def maskResiduesNOMAP(refMSA_file, numseq, alnlen, scores, x, formatout, final_file, seqType): ''' Masks poorly aligned residues whose score is <x. Will NOT mask gaps.''' new='?' parsed = AlignIO.read(refMSA_file, 'fasta') newseqs=[] numres=0 totalmasked=0 maskedMSA=MultipleSeqAlignment([]) for row in range(numseq): newseq='' for position in range(alnlen): thispos=str(parsed[row].seq[position]) if thispos=='-': newseq=newseq+parsed[row].seq[position] else: numres+=1 thescore=scores[row][position] if float(thescore)<float(x): #mask if below threshold. newseq=newseq+new totalmasked+=1 else: #or, keep that position newseq=newseq+parsed[row].seq[position] newseqs.append(newseq) for i in range(numseq): if str(seqType)=='protein': aln_record=SeqRecord(Seq(newseqs[i],generic_protein), id=str(i+1), description='') elif str(seqType)=='dna': aln_record=SeqRecord(Seq(newseqs[i],generic_dna), id=str(i+1), description='') maskedMSA.append(aln_record) outhandle=open(final_file, 'w') outhandle.write(maskedMSA.format(str(formatout))) outhandle.close()
def show_alignments(ali_path, out_format):
for aln in np.load(ali_path, allow_pickle=True):
if out_format == 'pir':
msa = MultipleSeqAlignment([
SeqRecord(Seq(aln[0], generic_protein),
id='Query',
name='',
description='sequence:::::::::'),
SeqRecord(Seq(aln[1], generic_protein),
id='Template',
name='',
description='structureX:::::::::')
])
else:
msa = MultipleSeqAlignment([
SeqRecord(Seq(aln[0], generic_protein),
id='Query',
name='',
description=''),
SeqRecord(Seq(aln[1], generic_protein),
id='Template',
name='',
description='')
])
print(msa.format(out_format))
def maskResiduesNOMAP(refMSA_file, numseq, alnlen, scores, x, formatout, final_file, seqType): ''' Masks poorly aligned residues whose score is <x. Will NOT mask gaps.''' new='?' parsed = AlignIO.read(refMSA_file, 'fasta') newseqs=[] numres=0 totalmasked=0 maskedMSA=MultipleSeqAlignment([]) for row in range(numseq): newseq='' for position in range(alnlen): thispos=str(parsed[row].seq[position]) if thispos=='-': newseq=newseq+parsed[row].seq[position] else: numres+=1 thescore=scores[row][position] if float(thescore)<float(x): #mask if below threshold. newseq=newseq+new totalmasked+=1 else: #or, keep that position newseq=newseq+parsed[row].seq[position] newseqs.append(newseq) for i in range(numseq): if str(seqType)=='protein': aln_record=SeqRecord(Seq(newseqs[i],generic_protein), id=str(i+1), description='') elif str(seqType)=='dna': aln_record=SeqRecord(Seq(newseqs[i],generic_dna), id=str(i+1), description='') maskedMSA.append(aln_record) outhandle=open(final_file, 'w') outhandle.write(maskedMSA.format(str(formatout))) outhandle.close()
def to_string(self, begin=None, end=None, **kwargs):
# type: (Union[int, None], Union[int, None], Dict[str, Any]) -> str
format = get_value(kwargs, "format", None)
if format == "pretty":
return self._to_string_pretty(begin, end, **kwargs)
# add markers as sequence records
seq_records = [
SeqRecord(Seq(m.to_string(begin, end)), id="#{}".format(m.name))
for m in self.list_msa_markers
]
if begin is not None or end is not None:
begin = begin if begin is not None else 0
end = end if end is not None else self.alignment_length()
# add actual sequences
for a in self.list_alignment_sequences:
if begin is not None or end is not None:
seq_records.append(a[begin:end])
else:
seq_records.append(a)
# create alignment with markers
alignment = MultipleSeqAlignment(seq_records)
return alignment.format("clustal")
def alignment_slicer(input, informat, outformat, SNPs, slide):
alignment = AlignIO.read(input, informat, alphabet = generic_dna)
alignment_seq_count = len(alignment)
first_seq = (alignment[0].seq)
length_alignment = len(first_seq)
chars_to_ignore = ['N']
start = 0
end = start + args.SNPs_in_window
while end <= length_alignment:
with open(input+'_site'+str(start)+'to'+str(end)+'.'+outformat, 'w') as output_handle:
