def save_merged_ligand_protein(self, raw_protein_file, raw_ligand_file):
"""
Merge structures from files and save to save directory
:param raw_ligand_file: absolute paths
:return:
"""
prot_st = next(StructureReader(raw_protein_file))
alpha = 'ABCDEFGHIJKMNOPQRST'
alpha_count = 0
for c in prot_st.chain:
if c.name.strip() == '': continue
c.name = alpha[alpha_count]
alpha_count += 1
if raw_ligand_file != '':
lig_st = next(StructureReader(raw_ligand_file))
# standardize chain naming
for c in lig_st.chain:
c.name = 'L'
merged_st = lig_st.merge(prot_st)
else:
merged_st = prot_st
st_wr = StructureWriter(self.path + self.raw_complex)
st_wr.append(merged_st)
st_wr.close()
def get_residues_from_radius(radius, ref):
asl = 'fillres ((within %i ligand ) and not (res.ptype "T3P"))' % (int(float(radius)))
st=StructureReader(ref).next()
atoms=evaluate_asl(st, asl)
if len(atoms)==0:
print "NO LIGAND IN REFERENCE FILE"
print "no ligand in reference file"
sys.exit()
pocket=st.extract(atoms)
resnames=[i.pdbres for i in pocket.residue]
resnums=[i.resnum for i in pocket.residue]
chains=[i.chain for i in pocket.residue]
data=dict()
for chain in set(chains):
for (n,c) in zip(resnums, chains):
if c not in data.keys():
data[c]=[]
if str(n) not in data[c]:
data[c].append(str(n))
selection=[]
for chain in data.keys():
selection.append(','.join(data[chain]) + '.%s' % chain)
if len(selection)==1:
selection=[selection,]
return selection
def write_grid_in_file(self,
mode,
grid_ligand='',
xyz=False,
receptor_file=''):
if mode == 'other_ligand':
st_2 = next(StructureReader(grid_ligand))
c2 = get_centroid(st_2)
x, y, z = c2[:3]
elif mode == 'xyz':
x, y, z = xyz
else:
st_2 = next(StructureReader(self.path + self.split_ligand))
c2 = get_centroid(st_2)
x, y, z = c2[:3]
if receptor_file == '':
use_file = self.split_protein
else:
use_file = receptor_file
with open(self.path + self.grid_in, 'w') as f:
f.write('GRID_CENTER {},{},{}\n'.format(x, y, z))
f.write('GRIDFILE {}\n'.format(self.grid_file))
f.write('INNERBOX 15,15,15\n')
f.write('OUTERBOX 30,30,30\n')
f.write('RECEP_FILE {}\n'.format(use_file))
def get_residues_from_radius(radius, ref):
asl = 'fillres ((within %i ligand ) and not (res.ptype "T3P"))' % (int(
float(radius)))
st = StructureReader(ref).next()
atoms = evaluate_asl(st, asl)
if len(atoms) == 0:
print "NO LIGAND IN REFERENCE FILE"
print "no ligand in reference file"
sys.exit()
pocket = st.extract(atoms)
resnames = [i.pdbres for i in pocket.residue]
resnums = [i.resnum for i in pocket.residue]
chains = [i.chain for i in pocket.residue]
data = dict()
for chain in set(chains):
for (n, c) in zip(resnums, chains):
if c not in data.keys():
data[c] = []
if str(n) not in data[c]:
data[c].append(str(n))
selection = []
for chain in data.keys():
selection.append(','.join(data[chain]) + '.%s' % chain)
if len(selection) == 1:
selection = [
selection,
]
return selection
def extract_prot_lig(self, structure_set_info):
'''
Setup and start running a set of docking task
structure_set_info: (list of dicts)
[{'folder', 'pdb', 'prot':['A','B'], 'lig':'L'}]
:return: (None)
for structure in structure_set_info:
with StructureReader(input_pdb) as st:
st = list(st)[0]
'''
print(structure_set_info)
for task in structure_set_info:
base = task['folder'] + '/' + task['name']
if task['method'] == 'by_chain':
#extract specific chains
with StructureReader(base + '.pdb') as st:
st = list(st)[0]
st_prot = st.chain[task['P']].extractStructure(
copy_props=True)
