mol):
#print('Processing Molecule ' + str(stats.read)
setupMolecule(mol)
hashcode = Chem.calcHashCode(mol)
if hashcode in xhashes:
stats.dropped += 1
print('Dropped Molecule ' + str(stats.read) + ': ' +
Chem.generateSMILES(mol) + ' ' + Chem.getName(mol),
file=sys.stderr)
else:
writer.write(mol)
stats.read += 1
if stats.read % 10000 == 0:
print('Processed ' + str(stats.read) + ' Molecules...',
file=sys.stderr)
print('', file=sys.stderr)
print('-- Summary --', file=sys.stderr)
print('Molecules processed: ' + str(stats.read), file=sys.stderr)
print('Molecules dropped: ' + str(stats.dropped), file=sys.stderr)
def __init__(self, structure: Chem.BasicMolecule = None):
self.shape: Shape.GaussianShape = Shape.GaussianShape()
self.shapeFunc: Shape.GaussianShapeFunction = Shape.GaussianShapeFunction(
)
self.coordinates: Math.Vector3DArray = Math.Vector3DArray()
self.ligands: List[Chem.BasicMolecule] = []
self.surfaceAtoms: Chem.BasicMolecule = Chem.BasicMolecule()
super(Protein, self).__init__()
if structure:
from MoleculeTools import sanitize_mol
self.assign(structure)
def remove_components(mol: Chem.BasicMolecule) -> Chem.BasicMolecule:
components = Chem.getComponents(mol)
largest_component = None # set default number of atoms and index
for comp in components:
if largest_component is None:
largest_component = comp
elif comp.numAtoms > largest_component.numAtoms:
largest_component = comp
new_mol = Chem.BasicMolecule()
new_mol.assign(largest_component)
if Chem.hasStructureData(mol):
Chem.setStructureData(new_mol, Chem.getStructureData(mol))
return new_mol
def loadCDFMolecule(fname):
mol = Chem.BasicMolecule()
cdf_reader = Chem.FileCDFMoleculeReader(fname)
if not cdf_reader.read(mol):
return None
Chem.calcImplicitHydrogenCounts(mol, False)
Chem.perceiveHybridizationStates(mol, False)
Chem.setAtomSymbolsFromTypes(mol, False)
Chem.perceiveSSSR(mol, False)
Chem.setRingFlags(mol, False)
Chem.setAromaticityFlags(mol, False)
return mol
def getSurfaceExposedAtoms(self, copy=True) -> Chem.BasicMolecule:
"""
Get a list of CDPL Molecules located on the protein surface.
:param copy: Whether to return a copy of the surface atoms or the surface atom object itself.
:return:
"""
from ProteinTools import getSurfaceAtoms
self.surfaceAtoms.assign(getSurfaceAtoms(self))
if copy:
surfaceAtoms = Chem.BasicMolecule()
surfaceAtoms.assign(self.surfaceAtoms)
return surfaceAtoms
return self.surfaceAtoms
def CDPLmolFromSdf(sdf_path, conformation):
'''
generates a single CDPL Molecule from an sdf-file. If conformations is true, then
one random conformation will be generated. \n
Input: \n
sdf_path (string): path to the sdf file \n
conformation (boolean): generates one 3d conformation according to MMFF94 \n
Return: \n
(CDPL BasicMolecule): the corresponding CDPL BasicMolecule
'''
mol = Chem.BasicMolecule()
ifs = Base.FileIOStream(sdf_path, 'r')
sdf_reader = Chem.SDFMoleculeReader(ifs)
if not sdf_reader.read(mol):
log.error("COULD NOT READ SDF", sdf_path)
return False
if conformation:
return _CDPLgenerateConformation(mol)
return mol
def CDPLmolFromSmiles(smiles_path, conformation):
'''
generates a CDPL Molecule from smiles. If confromations is true, then
one random conformation will be generated with explicit hydrogens. \n
Input: \n
smiles (string): smiles string \n
conformation (boolean): generates one 3d conformation according to MMFF94 \n
Return: \n
(CDPL BasicMolecule): the corresponding CDPL BasicMolecule
'''
mol = Chem.BasicMolecule()
if ".smi" in smiles_path:
smi_reader = Chem.FileSMILESMoleculeReader(smiles_path)
if not smi_reader.read(mol):
log.error("COULD NOT READ Smiles", smiles_path)
return False
else:
mol = Chem.parseSMILES(smiles_path)
if conformation:
return _CDPLgenerateConformation(mol)
else:
return mol
def _CDPLreadFromPDBFile(pdb_file):
'''
PRIVAT METHOD
reads a pdb file and is used by the CDPLreadProteinFile method.
