def concatenateSimpleBed (fileList):
a = []
for i in fileList:
a.extend(SimpleBed(i))
return mergeIntervals(a)
def compareTwo(a, b, effectiveGenomeSize, validChrs):
print "Compare 2"
aBed = SimpleBed(a, validChrs=validChrs)
bBed = SimpleBed(b, validChrs=validChrs)
print "Determine overlap"
overlappingbp = determineOverlap(aBed, bBed)
bpA = numbBP(aBed)
bpB = numbBP(bBed)
print "A percent overlapped:" + str(
100.0 * float(overlappingbp) / float(bpA)) + ", Expected:" + str(
100.0 * float(bpB) / float(effectiveGenomeSize))
print "B percent overlapped:" + str(
100.0 * float(overlappingbp) / float(bpB)) + ", Expected:" + str(
100.0 * float(bpA) / float(effectiveGenomeSize))
print "Pvalue" + pValue(overlappingbp, bpA, bpB, effectiveGenomeSize)
def __init__(self, build):
super(ChromosomeGaps, self).__init__(
os.path.expanduser("~/mount/publicdata/" + build + "/gaps." +
build))
self.chrmgaps = collections.defaultdict(int)
self.individualgaps = SimpleBed(
os.path.expanduser("~/mount/publicdata/" + build + "/gaps." +
build))
for (chr, start, stop) in self.individualgaps:
self.chrmgaps[chr] += (stop - start)
def compareTwo(a, b, numbRandomRegions, cores):
randomOverlap = []
randomBPOverlap = []
randomPercentOverlap = []
aBed = SimpleBed(a)
bBed = SimpleBed(b)
params = (aBed, bBed)
pool = Pool(processes=cores)
result = pool.map(doOverlap, [params] * numbRandomRegions)
for r, rbp, rpercent, _, _ in result:
randomOverlap.append(r)
randomBPOverlap.append(rbp)
randomPercentOverlap.append(rpercent)
bTree = BedIntervalTree(b)
overlapping, overlappingbp, overlappingPercent, avgOverlappingLength, avgNonOverlappingLength = determineOverlap(
aBed, bTree)
#########
random = sum(randomOverlap) / float(len(randomOverlap))
print "Average Overlapping Length:" + str(avgOverlappingLength)
print "Average NonOverlapping Length:" + str(avgNonOverlappingLength)
print
print "Overlap 1bp: " + str(overlapping) + " / " + str(
len(aBed)) + " (" + str(100.0 * overlapping / float(len(aBed))) + "%)"
print "Overlap 1bp Random: " + str(random) + " / " + str(
len(aBed)) + " (" + str(100.0 * random / float(len(aBed))) + "%)"
print "FC 1bp: " + str(overlapping) + " / " + str(random) + " = " + str(
overlapping / float(random))
#print "Max FC 1bp:"+ str(len(aBed)) + " / " + str(random) + " = " + str(len(aBed)/float(random))
print "Pvalues (corr/anticorr):" + str(pValue(overlapping, randomOverlap))
##########
# print
#
# ##########
#
# randomPercent = sum(randomPercentOverlap)/float(len(randomPercentOverlap))
#
# print "Overlap 50%: " + str(overlappingPercent) + " / " + str(len(aBed)) + " ("+str(100.0*overlappingPercent/float(len(aBed)))+"%)"
# print "Overlap 50% Random: " + str(randomPercent) + " / " + str(len(aBed)) + " ("+str(100.0*randomPercent/float(len(aBed)))+"%)"
# print "FC 50%: "+str(overlappingPercent) + " / " + str(randomPercent) + " = " + str(overlappingPercent/float(randomPercent))
#
# #print "Max FC 1bp:"+ str(len(aBed)) + " / " + str(randomPercent) + " = " + str(len(aBed)/float(randomPercent))
#
# print "Pvalues (corr/anticorr):" + str(pValue(overlappingPercent,randomPercentOverlap))
#
# ##########
print
##########
randombp = sum(randomBPOverlap) / float(len(randomBPOverlap))
