def read_dir(dir,mark,loc,ext,*spl):
try:
direct=os.popen('ls '+dir +'/*'+mark+'*'+ext)
except:
direct=os.popen('ls '+dir +'/'+mark+'*'+ext)
data=[]
files=[]
for fl in direct.readlines():
if loc in fl:
#print fl
if len(spl)>0:
tmpdata=read.read_dat(fl[:-1],str(spl[0]))
else:
tmpdata=read.read_dat(fl[:-1])
data.append(tmpdata)
#location=fl[::-1].index('/')
libst=['A','HS','E']
ind=0
run=True
while run and ind<len(libst):
try:
Ai=fl.index(libst[ind])
run=False
Aend=Ai+fl[Ai:].index('.')
except:
ind+=1
fl=read.tabless(fl)
files.append(fl[Ai:Aend])
return files, data
def read_dir(dir,mark,loc,ext,*spl):
try:
direct=os.popen('ls '+dir +'/*'+mark+'*'+ext)
except:
direct=os.popen('ls '+dir +'/'+mark+'*'+ext)
data=[]
files=[]
for fl in direct.readlines():
if loc in fl:
#print fl
if len(spl)>0:
tmpdata=read.read_dat(fl[:-1],str(spl[0]))
else:
tmpdata=read.read_dat(fl[:-1])
data.append(tmpdata)
#location=fl[::-1].index('/')
libst=['A','HS','E']
ind=0
run=True
while run and ind<len(libst):
try:
Ai=fl.index(libst[ind])
run=False
Aend=Ai+fl[Ai:].index('.')
except:
ind+=1
fl=read.tabless(fl)
files.append(fl[Ai:Aend])
return files, data
def readall_bedintersect(dir,mark,*flagstat):
direct=os.popen('ls '+dir +'/*'+mark+'*.bed')
data=[]
files=[]
for fl in direct.readlines():
fl=fl.strip()
tmpdata=read.read_bed_intersect(fl)
data.append(tmpdata)
ind=0
run=True
fl=fl[-fl[::-1].index('/'):fl.index('.')]
print fl
files.append(fl)
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
for fl in range(len(files)):
bool=False
for i in normalization:
if files[fl] in i[0]:
data[fl]=norm_bed(data[fl],i[-1])
bool=True
break
if not bool:
print files[fl], 'not found'
return files, data
def readall_bedintersect(dir,mark,*flagstat):
direct=os.popen('ls '+dir +'/*'+mark+'*.bed')
data=[]
files=[]
for fl in direct.readlines():
fl=fl.strip()
tmpdata=read.read_bed_intersect(fl)
data.append(tmpdata)
ind=0
run=True
fl=fl[-fl[::-1].index('/'):fl.index('.')]
print fl
files.append(fl)
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
for fl in range(len(files)):
bool=False
for i in normalization:
if files[fl] in i[0]:
data[fl]=norm_bed(data[fl],i[-1])
bool=True
break
if not bool:
print files[fl], 'not found'
return files, data
def CPG_RPKM(RPKM, CPG, lim):
CGI=[]
cpg=read.read_dat(CPG,'\t')
for i in cpg:
CGI.append(i[0])
genes=RPKM[0][1:]
libs=[]
mat=[]
for i in RPKM[1:]:
mat.append(i[1:])
libs.append(i[0])
mat=analyse.data2arr(mat)
genes=np.array(genes)
allave=[]
cgiave=[]
print len(mat)
for i in range(len(mat)):
gntmp=genes[mat[i,:]>lim]
tmp=mat[i,:][mat[i,:]>lim]
allave.append(np.mean(tmp))
cgirpkm=[]
for j in xrange(len(gntmp)):
if gntmp[j] in CGI:
cgirpkm.append(tmp[j])
cgiave.append(np.mean(cgirpkm))
print len(gntmp), len(cgirpkm)
allave=np.array(allave)
cgiave=np.array(cgiave)
return allave,cgiave, libs
def CPG_RPKM(RPKM, CPG, lim):
CGI = []
cpg = read.read_dat(CPG, '\t')
for i in cpg:
CGI.append(i[0])
genes = RPKM[0][1:]
libs = []
mat = []
for i in RPKM[1:]:
mat.append(i[1:])
libs.append(i[0])
mat = analyse.data2arr(mat)
genes = np.array(genes)
allave = []
cgiave = []
print len(mat)
for i in range(len(mat)):
gntmp = genes[mat[i, :] > lim]
tmp = mat[i, :][mat[i, :] > lim]
allave.append(np.mean(tmp))
cgirpkm = []
for j in xrange(len(gntmp)):
if gntmp[j] in CGI:
cgirpkm.append(tmp[j])
cgiave.append(np.mean(cgirpkm))
print len(gntmp), len(cgirpkm)
allave = np.array(allave)
cgiave = np.array(cgiave)
return allave, cgiave, libs
def gene_rpkm_compare(thersh, rpkmdata, rpkmfiles, libs, *enslist):
genelist = []
files = []
for ind in xrange(len(rpkmdata)):
files.append(rpkmfiles[ind][:rpkmfiles[ind].index('.')])
libs = read.read_dat(libs, '\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0]:
files[ind] = lib[1] + '-' + lib[-1]
break
pval = []
foldchange = []
for igene in range(len(rpkmdata[0])):
rpkmRT = []
rpkmnormal = []
for ind in xrange(len(rpkmdata)):
if 'RT' in files[ind]:
rpkmRT.append(rpkmdata[ind][igene][2])
elif 'Cancer' not in files[ind]:
rpkmnormal.append(rpkmdata[ind][igene][2])
rpkmRT = np.array(rpkmRT)
#rpkmRT=np.log(rpkmRT)
rpkmnormal = np.array(rpkmnormal)
#rpkmnormal=np.log(rpkmnormal)
a = ss.ks_2samp(rpkmRT, rpkmnormal)
if a[1] < thersh / np.float(20000) and np.mean(rpkmnormal) + np.mean(
rpkmRT) > 1.:
genelist.append(rpkmdata[0][igene][0])
#print rpkmdata[0][igene][0], 'pvalue RT=',(a[1])
pval.append(-np.log(a[1] * np.float(20000)))
#if np.mean(rpkmnormal)<0:
#print rpkmnormal
#rpkmRT=np.array(rpkmRT)
#rpkmnormal=np.log(rpkmnormal)
foldchange.append(np.log(
np.median(rpkmnormal) / np.median(rpkmRT)))
print 'RT sample:', len(rpkmRT), ', Normal sample:', len(rpkmnormal)
if len(enslist) > 0:
enslist = read.read_dat(enslist[0])
genelist2 = ens_genes(genelist, enslist)
#print genelist2
return genelist, foldchange, pval
def gene_rpkm_compare(thersh,rpkmdata,rpkmfiles,libs,*enslist):
genelist=[]
files=[]
for ind in xrange(len(rpkmdata)):
files.append(rpkmfiles[ind][:rpkmfiles[ind].index('.')])
