def import_redundant_nucleotides(self, pdbs, recalculate=False):
"""
"""
try:
logging.info("Importing redundant nucleotides")
if not recalculate:
recalculate = self.config["recalculate"]["redundant_nts"]
if recalculate:
pdb_list = pdbs
self.__delete_old_data(pdbs)
else:
pdb_list = self.filter_out_analyzed_pdbs(pdbs, "redundant_nts")
if pdb_list:
MotifAtlasBaseClass._setup_matlab(self)
for pdb_file in pdb_list:
logging.info("Running matlab on %s", pdb_file)
ifn, err_msg = self.mlab.loadRedundantNucleotides(pdb_file, nout=2)
if err_msg == "":
self.__import_temporary_file(ifn, pdb_file)
else:
MotifAtlasBaseClass._crash(self, err_msg)
self.mark_pdb_as_analyzed(pdb_file, "redundant_nts")
logging.info("%s", "=" * 40)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self, e)
def import_loops(self, Loops, l, pdb_id, loop_type):
"""
"""
try:
if Loops == 0:
self.mark_pdb_as_analyzed(pdb_id, loop_type)
return
for i in xrange(l):
loop_id = self._get_loop_id(Loops[i].AllLoops_table.full_id,
pdb_id, loop_type)
Loops[i].Filename = loop_id
session.merge(
AllLoops(id = loop_id,
type = loop_type,
pdb = pdb_id,
sequential_id = loop_id[-3:],
length = int(Loops[i].NumNT[0][0]),
seq = Loops[i].AllLoops_table.seq,
r_seq = Loops[i].AllLoops_table.r_seq,
nwc_seq = Loops[i].AllLoops_table.nwc,
r_nwc_seq = Loops[i].AllLoops_table.r_nwc,
pdb_file = Loops[i].PDBFilename,
nt_ids = Loops[i].AllLoops_table.full_id,
loop_name = Loops[i].AllLoops_table.loop_name))
self.save_mat_files(Loops)
self.mark_pdb_as_analyzed(pdb_id, loop_type)
logging.info('%s from %s successfully imported', loop_type, pdb_id)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self,e)
def import_best_chains_and_models(self, pdbs, recalculate=False):
"""
"""
try:
logging.info("Importing best chains and models")
if not recalculate:
recalculate = self.config["recalculate"]["best_chains_and_models"]
if recalculate:
pdb_list = pdbs
self.__delete_old_data(pdbs)
else:
pdb_list = self.filter_out_analyzed_pdbs(pdbs, "best_chains_and_models")
if pdb_list:
MotifAtlasBaseClass._setup_matlab(self)
for pdb_file in pdb_list:
logging.info("Running matlab on %s", pdb_file)
# 'ABC', '1,2', ''
best_chains, best_models, err_msg = self.mlab.loadBestChainsAndModels(pdb_file, nout=3)
best_chains = ",".join(list(best_chains))
if err_msg == "":
self.__import_into_db(pdb_file, best_chains, best_models)
else:
MotifAtlasBaseClass._crash(self, err_msg)
self.mark_pdb_as_analyzed(pdb_file, "best_chains_and_models")
logging.info("%s", "=" * 40)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self, e)
def save_mat_files(self,Loops):
"""Pass the Loops structure array back to matlab so that it can
save the .mat files in the specified location."""
MotifAtlasBaseClass._setup_matlab(self)
[status, err_msg] = self.mlab.aSaveLoops(Loops,
self.config['locations']['loops_mat_files'],
nout=2)
if status == 0:
logging.info('mat files saved')
else:
MotifAtlasBaseClass._crash(self,err_msg)
def check_loop_quality(self, pdbs):
"""
"""
try:
logging.info('Loop Quality Assurance')
release = LoopRelease(mode=self.config['release_mode']['loops'])
for pdb_id in pdbs:
self.loop_qa(pdb_id, release.id)
if pdbs:
session.add(release)
session.commit()
logging.info('Loop QA complete')
logging.info('%s', '='*40)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self,e)
def extract_and_import_loops(self, pdbs, recalculate=None):
"""Loops over `pdbs`, extracts and imports all loops"""
try:
for loop_type in self.loop_types:
logging.info('Extracting %s' % loop_type)
if recalculate is None:
recalculate = self.config['recalculate'][loop_type]
if recalculate:
pdb_list = pdbs[:]
else:
pdb_list = self.filter_out_analyzed_pdbs(pdbs, loop_type)
for pdb_id in pdb_list:
logging.info('Extracting %s from %s', loop_type, pdb_id)
(Loops,l) = self.extract_loops(pdb_id, loop_type)
self.import_loops(Loops, l, pdb_id, loop_type)
logging.info('%s', '='*40)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self,e)
def load_loop_positions(self):
"""update loop_positions table by loading data from the mat files
stored in the PrecomputedData folder"""
