def place_native_pdb(native_pdb,
native_mtz,
f_label,
sigf_label,
cleanup=True):
"""Place native_pdb into data from native_mtz using phaser with f_label and sigf_label"""
# get crystal info from pdb
sym_obj = iotbx.pdb.pdb_input(file_name=native_pdb).crystal_symmetry()
hall_symbol = sym_obj.space_group().type().hall_symbol()
unit_cell = sym_obj.unit_cell().parameters()
molecular_weight = pdb_edit.molecular_weight(native_pdb)
# Get data from MTZ file
mtz_object = _check_mtz(native_mtz, labels=[f_label, sigf_label])
miller_indices = mtz_object.extract_miller_indices()
fp_obs = mtz_object.get_column(f_label).extract_values().as_double()
sigfp_obs = mtz_object.get_column(sigf_label).extract_values().as_double()
# Create root_name for all files from native + mr_mtz
name1 = os.path.splitext(os.path.basename(native_pdb))[0]
name2 = os.path.splitext(os.path.basename(native_mtz))[0]
root_name = "{0}_{1}".format(name1, name2)
# Run phaser to position native
phaser_input = phaser.InputMR_AUTO()
phaser_input.setSPAC_HALL(hall_symbol)
phaser_input.setCELL6(unit_cell)
phaser_input.setREFL_F_SIGF(miller_indices, fp_obs, sigfp_obs)
# Make sure labels same as native_mtz
phaser_input.setLABI_F_SIGF(f_label, sigf_label)
phaser_input.setROOT(root_name)
phaser_input.addENSE_PDB_ID("native", native_pdb, 0.0)
phaser_input.addCOMP_PROT_MW_NUM(molecular_weight, 1)
phaser_input.addSEAR_ENSE_NUM("native", 1)
phaser_input.setMUTE(True)
phaser_run = phaser.runMR_AUTO(phaser_input)
del phaser_input
if not phaser_run.Success():
raise RuntimeError("PHASER failed: {0} : {1}".format(
phaser_run.ErrorName(), phaser_run.ErrorMessage()))
if phaser_run.foundSolutions():
#print "Phaser has found MR solutions"
#print "Top LLG = %f" % phaser_run.getTopLLG()
#print "Top PDB file = %s" % phaser_run.getTopPdbFile()
#print "Top MTZ file = %s" % phaser_run.getTopMtzFile()
native_placed_mtz = phaser_run.getTopMtzFile()
else:
raise RuntimeError("PHASER could not place native_pdb")
if cleanup:
os.unlink(phaser_run.getTopPdbFile())
return native_placed_mtz
def place_native_pdb(native_pdb, native_mtz, f_label, sigf_label, cleanup=True):
"""Place native_pdb into data from native_mtz using phaser with f_label and sigf_label"""
# get crystal info from pdb
sym_obj = iotbx.pdb.pdb_input(file_name=native_pdb).crystal_symmetry()
hall_symbol = sym_obj.space_group().type().hall_symbol()
unit_cell = sym_obj.unit_cell().parameters()
molecular_weight = pdb_edit.molecular_weight(native_pdb)
# Get data from MTZ file
mtz_object = _check_mtz(native_mtz, labels = [f_label, sigf_label])
miller_indices = mtz_object.extract_miller_indices()
fp_obs = mtz_object.get_column(f_label).extract_values().as_double()
sigfp_obs = mtz_object.get_column(sigf_label).extract_values().as_double()
# Create root_name for all files from native + mr_mtz
name1 = os.path.splitext(os.path.basename(native_pdb))[0]
name2 = os.path.splitext(os.path.basename(native_mtz))[0]
root_name = "{0}_{1}".format(name1, name2)
# Run phaser to position native
phaser_input = phaser.InputMR_AUTO()
phaser_input.setSPAC_HALL(hall_symbol)
phaser_input.setCELL6(unit_cell)
phaser_input.setREFL_F_SIGF(miller_indices, fp_obs, sigfp_obs)
# Make sure labels same as native_mtz
phaser_input.setLABI_F_SIGF(f_label,sigf_label)
phaser_input.setROOT(root_name)
phaser_input.addENSE_PDB_ID("native", native_pdb, 0.0)
phaser_input.addCOMP_PROT_MW_NUM(molecular_weight, 1)
phaser_input.addSEAR_ENSE_NUM("native", 1)
phaser_input.setMUTE(True)
phaser_run = phaser.runMR_AUTO(phaser_input)
del phaser_input
if not phaser_run.Success():
raise RuntimeError("PHASER failed: {0} : {1}".format(phaser_run.ErrorName(), phaser_run.ErrorMessage()))
if phaser_run.foundSolutions():
native_placed_mtz = phaser_run.getTopMtzFile()
else:
raise RuntimeError("PHASER could not place native_pdb")
if cleanup:
os.unlink(phaser_run.getTopPdbFile())
return native_placed_mtz
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 run(self):
"""Function to run molecular replacement using PHASER
Returns
-------
file
Output pdb file
