def run(self):
''' Run pacbio_smrtanalysis RS_sequencing'''
original_dir = os.getcwd()
os.chdir(self.working_directory)
no_bsub_option = "--no_bsub" if self.no_bsub else ""
# pacbio_smrtanalysis --memory 6 --reference /path/to/reference.fa RS_Resequencing Outputdir *.bax.h5
command = " ".join([
self.pacbio_exec, "--memory 6", no_bsub_option, "--reference",
self.input_file, "RS_Resequencing", self.output_directory,
self.read_data + "/*.bax.h5"
])
if (os.path.getsize(self.input_file)):
fastaqutils.syscall(command)
# subprocess.call(command, shell=True)
self._produce_summary(command)
else:
self._produce_summary("File empty: " + self.input_file)
# Wait for bsub to finish. Look into using process.wait() here
# while not os.path.exists(self._build_default_filename()):
# time.sleep(30)
#
# # Move results file and produce summary
# shutil.move(self._build_default_filename(), self._build_final_filename())
# We have abandoned this effort of trying to wait for the bsub quiver job to finish
# When run in the pipeline, we can carry out this step as a separate task
# When run on the command line the user will just have to live with the quiver output (consensus.fasta)
# Clean up quiver directory
# if not self.debug and os.path.exists(os.path.join(self.working_directory, self.output_directory)):
# shutil.rmtree(os.path.join(self.working_directory, self.output_directory))
os.chdir(original_dir)
def test_system_call(self):
'''Test that system call appears to work and die as it should'''
test_file = os.path.join(data_dir, 'utils_test_system_call.txt')
tmp_out = 'utils_test_syscall.tmp'
utils.syscall('cat ' + test_file + ' > ' + tmp_out)
self.assertTrue(filecmp.cmp(tmp_out, test_file))
os.unlink(tmp_out)
with self.assertRaises(utils.Error):
utils.syscall('thisisveryunlikelytoebarealcommandandshouldthrowerror')
utils.syscall('echo "this is not the right string" > ' + tmp_out)
self.assertFalse(filecmp.cmp(tmp_out, test_file))
os.unlink(tmp_out)
s = utils.syscall_get_stdout('echo bingo')
self.assertListEqual(["bingo"], s)
with self.assertRaises(utils.Error):
utils.syscall_get_stdout('thisisveryunlikelytoebarealcommandandshouldthrowerror')
def test_system_call(self):
'''Test that system call appears to work and die as it should'''
test_file = os.path.join(data_dir, 'utils_test_system_call.txt')
tmp_out = 'utils_test_syscall.tmp'
utils.syscall('cat ' + test_file + ' > ' + tmp_out)
self.assertTrue(filecmp.cmp(tmp_out, test_file))
os.unlink(tmp_out)
with self.assertRaises(utils.Error):
utils.syscall(
'thisisveryunlikelytoebarealcommandandshouldthrowerror')
utils.syscall('echo "this is not the right string" > ' + tmp_out)
self.assertFalse(filecmp.cmp(tmp_out, test_file))
os.unlink(tmp_out)
s = utils.syscall_get_stdout('echo bingo')
self.assertListEqual(["bingo"], s)
with self.assertRaises(utils.Error):
utils.syscall_get_stdout(
'thisisveryunlikelytoebarealcommandandshouldthrowerror')
help='Name of reference fasta file',
metavar='reference.fa')
parser.add_argument('outprefix', help='Prefix of output files')
options = parser.parse_args()
ref_seqs = {}
tasks.file_to_dict(options.ref_fa, ref_seqs)
nucmer_out_prefix = options.outprefix + '.nucmer'
nucmer_out_delta = nucmer_out_prefix + '.delta'
nucmer_out_filter = nucmer_out_prefix + '.delta-filter'
nucmer_out_coords = nucmer_out_filter + '.coords'
