def create_maf_distribution(
seqs, distrib_fhand=None, plot_fhand=None, summary_fhand=None, groups=None, group_kind=None
):
"It creates the distribution of the maf (not takes in account ref allele)"
title = "maf"
if groups and group_kind:
title = "maf (%s: %s)" % (group_kind, ",".join(groups))
mafs = CachedArray("f")
for seq in seqs:
for snv in seq.get_features("snv"):
maf = calculate_maf_frequency(snv, groups=groups, group_kind=group_kind)
if maf:
mafs.append(maf)
if list(mafs):
create_distribution(
mafs,
labels={"title": title},
distrib_fhand=distrib_fhand,
bins=None,
plot_fhand=plot_fhand,
range_=None,
summary_fhand=summary_fhand,
calculate_freqs=False,
remove_outliers=False,
)
def create_het_distribution(
seqs, distrib_fhand=None, plot_fhand=None, summary_fhand=None, group_kind=None, groups=None, ploidy=2
):
"""It creates the distribution of the heterozygosity
(not takes in account ref allele)"""
title = "heterozygosity"
if groups and group_kind:
title = "heterozygosity (%s: %s)" % (group_kind, ",".join(groups))
hets = CachedArray("f")
for seq in seqs:
for snv in seq.get_features("snv"):
if not group_kind and "heterozygosity" in snv.qualifiers:
het = snv.qualifiers["heterozygosity"]
else:
het = calculate_heterozygosity(snv, ploidy, group_kind=group_kind, groups=groups)
if het is not None:
hets.append(het)
if list(hets):
create_distribution(
hets,
labels={"title": title},
distrib_fhand=distrib_fhand,
bins=None,
plot_fhand=plot_fhand,
range_=None,
summary_fhand=summary_fhand,
calculate_freqs=False,
remove_outliers=False,
)
def create_pic_distribution(
seqs, distrib_fhand=None, plot_fhand=None, summary_fhand=None, read_groups=None, group_kind=None, groups=None
):
"It creates the distribution of the pic (not takes in account ref allele)"
title = "pic"
if groups and group_kind:
title = "pic (%s: %s)" % (group_kind, ",".join(groups))
pics = CachedArray("f")
for seq in seqs:
for snv in seq.get_features("snv"):
if not group_kind and "pic" in snv.qualifiers:
pic = snv.qualifiers["pic"]
else:
pic = calculate_pic(snv, group_kind=group_kind, groups=groups)
if pic is not None:
pics.append(pic)
if list(pics):
create_distribution(
pics,
labels={"title": title},
distrib_fhand=distrib_fhand,
bins=None,
plot_fhand=plot_fhand,
range_=None,
summary_fhand=summary_fhand,
calculate_freqs=False,
remove_outliers=False,
)
def test_basic_distribution(self):
'It tests the distribution'
summary_fhand = StringIO()
distrib_fhand = StringIO()
numbers = CachedArray(typecode='I')
numbers.extend([1, 2, 3, 4, 5, 6, 7, 8, 9, 101, 2, 3, 4, 5, 6, 7, 8, 9])
create_distribution(numbers, distrib_fhand=distrib_fhand,
summary_fhand=summary_fhand)
result = '''Statistics for histogram
-------------------------
minimum: 1
maximum: 101
average: 10.5556
variance: 486.9136
sum: 190
items: 18'''
assert result in summary_fhand.getvalue()
def main():
'''The main section'''
# Get parameters
infhand, outfhand, do_incompat, low_memory = set_parameters()
bins = 20
range_ = (95, 100)
# Parse blast results
blasts = BlastParser(infhand)
# The values for the distribution
score_keys = ['similarity']
if do_incompat:
score_keys.append('d_incompatibility')
scores = alignment_results_scores(blasts, score_keys)
# The distribution
if do_incompat:
#distrib, x_edges, y_edges = numpy.histogram2d(scores[0], scores[1],
# bins=bins)
distrib = numpy.histogram2d(scores[0], scores[1], bins=bins)[0]
else:
result = create_distribution(scores, range=range_, bins=bins,
low_memory=low_memory)
distrib = result['distrib'][0]
bin_edges = result['bin_edges'][0]
# The drawing
if do_incompat:
#fig = pylab.figure()
pylab.figure()
#axes = Axes3D(fig)
#axes.plot_surface(x_edges[:-1], y_edges[:-1], distrib)
#axes = pylab.subplot(111)
pylab.subplot(111)
image = pylab.imshow(distrib)
image.set_interpolation('bilinear')
pylab.show()
else:
draw_scatter(x_axe=bin_edges[:-1], y_axe=distrib, fhand=outfhand)
return
def bam_distribs(bam_fhand, kind, basename=None, range_=None,
grouping=None, sample_size=None, summary_fhand=None,
labels=None, plot_file_format='svg'):
'''It makes the bam coverage distribution.
It can make the distribution taking into account any of the readgroup items:
platform, sample and library
'''
value_calculator = {'coverage':_get_bam_coverage,
'mapq':_get_bam_mapping_quality,
'edit_distance': _get_bam_edit_distance}
coverage_labels = {'title': "Coverage for %s %s",
'xlabel': 'Coverage',
'ylabel': 'Num. of positions',
'sum':None,
'items':'total sequence length'
}
mapping_labels = {'title': "Mapping qualities for %s %s",
'xlabel': "mapping quality",
'ylabel': 'Num. of reads',
'sum':None, 'items':'number reads in the sam file'
}
edit_distance_labels = {'title': "Edit distances for %s %s",
'xlabel': "edit distance",
'ylabel': 'Num. of reads',
'sum':None, 'items':'number reads in the sam file'
}
plot_labels = {'coverage': coverage_labels,
'mapq':mapping_labels,
'edit_distance': edit_distance_labels}
if sample_size is not None:
sampled_bam_fhand = NamedTemporaryFile(suffix='.bam')
sample_bam(bam_fhand, sampled_bam_fhand, sample_size)
sample_fpath = sampled_bam_fhand.name
else:
sample_fpath = bam_fhand.name
create_bam_index(bam_fpath=sample_fpath)
bam = pysam.Samfile(sample_fpath, 'rb')
rgs = get_read_group_info(bam)
if grouping is None:
platforms = set([rg['PL'] for rg in rgs.values()])
if len(platforms) > 1:
grouping = 'PL'
else:
grouping = 'SM'
item_values = value_calculator[kind](bam, rgs, grouping)
results = {}
for group_name, values in item_values.items():
if basename is None:
distrib_fhand = None
plot_fhand = None
else:
distrib_fhand = open('%s.%s_%s.dat' % (basename, kind, group_name),
'w')
plot_fhand = open('%s.%s_%s.%s' % (basename, kind, group_name,
plot_file_format),
'w')
if grouping == 'PL':
grouping = 'platform'
elif grouping == 'SM':
grouping = 'sample'
labels = copy.deepcopy(plot_labels[kind])
labels['title'] = labels['title'] % (grouping, group_name)
remove_outliers = True if kind == 'coverage' else False
distrib = create_distribution(values, labels=labels,
distrib_fhand=distrib_fhand,
plot_fhand=plot_fhand,
range_=range_,
summary_fhand=summary_fhand,
remove_outliers=remove_outliers)
results[(grouping, group_name)] = distrib
return results
