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 load_impmodel_from_cmm(f_name, rand_init=None, radius=None):
'''
Loads an IMPmodel object using an cmm file of the form:
::
<marker_set name="1">
<marker id="1" x="7347.50964739" y="-7743.92836303" z="-8283.39749204" r="0.00990099009901" g="0" b="0.990099009901" radius="500.0" note="1"/>
<marker id="2" x="7647.90254377" y="-7308.1816344" z="-7387.75932893" r="0.019801980198" g="0" b="0.980198019802" radius="500.0" note="2"/>
<link id1="1" id2="2" r="1" g="1" b="1" radius="250.0"/>
</marker_set>
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
'''
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel((('x', {}), ('y', {}), ('z', {}), ('rand_init', rand_init),
('index', 0), ('objfun', 0), ('radius', radius)))
expr = compil(
' x="([0-9.-]+)" y="([0-9.-]+)" z="([0-9.-]+)".* radius="([0-9.]+)"')
for xxx, yyy, zzz, radius in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
if not model['radius']:
model['radius'] = float(radius)
return model
def load_impmodel_from_xyz_OLD(f_name, rand_init=None, radius=None,
chromosome='UNKNOWN', start=0, resolution=1):
"""
Loads an IMPmodel object using an xyz file of the form:
::
p1 1 44.847 412.828 -162.673
p2 2 -55.574 396.869 -129.782
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
"""
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel((('x', []), ('y', []), ('z', []), ('rand_init', rand_init),
('objfun', None), ('radius', radius)))
expr = compil('p[0-9]+\s+[0-9]+\s+([0-9.-]+)\s+([0-9.-]+)\s+([0-9.-]+)')
for xxx, yyy, zzz in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
model['description'] = {'chromosome':chromosome,
'start': start, 'resolution': resolution}
return model
def write(self, outfile=None, item='seq', reverse=False, width=60, descr=False):
"""
Write sequence object to file in fasta format
:argument None outfile: path to outfile, if None than, print to stdout
:argument seq item: what to put in place of sequence
:argument False reverse: wether to reverse or not the sequence
:argument 60 width: number of sites per line when printing sequence
:argument False descr: put description of sequence also, not recommended if you are not sure how the aligner will read it.
"""
if outfile:
out = open (outfile, 'w')
else:
out = stdout
wsub = compil('([A-Za-z-]{'+str(width)+'})')
for elt in self:
if decr:
out.write ('>%s |%s\n' % (elt, self[elt]['descr']))
else:
out.write ('>%s\n' % (elt))
seq = self[elt][item][::-1] if reverse else self[elt][item]
seq = seq if type(seq) is str else ''.join(seq)
out.write ('%s\n' % (sub(wsub, '\\1\n', seq)))
if outfile:
out.close()
def load_impmodel_from_cmm(f_name, rand_init=None, radius=None):
'''
Loads an IMPmodel object using an cmm file of the form:
::
<marker_set name="1">
<marker id="1" x="7347.50964739" y="-7743.92836303" z="-8283.39749204" r="0.00990099009901" g="0" b="0.990099009901" radius="500.0" note="1"/>
<marker id="2" x="7647.90254377" y="-7308.1816344" z="-7387.75932893" r="0.019801980198" g="0" b="0.980198019802" radius="500.0" note="2"/>
<link id1="1" id2="2" r="1" g="1" b="1" radius="250.0"/>
</marker_set>
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
'''
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel((('x', []), ('y', []), ('z', []), ('rand_init', rand_init),
('index', 0), ('objfun', 0), ('radius', radius)))
expr = compil(
' x="([0-9.-]+)" y="([0-9.-]+)" z="([0-9.-]+)".* radius="([0-9.]+)"')
for xxx, yyy, zzz, radius in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
if not model['radius']:
model['radius'] = float(radius)
return model
def load_impmodel_from_xyz_OLD(f_name, rand_init=None, radius=None,
chromosome='UNKNOWN', start=0, resolution=1):
"""
Loads an IMPmodel object using an xyz file of the form:
::
p1 1 44.847 412.828 -162.673
p2 2 -55.574 396.869 -129.782
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
"""
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel((('x', []), ('y', []), ('z', []), ('rand_init', rand_init),
('objfun', None), ('radius', radius)))
expr = compil('p[0-9]+\s+[0-9]+\s+([0-9.-]+)\s+([0-9.-]+)\s+([0-9.-]+)')
for xxx, yyy, zzz in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
model['description'] = {'chromosome':chromosome,
'start': start, 'resolution': resolution}
return model
def read_fasta(infile):
'''
read file in fasta format and yield each sequence
'''
nam = None
descr = None
seq = ''
blank_re = compil('[ \t]')
for line in open(infile):
line = line.strip()
if line.startswith('>'):
if nam is not None:
if seq == '':
print >> stderr, 'ERROR: no sequence for ', str(nam)
exit()
yield { 'name' : nam,
'descr' : descr,
'seq' : seq
}
items = blank_re.split(line, maxsplit=1)
nam = items[0].lstrip('>')
descr = items[1] if len (items) == 2 else None
seq = ''
continue
seq += blank_re.sub('', line)
if seq == '' and nam is not None:
print >> stderr, 'ERROR: no sequence for ', str(nam)
exit()
elif seq == '':
print >> stderr, 'ERROR: presence of repeated names'
exit()
yield { 'name' : nam,
'descr' : descr,
'seq' : seq
}
def write(self,
outfile=None,
item='seq',
reverse=False,
width=60,
descr=False):
"""
Write sequence object to file in fasta format
:argument None outfile: path to outfile, if None than, print to stdout
:argument seq item: what to put in place of sequence
:argument False reverse: wether to reverse or not the sequence
:argument 60 width: number of sites per line when printing sequence
:argument False descr: put description of sequence also, not recommended if you are not sure how the aligner will read it.
