def scan_sequence(seq,
seq_gc_bin,
motifs,
nreport,
scan_rc,
motifs_meanstd=None,
zscore=False):
ret = []
# scan for motifs
for motif, cutoff in motifs:
if cutoff is None:
ret.append([])
else:
if zscore:
m_mean, m_std = motifs_meanstd[seq_gc_bin][motif.id]
result = pwmscan(seq, motif.logodds, motif.pwm_min_score(),
nreport, scan_rc)
result = [[(row[0] - m_mean) / m_std, row[1], row[2]]
for row in result]
result = [row for row in result if row[0] >= cutoff]
else:
result = pwmscan(seq, motif.logodds, cutoff, nreport, scan_rc)
if cutoff <= motif.pwm_min_score() and len(result) == 0:
result = [[motif.pwm_min_score(), 0, 1]] * nreport
ret.append(result)
return ret
def pwm_scan_to_gff(self,
fa,
gfffile,
cutoff=0.9,
nreport=50,
scan_rc=True,
append=False):
if append:
out = open(gfffile, "a")
else:
out = open(gfffile, "w")
c = self.pwm_min_score() + (self.pwm_max_score() -
self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {-1: "-", "-1": "-", "-": "-", "1": "+", 1: "+", "+": "+"}
gff_line = ("{}\tpwmscan\tmisc_feature\t{}\t{}\t{}\t{}\t.\t"
"motif_name \"{}\" ; motif_instance \"{}\"\n")
for name, seq in fa.items():
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score, pos, strand in result:
out.write(
gff_line.format(name, pos, pos + len(pwm), score,
strandmap[strand], self.id,
seq[pos:pos + len(pwm)]))
out.close()
def pwm_scan_to_gff(self, fa, gfffile, cutoff=0.9, nreport=50, scan_rc=True, append=False):
if append:
out = open(gfffile, "a")
else:
out = open(gfffile, "w")
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {-1:"-","-1":"-","-":"-","1":"+",1:"+","+":"+"}
gff_line = ("{}\tpwmscan\tmisc_feature\t{}\t{}\t{}\t{}\t.\t"
"motif_name \"{}\" ; motif_instance \"{}\"\n")
for name, seq in fa.items():
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score, pos, strand in result:
out.write(gff_line.format(
name,
pos,
pos + len(pwm),
score,
strandmap[strand],
self.id,
seq[pos:pos + len(pwm)]
))
out.close()
def pwm_scan(self, fa, cutoff=0.9, nreport=50, scan_rc=True):
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
matches = {}
for id, seq in fa.items():
matches[id] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for _,pos,strand in result:
matches[id].append(pos)
return matches
def pwm_scan_score(self, fa, cutoff=0, nreport=1, scan_rc=True):
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
matches = {}
for name, seq in fa.items():
matches[name] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score,_,_ in result:
matches[name].append(score)
return matches
def scan_sequence(seq, motifs, nreport, scan_rc):
ret = []
# scan for motifs
for motif, cutoff in motifs:
result = pwmscan(seq, motif.pwm, cutoff, nreport, scan_rc)
ret.append(result)
# return results
return ret
def scan_sequence(seq, motifs, nreport, scan_rc):
ret = []
# scan for motifs
for motif, cutoff in motifs:
result = pwmscan(seq, motif.pwm, cutoff, nreport, scan_rc)
ret.append(result)
# return results
return ret
def pwm_scan_all(self, fa, cutoff=0.9, nreport=50, scan_rc=True):
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
matches = {}
for name, seq in fa.items():
matches[name] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score,pos,strand in result:
matches[name].append((pos,score,strand))
return matches
def pwm_scan_score(self, fa, cutoff=0, nreport=1, scan_rc=True):
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {"+":"+",1:"+","1":"+","-":"-",-1:"-","-1":"-"}
matches = {}
for name, seq in fa.items():
matches[name] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score,pos,strand in result:
matches[name].append(score)
return matches
def pwm_scan_score(self, fa, cutoff=0, nreport=1, scan_rc=True):
from gimmemotifs.c_metrics import pwmscan
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {"+":"+",1:"+","1":"+","-":"-",-1:"-","-1":"-"}
matches = {}
for id, seq in fa.items():
matches[id] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score,pos,strand in result:
matches[id].append(score)
return matches
def pwm_scan_score(self, fa, cutoff=0, nreport=1, scan_rc=True):
c = self.pwm_min_score() + (self.pwm_max_score() -
self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {"+": "+", 1: "+", "1": "+", "-": "-", -1: "-", "-1": "-"}
