def get_mtz_info(datafile):
"""
Get unit cell and SG from input mtz
"""
sg = False
cell = False
vol = False
# Convert from unicode
datafile = convert_unicode(datafile)
# Read datafile
data = iotbx_mtz.object(datafile)
# Derive space group from datafile
sg = fix_R3_sg(data.space_group_name().replace(" ", ""))
# Wrangle the cell parameters
cell = [round(x,3) for x in data.crystals()[0].unit_cell_parameters() ]
# The volume
vol = data.crystals()[0].unit_cell().volume()
return (sg, cell, vol)
def get_mtz_info(data_file):
"""
Get unit cell and SG from input mtz
"""
sg = False
cell = False
vol = False
# Convert from unicode
data_file = convert_unicode(data_file)
# Read data_file
data = iotbx_mtz.object(data_file)
# Derive space group from data_file
sg = fix_R3_sg(data.space_group_name().replace(" ", ""))
# Wrangle the cell parameters
cell = [round(x, 3) for x in data.crystals()[0].unit_cell_parameters()]
# The volume
vol = data.crystals()[0].unit_cell().volume()
return (sg, cell, vol)
def get_res(data_file):
"""Return resolution limit of dataset"""
data_file = convert_unicode(data_file)
data = iotbx_mtz.object(data_file)
return float(data.max_min_resolution()[-1])
def get_res(datafile):
"""Return resolution limit of dataset"""
datafile = convert_unicode(datafile)
data = iotbx_mtz.object(datafile)
return float(data.max_min_resolution()[-1])
def get_spacegroup_info(cif_file):
"""Get info from PDB of mmCIF file"""
# print "get_spacegroup_info", cif_file, os.getcwd()
cif_file = convert_unicode(cif_file)
if cif_file[-3:].lower() == "cif":
fail = False
cif_spacegroup = False
try:
input_file = open(cif_file, "r").read(20480)
for line in input_file.split('\n'):
if "_symmetry.space_group_name_H-M" in line:
cif_spacegroup = line[32:].strip()[1:-1].upper().replace(" ", "")
if "_pdbx_database_status.pdb_format_compatible" in line:
if line.split()[1] == "N":
fail = True
except IOError:
return False
if fail:
return False
else:
return cif_spacegroup
else:
return str(iotbx_pdb.input(cif_file).crystal_symmetry().space_group_info()).upper().replace(" ", "")
def get_spacegroup_info(struct_file):
"""Get info from PDB of mmCIF file"""
# print "get_spacegroup_info", struct_file, os.getcwd()
struct_file = convert_unicode(struct_file)
if struct_file[-3:].lower() == "cif":
fail = False
cif_spacegroup = False
try:
input_file = open(struct_file, "r").read(20480)
for line in input_file.split('\n'):
if "_symmetry.space_group_name_H-M" in line:
cif_spacegroup = line[32:].strip()[1:-1].upper().replace(
" ", "")
# print cif_spacegroup
if "_pdbx_database_status.pdb_format_compatible" in line:
if line.split()[1] == "N":
fail = True
except IOError:
return False
if fail:
return False
else:
return cif_spacegroup
else:
return str(
iotbx_pdb.input(struct_file).crystal_symmetry().space_group_info()
).upper().replace(" ", "")
def wrapper(**kwargs):
os.chdir(kwargs.get('work_dir', os.getcwd()))
if not kwargs.get('script', False):
# Pop out the launcher
launcher = kwargs.pop('launcher', None)
# Pop out the batch_queue
batch_queue = kwargs.pop('batch_queue', None)
# Pop out the results_queue
result_queue = kwargs.pop('result_queue', None)
# Create a unique identifier for Phaser results
#kwargs['output_id'] = 'Phaser_%d' % random.randint(0, 10000)
# Grab the RAPD python path (if available)
rapd_python = kwargs.pop('rapd_python', 'rapd.python')
# Signal to launch run
kwargs['script'] = True
if kwargs.get('pool', False):
# If running on local machine. Launcher will be 'utils.processes.local_subprocess'
pool = kwargs.pop('pool')
f = write_script(kwargs)
proc = pool.apply_async(
launcher,
kwds={
"command":
"%s %s" % (rapd_python, f),
"logfile":
os.path.join(convert_unicode(kwargs.get('work_dir')),
'rapd_phaser.log'),
"result_queue":
result_queue,
})
#return (proc, 'junk', kwargs['output_id'])
return (proc, 'junk')
else:
# If running on computer cluster. Launcher will be sites.cluster.(site_name).process_cluster
f = write_script(kwargs)
pid_queue = Queue()
proc = Process(target=launcher,
kwargs={
"command":
"%s %s" % (rapd_python, f),
"logfile":
os.path.join(kwargs.get('work_dir'),
'rapd_phaser.log'),
"pid_queue":
pid_queue,
"batch_queue":
batch_queue,
})
proc.start()
#return (proc, pid_queue.get(), kwargs['output_id'])
return (proc, pid_queue.get())
else:
# Remove extra input params used to setup job
l = ['script', 'test']
for k in l:
# pop WON'T error out if key not found!
_ = kwargs.pop(k, None)
# Just launch job
return func(**kwargs)
def run_ellg():
res0 = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if f[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(f), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(f), 0.7)
# i.addSEAR_ENSE_NUM("junk",5)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
res0 = r1.get_target_resolution('model')
del (r1)
return res0
def run_ellg():
new_res = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if mmcif[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(mmcif), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(mmcif), 0.7)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
# If it worked use the recommended resolution
new_res = round(r1.get_target_resolution('model'), 1)
del(r1)
return new_res
def run_ellg():
res0 = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if f[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(f), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(f), 0.7)
# i.addSEAR_ENSE_NUM("junk",5)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
res0 = r1.get_target_resolution('model')
del(r1)
return res0
def run_ellg():
new_res = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
# Read in CIF file
if struct_file[-3:] in ('cif', ):
i0.addENSE_CIT_ID("model", convert_unicode(struct_file), 0.7)
# Read in PDB file
else:
i0.addENSE_PDB_ID("model", convert_unicode(struct_file), 0.7)
try:
r1 = phaser.runMR_ELLG(i0)
except RuntimeError as e:
# print "Hit error"
# Known CIF error - convert to pdb and retry
if struct_file[-3:] in ('cif', ):
# print "Convert to pdb"
pdb.cif_as_pdb((struct_file, ))
pdb_file = struct_file.replace(".cif", ".pdb")
i1 = phaser.InputMR_ELLG()
i1.setSPAC_HALL(r.getSpaceGroupHall())
i1.setCELL6(r.getUnitCell())
i1.setMUTE(True)
i1.setREFL_DATA(r.getDATA())
i1.addENSE_PDB_ID("model", convert_unicode(pdb_file), 0.7)
r1 = phaser.runMR_ELLG(i1)
else:
raise e
# print r1.logfile()
if r1.Success():
# If it worked use the recommended resolution
new_res = round(r1.get_target_resolution('model'), 1)
del (r1)
return new_res
def __init__(self, command, output=None, tprint=False, logger=None):
# If the logging instance is passed in...
if logger:
self.logger = logger
else:
# Otherwise get the logger Instance
self.logger = logging.getLogger("RAPDLogger")
self.logger.debug("__init__")
# Store tprint for use throughout
if tprint:
self.tprint = tprint
# Dead end if no tprint passed
else:
def func(arg=False, level=False, verbosity=False, color=False):
pass
self.tprint = func
# Stopwatch
self.start_time = time.time()
# Store inputs
self.input = command
self.output = output
self.logger = logger
# pprint(command)
# Params
self.working_dir = self.input["directories"].get("work", os.getcwd())
self.test = self.input["preferences"].get("test", False)
self.sample_type = self.input["preferences"].get("type", "protein")
self.solvent_content = self.input["preferences"].get(
"solvent_content", 0.55)
self.cluster_use = self.input["preferences"].get("cluster", False)
self.clean = self.input["preferences"].get("clean", True)
self.gui = self.input["preferences"].get("gui", True)
# self.controller_address = self.input[0].get("control", False)
self.verbose = self.input["preferences"].get("verbose", False)
self.datafile = xutils.convert_unicode(
self.input["input_data"].get("datafile"))
Process.__init__(self, name="PDBQuery")
self.start()
def __init__(self, command, output=None, tprint=False, logger=None):