libs=read.read_dat(libs,'\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0]:
files[ind]=lib[1]+'-'+lib[-1]
break
pval=[]
foldchange=[]
for igene in range(len(rpkmdata[0])):
rpkmRT=[]
rpkmnormal=[]
for ind in xrange(len(rpkmdata)):
if 'RT' in files[ind]:
rpkmRT.append(rpkmdata[ind][igene][2])
elif 'Cancer' not in files[ind]:
rpkmnormal.append(rpkmdata[ind][igene][2])
rpkmRT=np.array(rpkmRT)
#rpkmRT=np.log(rpkmRT)
rpkmnormal=np.array(rpkmnormal)
#rpkmnormal=np.log(rpkmnormal)
a=ss.ks_2samp(rpkmRT,rpkmnormal)
if a[1]<thersh/np.float(20000) and np.mean(rpkmnormal)+np.mean(rpkmRT)>1.:
genelist.append(rpkmdata[0][igene][0])
#print rpkmdata[0][igene][0], 'pvalue RT=',(a[1])
pval.append(-np.log(a[1]*np.float(20000)))
#if np.mean(rpkmnormal)<0:
#print rpkmnormal
#rpkmRT=np.array(rpkmRT)
#rpkmnormal=np.log(rpkmnormal)
foldchange.append(np.log(np.median(rpkmnormal)/np.median(rpkmRT)))
print 'RT sample:',len(rpkmRT), ', Normal sample:',len(rpkmnormal)
if len(enslist)>0:
enslist=read.read_dat(enslist[0])
genelist2=ens_genes(genelist,enslist)
#print genelist2
return genelist, foldchange, pval
def pl_PET(dr, table):
def pl(tmpdata):
ls = []
for i in tmpdata:
ls.append(float(i[1]))
return np.array(ls)
lable = [
'H3K4me1', 'H3K4me3', 'H3K9me3', 'H3K27me3', 'H3K36me3', 'H3K27ac'
]
tab = read.read_dat(table)
libs = []
for i in range(len(lable)):
for j in range(len(tab[0])):
if lable[i] in tab[0][j]:
libs.append([])
for k in range(1, 17):
libs[-1].append(tab[k][j])
num = len(libs[0])
data = []
ind = 0
dr = os.popen('ls ' + dr + '*' + 'A' + '*.dist')
for fl in dr.readlines():
for i in range(len(lable)):
fig = plt.figure(i)
plt.title(lable[i])
for lib in libs[i]:
#print lib, lable[i]
if lib in fl[:-1]:
#print lib,fl[:-1],lable[i]
tmpdata = pl(read.read_dat(fl[:-1]))
data.append(tmpdata)
plt.plot(range(50, 550),
tmpdata[50:550] / np.max(tmpdata[50:550]),
's-',
label=lib)
#libs.append(lib)
plt.legend(prop={'size': 14})
print i
for i in range(len(lable)):
fig = plt.figure(i)
fig.savefig(lable[i] + '.pdf', bbox_inches='tight')
return data
def readall_coverage(dir,mark,libs):
ext='.coverage'
try:
direct=os.popen('ls '+dir +'/*'+mark+'*'+ext)
except:
direct=os.popen('ls '+dir +'/'+mark+'*'+ext)
data=[]
files=[]
for fl in direct.readlines():
#print fl
tmpdata=read.read_dat(fl[:-1],'\t')
data.append(tmpdata)
files.append(fl.split('/')[-1][0:-len('.coverage')-1])
genes=[]
for i in data[0]:
genes.append(i[3])
mat=[]
#\genes=np.array(genes)
for i in data:
mat.append(np.zeros((len(genes)),np.float))
for j in i:
try:
mat[-1][genes.index(j[3])]+=np.float(j[4])
#print j[2]
except:
pass
mat=np.array(mat)
print 'ending'
libs=read.read_dat(libs,'\t')
for i in libs:
for j in range(len(files)):
if i[-1] in files[j]:
files[j]=i[0]
if 'hg19v69_genes.TSS_2000.pc.' in files[j]:
ind=files[j].index('hg19v69_genes.TSS_2000.pc.')
files[j]=files[j][ind]+files[j][ind+len('hg19v69_genes.TSS_2000.pc.'):]
#break
return files,mat,genes
def readall_coverage(dir,mark,libs):
ext='.coverage'
try:
direct=os.popen('ls '+dir +'/*'+mark+'*'+ext)
except:
direct=os.popen('ls '+dir +'/'+mark+'*'+ext)
data=[]
files=[]
for fl in direct.readlines():
#print fl
tmpdata=read.read_dat(fl[:-1],'\t')
data.append(tmpdata)
files.append(fl.split('/')[-1][0:-len('.coverage')-1])
genes=[]
for i in data[0]:
genes.append(i[3])
mat=[]
#\genes=np.array(genes)
for i in data:
mat.append(np.zeros((len(genes)),np.float))
for j in i:
try:
mat[-1][genes.index(j[3])]+=np.float(j[4])
#print j[2]
except:
pass
mat=np.array(mat)
print 'ending'
libs=read.read_dat(libs,'\t')
for i in libs:
for j in range(len(files)):
if i[-1] in files[j]:
files[j]=i[0]
if 'hg19v69_genes.TSS_2000.pc.' in files[j]:
ind=files[j].index('hg19v69_genes.TSS_2000.pc.')
files[j]=files[j][ind]+files[j][ind+len('hg19v69_genes.TSS_2000.pc.'):]
#break
return files,mat,genes
def pl_PET(dr,table):
def pl(tmpdata):
ls=[]
for i in tmpdata:
ls.append(float(i[1]))
return np.array(ls)
lable=['H3K4me1','H3K4me3','H3K9me3','H3K27me3','H3K36me3','H3K27ac']
tab=read.read_dat(table)
libs=[]
for i in range(len(lable)):
for j in range(len(tab[0])):
if lable[i] in tab[0][j]:
libs.append([])
for k in range(1,17):
libs[-1].append(tab[k][j])
num=len(libs[0])
data=[]
ind=0
dr=os.popen('ls '+ dr +'*'+'A'+'*.dist')
for fl in dr.readlines():
for i in range(len(lable)):
fig=plt.figure(i)
plt.title(lable[i])
for lib in libs[i]:
#print lib, lable[i]
if lib in fl[:-1]:
#print lib,fl[:-1],lable[i]
tmpdata=pl(read.read_dat(fl[:-1]))
data.append(tmpdata)
plt.plot(range(50,550),tmpdata[50:550]/np.max(tmpdata[50:550]),'s-',label=lib)
#libs.append(lib)
plt.legend(prop={'size':14})
print i
for i in range(len(lable)):
fig=plt.figure(i)
fig.savefig(lable[i]+'.pdf', bbox_inches='tight')
return data
def write_QC1(fl,ordered,targetid,ref):
#fread=open(fl,'r')
lable=['H3K4me1','H3K4me3','H3K9me3','H3K27me3','H3K36me3','H3K27ac','Input DNA']
data=read.read_dat(fl)
num=len(ordered)/len(lable)
ordered=list(ordered)
ln=data[0]
i=0
for tmpi in range(1,len(data)):
tmpdata=data[tmpi]
libid=str(tmpdata[2])
if libid in ordered:
ind=ordered.index(libid)
tmpdata.append(lable[int(ind/float(num))])
idi=0
for idtmp in range(1,len(targetid)):
for libtmp in targetid[idtmp]:
if str(libid[-4:]) in str(libtmp):
idi=idtmp
tmp=targetid[idi][-1]
for itm in ref:
if tmp in itm[0]:
tmp=itm[-1]
if idi==0:
print 'aha'
tmpdata.append(str(tmp))
if float(tmpdata[5])<20000000:
tmpdata.append('failed')
else:
tmpdata.append('passed')
#data[tmpi]=tmpdata
fwrite=open('table.txt','w')
for i in ln:
print >> fwrite, str(i),
print '\n'
for tmpdata in data:
for i in tmpdata:
try:
print >> fwrite, float(i),'\t',
except:
print >> fwrite, str(i),'\t',
print >> fwrite, '\n',
fwrite.close()
return data
def readall_bed(dir,mark,libs,*flagstat):
direct=os.popen('ls '+dir +'/*'+mark+'*.coverage')
data=[]
files=[]
for fl in direct.readlines():
fl=fl.strip()
for alib in libs:
if alib in fl and '#' not in alib:
tmpdata=read.read_bed(fl)
data.append(tmpdata)
#location=fl[::-1].index('/')
libst=['A','HS','E']
ind=0
run=True
while run and ind<len(libst):
try:
Ai=fl.index(libst[ind])
run=False
Aend=Ai+fl[Ai:].index('.')
