def main():
'''
'''
options = parse_options()
# THIS OPTION WILL PRODUCE MORE VERBOSE OUTPUT
if options.debug: logger.setLevel(logging.DEBUG)
if options.output: fh = open(options.output,'wb')
else: fh = sys.stdout
# CHOOSE HOW THE OUPTPUT DATA WILL BE WRITTEN
if options.format == 'csv':
writer = csv.writer(fh, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
HEADER = True
# iterate through all mol2 files in inputdir
for mol2file in glob.glob(os.path.join(options.inputligdir, '*.mol2')):
#lig_path = os.path.join(option.inputdir, file)
lig_path = mol2file
protein_path = options.inputpdb
if not os.path.isfile(protein_path):
logger.fatal("The protein file does not exist.".format(options.inputpdb))
sys.exit(1)
protein = get_molecule(protein_path)
ligand = get_molecule(lig_path)
# calculate descriptor based on the sum of interacting element pairs
if options.descriptor == 'elements':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.element_descriptor(protein, ligand,
binsize=options.binsize)
# calculate descriptor based on the sum of interacting element pairs
elif options.descriptor == 'sybyl':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.sybyl_atom_type_descriptor(protein, ligand,
binsize=options.binsize)
# calculate descriptor using structural interaction fingerprints
elif options.descriptor == 'credo':
# get the protein-ligand structural interaction fingerprint
descriptor, labels = contacts.sift_descriptor(protein, ligand,
binsize=options.binsize)
if HEADER:
# UPDATE COLUMN LABELS
labels.append('ligand')
writer.writerow(labels)
HEADER = False
if options.format == 'csv':
#ligandname = "\"" + os.path.basename(lig_path).split('.')[0] + "\""
ligandname = os.path.basename(lig_path).split('.')[0]
#print(ligandname)
# FIRST COLUMN OF OUTPUT ROW
row = descriptor.tolist() + [ligandname]
writer.writerow(row)
def main():
'''
'''
options = parse_options()
# THIS OPTION WILL PRODUCE MORE VERBOSE OUTPUT
if options.debug: logger.setLevel(logging.DEBUG)
pdbbindconf = config['standard']
data = parse_index(options.pdbbind, options.index)
if options.output: fh = open(options.output, 'wb')
else: fh = sys.stdout
# CHOOSE HOW THE OUPTPUT DATA WILL BE WRITTEN
if options.format == 'csv':
writer = csv.writer(fh,
delimiter=',',
quotechar='"',
quoting=csv.QUOTE_MINIMAL)
HEADER = True
# ITERATE THROUGH ALL PROTEIN-LIGAND COMPLEXES
for pdb in data:
# NORMALISE ACTIVITY TO NANOMOLAR
pkd = get_pkd(float(data[pdb]['value']), data[pdb]['unit'])
# THE PDBBIND DIRECTORY CONTAINING ALL THE STRUCTURES FOR THIS PDB ENTRY
entry_dir = os.path.join(options.pdbbind, pdb)
# CHECK IF THE DIRECTORY ACTUALLY EXISTS
if not os.path.exists(entry_dir):
logger.error(
"The PDBbind directory for PDB entry {0} does not exist.".
format(pdb))
continue
# CREATE THE PATHS TO THE PROTEIN AND LIGAND USING THE SPECIFIC _<POCKET,PROTEIN,LIGAND,ZINC> LABEL
prot_path = os.path.join(
entry_dir, '{0}_{1}.pdb'.format(pdb, pdbbindconf['protein']))
lig_path = os.path.join(
entry_dir, '{0}_{1}.mol2'.format(pdb, pdbbindconf['ligand']))
if not os.path.exists(prot_path):
logger.error(
"The protein pocket structure for PDB entry {0} cannot be found."
.format(pdb))
continue
elif not os.path.exists(lig_path):
logger.error(
"The ligand structure for PDB entry {0} cannot be found.".
