def __call__(self, cpus=None):
# Check it is installed #
self.check_installed()
# Check version #
assert "v0.11.9" in sh.fastqc('-version')
# Number of cores #
if cpus is None: cpus = min(multiprocessing.cpu_count(), 32)
# Make a temporary directory #
self.tmp_dir = new_temp_dir()
# Run it #
sh.fastqc(self.source, '-o', self.tmp_dir, '-t', cpus, '--extract')
# Get location of results #
components = self.source.prefix.split('.')
if components[-1] == 'gz': components.pop()
if components[-1] == 'fastq': components.pop()
# Reassemble the components #
created_name = '.'.join(components) + '_fastqc/'
# This will be the name of the directory that fastqc created #
created_dir = self.tmp_dir + created_name
# Move results #
if self.dest.exists: shutil.rmtree(self.dest)
shutil.move(created_dir, self.dest)
self.tmp_dir.remove()
# Return #
return self.results
def raw(self):
# Link #
if os.path.lexists(self.p.raw_fastq): os.remove(self.p.raw_fastq)
os.symlink(self.path, self.p.raw_fastq)
# Fast QC #
sh.fastqc(self.p.raw_fastq, '-o', self.p.raw_dir, '-q')
if os.path.exists(self.p.raw_dir + 'raw_fastqc.zip'): os.remove(self.p.raw_dir + 'raw_fastqc.zip')
def raw(self):
# Call extraction #
shell_output('sffinfo -s %s > %s' % (self.p.raw_sff, self.p.raw_fasta))
shell_output('sffinfo -q %s > %s' % (self.p.raw_sff, self.p.raw_qual))
shell_output('sffinfo -m %s > %s' % (self.p.raw_sff, self.p.manifest))
# Convert #
sh.fasta_to_fastq(self.p.raw_fasta, self.p.raw_qual, self.p.raw_fastq)
# Fast QC #
sh.fastqc(self.p.raw_fastq, '-o', self.p.raw_dir, '-q')
if os.path.exists(self.p.raw_dir + 'raw_fastqc.zip'): os.remove(self.p.raw_dir + 'raw_fastqc.zip')
def run(self):
# Sliding window procedure #
clean(infile=self.parent.p.raw_fastq, outfile=self.parent.p.clean_fastq,
keeptrim=self.parent.p.trim_fastq, format='fastq',
minlength=150, threshold=21, windowsize=20)
# Back to FASTA #
sh.fastq_to_fasta(self.parent.p.clean_fastq, self.parent.p.clean_fasta, self.parent.p.clean_qual)
# Fast QC #
sh.fastqc(self.parent.p.clean_fastq, '-o', self.parent.p.clean_dir, '-q')
os.remove(self.parent.p.clean_dir + 'clean_fastqc.zip')
def fastqc(self):
# Call #
sh.fastqc(self.path, '-q')
# Paths #
zip_file = self.prefix_path + '_fastqc.zip'
report_dir = self.prefix_path + '_fastqc/'
#images_dir = self.prefix_path + '_fastqc/Images/'
# Clean up #
os.remove(zip_file)
# Return #
return report_dir
def run(self):
if self.dest is None:
sh.fastqc(self.source, '-q')
os.remove(self.source.prefix_path + '_fastqc.zip')
if self.dest is not None:
if self.dest.exists: self.dest.remove()
self.tmp_dir = new_temp_dir()
sh.fastqc(self.source, '-q', '-o', self.tmp_dir)
created_dir = self.tmp_dir + self.source.prefix.split('.')[0] + '_fastqc/'
shutil.move(created_dir, self.dest)
self.tmp_dir.remove()
return self.results
def fastqc(self, directory):
# Symbolic link #
tmp_dir = tempfile.mkdtemp() + '/'
if self.gziped: sym_fwd_path = tmp_dir + 'fwd.fastq.gz'
else: sym_fwd_path = tmp_dir + 'fwd.fastq'
if self.gziped: sym_rev_path = tmp_dir + 'rev.fastq.gz'
else: sym_rev_path = tmp_dir + 'rev.fastq'
os.symlink(self.fwd_path, sym_fwd_path)
os.symlink(self.rev_path, sym_rev_path)
# Call #
sh.fastqc(sym_fwd_path, '-q')
sh.fastqc(sym_rev_path, '-q')
# Move #
shutil.move(tmp_dir + 'fwd_fastqc/', directory + "fwd_fastqc/")
shutil.move(tmp_dir + 'rev_fastqc/', directory + "rev_fastqc/")
# Clean up #
shutil.rmtree(tmp_dir)
def fastqc(self, directory):
# Symbolic link #
tmp_dir = tempfile.mkdtemp() + '/'
if self.gziped: sym_fwd_path = tmp_dir + 'fwd.fastq.gz'
else: sym_fwd_path = tmp_dir + 'fwd.fastq'
if self.gziped: sym_rev_path = tmp_dir + 'rev.fastq.gz'
else: sym_rev_path = tmp_dir + 'rev.fastq'
os.symlink(self.fwd_path, sym_fwd_path)
os.symlink(self.rev_path, sym_rev_path)
# Call #
sh.fastqc(sym_fwd_path, '-q')
sh.fastqc(sym_rev_path, '-q')
# Move #
shutil.rmtree(directory + "fwd_fastqc/", ignore_errors=True)
shutil.rmtree(directory + "rev_fastqc/", ignore_errors=True)
shutil.move(tmp_dir + 'fwd_fastqc/', directory + "fwd_fastqc/")
shutil.move(tmp_dir + 'rev_fastqc/', directory + "rev_fastqc/")
# Clean up #
shutil.rmtree(tmp_dir)
def assembly_fastqc(self):
sh.fastqc(self.assembled.path, '-q')
os.remove(os.path.splitext(self.assembled.path)[0] + '_fastqc.zip')
def barcode_fastqc(self):
sh.fastqc(self.fwd_path, '-q')
os.remove(os.path.splitext(self.fwd_path)[0] + '_fastqc.zip')
sh.fastqc(self.rev_path, '-q')
os.remove(os.path.splitext(self.rev_path)[0] + '_fastqc.zip')
def check(self):
assert sh.fastqc('--v', )
def assembly_fastqc(self):
sh.fastqc(self.assembled.path, '-q')
os.remove(os.path.splitext(self.assembled.path)[0] + '_fastqc.zip')
def barcode_fastqc(self):
sh.fastqc(self.fwd_path, '-q')
os.remove(os.path.splitext(self.fwd_path)[0] + '_fastqc.zip')
sh.fastqc(self.rev_path, '-q')
os.remove(os.path.splitext(self.rev_path)[0] + '_fastqc.zip')