except:
ind+=1
Ai=fl.index('/')
Aend=len(fl)
#print libst[ind-1] , 'tryed';
fl = fl[Ai:Aend]
#print fl
if mark in fl:
fl=fl[fl.index(mark):]
fl=fl[:len(mark)+1+fl[len(mark)+1:].index('.')]
#print fl
files.append(fl.strip())
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
for fl in range(len(files)):
bool=False
for i in normalization:
if files[fl] in i[0]:
data[fl]=norm_bed(data[fl],i[-1])
#print i[-1]
bool=True
break
if not bool:
print files[fl], 'not found'
return files, data
def write_QC1(fl, ordered, targetid, ref):
# fread=open(fl,'r')
lable = ["H3K4me1", "H3K4me3", "H3K9me3", "H3K27me3", "H3K36me3", "H3K27ac", "Input DNA"]
data = read.read_dat(fl)
num = len(ordered) / len(lable)
ordered = list(ordered)
ln = data[0]
i = 0
for tmpi in range(1, len(data)):
tmpdata = data[tmpi]
libid = str(tmpdata[2])
if libid in ordered:
ind = ordered.index(libid)
tmpdata.append(lable[int(ind / float(num))])
idi = 0
for idtmp in range(1, len(targetid)):
for libtmp in targetid[idtmp]:
if str(libid[-4:]) in str(libtmp):
idi = idtmp
tmp = targetid[idi][-1]
for itm in ref:
if tmp in itm[0]:
tmp = itm[-1]
if idi == 0:
print "aha"
tmpdata.append(str(tmp))
if float(tmpdata[5]) < 20000000:
tmpdata.append("failed")
else:
tmpdata.append("passed")
# data[tmpi]=tmpdata
fwrite = open("table.txt", "w")
for i in ln:
print >> fwrite, str(i),
print "\n"
for tmpdata in data:
for i in tmpdata:
try:
print >> fwrite, float(i), "\t",
except:
print >> fwrite, str(i), "\t",
print >> fwrite, "\n",
fwrite.close()
return data
def readall_bed(dir,mark,libs,*flagstat):
direct=os.popen('ls '+dir +'/*'+mark+'*.coverage')
data=[]
files=[]
for fl in direct.readlines():
fl=fl.strip()
for alib in libs:
if alib in fl and '#' not in alib:
tmpdata=read.read_bed(fl)
data.append(tmpdata)
#location=fl[::-1].index('/')
libst=['A','HS','E']
ind=0
run=True
while run and ind<len(libst):
try:
Ai=fl.index(libst[ind])
run=False
Aend=Ai+fl[Ai:].index('.')
except:
ind+=1
Ai=fl.index('/')
Aend=len(fl)
#print libst[ind-1] , 'tryed';
fl = fl[Ai:Aend]
#print fl
if mark in fl:
fl=fl[fl.index(mark):]
fl=fl[:len(mark)+1+fl[len(mark)+1:].index('.')]
#print fl
files.append(fl.strip())
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
for fl in range(len(files)):
bool=False
for i in normalization:
if files[fl] in i[0]:
data[fl]=norm_bed(data[fl],i[-1])
#print i[-1]
bool=True
break
if not bool:
print files[fl], 'not found'
return files, data
def read_dist(dr):
lib=[]
dist=[]
direct=os.popen('ls '+dr +'/*.dist')
#print direct,direct[0], len(direct)
for fl in direct.readlines():
#print fl
data=read.read_dat(fl[:-1])
tmp=[]
for line in data:
tmp.append(float(line[-2]))
dist.append(np.array(tmp))
lib.append(fl[1+len(dr):len(dr)+7])
return lib,dist
def read_dist(dr):
lib=[]
dist=[]
direct=os.popen('ls '+dr +'/*.dist')
#print direct,direct[0], len(direct)
for fl in direct.readlines():
#print fl
data=read.read_dat(fl[:-1])
tmp=[]
for line in data:
tmp.append(float(line[-2]))
dist.append(np.array(tmp))
lib.append(fl[1+len(dr):len(dr)+7])
return lib,dist
def read_alldir(dir,ext,*include):
#h3k27files, allh3k27=read.read_alldir('rhabdoid/coverage/TSS_2000_all','coverage','H3K27me3')
data=[]
files=[]
spl='\t'
if len(include)>0:
include='*'+str(include[0])+'*'
else:
include='*'
direct=os.popen('ls '+dir +'/'+include+ext)
for fl in direct.readlines():
tmpdata=read.read_dat(fl[:-1],str(spl))
data.append(tmpdata)
files.append(fl[-fl[::-1].index('/'):][:-1])
return files, data
def read_alldir(dir,ext,*include):
#h3k27files, allh3k27=read.read_alldir('rhabdoid/coverage/TSS_2000_all','coverage','H3K27me3')
data=[]
files=[]
spl='\t'
if len(include)>0:
include='*'+str(include[0])+'*'
else:
include='*'
direct=os.popen('ls '+dir +'/'+include+ext)
for fl in direct.readlines():
tmpdata=read.read_dat(fl[:-1],str(spl))
data.append(tmpdata)
files.append(fl[-fl[::-1].index('/'):][:-1])
return files, data
def readall_genescore(dr,mark,lib,genes):
direct=os.popen('ls '+dr +'*'+mark+'*.coverage')
data=[]
mylib=[]
vs=[]
for i in xrange(len(lib)):
if '#' not in lib[i]:
if '-' in lib[i]:
mylib.append(lib[i].split('-')[0])
else:
mylib.append(lib[i])
data={}
ind=-1
for fl in direct.readlines():
ind+=1
for alib in mylib:
if alib in fl:
found=True
tmpdata=read.read_dat(fl[:-1],'\t')
data[alib]=tmpdata
break
mydata=[]
for alib in mylib:
#if '#' not in lib:
mydata.append(data[alib])
genenum=len(genes)
libnum=len(mylib)
mat=np.zeros((genenum,libnum),np.float)
for genei in xrange(genenum):
for libi in xrange(libnum):
found=False
for line in mydata[libi]:
if genes[genei] in line[3]:
mat[genei,libi]=np.float(line[-2])
found=True
break
#if found:
# print 'found'
#else:
# print genes[genei],mylib[libi]
return mat
def readall_genescore(dr,mark,lib,genes):
