def _align_to_mature(seq, hairpin, mature):
"""Get alignment between seq and mature"""
mirna = get_mature_sequence(hairpin, mature)
hit = align(seq, mirna)
start = hit[0][:8].count("-") - 4 + int(mature[0])
logger.debug("PROST::align:sequence to mature %s" % hit[0])
logger.debug("PROST::align:start: %s -> %s" % (mature[0], start))
return start
def test_locala(self):
"""testing pairwise alignment"""
from mirtop.mirna.realign import align
print "\nExamples of perfect match, deletion, mutation"
print align("TGAGGTAGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0]
print align("TGAGGTGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0]
print align("TGAGGTAGTAGGCTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0]
def test_locala(self):
"""testing pairwise alignment"""
from mirtop.mirna.realign import align
print("\nExamples of perfect match, deletion, mutation")
print(align("TGAGTAGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGAGGTGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGAGGTAGTAGGCTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGANTAGTAGNTTGTATNGTT", "TGAGTAGTAGGTTGTATAGTTT")[0])
print(align("TGANTAGTNGNTTGTATNGTT", "TGAGTATAGGCCTTGTATAGTT")[0])
print(align("NCANAGTCCAAGNTCATN", "TCATAGTCCAAGGTCATG")[0])
def test_locala(self):
"""testing pairwise alignment"""
from mirtop.mirna.realign import align
print("\nExamples of perfect match, deletion, mutation")
print(align("TGAGTAGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGAGGTGTAGGTTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGAGGTAGTAGGCTGTATAGTT", "TGAGGTAGTAGGTTGTATAGTT")[0])
print(align("TGANTAGTAGNTTGTATNGTT", "TGAGTAGTAGGTTGTATAGTTT")[0])
print(align("TGANTAGTNGNTTGTATNGTT", "TGAGTATAGGCCTTGTATAGTT")[0])
print(align("NCANAGTCCAAGNTCATN", "TCATAGTCCAAGGTCATG")[0])
def tune(seq, precursor, start, cigar):
"""
The actual fn that will realign the sequence
"""
if cigar:
seq, mature = cigar_correction(cigar, seq, precursor[start:])
else:
seq, mature, score, p, size = align(seq,
precursor[start:start + len(seq)])
cigar = make_cigar(seq, mature)
if seq.startswith("-"):
seq = seq[1:]
if seq.endswith("-"):
seq = seq[:-1]
logger.debug("TUNE:: %s %s %s" % (cigar, seq, mature))
error = set()
pattern_addition = [[1, 1, 0], [1, 0, 1], [0, 1, 0], [0, 1, 1], [0, 0, 1],
[1, 1, 1]]
for pos in range(0, len(seq)):
if seq[pos] != mature[pos]:
error.add(pos)
subs, add = [], []
for e in error:
if e < len(seq) - 3:
subs.append([e, seq[e], mature[e]])
pattern, error_add = [], []
for e in range(len(seq) - 3, len(seq)):
if e in error:
pattern.append(1)
error_add.append(e)
else:
pattern.append(0)
for p in pattern_addition:
if pattern == p:
add = seq[error_add[0]:].replace("-", "")
break
if not add and error_add:
for e in error_add:
subs.append([e, seq[e], mature[e]])
return subs, add, make_cigar(seq, mature)
def create_iso(name, mir, seq, numsim, exp):
data = dict()
reads = dict()
full_read = list()
clean_read = list()
seen = set()
for mirna in mir[name]:
info = mir[name][mirna]
mirSeq = seq[info[0]:info[1] + 1]
for rand in range(int(numsim)):
# expression
e = 1
if exp:
trial = random.randint(1, 100)
p = random.randint(1, 50) / 50.0
e = numpy.random.negative_binomial(trial, p, 1)[0]
iso = realign.isomir()
randSeq, iso.start, iso.t5, iso.t3, iso.subs, iso.add = variation(
info, seq)
if randSeq in seen:
continue
seen.add(randSeq)
iso.end = iso.start + len(randSeq)