format(pdb))
continue
protein = get_molecule(prot_path)
ligand = get_molecule(lig_path)
# CALCULATE DESCRIPTOR USING STRUCTURAL INTERACTION FINGERPRINTS
if options.descriptor == 'credo':
# GET THE PROTEIN-LIGAND STRUCTURAL INTERACTION FINGERPRINT
descriptor, labels = contacts.sift_descriptor(
protein, ligand, binsize=options.binsize)
# CALCULATE DESCRIPTOR BASED ON THE SUM OF INTERACTING ELEMENT PAIRS
elif options.descriptor == 'elements':
# CALCULATE ELEMENT PAIR DESCRIPTOR FOR THIS COMPLEX
descriptor, labels = contacts.element_descriptor(
protein, ligand, binsize=options.binsize)
# CALCULATE DESCRIPTOR BASED ON THE SUM OF INTERACTING ELEMENT PAIRS
elif options.descriptor == 'sybyl':
# CALCULATE ELEMENT PAIR DESCRIPTOR FOR THIS COMPLEX
descriptor, labels = contacts.sybyl_atom_type_descriptor(
protein, ligand, binsize=options.binsize)
if HEADER:
# UPDATE COLUMN LABELS
labels.insert(0, 'pKd/pKi')
labels.append('pdb')
writer.writerow(labels)
HEADER = False
if options.format == 'csv':
# KEEP ONLY THE TWO MOST SIGNIFICANT BITS
pkdstring = "{0:.2f}".format(pkd)
# FIRST COLUMN OF OUTPUT ROW
row = [pkdstring] + descriptor.tolist() + [pdb]
writer.writerow(row)
def main():
"""
"""
options = parse_options()
# this option will produce more verbose output
if options.debug: logger.setLevel(logging.DEBUG)
csarconf = config['csar']
if options.output: fh = open(options.output, 'wb')
else: fh = sys.stdout
# choose how the ouptput data will be written
if options.format == 'csv':
writer = csv.writer(fh,
delimiter=',',
quotechar='"',
quoting=csv.QUOTE_MINIMAL)
HEADER = True
# iterate through all numbered directories
for directory in os.listdir(csarconf['directory']):
entrydir = os.path.join(csarconf['directory'], directory)
# parse kd.dat to get the pKd
kddat_path = os.path.join(entrydir, 'kd.dat')
# exit if kd.dat is missing
if not os.path.isfile(kddat_path):
logger.fatal(
"CSAR directory {} does not contain kd.dat file.".format(
directory))
sys.exit(1)
entry, pdb, pkd = open(kddat_path).read().strip().replace(
' ', '').split(',')
protein_path = glob.glob(os.path.join(entrydir,
'*_complex.mol2')).pop()
protein = get_molecule(str(protein_path))
ligand = extract_ligand(protein.OBMol)
# calculate descriptor based on the sum of interacting element pairs
if options.descriptor == 'elements':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.element_descriptor(
protein, ligand, binsize=options.binsize)
# calculate descriptor based on the sum of interacting element pairs
elif options.descriptor == 'sybyl':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.sybyl_atom_type_descriptor(
protein, ligand, binsize=options.binsize)
# calculate descriptor using structural interaction fingerprints
elif options.descriptor == 'credo':
# get the protein-ligand structural interaction fingerprint
descriptor, labels = contacts.sift_descriptor(
protein, ligand, binsize=options.binsize)
if HEADER:
# UPDATE COLUMN LABELS
labels.insert(0, 'pKd/pKi')
labels.append('pdb')
writer.writerow(labels)
HEADER = False
if options.format == 'csv':
# FIRST COLUMN OF OUTPUT ROW
row = [pkd] + descriptor.tolist() + [pdb]
writer.writerow(row)
def main():
"""
"""
options = parse_options()
# this option will produce more verbose output
if options.debug: logger.setLevel(logging.DEBUG)
csarconf = config['csar']
if options.output: fh = open(options.output,'wb')
else: fh = sys.stdout
# choose how the ouptput data will be written
if options.format == 'csv':
writer = csv.writer(fh, delimiter=',', quotechar='"',
quoting=csv.QUOTE_MINIMAL)
HEADER = True
# iterate through all numbered directories
for directory in os.listdir(csarconf['directory']):
entrydir = os.path.join(csarconf['directory'], directory)
# parse kd.dat to get the pKd
kddat_path = os.path.join(entrydir, 'kd.dat')
# exit if kd.dat is missing
if not os.path.isfile(kddat_path):
logger.fatal("CSAR directory {} does not contain kd.dat file."
