def pysam_install():
_subprocess.call([_sys.executable, 'setup.py', 'build'])
for f in _os.listdir('build/lib.linux-x86_64-2.7/pysam'):
if f[-3:] == '.so':
#print(f)
_shutil.copy('build/lib.linux-x86_64-2.7/pysam/{}'.format(f), 'pysam/{}'.format(f))
try:
_shutil.rmtree('../pysam')
except OSError:
pass
_os.rename('pysam', '../pysam')
def pysam_install():
_subprocess.call([_sys.executable, 'setup.py', 'build'])
for f in _os.listdir('build/lib.linux-x86_64-2.7/pysam'):
if f[-3:] == '.so':
#print(f)
_shutil.copy('build/lib.linux-x86_64-2.7/pysam/{}'.format(f),
'pysam/{}'.format(f))
try:
_shutil.rmtree('../pysam')
except OSError:
pass
_os.rename('pysam', '../pysam')
def generateReads(self, path_to_exe = False,
paths_to_genomes = False,
readcov = 60,
readlen = 100,
fraglen = 350,
sterrfraglen = 20,
model = 4,
max_cpus = -1):
'''
Call GemSIM to generate reads
Need to have written genome sequences to generate from, possibly with
generated SNPs, small indels and large deletions.
'''
#max_cpus etc
if paths_to_genomes:
use_genomes = sorted(paths_to_genomes)
elif hasattr(self, 'written_genomes'):
use_genomes = sorted(self.written_genomes)
else:
raise ValueError('provide either paths_to_genomes or generate some then .writeSequences()')
if not path_to_exe:
path_to_exe = _get_exe_path('gemsim')
comment2 = '''
to generate reads put GemSIM v1.6 into subfolder GemSIM_v1.6 and issue these commands:
GemSIM_v1.6/GemReads.py -r LESB58_for_GemSim_01.fasta -n 1980527 -l d -u 350 -s 20 -m GemSIM_v1.6/models/ill100v4_p.gzip -c -q 33 -p -o GemSimLESB58_01
'''
num_pairs = len(self.genome.sequence) * readcov / (readlen*2)
if model == 4:
path_to_model = _os.path.sep.join(path_to_exe.split(_os.path.sep)[:-1] + ['models','ill100v4_p.gzip'])
elif model == 5:
path_to_model = _os.path.sep.join(path_to_exe.split(_os.path.sep)[:-1] + ['models','ill100v5_p.gzip'])
print('Using error model: {}'.format(path_to_model))
print('Generating {:,} {}bp read pairs for {}x coverage depth of a {}bp genome ({})'.format(
num_pairs, readlen, readcov, len(self.genome.sequence), self.genome.id))
processes = set()
max_processes = _decide_max_processes( max_cpus )
import time
start = time.time()
out_raw = []
for i,genome_in in enumerate(use_genomes):
# could use per genome length . . less consistent than using reference
# genome_len = len(_SeqIO.read(genome_in,'fasta').seq)
# num_pairs = genome_len * readcov / (readlen*2)
outprefix = 'GemSim_{}_{:02d}'.format(self.genome.id, i+1)
cmd = [path_to_exe,
'-r', genome_in,
'-n', num_pairs,
'-l', 'd', '-u', fraglen, '-s', sterrfraglen,
'-m', path_to_model,
'-c',
'-q', 33, '-p',
'-o', outprefix]
out_raw += [outprefix+'_fir.fastq', outprefix+'_sec.fastq']
# this would be better to rename and compress all in one
# maybe as a shell script? Then resuming (--force) would be easier.
if _os.path.exists(outprefix+'_fir.fastq') and \
_os.path.exists(outprefix+'_sec.fastq'):
print('Found output for {}_fir.fastq (and sec), not regenerating, '\
'delete these to start from scratch'.format(outprefix))
else:
cmd = map(str,cmd)
print(' '.join(cmd))
processes.add( _subprocess.Popen(cmd, shell=False) )
if len(processes) >= max_processes:
(pid, exit_status) = _os.wait()
processes.difference_update(
[p for p in processes if p.poll() is not None])
# Check if all the child processes were closed
for p in processes:
if p.poll() is None:
p.wait()
missing = []
for o in out_raw:
if not _os.path.exists(o):
missing += [o]
assert len(missing) == 0, 'Could not find:\n{}'.format('\n'.join(missing))
print('all finished after {} minutes'.format(int(round((time.time() - start)/60.0))))
outdir = _os.path.sep.join(['simulated_reads',self.genome.id])
try:
_os.makedirs(outdir)
except OSError:
pass
for o in out_raw:
new = _os.path.sep.join([outdir, o.replace('fir','R1').replace('sec','R2')])
print('{} ==> {}'.format(o, new))
_os.rename(o, new)
cmd = ['gzip', new]
print(' '.join(cmd))
_subprocess.call(cmd)
def generateReads(self,
path_to_exe=False,
paths_to_genomes=False,
readcov=60,
readlen=100,
fraglen=350,
sterrfraglen=20,
model=4,
max_cpus=-1):
'''
Call GemSIM to generate reads
Need to have written genome sequences to generate from, possibly with
generated SNPs, small indels and large deletions.