# If the logging instance is passed in...
if logger:
self.logger = logger
else:
# Otherwise get the logger Instance
self.logger = logging.getLogger("RAPDLogger")
self.logger.debug("__init__")
# Store tprint for use throughout
if tprint:
self.tprint = tprint
# Dead end if no tprint passed
else:
def func(arg=False, level=False, verbosity=False, color=False):
pass
self.tprint = func
# Stopwatch
self.start_time = time.time()
# Store inputs
self.input = command
self.output = output
self.logger = logger
# pprint(command)
# Params
self.working_dir = self.input["directories"].get("work", os.getcwd())
self.test = self.input["preferences"].get("test", False)
self.sample_type = self.input["preferences"].get("type", "protein")
self.solvent_content = self.input["preferences"].get("solvent_content", 0.55)
self.cluster_use = self.input["preferences"].get("cluster", False)
self.clean = self.input["preferences"].get("clean", True)
self.gui = self.input["preferences"].get("gui", True)
# self.controller_address = self.input[0].get("control", False)
self.verbose = self.input["preferences"].get("verbose", False)
self.datafile = xutils.convert_unicode(self.input["input_data"].get("datafile"))
Process.__init__(self, name="PDBQuery")
self.start()
def wrapper(**kwargs):
os.chdir(kwargs.get('work_dir', os.getcwd()))
if not kwargs.get('script', False):
# Pop out the launcher
launcher = kwargs.pop('launcher', None)
# Pop out the batch_queue
batch_queue = kwargs.pop('batch_queue', None)
# Create a unique identifier for Phaser results
kwargs['output_id'] = 'Phaser_%d' % random.randint(0, 10000)
# Signal to launch run
kwargs['script'] = True
if kwargs.get('pool', False):
# If running on local machine
pool = kwargs.pop('pool')
f = write_script(kwargs)
new_kwargs = {"command": "rapd2.python %s" % f,
"logfile": os.path.join(convert_unicode(kwargs.get('work_dir')), 'rapd_phaser.log'),
}
proc = pool.apply_async(launcher, kwds=new_kwargs,)
return (proc, 'junk', kwargs['output_id'])
else:
# If running on computer cluster
f = write_script(kwargs)
pid_queue = Queue()
proc = Process(target=launcher,
kwargs={"command": "rapd2.python %s" % f,
"pid_queue": pid_queue,
"batch_queue": batch_queue,
"logfile": os.path.join(kwargs.get('work_dir'), 'rapd_phaser.log'),
})
proc.start()
return (proc, pid_queue.get(), kwargs['output_id'])
else:
# Remove extra input params used to setup job
l = ['script', 'test']
for k in l:
# pop WON'T error out if key not found!
_ = kwargs.pop(k, None)
# Just launch job
return func(**kwargs)
def run_phaser(
data_file,
struct_file,
spacegroup,
db_settings=False,
tag=False,
work_dir=False,
adf=False,
name=False,
ncopy=1,
cell_analysis=False,
resolution=False,
full=False,
):
"""
Run Phaser and passes results back to RAPD Redis DB
**Requires Phaser src code!**
data_file - input data as mtz (required)
struct_file - input search model path in mmCIF or PDB format (required)
spacegroup - The space group to run MR (required)
tag - a Redis key where the results are sent (cluster mode)
db_settings - Redis connection settings for sending results (cluster mode)
work_dir - working directory (defaults to current working dir)
name - root name for output files (defaults to spacegroup)
ncopy - number of molecules to search for
cell_analysis - internal RAPD signal so all possible SG's are searched
resolution - high res limit to run MR (float)
full - signal to run more comprehensive MR
"""
phaser_log = False
# Change to work_dir
if not work_dir:
work_dir = os.getcwd()
os.chdir(work_dir)
if not name:
name = spacegroup
# # Handle CIF file input -> PDB
# if struct_file[-3:] == "cif":
# pdb.cif_as_pdb(struct_file)
# struct_file = struct_file.replace(".cif", ".pdb")
# Read the dataset
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(data_file))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r1 = phaser.runMR_DAT(i)
# Need to determine Phaser version for keyword changes!
version = re.search(r'Version:\s*([\d.]+)', r1.logfile()).group(1)
if r1.Success():
i = phaser.InputMR_AUTO()
# i.setREFL_DATA(r1.getREFL_DATA())
# i.setREFL_DATA(r1.DATA_REFL())
i.setREFL_F_SIGF(r1.getMiller(), r1.getFobs(), r1.getSigFobs())
i.setCELL6(r1.getUnitCell())
if struct_file[-3:].lower() == "cif":
#i.addENSE_CIF_ID('model', cif, 0.7)
### Typo in PHASER CODE!!! <<<CIT>>> ###
i.addENSE_CIT_ID('model', convert_unicode(struct_file), 0.7)
else:
i.addENSE_PDB_ID('model', convert_unicode(struct_file), 0.7)
i.addSEAR_ENSE_NUM("model", ncopy)
i.setSPAC_NAME(spacegroup)
if cell_analysis:
i.setSGAL_SELE("ALL")
# Set it for worst case in orth
# number of processes to run in parallel where possible
i.setJOBS(1)
else:
i.setSGAL_SELE("NONE")
if full:
# Picks own resolution
# Round 2, pick best solution as long as less that 10% clashes
i.setPACK_SELE("PERCENT")
i.setPACK_CUTO(0.1)
#command += "PACK CUTOFF 10\n"
else:
# For first round and cell analysis
# Only set the resolution limit in the first round or cell analysis.
if resolution:
i.setRESO_HIGH(resolution)
else:
i.setRESO_HIGH(6.0)
# If Phaser version < 2.6.0
if int(version.split('.')[1]) <= 6:
i.setSEAR_DEEP(False)
else:
i.setSEAR_METH("FAST")
# Don"t seem to work since it picks the high res limit now.
# Get an error when it prunes all the solutions away and TF has no input.
# command += "PEAKS ROT SELECT SIGMA CUTOFF 4.0\n"
# command += "PEAKS TRA SELECT SIGMA CUTOFF 6.0\n"
# Turn off pruning in 2.6.0
i.setSEAR_PRUN(False)
# Choose more top peaks to help with getting it correct.
i.setPURG_ROTA_ENAB(True)
i.setPURG_ROTA_NUMB(3)
#command += "PURGE ROT ENABLE ON\nPURGE ROT NUMBER 3\n"
i.setPURG_TRAN_ENAB(True)
i.setPURG_TRAN_NUMB(1)
#command += "PURGE TRA ENABLE ON\nPURGE TRA NUMBER 1\n"
