def bootstrap(n, chr_len, sra_loci, feature_loci):
data = ol.overlap(sra_loci, feature_loci, percentage=True)
boot = []
for i in range(n):
resampled_loci = resample(chr_len, sra_loci)
ol_perc = ol.overlap(resampled_loci, feature_loci, percentage=True)
boot.append(ol_perc)
z = z_score(data, boot)
return z
def bootstrap(n,chr_len, sra_loci, feature_loci):
data = ol.overlap(sra_loci, feature_loci, percentage = True)
boot = []
for i in range(n):
resampled_loci = resample(chr_len, sra_loci)
ol_perc = ol.overlap(resampled_loci, feature_loci,
percentage = True)
boot.append(ol_perc)
z = z_score(data, boot)
return z
def count(cls, loci, norm_factor, distinct=True):
'''count sRNA reads within given collection of loci
loci: [chr, start, end] MUST BE SORTED'''
locus_chr = set([i[0] for i in loci])
sRNA_chr = set([i[0] for i in cls])
chr = list(locus_chr & sRNA_chr)
chr.sort()
count = {}
for c in chr:
count[c] = []
a = filter(lambda x: x[0] == c, cls)
b = filter(lambda x: x[0] == c, loci)
aa = [[i[1], i[2]] for i in a]
bb = [[i[1], i[2]] for i in b]
o = ol.overlap(aa, bb)
for i in range(len(o)):
if o[i]:
read = 0
for j in o[i]:
if distinct == True:
read += float(1) / norm_factor
else:
read += float(a[j][3]) / (a[j][4] * norm_factor)
count[c].append(read)
else:
count[c].append(0)
return count
def count(cls, loci, norm_factor, distinct = True):
'''count sRNA reads within given collection of loci
loci: [chr, start, end] MUST BE SORTED'''
locus_chr = set([i[0] for i in loci])
sRNA_chr = set([i[0] for i in cls])
chr = list(locus_chr & sRNA_chr)
chr.sort()
count = {}
for c in chr:
count[c] = []
a = filter(lambda x: x[0] == c, cls)
b = filter(lambda x: x[0] == c, loci)
aa = [[i[1],i[2]] for i in a]
bb = [[i[1],i[2]] for i in b]
o = ol.overlap(aa, bb)
for i in range(len(o)):
if o[i]:
read = 0
for j in o[i]:
if distinct == True:
read += float(1)/norm_factor
else:
read += float(a[j][3])/(a[j][4]*norm_factor)
count[c].append(read)
else:
count[c].append(0)
return count
def locus_avg_sRNA(cls, locus, flank, window, norm_factor):
'''calculate the average distinct sRNA density in given loci'''
cls.sort_map(locus)
locus_chr = list(set([i[0] for i in locus]))
sRNA_chr = list(set([i[0] for i in cls]))
chr = locus_chr & sRNA_chr
chr.sort()
locus_num = len(locus)
point = flank / window
up = [0] * point
trans = [0] * point
down = [0] * point
for c in chr:
a = filter(lambda x: x[0] == c, cls)
b = filter(lambda x: x[0] == c, locus)
o = ol.overlap([i[1:4] for i in a],
[[i[1] - flank, i[2] + flank, i[3], i[4]]
for i in b])
for i in range(len(o)):
if o[i]:
locus_len = b[i][2] - b[i][1] + 1 + flank * 2
each_locus = [0] * locus_len
for j in o[i]:
rel_coord = [
a[j][1] - b[i][1] + flank,
a[j][2] - b[i][1] + flank
]
for k in range(max(0, rel_coord[0]),
min(locus_len, rel_coord[1]) + 1):
each_locus[k] += 1
if b[i][3] is "-":
each_locus.reverse()
up_list = each_locus[0:flank]
trans_list = each_locus[flank:locus_len - flank]
down_list = each_locus[locus_len - flank:]
for i in range(point):
up[i] += float(
sum(up_list[i * window:(i + 1) * window])) / window
down[i] += float(
sum(down_list[i * window:(i + 1) *
window])) / window
a = len(trans_list) / point
if a >= 1:
for i in range(point):
trans[i] += float(
sum(trans_list[i * a:(i + 1) * a])) / a
return [i / (locus_num * norm_factor) for i in up + trans + down]
def locus_avg_sRNA(cls, locus, flank, window, norm_factor):
'''calculate the average distinct sRNA density in given loci'''
cls.sort_map(locus)
locus_chr = list(set([i[0] for i in locus]))
sRNA_chr = list(set([i[0] for i in cls]))
chr = locus_chr & sRNA_chr
chr.sort()
locus_num = len(locus)
point = flank/window
up = [0]*point
trans = [0]*point
down = [0]*point
for c in chr:
a = filter(lambda x: x[0] == c, cls)
b = filter(lambda x: x[0] == c, locus)
o = ol.overlap([i[1:4] for i in a],
[[i[1]-flank, i[2]+flank, i[3],i[4]] for i in b])
for i in range(len(o)):
if o[i]:
locus_len = b[i][2]-b[i][1]+1+flank*2
each_locus = [0] * locus_len
for j in o[i]:
rel_coord = [a[j][1]-b[i][1]+flank,
a[j][2]-b[i][1]+flank]
for k in range(max(0,rel_coord[0]),
