def process(molid, smiles):
if molid is _END_MOLID:
writer.close()
return
try:
mol = to_rdkit_mol(smiles)
fpsinfo = {}
# N.B. We won't actually use rdkit hash, so we won't ask for nonzero values...
# Is there a way of asking rdkit to give us this directly?
AllChem.GetMorganFingerprint(mol, max_radius, bitInfo=fpsinfo, useFeatures=fcfp)
counts = defaultdict(int)
centers = defaultdict(list)
for bit_descs in fpsinfo.values():
for center, radius in bit_descs:
cansmiles = explain_circular_substructure(mol, center, radius)
counts[cansmiles] += 1
centers[cansmiles].append((center, radius))
if write_centers:
features_strings = ['%s %d %s' % (cansmiles,
count,
' '.join(['%d %d' % (c, r) for c, r in centers[cansmiles]]))
for cansmiles, count in counts.iteritems()]
else:
features_strings = ['%s %d' % (cansmiles, count) for cansmiles, count in counts.iteritems()]
writer.write('%s\t%s\n' % (molid, '\t'.join(features_strings)))
except:
info('Failed molecule %s: %s' % (molid, smiles))
writer.write('%s\t*FAILED*\n' % molid)
def process(molid, smiles):
if molid is _END_MOLID:
h5.close()
return
ne = len(molids)
try:
molids.resize((ne + 1,))
molids[ne] = molid
mol = to_rdkit_mol(smiles)
descs.resize((ne + 1, nf))
descs[ne, :] = computer.compute(mol)[0]
except:
info('Failed molecule %s: %s' % (molid, smiles))
descs[ne, :] = [np.nan] * nf
def save_from_smiles_iterator(self, it):
"""Creates the catalog from the (molid, smiles) iterator, possibly overwriting the present files."""
molids = []
coords = []
base = 0
with open(op.join(self._root, 'molsdata'), 'wb') as writer:
for molid, smiles in it:
mol = to_rdkit_mol(smiles, molid=molid)
if mol is None:
molids.append(molid)
coords.append((-1, 0))
else:
moldata = mol.ToBinary()
molids.append(molid)
coords.append((base, len(moldata)))
base += len(moldata)
writer.write(moldata)
with open(self._molids_file, 'wt') as writer:
for molid in molids:
writer.write(molid + '\n')
np.save(self._coords_file, np.array(coords))