st_prot.write(base + '_prot.mae')
st_lig = st.chain[task['L']].extractStructure(
copy_props=True)
st_lig.write(base + '_lig.mae')
if task['method'] == 'by_res':
#extract specific chains
with StructureReader(base + '.pdb') as st:
st = list(st)[0]
lig = [
a.index for a in st.atom
if a.getResidue().pdbres == task['L']
]
print(lig)
st_lig = st.extract(lig)
st_lig.write(base + '_lig.mae')
st_prot = st.extract([
a.index for a in st.atom
if a.getResidue().pdbres != task['L']
])
st_prot.write(base + '_prot.mae')
if task['method'] == 'from_PDBBind':
print(task['method'])
name_lower = task['name'].lower()
with StructureReader(task['folder'] + '/' + name_lower +
'_protein.pdb') as st:
st = list(st)[0]
st.write(base + '_prot.mae')
with StructureReader(task['folder'] + '/' + name_lower +
'_ligand.mol2') as st:
st = list(st)[0]
st.write(base + '_lig.mae')
return False
def create_feature_vector(protein, ligand, pickle_file, combind_root):
ending_1 = '{}/structures/aligned_files/{}/{}_out.mae'.format(
protein, ligand, ligand)
s = list(StructureReader(combind_root + ending_1))[0]
rot_s = list(StructureReader(combind_root + '/' + ending_1))[0]
residues = get_all_res(s, rot_s)
print(residues)
outfile = open(pickle_file, 'wb')
pickle.dump(residues, outfile)
outfile.close()
def check_selection_boolean(ref, target):
st=StructureReader(ref).next()
atoms=evaluate_asl(st, 'all')
sub=st.extract(atoms)
resnames=[i.pdbres.rstrip() for i in sub.residue]
if target in resnames:
verdict=True
else:
verdict=False
return verdict
def check_selection_boolean(ref, target):
st = StructureReader(ref).next()
atoms = evaluate_asl(st, 'all')
sub = st.extract(atoms)
resnames = [i.pdbres.rstrip() for i in sub.residue]
if target in resnames:
verdict = True
else:
verdict = False
return verdict
def load_pdb_struct(path,residue_cls = BaseResidue):
st = StructureReader(path).next()
st = mystructure(st)
if residue_cls is not None:
residues = map(residue_cls,st.residue)
#use custom Residue class
st.residues = residues
st.src = path
return st
def compute_protein_mcss(protein, ligands, save_folder):
init_file = '{}/{}_mcss.csv'.format(save_folder, protein)
for i in range(len(ligands)):
for j in range(i + 1, len(ligands)):
l1, l2 = ligands[i], ligands[j]
l1_path = '/scratch/PI/rondror/combind/bpp_data/MAPK14/ligands/prepared_ligands/{}_lig/{}_lig.mae'.format(
l1, l1)
l2_path = '/scratch/PI/rondror/combind/bpp_data/MAPK14/ligands/prepared_ligands/{}_lig/{}_lig.mae'.format(
l2, l2)
mcss_types_file = 'mcss_type_file.typ'
mcss = MCSS(l1, l2)
with StructureReader(l1_path) as ligand1, StructureReader(
l2_path) as ligand2:
ligands = {l1: next(ligand1), l2: next(ligand2)}
mcss.compute_mcss(ligands, init_file, mcss_types_file)
def generate_tree(maefile, origin, charge):
"""
A helper function to generate tree structure from a given maefile and origin
:type maefile: str
:param maefile: name of the maefile, endwith .mae
:type origin: int
:param origin: atom of interest
:type charge: bool
:param charge: use ESP charge for atomic descriptor
return tree structure
"""
print(f'Processing {maefile}')
st = next(StructureReader(maefile))
sorted_atoms = sort_atom_by_atomic(st)
print(f'Using atom {origin} ({st.atom[origin].element}) as origin.\n')
if charge:
try:
root = Node(None, origin, st.atom[origin].element,
st.atom[origin].property['r_j_ESP_Charges'], None,
set())
except:
sys.exit(
'Warning: You need to have ESP charge in the maefile in order to run charge option!!!'