aln = realign.align(randSeq, seq[iso.start:iso.end])
iso.cigar = realign.make_cigar(aln[0], aln[1])
iso.mirna = mirna
query_name = "%s.%s.%s" % (mirna, iso.format_id("."), randSeq)
reads[query_name] = realign.hits()
reads[query_name].set_sequence(randSeq)
reads[query_name].counts = e
reads[query_name].set_precursor(name, iso)
full_read.extend(create_read(randSeq, e))
clean_read.append([
randSeq,
e,
])
# print [randSeq, mutLab, addTag, t5Lab, t3Lab, mirSeq]
# data[randSeq] = [exp, iso] # create real object used in code to generate GFF
write_fastq(full_read, full_fq)
write_collapse_fastq(clean_read, clean_fq)
gff = body.create(reads, "miRBase21", "sim1")
return gff
def create_iso(name, mir, seq, numsim, exp):
data = dict()
reads = dict()
full_read = list()
clean_read = list()
seen = set()
for mirna in mir[name]:
info = mir[name][mirna]
mirSeq = seq[info[0]:info[1] + 1]
for rand in range(int(numsim)):
# expression
e = 1
if exp:
trial = random.randint(1, 100)
p = random.randint(1, 50) / 50.0
e = numpy.random.negative_binomial(trial, p, 1)[0]
iso = realign.isomir()
randSeq, iso.start, iso.t5, iso.t3, iso.subs, iso.add = variation(info, seq)
if randSeq in seen:
continue
seen.add(randSeq)
iso.end = iso.start + len(randSeq)
aln = realign.align(randSeq, seq[iso.start:iso.end])
iso.cigar = realign.make_cigar(aln[0], aln[1])
iso.mirna = mirna
query_name = "%s.%s.%s" % (mirna, iso.format_id("."), randSeq)
reads[query_name] = realign.hits()
reads[query_name].set_sequence(randSeq)
reads[query_name].counts = e
reads[query_name].set_precursor(name, iso)
full_read.extend(create_read(randSeq, e))
clean_read.append([randSeq, e,])
# print([randSeq, mutLab, addTag, t5Lab, t3Lab, mirSeq])
# data[randSeq] = [exp, iso] # create real object used in code to generate GFF
write_fastq(full_read, full_fq)
write_collapse_fastq(clean_read, clean_fq)
gff = body.create(reads, "miRBase21", "sim1")
return gff
def tune(seq, precursor, start, cigar):
"""
The actual fn that will realign the sequence to find the nt changes
at 5', 3' sequence and nt variations.
Args:
*seq (str)*: sequence of the read.
*precursor (str)*: sequence of the precursor.
*start (int)*: start position of sequence on the precursor, +1.
*cigar (str)*: similar to SAM CIGAR attribute.
Returns:
*list* with:
subs (list): substitutions
add (list): nt added to the end
cigar (str): updated cigar
"""
end = len(seq)
if start < 0:
end = end + start
start = 0
if cigar:
seq, mature = cigar_correction(cigar, seq, precursor[start:])
else:
seq, mature, score, p, size = align(seq, precursor[start:start + end])
cigar = make_cigar(seq, mature)
if seq.startswith("-"):
seq = seq[1:]
if seq.endswith("-"):
seq = seq[:-1]
logger.debug("TUNE:: %s %s %s" % (cigar, seq, mature))
error = set()
for pos in range(0, len(seq)):
if seq[pos] != mature[pos]:
error.add(pos)
subs, add = [], []
prob = 0
add_position = []
for e in range(len(seq) - 1, len(seq) - 6, -1):
if e in error:
prob = 1
if prob == 1:
add.append(seq[e])
add_position.append(e)
if e not in error and prob == 0 and seq[e] in ["A", "T"]:
add.append(seq[e])
add_position.append(e)
continue
if e not in error:
if add:
add.pop()
add_position.pop()
if prob == 0:
add = []
add_position = []
break
for e in error:
if e not in add_position:
subs.append([e, seq[e], mature[e]])
logger.debug("TUNE:: %s %s" % (subs, add))
return subs, "".join(add), make_cigar(seq, mature)