direct=os.popen('ls '+dr +'*'+mark+'*.coverage')
data=[]
mylib=[]
vs=[]
for i in xrange(len(lib)):
if '#' not in lib[i]:
if '-' in lib[i]:
mylib.append(lib[i].split('-')[0])
else:
mylib.append(lib[i])
data={}
ind=-1
for fl in direct.readlines():
ind+=1
for alib in mylib:
if alib in fl:
found=True
tmpdata=read.read_dat(fl[:-1],'\t')
data[alib]=tmpdata
break
mydata=[]
for alib in mylib:
#if '#' not in lib:
mydata.append(data[alib])
genenum=len(genes)
libnum=len(mylib)
mat=np.zeros((genenum,libnum),np.float)
for genei in xrange(genenum):
for libi in xrange(libnum):
found=False
for line in mydata[libi]:
if genes[genei] in line[3]:
mat[genei,libi]=np.float(line[-2])
found=True
break
#if found:
# print 'found'
#else:
# print genes[genei],mylib[libi]
return mat
def write_bed_loc(projlist, tresh, coverfile):
data = read.read_dat(coverfile, "\t")
f = open("out.bed", "w")
ind = 0
print len(data), len(data[0])
genes = []
for i in xrange(len(projlist)):
if abs(projlist[i]) > tresh:
ind += 1
stout = ""
genes.append(str(data[i][3]))
for j in range(3):
stout += str(data[i][j]) + "\t"
print >> f, stout[:-1]
print "num of peaks: ", ind
f.close()
return genes
def write_bed_loc(projlist,tresh,coverfile):
data=read.read_dat(coverfile,'\t')
f=open('out.bed','w')
ind=0
print len(data), len(data[0])
genes=[]
for i in xrange(len(projlist)):
if abs(projlist[i])>tresh:
ind+=1
stout=''
genes.append(str(data[i][3]))
for j in range(3):
stout+=str(data[i][j])+'\t'
print >> f, stout[:-1]
print 'num of peaks: ',ind
f.close()
return genes
def mk_cluster(dr,mark,loc,*perc):
#writes the tanle for clustering
header=[]
if len(perc)>0:
f=open('cluster-'+mark+'-'+loc+'-'+str(perc[0])+'.txt','w')
else:
f=open('cluster-'+mark+'-'+loc+'.txt','w')
direct=os.popen('ls '+dr +'/*'+mark+'*.coverage')
for fl in direct.readlines():
#print fl
libst=['A','HS','E']
ind=0
run=True
while run and ind<=len(libst):
try:
Ai=fl.index(libst[ind])
run=False
except:
ind+=1
Aend=Ai+fl[Ai:].index('.')
print >> f, '\t',fl[Ai:Aend],
header.append(fl[Ai:Aend])
print >> f, '\n',
direct=os.popen('ls '+dr +'/*'+mark+'*.coverage')
data=[]
for fl in direct.readlines():
if loc in fl:
tmpdata=read.read_dat(fl[:-1],'\t')
if len(tmpdata)<1:
print len(tmpdata),fl[:-1]
else:
data.append(tmpdata)
print len(data),len(data[0])
enrich=[]
lable=[]
table=[]
for line in range(len(data[0])):
table.append([])
#lable.append([])
for tmpdata in data:
#print (tmpdata[0])
table[-1].append(float(tmpdata[line][-2]))
table[-1]=np.array(table[-1])
tmp2=''
for i in range(4):
tmp2=tmp2+str(tmpdata[line][i])+'_'
tmp2=tmp2[:-1]+'_1'
lable.append(tmp2)
table=np.array(table)
print len(table[0,:]), len(table[:,0])
if len(perc)>0:
for row in range(len(table[0,:])):
tmp=np.percentile(table[row,:],float(perc[0]))
print tmp,
table[row,:][table[row,:]<tmp]=0.
enrich=[]
for i in range(len(table[:,0])):
enrich.append('')
for j in range(len(table[0,:])):
enrich[-1]+=str(table[i,j])+'\t'
for line in range(len(data[0])):
print >> f, lable[line],enrich[line]
#print lable[:3],enrich[:3]
f.close()
return lable,header,table
def read_wig(file,*flagstat):
elts = {}
f = open(file,'r')
if '.gz' in file:
f = gzip.open(file)
else:
f = open(file,'r')
s = f.readline()
print s
s = f.readline()
ind=0
cover1=int(s[:-1])
f.close()
if '.gz' in file:
f = gzip.open(file)
else:
f = open(file,'r')
count=0
data=f.read()
data=data.splitlines()
print 'reading done'
for s in data:
#count+=1
#if count >10000000:
# break
try:
cover=int(s)
if cover==cover1:
reg+=step
else:
begin=start+ind*step
end=begin+reg
start=end
ind=0
elts[chrom].append([begin,end,cover1])
cover1=cover
reg=step
#ind+=step
except:
ind=0
s = s.split()
s=map(str,s)
tmp=s[1]
chrom=tmp[6:]
tmp=s[2]
start=int(tmp[6:])
tmp=s[3]
step=int(tmp[5:])
reg=step
if not elts.has_key(chrom):
elts[chrom] = []
#s = f.readline()
f.close()
for key in elts.keys():
elts[key]=np.array(elts[key])
try:
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
fl=file[0:6]
print fl
bool=False
for i in normalization:
if fl in i[0]:
elts=norm_wig(elts,i[-1])
#print i[-1]
bool=True
break
if not bool:
print fl, 'not found'
except:
pass
return elts
import read, write, bedtools, analyse
import matplotlib.pyplot as plt
dirPD = str(sys.argv[1])
dirin = str(sys.argv[2])
fl = str(sys.argv[3])
fl = dirin + fl
mark = str(sys.argv[4])
sample1 = str(sys.argv[5])
sample2 = str(sys.argv[6])
field = int(sys.argv[7])
libsfile = str(sys.argv[8])
hugo = str(sys.argv[9])
libsdata = read.read_dat(libsfile, '\t')
genelist = []
genelist = read.read_dat(hugo)
if os.path.isfile(fl) and os.path.getsize(fl) > 0:
print fl, field, libsfile
nm, libsRNA, gene, genex, libs = analyse.heat_rna(fl, genelist, field,
libsfile)
rpkmmat = nm * 1.
