def generatecouplemat(pathTEcouple,repository,speciesname,endofthefile,chrsizepath,resolution):
resolution=int(resolution) #cast of argv
Matpath=repository+speciesname+".TEcouplematrix"
TEposdict=generateTElistedposdict(repository,speciesname,endofthefile,resolution)
print(TEposdict.keys())
filecouple=open(pathTEcouple,"r")
chrlist=utils.dictchr[speciesname] #help us to loop only on good chr
sizedict=utils.loadchrsizedict(chrsizepath,resolution)
matcolsize=sizedict["HicChrBeginchrX"] #size of the matrix : ["HicUsedTotalSize"] or #TODO here
l=filecouple.readline()
#initialisation of the loop
ls=l.split()
vi=np.zeros((1,matcolsize))
vi=completetheline(ls[0],TEposdict[ls[0]],sizedict,vi,1)
vi=completetheline(ls[1],TEposdict[ls[1]],sizedict,vi,-1)
mat=vi
l=filecouple.readline()
i=0
while l:
ls=l.split()
print("===>line",i)
vi=np.zeros((1,matcolsize))
vi=completetheline(ls[0],TEposdict[ls[0]],sizedict,vi,1)
vi=completetheline(ls[1],TEposdict[ls[1]],sizedict,vi,1)
mat=np.concatenate((mat,vi),axis=0)
l=filecouple.readline()
i+=1
filecouple.close()
toto={}#just to be un matlab savemat format
toto['mat']=mat
scpio.savemat(Matpath,toto)
def gccontentforonechr(fastarep,chrsizepath,species,GCoutrep,resolution):
resolution=int(resolution)
sizedict=utils.loadchrsizedict(chrsizepath,resolution)
matcolsize=sizedict["HicUsedTotalSize"] #HicUsedTotalSize HicChrBeginchr2
mat=np.zeros(matcolsize)
k=0
for i in utils.dictchr[species]:
print(i)
o=open(fastarep+i+".fa","r")
l=o.readline()
s=""
#init
l=l.split(" ")
l=o.readline()
while l:
l=l.replace("\n","")
s+=l
l=o.readline()
o.close()
i=0
j=resolution
if resolution>len(s):
mat[k]=gcpercent(s)
k+=1
while j<len(s):
mat[k]=gcpercent(s[i:j])
i+=resolution
j+=resolution
k+=1
toto={}#just to be un matlab savemat format
toto['mat']=mat
scpio.savemat(GCoutrep+str(resolution)+"pbGCvec.mat",toto)
utils.savematrixasfilelist3(mat,GCoutrep+str(resolution)+"pbGCvec.csv")
def generateTErecurencematrixHicComparableForOneChr(repository,endofthefile,achr,resolution,chrsizepath,species):
dumpmatname=repository+achr+endofthefile+"dumpmatHiCbin.dump"
densityname=repository+achr+endofthefile
if op.exists(dumpmatname+".npy"):
print("===> Matrice deja existante")
thematrix=np.load(open(dumpmatname+".npy","rb"))
else:
print("===>Constitution de la matrice")
filename=repository+achr+endofthefile
print("resolution:",resolution)
sizedict=utils.loadchrsizedict(chrsizepath,resolution)
#find the size of the list of list base on distance
sizeList=findsizeforHicListofList(sizedict,utils.dictchr[species],achr)
print("taille de la liste:",sizeList)
setlist=generateHicListofpattern(filename,sizeList,resolution)
transformHiClistofpatterninDensity(setlist,"LTR/Gypsy",densityname+"Gypsy")
transformHiClistofpatterninDensity(setlist,"LINE/CR1",densityname+"CR1")
transformHiClistofpatterninDensity(setlist,"DNA/P",densityname+"DNAP")
transformHiClistofpatterninDensity(setlist,"RC/Helitron",densityname+"helitron")
print("taille de la setlist",len(setlist))
L=sizeList #in theory len(setlist)==sizeList
#Lset=windowsize #juste for speed writing
print("==>Taille de la matrice:",L)
thematrix=np.zeros((L,L))
i=0
j=0
while i<L:
print(i)
while j<L:
if j>=i:
val=distancebetweenunequallist3(setlist[i],setlist[j])
thematrix[i,j]=val
thematrix[j,i]=val
j+=1
j=0
i+=1
print("====>Dumping de la matrice")
#pickle.dump(thematrix,open(dumpmatname,"wb")) #other option of dump but with bigger file
#toto={}#just to be un matlab savemat format
#toto['mat']=thematrix
#scpio.savemat(repository+achr+endofthefile+"TErecurencematrix.mat",toto)
np.save(dumpmatname, thematrix)
"""cormat=np.corrcoef(thematrix)
