def test_make_fastq_links(self):
"""Test making fastq links"""
# Assume Illumina/SciLife data structure
tl = target_generator(indir=self.project)
fql = make_fastq_links(tl, indir=self.project, outdir="tmp")
self.assertTrue(os.path.lexists(os.path.join("tmp", os.path.relpath(tl[0][2], self.project) + "_R1_001.fastq.gz")))
self.assertTrue(os.path.lexists(os.path.join("tmp", os.path.dirname(os.path.relpath(tl[0][2], self.project)),
"SampleSheet.csv")))
def _setup(self):
# List requirements for completion, consisting of classes above
if self.indir is None:
logger.error("Need input directory to run")
self.targets = []
if self.outdir is None:
self.outdir = self.indir
self.targets = [tgt for tgt in self.target_iterator()]
if self.outdir != self.indir and self.targets:
self.targets = make_fastq_links(self.targets, self.indir, self.outdir)
# Finally register targets in backend
backend.__global_vars__["targets"] = self.targets
def _setup(self):
# List requirements for completion, consisting of classes above
if self.indir is None:
logger.error("Need input directory to run")
self.targets = []
if self.outdir is None:
self.outdir = self.indir
self.targets = [tgt for tgt in self.target_iterator()]
if self.outdir != self.indir:
self.targets = make_fastq_links(self.targets, self.indir, self.outdir)
# Finally register targets in backend
backend.__global_vars__["targets"] = self.targets
def requires(self):
if not self.indir:
return
if self.outdir is None:
self.outdir = self.indir
tgt_fun = self.set_target_generator_function()
if not tgt_fun:
return []
targets = tgt_fun(self.indir, sample=self.sample, flowcell=self.flowcell, lane=self.lane)
if self.outdir != self.indir:
targets = make_fastq_links(targets, self.indir, self.outdir)
picard_metrics_targets = ["{}.{}".format(x[1], "sort.merge.dup") for x in targets]
return [PicardMetrics(target=tgt) for tgt in picard_metrics_targets]
tgt_gen_fun = hdl
# Collect information about what samples to run, and on how many
# nodes. This is somewhat convoluted since ratatosk_run_scilife
# also collects sample information, but this step is necessary as
# we need to wrap ratatosk_run_scilife.py in drmaa
targets = tgt_gen_fun(indir=pargs.indir, sample=pargs.sample,
flowcell=pargs.flowcell, lane=pargs.lane)
# After getting run list, if output directory is different to
# input directory, link raw data files to output directory and
# remember to use this directory for ratatosk tasks. In this way
# we actually can run on subsets of sample runs or flowcells
if not pargs.outdir:
pargs.outdir = pargs.indir
if pargs.outdir != pargs.indir:
targets = make_fastq_links(targets, pargs.indir, pargs.outdir)
# Group samples
sorted_samples = sorted(targets, key=lambda t:t.sample_id())
samples = {}
for k, g in itertools.groupby(sorted_samples, key=lambda t:t.sample_id()):
samples[k] = list(g)
# Initialize command
if pargs.sample_target_suffix or pargs.run_target_suffix:
cmd = [RATATOSK_RUN, "GenericWrapper", '--workers', pargs.workers,
'--scheduler-host', pargs.scheduler_host, '--task', pargs.task]
else:
cmd = [RATATOSK_RUN, pargs.task, '--indir', pargs.indir, '--outdir', pargs.outdir,
'--workers', pargs.workers, '--scheduler-host', pargs.scheduler_host]
if pargs.config_file:
logging.info("setting config to {}".format(pargs.config_file))