plt.xlabel(sample1 + ' vs. ' + sample2 + ' rpkm')
plt.savefig(dirin + sample1 + '-' + sample2 + '-rpkm.pdf',
bbox_inches='tight')
print dirin + sample1 + '-' + sample2 + '-rpkm.pdf'
plt.close()
import os, sys
import commands
import read
import write
import numpy as np
tmp=sys.argv
fl=tmp[1]
table=read.read_dat("peaks.txt",'\t')
data=read.read_dat(fl,'\t')
mark=map(str,tmp[2].split('_'))
#mark=mark[mark.index('_')+1:]
mark=mark[-1]
vecs=['pcDNA','K4E','Y69H','D83V']
#print data[1]D83V
for row in table:
for tmp in row:
if "bwa-0." in tmp:
if tmp in data[1][5]:
#print data[1][5]
cellline = data[-1][-1]
cellline = cellline [:cellline.index('_')]
print cellline
vs='WT'
out='table'+'-'+vs+'_all_vectors.txt'
f=open(out,'a')
tmp=fl[fl.index('/')+1:]
rep=int(tmp[-len('2_peaks.xls'):-len('2_peaks.xls')+1])
dirPD=str(sys.argv[1]) dirin=str(sys.argv[2]) fl=str(sys.argv[3]) fl=dirin+fl mark=str(sys.argv[4]) sample1=str(sys.argv[5]) sample2=str(sys.argv[6]) field=int(sys.argv[7]) libsfile=str(sys.argv[8]) hugo=str(sys.argv[9]) libsdata=read.read_dat(libsfile,'\t') genelist=[] genelist=read.read_dat(hugo) if os.path.isfile(fl) and os.path.getsize(fl) > 0: print fl,field,libsfile nm,libsRNA,gene,genex,libs=analyse.heat_rna(fl,genelist,field,libsfile) rpkmmat=nm*1. plt.xlabel(sample1+' vs. '+sample2+' rpkm') plt.savefig(dirin+sample1+'-'+sample2+'-rpkm.pdf', bbox_inches='tight') print dirin+sample1+'-'+sample2+'-rpkm.pdf' plt.close() # try:
import os, sys
import commands
import read
import write
import numpy as np
import time
import subprocess
table = read.read_dat("peaks.txt")
data = table
cells = ["D83V", "K4E", "pcDNA", "WT", "Y69H"]
marks = ["H3K27ac", "H3K4me3", "H3K9me3", "V5", "input"]
ran = 0
for irow in xrange(1, len(data)):
row = data[irow]
vector = row[-3]
solution = row[-1][:-2]
mymark = ""
if "input" in row[-2]:
mymark = "input"
else:
mymark = row[-2]
if '1' in row[-4]:
rep = '1'
else:
rep = '2'
bam = row[5]
if 'WT' not in vector and 'input' not in mymark and 'H3K4' not in mymark:
for ind in xrange(-1, 4):
try:
ii = data[irow + ind][-3]
import sys, os, numpy as np
sys.path.append('/Users/ssaberim/epigenomics/code')
import read,write,bedtools,analyse
ID=str(sys.argv[1])
drrna=str(sys.argv[2])
sample=str(sys.argv[3])
outfile=str(sys.argv[4])
field=int(sys.argv[5])
libsfile=str(sys.argv[6])
mark='RNA'
libsdata=read.read_dat(libsfile,'\t')
indmark=libsdata[0].index(mark)
list1=[]
for i in range(1,len(libsdata)):
if libsdata[i][field] in sample and len(libsdata[i][field])>0:
list1.append(libsdata[i][indmark])
hugo="/Users/ssaberim/epigenomics/resources/list.genes2.txt"
hugo=read.read_dat(hugo)
#print hugo[0]
ID=read.read_dat(ID,'\t')
for i in range(len(ID)):
for j in range(len(hugo)):
def mk_cluster(dr, mark, loc, *perc):
#writes the tanle for clustering
header = []
if len(perc) > 0:
f = open('cluster-' + mark + '-' + loc + '-' + str(perc[0]) + '.txt',
'w')
else:
f = open('cluster-' + mark + '-' + loc + '.txt', 'w')
direct = os.popen('ls ' + dr + '/*' + mark + '*.coverage')
for fl in direct.readlines():
#print fl
libst = ['A', 'HS', 'E']
ind = 0
run = True
while run and ind <= len(libst):
try:
Ai = fl.index(libst[ind])
run = False
except:
ind += 1
Aend = Ai + fl[Ai:].index('.')
print >> f, '\t', fl[Ai:Aend],
header.append(fl[Ai:Aend])
print >> f, '\n',
direct = os.popen('ls ' + dr + '/*' + mark + '*.coverage')
data = []
for fl in direct.readlines():
if loc in fl:
tmpdata = read.read_dat(fl[:-1], '\t')
if len(tmpdata) < 1:
print len(tmpdata), fl[:-1]
else:
data.append(tmpdata)
print len(data), len(data[0])
enrich = []
lable = []
table = []
for line in range(len(data[0])):
table.append([])
#lable.append([])
for tmpdata in data:
#print (tmpdata[0])
table[-1].append(float(tmpdata[line][-2]))
table[-1] = np.array(table[-1])
tmp2 = ''
for i in range(4):
tmp2 = tmp2 + str(tmpdata[line][i]) + '_'
tmp2 = tmp2[:-1] + '_1'
lable.append(tmp2)
table = np.array(table)
print len(table[0, :]), len(table[:, 0])
if len(perc) > 0:
for row in range(len(table[0, :])):
tmp = np.percentile(table[row, :], float(perc[0]))
print tmp,
table[row, :][table[row, :] < tmp] = 0.
enrich = []
for i in range(len(table[:, 0])):
enrich.append('')
for j in range(len(table[0, :])):
enrich[-1] += str(table[i, j]) + '\t'
for line in range(len(data[0])):
print >> f, lable[line], enrich[line]
#print lable[:3],enrich[:3]
f.close()
return lable, header, table
def gene_rpkm_corr_compare(thegene,thersh,rpkmdata,rpkmfiles,libs,*enslist):
#usage:
#genescompared,genesfc,pvals=\
#analyse.gene_rpkm_corr_compare('ENSG00000108799',0.000001,rpkm,files,'rhabdoid/RNA-ALL-key.txt','resources/list.genes2.txt')
genelist=[]
rpkmgene=[]
geneimp=[]
pvals=[]
files=[]
for ind in xrange(len(rpkmdata)):
for i in rpkmdata[ind]:
if thegene in i[0]:
rpkmgene.append(i[2])
files.append(rpkmfiles[ind][:rpkmfiles[ind].index('.')])
