def test_cache_index(self):
"""
Walk over a track after the index has been made but has been removed
from the filesystem.
"""
values = [(2, 392.0),
(3, 408.0),
(4, 420.0),
(5, 452.0),
(7, 466.0),
(8, 474.0),
(9, 479.0)]
b = [(r, p, v) for r in ('1', '13', 'MT') for (p, v) in values]
walker = wiggelen.walk(open_('b.wig'), force_index=True)
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
remove_indices(keep_cache=True)
walker = wiggelen.walk(open_('b.wig'))
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_walk_complex(self):
"""
Walk over a complex track.
"""
walker = wiggelen.walk(open_('complex.wig'))
for _ in walker:
pass
def test_walk_fixed_step_without_step(self):
"""
Walk over a fixed step wiggle track without `step` arguments.
According to the spec, `fixedStep` definitions require the `step`
argument. However, there seems to be real-world data where it is
missing and the UCSC Genome Browser can still work with it.
So we also support it.
Issue: https://github.com/martijnvermaat/wiggelen/issues/1
"""
c = [
('chr', 1, 64.), ('chr', 2, 64.), ('chr', 3, 65.), ('chr', 4, 66.),
('chr', 5, 66.), ('chr', 6, 66.), ('chr', 7, 69.), ('chr', 8, 70.),
('chr', 9, 71.), ('chr', 10, 71.), ('chr', 11, 71.),
('chr', 12, 71.), ('chr', 13, 71.), ('chr', 14, 71.),
('chr', 15, 71.), ('chr', 16, 71.), ('chr', 17, 71.),
('chr', 18, 71.), ('chr', 19, 73.), ('chr', 20, 73.),
('chr', 21, 73.), ('chr', 22, 73.), ('chr', 23, 73.),
('chr', 24, 73.), ('chr', 25, 73.), ('chr', 26, 74.),
('chr', 27, 75.), ('chr', 28, 75.), ('chr', 29, 75.),
('chr', 30, 75.), ('chr', 31, 76.)
]
walker = wiggelen.walk(open_('fixedstep-without-step.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def find_max(regions, wig):
# find max within a given region. Assumes regions are ordered and non-overlapping
# find max by walking through wig file
# Let's also calculate a new score sum just to be extra sure it's right. The
# old score sum was done during merging regions, so it could have been messed up
regions_with_max = []
start, end, score_sum = regions.pop(0)
max_value = -1
max_position = None
new_score_sum = 0
for chrom, position, value in fill(walk(wig)):
if start <= position and position <= end:
new_score_sum += value
if value >= max_value:
max_value = value
max_position = position
if position >= end and max_value != -1:
# position is past region and max_value has been recorded
region_with_max = (start, end, new_score_sum, max_value,
max_position)
regions_with_max.append(region_with_max)
# grab new region and reset max
if len(regions) == 0: # break when there are no more regions
break
start, end, score_sum = regions.pop(0)
max_value = -1
max_position = None
new_score_sum = 0
return regions_with_max
def test_walk_complex(self):
"""
Walk over a complex track.
"""
walker = wiggelen.walk(open_('complex.wig'))
for _ in walker:
pass
def test_store_index(self):
"""
Walk over a track after the index has been made.
"""
values = [(2, 392.0), (3, 408.0), (4, 420.0), (5, 452.0), (7, 466.0),
(8, 474.0), (9, 479.0)]
b = [(r, p, v) for r in ('1', '13', 'MT') for (p, v) in values]
walker = wiggelen.walk(open_('b.wig'), force_index=True)
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
walker = wiggelen.walk(open_('b.wig'))
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def read_wig(filename, norm=False):
wig_dict = {position : value for region, position, value in fill(walk(open(filename)))}
if norm:
print "Normalizing..."
wig_median = float(np.median(wig_dict.values()))
wig_norm = {x : wig_dict[x]/wig_median for x in wig_dict}
return wig_norm
else:
return wig_dict
def test_walk_single_region(self):
"""
Walk over a track with a single region.
