def test_nucl_per_position():
'It calculates the frec of each nucleotide per position'
seqs = []
seqs.append(SeqWithQuality(Seq('ACTG')))
seqs.append(SeqWithQuality(Seq('ACTG')))
seqs.append(SeqWithQuality(Seq('CATG')))
seqs.append(SeqWithQuality(Seq('CATGZ')))
seqs.append(SeqWithQuality(Seq('ACTT')))
seqs.append(SeqWithQuality(Seq('ACTTZT')))
fhand = NamedTemporaryFile(suffix='.svg')
stats = create_nucleotide_freq_histogram(seqs, fhand, title='test')
assert stats == {
'A': [0.66666666666666663, 0.33333333333333331, 0.0, 0.0, 0, 0.0],
'C': [0.33333333333333331, 0.66666666666666663, 0.0, 0.0, 0, 0.0],
'T': [0.0, 0.0, 1.0, 0.33333333333333331, 0, 1.0],
'G': [0.0, 0.0, 0.0, 0.66666666666666663, 0, 0.0]}
fhand.flush()
svg = open(fhand.name, 'r').read()
assert '<!-- Created with matplotlib (http://matplotlib' in svg
def _do_seq_distrib_for_pair(self, pair):
'It does the distribution for a pair of cleaned and raw seqs'
get_stats_dir = lambda seq_type: os.path.join(self._get_project_path(),
BACKBONE_DIRECTORIES['%s_reads_stats' % seq_type])
#the statistics for both clean and raw sequences
lengths = {}
quals = {}
for seq_type in ('raw', 'cleaned'):
stats_dir = get_stats_dir(seq_type)
stats_fpath = os.path.join(stats_dir,
BACKBONE_BASENAMES['statistics_file'])
stats_fhand = open(stats_fpath, 'a')
if seq_type in pair:
fpath = pair[seq_type].last_version
basename = pair[seq_type].basename
# nucleotide freq per position
out_fpath = os.path.join(stats_dir, basename + '.freq_position')
if not os.path.exists(out_fpath):
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
seqs = seqs_in_file(open(fpath))
create_nucleotide_freq_histogram(seqs,
fhand=open(plot_fpath, 'w'))
#the names for the output files
out_fpath = os.path.join(stats_dir, basename + '.length')
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = out_fpath + '.dat'
if os.path.exists(plot_fpath):
continue
lengths_, quals_ = self._get_lengths_quals_from_file(fpath)
lengths[seq_type] = lengths_
quals[seq_type] = quals_
#the distributions for the lengths
distrib = lengths_.calculate_distribution()
lengths_.draw_distribution(distrib, labels=PLOT_LABELS['seq_length'],
distrib_fhand=open(distrib_fpath, 'w'),
plot_fhand=open(plot_fpath, 'w'))
#the distributions for the quals
out_fpath = os.path.join(stats_dir, basename + '.qual')
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = out_fpath + '.dat'
if quals_.count != 0:
distrib = quals_.calculate_distribution()
quals_.draw_distribution(distrib,
labels=PLOT_LABELS['seq_qual'],
plot_fhand=open(plot_fpath, 'w'),
distrib_fhand=open(distrib_fpath, 'w'))
#the statistics for the statistics file
self._write_statistics(stats_fhand, fpath, lengths_, quals_)
#the statistics for the differences
if 'raw' in pair and 'cleaned' in pair:
fpath = pair['cleaned'].last_version
basename = pair['cleaned'].basename
#the names for the output files
stats_dir = get_stats_dir('cleaned')
out_fpath = os.path.join(stats_dir, basename + '.length.diff')
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = out_fpath + '.dat'
if not os.path.exists(plot_fpath) and lengths:
#the distributions for the lengths
lengths = lengths['raw'], lengths['cleaned']
self._do_diff_distrib_for_numbers(lengths,
plot_fhand= open(plot_fpath, 'w'),
distrib_fhand= open(distrib_fpath, 'w'),
dist_type='seq_length')
del lengths
#the distributions for the quals
