def main():
parser = argparse.ArgumentParser(description='RDKit constrained conformer generator')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-n', '--num', type=int, default=10, help='number of conformers to generate')
parser.add_argument('-r', '--refmol', help="Reference molecule file")
parser.add_argument('--refmolidx', help="Reference molecule index in file", type=int, default=1)
parser.add_argument('-c', '--core_smi', help='Core substructure. If not specified - guessed using MCS', default='')
args = parser.parse_args()
# Get the reference molecule
ref_mol_input, ref_mol_suppl = rdkit_utils.default_open_input(args.refmol, args.refmol)
counter = 0
# Get the specified reference molecule. Default is the first
for mol in ref_mol_suppl:
counter+=1
if counter == args.refmolidx:
ref_mol = mol
break
ref_mol_input.close()
if counter < args.refmolidx:
raise ValueError("Invalid refmolidx. " + str(args.refmolidx) + " was specified but only " + str(counter) + " molecules were present in refmol.")
# handle metadata
source = "constrained_conf_gen.py"
datasetMetaProps = {"source":source, "description": "Constrained conformer generation using RDKit " + rdBase.rdkitVersion}
clsMappings = {"EmbedRMS": "java.lang.Float"}
fieldMetaProps = [{"fieldName":"EmbedRMS", "values": {"source":source, "description":"Embedding RMS value"}}]
# Get the molecules
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
output, WRITER, output_base = rdkit_utils.\
default_open_output(args.output, "const_conf_gen", args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps)
inputs = 0
totalCount = 0
totalErrors = 0
for mol in suppl:
inputs += 1
if mol:
count, errors = generate_conformers(inputs, mol, args.num, ref_mol, WRITER, args.core_smi)
totalCount += count
totalErrors += errors
input.close()
WRITER.close()
if totalErrors > 0:
utils.log("WARNING:", totalErrors, "conformers failed to generate")
# write metrics
if args.meta:
metrics = {'__InputCount__':inputs, '__OutputCount__':totalCount, 'RDKitConstrainedConformer':totalCount}
if totalErrors > 0:
metrics['__ErrorCount__'] = totalErrors
utils.write_metrics(output_base, metrics)
def main():
# Example usage:
# python -m pipelines.xchem.xcos -f ../../data/mpro/hits-17.sdf.gz -i ../../data/mpro/poses.sdf.gz -o xcos
parser = argparse.ArgumentParser(description='XCos scoring with RDKit')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-f', '--fragments', required=True, help='Fragments to compare')
parser.add_argument('-ff', '--fragments-format', help='Fragments format')
parser.add_argument('-t', '--score-threshold', type=float, default=0.4,
help='Minimum shape overlay and feature map score required for scoring a bit to a fragment')
parser.add_argument('--no-gzip', action='store_true', help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--metrics', action='store_true', help='Write metrics')
args = parser.parse_args()
utils.log("XCos Args: ", args)
source = "xcos.py"
datasetMetaProps = {"source":source, "description": "XCos scoring using RDKit " + rdBase.rdkitVersion}
clsMappings = {}
fieldMetaProps = []
clsMappings[field_XCosRefMols] = "java.lang.String"
clsMappings[field_XCosNumHits] = "java.lang.Integer"
clsMappings[field_XCosScore1] = "java.lang.Float"
fieldMetaProps.append({"fieldName":field_XCosRefMols, "values": {"source":source, "description":"XCos reference fragments"}})
fieldMetaProps.append({"fieldName":field_XCosNumHits, "values": {"source":source, "description":"XCos number of hits"}})
fieldMetaProps.append({"fieldName":field_XCosScore1, "values": {"source":source, "description":"XCos score 1"}})
frags_input,frags_suppl = rdkit_utils.default_open_input(args.fragments, args.fragments_format)
inputs_file, inputs_supplr = rdkit_utils.default_open_input(args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(args.output,
'xcos', args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
# this does the processing
process(inputs_supplr, frags_suppl, writer, threshold=args.score_threshold)
writer.close()
def split(input, informat, fieldName, outputBase, writeMetrics):
"""Splits the input into separate files. The name of each file and the file the each record is written to
is determined by the fieldName parameter
"""
input, suppl = rdkit_utils.default_open_input(input, informat)
i = 0
written = 0
writers = {}
outputs = []
filenames = []
for mol in suppl:
i += 1
if mol is None: continue
if not mol.HasProp(fieldName):
utils.log("Skipping molecule", i, "- did not contain field",
fieldName)
continue
value = mol.GetProp(fieldName)
if value:
s = str(value)
if writers.has_key(s):
writer = writers[s]
else:
name = outputBase + s
output, writer = rdkit_utils.default_open_output_sdf(
name, outputBase, False, False)
filenames.append(name + '.sdf')
outputs.append(output)
writers[s] = writer
writer.write(mol)
written += 1
utils.log("Generated", len(writers), "outputs from", i, "records")
input.close()
for k in writers:
writers[k].close()
for o in outputs:
o.close()
if writeMetrics:
utils.write_metrics(outputBase, {
'__InputCount__': i,
'__OutputCount__': written,
'Splitter': i
})
return filenames
def main():
# Example usage
# python -m pipelines.xchem.featurestein_score -i ../../data/mpro/poses.sdf.gz -f mpro-fstein.p -o fstein
global fmaps
parser = argparse.ArgumentParser(description='FeatureStein scoring with RDKit')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-f', '--feat-map', help='Feature Map pickle to score with')
parser.add_argument('--no-gzip', action='store_true', help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--metrics', action='store_true', help='Write metrics')
args = parser.parse_args()
utils.log("FeatureStein Args: ", args)
source = "featurestein_score.py"
datasetMetaProps = {"source":source, "description": "FeatureStein scoring using RDKit " + rdBase.rdkitVersion}
clsMappings = {}
fieldMetaProps = []
clsMappings[field_FeatureSteinQualityScore] = "java.lang.Float"
clsMappings[field_FeatureSteinQuantityScore] = "java.lang.Float"
fieldMetaProps.append({"fieldName":field_FeatureSteinQualityScore, "values": {"source":source, "description":"FeatureStein quality score"},
"fieldName":field_FeatureSteinQuantityScore, "values": {"source":source, "description":"FeatureStein quantity score"}})
pkl_file = open(args.feat_map, 'rb')
fmaps = pickle.load(pkl_file)
utils.log('FeatureMap has', fmaps.GetNumFeatures(), "features")
inputs_file, inputs_supplr = rdkit_utils.default_open_input(args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(args.output,
'featurestein', args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
# this does the processing
total, success, errors = process(inputs_supplr, writer)