'''
#max_cpus etc
if paths_to_genomes:
use_genomes = sorted(paths_to_genomes)
elif hasattr(self, 'written_genomes'):
use_genomes = sorted(self.written_genomes)
else:
raise ValueError(
'provide either paths_to_genomes or generate some then .writeSequences()'
)
if not path_to_exe:
path_to_exe = _get_exe_path('gemsim')
comment2 = '''
to generate reads put GemSIM v1.6 into subfolder GemSIM_v1.6 and issue these commands:
GemSIM_v1.6/GemReads.py -r LESB58_for_GemSim_01.fasta -n 1980527 -l d -u 350 -s 20 -m GemSIM_v1.6/models/ill100v4_p.gzip -c -q 33 -p -o GemSimLESB58_01
'''
num_pairs = len(self.genome.sequence) * readcov / (readlen * 2)
if model == 4:
path_to_model = _os.path.sep.join(
path_to_exe.split(_os.path.sep)[:-1] +
['models', 'ill100v4_p.gzip'])
elif model == 5:
path_to_model = _os.path.sep.join(
path_to_exe.split(_os.path.sep)[:-1] +
['models', 'ill100v5_p.gzip'])
print('Using error model: {}'.format(path_to_model))
print(
'Generating {:,} {}bp read pairs for {}x coverage depth of a {}bp genome ({})'
.format(num_pairs, readlen, readcov, len(self.genome.sequence),
self.genome.id))
processes = set()
max_processes = _decide_max_processes(max_cpus)
import time
start = time.time()
out_raw = []
for i, genome_in in enumerate(use_genomes):
# could use per genome length . . less consistent than using reference
# genome_len = len(_SeqIO.read(genome_in,'fasta').seq)
# num_pairs = genome_len * readcov / (readlen*2)
outprefix = 'GemSim_{}_{:02d}'.format(self.genome.id, i + 1)
cmd = [
path_to_exe, '-r', genome_in, '-n', num_pairs, '-l', 'd', '-u',
fraglen, '-s', sterrfraglen, '-m', path_to_model, '-c', '-q',
33, '-p', '-o', outprefix
]
out_raw += [outprefix + '_fir.fastq', outprefix + '_sec.fastq']
# this would be better to rename and compress all in one
# maybe as a shell script? Then resuming (--force) would be easier.
if _os.path.exists(outprefix+'_fir.fastq') and \
_os.path.exists(outprefix+'_sec.fastq'):
print('Found output for {}_fir.fastq (and sec), not regenerating, '\
'delete these to start from scratch'.format(outprefix))
else:
cmd = map(str, cmd)
print(' '.join(cmd))
processes.add(_subprocess.Popen(cmd, shell=False))
if len(processes) >= max_processes:
(pid, exit_status) = _os.wait()
processes.difference_update(
[p for p in processes if p.poll() is not None])
# Check if all the child processes were closed
for p in processes:
if p.poll() is None:
p.wait()
missing = []
for o in out_raw:
if not _os.path.exists(o):
missing += [o]
assert len(missing) == 0, 'Could not find:\n{}'.format(
'\n'.join(missing))
print('all finished after {} minutes'.format(
int(round((time.time() - start) / 60.0))))
outdir = _os.path.sep.join(['simulated_reads', self.genome.id])
try:
_os.makedirs(outdir)
except OSError:
pass
for o in out_raw:
new = _os.path.sep.join(
[outdir, o.replace('fir', 'R1').replace('sec', 'R2')])
print('{} ==> {}'.format(o, new))
_os.rename(o, new)
cmd = ['gzip', new]
print(' '.join(cmd))
_subprocess.call(cmd)