def store_atomic_data_for_contigs_and_splits(self, sample_id, contigs, db):
self.progress.new('Storing atomic_data')
num_contigs = pp(len(contigs))
cur_contig = 1
# this loop will get atomic_data information from Contig instanes and store them into the db
# at once. this was broken down into about 10 functions, but this structure seems to be the most efficient
# although it looks crappy:
for contig_name in contigs:
self.progress.update("Processing contig %s of %s" % (pp(cur_contig), num_contigs))
contig = contigs[contig_name]
contig_atomic_data = contig.get_atomic_data_dict()
self.atomic_data_contigs[contig.name] = {'contig': contig.name}
for atomic_data_field in t.atomic_data_table_structure[1:]:
self.atomic_data_contigs[contig.name][atomic_data_field] = contig_atomic_data[atomic_data_field]
# contig is done, deal with splits in it:
for split in contig.splits:
split_atomic_data = split.get_atomic_data_dict()
self.atomic_data_splits[split.name] = {'contig': split.name}
for atomic_data_field in t.atomic_data_table_structure[1:]:
self.atomic_data_splits[split.name][atomic_data_field] = split_atomic_data[atomic_data_field]
self.progress.update("Generating tables ...")
gen_atomic_data_tables_for_contigs_and_splits(self.atomic_data_splits, self.atomic_data_contigs, db)
self.progress.end()
def clusterize(self, parts):
# create a 8 digits random identifier for cluster jobs:
identifier = ''.join(
random.choice(string.ascii_uppercase) for x in range(10))
for part in parts:
command = self.command % {'binary': self.binary, 'part': part}
# create sh file
shell_script = part + '.sh'
open(shell_script, 'w').write(QSUB_SCRIPT % {
'log': part + '.log',
'identifier': identifier,
'command': command
})
# submit script to cluster
utils.run_command('qsub %s' % shell_script)
while True:
qstat_info = self.get_qstat_info(identifier)
total_processes = sum(qstat_info.values())
if total_processes == 0:
break
self.progress.update(
'Qstat Info :: Total Jobs: %s, %s' %
(pp(total_processes), ', '.join(
['%s: %s' % (x, pp(qstat_info[x])) for x in qstat_info])))
time.sleep(5)
return True
def clusterize(self, parts):
# create a 8 digits random identifier for cluster jobs:
identifier = ''.join(random.choice(string.ascii_uppercase) for x in range(10))
for part in parts:
command = self.command % {'binary': self.binary, 'part': part}
# create sh file
shell_script = part + '.sh'
open(shell_script, 'w').write(QSUB_SCRIPT % {'log': part + '.log',
'identifier': identifier,
'command': command})
# submit script to cluster
utils.run_command('qsub %s' % shell_script)
while True:
qstat_info = self.get_qstat_info(identifier)
total_processes = sum(qstat_info.values())
if total_processes == 0:
break
self.progress.update('Qstat Info :: Total Jobs: %s, %s' % (pp(total_processes),
', '.join(['%s: %s' % (x, pp(qstat_info[x])) for x in qstat_info])))
time.sleep(5)
return True
def store_atomic_data_for_contigs_and_splits(self, sample_id, contigs, db):
self.progress.new('Storing atomic_data')
num_contigs = pp(len(contigs))
cur_contig = 1
# this loop will get atomic_data information from Contig instanes and store them into the db
# at once. this was broken down into about 10 functions, but this structure seems to be the most efficient
# although it looks crappy:
for contig_name in contigs:
self.progress.update("Processing contig %s of %s" %
(pp(cur_contig), num_contigs))
contig = contigs[contig_name]
contig_atomic_data = contig.get_atomic_data_dict()
self.atomic_data_contigs[contig.name] = {'contig': contig.name}
for atomic_data_field in t.atomic_data_table_structure[1:]:
self.atomic_data_contigs[contig.name][
atomic_data_field] = contig_atomic_data[atomic_data_field]
# contig is done, deal with splits in it:
for split in contig.splits:
split_atomic_data = split.get_atomic_data_dict()
self.atomic_data_splits[split.name] = {'contig': split.name}
for atomic_data_field in t.atomic_data_table_structure[1:]:
self.atomic_data_splits[
split.name][atomic_data_field] = split_atomic_data[
atomic_data_field]
self.progress.update("Generating tables ...")
gen_atomic_data_tables_for_contigs_and_splits(self.atomic_data_splits,
self.atomic_data_contigs,
db)
self.progress.end()
def split_input_file(self):
parts = []
next_part = 1
part_obj = None
if self.input_is_fasta:
fasta = u.SequenceSource(self.input_file_path)
while fasta.next():
if (fasta.pos - 1) % self.num_entries_per_file == 0:
self.progress.update('Creating part: ~ %s' %
(pp(next_part)))
if part_obj:
part_obj.close()
file_path = os.path.join(self.tmp_dir,
'part-%08d.fa' % next_part)
parts.append(file_path)
next_part += 1
part_obj = open(file_path, 'w')
part_obj.write('>%s\n' % fasta.id)
part_obj.write('%s\n' % fasta.seq)
if part_obj:
part_obj.close()
return parts
def split_input_file(self):
parts = []
next_part = 1
part_obj = None
if self.input_is_fasta:
fasta = u.SequenceSource(self.input_file_path)
while next(fasta):
if (fasta.pos - 1) % self.num_entries_per_file == 0:
self.progress.update('Creating part: ~ %s' % (pp(next_part)))
if part_obj:
part_obj.close()
file_path = os.path.join(self.tmp_dir, 'part-%08d.fa' % next_part)
parts.append(file_path)
next_part += 1
part_obj = open(file_path, 'w')
part_obj.write('>%s\n' % fasta.id)
part_obj.write('%s\n' % fasta.seq)
if part_obj:
part_obj.close()
return parts