# run nucmer of contigs vs ref
utils.syscall(' '.join([
'nucmer', options.nucmer_options, '-p', nucmer_out_prefix, options.ref_fa,
options.contigs_fa
]))
utils.syscall(' '.join([
'delta-filter', '-i 98 -l 180 -q', nucmer_out_delta, '>', nucmer_out_filter
]))
utils.syscall(' '.join(
['show-coords', '-dTlro', nucmer_out_filter, '>', nucmer_out_coords]))
# load hits into hash. key=ref_name, value=another hash with key=qry_name, value=list of hit positions in that ref seq
nucmer_hits = {}
contigs_to_print = {}
nucmer_reader = nucmer_file_reader(nucmer_out_coords)
for hit in nucmer_reader:
if hit.ref_name not in nucmer_hits:
if (not sam1.is_reverse) and (not sam2.is_reverse):
nodes[sam2.qname].add(sam1.qname)
elif sam1.is_reverse and sam2.is_reverse:
nodes[sam1.qname].add(sam2.qname)
else:
nodes[sam1.qname].add(sam2.qname)
nodes[sam2.qname].add(sam1.qname)
# make pdf of the graph using graphviz
cmd = 'echo "digraph G {'
first = True
for node, l in sorted(nodes.items()):
if first:
first = False
else:
cmd += ';'
if len(l):
cmd += ';'.join([node + '->' + x for x in l])
else:
cmd += node
cmd += '}" | dot -Tpdf > ' + options.outprefix + '.pdf'
make_graph = options.outprefix + '.make_graph.sh'
f = utils.open_file_write(make_graph)
print(cmd, file=f)
utils.close(f)
utils.syscall('bash ' + make_graph)
parser.add_argument("--nucmer_options", help="Options when running nucmer [%(default)s]", default="")
parser.add_argument("contigs_fa", help="Name of contigs fasta file", metavar="contigs.fa")
parser.add_argument("ref_fa", help="Name of reference fasta file", metavar="reference.fa")
parser.add_argument("outprefix", help="Prefix of output files")
options = parser.parse_args()
ref_seqs = {}
tasks.file_to_dict(options.ref_fa, ref_seqs)
nucmer_out_prefix = options.outprefix + ".nucmer"
nucmer_out_delta = nucmer_out_prefix + ".delta"
nucmer_out_filter = nucmer_out_prefix + ".delta-filter"
nucmer_out_coords = nucmer_out_filter + ".coords"
# run nucmer of contigs vs ref
utils.syscall(" ".join(["nucmer", options.nucmer_options, "-p", nucmer_out_prefix, options.ref_fa, options.contigs_fa]))
utils.syscall(" ".join(["delta-filter", "-i 98 -l 180 -q", nucmer_out_delta, ">", nucmer_out_filter]))
utils.syscall(" ".join(["show-coords", "-dTlro", nucmer_out_filter, ">", nucmer_out_coords]))
# load hits into hash. key=ref_name, value=another hash with key=qry_name, value=list of hit positions in that ref seq
nucmer_hits = {}
contigs_to_print = {}
nucmer_reader = nucmer_file_reader(nucmer_out_coords)
for hit in nucmer_reader:
if hit.ref_name not in nucmer_hits:
nucmer_hits[hit.ref_name] = {}
if hit.qry_name not in nucmer_hits[hit.ref_name]:
nucmer_hits[hit.ref_name][hit.qry_name] = []
if (not sam1.is_reverse) and (not sam2.is_reverse):
nodes[sam2.qname].add(sam1.qname)
elif sam1.is_reverse and sam2.is_reverse:
nodes[sam1.qname].add(sam2.qname)
else:
nodes[sam1.qname].add(sam2.qname)
nodes[sam2.qname].add(sam1.qname)
# make pdf of the graph using graphviz
cmd = 'echo "digraph G {'
first = True
for node, l in sorted(nodes.items()):
if first:
first = False
else:
cmd += ";"
if len(l):
cmd += ";".join([node + "->" + x for x in l])
else:
cmd += node
cmd += '}" | dot -Tpdf > ' + options.outprefix + ".pdf"
make_graph = options.outprefix + ".make_graph.sh"
f = utils.open_file_write(make_graph)
print(cmd, file=f)
utils.close(f)
utils.syscall("bash " + make_graph)