# Only keep the top after refinement.
i.setPURG_RNP_ENAB(True)
i.setPURG_RNP_NUMB(1)
#command += "PURGE RNP ENABLE ON\nPURGE RNP NUMBER 1\n"
i.setROOT(convert_unicode(name))
# i.setMUTE(False)
i.setMUTE(True)
# Delete the setup results
# del(r)
# launch the run
# r = phaser.runMR_AUTO(i)
try:
r = phaser.runMR_AUTO(i)
except RuntimeError as e:
# print "Hit error"
# Known CIF error - convert to pdb and retry
if struct_file[-3:] in ('cif', ):
# print "Convert to pdb"
pdb.cif_as_pdb((struct_file, ))
pdb_file = struct_file.replace(".cif", ".pdb")
i = phaser.InputMR_AUTO()
# i.setREFL_DATA(r1.getREFL_DATA())
# i.setREFL_DATA(r1.DATA_REFL())
i.setREFL_F_SIGF(r1.getMiller(), r1.getFobs(), r1.getSigFobs())
i.setCELL6(r1.getUnitCell())
i.addENSE_PDB_ID('model', convert_unicode(pdb_file), 0.7)
i.addSEAR_ENSE_NUM("model", ncopy)
i.setSPAC_NAME(spacegroup)
if cell_analysis:
i.setSGAL_SELE("ALL")
# Set it for worst case in orth
# number of processes to run in parallel where possible
i.setJOBS(1)
else:
i.setSGAL_SELE("NONE")
if full:
# Picks own resolution
# Round 2, pick best solution as long as less that 10% clashes
i.setPACK_SELE("PERCENT")
i.setPACK_CUTO(0.1)
#command += "PACK CUTOFF 10\n"
else:
# For first round and cell analysis
# Only set the resolution limit in the first round or cell analysis.
if resolution:
i.setRESO_HIGH(resolution)
else:
i.setRESO_HIGH(6.0)
# If Phaser version < 2.6.0
if int(version.split('.')[1]) <= 6:
i.setSEAR_DEEP(False)
else:
i.setSEAR_METH("FAST")
# Don"t seem to work since it picks the high res limit now.
# Get an error when it prunes all the solutions away and TF has no input.
# command += "PEAKS ROT SELECT SIGMA CUTOFF 4.0\n"
# command += "PEAKS TRA SELECT SIGMA CUTOFF 6.0\n"
# Turn off pruning in 2.6.0
i.setSEAR_PRUN(False)
# Choose more top peaks to help with getting it correct.
i.setPURG_ROTA_ENAB(True)
i.setPURG_ROTA_NUMB(3)
#command += "PURGE ROT ENABLE ON\nPURGE ROT NUMBER 3\n"
i.setPURG_TRAN_ENAB(True)
i.setPURG_TRAN_NUMB(1)
#command += "PURGE TRA ENABLE ON\nPURGE TRA NUMBER 1\n"
# Only keep the top after refinement.
i.setPURG_RNP_ENAB(True)
i.setPURG_RNP_NUMB(1)
#command += "PURGE RNP ENABLE ON\nPURGE RNP NUMBER 1\n"
i.setROOT(convert_unicode(name))
# i.setMUTE(False)
i.setMUTE(True)
# Delete the setup results
# del(r)
# launch the run
r = phaser.runMR_AUTO(i)
else:
raise e
if r.Success():
# print r
pass
#if r.foundSolutions():
#print "Phaser has found MR solutions"
#print "Top LLG = %f" % r.getTopLLG()
#print "Top PDB file = %s" % r.getTopPdbFile()
#else:
#print "Phaser has not found any MR solutions"
else:
print "Job exit status FAILURE"
print r.ErrorName(), "ERROR :", r.ErrorMessage()
# Save log files for debugging
phaser_log = r.logfile()
with open('phaser.log', 'w') as log:
log.write(r.logfile())
log.close()
if r.foundSolutions():
rfz = None
tfz = None
tncs = False
# Parse results
for p in r.getTopSet().ANNOTATION.split():
# print p
# For v 2.8.3
# RF*0\nTF*0\nLLG=30699\nTFZ==174.8\nPAK=0\nLLG=30699\nTFZ==174.8\n
if p.count('RFZ'):
if p.count('=') in [1]:
rfz = float(p.split('=')[-1])
if p.count('RF*0'):
rfz = "NC"
if p.count('TFZ'):
if p.count('=') in [1]:
tfz = p.split('=')[-1]
if tfz == '*':
tfz = 'arbitrary'
else:
tfz = float(tfz)
if p.count('TF*0'):
tfz = "NC"
tncs_test = [
1 for line in r.getTopSet().unparse().splitlines()
if line.count("+TNCS")
]
tncs = bool(len(tncs_test))
mtz_file = os.path.join(work_dir, r.getTopMtzFile())
phaser_result = {
"ID": name,
"solution": r.foundSolutions(),
"pdb_file": os.path.join(work_dir, r.getTopPdbFile()),
"mtz": mtz_file,
"gain": float(r.getTopLLG()),
"rfz": rfz,
# "tfz": r.getTopTFZ(),
"tfz": tfz,
"clash": r.getTopSet().PAK,
"dir": os.getcwd(),
"spacegroup":
r.getTopSet().getSpaceGroupName().replace(' ', ''),
"tNCS": tncs,
"nmol": r.getTopSet().NUM,
"adf": None,
"peak": None,
}
# Calculate 2Fo-Fc & Fo-Fc maps
# foo.mtz begets foo_2mFo-DFc.ccp4 & foo__mFo-DFc.ccp4
local_subprocess(command="phenix.mtz2map %s" % mtz_file,
logfile='map.log',
shell=True)
# Map files should now exist
map_2_1 = mtz_file.replace(".mtz", "_2mFo-DFc.ccp4")
map_1_1 = mtz_file.replace(".mtz", "_mFo-DFc.ccp4")
# Make sure the maps exist and then package them
if os.path.exists(map_2_1):
# Compress the map
arch_prod_file, arch_prod_hash = archive.compress_file(map_2_1)
# Remove the map that was compressed
os.unlink(map_2_1)
# Store information
map_for_display = {
"path": arch_prod_file,
"hash": arch_prod_hash,
"description": "map_2_1"
}
phaser_result["map_2_1"] = map_for_display
if os.path.exists(map_1_1):
# Compress the map
arch_prod_file, arch_prod_hash = archive.compress_file(map_1_1)
# Remove the map that was compressed
os.unlink(map_1_1)
# Store information
map_for_display = {
"path": arch_prod_file,
"hash": arch_prod_hash,
"description": "map_1_1"
}
phaser_result["map_1_1"] = map_for_display
# If PDB exists, package that too
if phaser_result.get("pdb_file", False):
if os.path.exists(phaser_result.get("pdb_file")):
# Compress the file
arch_prod_file, arch_prod_hash = archive.compress_file(
phaser_result.get("pdb_file"))
# Remove the map that was compressed
# os.unlink(phaser_result.get("pdb"))
# Store information
pdb_for_display = {
"path":
arch_prod_file,
"hash":
arch_prod_hash,
"description":
os.path.basename(phaser_result.get("pdb_file"))
}
phaser_result["pdb"] = pdb_for_display
# Calc ADF map
if adf:
if os.path.exists(phaser_result.get(
"pdb_file", False)) and os.path.exists(
phaser_result.get("mtz", False)):
adf_results = calc_ADF_map(data_file=data_file,
mtz=phaser_result["mtz"],
pdb=phaser_result["pdb_file"])
if adf_results.get("adf"):
phaser_result.update({
"adf":
os.path.join(work_dir, adf_results.get("adf"))
})
if adf_results.get("peak"):
phaser_result.update({
"peak":
os.path.join(work_dir, adf_results.get("peak"))
})
#phaser_result.update({"adf": adf_results.get("adf", None),
# "peak": adf_results.get("peak", None),})
# print "1"
# print name
# New procedure for making tar of results
# Create directory
# Remove the run # from the name
# new_name = name[:-2] #
new_name = phaser_result.get("ID") #
# print new_name
os.mkdir(new_name)
# # Go through and copy files to archive directory
file_types = ("pdb_file", "mtz", "adf", "peak")
for file_type in file_types:
# print file_type
target_file = phaser_result.get(file_type, False)
# print target_file
if target_file:
if os.path.exists(target_file):
# Copy the file to the directory to be archived
shutil.copy(target_file, new_name + "/.")