file
Output log file
"""
# Make a note of the current working directory
current_work_dir = os.getcwd()
# Change to the PHASER working directory
if os.path.exists(self.work_dir):
os.chdir(self.work_dir)
else:
os.makedirs(self.work_dir)
os.chdir(self.work_dir)
# Copy hklin and pdbin to working dire for efficient running of PHASER
hklin = os.path.join(self.work_dir, os.path.basename(self.hklin))
shutil.copyfile(self.hklin, hklin)
pdbin = os.path.join(self.work_dir, os.path.basename(self.pdbin))
shutil.copyfile(self.pdbin, pdbin)
i = InputMR_DAT()
i.setHKLI(hklin)
if self.hires:
i.setHIRES(self.hires)
if self.autohigh:
i.setRESO_AUTO_HIGH(self.autohigh)
if self.i != "None" and self.sigi != "None":
i.setLABI_I_SIGI(self.i, self.sigi)
elif self.f != "None" and self.sigf != "None":
i.setLABI_F_SIGF(self.f, self.sigf)
else:
msg = "No flags for intensities or amplitudes have been provided"
raise RuntimeError(msg)
i.setSGAL_SELE(SGAlternatives[self.sgalternative].value)
i.setMUTE(True)
r = runMR_DAT(i)
if r.Success():
i = InputMR_AUTO()
i.setJOBS(1)
i.setREFL_DATA(r.getREFL_DATA())
i.setROOT("phaser_mr_output")
i.addENSE_PDB_ID("PDB", pdbin, 0.7)
i.setENSE_DISA_CHEC('PDB', True)
i.setCOMP_BY("SOLVENT")
i.setCOMP_PERC(self.solvent)
# nmol set to one for testing
i.addSEAR_ENSE_NUM('PDB', 1)
i.setSGAL_SELE(SGAlternatives[self.sgalternative].value)
if self.timeout != 0:
i.setKILL_TIME(self.timeout)
i.setMUTE(True)
del(r)
r = runMR_AUTO(i)
with open(self.logfile, 'w') as f:
f.write(r.summary())
shutil.move(r.getTopPdbFile(), self.pdbout)
# Output original mtz with a change of basis if needed
original_space_group, _, _ = mtz_util.crystal_data(self.hklin)
space_group, _, _ = mtz_util.crystal_data(r.getTopMtzFile())
if original_space_group != space_group:
mtz_util.reindex(self.hklin, self.hklout, space_group)
else:
shutil.copyfile(self.hklin, self.hklout)
# Return to original working directory
os.chdir(current_work_dir)
# Delete any files copied across
if os.path.isfile(os.path.join(self.work_dir, os.path.basename(self.hklin))):
os.remove(os.path.join(self.work_dir, os.path.basename(self.hklin)))
if os.path.isfile(os.path.join(self.work_dir, os.path.basename(self.pdbin))):
os.remove(os.path.join(self.work_dir, os.path.basename(self.pdbin)))
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(self):
"""Function to run molecular replacement using PHASER
Returns
-------
file
Output pdb file
file
Output log file
"""
# Make a note of the current working directory
current_work_dir = os.getcwd()
# Change to the PHASER working directory
if os.path.exists(self.work_dir):
os.chdir(self.work_dir)
else:
os.makedirs(self.work_dir)
os.chdir(self.work_dir)
# Copy hklin and pdbin to working dire for efficient running of PHASER
hklin = os.path.join(self.work_dir, os.path.basename(self.hklin))
shutil.copyfile(self.hklin, hklin)
pdbin = os.path.join(self.work_dir, os.path.basename(self.pdbin))
shutil.copyfile(self.pdbin, pdbin)
i = InputMR_DAT()
i.setHKLI(hklin)
if self.hires:
i.setHIRES(self.hires)
if self.autohigh:
i.setRESO_AUTO_HIGH(self.autohigh)
if self.i != "None" and self.sigi != "None":
i.setLABI_I_SIGI(self.i, self.sigi)
elif self.f != "None" and self.sigf != "None":
i.setLABI_F_SIGF(self.f, self.sigf)
else:
msg = "No flags for intensities or amplitudes have been provided"
raise RuntimeError(msg)
i.setSGAL_SELE(SGAlternatives[self.sgalternative].value)
i.setMUTE(True)
r = runMR_DAT(i)
if r.Success():
i = InputMR_AUTO()
i.setJOBS(1)
i.setREFL_DATA(r.getREFL_DATA())
i.setROOT("phaser_mr_output")
i.addENSE_PDB_ID("PDB", pdbin, 0.7)
i.setCOMP_BY("SOLVENT")
i.setCOMP_PERC(self.solvent)
i.addSEAR_ENSE_NUM('PDB', self.nmol)
i.setSGAL_SELE(SGAlternatives[self.sgalternative].value)
if self.timeout != 0:
i.setKILL_TIME(self.timeout)
i.setMUTE(True)
del (r)
r = runMR_AUTO(i)
with open(self.logfile, 'w') as f:
f.write(r.summary())
shutil.move(r.getTopPdbFile(), self.pdbout)
# Output original mtz with a change of basis if needed
space_group, _, _ = mtz_util.crystal_data(r.getTopMtzFile())
