def deps(self):
longorfs = which('TransDecoder.LongOrfs')
if longorfs is None:
raise InstallationError('TransDecoder.LongOrfs not found.')
if self.logger:
self.logger.debug('TransDecoder.LongOrfs:' + longorfs)
return longorfs
def deps(self):
cmscan = which('cmscan')
if cmscan is None:
raise InstallationError('cmscan not found.')
if self.logger:
self.logger.debug('cmscan:' + cmscan)
return cmscan
def deps(self):
cmpress = which('cmpress')
if cmpress is None:
raise InstallationError('cmpress not found.')
if self.logger:
self.logger.debug('cmpress:' + cmpress)
return cmpress
def deps(self):
hmmpress = which('hmmpress')
if hmmpress is None:
raise InstallationError('hmmpress not found.')
if self.logger is not None:
self.logger.debug('hmmpress:' + hmmpress)
return hmmpress
def deps(self):
predict = which('TransDecoder.Predict')
if predict is None:
raise InstallationError('TransDecoder.Predict not found.')
else:
if self.logger:
logger.debug('TransDecoder.Predict:' + predict)
return predict
def deps(self):
buscov2 = which('BUSCO.py')
buscov3 = which('run_BUSCO.py')
tblastn = which('tblastn')
makeblastdb = which('makeblastdb')
if buscov2 is None and buscov3 is None:
raise InstallationError('BUSCO not found. NOTE: '\
'dammit 1.0 requires BUSCO v2 or greater')
if tblastn is None:
raise InstallationError('tblastn not found, required for BUSCO.')
if makeblastdb is None:
raise InstallationError(
'makeblastdb not found, required for BUSCO.')
if self.logger:
logger.debug('BUSCO:' + busco)
return buscov3 if buscov3 is not None else buscov2
def deps(self):
lastal = which('lastdb')
if lastal is None:
raise InstallationError('lastal not found.')
version = check_version(lastal)
if version < 600:
raise InstallationError('lastal version {0} < 600, please'\
' update'.format(version))
if self.logger:
self.logger.debug('lastal:' + lastal)
return lastal
def check_parallel(logger=None):
parallel = which('parallel')
if parallel is None:
raise InstallationError('parallel not found.')
else:
try:
version_string = subprocess.check_output(['parallel', '--version'])
except subprocess.CalledProcessError as e:
raise InstallationError('Error checking parallel '\
'version: [{0}] {1}'.format(e.returncode, e.output))
except OSError as e:
raise InstallationError('Error checking parallel version: '\
'[{0}] {1}'.format(e.errno, str(e)))
else:
version = version_string.strip().split()[2]
if logger:
logger.debug('parallel version:{0}'.format(version))
if int(version) < 20150000:
raise InstallationError('parallel version {0} < 20150000, '\
'please update'.format(version))
if logger:
logger.debug('parallel:' + parallel)
return parallel
def parallel_fasta(input_filename,
output_filename,
command,
n_jobs,
sshloginfile=None,
check_dep=True,
logger=None):
'''Given an input FASTA source, target, shell command, and number of jobs,
construct a gnu-parallel command to act on the sequences.
Args:
input_filename (str): The source FASTA.
output_filename (str): The target.
command (list): The shell command (in subprocess format).
n_jobs (int): Number of cores or nodes to split to.
sshloginfile (str): Path to file with node addresses.
check_dep (bool): If True, check for the gnu-parallel executable.
logger (logging.Logger): A logger to use.
Returns:
str: The constructed shell command.
'''
exc = which('parallel') if not check_dep else check_parallel(logger=logger)
cmd = [
'cat', input_filename, '|', exc, '--round-robin', '--pipe', '-L', 2,
'-N', 10000, '--gnu'
]
if sshloginfile is not None:
cmd.extend(['--sshloginfile', sshloginfile, '--workdir $PWD'])
else:
cmd.extend(['-j', n_jobs])
cmd.extend(['-a', input_filename])
if isinstance(command, list):
command = ' '.join(command)
cmd.extend([command, '>', output_filename])
return ' '.join(map(str, cmd))