min(locus_len,rel_coord[1])+1):
each_locus[k] += 1
if b[i][3] is "-":
each_locus.reverse()
up_list = each_locus[0:flank]
trans_list = each_locus[flank:locus_len-flank]
down_list = each_locus[locus_len-flank:]
for i in range(point):
up[i] += float(sum(up_list[i*window:(i+1)*window]))/window
down[i] += float(sum(down_list[i*window:(i+1)*window]))/window
a = len(trans_list)/point
if a >= 1:
for i in range(point):
trans[i]+= float(sum(trans_list[i*a:(i+1)*a]))/a
return [i/(locus_num * norm_factor) for i in up + trans + down]
def find(cls, locus, remove=False):
'''find all sRNAs mapped to a given genomic region'''
# locus needs to be sorted before doing overlap
cls.sort_map(locus)
locus_chr = list(set([i[0] for i in locus]))
sRNA_chr = list(set([i[0] for i in cls]))
locus_chr.sort()
sRNA_chr.sort()
new_list = []
for c in sRNA_chr:
a = filter(lambda x: x[0] == c, cls)
del_index = []
if c not in locus_chr: # if no need to search that chromosome:
if remove == True: # no sRNA to be removed from sRNA_map
new_list = new_list + a[:]
next
elif c in locus_chr:
b = filter(lambda x: x[0] == c, locus)
# extract start and end, overlap them
o = ol.overlap([i[1:3] for i in a], [i[1:3] for i in b])
if all([len(i) == 0 for i in o
]): # if no sRNAs were found to match any given loci
new_list = new_list + a[:]
print 'no sRNAs were found on Chr%s' % c
else:
for i in xrange(len(o)): # for each locus
sRNA = []
try: # if agi is provided
agi = b[i][3]
if remove == False:
print '@' + agi
except IndexError:
pass
if o[i]: # if there is at least one sRNA in that locus
del_index = del_index + o[i]
for j in o[i]:
sRNA.append(a[j])
if remove == False:
for s in sRNA:
print 'chr%s\t%d\t%d\t%d\t%d\t%s' % s
else: # if no sRNA is mapped to that locus
if remove == False:
print 'NA'
if remove == True: # if sRNAs need to be removed from sRNA_map
for k in xrange(len(del_index) - 1, -1,
-1): # del from the largest index
try:
del a[del_index[k]]
except IndexError:
print c, len(a), k
new_list = new_list + a[:]
if remove == False:
for c in locus_chr:
if c not in sRNA_chr: ## if acc of a locus is not in sRNA list
b = filter(lambda x: x[0] == c, locus)
for i in b:
try:
agi = i[3]
print '@' + agi
print 'No accession found'
except IndexError:
pass
elif remove == True:
cls.sort_map(new_list)
return new_list[:]
def find(cls, locus, remove = False):
'''find all sRNAs mapped to a given genomic region'''
# locus needs to be sorted before doing overlap
cls.sort_map(locus)
locus_chr = list(set([i[0] for i in locus]))
sRNA_chr = list(set([i[0] for i in cls]))
locus_chr.sort()
sRNA_chr.sort()
new_list = []
for c in sRNA_chr:
a = filter(lambda x: x[0] == c, cls)
del_index = []
if c not in locus_chr: # if no need to search that chromosome:
if remove == True: # no sRNA to be removed from sRNA_map
new_list = new_list + a[:]
next
elif c in locus_chr:
b = filter(lambda x: x[0] == c, locus)
# extract start and end, overlap them
o = ol.overlap([i[1:3] for i in a],[i[1:3] for i in b])
if all([len(i) == 0 for i in o]): # if no sRNAs were found to match any given loci
new_list = new_list + a[:]
print 'no sRNAs were found on Chr%s' % c
else:
for i in xrange(len(o)): # for each locus
sRNA = []
try: # if agi is provided
agi = b[i][3]
if remove == False:
print '@' + agi
except IndexError:
pass
if o[i]: # if there is at least one sRNA in that locus
del_index = del_index + o[i]
for j in o[i]:
sRNA.append(a[j])
if remove == False:
for s in sRNA:
print 'chr%s\t%d\t%d\t%d\t%d\t%s' % s
else: # if no sRNA is mapped to that locus
if remove == False:
print 'NA'
if remove == True: # if sRNAs need to be removed from sRNA_map
for k in xrange(len(del_index)-1, -1, -1): # del from the largest index
try:
del a[del_index[k]]
except IndexError:
print c, len(a), k
new_list = new_list + a[:]
if remove == False:
for c in locus_chr:
if c not in sRNA_chr: ## if acc of a locus is not in sRNA list
b = filter(lambda x: x[0] == c, locus)
for i in b:
try:
agi = i[3]
print '@' + agi
print 'No accession found'
except IndexError:
pass
elif remove == True:
cls.sort_map(new_list)
return new_list[:]