Input: \n
pdb_file (string): the path to the pdb file \n
Return: \n
(CDPL BasicMolecule): the corresponding pdb molecule
'''
ifs = Base.FileIOStream(pdb_file, 'r')
pdb_reader = Biomol.PDBMoleculeReader(ifs)
pdb_mol = Chem.BasicMolecule()
Biomol.setPDBApplyDictAtomBondingToNonStdResiduesParameter(
pdb_reader, False
) #TODO Should this be there for the pdb readin? or also in the config?
if not pdb_reader.read(pdb_mol):
log.error("COULD NOT READ PDB", pdb_file)
return False
return pdb_mol
def process(sdf_file, psd_file_path):
ifs = Base.FileIOStream(sdf_file, 'r')
reader = Chem.SDFMoleculeReader(ifs)
mol = Chem.BasicMolecule()
Chem.setMultiConfImportParameter(reader, True)
psd_creator = Pharm.PSDScreeningDBCreator(
psd_file_path, Pharm.PSDScreeningDBCreator.CREATE, True)
i = 0
t0 = time.clock()
while reader.read(mol):
setupMolecule(mol)
psd_creator.process(mol)
i += 1
if i % 100 == 0:
print 'Processed ' + str(i) + ' molecules (' + str(
time.clock() - t0), 's elapsed)...'
t0 = time.clock()
mol.clear()
print ''
print '-- Summary --'
print 'Molecules processed: ' + str(psd_creator.numProcessed)
print 'Molecules rejected: ' + str(psd_creator.numRejected)
print 'Molecules deleted: ' + str(psd_creator.numDeleted)
print 'Molecules inserted: ' + str(psd_creator.numInserted)
psd_creator.close()
def generate_ph(pdb, key):
ifs = Base.FileIOStream(pdb, 'r')
tlc = self.ligand_3_letter_code
pdb_reader = Biomol.PDBMoleculeReader(ifs)
pdb_mol = Chem.BasicMolecule()
print '- Reading input: ', pdb, ' ...'
if not pdb_reader.read(pdb_mol):
print '!! Could not read input molecule'
return
print '- Processing macromolecule', pdb, ' ...'
i = 0
while i < pdb_mol.getNumBonds():
bond = pdb_mol.getBond(i)
if Chem.isMetal(bond.atoms[0]) or Chem.isMetal(bond.atoms[1]):
pdb_mol.removeBond(i)
else:
i += 1
Chem.calcImplicitHydrogenCounts(pdb_mol, True)
Chem.perceiveHybridizationStates(pdb_mol, True)
Chem.makeHydrogenComplete(pdb_mol)
Chem.setAtomSymbolsFromTypes(pdb_mol, False)
Chem.calcImplicitHydrogenCounts(pdb_mol, True)
Biomol.setHydrogenResidueSequenceInfo(pdb_mol, False)
Chem.setRingFlags(pdb_mol, True)
Chem.setAromaticityFlags(pdb_mol, True)
Chem.generateHydrogen3DCoordinates(pdb_mol, True)
ligand = Chem.Fragment()
print '- Extracting ligand ', tlc, ' ...'
for atom in pdb_mol.atoms:
if Biomol.getResidueCode(atom) == tlc:
Biomol.extractResidueSubstructure(atom, pdb_mol, ligand,
False)
break
if ligand.numAtoms == 0:
print '!! Could not find ligand', tlc, 'in input file'
return
Chem.perceiveSSSR(ligand, True)
lig_env = Chem.Fragment()
Biomol.extractEnvironmentResidues(ligand, pdb_mol, lig_env, 7.0)
Chem.perceiveSSSR(lig_env, True)
print '- Constructing pharmacophore ...'