break
libs=read.read_dat(libs,'\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0] :
files[ind]=lib[1]+'-'+lib[-1]
break
rpkmgeneRT=[]
rpkmgenenormal=[]
for ind in xrange(len(rpkmgene)):
#if 'Blood' not in files[ind]:
#print files[ind]
if 'RT' in files[ind]:
rpkmgeneRT.append(rpkmgene[ind])
else:
rpkmgenenormal.append(rpkmgene[ind])
rpkmgeneRT=np.array(rpkmgeneRT)
rpkmgenenormal=np.array(rpkmgenenormal)
print len(rpkmgeneRT), len(rpkmgenenormal)
for igene in range(len(rpkmdata[0])):
rpkmRT=[]
rpkmnormal=[]
for ind in xrange(len(rpkmdata)):
if 'Blood' not in files[ind]:
if 'RT' in files[ind] :
rpkmRT.append(rpkmdata[ind][igene][2])
else:
rpkmnormal.append(rpkmdata[ind][igene][2])
#print len(rpkmRT)
#print len(rpkmnormal), rpkmnormal[0]
rpkmRT=np.array(rpkmRT)
rpkmnormal=np.array(rpkmnormal)
a=ss.pearsonr(rpkmRT,rpkmgeneRT)
b=ss.pearsonr(rpkmnormal,rpkmgenenormal)
mn=np.mean(rpkmRT)/np.mean(rpkmnormal)
if b[1]<thersh and (a[1]/b[1])>thersh:
genelist.append(rpkmdata[0][igene][0])
if mn>1. or mn <1.:
#print rpkmdata[0][igene][0], 'correlation for normal=%.3f, pvalue normal=%.5f, pvalue RT=%.5f, foldchange=%.3f'%(b[0],b[1],a[1],mn)
geneimp.append(mn)
pvals.append(np.log(a[1]/b[1]))
return genelist , geneimp, pvals
import os, sys
import commands
import read
import numpy as np
import time
import subprocess
import write
table=read.read_dat("peaks.txt",'\t')
data=table
cells=["D83V","K4E", "pcDNA","WT","Y69H"]
marks=["H3K27ac","H3K4me3","H3K9me3","V5","input"]
ran=0
for irow in xrange(2,len(data)):
row=data[irow]
vector=row[-3]
solution=row[-1][:-2]
mymark=""
#print row[-2]
if "input" in row:
mymark="input"
else:
mymark=row[-2]
if '1' in row[-4]:
rep='1'
else:
rep='2'
bam=row[5]
if 'WT' not in vector and 'input' not in mymark and mymark in marks:
for ind in xrange(-1,4):
try:
ii=data[irow+ind][-3]
def gene_rpkm_corr_compare(thegene, thersh, rpkmdata, rpkmfiles, libs,
*enslist):
#usage:
#genescompared,genesfc,pvals=\
#analyse.gene_rpkm_corr_compare('ENSG00000108799',0.000001,rpkm,files,'rhabdoid/RNA-ALL-key.txt','resources/list.genes2.txt')
genelist = []
rpkmgene = []
geneimp = []
pvals = []
files = []
for ind in xrange(len(rpkmdata)):
for i in rpkmdata[ind]:
if thegene in i[0]:
rpkmgene.append(i[2])
files.append(rpkmfiles[ind][:rpkmfiles[ind].index('.')])
break
libs = read.read_dat(libs, '\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0]:
files[ind] = lib[1] + '-' + lib[-1]
break
rpkmgeneRT = []
rpkmgenenormal = []
for ind in xrange(len(rpkmgene)):
#if 'Blood' not in files[ind]:
#print files[ind]
if 'RT' in files[ind]:
rpkmgeneRT.append(rpkmgene[ind])
else:
rpkmgenenormal.append(rpkmgene[ind])
rpkmgeneRT = np.array(rpkmgeneRT)
rpkmgenenormal = np.array(rpkmgenenormal)
print len(rpkmgeneRT), len(rpkmgenenormal)
for igene in range(len(rpkmdata[0])):
rpkmRT = []
rpkmnormal = []
for ind in xrange(len(rpkmdata)):
if 'Blood' not in files[ind]:
if 'RT' in files[ind]:
rpkmRT.append(rpkmdata[ind][igene][2])
else:
rpkmnormal.append(rpkmdata[ind][igene][2])
#print len(rpkmRT)
#print len(rpkmnormal), rpkmnormal[0]
rpkmRT = np.array(rpkmRT)
rpkmnormal = np.array(rpkmnormal)
a = ss.pearsonr(rpkmRT, rpkmgeneRT)
b = ss.pearsonr(rpkmnormal, rpkmgenenormal)
mn = np.mean(rpkmRT) / np.mean(rpkmnormal)
if b[1] < thersh and (a[1] / b[1]) > thersh:
genelist.append(rpkmdata[0][igene][0])
if mn > 1. or mn < 1.:
#print rpkmdata[0][igene][0], 'correlation for normal=%.3f, pvalue normal=%.5f, pvalue RT=%.5f, foldchange=%.3f'%(b[0],b[1],a[1],mn)
geneimp.append(mn)
pvals.append(np.log(a[1] / b[1]))
return genelist, geneimp, pvals
def gene_rpkm_lineage(hugo, rpkmdata, rpkmfiles, allgenes, libs, *plotbool):
genelist = []
for i in range(len(hugo)):
genelist.append(hugo[i])
genelist = ens_genes(genelist, allgenes)
rpkmls = []
files = []
for gene in genelist:
rpkmls.append([])
for ind in xrange(len(rpkmdata)):
for i in rpkmdata[ind]:
if gene in i[0]:
rpkmls[-1].append(i[2])
if len(rpkmls) == 1:
files.append(
rpkmfiles[ind][:rpkmfiles[ind].index('.')])
