def makeseq(chromlen, binlength, species):
x = cis.interval(genome=species)
x.chrom = []
x.start = []
x.end = []
# x.strand = []
inf = open(chromlen) #open("/mnt/data/static_libraries/chromLen/hg19.len")
#outf = open("hg19_%s_bin.bed"%(binlength),'w')
for line in inf:
start = 1
end = int(binlength)
while end < int(line.split()[1]):
# print start,end
#outf.write("\t".join([line.split()[0],str(start),str(end)])+"\n")
x.chrom.append(line.split()[0])
x.start.append(max(1, start - 100))
#print line
x.end.append(end + 100)
# x.end.append(min(end+100,int(line.split()[1])))
start += int(binlength)
end += int(binlength)
x.chrom.append(line.split()[0])
x.start.append(start)
x.end.append(int(line.split()[1]))
x.getSequence()
return x
def makeseq(bedfile,species):
x = cis.interval(genome=species)
x.chrom = []
x.start = []
x.end = []
# x.strand = []
inf = open(bedfile)#open("/mnt/data/static_libraries/chromLen/hg19.len")
#outf = open("hg19_%s_bin.bed"%(binlength),'w')
for line in inf:
#start = 1
#end = int(binlength)
#while end < int(line.split()[1]):
# print start,end
#outf.write("\t".join([line.split()[0],str(start),str(end)])+"\n")
x.chrom.append(line.split()[0])
x.start.append(int(line.split()[1]))
x.end.append(int(line.split()[2]))
# start += int(binlength)
# end += int(binlength)
#x.chrom.append(line.split()[0])
#x.start.append(start)
#x.end.append(int(line.split()[1]))
x.getSequence()
return x
def getsignal(inputfile,
outputfile,
BGmatrix,
pcut,
ncut,
pspan,
fetch_length=100,
gen='hg19'):
# p=BwIO(pcut)
# chrom_len = {}
# for i in p.chromosomeTree['nodes']:
# chrom_len[i['key']] = i['chromSize']
pcutbw = BigWigFile(open(pcut, 'rb'))
ncutbw = BigWigFile(open(ncut, 'rb'))
inf = open(inputfile)
testll = inf.readline().split()
ml = int(testll[2]) - int(testll[1])
pspan = pspan - ml / 2
inf.seek(0)
X = c.interval(genome=gen)
X.chrom, X.start, X.end, X.val = [], [], [], []
pBG, nBG = readBG(BGmatrix)
for line in inf:
ll = line.split()
# if not chrom_len.has_key(ll[0]):
# continue
X.chrom.append(ll[0])
X.start.append(int(ll[1]) - pspan - 3 + 1)
X.end.append(int(ll[2]) + pspan + 3 + 1)
X.val.append(ll[5])
inf.close()
X.getSequence()
outf = open(outputfile, 'w')
for i, elem in enumerate(X.seq):
pchrm = X.chrom[i]
pstart = X.start[i] - 1 + 3 + pspan
pend = X.end[i] - 1 - 3 - pspan
seq = X.seq[i]
strand = X.val[i]
pll = [pchrm, pstart, pend, strand]
pout = make_cut(pcutbw, pll, pspan, fetch_length)
nout = make_cut(ncutbw, pll, pspan, fetch_length)
if pll[3] == "-":
pout, nout = nout, pout
if pout == 'NA':
continue
if 'N' in seq.upper():
continue
#print 1
pseq = seq[:-1]
nseq = seq[1:]
p = []
n = []
for k in range(len(pseq) + 1 - 6):
p.append(pBG[pseq[k:k + 6].upper()])
n.append(nBG[nseq[k:k + 6].upper()])
if strand != '-':
pbglist = p
nbglist = n
else:
pbglist = n[::-1]
nbglist = p[::-1]
#print nbglist
newll = [pchrm, pstart, pend, strand] + pout + nout + pbglist + nbglist
#print len(pout),len(nout),len(pbglist),len(nbglist),len(newll)
outf.write("\t".join(map(str, newll)) + "\n")
outf.close()