lig_pharm = Pharm.BasicPharmacophore()
env_pharm = Pharm.BasicPharmacophore()
pharm_gen = Pharm.DefaultPharmacophoreGenerator(False)
pharm_gen.generate(ligand, lig_pharm)
pharm_gen.generate(lig_env, env_pharm)
analyzer = Pharm.DefaultInteractionAnalyzer()
interactions = Pharm.FeatureMapping()
analyzer.analyze(lig_pharm, env_pharm, interactions)
#------------------------- XVOLS
int_env_ftrs = Pharm.FeatureSet()
Pharm.getFeatures(int_env_ftrs, interactions, False)
int_core_ftrs = Pharm.FeatureSet()
Pharm.getFeatures(int_core_ftrs, interactions, True)
int_pharm = Pharm.BasicPharmacophore(int_core_ftrs)
for ftr in int_env_ftrs:
if Pharm.getType(
ftr
) == Pharm.FeatureType.H_BOND_DONOR or Pharm.getType(
ftr) == Pharm.FeatureType.H_BOND_ACCEPTOR:
Pharm.setTolerance(ftr, 1.0)
else:
Pharm.setTolerance(ftr, 1.5)
Pharm.createExclusionVolumes(int_pharm, int_env_ftrs, 0.0, 0.1,
False)
int_env_ftr_atoms = Chem.Fragment()
Pharm.getFeatureAtoms(int_env_ftrs, int_env_ftr_atoms)
int_residue_atoms = Chem.Fragment()
Biomol.extractResidueSubstructures(int_env_ftr_atoms, lig_env,
int_residue_atoms, True)
Chem.makeHydrogenDeplete(int_residue_atoms)
def isAlphaAtom(atom):
return Biomol.getResidueAtomName(atom) == 'CA'
Chem.removeAtomsIfNot(int_residue_atoms, isAlphaAtom)
Pharm.createExclusionVolumes(int_pharm, int_residue_atoms,
Chem.Atom3DCoordinatesFunctor(), 1.0,
2.0, False)
features_in_ph = []
for int_ftr in int_pharm:
if Pharm.hasSubstructure(int_ftr) == False:
continue
elif ftype_names[Pharm.getType(int_ftr)] == 'XV':
continue
feature_id = generate_key(int_ftr)
features_in_ph.append(str(feature_id))
self.unique_feature_vector.add(str(feature_id))
int_pharm.fv = features_in_ph
int_pharm.path_to_pdb = pdb
return int_pharm
import CDPL.Chem as Chem
import CDPL.Math as Math
def process():
if len(sys.argv) < 4:
print('Usage:', sys.argv[0], 'training-set.sdf logP-data regression-coeff-file', file=sys.stderr)
sys.exit(2)
struct_is = Base.FileIOStream(sys.argv[1], 'r')
exp_logp_is = Base.FileIOStream(sys.argv[2], 'r')
coeff_os = Base.FileIOStream(sys.argv[3], 'w')
mlr_model = Math.DMLRModel()
sdf_reader = Chem.SDFMoleculeReader(struct_is)
mol = Chem.BasicMolecule()
xlogp_calc = Chem.XLogPCalculator()
histo = Math.DVector()
histo.resize(Chem.XLogPCalculator.FEATURE_VECTOR_SIZE)
Chem.setMultiConfImportParameter(sdf_reader, False)
while sdf_reader.read(mol):
exp_logp = float(exp_logp_is.readline())
Chem.perceiveComponents(mol, False)
Chem.perceiveSSSR(mol, False)
Chem.setRingFlags(mol, False)
Chem.calcImplicitHydrogenCounts(mol, False)
Chem.perceiveHybridizationStates(mol, False)
def process():
if len(sys.argv) < 3:
print >> sys.stderr, 'Usage:', sys.argv[
0], '[input.cdf] [output directory]'
sys.exit(2)
in_fname = path.splitext(path.basename(sys.argv[1]))[0]
mol = Chem.BasicMolecule()
cdf_reader = Chem.FileCDFMoleculeReader(sys.argv[1])
pvd_file = open(path.join(sys.argv[2], in_fname + '.pvd'), 'w')
Util.writePVDHeader(pvd_file)
print >> sys.stderr, '- Processing CDF-file:', sys.argv[1], '...'