# # Create the archive
archive_result = archive.create_archive(new_name)
archive_result["description"] = '%s_files' % new_name
phaser_result["tar"] = archive_result
# print "2"
else:
phaser_result = {
"ID": name,
"solution": False,
"message": "No solution",
"spacegroup": spacegroup
}
# Add the phaser log
if phaser_log:
phaser_result.update({"logs": {"phaser": phaser_log}})
# print "3"
if db_settings and tag:
print "db_settings and tag"
# Connect to Redis
redis = connect_to_redis(db_settings)
# Key should be deleted once received, but set the key to expire in 24 hours just in case.
redis.setex(tag, 86400, json.dumps(phaser_result))
# Do a little sleep to make sure results are in Redis for postprocess_phaser
time.sleep(0.1)
else:
# print "Printing phaser_result"
# Print the result so it can be seen thru the queue by reading stdout
# print phaser_result
print json.dumps(phaser_result)
def run_phaser_module_OLD(datafile, inp=False):
"""
Run separate module of Phaser to get results before running full job.
Setup so that I can read the data in once and run multiple modules.
"""
# if self.verbose:
# self.logger.debug('Utilities::runPhaserModule')
target_resolution = 0.0
z = 0
solvent_content = 0.0
def run_ellg():
res0 = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if f[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(f), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(f), 0.7)
# i.addSEAR_ENSE_NUM("junk",5)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
res0 = r1.get_target_resolution('model')
del(r1)
return res0
def run_cca():
z0 = 0
sc0 = 0.0
i0 = phaser.InputCCA()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
# Have to set high res limit!!
i0.setRESO_HIGH(res0)
if np > 0:
i0.addCOMP_PROT_NRES_NUM(np, 1)
if na > 0:
i0.addCOMP_NUCL_NRES_NUM(na, 1)
r1 = phaser.runCCA(i0)
#print r1.logfile()
if r1.Success():
z0 = r1.getBestZ()
sc0 = 1-(1.23/r1.getBestVM())
del(r1)
return (z0, sc0)
def run_ncs():
i0 = phaser.InputNCS()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setREFL_DATA(r.getDATA())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setLABI_F_SIGF(f,sigf)
i0.setMUTE(True)
# i0.setVERB(True)
r1 = phaser.runNCS(i0)
print r1.logfile()
print r1.loggraph().size()
print r1.loggraph().__dict__.keys()
#print r1.getCentricE4()
if r1.Success():
return(r1)
def run_ano():
#from cStringIO import StringIO
i0 = phaser.InputANO()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
# i0.setREFL(p.getMiller(),p.getF(),p.getSIGF())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setREFL_F_SIGF(p.getMiller(),p.getIobs(),p.getSigIobs())
i0.setREFL_DATA(r.getDATA())
i0.setMUTE(True)
r1 = phaser.runANO(i0)
print r1.loggraph().__dict__.keys()
print r1.loggraph().size()
print r1.logfile()
"""
o = phaser.Output()
redirect_str = StringIO()
o.setPackagePhenix(file_object=redirect_str)
r1 = phaser.runANO(i0,o)
"""
if r1.Success():
print 'SUCCESS'
return(r1)
# Setup which modules are run
matthews = False
if inp:
ellg = True
ncs = False
if type(inp) == str:
f = inp
else:
np, na, res0, f = inp
matthews = True
else:
ellg = False
ncs = True
# Read the dataset
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(datafile))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r = phaser.runMR_DAT(i)
if r.Success():
if ellg:
target_resolution = run_ellg()
if matthews:
z, solvent_content = run_cca()
if ncs:
n = run_ncs()
if matthews:
# Assumes ellg is run as well.
# return (z,sc,res)
return {"z": z,
"solvent_content": solvent_content,
"target_resolution": target_resolution}
elif ellg:
# ellg run by itself
# return target_resolution
return {"target_resolution": target_resolution}
else:
# NCS
return n
def run_phaser_module(data_file,
result_queue=False,
cca=False,
tncs=False,
ellg=False,
struct_file=False,
dres=False,
np=0,
na=0):
"""
Run separate module of Phaser to get results before running full job.
Setup so that I can read the data in once and run multiple modules.
data_file - input dataset mtz file
result_queue - pass results to queue
cca - Run CCA to determine number of molecules in AU, and solvent content (Matthew's Coefficient calc)
tncs - Run Anisotropy and tNCS correction on CID plots
ellg - Run analysis to determonine optimum Phaser resolution MR.
struct_file - input struct_file file. Could be a PDB or mmCIF file
dres - resolution of dataset (ELLG, CCA)
np - default number of protein residues (CCA)
na - default number of nucleic acid residues (CCA)
"""
target_resolution = 0.0
z = 0
solvent_content = 0.0
def run_ellg():
new_res = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
# Read in CIF file
if struct_file[-3:] in ('cif', ):
i0.addENSE_CIT_ID("model", convert_unicode(struct_file), 0.7)
# Read in PDB file
else:
i0.addENSE_PDB_ID("model", convert_unicode(struct_file), 0.7)
try:
r1 = phaser.runMR_ELLG(i0)
except RuntimeError as e:
# print "Hit error"
# Known CIF error - convert to pdb and retry
if struct_file[-3:] in ('cif', ):
# print "Convert to pdb"
pdb.cif_as_pdb((struct_file, ))
pdb_file = struct_file.replace(".cif", ".pdb")
i1 = phaser.InputMR_ELLG()
i1.setSPAC_HALL(r.getSpaceGroupHall())
i1.setCELL6(r.getUnitCell())
i1.setMUTE(True)
i1.setREFL_DATA(r.getDATA())
i1.addENSE_PDB_ID("model", convert_unicode(pdb_file), 0.7)
r1 = phaser.runMR_ELLG(i1)
else:
raise e
# print r1.logfile()
if r1.Success():
# If it worked use the recommended resolution
new_res = round(r1.get_target_resolution('model'), 1)
del (r1)
return new_res
def run_cca(res):
z0 = 0
sc0 = 0.0
i0 = phaser.InputCCA()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
# Have to set high res limit!!
i0.setRESO_HIGH(res)
if np > 0:
i0.addCOMP_PROT_NRES_NUM(np, 1)
if na > 0:
i0.addCOMP_NUCL_NRES_NUM(na, 1)
r1 = phaser.runCCA(i0)
#print r1.logfile()
#print dir(r1)
if r1.Success():
z0 = r1.getBestZ()
try:
sc0 = round(1 - (1.23 / r1.getBestVM()), 2)
except ZeroDivisionError:
sc0 = 0
del (r1)
return (z0, sc0)
def run_tncs():
# CAN'T GET READABLE loggraph?!?
i0 = phaser.InputNCS()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setREFL_DATA(r.getDATA())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setLABI_F_SIGF(f,sigf)
i0.setMUTE(True)
# i0.setVERB(True)
r1 = phaser.runNCS(i0)
print dir(r1)
print r1.logfile()
# for l in r1.loggraph():
# print l
print r1.loggraph().size()
print r1.output_strings
#print r1.hasTNCS()
#print r1.summary()
print r1.warnings()
print r1.ErrorMessage()
#print r1.getCentricE4()
if r1.Success():
return (r1.loggraph())
def run_ano():
#from cStringIO import StringIO
i0 = phaser.InputANO()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
# i0.setREFL(p.getMiller(),p.getF(),p.getSIGF())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setREFL_F_SIGF(p.getMiller(),p.getIobs(),p.getSigIobs())
i0.setREFL_DATA(r.getDATA())
i0.setMUTE(True)
r1 = phaser.runANO(i0)
print r1.loggraph().__dict__.keys()
print r1.loggraph().size()
print r1.logfile()
"""
o = phaser.Output()
redirect_str = StringIO()
o.setPackagePhenix(file_object=redirect_str)
r1 = phaser.runANO(i0,o)
"""
if r1.Success():
print 'SUCCESS'
return (r1)
# MAIN
# Setup which modules are run
# Read input MTZ file
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(data_file))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r = phaser.runMR_DAT(i)
if r.Success():
if ellg:
target_resolution = run_ellg()
if cca:
# Assumes ellg is run as well.
z, solvent_content = run_cca(target_resolution)
if tncs:
n = run_tncs()
if cca:
out = {
"z": z,
"solvent_content": solvent_content,
"target_resolution": target_resolution
}
if result_queue:
result_queue.put(out)
else:
return out
elif ellg:
# ellg run by itself
out = {"target_resolution": target_resolution}
if result_queue:
result_queue.put(out)
else:
return out
else:
# tNCS
out = n
if result_queue:
result_queue.put(out)
else:
return out
"""
def run_phaser_module(data_file,
result_queue=False,
cca=False,
tncs=False,
ellg=False,
mmcif=False,
dres=False,
np=0,
na=0,):
"""
Run separate module of Phaser to get results before running full job.
Setup so that I can read the data in once and run multiple modules.
data_file - input dataset mtz file
result_queue - pass results to queue
cca - Run CCA to determine number of molecules in AU, and solvent content (Matthew's Coefficient calc)
tncs - Run Anisotropy and tNCS correction on CID plots
ellg - Run analysis to determonine optimum Phaser resolution MR.
mmcif - input mmcif file. Could also be a PDB file
dres - resolution of dataset (ELLG, CCA)
np - default number of protein residues (CCA)
na - default number of nucleic acid residues (CCA)
"""
target_resolution = 0.0
z = 0
solvent_content = 0.0
def run_ellg():
new_res = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if mmcif[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(mmcif), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(mmcif), 0.7)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
# If it worked use the recommended resolution
new_res = round(r1.get_target_resolution('model'), 1)
del(r1)
return new_res
def run_cca(res):
z0 = 0
sc0 = 0.0
i0 = phaser.InputCCA()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
# Have to set high res limit!!
i0.setRESO_HIGH(res)
if np > 0:
i0.addCOMP_PROT_NRES_NUM(np, 1)
if na > 0:
i0.addCOMP_NUCL_NRES_NUM(na, 1)
r1 = phaser.runCCA(i0)
#print r1.logfile()
#print dir(r1)
if r1.Success():
z0 = r1.getBestZ()
sc0 = round(1-(1.23/r1.getBestVM()), 2)
del(r1)
return (z0, sc0)
def run_tncs():
# CAN'T GET READABLE loggraph?!?
i0 = phaser.InputNCS()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setREFL_DATA(r.getDATA())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setLABI_F_SIGF(f,sigf)
i0.setMUTE(True)
# i0.setVERB(True)
r1 = phaser.runNCS(i0)
print dir(r1)
print r1.logfile()
# for l in r1.loggraph():
# print l
print r1.loggraph().size()
print r1.output_strings
#print r1.hasTNCS()
#print r1.summary()
print r1.warnings()
print r1.ErrorMessage()
#print r1.getCentricE4()
if r1.Success():
return(r1.loggraph())
def run_ano():
#from cStringIO import StringIO
i0 = phaser.InputANO()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
# i0.setREFL(p.getMiller(),p.getF(),p.getSIGF())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setREFL_F_SIGF(p.getMiller(),p.getIobs(),p.getSigIobs())
i0.setREFL_DATA(r.getDATA())
i0.setMUTE(True)
r1 = phaser.runANO(i0)
print r1.loggraph().__dict__.keys()
print r1.loggraph().size()
print r1.logfile()
"""
o = phaser.Output()
redirect_str = StringIO()
o.setPackagePhenix(file_object=redirect_str)
r1 = phaser.runANO(i0,o)
"""
if r1.Success():
print 'SUCCESS'
return(r1)
# MAIN
# Setup which modules are run
# Read input MTZ file
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(data_file))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r = phaser.runMR_DAT(i)
if r.Success():
if ellg:
target_resolution = run_ellg()
if cca:
# Assumes ellg is run as well.
z, solvent_content = run_cca(target_resolution)
if tncs:
n = run_tncs()
if cca:
out = {"z": z,
"solvent_content": solvent_content,
"target_resolution": target_resolution}
if result_queue:
result_queue.put(out)
else:
return out
elif ellg:
# ellg run by itself
out = {"target_resolution": target_resolution}
if result_queue:
result_queue.put(out)
else:
return out
else:
# tNCS
out = n
if result_queue:
result_queue.put(out)
else:
return out
"""
def run_phaser(datafile,
spacegroup,
output_id,
db_settings,
work_dir=False,
cif=False,
pdb=False,
name=False,
ncopy=1,
cell_analysis=False,
resolution=False,
large_cell=False,
run_before=False,
):
"""
Run Phaser and passes results back to RAPD Redis DB
**Requires Phaser src code!**
datafile - input data as mtz
spacegroup - The space group to run MR
output_id - a Redis key where the results are sent
db_settings - Redis connection settings for sending results
work_dir - working directory
cif - input search model path in mmCIF format (do not use with 'pdb')
pdb - input search model path in PDB format (do not use with 'cif')
name - root name for output files
ncopy - number of molecules to search for
cell_analysis - internal RAPD signal so all possible SG's are searched
resolution - high res limit to run MR (float)
large_cell - optimizes parameters to speed up MR with large unit cell.
run_before - signal to run more comprehensive MR
"""
# Change to work_dir
if not work_dir:
work_dir = os.getcwd()
os.chdir(work_dir)
if not name:
name = spacegroup
# Connect to Redis
redis = connect_to_redis(db_settings)
# Read the dataset
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(datafile))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r = phaser.runMR_DAT(i)
if r.Success():
i = phaser.InputMR_AUTO()
# i.setREFL_DATA(r.getREFL_DATA())
# i.setREFL_DATA(r.DATA_REFL())
i.setREFL_F_SIGF(r.getMiller(), r.getFobs(), r.getSigFobs())
i.setCELL6(r.getUnitCell())
if cif:
#i.addENSE_CIF_ID('model', cif, 0.7)
### Typo in PHASER CODE!!!###
i.addENSE_CIT_ID('model', convert_unicode(cif), 0.7)
if pdb:
i.addENSE_PDB_ID('model', convert_unicode(pdb), 0.7)
i.addSEAR_ENSE_NUM("model", ncopy)
i.setSPAC_NAME(spacegroup)
if cell_analysis:
i.setSGAL_SELE("ALL")
# Set it for worst case in orth
# number of processes to run in parallel where possible
i.setJOBS(1)
else:
i.setSGAL_SELE("NONE")
if run_before:
# Picks own resolution
# Round 2, pick best solution as long as less that 10% clashes
i.setPACK_SELE("PERCENT")
i.setPACK_CUTO(0.1)
#command += "PACK CUTOFF 10\n"
else:
# For first round and cell analysis
# Only set the resolution limit in the first round or cell analysis.
if resolution:
i.setRESO_HIGH(resolution)
else:
# Otherwise it runs a second MR at full resolution!!
# I dont think a second round is run anymore.
# command += "RESOLUTION SEARCH HIGH OFF\n"
if large_cell:
i.setRESO_HIGH(6.0)
else:
i.setRESO_HIGH(4.5)
i.setSEAR_DEEP(False)
