def create_analysis_dir(project,
top_dir=None,
merge_replicates=False,
keep_names=False,
dry_run=False):
"""Create and populate analysis directory for an IlluminaProject
Creates a new directory and populates either with links to FASTQ
files, or with 'merged' FASTQ files created by concatenating
multiple FASTQs for each sample (which can happen for multiplexed
runs where samples are split across multiple lanes).
Project directory names are made up of the project name and then
the experiment type, or just the project name if experiment type
is not set.
Arguments:
project : populated IlluminaProject object
top_dir : parent directory to create analysis subdirectory
under. Defaults to cwd if not explicitly specified
merge_replicates: if True then creates a single FASTQ file for
each sample by merging multiple FASTQs together
keep_names: if True then links to FASTQ files will have the same
names as the original files; by default links use the
shortest unique name
dry_run : if True then report what would be done but don't
actually perform any action
Returns:
Name of the project directory.
"""
project_dir = os.path.join(top_dir,project.full_name)
print "Creating analysis directory for project '%s'..." % project.full_name
# Check for & create directory
if os.path.exists(project_dir):
print "-> %s already exists" % project_dir
else:
print "Making analysis directory for %s" % project.name
if not dry_run:
bcf_utils.mkdir(project_dir,mode=0775)
# Make an empty ScriptCode directory
scriptcode_dir = os.path.join(project_dir,"ScriptCode")
if os.path.exists(scriptcode_dir):
print "'ScriptCode' directory %s already exists" % scriptcode_dir
else:
print "Making 'ScriptCode' directory for %s" % project.name
if not dry_run:
bcf_utils.mkdir(scriptcode_dir,mode=0775)
# Check for & create links to fastq files
if not merge_replicates:
for sample in project.samples:
fastq_names = IlluminaData.get_unique_fastq_names(sample.fastq)
for fastq in sample.fastq:
fastq_file = os.path.join(sample.dirn,fastq)
if keep_names:
fastq_ln = os.path.join(project_dir,fastq)
else:
fastq_ln = os.path.join(project_dir,fastq_names[fastq])
if os.path.exists(fastq_ln):
logging.error("Failed to link to %s: %s already exists" %
(fastq_file,os.path.basename(fastq_ln)))
else:
print "Linking to %s" % fastq
if not dry_run:
bcf_utils.mklink(fastq_file,fastq_ln,relative=True)
else:
# Merge files for replicates within each sample
for sample in project.samples:
replicates = {}
# Gather replicates to be merged
for fastq in sample.fastq:
fastq_data = IlluminaData.IlluminaFastq(fastq)
name = "%s_%s_R%d" % (fastq_data.sample_name,
fastq_data.barcode_sequence,
fastq_data.read_number)
if name not in replicates:
replicates[name] = []
replicates[name].append(os.path.join(sample.dirn,fastq))
# Sort into order
replicates[name].sort()
# Report detected replicates
print "Sample %s" % sample.name
for name in replicates:
print "\tReplicate '%s'" % name
for fastq in replicates[name]:
print "\t\t%s" % fastq
# Do the merge
for name in replicates:
merged_fastq = os.path.join(project_dir,name+'.fastq')
bcf_utils.concatenate_fastq_files(merged_fastq,replicates[name])
# Return directory name
return project_dir
p.add_argument('-v',
'--verbose',
action="store_true",
dest="verbose",
default=False,
help="verbose output")
p.add_argument('fastqs',
metavar="FASTQ",
nargs='+',
help="Input FASTQ to concatenate")
p.add_argument('fastq_out',
metavar="FASTQ_OUT",
help="Output FASTQ with concatenated reads")
args = p.parse_args()
# Sort out inputs
if len(args.fastqs) < 2:
p.error("Need to supply at least 2 input fastqs plus output name")
# Check inputs exist
for fq in args.fastqs:
if not os.path.exists(fq):
logging.critical("Input file '%s' not found" % fq)
sys.exit(1)
# Run the concatenation
try:
concatenate_fastq_files(args.fastq_out,
args.fastqs,
verbose=args.verbose)
except Exception as ex:
logging.critical("Failed with exception: %s" % ex)
sys.exit(1)
if __name__ == "__main__":
# Handle command line
p = optparse.OptionParser(
usage="%prog [OPTIONS] FASTQ [FASTQ...] FASTQ_OUT",
description="Concatenate reads from one or more input Fastq files "
"into a single new file FASTQ_OUT.",
version=__version__)
p.add_option('-v','--verbose',
action="store_true",dest="verbose",
default=False,
help="verbose output")
opts,args = p.parse_args()
# Sort out inputs
if len(args) < 2:
p.error("Need to supply at least 2 input fastqs plus output name")
fastq_out = args[-1]
fastqs = args[:-1]
# Check inputs exist
for fq in fastqs:
if not os.path.exists(fq):
logging.critical("Input file '%s' not found" % fq)
sys.exit(1)
# Run the concatenation
try:
concatenate_fastq_files(fastq_out,fastqs,
verbose=opts.verbose)
except Exception as ex:
logging.critical("Failed with exception: %s" % ex)
sys.exit(1)
shutil.copy(fastq_file, dst)
# Verify against sample sheet
if options.sample_sheet is not None:
if IlluminaData.verify_run_against_sample_sheet(
illumina_data, options.sample_sheet):
print "Verification against sample sheet '%s': OK" % \
options.sample_sheet
status = 0
else:
logging.error("Verification against sample sheet '%s': FAILED" %
options.sample_sheet)
status = 1
sys.exit(status)
# Merge multiple fastqs in each sample
if options.merge_fastqs:
for project in illumina_data.projects:
for sample in project.samples:
for read in (1, 2):
# Concatenate fastqs for this read
fastq_merged = sample.name
if sample.paired_end:
fastq_merged += "_R%d" % read
fastq_merged += ".fastq.gz"
bcf_utils.concatenate_fastq_files(fastq_merged,
sample.fastq_subset(
read_number=read,
full_path=True),
bufsize=1024 * 1024)
# Verify against sample sheet
if options.sample_sheet is not None:
if IlluminaData.verify_run_against_sample_sheet(illumina_data,options.sample_sheet):
print "Verification against sample sheet '%s': OK" % \
options.sample_sheet
status = 0
else:
logging.error("Verification against sample sheet '%s': FAILED" %
options.sample_sheet)
status = 1
sys.exit(status)
# Merge multiple fastqs in each sample
if options.merge_fastqs:
for project in illumina_data.projects:
for sample in project.samples:
for read in (1,2):
# Concatenate fastqs for this read
fastq_merged = sample.name
if sample.paired_end:
fastq_merged += "_R%d" % read
fastq_merged += ".fastq.gz"
bcf_utils.concatenate_fastq_files(fastq_merged,
sample.fastq_subset(read_number=read,
full_path=True),
bufsize=1024*1024)