matches = {}
for name, seq in fa.items():
matches[name] = []
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score, pos, strand in result:
matches[name].append(score)
return matches
def pwm_scan(self, fa, cutoff=0.9, scan_strand=None, nreport=50):
from gimmemotifs.c_metrics import pwmscan
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {"+":"+",1:"+","1":"+","-":"-",-1:"-","-1":"-"}
matches = {}
for id, seq in fa.items():
matches[id] = []
result = pwmscan(seq.upper(), pwm, c, nreport)
for score,pos,strand in result:
if not scan_strand or (scan_strand and strandmap[scan_strand] == strandmap[strand]):
matches[id].append(pos)
return matches
def pwm_scan_all(self, fa, cutoff=0.9, nreport=50, scan_rc=True):
from gimmemotifs.c_metrics import pwmscan
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {"+": "+", 1: "+", "1": "+", "-": "-", -1: "-", "-1": "-"}
matches = {}
for id, seq in fa.items():
matches[id] = []
# sys.stderr.write("{0}\t{1}\t{2}\t{3}\t{4}\t{5}".format(seq.upper(), pwm, c, nreport, scan_rc))
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score, pos, strand in result:
matches[id].append((pos, score, strand))
return matches
def scan_sequence(seq, motifs, nreport, scan_rc):
ret = []
# scan for motifs
for motif, cutoff in motifs:
if cutoff is None:
ret.append([])
else:
result = pwmscan(seq, motif.logodds, cutoff, nreport, scan_rc)
if cutoff <= motif.pwm_min_score() and len(result) == 0:
result = [[motif.pwm_min_score(), 0, 1]] * nreport
ret.append(result)
# return results
return ret
def seqcor(m1, m2, seq=None):
l1 = len(m1)
l2 = len(m2)
l = max(l1, l2)
if seq is None:
seq = RANDOM_SEQ
# Scan random sequence
result1 = pwmscan(seq, m1.pwm, m1.pwm_min_score(), len(seq), False, True)
result2 = pwmscan(seq, m2.pwm, m2.pwm_min_score(), len(seq), False, True)
result1 = np.array(result1)
result2 = np.array(result2)
# Return maximum correlation
c = []
for i in range(l1):
c.append(1 -
distance.correlation(result1[:L - l - i], result2[i:L - l]))
for i in range(l2):
c.append(1 -
distance.correlation(result1[i:L - l], result2[:L - l - i]))
return max(c)
def seqcor(m1,m2):
l1 = len(m1)
l2 = len(m2)
l = max(l1, l2)
# Create random sequence
nucs = []
L = 10 ** 4
for i in range(L):
nucs.append(random.choice(['A', 'C', 'T', 'G']))
random_seq = "".join(nucs)
# Scan random sequence
result1 = pwmscan(random_seq.upper(), m1.pwm, m1.pwm_min_score(), len(random_seq), False, True)
result2 = pwmscan(random_seq.upper(), m2.pwm, m2.pwm_min_score(), len(random_seq), False, True)
# Return maximum correlation
c = []
for i in range(l1):
c.append(1 - distance.correlation(result1[:L-l-i],result2[i:L-l]))
for i in range(l2):
c.append(1 - distance.correlation(result1[i:L-l],result2[:L-l-i]))
return max(c)
def seqcor(m1,m2):
l1 = len(m1)
l2 = len(m2)
l = max(l1, l2)
# Create random sequence
nucs = []
L = 10 ** 4
for i in range(L):
nucs.append(random.choice(['A', 'C', 'T', 'G']))
random_seq = "".join(nucs)
# Scan random sequence
result1 = pwmscan(random_seq.upper(), m1.pwm, m1.pwm_min_score(), len(random_seq), False, True)
result2 = pwmscan(random_seq.upper(), m2.pwm, m2.pwm_min_score(), len(random_seq), False, True)
# Return maximum correlation
c = []
for i in range(l1):
c.append(1 - distance.correlation(result1[:L-l-i],result2[i:L-l]))
for i in range(l2):
c.append(1 - distance.correlation(result1[i:L-l],result2[:L-l-i]))
return max(c)
def pwm_scan_to_gff(self, fa, gfffile, cutoff=0.9, rc=True, nreport=50, append=False):
from gimmemotifs.c_metrics import pwmscan
if append:
out = open(gfffile, "a")
else:
out = open(gfffile, "w")
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {-1:"-","-1":"-","-":"-","1":"+",1:"+","+":"+"}
for id, seq in fa.items():
result = pwmscan(seq.upper(), pwm, c, nreport)
for score, pos, strand in result:
out.write("%s\tpwmscan\tmisc_feature\t%s\t%s\t%s\t%s\t.\tmotif_name \"%s\" ; motif_instance \"%s\"\n" %
(id, pos, pos + len(pwm), score, strandmap[strand], self.id, seq[pos:pos + len(pwm)]))
out.close()
def pwm_scan_to_gff(self, fa, gfffile, cutoff=0.9, nreport=50, scan_rc=True, append=False):
