def main():
'The script itself'
#set parameters
work_dir, output, reference = set_parameters()
# make a working tempfir
temp_dir = NamedTemporaryDir()
# add readgroup tag to each alignment in bam
add_header_and_tags_bams(work_dir, temp_dir.name)
# Prepare files to merge
sams = get_opened_sams_from_dir(temp_dir.name)
temp_sam = NamedTemporaryFile()
# merge all the sam in one
merge_sam(sams, temp_sam, reference)
# Convert sam into a bam,(Temporary)
temp_bam = NamedTemporaryFile(suffix='.bam')
sam2bam(temp_sam.name, temp_bam.name)
# finally we need to order the bam
sort_bam_sam(temp_bam.name, output)
# and make and index of the bam
call(['samtools', 'index', output], raise_on_error=True)
temp_dir.close()
def map_reads_with_bwa(reference_fpath, reads_fpath, bam_fpath, parameters):
'It maps the reads to the reference using bwa and returns a bam file'
colorspace = parameters['colorspace']
reads_length = parameters['reads_length']
threads = parameters['threads']
java_conf = parameters['java_conf']
tmp_dir = parameters['tmp_dir'] if 'tmp_dir' in parameters else None
threads = get_num_threads(threads)
#the reference should have an index
bwt_fpath = reference_fpath + '.bwt'
if not os.path.exists(bwt_fpath):
create_bwa_reference(reference_fpath, color=colorspace)
output_ali = 'output.ali'
bam_file_bam = 'bam_file.bam'
output_sai = 'output.sai'
if reads_length == 'short':
cmd = ['bwa', 'aln', reference_fpath, reads_fpath,
'-t', str(threads)]
if colorspace:
cmd.append('-c')
sai_fhand = NamedTemporaryFile(dir=tmp_dir, suffix=output_sai, mode='wb')
call(cmd, stdout=sai_fhand, raise_on_error=True)
cmd = ['bwa', 'samse', reference_fpath, sai_fhand.name, reads_fpath]
ali_fhand = NamedTemporaryFile(dir=tmp_dir, suffix=output_ali, mode='w')
call(cmd, stdout=ali_fhand, raise_on_error=True)
elif reads_length == 'long':
cmd = ['bwa', 'dbwtsw', reference_fpath, reads_fpath,
'-t', str(threads)]
ali_fhand = NamedTemporaryFile(dir=tmp_dir, suffix=output_ali)
call(cmd, stdout=ali_fhand, raise_on_error=True)
else:
raise ValueError('Reads length: short or long')
if 'unmapped_fhand' in parameters and parameters['unmapped_fhand'] is not None:
out_ali_fhand = NamedTemporaryFile(dir=tmp_dir, suffix=output_ali)
get_out_unmapped(ali_fhand, parameters['unmapped_fhand'], out_ali_fhand)
ali_fhand = out_ali_fhand
# From sam to Bam
# unsorted_bam = os.path.join(temp_dir.name, bam_file_bam)
unsorted_bam = NamedTemporaryFile(dir=tmp_dir, suffix=bam_file_bam)
sam2bam(ali_fhand.name, unsorted_bam.name)
# sort bam file
sort_bam_sam(unsorted_bam.name, bam_fpath, sort_method='coordinate',
java_conf=java_conf, strict_validation=False, tmp_dir=tmp_dir)
def test_sort_bam():
'It test that we can sort bams using picard'
#sort sam
sam_fpath = os.path.join(TEST_DATA_DIR, 'samtools', 'seqs.sam')
sorted_samfhand = NamedTemporaryFile(suffix='.sam')
sort_bam_sam(sam_fpath, sorted_samfhand.name)
#sort bam
bam_fpath = os.path.join(TEST_DATA_DIR, 'samtools', 'seqs.bam')
sorted_bamfhand = NamedTemporaryFile(suffix='.bam')
sort_bam_sam(bam_fpath, sorted_bamfhand.name)
#sort bam to sam
bam_fpath = os.path.join(TEST_DATA_DIR, 'samtools', 'seqs.bam')
sorted_samfhand = NamedTemporaryFile(suffix='.sam')
sort_bam_sam(bam_fpath, sorted_samfhand.name)
#sort sam
sam_fpath = os.path.join(TEST_DATA_DIR, 'samtools', 'seqs.sam')
sorted_bamfhand = NamedTemporaryFile(suffix='.bam')
sort_bam_sam(sam_fpath, sorted_bamfhand.name)
def run(self):
'''It runs the analysis.'''
self._log({'analysis_started':True})
settings = self._project_settings
project_path = settings['General_settings']['project_path']
tmp_dir = settings['General_settings']['tmpdir']
inputs = self._get_input_fpaths()
bam_paths = inputs['bams']
reference_path = inputs['reference']
output_dir = self._create_output_dirs()['result']
merged_bam_path = VersionedPath(os.path.join(output_dir,
BACKBONE_BASENAMES['merged_bam']))
merged_bam_fpath = merged_bam_path.next_version
#Do we have to add the default qualities to the sam file?
#do we have characters different from ACTGN?
add_qualities = settings['Sam_processing']['add_default_qualities']
#memory for the java programs
java_mem = settings['Other_settings']['java_memory']
picard_path = settings['Other_settings']['picard_path']
if add_qualities:
default_sanger_quality = settings['Other_settings']['default_sanger_quality']
default_sanger_quality = int(default_sanger_quality)
else:
default_sanger_quality = None
temp_dir = NamedTemporaryDir()
for bam_path in bam_paths:
bam_basename = bam_path.basename
temp_sam = NamedTemporaryFile(prefix='%s.' % bam_basename,
suffix='.sam')
sam_fpath = os.path.join(temp_dir.name, bam_basename + '.sam')
bam2sam(bam_path.last_version, temp_sam.name)
sam_fhand = open(sam_fpath, 'w')
# First we need to create the sam with added tags and headers
add_header_and_tags_to_sam(temp_sam, sam_fhand)
temp_sam.close()
sam_fhand.close()
#the standardization
temp_sam2 = NamedTemporaryFile(prefix='%s.' % bam_basename,
suffix='.sam', delete=False)
standardize_sam(open(sam_fhand.name), temp_sam2,
default_sanger_quality,
add_def_qual=add_qualities,
only_std_char=True)
temp_sam2.flush()
shutil.move(temp_sam2.name, sam_fhand.name)
temp_sam2.close()
get_sam_fpaths = lambda dir_: [os.path.join(dir_, fname) for fname in os.listdir(dir_) if fname.endswith('.sam')]
# Once the headers are ready we are going to merge
sams = get_sam_fpaths(temp_dir.name)
sams = [open(sam) for sam in sams]
temp_sam = NamedTemporaryFile(suffix='.sam')
reference_fhand = open(reference_path.last_version)
try:
merge_sam(sams, temp_sam, reference_fhand)
except Exception:
if os.path.exists(merged_bam_fpath):
os.remove(merged_bam_fpath)
raise
reference_fhand.close()
# close files
for sam in sams:
sam.close()
# Convert sam into a bam,(Temporary)
temp_bam = NamedTemporaryFile(suffix='.bam')
sam2bam(temp_sam.name, temp_bam.name)
# finally we need to order the bam
#print 'unsorted.bam', temp_bam.name
#raw_input()
sort_bam_sam(temp_bam.name, merged_bam_fpath,
java_conf={'java_memory':java_mem,
'picard_path':picard_path}, tmp_dir=tmp_dir )
temp_bam.close()
temp_sam.close()
create_bam_index(merged_bam_fpath)
self._log({'analysis_finished':True})