# Don"t seem to work since it picks the high res limit now.
# Get an error when it prunes all the solutions away and TF has no input.
# command += "PEAKS ROT SELECT SIGMA CUTOFF 4.0\n"
# command += "PEAKS TRA SELECT SIGMA CUTOFF 6.0\n"
# Turn off pruning in 2.6.0
i.setSEAR_PRUN(False)
# Choose more top peaks to help with getting it correct.
i.setPURG_ROTA_ENAB(True)
i.setPURG_ROTA_NUMB(3)
#command += "PURGE ROT ENABLE ON\nPURGE ROT NUMBER 3\n"
i.setPURG_TRAN_ENAB(True)
i.setPURG_TRAN_NUMB(1)
#command += "PURGE TRA ENABLE ON\nPURGE TRA NUMBER 1\n"
# Only keep the top after refinement.
i.setPURG_RNP_ENAB(True)
i.setPURG_RNP_NUMB(1)
#command += "PURGE RNP ENABLE ON\nPURGE RNP NUMBER 1\n"
i.setROOT(convert_unicode(name))
# i.setMUTE(False)
i.setMUTE(True)
# Delete the setup results
del(r)
# launch the run
r = phaser.runMR_AUTO(i)
if r.Success():
if r.foundSolutions():
print "Phaser has found MR solutions"
#print "Top LLG = %f" % r.getTopLLG()
#print "Top PDB file = %s" % r.getTopPdbFile()
else:
print "Phaser has not found any MR solutions"
else:
print "Job exit status FAILURE"
print r.ErrorName(), "ERROR :", r.ErrorMessage()
with open('phaser.log', 'w') as log:
log.write(r.logfile())
log.close()
with open('phaser_sum.log', 'w') as log:
log.write(r.summary())
log.close()
if r.foundSolutions():
rfz = None
tfz = None
tncs = False
# Parse results
for p in r.getTopSet().ANNOTATION.split():
if p.count('RFZ'):
if p.count('=') in [1]:
rfz = float(p.split('=')[-1])
if p.count('RF*0'):
rfz = "NC"
if p.count('TFZ'):
if p.count('=') in [1]:
tfz = p.split('=')[-1]
if tfz == '*':
tfz = 'arbitrary'
else:
tfz = float(tfz)
if p.count('TF*0'):
tfz = "NC"
tncs_test = [1 for line in r.getTopSet().unparse().splitlines()
if line.count("+TNCS")]
tncs = bool(len(tncs_test))
phaser_result = {"ID": name,
"solution": r.foundSolutions(),
"pdb": r.getTopPdbFile(),
"mtz": r.getTopMtzFile(),
"gain": float(r.getTopLLG()),
"rfz": rfz,
# "tfz": r.getTopTFZ(),
"tfz": tfz,
"clash": r.getTopSet().PAK,
"dir": os.getcwd(),
"spacegroup": r.getTopSet().getSpaceGroupName().replace(' ', ''),
"tNCS": tncs,
"nmol": r.getTopSet().NUM,
"adf": None,
"peak": None,
}
# make tar.bz2 of result files
# l = ['pdb', 'mtz', 'adf', 'peak']
# archive = "%s.tar.bz2" % name
# with tarfile.open(archive, "w:bz2") as tar:
# for f in l:
# fo = phaser_result.get(f, False)
# if fo:
# if os.path.exists(fo):
# tar.add(fo)
# tar.close()
# phaser_result['tar'] = os.path.join(work_dir, archive)
# New procedure for making tar of results
# Create directory
os.mkdir(name)
# Go through and copy files to archive directory
file_types = ("pdb", "mtz", "adf", "peak")
for file_type in file_types:
target_file = phaser_result.get(file_type, False)
if target_file:
if os.path.exists(target_file):
# Copy the file to the directory to be archived
shutil.copy(target_file, name+"/.")
# Create the archive
archive_result = archive.create_archive(name)
archive_result["description"] = name
phaser_result["tar"] = archive_result
phaser_result["pdb_file"] = os.path.join(work_dir, r.getTopPdbFile())
else:
phaser_result = {"ID": name,
"solution": False,
"message": "No solution"}
# Print the result so it can be seen in the rapd._phaser.log if needed
print phaser_result
# Key should be deleted once received, but set the key to expire in 24 hours just in case.
redis.setex(output_id, 86400, json.dumps(phaser_result))
# Do a little sleep to make sure results are in Redis for postprocess_phaser
time.sleep(0.1)
def __init__(self,
site,
command,
processed_results=False,
tprint=False,
logger=False,
verbosity=False):
"""Initialize the plugin"""
Thread.__init__(self)
# If the logging instance is passed in...
if logger:
self.logger = logger
else:
# Otherwise get the logger Instance
self.logger = logging.getLogger("RAPDLogger")
self.logger.debug("__init__")
# Keep track of start time
self.start_time = time.time()
# Store tprint for use throughout
if tprint:
self.tprint = tprint
# Dead end if no tprint passed
else:
def func(arg=False, level=False, verbosity=False, color=False):
"""Dummy function"""
pass
self.tprint = func
# Used for sending results back to DB referencing a dataset
self.processed_results = processed_results
# Some logging
self.logger.info(command)
self.verbose = verbosity
# Store passed-in variables
self.site = site
self.command = command
self.preferences = self.command.get("preferences", {})
# Params
self.working_dir = self.command["directories"].get("work", os.getcwd())
self.test = self.preferences.get("test", False)
#self.test = self.preferences.get("test", True) # Limit number of runs on cluster
#self.sample_type = self.preferences.get("type", "protein")
#self.solvent_content = self.preferences.get("solvent_content", 0.55)
self.clean = self.preferences.get("clean", True)
# self.verbose = self.command["preferences"].get("verbose", False)
self.data_file = xutils.convert_unicode(
self.command["input_data"].get("data_file"))
# Used for setting up Redis connection
self.db_settings = self.command["input_data"].get("db_settings")
#self.nproc = self.preferences.get("nproc", 1)
# If no launcher is passed in, use local_subprocess in a multiprocessing.Pool
self.computer_cluster = xutils.load_cluster_adapter(self)
if self.computer_cluster:
self.launcher = self.computer_cluster.process_cluster
self.batch_queue = self.computer_cluster.check_queue(
self.command.get('command'))
else:
self.launcher = local_subprocess
self.pool = mp_pool(self.preferences.get("nproc", cpu_count() - 1))
self.manager = mp_manager()
# Setup a multiprocessing pool if not using a computer cluster.
#if not self.computer_cluster:
# self.pool = mp_pool(self.nproc)
# Set Python path for subcontractors.rapd_phaser
self.rapd_python = "rapd.python"
if self.site:
if hasattr(self.site, "RAPD_PYTHON_PATH"):
self.rapd_python = self.site.RAPD_PYTHON_PATH
def run_phaser_module_OLD(data_file, inp=False):
"""
Run separate module of Phaser to get results before running full job.
Setup so that I can read the data in once and run multiple modules.
