def run_alignments(path, cpus=1, quiet=False, tries=2):
"""
"""
procs = []
files = []
aligners = [ali for ali in BINARIES if 'fun' in BINARIES[ali]]
for ali in sorted(
aligners,
key=lambda x: ['probcons', 'dialign', 'muscle', 'mafft'].index(x)):
if not quiet:
print 'Aligning with: ' + ali
for sense in ['prot', 'torp'][:tries]:
if not quiet:
print ' -> ' + ('sense' if sense == 'prot' else 'anti-sense')
files.append('%s.fasta_%s' % (sense, ali))
procs.append(BINARIES[ali]['fun'](path + '/%s.fasta' % (sense)))
if len(procs) < cpus: continue
while len(procs) != 0:
out, err = procs.pop(0).communicate()
if 'ERROR' in err:
print >> stderr, out, err
raise Exception('\nERROR: running alignments')
while len(procs) != 0:
out, err = procs.pop(0).communicate()
if 'ERROR' in err:
print >> stderr, out, err
raise Exception('\nERROR: running alignments')
for fil in files:
if fil.startswith('torp'):
seqs = parse_fasta(path + '/' + fil)
for seq in seqs:
seqs[seq]['seq'] = seqs[seq]['seq'][::-1]
write_rfasta(seqs, path + '/' + fil)
def run_alignments(path, cpus=1, quiet=False, tries=2):
"""
"""
procs = []
files = []
aligners = [ali for ali in BINARIES if 'fun' in BINARIES[ali]]
for ali in sorted(aligners, key=lambda x:['probcons', 'dialign',
'muscle','mafft'].index(x)):
if not quiet:
print 'Aligning with: ' + ali
for sense in ['prot', 'torp'][:tries]:
if not quiet:
print ' -> ' + ('sense' if sense == 'prot' else 'anti-sense')
files.append('%s.fasta_%s' % (sense, ali))
procs.append (BINARIES[ali]['fun'](path + '/%s.fasta' % (sense)))
if len (procs) < cpus: continue
while len (procs) != 0:
out, err = procs.pop(0).communicate()
if 'ERROR' in err:
print >> stderr, out, err
raise Exception ('\nERROR: running alignments')
while len (procs) != 0:
out, err = procs.pop(0).communicate()
if 'ERROR' in err:
print >> stderr, out, err
raise Exception ('\nERROR: running alignments')
for fil in files:
if fil.startswith('torp'):
seqs = parse_fasta (path + '/' + fil)
for seq in seqs:
seqs[seq]['seq'] = seqs[seq]['seq'][::-1]
write_rfasta (seqs, path + '/' + fil)
def trim_columns(sequences, opts, tmp_dir):
aali_path = tmp_dir + '/aligned.fasta'
write_rfasta(sequences, aali_path, what='aa_ali')
trimcl_path = tmp_dir + '/trimmed.fasta'
if opts.trimcol == 'specific':
cmds = [BINARIES['trimal']['bin'], '-in' , aali_path,
'-out', trimcl_path, '-gt' , str (opts.gaptreshold),
'-st' , str (opts.similarity), '-colnumbering']
else:
cmds = [BINARIES['trimal']['bin'], '-in' , aali_path,
'-out', trimcl_path, '-' + opts.trimcol,
'-colnumbering']
proc = Popen(cmds, stdout=PIPE, stderr=PIPE)
(keeplist, err) = proc.communicate()
LOG.append('')
if 'ERROR' in err:
exit('ERROR: trimming columns:\n' + err)
keeplist = str (keeplist).strip().split(', ')
algt = get_alignment(sequences, typ=('aa_ali' if opts.aa else 'codon'))
nnn = compil('[A-Z]{3}')
if opts.nogap:
for (col, num) in zip (algt, range (len (algt))):
if not str(num) in keeplist:
algt[num] = [ nnn.sub('', x) for x in col ]
algt[num] = [ compil('---').sub('', x) for x in algt[num]]
else:
for (col, num) in zip (algt, range (len (algt))):
if not str(num) in keeplist:
algt[num] = [ nnn.sub('NNN', x) for x in col ]
for (key, seq) in zip (sorted (sequences.keys()), zip (*algt)):
sequences[key]['codon'] = seq
def main():
"""
main function
"""
opts = get_options()
genetic_code = None if opts.aa else get_genetic_code(opts.code)
sequences = parse_fasta(opts.fastafile, genetic_code)
tmp_dir = dirname(
opts.outfile) + ('/tmp' if '/' in opts.outfile else 'tmp')
Popen('mkdir -p ' + tmp_dir, shell=True).communicate()
### if we need to align:
# write sense and anti-sense translated sequences
this = 'seq' if opts.aa else 'prot'
write_rfasta(sequences, tmp_dir + '/prot.fasta', what=this)
if opts.align == 2:
write_rfasta(sequences, tmp_dir + '/torp.fasta', what=this, rev=True)
# run alignment
if opts.align:
run_alignments(tmp_dir, opts.cpus, opts.quiet, opts.align)
# merge all in one, keep only sites with score better than m_coffee cut
aligners = [ali for ali in BINARIES if 'fun' in BINARIES[ali]]
if len(aligners) > 1 or opts.align == 2:
merge_mcoffee(tmp_dir, opts.mcoffee_cut, sequences, aa=opts.aa)
else:
aa_ali = parse_fasta(tmp_dir + '/prot.fasta_' + aligners[0])
for seq in sequences:
sequences[seq]['aa_ali'] = aa_ali[seq]['seq']
for elt in xrange(len(sequences[seq]['aa_ali'])):
if sequences[seq]['aa_ali'][elt] == '-':
sequences[seq]['codon'].insert(elt, '---')
continue
# trimal
if opts.trimseq:
trim_sequences(tmp_dir,
opts.outfile,
sequences,
opts.trimseq,
quiet=opts.quiet)
if opts.trimcol != 'None':
trim_columns(sequences, opts, tmp_dir)
# write codon sequences
if opts.aa:
write_fasta(sequences, opts.outfile, what='aa_ali')
else:
write_fasta(sequences, opts.outfile, what='codon')
# print map
if opts.printmap:
printmap(sequences, opts.outfile + '.map', opts.pymap)
def main():
"""
main function
"""
opts = get_options()
genetic_code = None if opts.aa else get_genetic_code (opts.code)
sequences = parse_fasta (opts.fastafile, genetic_code)
tmp_dir = dirname(opts.outfile) + ('/tmp' if '/' in opts.outfile else 'tmp')
Popen('mkdir -p ' + tmp_dir, shell=True).communicate()
### if we need to align:
# write sense and anti-sense translated sequences
this = 'seq' if opts.aa else 'prot'
write_rfasta(sequences, tmp_dir + '/prot.fasta', what=this)
if opts.align == 2:
write_rfasta(sequences, tmp_dir + '/torp.fasta', what=this, rev=True)
# run alignment
if opts.align:
run_alignments(tmp_dir, opts.cpus, opts.quiet, opts.align)
# merge all in one, keep only sites with score better than m_coffee cut
aligners = [ali for ali in BINARIES if 'fun' in BINARIES[ali]]
if len(aligners) > 1 or opts.align == 2:
merge_mcoffee(tmp_dir, opts.mcoffee_cut, sequences, aa=opts.aa)
else:
aa_ali = parse_fasta(tmp_dir + '/prot.fasta_' + aligners[0])
for seq in sequences:
sequences[seq]['aa_ali'] = aa_ali[seq]['seq']
for elt in xrange(len(sequences[seq]['aa_ali'])):
if sequences[seq]['aa_ali'][elt] == '-':
sequences[seq]['codon'].insert(elt, '---')
continue
# trimal
if opts.trimseq:
trim_sequences(tmp_dir, opts.outfile, sequences,
opts.trimseq, quiet=opts.quiet)
if opts.trimcol != 'None':
trim_columns(sequences, opts, tmp_dir)
# write codon sequences
if opts.aa:
write_fasta(sequences, opts.outfile, what='aa_ali')
else:
write_fasta(sequences, opts.outfile, what='codon')
# print map
if opts.printmap:
printmap(sequences, opts.outfile + '.map', opts.pymap)
def trim_columns(sequences, opts, tmp_dir):
aali_path = tmp_dir + '/aligned.fasta'
write_rfasta(sequences, aali_path, what='aa_ali')
trimcl_path = tmp_dir + '/trimmed.fasta'
if opts.trimcol == 'specific':
cmds = [
BINARIES['trimal']['bin'], '-in', aali_path, '-out', trimcl_path,
'-gt',
str(opts.gaptreshold), '-st',
str(opts.similarity), '-colnumbering'
]
else:
cmds = [
BINARIES['trimal']['bin'], '-in', aali_path, '-out', trimcl_path,
'-' + opts.trimcol, '-colnumbering'
]
proc = Popen(cmds, stdout=PIPE, stderr=PIPE)
(keeplist, err) = proc.communicate()
LOG.append('')
if 'ERROR' in err:
exit('ERROR: trimming columns:\n' + err)
keeplist = str(keeplist).strip().split(', ')
algt = get_alignment(sequences)
nnn = compil('[A-Z]{3}')
if opts.nogap:
for (col, num) in zip(algt, range(len(algt))):
if not str(num) in keeplist:
algt[num] = [nnn.sub('', x) for x in col]
algt[num] = [compil('---').sub('', x) for x in algt[num]]
else:
for (col, num) in zip(algt, range(len(algt))):
if not str(num) in keeplist:
algt[num] = [nnn.sub('NNN', x) for x in col]
for (key, seq) in zip(sorted(sequences.keys()), zip(*algt)):
sequences[key]['codon'] = seq
