def check_MaxEntScan_5ss(pairs):
scores = []
for i in pairs:
score = [0, 0]
[ref, alt] = i
score[0] = maxent_fast.score5(ref, matrix=matrix5)
score[1] = maxent_fast.score5(alt, matrix=matrix5)
scores.append(score[1] - score[0])
return scores
def score_ss_seq(matrix5, matrix3, seq_ss, seq_ss_mut, ss):
if 'N' in seq_ss or 'N' in seq_ss_mut or 'n' in seq_ss or 'n' in seq_ss_mut:
return '.', '.', '.'
if ss == 'ss5':
score_wt = score5(seq_ss, matrix=matrix5)
score_mut = score5(seq_ss_mut, matrix=matrix5)
return str(round(score_wt, 3)), str(round(score_mut, 3)), str(
round((score_mut - score_wt) / score_wt, 3))
elif ss == 'ss3':
score_wt = score3(seq_ss, matrix=matrix3)
score_mut = score3(seq_ss_mut, matrix=matrix3)
return str(round(score_wt, 3)), str(round(score_mut, 3)), str(
round((score_mut - score_wt) / score_wt, 3))
else:
return '.', '.', '.'
def runMaxEntScan(sequence, donor=False, usePerl=False):
"""Run maxEntScan on the indicated sequence. Run score5.pl on candidate donor sequences,
score3.pl on candidate acceptor sequences. Return the score"""
if usePerl:
(fd, tmpfile) = tempfile.mkstemp()
fp = os.fdopen(fd, "w")
fp.write(sequence)
fp.close()
if donor:
pipe = subprocess.Popen([
"perl",
os.path.join(os.path.dirname(__file__), 'score5.pl'), tmpfile
],
stdout=subprocess.PIPE)
else:
pipe = subprocess.Popen([
"perl",
os.path.join(os.path.dirname(__file__), 'score3.pl'), tmpfile
],
stdout=subprocess.PIPE)
result = pipe.stdout.read()
entScore = re.findall("[+-]?\d+(?:\.\d+)?", str(result))
os.remove(tmpfile)
return (float(entScore[0]))
else:
# use the fastest available maxentpy implementation
# we round to two decimal places to match the perl script's output
if donor:
return round(maxent_fast.score5(str(sequence), matrix=matrix5), 2)
else:
return round(maxent_fast.score3(str(sequence), matrix=matrix3), 2)
def compute_mes(interval, matrix5, matrix3, genome):
genome = pysam.FastaFile(genome)
# 5ss 3bases in exon and 6 bases in intron
# 3ss 20 bases in intron and 3 bases in exon
if interval['strand'] == '+':
seq5 = genome.fetch(interval.chrom, interval.end - 3,
interval.end + 6).upper()
seq3 = genome.fetch(interval.chrom, interval.start - 20,
interval.start + 3).upper()
else:
seq5 = reverse_complement(
genome.fetch(interval.chrom, interval.start - 6,
interval.start + 3).upper())
seq3 = reverse_complement(
genome.fetch(interval.chrom, interval.end - 3,
interval.end + 20).upper())
name_format_str = '{seq5}:{mes5}|{seq3}:{mes3}'
if set(seq5).issubset('ACGT') and set(seq3).issubset('ACGT'):
mes5 = maxent_fast.score5(seq5, matrix=matrix5)
mes3 = maxent_fast.score3(seq3, matrix=matrix3)
interval['seq5'] = seq5
interval['mes5'] = mes5
interval['seq3'] = seq3
interval['mes3'] = mes3
else:
interval['seq5'] = seq5
interval['mes5'] = 'NA'
interval['seq3'] = seq3
interval['mes3'] = 'NA'
return interval
def runMaxEntScan(sequence, donor=False, usePerl=False):
"""Run maxEntScan on the indicated sequence. Run score5.pl on candidate donor sequences,
score3.pl on candidate acceptor sequences. Return the score"""
if usePerl:
(fd, tmpfile) = tempfile.mkstemp()
fp = os.fdopen(fd, "w")
fp.write(sequence)
fp.close()
if donor:
pipe = subprocess.Popen(["perl", os.path.join(os.path.dirname(__file__), 'score5.pl'), tmpfile], stdout=subprocess.PIPE)
else:
pipe = subprocess.Popen(["perl", os.path.join(os.path.dirname(__file__), 'score3.pl'), tmpfile], stdout=subprocess.PIPE)
result = pipe.stdout.read()
entScore = re.findall("[+-]?\d+(?:\.\d+)?", str(result))
os.remove(tmpfile)
return(float(entScore[0]))
else:
# use the fastest available maxentpy implementation
# we round to two decimal places to match the perl script's output
if donor:
return round(maxent_fast.score5(str(sequence), matrix=matrix5), 2)
else:
return round(maxent_fast.score3(str(sequence), matrix=matrix3), 2)
def compute_five_score(dna):
if len(dna) != 9:
print("ERROR INCORRECT LENGTH: %s" % dna)
return maxent_fast.score5(dna, matrix=mat5)