def entropy_analysis(alignment_path, output_file = None, verbose = True, uniqued = False, freq_from_defline = None, weighted = False, qual_stats_dict = None, amino_acid_sequences = False):
if freq_from_defline == None:
freq_from_defline = lambda x: int([t.split(':')[1] for t in x.split('|') if t.startswith('freq')][0])
lines = []
previous_alignment_length = None
progress = Progress()
progress.verbose = verbose
alignment = u.SequenceSource(alignment_path)
progress.new('Processing the Alignment')
# processing the alignment file..
while alignment.next():
# check the alignment lengths along the way:
if previous_alignment_length:
if previous_alignment_length != len(alignment.seq):
raise EntropyError, "Not all reads have the same length."
# print out process info
if alignment.pos % 10000 == 0:
progress.update('Reads processed: %s' % (pretty_print(alignment.pos)))
# fill 'lines' variable
if not uniqued:
lines.append(alignment.seq)
else:
try:
frequency = freq_from_defline(alignment.id)
except IndexError:
raise EntropyError, "Reads declared as unique, but they do not have proper deflines. See help for --uniqued."
for i in range(0, frequency):
lines.append(alignment.seq)
previous_alignment_length = len(alignment.seq)
progress.end()
if verbose:
run.info('Number of reads', pretty_print(alignment.pos))
alignment.close()
# entropy analysis
progress.new('Entropy Analysis')
entropy_tpls = []
for position in range(0, len(lines[0])):
progress.update(P(int(position + 1), len(lines[0])))
if len(set([x[position] for x in lines])) == 1:
entropy_tpls.append((position, 0.0),)
else:
column = "".join([x[position] for x in lines])
if weighted:
if not qual_stats_dict:
raise EntropyError, "Weighted entropy is selected, but no qual stats are provided"
e = entropy(column, l_qual = qual_stats_dict[position], amino_acid_sequences = amino_acid_sequences)
else:
e = entropy(column, amino_acid_sequences = amino_acid_sequences)
if e < 0.00001:
entropy_tpls.append((position, 0.0),)
else:
entropy_tpls.append((position, e),)
sorted_entropy_tpls = sorted(entropy_tpls, key=operator.itemgetter(1), reverse=True)
progress.end()
if verbose:
entropy_components_larger_than_0 = [e[1] for e in entropy_tpls if e[1] > 0]
if entropy_components_larger_than_0:
run.info('Entropy analysis', 'Done (total of %d components greater than 0, mean: %.2f, max: %.2f, min: %.2f).' \
% (len(entropy_components_larger_than_0),
numpy.mean(entropy_components_larger_than_0),
numpy.max(entropy_components_larger_than_0),
numpy.min(entropy_components_larger_than_0)))
else:
run.info('Entropy analysis', 'None of the nucleotide positions posessed any entropy!')
if output_file:
entropy_output = open(output_file, 'w')
for _component, _entropy in sorted_entropy_tpls:
entropy_output.write('%d\t%.4f\n' % (_component, _entropy))
if verbose:
run.info('Entropy analysis output file path', output_file)
entropy_output.close()
return [x[1] for x in entropy_tpls]
def entropy_analysis(alignment_path,
output_file=None,
verbose=True,
uniqued=False,
freq_from_defline=None,
weighted=False,
qual_stats_dict=None,
amino_acid_sequences=False):
if freq_from_defline == None:
freq_from_defline = lambda x: int(
[t.split(':')[1] for t in x.split('|') if t.startswith('freq')][0])
lines = []
previous_alignment_length = None
progress = Progress()
progress.verbose = verbose
alignment = u.SequenceSource(alignment_path)
progress.new('Processing the Alignment')
# processing the alignment file..
while alignment.next():
# check the alignment lengths along the way:
if previous_alignment_length:
if previous_alignment_length != len(alignment.seq):
raise EntropyError, "Not all reads have the same length."
# print out process info
if alignment.pos % 10000 == 0:
progress.update('Reads processed: %s' %
(pretty_print(alignment.pos)))
# fill 'lines' variable
if not uniqued:
lines.append(alignment.seq)
else:
try:
frequency = freq_from_defline(alignment.id)
except IndexError:
raise EntropyError, "Reads declared as unique, but they do not have proper deflines. See help for --uniqued."
for i in range(0, frequency):
lines.append(alignment.seq)
previous_alignment_length = len(alignment.seq)
progress.end()
if verbose:
run.info('Number of reads', pretty_print(alignment.pos))
alignment.close()
# entropy analysis
progress.new('Entropy Analysis')
entropy_tpls = []
for position in range(0, len(lines[0])):
progress.update(P(int(position + 1), len(lines[0])))
if len(set([x[position] for x in lines])) == 1:
entropy_tpls.append((position, 0.0), )
else:
column = "".join([x[position] for x in lines])
if weighted:
if not qual_stats_dict:
raise EntropyError, "Weighted entropy is selected, but no qual stats are provided"
e = entropy(column,
l_qual=qual_stats_dict[position],
amino_acid_sequences=amino_acid_sequences)
else:
e = entropy(column, amino_acid_sequences=amino_acid_sequences)
if e < 0.00001:
entropy_tpls.append((position, 0.0), )
else:
entropy_tpls.append((position, e), )
sorted_entropy_tpls = sorted(entropy_tpls,
key=operator.itemgetter(1),
reverse=True)
progress.end()
if verbose:
entropy_components_larger_than_0 = [
e[1] for e in entropy_tpls if e[1] > 0
]
if entropy_components_larger_than_0:
run.info('Entropy analysis', 'Done (total of %d components greater than 0, mean: %.2f, max: %.2f, min: %.2f).' \
% (len(entropy_components_larger_than_0),
numpy.mean(entropy_components_larger_than_0),
numpy.max(entropy_components_larger_than_0),
numpy.min(entropy_components_larger_than_0)))
else:
run.info('Entropy analysis',
'None of the nucleotide positions posessed any entropy!')
if output_file:
entropy_output = open(output_file, 'w')
for _component, _entropy in sorted_entropy_tpls:
entropy_output.write('%d\t%.4f\n' % (_component, _entropy))
if verbose:
run.info('Entropy analysis output file path', output_file)
entropy_output.close()
return [x[1] for x in entropy_tpls]
#
# Please read the COPYING file.
import numpy as np
from scipy import log2 as log
import matplotlib.pyplot as plt
from Oligotyping.utils.random_colors import get_list_of_colors
from Oligotyping.utils.utils import get_unique_sequences_from_FASTA
from Oligotyping.utils.utils import Progress
from Oligotyping.utils.utils import Run
from Oligotyping.utils.utils import NUCL_COLORS
run = Run()
progress = Progress()
def entropy_distribution_bar(alignment, entropy_values, output_file, quick = False, no_display = False, qual_stats_dict = None, weighted = False, verbose = False):
progress.verbose = verbose
progress.new('Entropy Distribution Figure')
progress.update('Computing ')
y_maximum = max(entropy_values) + (max(entropy_values) / 10.0)
y_maximum = 1 if y_maximum < 1 else y_maximum
number_of_uniques_to_show = int(y_maximum * 100)
if alignment == None:
quick = True