print "BP: " + str(overlappingbp) + " / " + str(numbBP(aBed)) + " (" + str(
100.0 * overlappingbp / float(numbBP(aBed))) + "%)"
print "Overlap BP Random: " + str(randombp) + " / " + str(
numbBP(aBed)) + " (" + str(
100.0 * randombp / float(numbBP(aBed))) + "%)"
print "FC BP: " + str(overlappingbp) + " / " + str(randombp) + " = " + str(
overlappingbp / float(randombp))
#print "Max FC BP:"+ str(numbBP(aBed)) + " / " + str(randombp) + " = " + str(numbBP(aBed)/float(randombp))
print "Pvalues:" + str(pValue(overlappingbp, randomBPOverlap))
for i in range(len(intervals)):
for j in range(i, len(intervals)):
if i == j:
continue
else:
if (intervals[j].start < intervals[i].stop and intervals[i].start < intervals[j].stop):
intervals[i].start = min(intervals[i].start, intervals[j].start)
intervals[i].stop = max(intervals[i].stop, intervals[j].stop)
del intervals[j]
mergeIntervals(intervals)
return
allintervals = {}
for arg in args:
treatment = SimpleBed(arg)
for (chr, start, stop) in treatment:
interval = Interval(start, stop)
if chr not in allintervals:
allintervals[chr] = []
allintervals[chr].append(interval)
for chr in allintervals:
mergeIntervals(allintervals[chr])
treatments = {}
for arg in args:
treatments[arg] = BedTreatment(arg)
prettyheader = ""
assert outputfile != None
assert genomeBuild != None
assert replicateAnnotation != None
assert replicateTypes != None
if len(replicateAnnotation)%len(replicateTypes) == 0:
print "Unbalanced length of replicant annotators..."
if (len(replicateAnnotation)/2) != len(replicateTypes):
print "Unbalanced counts of replicant annotators to replicant types..."
if genomeBuild not in ("hg19", "hg18", "mm9"):
genomeBuild = "hg19"
print "Genome build type unacceptable. Defaulting to genome hg19..."
print "Loading regions file into memory..."
regions = SimpleBed(regionsfile)
print "Loading methylation data into memory..."
methdata = PooledAggregateTree(methdatafile)
print "Loading genome into structure..."
genome = Genome(genomeBuild)
def methGetP(rep1, rep2):
'''
Calculate fishers exact test (p-value) from the methylated and unmethylated
scores between two sets of replicates
'''
oddsRatio, p = scipy.stats.fisher_exact([rep1, rep2]) #@UnusedVariable
return p
def compareTwoFiles(a, b):
aBed = SimpleBed(a)
bTree = BedIntervalTree(b)
return determineOverlap(aBed, bTree)
a = arg
elif opt == "-b":
b = arg
elif opt == "-c":
c = arg
elif opt == "-d":
d = arg
elif opt == "-e":
e = arg
elif opt == "-n":
numberTrials = int(arg)
elif opt == "--cores":
cores = int(arg)
#Gets the overlap values
aa = numbBP(SimpleBed(a))
bb = numbBP(SimpleBed(b))
cc = numbBP(SimpleBed(c))
dd = numbBP(SimpleBed(d))
ee = numbBP(SimpleBed(e))
ab, abBed = compareTwoFiles(a, b)
ac, acBed = compareTwoFiles(a, c)
ad, adBed = compareTwoFiles(a, d)
ae, aeBed = compareTwoFiles(a, e)
bc, bcBed = compareTwoFiles(b, c)
bd, bdBed = compareTwoFiles(b, d)
be, beBed = compareTwoFiles(b, e)
cd, cdBed = compareTwoFiles(c, d)
ce, ceBed = compareTwoFiles(c, e)
de, deBed = compareTwoFiles(d, e)
# UCSC table browser - Mapping and Sequencing - Chromosome Bands
# Download all columns with exception of "gieStain"
gBanding = ExtendedBed(os.path.expanduser("/mnt/50tb/publicdata/" + assembly +