break
files = np.array(files)
rpkmls = data2arr(rpkmls)
sortedinds = rpkmls[0].argsort()
files = files[sortedinds]
files = files[::-1]
for i in range(len(genelist)):
rpkmls[i] = rpkmls[i][sortedinds]
rpkmls[i] = rpkmls[i][::-1]
normalfiles = []
normalrpkms = []
RTfiles = []
RTrpkms = []
libs = read.read_dat(libs, '\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0]:
files[ind] = lib[1] + '-' + lib[-1]
break
for i in range(len(genelist)):
rpkmls[i] = np.log(rpkmls[i] + 0.001) / np.log(10)
for i in range(len(genelist)):
normalrpkms.append([])
RTrpkms.append([])
for ind in range(len(files)):
if 'Cancer' not in files[ind] and 'ES' not in files[ind]:
print files[ind]
normalfiles.append(files[ind])
for i in range(len(genelist)):
normalrpkms[i].append(rpkmls[i][ind])
elif 'RT' in files[ind]:
RTfiles.append(files[ind])
for i in range(len(genelist)):
RTrpkms[i].append(rpkmls[i][ind])
plt.figure()
for i in range(len(genelist)):
if i < 7:
plt.plot(xrange(len(list(files))), rpkmls[i], 's-', label=hugo[i])
else:
plt.plot(xrange(len(list(files))), rpkmls[i], '^-', label=hugo[i])
plt.legend(loc=1)
plt.xticks(range(len(files)), files)
plt.xticks(rotation=90)
plt.xticks(fontsize=8)
plt.ylabel('$log_{10}$ $RPKM$')
plt.grid()
plt.show()
def heatmap(rpkmls, hugo, txt):
rpkmls = data2arr(rpkmls)
norm, matrix, dist = analyse.all_corr(rpkmls.T)
plt.matshow((matrix), cmap='RdYlBu', vmax=1, vmin=-1)
plt.yticks(range(len(hugo)), hugo)
plt.xticks(range(len(hugo)), hugo)
plt.xticks(rotation=90)
plt.xlabel('Correlation-' + txt)
plt.colorbar()
#plt.matshow(np.log(pval),cmap='PuBu',vmax=-5,vmin=-25)
#plt.yticks(range(len(hugo)),hugo)
#plt.xticks(range(len(hugo)),hugo)
#plt.xlabel('P-Value $\log$ -'+txt)
#plt.xticks(rotation=90)
#plt.colorbar()
if len(plotbool) > 0:
if plotbool[0]:
heatmap(rpkmls, hugo, 'All Samples')
heatmap(normalrpkms, hugo, 'Normal Samples')
heatmap(RTrpkms, hugo, 'RT Samples')
for i in range(len(hugo)):
print hugo[i], ss.ks_2samp(normalrpkms[i],
RTrpkms[i])[1] * np.float(
len(allgenes)), 'KS'
print hugo[i], ss.ttest_ind(normalrpkms[i],
RTrpkms[i])[1] * np.float(
len(allgenes)), 'T-test'
#fig=plt.figure()
#plt.plot()
return rpkmls, files, data2arr(normalrpkms), np.array(
normalfiles), genelist
import os, sys
import commands
import read
import write
import numpy as np
tmp=sys.argv
fl=tmp[1]
#vs=str(tmp[3])
mark=map(str,tmp[2].split('-'))
mark=mark[1]
vs='input'
data=read.read_dat(fl)
out='table'+'-'+vs+'.txt'
f=open(out,'a')
tmp=fl[fl.index('/')+1:]
solution=tmp.split('_')[1]
rep=tmp.split('_')[2]
tmp = str(data[-1][-1])
tmp=tmp.split('_')
rank=tmp[2]
coverage=tmp[-1]
vec=tmp[0]
try:
print >> f, rep,mark,vec,solution,vs,coverage
except:
print vec
f.close()
def read_wig(file,*flagstat):
elts = {}
f = open(file,'r')
if '.gz' in file:
f = gzip.open(file)
else:
f = open(file,'r')
s = f.readline()
print s
s = f.readline()
ind=0
cover1=int(s[:-1])
f.close()
if '.gz' in file:
f = gzip.open(file)
else:
f = open(file,'r')
count=0
data=f.read()
data=data.splitlines()
print 'reading done'
for s in data:
#count+=1
#if count >10000000:
# break
try:
cover=int(s)
if cover==cover1:
reg+=step
else:
begin=start+ind*step
end=begin+reg
start=end
ind=0
elts[chrom].append([begin,end,cover1])
cover1=cover
reg=step
#ind+=step
except:
ind=0
s = s.split()
s=map(str,s)
tmp=s[1]
chrom=tmp[6:]
tmp=s[2]
start=int(tmp[6:])
tmp=s[3]
step=int(tmp[5:])
reg=step
if not elts.has_key(chrom):
elts[chrom] = []
#s = f.readline()
f.close()
for key in elts.keys():
elts[key]=np.array(elts[key])
try:
if len(flagstat)>0:
flagstat=str(flagstat[0])
normalization=read.read_dat(flagstat,'\t')
fl=file[0:6]
print fl
bool=False
for i in normalization:
if fl in i[0]:
elts=norm_wig(elts,i[-1])
#print i[-1]
bool=True
break
if not bool:
print fl, 'not found'
except:
pass
return elts
def heat_rna(fl,genelist,field,libfile):
data=read.read_dat(fl)
data=data[:-1]
if [''] in data:
data.delete([['']])
lib=[]
lib2=[]
gene=[]
for i in data:
try:
if i[2] not in gene:
gene.append(i[2])
except:
print i
libstable=read.read_dat(libfile,'\t')
for i in data:
try:
if i[0]+'-'+i[1] not in lib:
lib.append(i[0]+'-'+i[1])
lib2.append(i[0]+'-'+i[1])
if len(libfile)>0:
for ID in libstable:
if i[0] in ID:
lib2[-1]=ID[0]+'-'+ID[field]
break
except:
print i
pheno=[]
pnum=-1.5
hist=''
for i in lib:
j=i.split('-')
if j[1] != hist:
pnum+=1
pheno.append(pnum)
hist=j[1]
#print hist,pnum
mat=np.zeros((len(gene),len(lib)),np.float)
for i in data:
mat[gene.index(i[2]),lib.index(i[0]+'-'+i[1])]=np.float(i[3])
proj=np.dot(mat,pheno)
for i in xrange(len(proj)):
norm=np.sqrt(np.dot(mat[i,:],mat[i,:]))
if norm!=0.:
proj[i]/=norm
#proj=np.array(proj)
inds=sortedinds=proj.argsort()
mat=mat[inds]
gene=np.array(gene)
gene=gene[inds]
inds=np.any(mat != 0, axis=1)
mat=mat[inds]
gene=np.array(gene)
gene=gene[inds]
#for i in mat:
# if np.mean(abs(i))==0:
# print i
nm=norm_max(mat)
#for i in nm:
# if np.mean(abs(i))==0:
# print i
genex=gene
if len(genelist)>0:
lb=ens_genes(gene,genelist)
genex=lb
else:
return nm,lib2,gene,genex,lib
b=plt.matshow(nm,aspect='auto',cmap='RdYlBu')
if len(libfile)>0:
plt.xticks(range(len(lib2)),lib2)
else:
plt.xticks(range(len(lib)),lib)
plt.colorbar()
plt.xticks(rotation=90)
plt.yticks(range(len(genex)),genex)
plt.yticks(fontsize=8)
mytemplate(nm)
plt.xlabel(lb)
lb=fl.split('/')[-1][:-5]
return nm,lib2,gene,genex,lib
def heat_rna(fl, genelist, field, libfile):
data = read.read_dat(fl)
data = data[:-1]
if [''] in data:
data.delete([['']])
lib = []
lib2 = []
gene = []
for i in data:
try:
if i[2] not in gene:
gene.append(i[2])
except:
print i
libstable = read.read_dat(libfile, '\t')
for i in data:
try:
if i[0] + '-' + i[1] not in lib:
lib.append(i[0] + '-' + i[1])
lib2.append(i[0] + '-' + i[1])
if len(libfile) > 0:
for ID in libstable:
if i[0] in ID:
lib2[-1] = ID[0] + '-' + ID[field]
break
except:
print i
pheno = []
pnum = -1.5
hist = ''
for i in lib:
j = i.split('-')
if j[1] != hist:
pnum += 1
pheno.append(pnum)
hist = j[1]
#print hist,pnum
mat = np.zeros((len(gene), len(lib)), np.float)
for i in data:
mat[gene.index(i[2]), lib.index(i[0] + '-' + i[1])] = np.float(i[3])
proj = np.dot(mat, pheno)
for i in xrange(len(proj)):
norm = np.sqrt(np.dot(mat[i, :], mat[i, :]))
if norm != 0.:
proj[i] /= norm
#proj=np.array(proj)
inds = sortedinds = proj.argsort()
mat = mat[inds]
gene = np.array(gene)
gene = gene[inds]
inds = np.any(mat != 0, axis=1)
mat = mat[inds]
gene = np.array(gene)
gene = gene[inds]
#for i in mat:
# if np.mean(abs(i))==0:
# print i
nm = norm_max(mat)
#for i in nm:
# if np.mean(abs(i))==0:
# print i
genex = gene
if len(genelist) > 0:
lb = ens_genes(gene, genelist)
genex = lb
else:
return nm, lib2, gene, genex, lib
b = plt.matshow(nm, aspect='auto', cmap='RdYlBu')
if len(libfile) > 0:
plt.xticks(range(len(lib2)), lib2)
else:
plt.xticks(range(len(lib)), lib)
plt.colorbar()
plt.xticks(rotation=90)
plt.yticks(range(len(genex)), genex)
plt.yticks(fontsize=8)
mytemplate(nm)
plt.xlabel(lb)
lb = fl.split('/')[-1][:-5]
return nm, lib2, gene, genex, lib
def gene_rpkm_lineage(hugo,rpkmdata,rpkmfiles,allgenes,libs,*plotbool):
genelist=[]
for i in range(len(hugo)):
genelist.append(hugo[i])
genelist=ens_genes(genelist,allgenes)
rpkmls=[]
files=[]
for gene in genelist:
rpkmls.append([])
for ind in xrange(len(rpkmdata)):
for i in rpkmdata[ind]:
if gene in i[0]:
rpkmls[-1].append(i[2])
if len(rpkmls)==1:
files.append(rpkmfiles[ind][:rpkmfiles[ind].index('.')])