#print "received", gfffile, cutoff, nreport, scan_rc, append
from gimmemotifs.c_metrics import pwmscan
if append:
out = open(gfffile, "a")
else:
out = open(gfffile, "w")
c = self.pwm_min_score() + (self.pwm_max_score() - self.pwm_min_score()) * cutoff
pwm = self.pwm
strandmap = {-1:"-","-1":"-","-":"-","1":"+",1:"+","+":"+"}
for id, seq in fa.items():
result = pwmscan(seq.upper(), pwm, c, nreport, scan_rc)
for score, pos, strand in result:
out.write("%s\tpwmscan\tmisc_feature\t%s\t%s\t%s\t%s\t.\tmotif_name \"%s\" ; motif_instance \"%s\"\n" %
(id, pos, pos + len(pwm), score, strandmap[strand], self.id, seq[pos:pos + len(pwm)]))
out.close()
parser = OptionParser()
parser.add_option("-p", "--pwmfile", dest="pwmfile", help="File with pwms", metavar="FILE")
parser.add_option("-i", "--inputfile", dest="inputfile", help="FASTA file with background sequences", metavar="FILE")
parser.add_option("-f", "--fpr", dest="fpr", help="Desired fpr", type="float", metavar="FLOAT")
(options, args) = parser.parse_args()
if not options.pwmfile or not options.inputfile or not options.fpr:
parser.print_help()
exit()
if options.fpr < 0 or options.fpr > 1:
print "Please specify a FPR between 0 and 1"
sys.exit()
f = Fasta(options.inputfile)
motifs = pwmfile_to_motifs(options.pwmfile)
print "Motif\tScore\tCutoff"
for motif in motifs:
pwm = motif.pwm
scores = []
min_score = motif.pwm_min_score()
for name,seq in f.items():
result = pwmscan(seq.upper(), pwm, min_score, 1, True)
score = result[0][0]
scores.append(score)
opt_score = scoreatpercentile(scores, 100 - (100 * options.fpr))
cutoff = (opt_score - min_score) / (motif.pwm_max_score() - min_score)
print "%s\t%s\t%s" % (motif.id, opt_score , cutoff)
if options.nreport:
nreport = int(options.nreport)
cutoff = float(options.cutoff)
motifs = pwmfile_to_motifs(options.pwmfile)
bed = options.bed
f = Fasta(inputfile)
strandmap = {-1:"-",1:"+"}
for (id,seq) in f.items():
for motif in motifs:
pwm = motif.pwm
c = motif.pwm_min_score() + (motif.pwm_max_score() - motif.pwm_min_score()) * cutoff
result = pwmscan(seq.upper(), pwm, c, nreport, options.scan_rc)
for (score, pos, strand) in result:
if bed:
first = id.split(" ")[0]
(chr,loc) = first.split(":")
if loc:
(start, end) = map(int, loc.split("-"))
print "%s\t%s\t%s\t%s" % (chr, start + pos, start + pos + len(pwm) , score)
else:
print "%s\t%s\t%s\t%s" % (id, pos, pos + len(pwm), score)
else:
print "%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\tmotif_name \"%s\" ; motif_instance \"%s\"" % (
id,
"pwmscan",
"misc_feature",
pos, pos + len(pwm) ,
if options.nreport:
nreport = int(options.nreport)
cutoff = float(options.cutoff)
motifs = pwmfile_to_motifs(options.pwmfile)
bed = options.bed
f = Fasta(inputfile)
strandmap = {-1:"-",1:"+"}
for (id,seq) in f.items():
for motif in motifs:
pwm = motif.pwm
c = motif.pwm_min_score() + (motif.pwm_max_score() - motif.pwm_min_score()) * cutoff
result = pwmscan(seq.upper(), pwm, c, nreport)
for (score, pos, strand) in result:
if bed:
first = id.split(" ")[0]
(chr,loc) = first.split(":")
if loc:
(start, end) = map(int, loc.split("-"))
print "%s\t%s\t%s\t%s" % (chr, start + pos, start + pos + len(pwm) , score)
else:
print "%s\t%s\t%s\t%s" % (id, pos, pos + len(pwm), score)
else:
print "%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\tmotif_name \"%s\" ; motif_instance \"%s\"" % (
id,
"pwmscan",
"misc_feature",
pos, pos + len(pwm) ,
dest="fpr",
help="Desired fpr",
type="float",
metavar="FLOAT")
(options, args) = parser.parse_args()
if not options.pwmfile or not options.inputfile or not options.fpr:
parser.print_help()
exit()
if options.fpr < 0 or options.fpr > 1:
print "Please specify a FPR between 0 and 1"
sys.exit()
f = Fasta(options.inputfile)
motifs = pwmfile_to_motifs(options.pwmfile)
print "Motif\tScore\tCutoff"
for motif in motifs:
pwm = motif.pwm
scores = []
min_score = motif.pwm_min_score()
for name, seq in f.items():
result = pwmscan(seq.upper(), pwm, min_score, 1, True)
score = result[0][0]
scores.append(score)
opt_score = scoreatpercentile(scores, 100 - (100 * options.fpr))
cutoff = (opt_score - min_score) / (motif.pwm_max_score() - min_score)
print "%s\t%s\t%s" % (motif.id, opt_score, cutoff)