if not cdf_reader.read(mol):
print '!! Could not read file'
sys.exit(2)
backbone_atoms = []
for atom in mol.atoms:
if Biomol.isPDBBackboneAtom(atom) and Biomol.getResidueAtomName(
atom) == 'C':
backbone_atoms.append(atom)
bond_list = []
for bond in mol.bonds:
if Biomol.getResidueCode(
bond.getAtom(0)) == 'HOH' or Biomol.getResidueCode(
bond.getAtom(1)) == 'HOH':
continue
if Chem.getType(bond.getAtom(0)) == Chem.AtomType.H or Chem.getType(
bond.getAtom(1)) == Chem.AtomType.H:
continue
bond_list.append(bond)
num_confs = Chem.getNumConformations(mol)
num_coords = len(bond_list) * 4 + (
len(backbone_atoms) * SPLINE_POINTS_PER_BB_ATOM - 1) * 2
bond_ctr = Math.Vector3D()
i = 0
while i < num_confs:
line_x_coords = numpy.ndarray(num_coords, numpy.float32)
line_y_coords = numpy.ndarray(num_coords, numpy.float32)
line_z_coords = numpy.ndarray(num_coords, numpy.float32)
atom_types = numpy.ndarray(num_coords, numpy.uint32)
spline_ctrl_points = numpy.ndarray((len(backbone_atoms), 3),
numpy.float32)
j = 0
for atom in backbone_atoms:
atom_pos = Chem.getConformer3DCoordinates(atom, i)
spline_ctrl_points[j, 0] = atom_pos(0)
spline_ctrl_points[j, 1] = atom_pos(1)
spline_ctrl_points[j, 2] = atom_pos(2)
j += 1
spline_pts = spline(spline_ctrl_points,
len(backbone_atoms) * SPLINE_POINTS_PER_BB_ATOM)
j = 0
k = 0
while k < (len(backbone_atoms) * SPLINE_POINTS_PER_BB_ATOM - 1):
line_x_coords[j] = spline_pts[0][k]
line_y_coords[j] = spline_pts[1][k]
line_z_coords[j] = spline_pts[2][k]
atom_types[j] = 0
j += 1
line_x_coords[j] = spline_pts[0][k + 1]
line_y_coords[j] = spline_pts[1][k + 1]
line_z_coords[j] = spline_pts[2][k + 1]
atom_types[j] = 0
j += 1
k += 1
for bond in bond_list:
atom1 = bond.getAtom(0)
atom2 = bond.getAtom(1)
atom1_pos = Chem.getConformer3DCoordinates(atom1, i)
atom2_pos = Chem.getConformer3DCoordinates(atom2, i)
atom1_type = Chem.getType(atom1)
atom2_type = Chem.getType(atom2)
bond_ctr.assign(atom1_pos)
bond_ctr += atom2_pos
bond_ctr *= 0.5
line_x_coords[j] = atom1_pos(0)
line_y_coords[j] = atom1_pos(1)
line_z_coords[j] = atom1_pos(2)
atom_types[j] = atom1_type
j += 1
line_x_coords[j] = bond_ctr(0)
line_y_coords[j] = bond_ctr(1)
line_z_coords[j] = bond_ctr(2)
atom_types[j] = atom1_type
j += 1
line_x_coords[j] = bond_ctr(0)
line_y_coords[j] = bond_ctr(1)
line_z_coords[j] = bond_ctr(2)
atom_types[j] = atom2_type
j += 1
line_x_coords[j] = atom2_pos(0)
line_y_coords[j] = atom2_pos(1)
line_z_coords[j] = atom2_pos(2)
atom_types[j] = atom2_type
j += 1
line_x_coords.resize(j)
line_y_coords.resize(j)
line_z_coords.resize(j)
atom_types.resize(j)
out_fname = in_fname + '_frame_no_' + str(i)
out_path = path.join(sys.argv[2], out_fname)
line_data = {'atom_type': atom_types}
print >> sys.stderr, '- Writing structure data for frame', i, '...'
if not pyevtk.hl.linesToVTK(out_path,
line_x_coords,
line_y_coords,
line_z_coords,
pointData=line_data):
print '!! Could not write output file'
sys.exit(2)
Util.writePVDEntry(pvd_file, i, out_fname, 'vtu')
i += 1
Util.writePVDFooter(pvd_file)
def generate_ph(pdb, args, df_constructor, ts):
ifs = Base.FileIOStream(pdb, 'r')
tlc = args.ligand_three_letter_code
pdb_reader = Biomol.PDBMoleculeReader(ifs)
pdb_mol = Chem.BasicMolecule()
print '- Reading input: ', pdb, ' ...'
if not pdb_reader.read(pdb_mol):
print '!! Could not read input molecule'
return
print '- Processing macromolecule', pdb, ' ...'
