def run(bidirlist,fimofile,chipfile,bidirdict):
#====================================================================================================
#Create a dictionary for each file (bidirectional, fimo, chip), convert (chr,start,stop) to one list
#calculating chromosome offsets from lists above
fimodict = Functions.create_tup_uncut_fimo2(fimofile, True)
fimolist = list()
for chrom in fimodict:
if chrom in chromosomes:
i = chromosomes.index(chrom)
for interval in fimodict[chrom]:
fimolist.append((int(interval[0])+sum(sizes[0:i]),int(interval[1])+sum(sizes[0:i])))
chipdict = Functions.create_tup_dict(chipfile, False)
chiplist = list()
for chrom in chipdict:
if chrom in chromosomes:
i = chromosomes.index(chrom)
for interval in chipdict[chrom]:
chiplist.append((int(interval[0])+sum(sizes[0:i]),int(interval[1])+sum(sizes[0:i])))
#====================================================================================================
#Using intervals, compare bidirectionals with fimo sites with chip sites, populate bidirectional
#dictionary with overlapping fimo and chip sites
ST = intervals.comparison(bidirlist,fimolist,chiplist)
AB_Overlaps = ST.find_overlaps(0,1)
ABList = list()
FandB = ST.find_overlaps(0,1)
FandBList = list()
for O in AB_Overlaps:
for interval_original in O.overlaps:
ABList.append((interval_original.start,interval_original.stop,interval_original.INFO))
BC_Overlaps = ST.find_overlaps(0,2)
return bidirdict
def run(bidirectionalfile, DNAseFile):
bidirlist = Functions.parse_bidirfile(bidirectionalfile)
bidirdict = dict()
datapoints = []
for item in bidirlist:
chrom, start, stop, parameters = item
if chrom not in bidirdict:
bidirdict[chrom] = []
else:
bidirdict[chrom].append((start,stop,parameters))
dnasedict = Functions.create_tup_dict(DNAseFile, False)
for chrom in dnasedict:
if chrom in bidirdict:
bidirtree = node.tree(bidirdict[chrom])
for item in dnasedict[chrom]:
bidirsite = bidirtree.searchInterval(item)
if len(bidirsite) != 0:
start = float(item[0])
stop = float(item[1])
size = stop - start
datapoints.append((bidirsite[2][6],size))
return datapoints
def run(BidirFile, ChipFile, FimoFile):
#Give cutoff to i for motif calling
motiftoicutoff = 100
#Give size of window to look at ChIP reads
windowsize = 100
#Create list of TruePositives,TrueNegatives,FalsePositives,FalseNegatives
matrix = list()
BidirDict = Functions.create_tup_bidir(BidirFile)
ChipDict = Functions.create_tup_dict(ChipFile, False)
FimoDict = Functions.create_tup_uncut_fimo2(FimoFile, True)
#RandomizedDict = Functions.create_randomized_sites(windowsize)
#Calculate reads over background by generating equally spaced random sites that are
#2xwindowsize in length
#background = 0
#genomesize = 3234830000
#BackgroundDict = dict()
#for chrom in RandomizedDict:
# if chrom in ChipDict:
# RandomList = RandomizedDict[chrom]
# ChipList = ChipDict[chrom]
# BackgroundDict[chrom] = list()
# STBackground = intervals.comparison((RandomList,ChipList))
# for O in STBackground.find_overlaps(0,1):
# for interval_original in O.overlaps:
# if not interval_original.INFO == '':
# BackgroundDict[chrom].append(interval_original.INFO)
#Bulk of code. Will populate dictionaries (key = chrom) with ChIP coverage
#for motif sites not in bidirectionals (as defined by motifcutoff) and for
#motif sites overlapping bidir calls (within motifcutoff distance)
FnoBDict = dict()
FandBDict = dict()
for chrom in FimoDict:
if chrom in BidirDict:
if chrom in ChipDict:
FnoBDict[chrom] = list()
FandBDict[chrom] = list()
BidirSites = BidirDict[chrom]
#Apply motif distance to i cutoff to bidirectional sites (i.e.
#collapse bidirectional sites into cutoff window)
BidirPaddedSites = list()
for site in BidirSites:
mid = (site[0]+site[1])/2
start = mid-motiftoicutoff
stop = mid+motiftoicutoff
BidirPaddedSites.append((start,stop))
ChipSites = ChipDict[chrom]
FimoSites = FimoDict[chrom]
#Create interval tree between padded Bidir sites and FIMO motif
#sites. Compares motif sites that are at cutoff distance to
#bidir site.
ST1 = intervals.comparison((BidirPaddedSites,FimoSites))
FnoB = ST1.get_isolated(1)
#Find motifs that do not overlap a bidirectional, expand sites
#to mid+-windowsize
FnoBList = list()
for I in FnoB:
mid = (I.start+I.stop)/2
FnoBList.append((mid-windowsize,mid+windowsize))
#Find motifs that overlap a bidirectional, expand sites to
#mid+-windowsize
FandB = ST1.find_overlaps(0,1)
FandBList = list()
for O in FandB:
for interval_original in O.overlaps:
if not interval_original.INFO == '':
mid = (interval_original.start+interval_original.stop)/2
FandBList.append((mid-windowsize,mid+windowsize))
#Find ChIP sites over motif sites that do not overlap a bidir
ST2 = intervals.comparison(FnoBList, ChipSites)
FnoBChipOverlaps = ST2.find_overlaps(0,1)
#FnoBtotalsize = 0
for O in FnoBChipOverlaps:
for interval_original in O.overlaps:
if not interval_original.INFO == '':
FnoBDict[chrom].append(interval_original.INFO)
#FnoBtotalsize += interval_original.stop - interval_original.start
#Find ChIP coverage over motif sites that overlap a bidir
ST3 = intervals.comparison(FandBList,ChipSites)
FandBChipOverlaps = ST3.find_overlaps(0,1)
#FandBtotalsize = 0
for O in FandBChipOverlaps:
for interval_original in O.overlaps:
if not interval_original.INFO == '':
FandBDict[chrom].append(interval_original.INFO)
#FandBtotalsize += interval_original.stop - interval_original.start
return BackgroundDict, FnoBDict, FandBDict