"""
# if self.verbose:
# self.logger.debug('Utilities::runPhaserModule')
target_resolution = 0.0
z = 0
solvent_content = 0.0
def run_ellg():
res0 = 0.0
i0 = phaser.InputMR_ELLG()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
i0.setREFL_DATA(r.getDATA())
if f[-3:] in ('cif'):
i0.addENSE_CIT_ID('model', convert_unicode(f), 0.7)
else:
i0.addENSE_PDB_ID("model", convert_unicode(f), 0.7)
# i.addSEAR_ENSE_NUM("junk",5)
r1 = phaser.runMR_ELLG(i0)
#print r1.logfile()
if r1.Success():
res0 = r1.get_target_resolution('model')
del (r1)
return res0
def run_cca():
z0 = 0
sc0 = 0.0
i0 = phaser.InputCCA()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setMUTE(True)
# Have to set high res limit!!
i0.setRESO_HIGH(res0)
if np > 0:
i0.addCOMP_PROT_NRES_NUM(np, 1)
if na > 0:
i0.addCOMP_NUCL_NRES_NUM(na, 1)
r1 = phaser.runCCA(i0)
#print r1.logfile()
if r1.Success():
z0 = r1.getBestZ()
sc0 = 1 - (1.23 / r1.getBestVM())
del (r1)
return (z0, sc0)
def run_ncs():
i0 = phaser.InputNCS()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
i0.setREFL_DATA(r.getDATA())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setLABI_F_SIGF(f,sigf)
i0.setMUTE(True)
# i0.setVERB(True)
r1 = phaser.runNCS(i0)
print r1.logfile()
print r1.loggraph().size()
print r1.loggraph().__dict__.keys()
#print r1.getCentricE4()
if r1.Success():
return (r1)
def run_ano():
#from cStringIO import StringIO
i0 = phaser.InputANO()
i0.setSPAC_HALL(r.getSpaceGroupHall())
i0.setCELL6(r.getUnitCell())
# i0.setREFL(p.getMiller(),p.getF(),p.getSIGF())
# i0.setREFL_F_SIGF(r.getMiller(),r.getF(),r.getSIGF())
# i0.setREFL_F_SIGF(p.getMiller(),p.getIobs(),p.getSigIobs())
i0.setREFL_DATA(r.getDATA())
i0.setMUTE(True)
r1 = phaser.runANO(i0)
print r1.loggraph().__dict__.keys()
print r1.loggraph().size()
print r1.logfile()
"""
o = phaser.Output()
redirect_str = StringIO()
o.setPackagePhenix(file_object=redirect_str)
r1 = phaser.runANO(i0,o)
"""
if r1.Success():
print 'SUCCESS'
return (r1)
# Setup which modules are run
matthews = False
if inp:
ellg = True
ncs = False
if type(inp) == str:
f = inp
else:
np, na, res0, f = inp
matthews = True
else:
ellg = False
ncs = True
# Read the dataset
i = phaser.InputMR_DAT()
i.setHKLI(convert_unicode(data_file))
i.setLABI_F_SIGF('F', 'SIGF')
i.setMUTE(True)
r = phaser.runMR_DAT(i)
if r.Success():
if ellg:
target_resolution = run_ellg()
if matthews:
z, solvent_content = run_cca()
if ncs:
n = run_ncs()
if matthews:
# Assumes ellg is run as well.
# return (z,sc,res)
return {
"z": z,
"solvent_content": solvent_content,
"target_resolution": target_resolution
}
elif ellg:
# ellg run by itself
# return target_resolution
return {"target_resolution": target_resolution}
else:
# NCS
return n
def __init__(self,
command,
site=False,
processed_results=False,
tprint=False,
logger=False,
verbosity=False):
"""Initialize the plugin"""
Thread.__init__(self)
# If the logging instance is passed in...
if logger:
self.logger = logger
else:
# Otherwise get the logger Instance
self.logger = logging.getLogger("RAPDLogger")
self.logger.debug("__init__")
# Keep track of start time
self.start_time = time.time()
# Store tprint for use throughout
if tprint:
self.tprint = tprint
# Dead end if no tprint passed
else:
def func(arg=False, level=False, verbosity=False, color=False):
"""Dummy function"""
pass
self.tprint = func
# Used for sending results back to DB referencing a dataset
self.processed_results = processed_results
# Some logging
self.logger.info(command)
self.verbose = verbosity
# Store passed-in variables
self.site = site
self.command = command
self.preferences = self.command.get("preferences", {})
# Params
self.working_dir = self.command["directories"].get("work", os.getcwd())
#self.test = self.preferences.get("test", False)
self.test = self.preferences.get(
"test", True) # Limit number of runs on cluster
#self.sample_type = self.preferences.get("type", "protein")
#self.solvent_content = self.preferences.get("solvent_content", 0.55)
# Number of molecules specified
#self.nmol = self.preferences.get('nmol', False)
# Input data MTZ file
self.data_file = xutils.convert_unicode(
self.command["input_data"].get("data_file"))
# Input PDB/mmCIF file or PDB code.
self.struct_file = xutils.convert_unicode(
self.command["input_data"].get("struct_file"))
# Save preferences
self.clean = self.preferences.get("clean", True)
# Calc ADF for each solution (creates a lot of big map files).
self.adf = self.preferences.get("adf", False)
# Check if there is a computer cluster and load adapter.
self.computer_cluster = xutils.load_cluster_adapter(self)
if self.computer_cluster:
self.launcher = self.computer_cluster.process_cluster
self.batch_queue = self.computer_cluster.check_queue(
self.command.get('command'))
else:
# if NOT using a computer cluster setup a multiprocessing.pool and manager for queues.
self.launcher = local_subprocess
self.pool = mp_pool(self.preferences.get("nproc", cpu_count() - 1))
self.manager = mp_manager()
# Set Python path for subcontractors.rapd_phaser
self.rapd_python = "rapd.python"
if self.site and hasattr(self.site, "RAPD_PYTHON_PATH"):
self.rapd_python = self.site.RAPD_PYTHON_PATH
def get_pdb_info(struct_file, data_file, dres, matthews=True, chains=True):
"""Get info from PDB or mmCIF file"""
# Get rid of ligands and water so Phenix won't error.
np = 0
na = 0
nmol = 1
sc = 0.55
nchains = 0
res1 = 0.0
d = {}
l = []
# Read in the file
struct_file = convert_unicode(struct_file)
if struct_file[-3:].lower() == 'cif':
root = iotbx_mmcif.cif_input(
file_name=struct_file).construct_hierarchy()
else:
root = iotbx_pdb.input(struct_file).construct_hierarchy()
# Go through the chains
for chain in root.models()[0].chains():
# Number of protein residues
np1 = 0
# Number of nucleic acid residues
na1 = 0
# Sometimes Hetatoms are AA with same segid.
if l.count(chain.id) == 0:
l.append(chain.id)
repeat = False
nchains += 1
else:
repeat = True
# Count the number of AA and NA in pdb file.
for rg in chain.residue_groups():
if rg.atoms()[0].parent(
).resname in iotbx_pdb.common_residue_names_amino_acid:
np1 += 1
if rg.atoms()[0].parent(
).resname in iotbx_pdb.common_residue_names_rna_dna:
na1 += 1
# Not sure if I get duplicates?
if rg.atoms()[0].parent().resname in \
iotbx_pdb.common_residue_names_ccp4_mon_lib_rna_dna:
na1 += 1
# Limit to 10 chains?!?
if nchains < 10:
# Do not split up PDB if run from cell analysis
if chains and not repeat:
# Save info for each chain.
if np1 or na1:
# Write new pdb files for each chain.
temp = iotbx_pdb.hierarchy.new_hierarchy_from_chain(chain)
# Long was of making sure that user does not have directory named '.pdb' or
# '.cif'
#n = os.path.join(os.path.dirname(struct_file), "%s_%s.pdb" % \
n = os.path.join(os.path.dirname(struct_file), "%s_%s.cif" % \
(os.path.basename(struct_file)[:os.path.basename(struct_file).find('.')], \
chain.id))
#temp.write_pdb_file(file_name=n)