v,e=giveeigenvalueandvectorvalue(cormat,10)
utils.savematrixasfilelist3(e[:,0],repository+achr+"Corrv1")
utils.savematrixasfilelist3(e[:,1],repository+achr+"Corrv2")
utils.savematrixasfilelist3(e[:,2],repository+achr+"Corrv3")
utils.savematrixasfilelist3(e[:,3],repository+achr+"Corrv4")"""
print("===> Afichage de la matrice")
manipname=repository+achr+endofthefile+"recurenceplot.png"
print(thematrix.shape)
#print(np.where(thematrix<0))
utils.showamatwithcolorcode(thematrix,manipname,"bwr") #jet<bwr
return thematrix
def generateTEmatrixforaspecies(repository,speciesname,endofthefile,chrsizepath,resolution):
resolution=int(resolution) #cast of argv
TEpath=repository+speciesname+".TElist"
Matpath=repository+speciesname+".TEmatrix"
TEpath=repository+speciesname+".TElist"
DEpath=repository+speciesname+".TEdict"
DEfamilypath=repository+speciesname+".TEfamilydict"
#dict to return
if op.exists(TEpath):
TElist=utils.loadfilelist(TEpath)
TEfamilydict=utils.loadstrfiledict(DEfamilypath)
else:
TElist,TEfamilydict=utils.generatelistofTEforaspecies(repository,speciesname,endofthefile)
sizedict=utils.loadchrsizedict(chrsizepath,resolution)
matcolsize=sizedict["HicUsedTotalSize"] #size of the matrix : ["HicUsedTotalSize"] or ["HicChrBeginchrX"]#TODO here
mat=np.zeros((len(TElist),matcolsize))
chrlist=utils.dictchr[speciesname] #help us to loop only on good chr
#print(sizedict)
print(matcolsize)
print(chrlist)
#by chr file
for i in chrlist:
filename=repository+i+endofthefile
filein=open(filename,"r")
l=filein.readline()
l=filein.readline()
ls=l.split() #just to have the first chr size before loop
chrbegin=np.float(sizedict["HicChrBegin"+ls[0]])
#line of the file
while l:
ls=l.split()
#by bin
begin=(np.float(ls[4])/resolution)+chrbegin
end=(np.float(ls[5])/resolution)+chrbegin
i=0
#print(TElist[0],TElist[1])
while i<np.ceil(end-begin):
#print((end-begin),ls[1],ls[0],ls[4],ls[5],TElist.index(ls[1]),begin+i,chrbegin)
mat[TElist.index(ls[1]),begin+i]=1
i+=1
l=filein.readline()
filein.close()
toto={}#just to be un matlab savemat format
toto['mat']=sparse.csr_matrix(mat)
scpio.savemat(Matpath,toto)
def makeTEPieChart(repository, speciesname, endofthefile, chrsizepath):
outname = repository + speciesname + "TEproportion"
sizedict = utils.loadchrsizedict(chrsizepath, 1) # absolutely no binning stuff => resolution=1
# load some accelerator information that we supposed to have
TEpath = repository + speciesname + ".TElist"
DEfamilypath = repository + speciesname + ".TEfamilydict"
if op.exists(TEpath):
TElist = utils.loadfilelist(TEpath)
TEfamilydict = utils.loadstrfiledict(DEfamilypath)
else:
TElist, TEfamilydict = utils.generatelistofTEforaspecies(repository, speciesname, endofthefile)
ProportionDict, TEreversefamilitydict = reservefamilydict(TEfamilydict)
# reel algorythm
sumchrtot = 0
sumTE = 0 # helpfull tu calculed unmasked proportion
chrlist = utils.dictchr[speciesname] # in theory : chr here are same as chrsizepath #no bug ifelse
for z in chrlist:
filename = repository + z + endofthefile
print("=====>nom du fichier de chr: ", filename)
filein = open(filename, "r")
l = filein.readline() # first line : annotation
l = filein.readline()
while l:
ls = l.split()
val = float(ls[5]) - float(ls[4])
if val < 0:
print("Danger")
ProportionDict[TEfamilydict[ls[1]]] += val
sumTE += val
l = filein.readline()
sumchrtot += sizedict[z]
filein.close()
# print(ProportionDict)
ProportionDict["Autre"] = sumchrtot - sumTE
# save the dict
fout = open(outname, "w")
print("nom du repertoire de sortie", outname)
fout.write("TEtype\tQuantity\n")
for i in ProportionDict:
s = i + "\t" + str(ProportionDict[i]) + "\n"
fout.write(s)
fout.close()