ed = mtz_util.ExperimentalData(self.hklin)
ed.change_space_group(space_group)
ed.output_mtz(self.hklout)
# Return to original working directory
os.chdir(current_work_dir)
# Delete any files copied across
if os.path.isfile(os.path.join(self.work_dir, os.path.basename(self.hklin))):
os.remove(os.path.join(self.work_dir, os.path.basename(self.hklin)))
if os.path.isfile(os.path.join(self.work_dir, os.path.basename(self.pdbin))):
os.remove(os.path.join(self.work_dir, os.path.basename(self.pdbin)))
def run_phaser(root, xml_file, xmlroot, docid, output, logfile, data, formfactors, pdbin, copies, rms, pdbin_fixed,
seqin, ncs_copies, phaser_solvent, sgalternative, sg_list, input_hand, tncs,
search_lowres, search_highres, rot_peaks, rot_cluster_off,
rot_samp, tra_samp, search_down, tfz_solved, solutions, purge,
rescore_strands, rescore_residues, rescore_models, nproc):
input = phaser.InputMR_AUTO()
input.setRESO(search_highres, search_lowres)
# test OpenMP
if nproc > 1:
input.setJOBS(nproc)
else:
input.setJOBS(1)
if rot_samp is not None:
input.setSAMP_ROTA(rot_samp)
if tra_samp is not None:
input.setSAMP_TRAN(tra_samp)
input.setSPAC_HALL(data.getSpaceGroupHall())
input.setCELL6(data.getUnitCell())
input.setREFL_DATA(data.getDATA())
# electron formfactors
if formfactors == 'electron':
input.setFORM('ELECTRON')
input.addENSE_PDB_RMS(root, pdbin, rms)
input.setROOT(root)
# fixed solution
if pdbin_fixed is not None:
fixed_root = pdbin_fixed[:-4]
input.addENSE_PDB_RMS(fixed_root, pdbin_fixed, rms)
input.addSOLU_6DIM_ENSE(fixed_root, [0.0, 0.0, 0.0], False, [0.0, 0.0, 0.0], 0.0, False, False, False, 1.0, 1.0)
if phaser_solvent is not None:
input.setCOMP_SOLV()
input.setCOMP_PERC(phaser_solvent)
else:
input.addCOMP_PROT_SEQ_NUM(seqin, ncs_copies)
input.addSEAR_ENSE_NUM(root, copies)
if rot_peaks is not None:
input.setPEAK_ROTA_SELE('NUMBER')
input.setPEAK_ROTA_CUTO(rot_peaks)
if rot_cluster_off:
input.setPEAK_ROTA_CLUS(False)
if sgalternative:
input.setSGAL_SELE('ALL')
elif input_hand:
input.setSGAL_SELE('NONE')
elif sg_list is not None:
input.setSGAL_SELE('LIST')
for sg in sg_list.split(','):
input.addSGAL_TEST(sg)
else:
input.setSGAL_SELE('HAND')
input.setTOPF(solutions)
if rescore_strands or rescore_residues or rescore_models:
# we don't need to write out the files here
input.setXYZO(False)
input.setHKLO(False)
input.setXYZO_ENSE(False)
# turn off tNCS correction if required
if tncs:
# to turn restore default MR_AUTO behaviour needs this
input.setTNCS_PATT_PERC(99.9)
# in desparate cases search further down the RF and TF list
if search_down is not None:
input.setSEAR_DOWN_PERC(search_down)
input.setPURG_TRAN_PERC(75.00 - search_down)
# turn purging on if requested for multicopy searches
if purge is not None:
input.setPURG_RNP_ENAB(True)
input.setPURG_RNP_NUMB(purge)
# turn it off otherwise (so more solutions are output)
else:
input.setPURG_RNP_ENAB(False)
# these are not default but fragments should never clash.
input.setPACK_KEEP_HIGH_TFZ(False)
input.setSEAR_PRUN(False)
input.setTRAN_PACK_CUTO(0.0)
input.setPACK_CUTO(0.0)
# end non defaults
input.setZSCO_USE(True)
input.setZSCO_SOLV(tfz_solved)
if xml_file is not None and xmlroot is not None:
input.setOUTP_LEVE('LOGFILE') #i2
elif docid is not None:
input.setOUTP_LEVE('LOGFILE') #rvapi
else:
input.setOUTP_LEVE('SUMMARY') #command line
phaser_output = phaser.Output()
phaser_log = open(logfile, 'w')
output_object = CallbackObject(xml_file, xmlroot, docid, output)
phaser_output.setPhenixPackageCallback(output_object)
phaser_output.set_file_object(phaser_log)
mr = phaser.runMR_AUTO(input, phaser_output)
phaser_log.close()
if mr.Success():
if mr.foundSolutions() :
log.info('\nPhaser has found %d solution(s)' % mr.numSolutions())
if not rescore_strands or rescore_residues:
log.info('Top solution PDB file = %s' % mr.getTopPdbFile())
log.info('Top solution LLG = %f' % mr.getTopLLG())
return mr.getDotSol()
else:
log.info('\nPhaser failed to find any solutions')
return 'No solutions'
else:
log.info('\nPhaser failed to find any solutions')
return 'No solutions'