i = 0
while i < pdb_mol.getNumBonds():
bond = pdb_mol.getBond(i)
if Chem.isMetal(bond.atoms[0]) or Chem.isMetal(bond.atoms[1]):
pdb_mol.removeBond(i)
else:
i += 1
for a in pdb_mol.atoms:
Chem.setImplicitHydrogenCount(a, 0)
Chem.calcImplicitHydrogenCounts(pdb_mol, True)
Chem.perceiveHybridizationStates(pdb_mol, True)
Chem.makeHydrogenComplete(pdb_mol)
Chem.setAtomSymbolsFromTypes(pdb_mol, False)
Chem.calcImplicitHydrogenCounts(pdb_mol, True)
Biomol.setHydrogenResidueSequenceInfo(pdb_mol, False)
Chem.setRingFlags(pdb_mol, True)
Chem.setAromaticityFlags(pdb_mol, True)
Chem.generateHydrogen3DCoordinates(pdb_mol, True)
Chem.calcFormalCharges(pdb_mol, True)
ligand = Chem.Fragment()
print '- Extracting ligand ', tlc, ' ...'
for atom in pdb_mol.atoms:
if Biomol.getResidueCode(atom) == tlc:
Biomol.extractResidueSubstructure(atom, pdb_mol, ligand, False)
break
if ligand.numAtoms == 0:
print '!! Could not find ligand', tlc, 'in input file'
return
Chem.perceiveSSSR(ligand, True)
lig_env = Chem.Fragment()
Biomol.extractEnvironmentResidues(ligand, pdb_mol, lig_env, 7.0)
Chem.perceiveSSSR(lig_env, True)
print '- Constructing pharmacophore ...'
lig_pharm = Pharm.BasicPharmacophore()
env_pharm = Pharm.BasicPharmacophore()
pharm_gen = Pharm.DefaultPharmacophoreGenerator(True)
pharm_gen.generate(ligand, lig_pharm)
pharm_gen.generate(lig_env, env_pharm)
#Pharm.FilePMLFeatureContainerWriter('./test/lig_ph_' + str(ts) + '.pml').write(lig_pharm)
analyzer = Pharm.DefaultInteractionAnalyzer()
interactions = Pharm.FeatureMapping()
analyzer.analyze(lig_pharm, env_pharm, interactions)
df_constructor, interaction_at_ts = outputInteractions(
lig_pharm, env_pharm, interactions, df_constructor)
#Chem.FileSDFMolecularGraphWriter('./test/ligand_' + str(ts) + '.sdf').write(ligand)
return df_constructor, interaction_at_ts
def cleanStructures():
if len(sys.argv) < 5:
print('Usage:',
sys.argv[0],
'[input.sdf] [output.sdf] [dropped.sdf] [start_index] [[count]]',
file=sys.stderr)
sys.exit(2)
ifs = Base.FileIOStream(sys.argv[1], 'r')
ofs = Base.FileIOStream(sys.argv[2], 'w')
dofs = Base.FileIOStream(sys.argv[3], 'w')
offset = int(sys.argv[4])
count = 0
if len(sys.argv) > 5:
count = int(sys.argv[5])
reader = Chem.SDFMoleculeReader(ifs)
writer = Chem.SDFMolecularGraphWriter(ofs)
dwriter = Chem.SDFMolecularGraphWriter(dofs)
mol = Chem.BasicMolecule()
#Chem.setSMILESRecordFormatParameter(reader, 'SN')
stats = Stats()
stats.read = 0
stats.dropped = 0
stats.modified = 0
Chem.setMultiConfImportParameter(reader, False)
Chem.setMultiConfExportParameter(writer, False)
Chem.setMultiConfExportParameter(dwriter, False)
if offset > 0:
print('Skipping Molecules to Start Index ' + str(offset),
file=sys.stderr)
reader.setRecordIndex(offset)
#print('Finished Setting Record Index', file=sys.stderr)
stats.read = offset
while reader.read(mol):
#print('Processing Molecule ' + str(stats.read)
proc_mol = processMolecule(mol, stats)
if proc_mol:
writer.write(proc_mol)
else:
stats.dropped += 1
dwriter.write(mol)
print('Dropped Molecule ' + str(stats.read) + ': ' +
generateSMILES(mol) + ' ' + Chem.getName(mol),
file=sys.stderr)
stats.read += 1
if stats.read % 10000 == 0:
print('Processed ' + str(stats.read - offset) + ' Molecules...',
file=sys.stderr)
if count > 0 and (stats.read - offset) >= count:
break
print('', file=sys.stderr)
print('-- Summary --', file=sys.stderr)
print('Molecules processed: ' + str(stats.read - offset), file=sys.stderr)
print('Molecules dropped: ' + str(stats.dropped), file=sys.stderr)
print('Molecules modified: ' + str(stats.modified), file=sys.stderr)
def mol_from_smiles(smiles: str) -> Chem.BasicMolecule:
ifs = Base.StringIOStream(smiles)
mol = Chem.BasicMolecule()
r = Chem.SMILESMoleculeReader(ifs)
r.read(mol)
return sanitize_mol(mol)
def readPDBFromStream(stream: Base.IOStream):
r = Biomol.PDBMoleculeReader(stream)
mol = Chem.BasicMolecule()
r.read(mol)
return Protein(mol)
# remove ligand and water, just to be sure
protein.separateLigandFromProtein()
# calculate surface area for a single carbon atom
protein.getGaussianShape()
shapeFunc = protein.shapeFunc
carbonProteinIndex = None
for a in protein.atoms:
if Chem.getType(a) == 6:
carbonProteinIndex = protein.getAtomIndex(a)
break
surfAreaCarbonProtein = shapeFunc.calcSurfaceArea(carbonProteinIndex) # calculate the surface area contribution here?