)
else:
root = Node(None, origin, st.atom[origin].element, 0, None, set())
get_stereomatic_desc(st, root, sorted_atoms, charge)
return root
def main():
args = parse_args()
print("args", args)
if args.implementation == TORCHANI:
from torchani_calculator import torchani_calculator
calculator = torchani_calculator(args.network_type, args.numb_networks)
elif args.implementation == AES_ANI:
from ani_ase import ani_ase_calculator
calculator = ani_ase_calculator(args.network_type)
elif args.implementation == KHAN:
from khan_calculator import khan_calculator
calculator = khan_calculator(args.network_type, args.khan_network,
args.numb_networks)
strs = list(StructureReader(args.mae_file))
cvgd = optimize_structures(strs,
calculator.committee,
maxit=500,
thresh=0.01)
rd = rawdataset_from_st(strs)
energy_data = calculator.committee.compute_data(rd)
for st, data in zip(strs, energy_data):
st.property["r_j_ANI_ENERGY"] = data.energy
st.property["r_j_ANI_RHO"] = data.rho
base, ext = os.path.splitext(args.mae_file)
outfile = base + "_optimized" + ext
print("final structure written to file: %s" % outfile)
with StructureWriter(outfile) as writer:
writer.extend(strs)
def molecule_to_schrodinger_struct(molecule: Molecule):
"""
Convert a pymatgen Molecule object to a Schrodinger Structure object
Args:
molecule (pymatgen.core.structure.Molecule): Molecule to be converted
Returns:
struct: schrodinger.structure.Structure object
"""
# First need to convert molecule to file to use StructureReader
file_suffix = random.randint(1, 100000000)
molecule.to('sdf', "temp_conversion{}.sdf".format(file_suffix))
reader = StructureReader("temp_conversion{}.sdf".format(file_suffix))
# Assume only one structure (should be the case for a single SDF file)
struct = [r for r in reader][0]
for atom in struct.atom:
atom.formal_charge = 0
struct.atom[1].formal_charge = molecule.charge
Path("temp_conversion{}.sdf".format(file_suffix))
return struct
def aggregate_structures(data_path, save_location, scores_version):
protein = 'MAPK14'
print(protein)
#try to load the score file
scores_file = "{}/{}/scores/{}/pdb.sc".format(data_path, protein,
scores_version)
if os.path.isfile(scores_file):
ligand_folder = '{}/{}/structures/aligned_files'.format(
data_path, protein)
ligands = os.listdir(ligand_folder)
# write one mae files for all poses
with StructureWriter('{}/all_{}_pv.mae'.format(save_location,
protein)) as all:
#iterate over each scored ligand
for i, ligand in enumerate(ligands):
pose_file = "{}/{}/structures/aligned_files/{}/{}_out.mae".format(
data_path, protein, ligand, ligand)
#load the structure file containing docking results
pv = list(StructureReader(pose_file))
print(pv)
#add the protein structure but only for the first file
all.append(pv[0])
def compute_protein_mcss(ligands, pair_path):
init_file = '{}/{}-to-{}_mcss.csv'.format(pair_path, ligands[0],
ligands[1])
for i in range(len(ligands)):
for j in range(i + 1, len(ligands)):
l1, l2 = ligands[i], ligands[j]
l1_path = '{}/ligand_poses/{}_lig0.mae'.format(pair_path, l1)
new_l1_path = create_mcss_file(l1_path, l1, pair_path)
l2_path = '{}/{}_lig.mae'.format(pair_path, l2)
new_l2_path = create_mcss_file(l2_path, l2, pair_path)
mcss_types_file = 'mcss_type_file.typ'
mcss = MCSS(l1, l2)