inputs_file.close()
writer.flush()
writer.close()
output.close()
if args.metrics:
utils.write_metrics(output_base, {'__InputCount__':total, '__OutputCount__':success, '__ErrorCount__':errors, 'RDKitFeatureMap':success})
def main():
### command line args definitions #########################################
parser = argparse.ArgumentParser(description='Filter interactions')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-f',
'--group-by-field',
required=True,
help='Field to group records by (must be sequential)')
parser.add_argument('-s',
'--score-field',
required=True,
help='Field to use to rank records within a group')
parser.add_argument('-d',
'--score-descending',
action='store_true',
help='Sort records in descending order')
parser.add_argument('-x',
'--stats-fields',
nargs='*',
help='Field to use to for summary statistics')
parser.add_argument('-q',
'--quiet',
action='store_true',
help='Quiet mode')
parser.add_argument('--thin', action='store_true', help='Thin output mode')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
args = parser.parse_args()
utils.log("filter_interactions: ", args)
# handle metadata
source = "filter_interactions.py"
datasetMetaProps = {
"source": source,
"description": "Filter by interactions"
}
clsMappings = {
# "EnumChargesSrcMolUUID": "java.lang.String",
# "EnumChargesSrcMolIdx": "java.lang.Integer"
}
fieldMetaProps = [
# {"fieldName": "EnumChargesSrcMolUUID", "values": {"source": source, "description": "UUID of source molecule"}},
# {"fieldName": "EnumChargesSrcMolIdx", "values": {"source": source, "description": "Index of source molecule"}}
]
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'filter_interactions',
args.outformat,
thinOutput=False,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
report_file = open(output_base + '.report', 'wt')
count, total, errors = execute(suppl, writer, report_file,
args.group_by_field, args.score_field,
args.score_descending, args.stats_fields)
utils.log(count, total, errors)
if input:
input.close()
writer.flush()
writer.close()
output.close()
report_file.close()
# re-write the metadata as we now know the size
if args.outformat == 'json':
utils.write_squonk_datasetmetadata(output_base,
False,
clsMappings,
datasetMetaProps,
fieldMetaProps,
size=total)
if args.meta:
utils.write_metrics(
output_base, {
'__InputCount__': count,
'__OutputCount__': total,
'__ErrorCount__': errors,
'FilterInteractions': total
})
def main():
### command line args defintions #########################################
### Define the reactions available
poised_filter = True
if poised_filter == True:
from poised_filter import Filter
filter_to_use = Filter()
parser = argparse.ArgumentParser(description='RDKit rxn process')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-q',
'--quiet',
action='store_true',
help='Quiet mode')
parser.add_argument('-m',
'--multi',
action='store_true',
help='Output one file for each reaction')
parser.add_argument('-r',
'--reaction',
choices=filter_to_use.poised_reactions.keys(),
help='Name of reaction to be run')
parser.add_argument('-rl',
'--reagent_lib',
help="Input SD file, if not defined the STDIN is used")
parser.add_argument(
'-rlf',
'--reagent_lib_format',
choices=['sdf', 'json'],
help="Input format. When using STDIN this must be specified.")
args = parser.parse_args()
utils.log("Screen Args: ", args)
if not args.output and args.multi:
raise ValueError(
"Must specify output location when writing individual result files"
)
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
reagent_input, reagent_suppl = rdkit_utils.default_open_input(
args.reagent_lib, args.reagent_lib_format)
output, writer, output_base = rdkit_utils.default_open_output(
args.output, "rxn_maker", args.outformat)
i = 0
count = 0
if args.multi:
dir_base = os.path.dirname(args.output)
writer_dict = filter_to_use.get_writers(dir_base)
else:
writer_dict = None
dir_base = None
for mol in suppl:
i += 1
if mol is None: continue
# Return a dict/class here - indicating which filters passed
count = filter_to_use.perform_reaction(mol, args.reaction,
reagent_suppl, writer, count)
utils.log("Created", count, "molecules from a total of ", i,
"input molecules")
writer.flush()
writer.close()
if input:
input.close()
if output:
output.close()
# close the individual writers
if writer_dict:
for key in writer_dict:
writer_dict[key].close()
if args.meta:
utils.write_metrics(
output_base, {
'__InputCount__': i,
'__OutputCount__': count,
'RxnSmartsFilter': count
})
def main():
# Example usage:
# 1. Create keycloak token:
# export KEYCLOAK_TOKEN=$(curl -d "grant_type=password" -d "client_id=fragnet-search" -d "username=<username>" -d "password=<password>" \
# https://squonk.it/auth/realms/squonk/protocol/openid-connect/token 2> /dev/null | jq -r '.access_token')
#
# 2. Run the module:
# python -m pipelines.xchem.fragnet_expand -i ../../data/mpro/hits-17.sdf.gz --token $KEYCLOAK_TOKEN
parser = argparse.ArgumentParser(
description='Fragnet expand scoring with RDKit')
parameter_utils.add_default_input_args(parser)
parser.add_argument('--hac-min',
type=int,
default=3,
help='The min change in heavy atom count')
parser.add_argument('--hac-max',
type=int,
default=3,
help='The max change in heavy atom count')
parser.add_argument('--rac-min',
type=int,
default=1,
help='The min change in ring atom count')
parser.add_argument('--rac-max',
type=int,
default=1,
help='The max change in ring atom count')
parser.add_argument('--hops',
type=int,
default=1,
help='The number of graph traversals (hops)')
parser.add_argument(
'-s',
'--server',
default='https://fragnet-external.xchem-dev.diamond.ac.uk',
help='The fragnet search server')
parser.add_argument(
'--token',
help='Keycloak auth token (or specify as KEYCLOAK_TOKEN env variable')
parser.add_argument(
'--index-as-filename',
action='store_true',
help='Use the index as the file name instead of the molecule name')
args = parser.parse_args()
utils.log("FragnetExpand Args: ", args)
inputs_file, inputs_supplr = rdkit_utils.default_open_input(
args.input, args.informat)
if args.token:
auth_token = args.token
else:
auth_token = os.getenv('KEYCLOAK_TOKEN')
if not auth_token:
utils.log(
'WARNING: not authentication token found in environment variable KEYCLOAK_TOKEN'
)
# this does the processing
process(inputs_supplr,
hac_min=args.hac_min,
hac_max=args.hac_max,
rac_min=args.rac_min,
rac_max=args.rac_max,
hops=args.hops,
server=args.server,
token=auth_token,
index_as_filename=args.index_as_filename)
inputs_file.close()
def main():
global PDB_PATH, WRITER, THRESHOLD
parser = argparse.ArgumentParser(
description='PLI scoring - Docking calculation.')