# create a simple carbon molecule with coordinates
carbon = Chem.BasicMolecule()
cAtom = carbon.addAtom()
Chem.setType(cAtom, 6)
coords = Math.Vector3D()
coords.assign([1, 2, 3])
Chem.set3DCoordinates(cAtom, coords)
# calculate shape and surface area of carbon molecule
carbonShape, carbonShapeFunc = getGaussianShapeOfMolecule(carbon)
surfAreaCarbon = carbonShapeFunc.surfaceArea
# assert that contribution of carbon atom in protein is in fact the entire surface area of a single carbon
assert surfAreaCarbon == surfAreaCarbonProtein
# What am I missing here?
# Summing the surface area of all atoms in the protein yields the surface area of the protein. I find it hard to
def process():
if len(sys.argv) < 4:
print(
'Usage:',
sys.argv[0],
'[input torsion rules.xml] [structures.sdf] [output torsion histogram library.sdf]',
file=sys.stderr)
sys.exit(2)
tor_lib = ConfGen.TorsionLibrary()
try:
tor_lib.load(Base.FileIOStream(sys.argv[1], 'r'))
except:
print('Error while loading input torsion rules:',
sys.exc_info()[0],
file=sys.stderr)
sys.exit(2)
tor_matcher = ConfGen.TorsionRuleMatcher(tor_lib)
tor_matcher.findAllRuleMappings(True)
tor_matcher.findUniqueMappingsOnly(True)
tor_matcher.stopAtFirstMatchingRule(True)
mol = Chem.BasicMolecule()
mol_reader = Chem.FileSDFMoleculeReader(sys.argv[2])
Chem.setMultiConfImportParameter(mol_reader, False)
print('- Analyzing input structures...', file=sys.stderr)
i = 1
rule_to_angle_hists = {}
coords = Math.Vector3DArray()
while True:
try:
if not mol_reader.read(mol):
break
except IOError as e:
print('Error while reading input molecule',
i,
':',
e,
file=sys.stderr)
i += 1
continue
if i % 500 == 0:
print(' ... At input molecule', i, '...', file=sys.stderr)
Chem.initSubstructureSearchTarget(mol, False)
try:
Chem.get3DCoordinates(mol, coords)
except Base.ItemNotFound:
print('Could not get 3D-coordinates for molecule',
i,
file=sys.stderr)
i += 1
continue
for bond in mol.bonds:
if Chem.getRingFlag(bond):
continue
if Chem.isHydrogenBond(bond):
continue
if Chem.getExplicitBondCount(
bond.getBegin()) <= 1 or Chem.getExplicitBondCount(
bond.getEnd()) <= 1:
continue
tor_matcher.findMatches(bond, mol, False)
for match in tor_matcher:
processMatch(i, match, mol, coords, rule_to_angle_hists)
i += 1
print('- Processing torsion angle histograms...', file=sys.stderr)
processHistograms(tor_lib, rule_to_angle_hists)
print('- Writing output torsion library...', file=sys.stderr)
try:
tor_lib.save(Base.FileIOStream(sys.argv[3], 'w+'))
except:
print('Error while writing torsion library:',
sys.exc_info()[0],
file=sys.stderr)
sys.exit(2)
print('DONE!', file=sys.stderr)
def process():
if len(sys.argv) < 4:
print >> sys.stderr, 'Usage:', sys.argv[
0], '[input topology-file] [input coordinates-file] [output CDF-file]'
sys.exit(2)
print >> sys.stderr, '- Processing topology-file', sys.argv[
1], 'and coordinates-file', sys.argv[2], '...'
