def get_rmsd(smiles, pdb_path, ref_pdb_path):
smiles_mol = Chem.MolFromSmiles(smiles)
ref_mol = Chem.MolFromPDBFile(ref_pdb_path)
mol = Chem.MolFromPDBFile(pdb_path)
ref_mol = AllChem.AssignBondOrdersFromTemplate(smiles_mol,ref_mol)
mol = AllChem.AssignBondOrdersFromTemplate(smiles_mol, mol)
order = list(mol.GetSubstructMatches(ref_mol)[0])
mol = Chem.RenumberAtoms(mol, order)
indices = cpeptools.get_largest_ring(ref_mol)
assert len(set(indices) - set(cpeptools.get_largest_ring(mol))) == 0, "ring atom indices do not agree"
tmp_dir = tempfile.mkdtemp()
ref_pdb_filename = tempfile.mktemp(suffix=".pdb", dir = tmp_dir)
pdb_filename = tempfile.mktemp(suffix=".pdb", dir = tmp_dir)
Chem.MolToPDBFile(ref_mol, ref_pdb_filename)
Chem.MolToPDBFile(mol, pdb_filename)
ref = md.load(ref_pdb_filename)
compare = md.load(pdb_filename)
rmsd = md.rmsd(compare, ref, 0)
ring_rmsd = md.rmsd(compare, ref, 0, atom_indices = indices)
compare = compare.superpose(ref, 0, atom_indices = indices)
return rmsd, compare[np.argmin(rmsd)] , ring_rmsd, compare[np.argmin(ring_rmsd)] , ref
num_conf_list = []
mol_list = []
tmp = sys.argv[0][:-3]
tmp = tmp.split(".")[-1]
folder = "./" + tmp #remove the number at the beginning for folder names, this way scripts all starts with number, while folders do not
df = pd.read_csv("../reference/smiles.tsv", sep="\t", comment="%")
mol_dict = {
name: Chem.MolFromSmiles(val)
for name, val in zip(df.Name, df.Smiles)
}
for name in mol_dict:
print(name)
rsize = len(cpeptools.get_largest_ring(mol_dict[name]))
ref = mol_dict[name] #ref is mol from smiles
mol = ref
try:
mol = Chem.AddHs(mol)
except ValueError:
print("cannot add H to {}".format(name))
continue
mol.UpdatePropertyCache()
Chem.GetSymmSSSR(mol)
mol_list.append(mol)
path = folder + "/" + str(name) + "/"
path_list.append(path)
# this code generates different number of conformers given then ring size