"""
c = [('MT', 1, 364.0), ('MT', 6, 435.0), ('MT', 10, 485.0)]
walker = wiggelen.walk(open_('c.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_walk_fixed_step(self):
"""
Walk over a fixed step wiggle track.
"""
c = [('chr8', 1, 11), ('chr8', 2, 11), ('chr8', 6, 33),
('chr8', 7, 33), ('chr8', 11, 44), ('chr8', 12, 44)]
walker = wiggelen.walk(open_('fixedstep.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_walk_single_region(self):
"""
Walk over a track with a single region.
"""
c = [('MT', 1, 364.0),
('MT', 6, 435.0),
('MT', 10, 485.0)]
walker = wiggelen.walk(open_('c.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_sort_multiple_regions(self):
"""
Walk over a track with multiple regions and index.
"""
values = [(2, 392.0), (3, 408.0), (4, 420.0), (5, 452.0), (7, 466.0),
(8, 474.0), (9, 479.0)]
b = [(r, p, v) for r in ('1', '13', 'MT') for (p, v) in values]
walker = wiggelen.walk(open_('b.wig'), force_index=True)
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def main(infile, threshold, merge_dist, min_width, strand, outfile):
# list of tuples, (start, end, avg_exp)
called_regions = []
start = None
end = None
total_exp = 0
# fill function steps through every position, returns None if position not
# in original wig file
print("Calling preliminary regions...")
wig = open(infile)
for region, position, value in fill(walk(wig)):
if start is None:
# initialize start of new region to current position
start = position
if value is None:
if total_exp > 0: # if a region already exists, end it
called_regions.append((start, end, total_exp))
# reset start, end, and total_exp
start = None
end = None
total_exp = 0
elif value < threshold:
if total_exp > 0: # if a region already exists, end it
called_regions.append((start, end, total_exp))
# reset start, end, and total_exp
start = None
end = None
total_exp = 0
elif value >= threshold: # value exceeds threshold, continue region
total_exp += value
end = position
wig.close()
if total_exp != 0: # finished iterating but one last region
called_regions.append((start, end, total_exp))
print("Filtering out regions smaller than minimum width...")
# filter out regions that are below minimum width
filtered_regions = [
x for x in called_regions if x[1] - x[0] + 1 >= min_width
]
print("Merging regions...")
merged_regions = merge_regions(filtered_regions, merge_dist)
# find max region and re-do score sum
# open wig file again
print("Finding region max and calculating total score...")
regions_with_max = find_max(merged_regions, open(infile))
write_bed(regions_with_max, strand, outfile)
def test_walk_fixed_step(self):
"""
Walk over a fixed step wiggle track.
"""
c = [('chr8', 1, 11),
('chr8', 2, 11),
('chr8', 6, 33),
('chr8', 7, 33),
('chr8', 11, 44),
('chr8', 12, 44)]
walker = wiggelen.walk(open_('fixedstep.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_sort_multiple_regions(self):
"""
Walk over a track with multiple regions and index.
"""
values = [(2, 392.0),
(3, 408.0),
(4, 420.0),
(5, 452.0),
(7, 466.0),
(8, 474.0),
(9, 479.0)]
b = [(r, p, v) for r in ('1', '13', 'MT') for (p, v) in values]
walker = wiggelen.walk(open_('b.wig'), force_index=True)
for expected, item in zip(b, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
def test_walk_fixed_step_without_step(self):
"""
Walk over a fixed step wiggle track without `step` arguments.
According to the spec, `fixedStep` definitions require the `step`
argument. However, there seems to be real-world data where it is
missing and the UCSC Genome Browser can still work with it.
So we also support it.