u = MDAnalysis.Universe(sys.argv[1], sys.argv[2])
cdf_mol = Chem.BasicMolecule()
cdf_mol.reserveMemoryForAtoms(len(u.atoms))
cdf_mol.reserveMemoryForBonds(len(u.bonds))
print >> sys.stderr, '- Num. atoms:', len(u.atoms)
print >> sys.stderr, '- Num. bonds:', len(u.bonds)
num_frames = len(u.trajectory)
print >> sys.stderr, '- Num. frames:', num_frames
# construct atoms
print >> sys.stderr, '- Building atoms ...'
waters = {}
i = 0
for md_atom in u.atoms:
atom = cdf_mol.addAtom()
sym = MDAnalysis.topology.guessers.guess_atom_element(md_atom.name)
Chem.setSymbol(atom, sym.title())
Chem.setImplicitHydrogenCount(atom, 0)
Biomol.setChainID(atom, md_atom.segid)
if md_atom.resname == 'WAT':
Biomol.setResidueCode(atom, 'HOH')
else:
Biomol.setResidueCode(atom, md_atom.resname)
if Biomol.getResidueCode(atom) == 'HOH':
if md_atom.resid in waters:
waters[md_atom.resid].append(i)
else:
waters[md_atom.resid] = [i]
Biomol.setResidueSequenceNumber(atom, int(md_atom.resid))
Biomol.setResidueAtomName(atom, md_atom.name)
# fix positive charge on arginin nitrogen
if md_atom.resname == 'ARG' and md_atom.name == 'NH2':
Chem.setFormalCharge(atom, 1)
coords = []
for coord in md_atom.position:
coords.append(float(coord))
Chem.set3DCoordinates(atom, coords)
coords_array = Math.Vector3DArray()
coords_array.reserve(num_frames)
Chem.set3DCoordinatesArray(atom, coords_array)
Chem.setPEOECharge(atom, float(md_atom.charge))
i += 1
Chem.setAtomTypesFromSymbols(cdf_mol, True)
# construct bonds
print >> sys.stderr, '- Building bonds ...'
for md_bond in u.bonds:
cdf_mol.addBond(int(md_bond.atoms[0].index),
int(md_bond.atoms[1].index))
print >> sys.stderr, '- Building water atom bonds ...'
for water in waters.values():
if len(water) < 2:
continue
for atom_idx in water:
if Chem.getType(cdf_mol.atoms[atom_idx]) == Chem.AtomType.O:
if atom.numBonds > 1:
break
for atom_idx2 in water:
if Chem.getType(
cdf_mol.atoms[atom_idx2]) == Chem.AtomType.H:
cdf_mol.addBond(atom_idx, atom_idx2)
break
# make sane biomolecule
Chem.perceiveSSSR(cdf_mol, True)
Chem.setRingFlags(cdf_mol, True)
Chem.perceiveBondOrders(cdf_mol, True)
Chem.perceiveHybridizationStates(cdf_mol, True)
Chem.setAromaticityFlags(cdf_mol, True)
Chem.calcFormalCharges(cdf_mol, True)
# read timsteps and write cdf
print >> sys.stderr, '- Importing coordinates ...'
i = 0
traj_coords = []
atom_coords = Math.Vector3D()
for ts in u.trajectory:
print >> sys.stderr, '- Processing time step', i, '...'
for md_atom in u.atoms:
del traj_coords[:]
for coord in md_atom.position:
traj_coords.append(float(coord))
coords_array = Chem.get3DCoordinatesArray(
cdf_mol.getAtom(int(md_atom.index)))
atom_coords[0] = traj_coords[0]
atom_coords[1] = traj_coords[1]
atom_coords[2] = traj_coords[2]
coords_array.addElement(atom_coords)
i += 1
print >> sys.stderr, '- Writing output file:'
if not Chem.FileCDFMolecularGraphWriter(sys.argv[3]).write(cdf_mol):
print >> sys.stderr, '!! Could not write output file'
sys.exit(2)