def profile(inputfile,outputfile,pattern,strand):
x = cis.interval(genome='hg19')
x.chrom=[]
x.start=[]
x.end=[]
inf = open(inputfile)
for line in inf:
ll = line.split()
x.chrom.append(ll[0])
x.start.append(int(ll[1])-30)
x.end.append(int(ll[1])+2+31)
x.getSequence()
scores = []
for i in range(len(x.start)):
score = []
s = string.upper(x.seq[i])
if strand == "+":
for j in range(len(s)-2):
if s[j] == "A":
p1=0
elif s[j] == "C":
p1=1
elif s[j] == "G":
p1=2
elif s[j] == "T":
p1=3
else:
break
if s[j+1] == "A":
p2=0
elif s[j+1] == "C":
p2=1
elif s[j+1] == "G":
p2=2
elif s[j+1] == "T":
p2=3
else:
break
prob = pattern[p1][p2]
score.append(prob)
elif strand == "-":
for j in range(1,len(s)-1):
if s[j+1]=="A":
p1=3
elif s[j+1] == "C":
p1=2
elif s[j+1] == "G":
p1=1
elif s[j+1] == "T":
p1=0
else:
break
if s[j] == "A":
p2=3
elif s[j] == "C":
p2=2
elif s[j] == "G":
p2=1
elif s[j] == "T":
p2=0
else:
break
prob = pattern[p1][p2]
score.append(prob)
scores.append(score)
inf.close()
outf = open(outputfile,'w')
for score in scores:
if len(score)!=61:
continue
outf.write("\t".join(map(str,score))+"\n")
outf.close()
def profile(inputfile, outputfile, mode, strand):
A1 = [0] * 4 ##[ACGT]
C1 = [0] * 4
G1 = [0] * 4
T1 = [0] * 4
x = cis.interval(genome='hg19')
x.chrom = []
x.start = []
x.end = []
inf = open(inputfile)
for line in inf:
ll = line.split()
x.chrom.append(ll[0])
if mode == "peak":
x.start.append(int(ll[1]))
x.end.append(int(ll[2]) + 2)
elif mode == "cut":
if strand == "+":
x.start.append(int(ll[1]))
x.end.append(int(ll[1]) + 2)
else:
x.start.append(int(ll[2]))
x.end.append(int(ll[2]) + 2)
else:
print "mode wrong, only peak,cut availabe"
exit()
x.getSequence()
for i in range(len(x.start)):
s = string.upper(x.seq[i])
if strand == "+":
for j in range(len(s) - 1):
if s[j + 1] == "A":
p1 = 0
elif s[j + 1] == "C":
p1 = 1
elif s[j + 1] == "G":
p1 = 2
elif s[j + 1] == "T":
p1 = 3
else:
continue
if s[j] == "A":
A1[p1] += 1
elif s[j] == "C":
C1[p1] += 1
elif s[j] == "G":
G1[p1] += 1
elif s[j] == "T":
T1[p1] += 1
else:
continue
elif strand == "-":
s = s[::-1]
for j in range(len(s) - 1):
if s[j + 1] == "A":
p1 = 3
elif s[j + 1] == "C":
p1 = 2
elif s[j + 1] == "G":
p1 = 1
elif s[j + 1] == "T":
p1 = 0
else:
continue
if s[j] == "A":
T1[p1] += 1
elif s[j] == "C":
G1[p1] += 1
elif s[j] == "G":
C1[p1] += 1
elif s[j] == "T":
A1[p1] += 1
else:
continue
else:
print "strand only + and - "
exit()
inf.close()
outf = open(outputfile, 'w')
# outf.write("\t".join(['P','A','C','G','T'])+"\n")
# outf.write("\t".join(map(str,['A']+A1))+"\n")
# outf.write("\t".join(map(str,['C']+C1))+"\n")
# outf.write("\t".join(map(str,['G']+G1))+"\n")
# outf.write("\t".join(map(str,['T']+T1))+"\n")
outf.write("\t".join(map(str, A1)) + "\n")
outf.write("\t".join(map(str, C1)) + "\n")
outf.write("\t".join(map(str, G1)) + "\n")
outf.write("\t".join(map(str, T1)) + "\n")
# outf.write("total\t"+str(sum(A1)+sum(C1)+sum(G1)+sum(T1))+"\n")
outf.close()
def read_file(fp, mintags=0, maxtags=10, maxlines=100, select=100):
"""
Parse data
"""