with StructureReader(new_l1_path) as ligand1, StructureReader(
new_l2_path) as ligand2:
ligands = {l1: next(ligand1), l2: next(ligand2)}
mcss.compute_mcss(ligands, init_file, mcss_types_file)
os.system('rm {} {}'.format(new_l1_path, new_l2_path))
def _show_schrodinger_file(path, **kwargs):
from schrodinger.structure import StructureReader
view = NGLWidget()
cts = list(StructureReader(path))
for ct in cts:
view.add_component(SchrodingerStructure(ct), **kwargs)
return view
def load_poses(self, maxposes=10):
poses = []
for i, st in enumerate(StructureReader(self.folder+'/'+self.pose_viewer_file_name)):
if i > maxposes: break
if i == 0:
prot_st = st
continue
poses.append(st)
return prot_st, poses
def get_sequence_from_str(file):
s = list(StructureReader(file))[0]
str = ''
for m in list(s.molecule):
if len(m.residue) != 1:
for r in list(m.residue):
atom = list(r.atom)[0]
if (atom.chain == "A"): # to fix for MAPK14: 3GCU
str += atom.pdbcode
return str
def get_sequence_from_str(file, chains, protein):
s = list(StructureReader(file))[0]
str = ''
for m in list(s.molecule):
if len(m.residue) != 1:
for r in list(m.residue):
atom = list(r.atom)[0]
if atom.chain == chains[protein]:
str += atom.pdbcode
return str
def main():
coms = {}
with open(prot_file) as fp:
for line in tqdm(fp, desc='protein file'):
if line[0] == '#': continue
protein, target, start = line.strip().split()
if protein not in coms:
coms[protein] = {}
protein_folder = os.path.join(data_root,
protein + '/structures/aligned')
if start not in coms[protein]:
file = os.path.join(protein_folder, start + '_prot.mae')
s = list(StructureReader(file))[0]
coms[protein][start] = analyze.center_of_mass(s)
file = os.path.join(protein_folder, start + '_prot.mae')
s = list(StructureReader(file))[0]
coms[protein][start] = analyze.center_of_mass(s)
with open('com_after_aligned.pkl', 'wb') as f:
pickle.dump(coms, f)
def docking(grid):
ligands = os.listdir(grid)
similarities = []
lgr = logger()
cfg = CanvasFingerprintGenerator(lgr, 'Radial')
cfs = CanvasFingerprintSimilarity(lgr)
cfs.setMetric('Tanimoto')
for struc_ligand in ligands:
for ligand in ligands:
prepped_ligand_file = '/scratch/PI/rondror/combind/bpp_data/MAPK14/ligands/prepared_ligands/{}_lig/{}_lig.mae'.format(
ligand, ligand)
prepped_struc_ligand_file = '/scratch/PI/rondror/combind/bpp_data/MAPK14/ligands/prepared_ligands/{}_lig/{}_lig.mae'.format(
struc_ligand, struc_ligand)
lig = cfg.generate(next(StructureReader(prepped_ligand_file)))
struc_lig = cfg.generate(
next(StructureReader(prepped_struc_ligand_file)))
similarities.append(cfs.calculateSimilarity(lig, struc_lig))
outfile = open('/home/users/sidhikab/MAPK14_similarities', 'wb')
pickle.dump(similarities, outfile)
outfile.close()
def find_chain(protein, chains, ligand):
protein_folder = os.path.join(data_root, protein + '/structures/aligned')
primary_file = os.path.join(protein_folder, ligand + '_prot.mae')
s = list(StructureReader(primary_file))[0]
chain_list = []
for m in list(s.molecule):
for r in list(m.residue):
atom = list(r.atom)[0]
if atom.chain not in chain_list:
chain_list.append(atom.chain)
if protein not in chains:
chains[protein] = chain_list
else:
chains[protein] = set.intersection(set(chains[protein]),
set(chain_list))
def main():
coms = {}
with open(prot_file) as fp:
