def main(cmdline=None):
parser = ArgumentParser()
parser.add_argument('-l',
'--library',
required=True,
action='append',
help="library table to load")
parser.add_argument('-o', '--output', help='filename to write report to')
args = parser.parse_args(cmdline)
libraries = load_library_tables(args.library)
metrics = []
for library_id, library in libraries.iterrows():
genome_triple = genome_name_from_library(library)
filename = library.analysis_name + '-' + genome_triple + '_picard_markdup.metrics'
pathname = Path(library.analysis_dir) / filename
if pathname.exists():
picard_metric = parse_picard_metric(pathname,
library_id=library_id)
metrics.append(picard_metric)
else:
print('{} is missing. Skipping'.format(pathname))
metrics = pandas.DataFrame(metrics)
metrics.set_index('LIBRARY', inplace=True)
if args.output:
metrics.to_csv(args.output, sep='\t')
else:
print(metrics)
def test_genome_name_from_library_dict(self):
d = {
'genome': 'mm10',
'annotation': 'M21_minimal',
'sex': 'male',
}
self.assertEqual(models.genome_name_from_library(d), 'mm10-M21_minimal-male')
self.assertRaises(ValueError, models.genome_name_from_library, 10)
def build_hash_tree(library_filename):
table = load_library_tables([library_filename])
hashes = {}
for library_id, row in table.iterrows():
analysis_dir = row.analysis_dir
name = row.analysis_name + '-' + genome_name_from_library(
row) + '_genome.bam'
alignment = os.path.join(analysis_dir, name)
hashes[library_id] = hash_alignments(alignment)
return hashes
def link_genome_bams(libraries, output_dir):
for library_id, library in libraries.iterrows():
clean_library_id = sanitize_library_suffix(library_id)
target_dir = os.path.join(output_dir, clean_library_id)
if not os.path.exists(target_dir):
os.mkdir(target_dir)
name = make_bam_track_name(library, library.analysis_dir)
source_pathname = os.path.join(library.analysis_dir, name)
target_name = clean_library_id + genome_name_from_library(
library) + '_genome.bam'
cur_dir = os.getcwd()
os.chdir(target_dir)
if os.path.exists(source_pathname) and not os.path.exists(target_name):
print(source_pathname, '->', target_name)
os.symlink(source_pathname, target_name)
os.chdir(cur_dir)
def make_bigwig_track_name(library, signal_type, analysis_root):
"""Generate the base path where the bigwig track is
:param Series library: row from a library table DataFrame
:param str signal_type: either uniq or all to specify bigwig type.
:param str analysis_root: root directory to be searching for track files
:returns: list of paths of bigWig files relative to analysis_root
"""
assert signal_type in ('uniq', 'all')
genome_triplet = genome_name_from_library(library)
track_name = library.analysis_name + '-' + genome_triplet + '_' + signal_type + '.bw'
for pathname in [
os.path.join(library.analysis_dir, track_name),
os.path.join(analysis_root, track_name)
]:
if os.path.exists(pathname):
return return_subpath(pathname, analysis_root)
logger.warning("Couldn't find track file %s", track_name)
def link_rsem(libraries, output_dir):
for library_id, library in libraries.iterrows():
clean_library_id = sanitize_library_suffix(library_id)
target_dir = os.path.join(output_dir, clean_library_id)
if not os.path.exists(target_dir):
os.mkdir(target_dir)
source_dir = library.analysis_dir
cur_dir = os.getcwd()
os.chdir(target_dir)
for extension in [
'_anno_rsem.genes.results', '_anno_rsem.isoforms.results'
]:
suffix = '-' + genome_name_from_library(library) + extension
source_name = library_id + suffix
target_name = clean_library_id + suffix
source_pathname = os.path.join(source_dir, source_name)
if os.path.exists(
source_pathname) and not os.path.exists(target_name):
print(source_pathname, '->', target_name)
os.symlink(source_pathname, target_name)
os.chdir(cur_dir)
def make_bam_track_name(library, analysis_root=None):
"""Generate the base path where the bam track is.
:param Series library: row from a library table DataFrame
:param str analysis_root: root directory to be searching for track files
:returns: path of bam file relative to analysis_root
"""
genome_triplet = genome_name_from_library(library)
track_name = library.analysis_name + '-' + genome_triplet + '_genome.bam'
old_name = 'Aligned.sortedByCoord.out.bam'
to_check = [
os.path.join(library.analysis_dir, track_name),
os.path.join(analysis_root, track_name),
os.path.join(library.analysis_dir, old_name),
]
for pathname in to_check:
if os.path.exists(pathname):
bai = pathname + '.bai'
if not os.path.exists(bai):
logger.warning('Missing index file for {}'.format(pathname))
return return_subpath(pathname, analysis_root)
logger.warning("Couldn't find track file %s", track_name)
def test_genome_name_from_library_series(self):
mm10tsv = resource_filename(__name__, 'library-mm10-se.tsv')
mm10 = models.load_library_tables([mm10tsv])
self.assertEqual(models.genome_name_from_library(mm10.loc['12304']), 'mm10-M4-female')
self.assertEqual(models.genome_name_from_library(mm10.loc['12309']), 'mm10-M4-male')