parameter_utils.add_default_io_args(parser)
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('-pdb', '--pdb_file', help="PDB file for scoring")
parser.add_argument('-t',
'--threshold',
type=float,
help="The maximum score to allow",
default=None)
parser.add_argument(
'--threads',
type=int,
help="Number of threads to used. Default is the number of cores",
default=None)
parser.add_argument('--thin', action='store_true', help='Thin output mode')
args = parser.parse_args()
utils.log("PLI Args: ", args)
# Open up the input file
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
# Open the output file
s_now = datetime.datetime.utcnow().strftime("%d-%b-%Y %H:%M:%S UTC")
source = 'pipelines/docking/plip.py'
output, WRITER, output_base = \
rdkit_utils.default_open_output(args.output, "plip", args.outformat,
compress=not args.no_gzip,
thinOutput=args.thin,
valueClassMappings={'pliff_iscore': 'java.lang.Float',
'pliff_cscore': 'java.lang.Float',
'pliff_nb_score': 'java.lang.Float',
'pliff_gscore': 'java.lang.Float',
'pliff_score': 'java.lang.Float',
'pliff_tscore': 'java.lang.Float'},
datasetMetaProps={'created': s_now,
'source': source,
'description': 'PLI scoring of docked structures'}
)
PDB_PATH = args.pdb_file
if args.threshold:
THRESHOLD = args.threshold
# Iterate over the molecules
# WARNING - if using parallel processing the order of molecules is not preserved. Set args.threads to 1 to ensure this.
pool = ThreadPool(args.threads if args.
threads is not None else multiprocessing.cpu_count())
pool.map(run_dock, suppl)
pool.close()
pool.join()
# Close the file
WRITER.close()
if args.meta:
utils.write_metrics(output_base, {
'__InputCount__': COUNTER,
'__OutputCount__': SUCCESS,
'PLI': COUNTER
})
def main():
### command line args definitions #########################################
parser = argparse.ArgumentParser(description='Enumerate charges')
parser.add_argument('-i',
'--input',
help="Input file, if not defined the STDIN is used")
parser.add_argument(
'-if',
'--informat',
choices=['sdf', 'json', 'smi'],
help="Input format. When using STDIN this must be specified.")
parameter_utils.add_default_output_args(parser)
parser.add_argument(
'--fragment-method',
choices=['hac', 'mw'],
help=
'Approach to find biggest fragment if more than one (hac = biggest by heavy atom count, mw = biggest by mol weight)'
)
parser.add_argument(
"--minimize",
type=int,
default=0,
help="number of minimisation cycles when generating 3D molecules")
parser.add_argument('--enumerate-chirals',
help='Enumerate undefined chiral centers',
type=int,
default=0)
parser.add_argument('--smiles-field',
help='Add the SMILES as a field of this name')
parser.add_argument(
'--name-field',
help=
'Use this field in the input as the name field in the output. If not specified the SMILES is used'
)
parser.add_argument('--include-hydrogens',
action='store_true',
help='Include hydrogens in the output')
parser.add_argument("--min-charge",
type=int,
help="Minimum charge of molecule to process")
parser.add_argument("--max-charge",
type=int,
help="Maximum charge of molecule to process")
parser.add_argument("--num-charges",
type=int,
help="Maximum number of atoms with a charge")
parser.add_argument('-q',
'--quiet',
action='store_true',
help='Quiet mode')
parser.add_argument('--thin', action='store_true', help='Thin output mode')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
args = parser.parse_args()
utils.log("prepare_3d: ", args)
# handle metadata
source = "prepare_3d.py"
datasetMetaProps = {
"source": source,
"description": "Enumerate undefined stereo isomers and generate 3D"
}
clsMappings = {
"EnumChargesSrcMolUUID": "java.lang.String",
"EnumChargesSrcMolIdx": "java.lang.Integer"
}
fieldMetaProps = [{
"fieldName": "EnumChargesSrcMolUUID",
"values": {
"source": source,
"description": "UUID of source molecule"
}
}, {
"fieldName": "EnumChargesSrcMolIdx",
"values": {
"source": source,
"description": "Index of source molecule"
}
}]
oformat = utils.determine_output_format(args.outformat)
if args.informat == 'smi':
suppl = rdkit_utils.default_open_input_smiles(args.input,
delimiter='\t',
titleLine=False)
input = None
else:
input, suppl = rdkit_utils.default_open_input(args.input,
args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'enumerate_molecules',
args.outformat,
thinOutput=False,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
count, total, errors = execute(suppl,
writer,
minimize=args.minimize,
enumerate_chirals=args.enumerate_chirals,
smiles_field=args.smiles_field,
include_hs=args.include_hydrogens,
min_charge=args.min_charge,
max_charge=args.max_charge,
num_charges=args.num_charges)
utils.log(count, total, errors)
if input:
input.close()
writer.flush()
writer.close()
output.close()
# re-write the metadata as we now know the size
if oformat == 'json':
utils.write_squonk_datasetmetadata(output_base,
False,
clsMappings,
datasetMetaProps,
fieldMetaProps,
size=total)
if args.meta:
utils.write_metrics(
output_base, {
'__InputCount__': count,
'__OutputCount__': total,
'__ErrorCount__': errors,
'EnumerateChargesDimporphite': total
})
def main():
# Example usage
# python -m pipelines.xchem.calc_interactions -p ../../data/mpro/Mpro-x0387_0.pdb -i ../../data/mpro/hits-17.sdf.gz -o output
parser = argparse.ArgumentParser(description='Calculate interactions')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-p',
'--protein',
nargs='*',
help="File with protein (PDB or MOL2 format")