for line in tqdm(fp, desc='protein file'):
if line[0] == '#': continue
protein, target, start = line.strip().split()
if protein not in coms:
coms[protein] = {}
if start not in coms[protein]:
start_receptor_file = os.path.join(
data_root,
'{}/structures/aligned/{}_prot.mae'.format(protein, start))
start_ligand_file = os.path.join(
data_root,
'{}/structures/aligned/{}_lig.mae'.format(protein, start))
start_struct = list(StructureReader(start_receptor_file))[0]
start_lig = list(StructureReader(start_ligand_file))[0]
# print(protein, start)
coms[protein][start] = get_pocket_res(start_struct, start_lig)
if target not in coms[protein]:
target_receptor_file = os.path.join(
data_root, '{}/structures/aligned/{}_prot.mae'.format(
protein, target))
target_ligand_file = os.path.join(
data_root,
'{}/structures/aligned/{}_lig.mae'.format(protein, target))
target_struct = list(StructureReader(target_receptor_file))[0]
target_lig = list(StructureReader(target_ligand_file))[0]
# print(protein, target)
get_pocket_res(target_struct, target_lig)
with open('pocket_com.pkl', 'wb') as f:
pickle.dump(coms, f)
def main():
maefile = sys.argv[1]
origin = int(sys.argv[2])
print(f'Processing {maefile}')
st = next(StructureReader(maefile))
print(f'Using atom {origin} ({st.atom[origin].element}) as origin.\n')
# Index of atom origin of stereomatic network
sd = get_stereomatic_desc(st, origin, set([origin]))
print(sd)
with open(maefile.replace('.mae', '.json'), 'w') as fh:
json.dump(sd, fh, indent=4)
def _set_ligand_sizes(self, structure_file):
try:
refs = [st for st in StructureReader(structure_file)]
except:
print('Unable to read MCSS structure file for', self.l1, self.l2)
return None
if len(refs) != 2:
print('Wrong number of structures', self.l1, self.l2)
return None
ref1, ref2 = refs
n_l1_atoms = len([a for a in ref1.atom if a.element != 'H'])
n_l2_atoms = len([a for a in ref2.atom if a.element != 'H'])
self.n_l1_atoms = n_l1_atoms
self.n_l2_atoms = n_l2_atoms
def run_group(grouped_files, raw_root, index, clash_dir):
"""
checks mean distance of displacement for decoys for each protein, target, start group
:param grouped_files: (list) list of protein, target, start groups
:param raw_root: (string) directory where raw data will be placed
:param index: (int) group number
:param dist_dir: (string) directiory to place distances
:param max_poses: (int) maximum number of glide poses considered
:param max_decoys: (int) maximum number of decoys created per glide pose
:return:
"""
clash_dict = {}
for protein, target, start in grouped_files[index]:
protein_path = os.path.join(raw_root, protein)
pair_path = os.path.join(protein_path, '{}-to-{}'.format(target, start))
pose_path = os.path.join(pair_path, 'ligand_poses')
struct_path = os.path.join(pair_path, '{}_prot.mae'.format(start))
lig_path = os.path.join(pose_path, '{}_lig0.mae'.format(target))
s1 = list(StructureReader(struct_path))[0]
lig = list(StructureReader(lig_path))[0]
clash_dict[(protein, target, start)] = steric_clash.clash_volume(s1, struc2=lig)
outfile = open(os.path.join(clash_dir, '{}.pkl'.format(index)), 'wb')
pickle.dump(clash_dict, outfile)
def file_to_schrodinger_structure(filename: Union[str, Path]):
"""
Convert a file (sdf, pdb, sd, mol2, maestro, or maestro_text) to a
Schrodinger Structure object.