Issue: https://github.com/martijnvermaat/wiggelen/issues/1
"""
c = [('chr', 1, 64.),
('chr', 2, 64.),
('chr', 3, 65.),
('chr', 4, 66.),
('chr', 5, 66.),
('chr', 6, 66.),
('chr', 7, 69.),
('chr', 8, 70.),
('chr', 9, 71.),
('chr', 10, 71.),
('chr', 11, 71.),
('chr', 12, 71.),
('chr', 13, 71.),
('chr', 14, 71.),
('chr', 15, 71.),
('chr', 16, 71.),
('chr', 17, 71.),
('chr', 18, 71.),
('chr', 19, 73.),
('chr', 20, 73.),
('chr', 21, 73.),
('chr', 22, 73.),
('chr', 23, 73.),
('chr', 24, 73.),
('chr', 25, 73.),
('chr', 26, 74.),
('chr', 27, 75.),
('chr', 28, 75.),
('chr', 29, 75.),
('chr', 30, 75.),
('chr', 31, 76.)]
walker = wiggelen.walk(open_('fixedstep-without-step.wig'))
for expected, item in zip(c, walker):
assert_equal(expected, item)
assert_raises(StopIteration, next, walker)
return wig_sum
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('bed', help='bed file')
parser.add_argument('plus_wig', help='wig file for plus strand')
parser.add_argument('minus_wig', help='wig file for minus strand')
parser.add_argument('output_name', help='name of output file')
args = parser.parse_args()
plus_wig_dict = {
position: value
for region, position, value in fill(walk(open(args.plus_wig)))
}
minus_wig_dict = {
position: value
for region, position, value in fill(walk(open(args.minus_wig)))
}
outfile = open(args.output_name, 'w')
with open(args.bed) as infile:
for line in infile:
fields = line.strip().split()
chrom = fields[0]
start = int(fields[1])
end = int(fields[2])
name = fields[3]
import sys import wiggelen wig_fname = sys.argv[1] threshold = int(sys.argv[2]) # the lowest coverage of high-coverage region. # Read data frame from output of 'samtools depth'. wig_file = wiggelen.walk(open(wig_fname)) # Set the beginning state 'LOW'. state = 'LOW' reference_name = None # Search and print regions. for reference, position, value in wig_file: if state == 'LOW': if value >= threshold: reference_name = reference beg = position end = position state = 'HIGH' elif state == 'HIGH': # The next reference has started. if reference_name != reference: print(str(reference_name) + '\t' + str(beg) + '\t' + str(end)) state = 'LOW'
ofile = output_area + '/' + base + '.covered_bases.tsv'
of = open(ofile, 'wt')
L = fbed.readline().strip().split()
# assume order of bed and wig both in chromosome position order
n0 = 0
n1 = 0
pp1 = 0
pp2 = 0
c0 = 'A'
p1 = genome_pos(L[0], L[1])
p2 = genome_pos(L[0], L[2])
n = 0
LT = LT + p2 - p1
for x in wiggelen.walk(fw):
n = n + 1
wp = genome_pos(x[0], x[1])
if wp > p2: # write out interval line, read next interval (pp1,pp2,p1,p2), reset n1,n2
of.write('%s\t%s\t%d\t%d\t%d\n' %
(L[0], L[1], int(L[2]) - int(L[1]), n0, n1))
if (L[0] != c0):
print('%s\t%d' % (L[0], n))
c0 = L[0]
N0 = N0 + n0
N1 = N1 + n1
n0 = 0
n1 = x[2]
pp1 = p1
pp2 = p2
Modified by: Zaka Yuen, JCSMR, ANU
Created on Nov 2019
Tombo can only output methylation scores in a wig format.
This script is to:
-convert a wiggle format to a text format containing per-site scores for downstream analyses
"""
# Positive strand
r=[]
p=[]
v=[]
with open(snakemake.input[0],'r') as fh:
for region, position, value in wiggelen.walk(fh):
r.append(region)
p.append(position)
v.append(value)
d_plus = pd.DataFrame(list(zip(r,p,v)), columns =['names', 'start','values'])
d_plus["strand"]="+"
# Negative strand
r_m=[]
p_m=[]
v_m=[]
with open(snakemake.input[1],'r') as fh:
for region, position, value in wiggelen.walk(fh):
r_m.append(region)
p_m.append(position)