out_fpath = os.path.join(stats_dir, basename + '.qual.diff')
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = out_fpath + '.dat'
quals = quals['raw'], quals['cleaned']
if quals[0].count != 0 and quals[1].count != 0:
self._do_diff_distrib_for_numbers(quals,
plot_fhand= open(plot_fpath, 'w'),
distrib_fhand= open(distrib_fpath, 'w'),
dist_type='seq_qual')
del quals
for seq_type in ('raw', 'cleaned'):
if seq_type in pair:
stats_dir = get_stats_dir(seq_type)
fpath = pair[seq_type].last_version
basename = pair[seq_type].basename
#the names for the output files
out_fpath = os.path.join(stats_dir, basename + '.qual.boxplot')
plot_fpath = out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = out_fpath + '.dat'
if os.path.exists(plot_fpath):
continue
quals_ = self._get_quals_by_length_from_file(fpath)
if quals_ and quals_[0]:
#boxplot
draw_boxplot(quals_, fhand=open(plot_fpath, 'w'),
title=PLOT_LABELS['qual_boxplot']['title'],
xlabel=PLOT_LABELS['qual_boxplot']['xlabel'],
ylabel=PLOT_LABELS['qual_boxplot']['ylabel'],
stats_fhand=open(distrib_fpath, 'w'),
max_plotted_boxes=30)
del quals_
def _do_stats_for_file(self, path, stats_dir):
'It calculates the stats for a file os seqs'
stats_fpath = os.path.join(stats_dir,
BACKBONE_BASENAMES['statistics_file'])
stats_fhand = open(stats_fpath, 'a')
fpath = path.last_version
basename = path.basename
#output_files
freq_pos_out_fpath = os.path.join(stats_dir,
basename + '.freq_position')
length_out_fpath = os.path.join(stats_dir, basename + '.length')
qual_out_fpath = os.path.join(stats_dir, basename + '.qual')
qual_boxplot_out_fpath = os.path.join(stats_dir,
basename + '.qual.boxplot')
# nucleotide freq per position
if not os.path.exists(freq_pos_out_fpath):
plot_fpath = freq_pos_out_fpath + '.' + PLOT_FILE_FORMAT
seqs = seqs_in_file(open(fpath))
create_nucleotide_freq_histogram(seqs, fhand=open(plot_fpath, 'w'))
#Extract lengths and quals
if (not os.path.exists(length_out_fpath) or
not os.path.exists(qual_out_fpath)):
lengths_, quals_ = self._get_lengths_quals_from_file(fpath)
#the distributions for the lengths
if not os.path.exists(length_out_fpath):
plot_fpath = length_out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = length_out_fpath + '.dat'
distrib = lengths_.calculate_distribution()
lengths_.draw_distribution(distrib,
labels=PLOT_LABELS['seq_length'],
distrib_fhand=open(distrib_fpath, 'w'),
plot_fhand=open(plot_fpath, 'w'))
#the distributions for the quals
if not os.path.exists(qual_out_fpath) or quals_.count != 0:
plot_fpath = qual_out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = qual_out_fpath + '.dat'
distrib = quals_.calculate_distribution()
quals_.draw_distribution(distrib,
labels=PLOT_LABELS['seq_qual'],
plot_fhand=open(plot_fpath, 'w'),
distrib_fhand=open(distrib_fpath, 'w'))
# qual boxplot
if not os.path.exists(qual_boxplot_out_fpath):
quals = self._get_quals_by_length_from_file(fpath)
if (quals and quals[0]):
plot_fpath = qual_boxplot_out_fpath + '.' + PLOT_FILE_FORMAT
distrib_fpath = qual_boxplot_out_fpath + '.dat'
#boxplot
draw_boxplot(quals, fhand=open(plot_fpath, 'w'),
title=PLOT_LABELS['qual_boxplot']['title'],
xlabel=PLOT_LABELS['qual_boxplot']['xlabel'],
ylabel=PLOT_LABELS['qual_boxplot']['ylabel'],
stats_fhand=open(distrib_fpath, 'w'),
max_plotted_boxes=30)
#the statistics for the statistics file
self._write_statistics(stats_fhand, fpath, lengths_, quals_)