Args:
filename (str): Name of the file
Returns:
structures (list of schrodinger.structure.Structure objects)
"""
if isinstance(filename, Path):
fn = filename.as_posix()
else:
fn = filename
reader = StructureReader(fn)
structures = [s for s in reader]
return structures
def run_group(grouped_files, raw_root, index):
for protein, target, start in grouped_files[index]:
pair = '{}-to-{}'.format(target, start)
protein_path = os.path.join(raw_root, protein)
pair_path = os.path.join(protein_path, pair)
pose_path = os.path.join(pair_path, 'ligand_poses')
# comput mcss
compute_protein_mcss([target, start], pair_path)
# combine ligands
with StructureWriter('{}/{}_merge_pv.mae'.format(pair_path,
pair)) as all_file:
for file in os.listdir(pose_path):
if file[-3:] == 'mae':
pv = list(StructureReader(os.path.join(pose_path, file)))
all_file.append(pv[0])
# zip file
os.system('gzip -f {}/{}_merge_pv.mae'.format(pair_path, pair,
pair_path, pair))
def split(self):
#look for the opt_complex, if it doesn't exist, just use the prepped complex
usefile = ''
if (os.path.isfile(self.path + self.align_file)):
usefile = self.path + self.align_file
else:
usefile = self.path + self.prepped_complex
st = next(StructureReader(usefile))
prot_st = st.extract([a.index for a in st.atom if a.chain != 'L'])
prot_st.title = '{}_prot'.format(self.name)
lig_st = st.extract([a.index for a in st.atom if a.chain == 'L'])
lig_st.title = '{}_lig'.format(self.name)
prot_wr = StructureWriter(self.path + self.split_protein)
prot_wr.append(prot_st)
prot_wr.close()
lig_wr = StructureWriter(self.path + self.split_ligand)
lig_wr.append(lig_st)
lig_wr.close()
def main():
args = parse_args()
print("args", args)
if args.implementation == TORCHANI:
from torchani_calculator import torchani_calculator
calculator = torchani_calculator(args.network_type, args.numb_networks)
elif args.implementation == AES_ANI:
from ani_ase import ani_ase_calculator
calculator = ani_ase_calculator(args.network_type)
elif args.implementation == KHAN:
from khan_calculator import khan_calculator
calculator = khan_calculator(args.network_type, args.khan_network,
args.numb_networks)
#assert args.cif_file.endswith('.cif') or args.cif_file.endswith('.xyz')
if args.cif_file.endswith('.mae'):
st = next(StructureReader(args.cif_file))
if args.truncate_box:
# reduced size box
a = 12.0
deletions = []
for mol in st.molecule:
rcom = center_of_mass(st, mol.getAtomIndices())
inbox = [abs(x) < a / 2 for x in rcom]
if False in [abs(x) < a / 2 for x in rcom]:
deletions.extend(mol.getAtomIndices())
st.deleteAtoms(deletions)
rd = rawdataset_from_st([st])
ase_molecs = khan_molec_to_ase_atoms(rd.all_Xs)
atoms = ase_molecs[0]
if args.truncate_box:
# hard coded box
#a = 22.6868
vecs = [[a, 0.0, 0.0], [0.0, a, 0.0], [0.0, 0.0, a]]
atoms.set_cell(vecs)
atoms.set_pbc([True, True, True])
print("after set cell", atoms.get_pbc())
if args.cif_file.endswith('.cif'):
atoms = ase_io.read(args.cif_file)
if args.box_length is not None:
atoms.set_cell([
[args.box_length, 0.0, 0.0],
[0.0, args.box_length, 0.0],
[0.0, 0.0, args.box_length],
])
atoms.set_pbc([True, True, True])
if args.supercell is not None:
atoms = build_supercell(atoms, [args.supercell] * 3)
periodic = False not in atoms.get_pbc()
print("cell")
print(atoms.get_cell())
if periodic:
#atoms.wrap()
space_group = spacegroup.get_spacegroup(atoms)
print("Space group of crystal: %s" % space_group)
print("initial unit cell")
print(atoms.cell)
# this shows how to get unique atoms and there equivalents
#scaled_positions = atoms.get_scaled_positions()
#unique_positions = space_group.unique_sites(scaled_positions)
#all_positions, symmetry_map = space_group.equivalent_sites(unique_positions)
#symmetry_groups = defaultdict(list)
#for igroup, position in zip(symmetry_map, all_positions):
# symmetry_groups[igroup].append(position)
#print("unique positions")