"""
if outfile:
out = open(outfile, 'w')
else:
out = stdout
wsub = compil('([A-Za-z-]{' + str(width) + '})')
for elt in self:
if decr:
out.write('>%s |%s\n' % (elt, self[elt]['descr']))
else:
out.write('>%s\n' % (elt))
seq = self[elt][item][::-1] if reverse else self[elt][item]
seq = seq if type(seq) is str else ''.join(seq)
out.write('%s\n' % (sub(wsub, '\\1\n', seq)))
if outfile:
out.close()
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
def load_impmodel_from_xyz(f_name, rand_init=None, radius=None):
"""
Loads an IMPmodel object using an xyz file of the form:
::
# ID : some identifier
# SPECIES : None
# CELL TYPE : None
# EXPERIMENT TYPE : Hi-C
# RESOLUTION : 10000
# ASSEMBLY : None
# CHROMOSOME : 19
# START : 1
# END : 50
1 19:1-10000 44.847 412.828 -162.673
2 19:10001-20000 -55.574 396.869 -129.782
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
"""
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel(
(('x', []), ('y', []), ('z', []), ('rand_init', rand_init),
('index', 0), ('objfun', 0), ('radius', radius)))
expr = compil(
'[0-9]+\s[A-Za-z0-9_ ]+:[0-9]+-[0-9]+\s+([0-9.-]+)\s+([0-9.-]+)\s+([0-9.-]+)'
)
model['description'] = {}
for line in open(f_name):
if line.startswith('# '):
key, val = line.strip('# ').split(':')
model['description'][key.strip().lower()] = val.strip()
for xxx, yyy, zzz in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
return model
def load_impmodel_from_xyz(f_name, rand_init=None, radius=None):
"""
Loads an IMPmodel object using an xyz file of the form:
::
# ID : some identifier
# SPECIES : None
# CELL TYPE : None
# EXPERIMENT TYPE : Hi-C
# RESOLUTION : 10000
# ASSEMBLY : None
# CHROMOSOME : 19
# START : 1
# END : 50
1 19:1-10000 44.847 412.828 -162.673
2 19:10001-20000 -55.574 396.869 -129.782
:params f_name: path where to find the file
:params None rand_init: IMP random initial number used to generate the model
:param None radius: radius of each particle
:return: IMPmodel
"""
if not rand_init:
try:
rand_init = str(int(f_name.split('.')[-2]))
except:
rand_init = None
model = IMPmodel((('x', []), ('y', []), ('z', []), ('rand_init', rand_init),
('index', 0), ('objfun', 0), ('radius', radius)))
expr = compil('[0-9]+\s[A-Za-z0-9_ ]+:[0-9]+-[0-9]+\s+([0-9.-]+)\s+([0-9.-]+)\s+([0-9.-]+)')
model['description'] = {}
for line in open(f_name):
if line.startswith('# '):
key, val = line.strip('# ').split(':')
model['description'][key.strip().lower()] = val.strip()
for xxx, yyy, zzz in findall(expr, open(f_name).read()):
model['x'].append(float(xxx))
model['y'].append(float(yyy))
model['z'].append(float(zzz))
return model
def main():
'''
main function when called by command line.
'''
opts = get_options()
log = '\n\n'
gencode = _set_code(opts.code)
seqs = {}
for seq in read_fasta(opts.fastafile):
seq['trseq'] = translate(seq['seq'], gencode, stop=opts.remove_stop)
seqs[seq['name']] = seq
log += ' ' + str (len (seqs)) + ' sequences\n\n'
prot_path = opts.outfile + '_prot'
aali_path = opts.outfile + '_aa_ali'
ali_path = opts.outfile + '_ali'
trimsq_path = opts.outfile + '_trimseq'
trimcl_path = opts.outfile + '_trimcol'
score_path = opts.outfile + '_score'
map_path = opts.outfile + '_map'
todel = [prot_path]
write_fasta(seqs, prot_path, clean=True, typ='trseq')
if opts.only_translate:
exit()
###########
# ALIGN
if not opts.input_ali:
proc = Popen([opts.muscle_bin,
'-quiet',
'-noanchors',
'-maxiters' , '999',
'-maxhours' , '24 ',
'-maxtrees' , '100',
'-in' , prot_path,
'-out' , aali_path,
'-scorefile', score_path # must be last!!! because option...