"/G-Banding/cytogenetic.map.bed"),
defaultkeys=["chrm", "start", "stop", "band"],
forcekeys=True)
chromosomeEnds = ChromosomeEnds(assembly)
###
###
###
headerRow = ['id', 'chr', 'start', 'stop']
intervals = SimpleBed(infile)
# extra columns from the input file
headerRow.extend(intervals.header[3:])
headerRow.extend([
'In genes', "In genes + promotor", "intergenic", "iswholelyinintron",
"isinpromotor", "isindownstreampromotor", "isinupstreampromotor",
'overlapsTSS', 'In exons', 'in.cpg'
])
# "Ebox Motif", "Ebox Motif - Canonical A", "Ebox Motif - Canonical G", "AP1 Motif","AP-2 Motif"])
for genelist in genelists:
headerRow.append(genelist.getFullName())
headerRow.append(genelist.getFullName() + "-promotoronly")
genelists_bound[genelist.getFullName()] = set()
import getopt
if __name__ == "__main__":
try:
opts, args = getopt.getopt(sys.argv[1:], "b:a:o:", [])
except getopt.GetoptError, err:
# print help information and exit:
print str(err) # will print something like "option -a not recognized"
sys.exit(2)
bedfile = None
outputfile = None
for o, a in opts:
if o == "-a":
assembly = a
if o == "-b":
bedfile = SimpleBed(a)
if o == "-o":
outputfile = csv.writer(open(a, "w"), delimiter='\t')
ends = ChromosomeEnds(assembly)
assert bedfile != None
assert outputfile != None
for (chr, start, stop) in bedfile:
if int(start) > ends[chr]:
continue
else:
outputfile.writerow([chr, start, str(min(ends[chr], int(stop)))])
def overlappingBP(start, stop, intervals):
overlapBP = 0
for interval in intervals:
# how much of interval is overlapped with start, stop
overlapBP += max(
min(interval.end, stop) - max(interval.start, start), 0)
return overlapBP
for infile in args:
print infile
intervals = SimpleBed(infile)
print "Starting..."
for row in intervals:
chr, start, stop = row
# 0 length regions
if start == stop:
continue
size = stop - start
# genes
ingenes = mergeIntervals(genes.getIntervalsInRange(chr, start, stop))
excList = exclude[i]
print intList
#Test for excluded regions:
excludeUnionBed = None
excludeBp = 0
if (len(excList)) >0:
#Concatenates the excluded bed list and gets the union of its regions
excludeUnionBed = concatenateSimpleBed(excList)
#Outputs the union list to a file
excludeUnionBed = intervalListToFile(sys.path[0],excludeUnionBed)
if len(intList) == 1:
overlap = numbBP(SimpleBed(intList[0]))
raweights.append(overlap)
if excludeUnionBed != None:
excludeBp,_ = compareTwoFiles(intList[0],excludeUnionBed)
weights.append(overlap - excludeBp)
if len(intList) == 2:
overlap,bed = compareTwoFiles(intList[0],intList[1])
raweights.append(overlap)
if excludeUnionBed != None:
excludeBp,_ = compareListAndFile(bed,excludeUnionBed)
weights.append(overlap - excludeBp)
if len(intList) > 2:
_,bed = compareTwoFiles(intList[0],intList[1])
cores = 1
for opt, arg in opts:
if opt == "-a":
a = arg
elif opt == "-b":
b = arg
elif opt == "-c":
c = arg
elif opt == "-n":
numberTrials = int(arg)
elif opt == "--cores":
cores = int(arg)
#Gets the overlap values
aa = numbBP(SimpleBed(a))
bb = numbBP(SimpleBed(b))
cc = numbBP(SimpleBed(c))
ab, abBed = compareTwoFiles(a, b)
ac, _ = compareTwoFiles(a, c)
bc, _ = compareTwoFiles(b, c)
abc, _ = compareListAndFile(abBed, c)
print "Raw:"
print "a\t" + str(aa)