#for xyz in unique_positions:
# print(xyz)
#for igroup in sorted(symmetry_groups.keys()):
# print("positions in symmetry group %d" % igroup)
# for xyz in symmetry_groups[igroup]:
# print(xyz)
torsions = ()
if args.torsion is not None:
torsions = [(
float(args.torsion[0]),
int(args.torsion[1]),
int(args.torsion[2]),
int(args.torsion[3]),
)]
if args.optimize:
# cannot optimize cell until we implement a stress tensor
energy = optimize_molecule(
atoms,
calculator,
torsions=torsions,
thresh=0.05,
optimize_cell=args.relax_cell,
)
else:
start_time = time.time()
energy = single_point_energy(atoms, calculator)
print("--- %s seconds ---" % (time.time() - start_time))
if args.dynamics:
traj = molecular_dynamics(atoms,
calculator,
total_time=args.total_time,
time_step=args.time_step)
if args.cif_file.endswith('.mae'):
st = next(StructureReader(args.cif_file))
with StructureWriter("md_path.mae") as writer:
for m in traj:
st0 = st.copy()
st0.setXYZ(m.get_positions())
writer.append(st0)
else:
with open("md_path.xyz", "w") as fout:
xyz_lst = []
for atoms in traj:
xyz_lst.append("%d\n" % len(atoms))
species = atoms.get_chemical_symbols()
carts = atoms.get_positions()
for ch, xyz in zip(species, carts):
xyz_lst.append("%s %.8f %.8f %.8f" %
(ch, xyz[0], xyz[1], xyz[2]))
fout.write("\n".join(xyz_lst))
print("energy of cell (au):", energy)
if periodic:
space_group = spacegroup.get_spacegroup(atoms)
print("Final Space group of crystal: %s" % space_group)
print("energy of cell/volume (au):", energy / atoms.get_volume())
print("final unit cell")
print(atoms.cell)
print("Uncertainty %.4f (kcal/mol)" %
(calculator.uncertainty(atoms) * 627.509))
if periodic:
outfile = os.path.splitext(args.cif_file)[0] + "_optimized.cif"
else:
outfile = os.path.splitext(args.cif_file)[0] + "_optimized.xyz"
ase_io.write(outfile, atoms)
print("final structure written to file: %s" % outfile)
def read_pv_file(pvfile):
if not fileutils.is_poseviewer_file(pvfile):
raise PoseViewerFileInvalidError(pvfile)
reader = StructureReader(pvfile)
return (Receptor(reader.next()),
tuple(Pose(st, i) for i, st in enumerate(reader)))
for i, group in enumerate(grouped_files):
print('making grid', i)
with open('{}/run/grid{}_in.sh'.format(save, i), 'w') as f:
for s_file in group:
out_f = s_file[:12]
os.system('mkdir -p {}/{}'.format(save, out_f))
with open('{}/{}/{}.in'.format(save, out_f, out_f),
'w') as f_in:
if len(s_file) != 16:
continue
s = next(StructureReader(ligands + s_file[:4] +
'_lig.mae'))
c = get_centroid(s)
x, y, z = c[:3]
f_in.write('GRID_CENTER {},{},{}\n'.format(x, y, z))
f_in.write('GRIDFILE {}.zip\n'.format(out_f))
f_in.write('INNERBOX 15,15,15\n')
f_in.write('OUTERBOX 30,30,30\n')
f_in.write('RECEP_FILE {}/{}\n'.format(root, s_file))
f.write('#!/bin/bash\n')
f.write('cd {}/{}\n'.format(save, out_f))
f.write('$SCHRODINGER/glide -WAIT {}.in\n'.format(out_f))
os.chdir('{}/run'.format(save))
os.system(
'sbatch -p owners -t 02:00:00 -o grid{}.out grid{}_in.sh'.format(
def parse(pdbfile):
list(StructureReader(pdbfile))
return complete_a_list
if __name__ == '__main__':
folder = '/scratch/PI/rondror/combind/bpp_data/MAPK14/structures/aligned_files/'
ligands = os.listdir(folder)
ligands.remove("4DLI")
infile = open('MAPK14_pairwise_alignment', 'rb')
paired_strs = pickle.load(infile)
infile.close()
rmsds = []
for i in range(len(ligands)):
ending_1 = '{}/{}_out.mae'.format(ligands[i], ligands[i])
s1 = list(StructureReader(folder + ending_1))[0]
arr = []
for j in range(len(ligands)):
#to monitor progress
print(i, ligands[i], j, ligands[j])
ending_2 = '{}/{}_out.mae'.format(ligands[j], ligands[j])
s2 = list(StructureReader(folder + ending_2))[0]
(paired_str_s1, paired_str_s2) = paired_strs[i][j]
r_list_s1 = get_all_res(s1)
r_list_s2 = get_all_res(s2)
r_to_i_map_s1 = map_residues_to_align_index(