][:None if opts.score else -2], stdout=PIPE)
if proc.communicate()[1] is not None:
print >> stderr, proc.communicate()[0]
exit('\nERROR: runninge muscle')
log += ' Muscle command line: \n' + \
' '.join([opts.muscle_bin, '-quiet', '-noanchors', '-maxiters' , \
'999', '-maxhours', '24 ', '-maxtrees', '100', '-in', \
prot_path, '-out', aali_path, '-scorefile', \
score_path][:None if opts.score else -2]) + '\n\n'
else:
proc = Popen(['cp', prot_path, aali_path], stdout=PIPE)
if proc.communicate()[1] is not None:
print >> stderr, proc.communicate()[0]
exit('\nERROR: when skipping muscle.')
###########
# TRIM SEQS
if opts.trimseq != False:
todel.append(trimsq_path)
proc = Popen([opts.trimal_bin,
'-in' , aali_path,
'-out' , trimsq_path,
'-resoverlap', str (opts.trimseq[1]),
'-seqoverlap', str (opts.trimseq[2]),
'-cons' , '100'
], stdout=PIPE)
if proc.communicate()[1] is not None:
print >> stderr, proc.communicate()[0]
exit('\nERROR: runninge muscle')
for seq in read_fasta(trimsq_path):
seqs[seq['name']]['ali'] = seq['seq']
trimmed = filter (lambda x: not seqs[x].has_key('ali'), seqs)
if not opts.quiet:
print >> stderr, 'WARNING: trimmed sequences: \n\t' + \
'\n\t'.join(trimmed)
if len (trimmed) > 0:
log += '->trimmed sequences: \n\t' + \
'\n\t'.join(trimmed) + '\n'
else: log += '->no trimmed sequences\n'
for s in seqs.keys():
if s in trimmed:
del(seqs[s])
aali_path = trimsq_path
log += ' Trimal (sequences) command line: \n' + \
' '.join([opts.trimal_bin, '-in', aali_path, '-out',
trimsq_path, '-resoverlap', str (opts.trimseq[1]), \
'-seqoverlap', str (opts.trimseq[2]), '-cons', '100']) \
+ '\n\n'
else:
for seq in read_fasta(aali_path):
seqs[seq['name']]['ali'] = seq['seq']
###########
# CODON MAP
seqs = map2codons(seqs, opts.input_ali)
###########
# TRIM COLS
if opts.trimcol != 'None':
if opts.trimcol == 'specific':
todel.append(trimcl_path)
proc = Popen([opts.trimal_bin,
'-in' , aali_path,
'-out', trimcl_path,
'-gt' , str (opts.gaptreshold),
'-st' , str (opts.similarity),
'-colnumbering'
], stdout=PIPE)
(keeplist, err) = proc.communicate()
if err is not None:
exit('ERROR: trimming columns.')
log += ' Trimal (columns) command line: \n' + \
' '.join([opts.trimal_bin,
'-in', aali_path,
'-out', trimcl_path,
'-gt', str (opts.gaptreshold),
'-st', str (opts.similarity),
'-colnumbering'
]) + '\n'
else:
todel.append(trimcl_path)
proc = Popen([opts.trimal_bin,
'-in' , aali_path,
'-out', trimcl_path,
'-' + opts.trimcol,
'-colnumbering'
], stdout=PIPE)
(keeplist, err) = proc.communicate()
if err is not None:
exit('ERROR: trimming columns.')
log += ' Trimal (columns) command line: \n' + \
' '.join([opts.trimal_bin, '-in' , aali_path, '-out',
trimcl_path, '-' + opts.trimcol, \
'-colnumbering']) + '\n'
keeplist = str (keeplist).strip().split(', ')
algt = get_alignment(seqs)
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] = map (lambda x: nnn.sub('', x), col)
algt[num] = map (lambda x: compil('---').sub('', x), algt[num])
else:
for (col, num) in zip (algt, range (len (algt))):
if not str(num) in keeplist:
algt[num] = map (lambda x: nnn.sub('NNN', x), col)
for (key, seq) in zip (sorted (seqs.keys()), zip (*algt)):
seqs[key]['codons'] = ''.join(seq)
###########
# SEQ MAP
if opts.printmap:
_printmap(seqs, map_path, opts.pymap)
write_fasta(seqs, ali_path, clean=opts.clean, typ='codons')
Popen(['rm', '-f'] + todel, stdout=PIPE)
if opts.print_log:
print log