break
files=np.array(files)
rpkmls=data2arr(rpkmls)
sortedinds=rpkmls[0].argsort()
files=files[sortedinds]
files=files[::-1]
for i in range(len(genelist)):
rpkmls[i]=rpkmls[i][sortedinds]
rpkmls[i]=rpkmls[i][::-1]
normalfiles=[]
normalrpkms=[]
RTfiles=[]
RTrpkms=[]
libs=read.read_dat(libs,'\t')
for ind in range(len(files)):
for lib in libs:
if files[ind] in lib[0]:
files[ind]=lib[1]+'-'+lib[-1]
break
for i in range(len(genelist)):
rpkmls[i]=np.log(rpkmls[i]+0.001)/np.log(10)
for i in range(len(genelist)):
normalrpkms.append([])
RTrpkms.append([])
for ind in range(len(files)):
if 'Cancer' not in files[ind] and 'ES' not in files[ind]:
print files[ind]
normalfiles.append(files[ind])
for i in range(len(genelist)):
normalrpkms[i].append(rpkmls[i][ind])
elif 'RT' in files[ind]:
RTfiles.append(files[ind])
for i in range(len(genelist)):
RTrpkms[i].append(rpkmls[i][ind])
plt.figure()
for i in range(len(genelist)):
if i<7:
plt.plot(xrange(len(list(files))),rpkmls[i],'s-',label=hugo[i])
else:
plt.plot(xrange(len(list(files))),rpkmls[i],'^-',label=hugo[i])
plt.legend(loc=1)
plt.xticks(range(len(files)),files)
plt.xticks(rotation=90)
plt. xticks(fontsize=8)
plt.ylabel('$log_{10}$ $RPKM$')
plt.grid()
plt.show()
def heatmap(rpkmls,hugo,txt):
rpkmls=data2arr(rpkmls)
norm,matrix,dist=analyse.all_corr(rpkmls.T)
plt.matshow((matrix),cmap='RdYlBu',vmax=1,vmin=-1)
plt.yticks(range(len(hugo)),hugo)
plt.xticks(range(len(hugo)),hugo)
plt.xticks(rotation=90)
plt.xlabel('Correlation-'+txt)
plt.colorbar()
#plt.matshow(np.log(pval),cmap='PuBu',vmax=-5,vmin=-25)
#plt.yticks(range(len(hugo)),hugo)
#plt.xticks(range(len(hugo)),hugo)
#plt.xlabel('P-Value $\log$ -'+txt)
#plt.xticks(rotation=90)
#plt.colorbar()
if len(plotbool)>0:
if plotbool[0]:
heatmap(rpkmls,hugo,'All Samples')
heatmap(normalrpkms,hugo,'Normal Samples')
heatmap(RTrpkms,hugo,'RT Samples')
for i in range(len(hugo)):
print hugo[i],ss.ks_2samp(normalrpkms[i],RTrpkms[i])[1]*np.float(len(allgenes)), 'KS'
print hugo[i],ss.ttest_ind(normalrpkms[i],RTrpkms[i])[1]*np.float(len(allgenes)), 'T-test'
#fig=plt.figure()
#plt.plot()
return rpkmls,files,data2arr(normalrpkms),np.array(normalfiles),genelist
import sys,read, write,bedtools, analyse
mark=sys.argv[1]
sample1=sys.argv[2]
sample2=sys.argv[3]
num=int(sys.argv[4])
T=float(sys.argv[5])
dirout=sys.argv[6]
field=int(sys.argv[7])
libsfile=str(sys.argv[8])
TSSdir=str(sys.argv[9])
beds={}
libsdata=read.read_dat(libsfile,'\t')
indmark=libsdata[0].index(mark)
list1=[]
list2=[]
for i in xrange(1,len(libsdata)):
if sample1 in libsdata[i][field] and len(libsdata[i][indmark])>1:
list1.append(libsdata[i][indmark])
elif sample2 in libsdata[i][field] and len(libsdata[i][indmark])>1:
list2.append(libsdata[i][indmark])
beds={}
files1,beds[sample1]=read.readall_bed(TSSdir,mark,list1)
files2,beds[sample2]=read.readall_bed(TSSdir,mark,list2)
genes=read.read_gene_pos('rhabdoid/coverage/TSS_2000_all/hg19v69_genes.TSS_2000.pc.A03480.H3K27me3.GE02.coverage')
#!/usr/local/bin/env python
#coding: utf8
import sys, read, write, bedtools, analyse
mark = sys.argv[1]
sample1 = sys.argv[2]
sample2 = sys.argv[3]
num = int(sys.argv[4])
T = float(sys.argv[5])
dirout = sys.argv[6]
field = int(sys.argv[7])
libsfile = str(sys.argv[8])
TSSdir = str(sys.argv[9])
beds = {}
libsdata = read.read_dat(libsfile, '\t')
indmark = libsdata[0].index(mark)
list1 = []
list2 = []
for i in xrange(1, len(libsdata)):
if sample1 in libsdata[i][field] and len(libsdata[i][indmark]) > 1:
list1.append(libsdata[i][indmark])
elif sample2 in libsdata[i][field] and len(libsdata[i][indmark]) > 1:
list2.append(libsdata[i][indmark])
beds = {}
files1, beds[sample1] = read.readall_bed(TSSdir, mark, list1)
files2, beds[sample2] = read.readall_bed(TSSdir, mark, list2)
genes = read.read_gene_pos(
'rhabdoid/coverage/TSS_2000_all/hg19v69_genes.TSS_2000.pc.A03480.H3K27me3.GE02.coverage'