# TODO read in more general format
if 1:
CHR, START, END, NAME, SCORE, STRAND, COUNT50P, COUNT50M = 0, 1, 2, 3, 4, 5, 6, 7
CUT_START = 14
CUT_END = 414
else:
CHR, START, END, NAME, SCORE, COUNT50P, COUNT50M = 0, 1, 2, 3, 4, 5, 6
STRAND = None
CUT_START = 13
CUT_END = 413
k = 0
X = c.interval(genome='hg19')
X.chrom, X.start, X.end, X.strand, X.name, X.val = [], [], [], [], [], []
cuts_pos = []
cuts_neg = []
mid = int(0.5 * (CUT_END - CUT_START))
startoffset = (mid - int(0.5 * select))
endoffset = (mid - int(0.5 * select) + select)
for elem in fp.readlines():
if elem[0:3] == 'chr':
f = elem.split()
chr, start, end, seq, motifscore, n50p, n50m = f[CHR], int(
f[START]), int(f[END]), f[NAME], float(f[SCORE]), float(
f[COUNT50P]), float(f[COUNT50M])
if STRAND:
strand = f[STRAND]
else:
strand = '+'
if (strand == '+') and (n50p + n50m >= mintags) and (n50p + n50m <
maxtags):
## Shawn : only + motif included
X.chrom.append(f[0])
#X.start.append( start - (CUT_END-CUT_START)/2 )
#X.end.append( start + (CUT_END-CUT_START)/2 )
X.start.append(start - (CUT_END - CUT_START) / 2 + startoffset)
X.end.append(start - (CUT_END - CUT_START) / 2 + endoffset)
X.strand.append(strand)
X.name.append(seq)
X.val.append((motifscore, n50p + n50m))
k += 1
poscut = [float(z) for z in f[CUT_START:CUT_END]]
negcut = [
float(z) for z in f[CUT_END:2 * CUT_END + 2 - CUT_START]
]
##Shawn : negcut = [ float(z) for z in f[ CUT_END: 2*CUT_END-CUT_START ] ]
cuts_pos.append(poscut[startoffset:endoffset])
cuts_neg.append(negcut[startoffset:endoffset])
if k == maxlines:
break
X.getSequence()
#for i,elem in enumerate( X.seq ):
# print X.name[i], elem[ (CUT_END-CUT_START)/2: (CUT_END-CUT_START)/2 + 10 ], X.strand[i], '\t'.join( [ '%3.1f' % x for x in cuts_pos[i][ (CUT_END-CUT_START)/2: (CUT_END-CUT_START)/2 + 10 ] ] )
#X.seq[i] = elem[ (CUT_END-CUT_START)/2: (CUT_END-CUT_START)/2 + 10 ]
#cuts_pos[i] = cuts_pos[i][ (CUT_END-CUT_START)/2: (CUT_END-CUT_START)/2 + 10 ]
#return
return X, numpy.array(cuts_pos), numpy.array(cuts_neg)
def getsignal(inputfile,outputfile,BGmatrix,pcut,ncut,pspan,fetch_length=100,gen='hg19'):
p=BwIO(pcut)
chrom_len = {}
for i in p.chromosomeTree['nodes']:
chrom_len[i['key']] = i['chromSize']
pcutbw = BigWigFile(open(pcut, 'rb'))
ncutbw = BigWigFile(open(ncut, 'rb'))
inf = open(inputfile)
pp=[]
pm=[]
X = c.interval(genome=gen)
X.chrom,X.start,X.end,X.val = [],[],[],[]
pBG,nBG = readBG(BGmatrix)
for line in inf:
ll = line.split()
if not chrom_len.has_key(ll[0]):
continue
pout = make_cut(pcutbw,ll,pspan,fetch_length)
nout = make_cut(ncutbw,ll,pspan,fetch_length)
if ll[5] == "-":
pout,nout = nout,pout
if pout == 'NA':
continue
#print len(pout),len(nout),ll[:3]
pp.append(pout)
pm.append(nout)
X.chrom.append(ll[0])
X.start.append(int(ll[1])-pspan -3 + 1)
X.end.append(int(ll[2]) + pspan +3 + 1)
X.val.append(ll[5])
#total[ ( flength - span ) : ( flength + int(ll[2]) - int(ll[1]) + span ) ]
meanp = apply_mean(pp)
meanm = apply_mean(pm)
X.getSequence()
pbglist = []
nbglist = []
for i,elem in enumerate(X.seq):