# NOTE reading mol2 format seems to be problematical.
parser.add_argument('-pf',
'--protein-format',
choices=['pdb', 'mol2'],
help="Protein file format")
parser.add_argument('--strict',
action='store_true',
help='Strict filtering')
parser.add_argument(
'--exact-protein',
action='store_true',
help='Exact matching of hydrogens and charges for protein')
parser.add_argument(
'--exact-ligand',
action='store_true',
help='Exact matching of hydrogens and charges for ligand')
parser.add_argument('--keep-hs-protein',
action='store_true',
help='Keep hydrogens on the protein')
parser.add_argument('--keep-hs-ligand',
action='store_true',
help='Keep hydrogens on the ligand')
parser.add_argument('--key-hbond',
nargs='*',
help='List of canonical H-bond interactions to count')
parser.add_argument(
'--key-hydrophobic',
nargs='*',
help='List of canonical hydrophobic interactions to count')
parser.add_argument(
'--key-salt-bridge',
nargs='*',
help='List of canonical salt bridge interactions to count')
parser.add_argument(
'--key-pi-stacking',
nargs='*',
help='List of canonical pi stacking interactions to count')
parser.add_argument(
'--key-pi-cation',
nargs='*',
help='List of canonical pi cation interactions to count')
parser.add_argument(
'--key-halogen',
nargs='*',
help='List of canonical halogen bond interactions to count')
parser.add_argument(
'--rfscores',
nargs='*',
help="Pickle(s) for RFScore model e.g. RFScore_v1_pdbbind2016.pickle")
parser.add_argument(
'--nnscores',
nargs='*',
help="Pickle(s) for NNScore model e.g. NNScore_pdbbind2016.pickle")
parser.add_argument(
'--plecscores',
nargs='*',
help=
"Pickle(s) for PLECScore model e.g. PLEClinear_p5_l1_pdbbind2016_s65536.pickle"
)
parser.add_argument('-r', '--report-file', help="File for the report")
parser.add_argument('-c',
'--compare',
help="Compare interactions with this report")
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--metrics', action='store_true', help='Write metrics')
args = parser.parse_args()
utils.log("Calculate interactions Args: ", args)
key_inters = {}
if args.key_hbond:
key_inters[interactions.I_TYPE_HBOND] = args.key_hbond
if args.key_hydrophobic:
key_inters[interactions.I_TYPE_HYDROPHOBIC] = args.key_hydrophobic
if args.key_salt_bridge:
key_inters[interactions.I_TYPE_SALT_BRIDGE] = args.key_salt_bridge
if args.key_pi_stacking:
key_inters[interactions.I_TYPE_PI_STACKING] = args.key_pi_stacking
if args.key_pi_cation:
key_inters[interactions.I_TYPE_PI_CATION] = args.key_pi_cation
if args.key_halogen:
key_inters[interactions.I_TYPE_HALOGEN] = args.key_halogen
source = "calc_interactions.py using ODDT"
datasetMetaProps = {
"source": source,
"description": "Calculate interactions using ODDT"
}
clsMappings = {}
fieldMetaProps = []
clsMappings[interactions.I_NAME_HBOND] = "java.lang.String"
clsMappings[interactions.I_NAME_HALOGEN] = "java.lang.String"
clsMappings[interactions.I_NAME_HYDROPHOBIC] = "java.lang.String"
clsMappings[interactions.I_NAME_SALT_BRIDGE] = "java.lang.String"
clsMappings[interactions.I_NAME_PI_STACKING] = "java.lang.String"
clsMappings[interactions.I_NAME_PI_CATION] = "java.lang.String"
clsMappings['NumTotalInteractions'] = "java.lang.Integer"
clsMappings['NumKeyInteractions'] = "java.lang.Integer"
clsMappings['KeyInteractions'] = "java.lang.String"
fieldMetaProps.append({
"fieldName": interactions.I_NAME_HBOND,
"values": {
"source": source,
"description": "Hydrogen bond interactions"
},
"fieldName": interactions.I_NAME_HALOGEN,
"values": {
"source": source,
"description": "Halogen bond interactions"
},
"fieldName": interactions.I_NAME_HYDROPHOBIC,
"values": {
"source": source,
"description": "Hydrophobic interactions"
},
"fieldName": interactions.I_NAME_SALT_BRIDGE,
"values": {
"source": source,
"description": "Salt bridge interactions"
},
"fieldName": interactions.I_NAME_PI_STACKING,
"values": {
"source": source,
"description": "Pi stacking interactions"
},
"fieldName": interactions.I_NAME_PI_CATION,
"values": {
"source": source,
"description": "Pi cation interactions"
}
})
inputs_file, inputs_supplr = rdkit_utils.default_open_input(
args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'calc_interactions',
args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
# this does the processing
count, errors = process(args.protein,
args.input,
writer,
key_inters,
protein_format=args.protein_format,
filter_strict=args.strict,
exact_protein=args.exact_protein,
exact_ligand=args.exact_ligand,
keep_hs_protein=args.keep_hs_protein,
keep_hs_ligand=args.keep_hs_ligand,
report_file=args.report_file,
compare_file=args.compare,
rfscores=args.rfscores,
nnscores=args.nnscores,
plecscores=args.plecscores)
utils.log('Processing complete.', count, 'records processed.', errors,
'errors')
inputs_file.close()
writer.flush()
writer.close()
output.close()
#
if args.metrics:
utils.write_metrics(
output_base, {
'__InputCount__': total,
'__OutputCount__': count,
'__ErrorCount__': errors,
'ODDTInteraction': count
})
def main():
parser = argparse.ArgumentParser(description='Max SuCOS scores with RDKit')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-tm',
'--target-molecules',
help='Target molecules to compare against')
parser.add_argument('-tf',
'--targets-format',
help='Target molecules format')
parser.add_argument('-n',
'--name-field',
help='Name of field with molecule name')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--filter-value',
type=float,
help='Filter out values with scores less than this.')