.format(directory))
sys.exit(1)
entry, pdb, pkd = open(kddat_path).read().strip().replace(' ','').split(',')
protein_path = glob.glob(os.path.join(entrydir, '*_complex.mol2')).pop()
protein = get_molecule(str(protein_path))
ligand = extract_ligand(protein.OBMol)
# calculate descriptor based on the sum of interacting element pairs
if options.descriptor == 'elements':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.element_descriptor(protein, ligand,
binsize=options.binsize)
# calculate descriptor based on the sum of interacting element pairs
elif options.descriptor == 'sybyl':
# calculate element pair descriptor for this complex
descriptor, labels = contacts.sybyl_atom_type_descriptor(protein, ligand,
binsize=options.binsize)
# calculate descriptor using structural interaction fingerprints
elif options.descriptor == 'credo':
# get the protein-ligand structural interaction fingerprint
descriptor, labels = contacts.sift_descriptor(protein, ligand,
binsize=options.binsize)
if HEADER:
# UPDATE COLUMN LABELS
labels.insert(0,'pKd/pKi')
labels.append('pdb')
writer.writerow(labels)
HEADER = False
if options.format == 'csv':
# FIRST COLUMN OF OUTPUT ROW
row = [pkd] + descriptor.tolist() + [pdb]
writer.writerow(row)
def main():
'''
'''
options = parse_options()
# THIS OPTION WILL PRODUCE MORE VERBOSE OUTPUT
if options.debug: logger.setLevel(logging.DEBUG)
pdbbindconf = config['pdbbind']
data = parse_index(options.pdbbind, options.index)
if options.output: fh = open(options.output,'wb')
else: fh = sys.stdout
# CHOOSE HOW THE OUPTPUT DATA WILL BE WRITTEN
if options.format == 'csv':
writer = csv.writer(fh, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
HEADER = True
counter = 0
# ITERATE THROUGH ALL PROTEIN-LIGAND COMPLEXES
for pdb in data:
# THE PDBBIND DIRECTORY CONTAINING ALL THE STRUCTURES FOR THIS PDB ENTRY
entry_dir = os.path.join(options.pdbbind,pdb)
# CHECK IF THE DIRECTORY ACTUALLY EXISTS
if not os.path.exists(entry_dir):
logger.error("The PDBbind directory for PDB entry {0} does not exist.".format(pdb))
continue
# CREATE THE PATHS TO THE PROTEIN AND LIGAND USING THE SPECIFIC _<POCKET,PROTEIN,LIGAND,ZINC> LABEL
prot_path = os.path.join(entry_dir,'{0}_{1}.pdb'.format(pdb,pdbbindconf['protein']))
ref_lig_path = os.path.join(entry_dir,'{0}_{1}.mol2'.format(pdb,pdbbindconf['ligand']))
#for each protein, the ligand gets generated docking poses from x docking methods,
#
if not os.path.exists(prot_path):
logger.error("The protein pocket structure for PDB entry {0} cannot be found.".format(pdb))
continue
for score in dockingMethods:
pose_path = os.path.join(posesDir, score, pdb)
# \TODO: add pattern for each docking method, right now only works with gold
lig_pattern = "gold_soln"
# RMSD dict for all poses
counter = counter + 1
print("Calculating RMSDs for ligand " + pdb + ", docking method " + score)
RMSDs = calcRMSDPoses(ref_lig_path, pose_path, lig_pattern)
for pose in listFiles(pose_path, lig_pattern):
lig_path = os.path.join(posesDir, score, pdb, pose)
poseRMSD = RMSDs[pose]
poseID = pose.split('.')[0] + '_' + score
if not os.path.exists(lig_path):
logger.error("The ligand structure for PDB entry {0} cannot be found.".format(pdb))
continue
protein = get_molecule(prot_path)
ligand = get_molecule(lig_path)
# CALCULATE DESCRIPTOR USING STRUCTURAL INTERACTION FINGERPRINTS
if options.descriptor == 'credo':
# GET THE PROTEIN-LIGAND STRUCTURAL INTERACTION FINGERPRINT
descriptor, labels = contacts.sift_descriptor(protein, ligand, binsize=options.binsize)
# CALCULATE DESCRIPTOR BASED ON THE SUM OF INTERACTING ELEMENT PAIRS
elif options.descriptor == 'elements':
# CALCULATE ELEMENT PAIR DESCRIPTOR FOR THIS COMPLEX
descriptor, labels = contacts.element_descriptor(protein, ligand, binsize=options.binsize)
# CALCULATE DESCRIPTOR BASED ON THE SUM OF INTERACTING ELEMENT PAIRS
elif options.descriptor == 'sybyl':
# CALCULATE ELEMENT PAIR DESCRIPTOR FOR THIS COMPLEX
descriptor, labels = contacts.sybyl_atom_type_descriptor(protein, ligand, binsize=options.binsize)
if HEADER:
# UPDATE COLUMN LABELS
labels.insert(0,'RMSD')
labels.append('ligandID')
writer.writerow(labels)
HEADER = False
if options.format == 'csv':
# KEEP ONLY THE TWO MOST SIGNIFICANT BITS
#pkdstring = "{0:.2f}".format(pkd)
# FIRST COLUMN OF OUTPUT ROW
row = [poseRMSD] + descriptor.tolist() + [poseID]
writer.writerow(row)