# print 'start:', start
# print 'end:', end
alignment_iteration = MultipleSeqAlignment(alignment[:, start:end], alphabet=generic_dna)
if outformat.lower() == 'nexus':
n_alignments = []
alignment_iteration = alignment_iteration.format('nexus')
n_alignments.append(('site'+str(start)+'to'+str(end),Nexus.Nexus(alignment_iteration)))
combined = Nexus.combine(n_alignments)
combined.write_nexus_data(output_handle)
else:
AlignIO.write(alignment_iteration, output_handle, outformat)
# print alignment_iteration
start += args.slide
end += args.slide
else:
with open(input+'_site'+str(start)+'to'+str(length_alignment)+'.'+outformat, 'w') as output_handle:
n_alignments = []
# print 'now in else loop\n'
# print 'start:', start
# print 'end:', length_alignment
alignment_iteration = MultipleSeqAlignment(alignment[:, start:length_alignment], alphabet=generic_dna)
if outformat.lower() == 'nexus':
n_alignments = []
alignment_iteration = alignment_iteration.format('nexus')
n_alignments.append(('site'+str(start)+'to'+str(end),Nexus.Nexus(alignment_iteration)))
combined = Nexus.combine(n_alignments)
combined.write_nexus_data(output_handle)
else:
AlignIO.write(alignment_iteration, output_handle, outformat)
# print alignment_iteration
print "\ndone\n"
def test_alnRemoveGapOnlyCols(self):
s1 = SeqRecord(Seq('A-TT---TTAA---'),id='s1',name='s1')
s2 = SeqRecord(Seq('AATT---TTAA---'),id='s2',name='s2')
aln = MultipleSeqAlignment([s1, s2])
s1_nogap = SeqRecord(Seq('A-TTTTAA'),id='s1',name='s1')
s2_nogap = SeqRecord(Seq('AATTTTAA'),id='s2',name='s2')
alnnogap = MultipleSeqAlignment([s1_nogap, s2_nogap])
aln = MultipleSeqAlignment([s1, s2])
# Use format() to report, because the Align objects will be compared
# by hash values which will not be equal
self.assertEqual(
Milraa.alnRemoveGapOnlyCols(aln).format('fasta'),
alnnogap.format('fasta'))
def convert_a2m(ali):
fh = cStringIO.StringIO(ali)
msa = AlignIO.read(fh, 'fasta')
fh.close()
new_msa = []
for rec in msa:
new_seq = Seq(re.sub(r'[a-z.]', '', str(rec.seq)),
SingleLetterAlphabet())
new_rec = rec
new_rec.seq = new_seq
new_msa.append(new_rec)
new_msa = MultipleSeqAlignment(new_msa)
return new_msa.format('fasta')
del letters
print("testing reading and writing clustal format...")
test_dir = os.path.join(os.getcwd(), 'Clustalw')
test_names = ['opuntia.aln', 'cw02.aln']
test_files = []
for name in test_names:
test_files.append(os.path.join(test_dir, name))
for test_file in test_files:
# parse the alignment file and get an aligment object
alignment = AlignIO.read(test_file, "clustal")
# print the alignment back out
print(alignment.format("clustal"))
alignment = AlignIO.read(os.path.join(test_dir, test_names[0]), "clustal",
alphabet=Alphabet.Gapped(IUPAC.unambiguous_dna))
# test the base alignment stuff
print('all_seqs...')
for seq_record in alignment:
print('description: %s' % seq_record.description)
print('seq: %r' % seq_record.seq)
print('length: %i' % alignment.get_alignment_length())
print('Calculating summary information...')
align_info = AlignInfo.SummaryInfo(alignment)
dumb_consensus = align_info.dumb_consensus(ambiguous="N", threshold=0.6)
simple_consensus = align_info.simple_consensus(ambiguous="N", threshold=0.6)
del letters
print("testing reading and writing clustal format...")
test_dir = os.path.join(os.getcwd(), 'Clustalw')
test_names = ['opuntia.aln', 'cw02.aln']
test_files = []
for name in test_names:
test_files.append(os.path.join(test_dir, name))
for test_file in test_files:
# parse the alignment file and get an aligment object
alignment = AlignIO.read(test_file, "clustal")
# print the alignment back out
print(alignment.format("clustal"))
alignment = AlignIO.read(os.path.join(test_dir, test_names[0]),
"clustal",
alphabet=Alphabet.Gapped(IUPAC.unambiguous_dna))
# test the base alignment stuff
print('all_seqs...')
for seq_record in alignment:
print('description: %s' % seq_record.description)
print('seq: %r' % seq_record.seq)
print('length: %i' % alignment.get_alignment_length())
print('Calculating summary information...')