parser.add_argument('--filter-field',
help='Field to use to filter values.')
args = parser.parse_args()
utils.log("Max SuCOSMax Args: ", args)
source = "sucos_max.py"
datasetMetaProps = {
"source": source,
"description": "SuCOSMax using RDKit " + rdBase.rdkitVersion
}
clsMappings = {}
fieldMetaProps = []
clsMappings[field_SuCOSMax_Score] = "java.lang.Float"
clsMappings[field_SuCOSMax_FMScore] = "java.lang.Float"
clsMappings[field_SuCOSMax_ProtrudeScore] = "java.lang.Float"
clsMappings[field_SuCOSMax_Index] = "java.lang.Integer"
clsMappings[field_SuCOSCum_Score] = "java.lang.Float"
clsMappings[field_SuCOSCum_FMScore] = "java.lang.Float"
clsMappings[field_SuCOSCum_ProtrudeScore] = "java.lang.Float"
fieldMetaProps.append({
"fieldName": field_SuCOSMax_Score,
"values": {
"source": source,
"description": "SuCOS Max score"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSMax_FMScore,
"values": {
"source": source,
"description": "SuCOS Max Feature Map score"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSMax_ProtrudeScore,
"values": {
"source": source,
"description": "SuCOS Max Protrude score"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSMax_Index,
"values": {
"source": source,
"description": "SuCOS Max target index"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSCum_Score,
"values": {
"source": source,
"description": "SuCOS Cumulative score"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSCum_FMScore,
"values": {
"source": source,
"description": "SuCOS Cumulative Feature Map score"
}
})
fieldMetaProps.append({
"fieldName": field_SuCOSCum_ProtrudeScore,
"values": {
"source": source,
"description": "SuCOS Cumulative Protrude score"
}
})
if args.name_field:
clsMappings[field_SuCOSMax_Target] = "java.lang.String"
fieldMetaProps.append({
"fieldName": field_SuCOSMax_Target,
"values": {
"source": source,
"description": "SuCOS Max target name"
}
})
inputs_file, inputs_supplr = rdkit_utils.default_open_input(
args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'sucos-max',
args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
targets_file, targets_supplr = rdkit_utils.default_open_input(
args.target_molecules, args.targets_format)
count, total, errors = process(inputs_supplr, targets_supplr, writer,
args.name_field, args.filter_value,
args.filter_field)
inputs_file.close()
targets_file.close()
writer.flush()
writer.close()
output.close()
if args.meta:
utils.write_metrics(
output_base, {
'__InputCount__': count,
'__OutputCount__': total,
'__ErrorCount__': errors,
'RDKitSuCOS': total
})
def main():
### command line args defintions #########################################
parser = argparse.ArgumentParser(description='RDKit Butina Cluster')
parser.add_argument('-t', '--threshold', type=float, default=0.0, help='similarity threshold (1.0 means identical)')
parser.add_argument('-d', '--descriptor', type=str.lower, choices=list(descriptors.keys()), default='morgan2', help='descriptor or fingerprint type (default rdkit)')
parser.add_argument('-q', '--quiet', action='store_true', help='Quiet mode')
parser.add_argument('-n', '--num', type=int, help='maximum number to pick for diverse subset selection')
parser.add_argument('-s', '--seed-molecules', help='optional file containing any seed molecules that have already been picked')
parser.add_argument('--fragment-method', choices=['hac', 'mw'], default='hac', help='Approach to find biggest fragment if more than one (hac = biggest by heavy atom count, mw = biggest by mol weight)')
parser.add_argument('--output-fragment', action='store_true', help='Output the biggest fragment rather than the original molecule')
parameter_utils.add_default_io_args(parser)
args = parser.parse_args()
utils.log("MaxMinPicker Args: ", args)
descriptor = descriptors[args.descriptor]
if descriptor is None:
raise ValueError('No descriptor specified')
if not args.num and not args.threshold:
raise ValueError('--num or --threshold arguments must be specified, or both')
# handle metadata
source = "max_min_picker.py"
datasetMetaProps = {"source":source, "description": "MaxMinPicker using RDKit " + rdBase.rdkitVersion}
### generate fingerprints
fps = []
mols = []
errors = 0
# first the initial seeds, if specified
firstPicks = []
num_seeds = 0
if args.seed_molecules:
seedsInput,seedsSuppl = rdkit_utils.default_open_input(args.seed_molecules, None)
start = time.time()
errors += mol_utils.fragmentAndFingerprint(seedsSuppl, mols, fps, descriptor, fragmentMethod=args.fragment_method, outputFragment=args.output_fragment, quiet=args.quiet)
end = time.time()
seedsInput.close()
num_seeds = len(fps)
utils.log("Read", len(fps), "fingerprints for seeds in", end-start, "secs,", errors, "errors")
firstPicks = list(range(num_seeds))
# now the molecules to pick from
input,output,suppl,writer,output_base = rdkit_utils.default_open_input_output(args.input, args.informat, args.output, 'max_min_picker',
args.outformat, datasetMetaProps=datasetMetaProps)
# reset the mols list as we don't need the seeds, only the candidates
mols = []
start = time.time()
errs = mol_utils.fragmentAndFingerprint(suppl, mols, fps, descriptor, fragmentMethod=args.fragment_method, outputFragment=args.output_fragment, quiet=args.quiet)
end = time.time()
errors += errs
input.close()
num_fps = len(fps)
num_candidates = num_fps - num_seeds