if not self.mlab:
self._setup_matlab()
# loop over directories
for folder in os.listdir(self.precomputedData):
if re.search(self.pdb_regex, folder):
logging.info('Importing loop annotations from %s', folder)
else:
continue
[outputFile, err_msg] = self.mlab.loadLoopPositions(os.path.join(self.precomputedData, folder), nout=2)
if err_msg != '':
MotifAtlasBaseClass._crash(self, err_msg)
else:
reader = csv.reader(open(outputFile), delimiter=',', quotechar='"')
for row in reader:
(loop_id, position, nt_id, bulge, flanking, border) = row
existing = session.query(LoopPositions). \
filter(LoopPositions.loop_id==loop_id). \
filter(LoopPositions.position==position). \
filter(LoopPositions.border==border). \
first()
if existing:
if self.update:
existing.flanking = int(flanking)
existing.bulge = int(bulge)
existing.nt_id = nt_id
existing.border = int(border)
session.merge(existing)
else:
logging.info('Keeping existing annotations')
else:
session.add(LoopPositions(loop_id=loop_id,
position=position,
nt_id=nt_id,
flanking=int(flanking),
bulge=int(bulge),
border=int(border)))
session.commit()
os.remove(outputFile) # delete temporary csv file
def extract_loops(self, pdb_id, loop_type):
"""
"""
try:
MotifAtlasBaseClass._setup_matlab(self)
"""Loops - array of FR3D File structures. l - its length"""
[Loops, l, err_msg] = self.mlab.extractLoops(pdb_id, loop_type, nout=3)
if err_msg != '':
MotifAtlasBaseClass._crash(self,err_msg)
if Loops == 0:
logging.info('No %s in %s', loop_type, pdb_id)
return (0, 0)
else:
logging.info('Found %i loops', l)
return (Loops, l)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self,e)
def load_loop_searches(self):
"""
directory structure: loopSearchDir filesep IL_1S72_001 filesep IL_1S72_001_IL_1J5E_001.mat
"""
# loop over directories
for loop_id in os.listdir(self.loopSearchDir):
if re.search(self.loop_regex, loop_id):
logging.info('Importing %s searches', loop_id)
else:
continue
# read in No_candidates.txt if it exists
self._read_no_candidates_file(loop_id)
# get stored loop searches and list all matfiles
imported = self._get_imported_loop_searches(loop_id)
matfiles = self._get_saved_mat_files(os.path.join(self.loopSearchDir, loop_id, '*.mat'))
toImport = matfiles - imported;
if len(toImport) == 0:
continue
toImport = [os.path.join(self.loopSearchDir, loop_id, x + '.mat') for x in toImport]
if not self.mlab:
self._setup_matlab()
# run matlab to create a temporary csv file with results
[outputFile, err_msg] = self.mlab.loadLoopSearchFile(','.join(toImport), os.path.join(self.loopSearchDir, loop_id), nout=2)
if err_msg != '':
MotifAtlasBaseClass._crash(self, err_msg)
else:
reader = csv.reader(open(outputFile), delimiter=',', quotechar='"')
for row in reader:
(loop_id1, loop_id2, disc, nt_list1, nt_list2) = row
self._store_in_database(loop_id1, loop_id2, disc, nt_list1, nt_list2)
os.remove(outputFile) # delete temporary csv file
def import_interactions(self, pdbs, recalculate=False):
"""Determines what files need to be analyzed, deletes stored data if
necessary, loops over the pdbs, runs matlab on each of them
independently, matlab generates a temporary csv file, it's imported
and immediately deleted."""
try:
logging.info('Inside import_interactions')
if not recalculate:
recalculate = self.config['recalculate']['interactions']
if recalculate:
pdb_list = pdbs
self.__delete_interactions(pdbs)
else:
pdb_list = self.filter_out_analyzed_pdbs(pdbs,'interactions')
if pdb_list:
MotifAtlasBaseClass._setup_matlab(self)
for pdb_file in pdb_list:
logging.info('Running matlab on %s', pdb_file)
ifn, status, err_msg = self.mlab.loadInteractions(pdb_file,nout=3)
status = status[0][0]
if status == 0:
self.__import_interactions_from_csv(ifn, pdb_file)
elif status == 2: # no nucleotides in the pdb file
logging.info('Pdb file %s has no nucleotides', pdb_file)
else:
logging.warning('Matlab error code %i when analyzing %s',
status, pdb_file)
MotifAtlasBaseClass._crash(self,err_msg)
self.mark_pdb_as_analyzed(pdb_file,'interactions')
self.success = True
logging.info('%s', '='*40)
except:
e = sys.exc_info()[1]
MotifAtlasBaseClass._crash(self,e)