seq = X.seq[i]
strand = X.val[i]
if 'N' in seq.upper():
continue
pseq = seq[:-1]
nseq = seq[1:]
#if 'N' in pseq or 'N' in nseq:
# continue
p=[]
n=[]
for k in range(len(pseq) +1 - 6):
p.append(pBG[pseq[k:k+6].upper()])
n.append(nBG[nseq[k:k+6].upper()])
if strand != '-':
pbglist.append(p)
nbglist.append(n)
else:
pbglist.append(n[::-1])
nbglist.append(p[::-1])
#print nbglist
meanpbglist = apply_mean(pbglist)
meanmbglist = apply_mean(nbglist)
plot_template(meanp,meanm,meanpbglist,meanmbglist,outputfile)
def make_template(data, flank, pflank, topmotif, out, pbw, mbw, bgmatrix, gen):
w_plus_H = BigWigFile(open(pbw, 'rb'))
w_minus_H = BigWigFile(open(mbw, 'rb'))
i = 0
templatelist = []
pp = []
pm = []
inf = open(data)
l1st = inf.readline().split()
ml = int(l1st[2]) - int(l1st[1])
inf.seek(0)
for line in inf:
#if i >= topmotif:
# break
ll = line.split()
templatelist.append(ll)
inf.close()
templatelist.sort(key=lambda x: float(x[4]), reverse=True)
### for cut sitepro
for ll in templatelist:
p_sum = list(
w_plus_H.summarize(ll[0],
int(ll[1]) - flank,
int(ll[1]) + flank, 2 * flank).sum_data)
m_sum = list(
w_minus_H.summarize(ll[0],
int(ll[1]) - flank,
int(ll[1]) + flank, 2 * flank).sum_data)
if ll[5] == "+":
pp.append(p_sum[(flank + 1 + ml / 2 - pflank):(flank + 1 + ml / 2 +
pflank)])
pm.append(m_sum[(flank + 1 + ml / 2 - pflank):(flank + 1 + ml / 2 +
pflank)])
if ll[5] == '-':
pm.append(p_sum[::-1][(flank + 1 + ml / 2 - 1 - ml -
pflank):(flank + 1 + ml / 2 - 1 - ml +
pflank)])
pp.append(m_sum[::-1][(flank + 1 + ml / 2 - 1 - ml -
pflank):(flank + 1 + ml / 2 - 1 - ml +
pflank)])
print pp
print pm
meanp = apply_mean(pp)
meanm = apply_mean(pm)
allsum = sum(meanp) + sum(meanm)
P = []
M = []
for i in range(len(meanp)):
P.append(meanp[i]) #/allsum)
M.append(meanm[i]) #/allsum)
### for seqbias bg
pBG = {}
nBG = {}
inf = open(bgmatrix)
for line in inf:
ll = line.split()
name = ll[0]
pBG[name] = float(ll[1])
nBG[name] = float(ll[2])
inf.close()
X = c.interval(genome=gen)
X.chrom, X.start, X.end, X.val = [], [], [], []
for ll in templatelist:
X.chrom.append(ll[0])
X.start.append(int(ll[1]) + 1 - flank)
X.end.append(int(ll[1]) + 1 + flank)
X.val.append(ll[5])
X.getSequence()
pbglist = []
nbglist = []
for i, elem in enumerate(X.seq):
seq = X.seq[i]
strand = X.val[i]
if strand != '+' or 'N' in seq or 'n' in seq:
continue
pseq = seq[(flank + 1 + ml / 2 - pflank - 3):(flank + 1 + ml / 2 +
pflank + 2)]
nseq = seq[(flank + 1 + ml / 2 - pflank - 2):(flank + 1 + ml / 2 +
pflank + 3)]
#if 'N' in pseq or 'N' in nseq:
# continue
p = []
n = []
for k in range(len(pseq) + 1 - 6):
p.append(pBG[pseq[k:k + 6].upper()])
n.append(nBG[nseq[k:k + 6].upper()])
pbglist.append(p)
nbglist.append(n)
#print pbglist
#print nbglist
meanpbglist = apply_mean(pbglist)
meanmbglist = apply_mean(nbglist)
allsum = sum(meanpbglist) + sum(meanmbglist)
Plusbg = []
Minusbg = []
for i in range(len(meanpbglist)):
Plusbg.append(meanpbglist[i]) #/allsum)
Minusbg.append(meanmbglist[i]) #/allsum)
plot_template(P, M, Plusbg, Minusbg, out)