utils.log("Read", num_candidates, "fingerprints for candidates in", end-start, "secs,", errs, "errors")
if not args.num:
num_to_pick = num_candidates
elif args.num > num_candidates:
num_to_pick = num_candidates
utils.log("WARNING: --num argument (", args.num, ") is larger than the total number of candidates (", num_candidates, ") - resetting to", num_candidates)
else:
num_to_pick = args.num
### do picking
utils.log("MaxMinPicking with descriptor", args.descriptor, "and threshold", args.threshold, ",", num_seeds, "seeds,", num_candidates, "candidates", num_fps, "total")
start = time.time()
picks, thresh = performPick(fps, num_to_pick + num_seeds, args.threshold, firstPicks)
end = time.time()
num_picks = len(picks)
utils.log("Found", num_picks, "molecules in", end-start, "secs, final threshold", thresh)
utils.log("Picks:", list(picks[num_seeds:]))
del fps
# we want to return the results in the order they were in the input so first we record the order in the pick list
indices = {}
i = 0
for idx in picks[num_seeds:]:
indices[idx] = i
i += 1
# now do the sort
sorted_picks = sorted(picks[num_seeds:])
# now write out the mols in the correct order recording the value in the pick list as the PickIndex property
i = 0
for idx in sorted_picks:
mol = mols[idx - num_seeds] # mols array only contains the candidates
mol.SetIntProp("PickIndex", indices[idx] + 1)
writer.write(mol)
i += 1
utils.log("Output", i, "molecules")
writer.flush()
writer.close()
output.close()
if args.meta:
metrics = {}
status_str = "{} compounds picked. Final threshold was {}.".format(i, thresh)
if errors > 0:
metrics['__ErrorCount__'] = errors
status_str = status_str + " {} errors.".format(errors)
metrics['__StatusMessage__'] = status_str
metrics['__InputCount__'] = num_fps
metrics['__OutputCount__'] = i
metrics['RDKitMaxMinPicker'] = num_picks
utils.write_metrics(output_base, metrics)
default=10,
help='number of conformers to generate')
parser.add_argument('-r', '--refmol', help="Reference molecule file")
parser.add_argument('--refmolidx',
help="Reference molecule index in file",
type=int,
default=1)
parser.add_argument(
'-c',
'--core_smi',
help='Core substructure. If not specified - guessed using MCS',
default='')
args = parser.parse_args()
# Get the reference molecule
ref_mol_input, ref_mol_suppl = rdkit_utils.default_open_input(
args.refmol, args.refmol)
counter = 0
# Get the specified reference molecule. Default is the first
for mol in ref_mol_suppl:
counter += 1
if counter == args.refmolidx:
ref_mol = mol
break
ref_mol_input.close()
if counter < args.refmolidx:
raise ValueError("Invalid refmolidx. " + str(args.refmolidx) +
" was specified but only " + str(counter) +
" molecules were present in refmol.")
# handle metadata
def main():
### command line args defintions #########################################
parser = argparse.ArgumentParser(description='RDKit filter')
parser.add_argument(
'-f',
'--fragment',
choices=['hac', 'mw'],
help=
'Find single fragment if more than one (hac = biggest by heavy atom count, mw = biggest by mol weight)'
)
parser.add_argument('--hacmin', type=int, help='Min heavy atom count')
parser.add_argument('--hacmax', type=int, help='Max heavy atom count')
parser.add_argument('--mwmin', type=float, help='Min mol weight')
parser.add_argument('--mwmax', type=float, help='Max mol weight')
parser.add_argument('--rotbmin',
type=float,
help='Min rotatable bond count')
parser.add_argument('--rotbmax',
type=float,
help='Max rotatable bond count')
parser.add_argument('--logpmin', type=float, help='Min logP')
parser.add_argument('--logpmax', type=float, help='Max logP')
parser.add_argument('-l',
'--limit',
type=int,
help='Limit output to this many records')
parser.add_argument(
'-c',
'--chunksize',
type=int,
help=
'Split output into chunks of size c. Output will always be files. Names like filter1.sdf.gz, filter2.sdf.gz ...'
)
parser.add_argument(
'-d',
'--digits',
type=int,
default=0,
help=
'When splitting zero pad the file name to this many digits so that they are in sorted order. Names like filter001.sdf.gz, filter002.sdf.gz ...'
)
parser.add_argument('-r',
'--rename',
action='append',
help='Rename field (fromname:toname)')
parser.add_argument(
'-t',
'--transform',
action='append',
help='Transform field value(fieldname:regex:type). ' +
'Regex is in the form of /regex/substitution/ (the 3 slashes are required). '
+
'Type is of int, float, boolean or string. The type is optional and if not specified then string is assumed. '
+
'Transformation occurs after field renaming so specify the new name.')
parser.add_argument('--delete', action='append', help='Delete field')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
# WARNING: thin output is not appropriate when using --fragment
parser.add_argument('--thin', action='store_true', help='Thin output mode')
parser.add_argument(
'-q',
'--quiet',
action='store_true',
help='Quiet mode - suppress reporting reason for filtering')
parameter_utils.add_default_io_args(parser)
args = parser.parse_args()
utils.log("Filter Args: ", args)
field_renames = {}
if args.rename:
for t in args.rename:
parts = t.split(':')
if len(parts) != 2:
raise ValueError('Invalid field rename argument:', t)
field_renames[parts[0]] = parts[1]
if args.delete:
for f in args.delete:
field_renames[f] = None
field_regexes = {}
field_replacements = {}
field_types = {}