align_info = AlignInfo.SummaryInfo(alignment)
consensus = align_info.dumb_consensus()
taxon_sequence_threshold = 0.40
in_alignment=AlignIO.read(open(infile), "nexus", alphabet=Gapped(IUPAC.protein))
alignment_length = in_alignment.get_alignment_length();
out_alignment=MultipleSeqAlignment([], alphabet=Gapped(IUPAC.protein))
for seq_record in in_alignment:
missing_data = float(seq_record.seq.count("-"))/float(alignment_length)
print seq_record.id + "\t" + str(missing_data)
if missing_data < taxon_sequence_threshold:
out_alignment.append(seq_record)
outname=re.search("(/.+)\.nexus$",infile).group(1) + "." + str(taxon_sequence_threshold)
nexfile=open((outname + ".nexus"), "w")
phyfile=open((outname + ".phylip"), "w")
try:
nexfile.write(out_alignment.format("nexus"))
phyfile.write(out_alignment.format("phylip"))
except:
print "Could not write alignment " + infile + " ", sys.exc_info()[0]
nexfile.close()
phyfile.close()
if args.extract_ref: inddict['reference']='reference' seqdict['reference']=[refseq] #for ind in seqdict.keys(): # print ind,seqdict[ind] #generate alignment for ind in inddict.keys(): for i in range(1): # print ">"+ind+"_"+str(i) # print seqdict[ind][i] seqrec=SeqRecord(Seq(seqdict[ind][i], generic_dna), id=ind+"_"+str(i), description=ind+"_"+str(i)) # print seqrec seqrecords.append(seqrec) align = MultipleSeqAlignment(seqrecords) #print alignemnt in desired formats for f in formats: print "writing to "+f+"\n" OUT = open(args.out_prefix+'.'+f[:3],'w') OUT.write(align.format(f)) if f is 'nexus' and args.popmap: add_traits_block(p_list=pops_list, i_dict=inddict, nexhandle=OUT) OUT.close()
def debug_matching(gen,
primer_pair,
mf,
mr,
output_file,
hanging_primers=False):
"""
This function computes and displays a single alignment. Used for debugging purposes
"""
try:
assert (len(gen) == 1), "Multiple gen sequences detected"
assert (len(primer_pair) == 1), "Multiple primer pairs detected"
except (Exception) as e:
logging.error(e)
return
try:
g = copy.deepcopy(gen)
except:
g = gen
pass
template, discarded, raw_stats, cooked_stats = compute_gen_matching(
mf, mr, primer_pair, g, output_file, hanging_primers=hanging_primers)
if (template.empty):
logging.warning("No result")
return
match_result = template.loc[0]
pp = primer_pair[next(iter(primer_pair))]
gen = gen[next(iter(gen))]
fpos = match_result.at['F_pos'] - 1
if (fpos < 0):
gen = '-' * (-fpos) + gen
fpos = 0
len_primer = fpos + pp.flen + match_result.at['ampliconLen'] + pp.rlen
rem_len = len(gen) - (fpos + pp.flen + match_result.at['ampliconLen'] +
pp.rlen)
pp.f.seq = Seq(''.join(pp.f.seq))
pp.r.seq = Seq(''.join(pp.r.seq))
pp_aligned = '-' * fpos + pp.f.seq + '-' * match_result.at[
'ampliconLen'] + pp.r.seq + '-' * rem_len
pp_aligned = SeqRecord(pp_aligned)
pp_aligned.id = pp.id
align = MultipleSeqAlignment([gen, pp_aligned])
print(align.format("clustal"))
try:
with open(output_file + ".txt", 'w') as outfile:
outfile.write(align.format("clustal"))
print("Debug saved")
except (Exception) as e:
logging.error(e)
return
def _to_string_pretty(self, begin=None, end=None, **kwargs):
# type: (Union[int, None], Union[int, None], Dict[str, Any]) -> str
tag = get_value(kwargs, "tag", "", default_if_none=True)
self.change_marker("q3prime", new_symbol="*")
# add markers as sequence records
seq_records = [
SeqRecord(Seq(m.to_string(begin, end)), id="#{}".format(m.name))
for m in self.list_msa_markers
]
if begin is not None or end is not None:
begin = begin if begin is not None else 0
end = end if end is not None else self.alignment_length()
headers_old = [a.id for a in self.list_alignment_sequences]
headers_new = MSAType._format_headers_pretty(headers_old)
# add actual sequences
for i, a in enumerate(self.list_alignment_sequences):
a.id = headers_new[i]
if begin is not None or end is not None:
seq_records.append(a[begin:end])
else:
seq_records.append(a)
# create alignment with markers
alignment = MultipleSeqAlignment(seq_records)
output_string = alignment.format("clustal")
# Remove header
output_string = output_string.replace("_", " ")
output_string_array = output_string.split("\n")
def get_summary_statistics_line_for_alignment():
# type: () -> str
ref_position = self.get_mark_position("ref")
is_lorf = len(set(self[0][0:ref_position])) <= 1
def count_lorf_targets_near_position(position):
# type: (int) -> int
count = 0
for idx in range(1, self.number_of_sequences()):
j = 0
while True:
if position - j >= 0 and self[idx][position -
j].isupper():
if len(set(self[idx][0:position - j])) <= 1:
count += 1
break
if position + j < self.alignment_length(
) and self[idx][position + j].isupper():
if len(set(self[idx][0:position + j])) <= 1:
count += 1
break
j += 1
return count
num_targets_that_are_lorf = count_lorf_targets_near_position(
ref_position)
return "{}: LORF={}\tTargetLORF={}".format(
tag,
str(is_lorf)[0], num_targets_that_are_lorf)
output_string_array[0] = get_summary_statistics_line_for_alignment()
output_string = "\n".join(output_string_array)
return output_string