# Write chain as mmCIF file.
temp.write_mmcif_file(file_name=n)
d[chain.id] = {
'file': n,
'NRes': np1 + na1,
'MWna': na1 * 330,
'MWaa': np1 * 110,
'MW': na1 * 330 + np1 * 110
}
if matthews:
# Run Matthews Calc. on chain
#phaser_return = run_phaser_module((np1, na1, dres, n, data_file))
#phaser_return = run_phaser_module(data_file, (np1, na1, dres, n))
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
cca=True,
struct_file=n,
dres=dres,
np=np1,
na=na1)
d[chain.id].update({
'NMol':
phaser_return.get("z", nmol),
'SC':
phaser_return.get("solvent_content", sc),
'res':
phaser_return.get("target_resolution", res1)
})
else:
#res1 = run_phaser_module(n)
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
struct_file=n)
d[chain.id].update({
'res':
phaser_return.get("target_resolution", res1)
})
"""
d[chain.id] = {'file': n,
'NRes': np1+na1,
'MWna': na1*330,
'MWaa': np1*110,
'MW': na1*330+np1*110,
'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)}
"""
# Add up residue count
np += np1
na += na1
d['all'] = {
'file': struct_file,
'NRes': np + na,
'MWna': na * 330,
'MWaa': np * 110,
'MW': na * 330 + np * 110
}
# Run on entire PDB
if matthews:
#phaser_return = run_phaser_module((np, na, dres, struct_file, data_file))
#phaser_return = run_phaser_module(data_file, (np, na, dres, struct_file))
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
cca=True,
struct_file=struct_file,
dres=dres,
np=np,
na=na)
d['all'].update({
'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)
})
else:
#phaser_return = run_phaser_module((np, na, dres, struct_file, data_file))
#phaser_return = run_phaser_module(data_file, (np, na, dres, struct_file))
# phaser_return = run_phaser_module(data_file=data_file,
# ellg=True,
# struct_file=struct_file)
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
struct_file=struct_file)
d['all'].update({'res': phaser_return.get("target_resolution", res1)})
"""
d['all'] = {'file': struct_file,
'NRes': np+na,
'MWna': na*330,
'MWaa': np*110,
'MW': na*330+np*110,
'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)}
"""
return d
def get_pdb_info(cif_file, data_file, dres, matthews=True, chains=True):
"""Get info from PDB of mmCIF file"""
# Get rid of ligands and water so Phenix won't error.
np = 0
na = 0
nmol = 1
sc = 0.55
nchains = 0
res1 = 0.0
d = {}
l = []
# Read in the file
cif_file = convert_unicode(cif_file)
if cif_file[-3:].lower() == 'cif':
root = iotbx_mmcif.cif_input(file_name=cif_file).construct_hierarchy()
else:
root = iotbx_pdb.input(cif_file).construct_hierarchy()
# Go through the chains
for chain in root.models()[0].chains():
# Number of protein residues
np1 = 0
# Number of nucleic acid residues
na1 = 0
# Sometimes Hetatoms are AA with same segid.
if l.count(chain.id) == 0:
l.append(chain.id)
repeat = False
nchains += 1
else:
repeat = True
# Count the number of AA and NA in pdb file.
for rg in chain.residue_groups():
if rg.atoms()[0].parent().resname in iotbx_pdb.common_residue_names_amino_acid:
np1 += 1
if rg.atoms()[0].parent().resname in iotbx_pdb.common_residue_names_rna_dna:
na1 += 1
# Not sure if I get duplicates?
if rg.atoms()[0].parent().resname in \
iotbx_pdb.common_residue_names_ccp4_mon_lib_rna_dna:
na1 += 1
# Limit to 10 chains?!?
if nchains < 10:
# Do not split up PDB if run from cell analysis
if chains and not repeat:
# Save info for each chain.
if np1 or na1:
# Write new pdb files for each chain.
temp = iotbx_pdb.hierarchy.new_hierarchy_from_chain(chain)
# Long was of making sure that user does not have directory named '.pdb' or
# '.cif'
#n = os.path.join(os.path.dirname(cif_file), "%s_%s.pdb" % \
n = os.path.join(os.path.dirname(cif_file), "%s_%s.cif" % \
(os.path.basename(cif_file)[:os.path.basename(cif_file).find('.')], \
chain.id))
#temp.write_pdb_file(file_name=n)
temp.write_mmcif_file(file_name=n)
d[chain.id] = {'file': n,
'NRes': np1+na1,
'MWna': na1*330,
'MWaa': np1*110,
'MW': na1*330+np1*110}
if matthews:
# Run Matthews Calc. on chain
#phaser_return = run_phaser_module((np1, na1, dres, n, data_file))
#phaser_return = run_phaser_module(data_file, (np1, na1, dres, n))
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
cca=True,
mmcif=n,
dres=dres,
np=np1,
na=na1)
d[chain.id].update({'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)})
else:
#res1 = run_phaser_module(n)
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
mmcif=n)
d[chain.id].update({'res': phaser_return.get("target_resolution", res1)})
"""
d[chain.id] = {'file': n,
'NRes': np1+na1,
'MWna': na1*330,
'MWaa': np1*110,
'MW': na1*330+np1*110,
'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)}
"""
# Add up residue count
np += np1
na += na1
d['all'] = {'file': cif_file,
'NRes': np+na,
'MWna': na*330,
'MWaa': np*110,
'MW': na*330+np*110}
# Run on entire PDB
if matthews:
#phaser_return = run_phaser_module((np, na, dres, cif_file, data_file))
#phaser_return = run_phaser_module(data_file, (np, na, dres, cif_file))
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
cca=True,
mmcif=cif_file,
dres=dres,
np=np,
na=na)
d['all'].update({'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)})
else:
#phaser_return = run_phaser_module((np, na, dres, cif_file, data_file))
#phaser_return = run_phaser_module(data_file, (np, na, dres, cif_file))
phaser_return = run_phaser_module(data_file=data_file,
ellg=True,
mmcif=cif_file)
d['all'].update({'res': phaser_return.get("target_resolution", res1)})
"""
d['all'] = {'file': cif_file,
'NRes': np+na,
'MWna': na*330,
'MWaa': np*110,
'MW': na*330+np*110,
'NMol': phaser_return.get("z", nmol),
'SC': phaser_return.get("solvent_content", sc),
'res': phaser_return.get("target_resolution", res1)}
"""
return d
def __init__(self, command, processed_results=False, computer_cluster=False, tprint=False, logger=False, verbosity=False):
"""Initialize the plugin"""
Thread.__init__ (self)
# If the logging instance is passed in...
if logger:
self.logger = logger
else:
# Otherwise get the logger Instance
self.logger = logging.getLogger("RAPDLogger")
self.logger.debug("__init__")
# Keep track of start time
self.start_time = time.time()
# Store tprint for use throughout
if tprint:
self.tprint = tprint
# Dead end if no tprint passed
else:
def func(arg=False, level=False, verbosity=False, color=False):
"""Dummy function"""
pass
self.tprint = func
# Used for sending results back to DB referencing a dataset
self.processed_results = processed_results
# Some logging
self.logger.info(command)
self.verbose = verbosity
# Store passed-in variables
self.command = command
self.preferences = self.command.get("preferences", {})
# Params
self.working_dir = self.command["directories"].get("work", os.getcwd())
self.test = self.preferences.get("test", False)
self.sample_type = self.preferences.get("type", "protein")
self.solvent_content = self.preferences.get("solvent_content", 0.55)
self.clean = self.preferences.get("clean", True)
# self.verbose = self.command["preferences"].get("verbose", False)
self.datafile = xutils.convert_unicode(self.command["input_data"].get("datafile"))
# Used for setting up Redis connection
self.db_settings = self.command["input_data"].get("db_settings")
self.nproc = self.preferences.get("nproc", 1)
self.computer_cluster = self.preferences.get("computer_cluster", False)
# If no launcher is passed in, use local_subprocess in a multiprocessing.Pool
self.computer_cluster = computer_cluster
if self.computer_cluster:
self.launcher = self.computer_cluster.process_cluster
self.batch_queue = self.computer_cluster.check_queue(self.command.get('command'))
else:
self.launcher = local_subprocess
# Setup a multiprocessing pool if not using a computer cluster.
if not self.computer_cluster:
self.pool = mp_pool(self.nproc)