if args.transform:
for t in args.transform:
parts = t.split(':')
if len(parts) < 2 or len(parts) > 3:
raise ValueError('Invalid field transform argument:', t)
terms = parts[1].split('/')
utils.log("|".join(terms) + str(len(terms)))
field_regexes[parts[0]] = re.compile(terms[1])
field_replacements[parts[0]] = terms[2]
if len(parts) == 3:
t = parts[2]
else:
t = 'string'
field_types[parts[0]] = t
utils.log("Created transform of " + terms[1] + " to " + terms[2] +
" using type of " + t)
if args.delete:
for f in args.delete:
field_renames[f] = None
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
if args.chunksize:
chunkNum = 1
if args.output:
output_base = args.output
else:
output_base = 'filter'
output_base_chunk = output_base + str(chunkNum).zfill(args.digits)
output, writer, output_base_chunk = rdkit_utils.default_open_output(
output_base_chunk,
output_base_chunk,
args.outformat,
thinOutput=args.thin,
compress=not args.no_gzip)
else:
output, writer, output_base_chunk = rdkit_utils.default_open_output(
args.output,
"filter",
args.outformat,
thinOutput=args.thin,
compress=not args.no_gzip)
output_base = output_base_chunk
utils.log("Writing to " + output_base_chunk)
i = 0
count = 0
chunkNum = 1
for mol in suppl:
if args.limit and count >= args.limit:
break
i += 1
if mol is None: continue
if args.fragment:
mol = mol_utils.fragment(mol, args.fragment, quiet=args.quiet)
if not filter(mol,
minHac=args.hacmin,
maxHac=args.hacmax,
minMw=args.mwmin,
maxMw=args.mwmax,
minRotb=args.rotbmin,
maxRotb=args.rotbmax,
minLogp=args.logpmin,
maxLogp=args.logpmax,
quiet=args.quiet):
continue
if args.chunksize:
if count > 0 and count % args.chunksize == 0:
# new chunk, so create new writer
writer.close()
output.close()
chunkNum += 1
output_chunk_base = output_base + str(chunkNum).zfill(
args.digits)
utils.log("Writing to " + output_chunk_base)
output, writer, output_chunk_base = rdkit_utils.default_open_output(
output_chunk_base,
output_chunk_base,
args.outformat,
thinOutput=args.thin,
compress=not args.no_gzip)
for from_name in field_renames:
to_name = field_renames[from_name]
if mol.HasProp(from_name):
val = mol.GetProp(from_name)
mol.ClearProp(from_name)
if to_name:
mol.SetProp(to_name, val)
for fieldname in field_regexes:
p = mol.GetProp(fieldname)
if p is not None:
regex = field_regexes[fieldname]
q = regex.sub(field_replacements[fieldname], p)
t = field_types[fieldname]
if t == 'int':
mol.SetIntProp(fieldname, int(q))
elif t == 'float':
mol.SetDoubleProp(fieldname, float(q))
elif t == 'boolean':
mol.SetBoolProp(fieldname, bool(q))
else:
mol.SetProp(fieldname, q)
count += 1
writer.write(mol)
utils.log("Filtered", i, "down to", count, "molecules")
if args.chunksize:
utils.log("Wrote", chunkNum, "chunks")
if (args.digits > 0 and len(str(chunkNum)) > args.digits):
utils.log(
"WARNING: not enough digits specified for the number of chunks"
)
writer.flush()
writer.close()
input.close()
output.close()
if args.meta:
utils.write_metrics(output_base, {
'__InputCount__': i,
'__OutputCount__': count,
'RDKitFilter': i
})
def main():
# Example usage
# python -m pipelines.xchem.rmsd_filter -i ../../data/mpro/poses.sdf.gz -o output -c 0.5
parser = argparse.ArgumentParser(description='RSMD filter')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-c', '--cutoff-rmsd', type=float, help='RMSD cutoff')
parser.add_argument('-f',
'--field',
default='_Name',
help='Field to group records')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--metrics', action='store_true', help='Write metrics')
args = parser.parse_args()
utils.log("RSMD filter Args: ", args)
source = "rmsd_filter.py"
datasetMetaProps = {
"source": source,
"description": "RMSD filter " + rdBase.rdkitVersion
}
clsMappings = {}
fieldMetaProps = []
# clsMappings[field_FeatureSteinQualityScore] = "java.lang.Float"
# clsMappings[field_FeatureSteinQuantityScore] = "java.lang.Float"
# fieldMetaProps.append({"fieldName":field_FeatureSteinQualityScore, "values": {"source":source, "description":"FeatureStein quality score"},
# "fieldName":field_FeatureSteinQuantityScore, "values": {"source":source, "description":"FeatureStein quantity score"}})
inputs_file, inputs_supplr = rdkit_utils.default_open_input(
args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'rmsd_filter',
args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
# this does the processing
count, groups, kept, errors = process(inputs_supplr, writer, args.field,
args.cutoff_rmsd)
utils.log('Processing complete.', count, 'records processed with', groups,
'groups.', kept, 'records retained.', errors, 'errors')
inputs_file.close()
writer.flush()
writer.close()
output.close()
if args.metrics:
utils.write_metrics(
output_base, {
'__InputCount__': total,
'__OutputCount__': success,
'__ErrorCount__': errors,
'RDKitFeatureMap': success
})
def main():
### command line args defintions #########################################
parser = argparse.ArgumentParser(description='RDKit Butina Cluster Matrix')
parameter_utils.add_default_input_args(parser)
parser.add_argument(
'-o',
'--output',
help=
"Base name for output file (no extension). If not defined then SDTOUT is used for the structures and output is used as base name of the other files."
)
parser.add_argument('-of',
'--outformat',
choices=['tsv', 'json'],
default='tsv',
help="Output format. Defaults to 'tsv'.")
parser.add_argument('--meta',
action='store_true',
help='Write metadata and metrics files')
parser.add_argument(
'-t',
'--threshold',
type=float,
default=0.7,
help='Similarity clustering threshold (1.0 means identical)')
parser.add_argument(
'-mt',
'--matrixThreshold',
type=float,
default=0.5,
help='Threshold for outputting values (1.0 means identical)')
parser.add_argument('-d',
'--descriptor',
type=str.lower,
choices=list(cluster_butina.descriptors.keys()),
default='rdkit',
help='descriptor or fingerprint type (default rdkit)')
parser.add_argument('-m',
'--metric',
type=str.lower,
choices=list(cluster_butina.metrics.keys()),
default='tanimoto',
help='similarity metric (default tanimoto)')
parser.add_argument('-q',
'--quiet',
action='store_true',
help='Quiet mode')
args = parser.parse_args()
utils.log("Cluster Matrix Args: ", args)
descriptor = cluster_butina.descriptors[args.descriptor]
if descriptor is None:
raise ValueError('Invalid descriptor name ' + args.descriptor)
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
# handle metadata
source = "cluster_butina_matrix.py"
datasetMetaProps = {
"source": source,
"description": "Butina clustering using RDKit " + rdBase.rdkitVersion
}
clsMappings = {
"Cluster1": "java.lang.Integer",
"Cluster2": "java.lang.Integer",
"ID1": "java.lang.String",
"ID2": "java.lang.String",
"M1": "java.lang.String",
"M2": "java.lang.String",
"Similarity": "java.lang.Float"
}
fieldMetaProps = [{
"fieldName": "Cluster",
"values": {
"source": source,
"description": "Cluster number"
}
}]
fieldNames = collections.OrderedDict()
fieldNames['ID1'] = 'ID1'
fieldNames['ID2'] = 'ID2'
fieldNames['Cluster1'] = 'Cluster1'
fieldNames['Cluster2'] = 'Cluster2'
fieldNames['Similarity'] = 'Similarity'
fieldNames['M1'] = 'M1'
fieldNames['M2'] = 'M2'
writer,output_base = utils.\
create_simple_writer(args.output, 'cluster_butina_matrix',
args.outformat, fieldNames,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps)
### generate fingerprints
mols = [x for x in suppl if x is not None]
fps = [descriptor(x) for x in mols]
input.close()
### do clustering
utils.log("Clustering with descriptor", args.descriptor, "metric",
args.metric, "and threshold", args.threshold)
clusters, dists, matrix, = cluster_butina.ClusterFps(
fps, args.metric, 1.0 - args.threshold)
utils.log("Found", len(clusters), "clusters")
MapClusterToMols(clusters, mols)
if not args.quiet:
utils.log("Clusters:", clusters)
writer.writeHeader()
size = len(matrix)
#utils.log("len(matrix):", size)
count = 0
for i in range(size):
#utils.log("element",i, "has length", len(matrix[i]))
writer.write(create_values(mols, i, i, 1.0))
count += 1
for j in range(len(matrix[i])):
#utils.log("writing",i,j)
dist = matrix[i][j]
if dist > args.matrixThreshold:
# the matrix is the lower left segment without the diagonal
x = j
y = i + 1
writer.write(create_values(mols, x, y, dist))
writer.write(create_values(mols, y, x, dist))
count += 2
writer.write(create_values(mols, size, size, 1.0))
writer.writeFooter()
writer.close()
if args.meta:
utils.write_metrics(output_base, {
'__InputCount__': i,
'__OutputCount__': count,
'RDKitCluster': i
})
def main():
global WRITER, THRESHOLD
global PDB_PATH
parser = argparse.ArgumentParser(
description='SMoG2016 - Docking calculation.')
parameter_utils.add_default_io_args(parser)
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('-pdb', '--pdb_file', help="PDB file for scoring")
parser.add_argument('-t',
'--threshold',
help="The maximum score to allow",
default=None)
parser.add_argument(
'--threads',
type=int,
help="Number of threads to used. Default is the number of cores",
default=None)
parser.add_argument('--thin', action='store_true', help='Thin output mode')
args = parser.parse_args()
utils.log("SMoG2016 Args: ", args)
smog_path = "/usr/local/SMoG2016/"
if args.threshold:
THRESHOLD = float(args.threshold)
else:
THRESHOLD = None
PDB_PATH = "/tmp/pdb_file.pdb"
# Now copy it to prot_pdb.pdb -> silly SMOG bug requires underscore in the filename!
shutil.copy(args.pdb_file, PDB_PATH)
# Open up the input file
input, suppl = rdkit_utils.default_open_input(args.input, args.informat)
# Open the output file
output, WRITER, output_base = rdkit_utils.\
default_open_output(args.output, "SMoG2016",
args.outformat, compress=not args.no_gzip)
# Cd to the route of the action
# TODO - can this be done without changing dir? It gives problems in finding the input files and in writing the metrics
cwd = os.getcwd()
os.chdir(smog_path)
# Iterate over the molecules
# WARNING - if using parallel processing the order of molecules is not preserved. Set args.threads to 1 to ensure this.
if args.threads is None:
threads = multiprocessing.cpu_count()
else:
threads = args.threads
pool = ThreadPool(threads)
pool.map(run_dock, suppl)
# Close the file
WRITER.close()
os.chdir(cwd)
if args.meta:
utils.write_metrics(
output_base, {
'__InputCount__': COUNTER,
'__OutputCount__': SUCCESS,
'SMoG2016': COUNTER
})
utils.log("SMoG2016 complete")
def main():
# Example usage
# python -m pipelines.xchem.featurestein_generate_and_score -i ../../data/mpro/poses.sdf.gz -f ../../data/mpro/hits-17.sdf.gz -o output_fs
global fmaps
parser = argparse.ArgumentParser(
description='FeatureStein scoring with RDKit')
parameter_utils.add_default_io_args(parser)
parser.add_argument('-f',
'--fragments',
help='Fragments to use to generate the feature map')
parser.add_argument('-ff', '--fragments-format', help='Fragments format')
parser.add_argument(
'--no-gzip',
action='store_true',
help='Do not compress the output (STDOUT is never compressed')
parser.add_argument('--metrics', action='store_true', help='Write metrics')
args = parser.parse_args()
utils.log("FeatureStein Args: ", args)
source = "featurestein_generate_and_score.py"
datasetMetaProps = {
"source": source,
"description":
"FeatureStein scoring using RDKit " + rdBase.rdkitVersion
}
clsMappings = {}
fieldMetaProps = []
clsMappings[field_FeatureSteinQualityScore] = "java.lang.Float"
clsMappings[field_FeatureSteinQuantityScore] = "java.lang.Float"
fieldMetaProps.append({
"fieldName": field_FeatureSteinQualityScore,
"values": {
"source": source,
"description": "FeatureStein quality score"
},
"fieldName": field_FeatureSteinQuantityScore,
"values": {
"source": source,
"description": "FeatureStein quantity score"
}
})
# generate the feature maps
frags_input, frags_suppl = rdkit_utils.default_open_input(
args.fragments, args.fragments_format)
fmaps = create_featuremap(frags_suppl)
frags_input.close()
# read the ligands to be scored
inputs_file, inputs_supplr = rdkit_utils.default_open_input(
args.input, args.informat)
output, writer, output_base = rdkit_utils.default_open_output(
args.output,
'featurestein',
args.outformat,
valueClassMappings=clsMappings,
datasetMetaProps=datasetMetaProps,
fieldMetaProps=fieldMetaProps,
compress=not args.no_gzip)
# do the scoring
total, success, errors = score_molecules(inputs_supplr, writer)
utils.log('Scored', success, 'molecules.', errors, 'errors.')
inputs_file.close()
writer.flush()
writer.close()
output.close()
if args.metrics:
utils.write_metrics(
output_base, {
'__InputCount__': total,
'__OutputCount__': success,
'__ErrorCount__': errors,
'RDKitFeatureMap': success
})
