def main(environment_file,
sample_mapping_file=None,
unit_mapping_file=None,
min_abundance=0,
min_sum_normalized_percent=1):
samples_dict = utils.get_samples_dict_from_environment_file(
environment_file)
oligos = utils.get_oligos_sorted_by_abundance(samples_dict,
min_abundance=min_abundance)
unit_counts, unit_percents = utils.get_unit_counts_and_percents(
oligos, samples_dict)
if sample_mapping_file:
sample_mapping = utils.get_sample_mapping_dict(sample_mapping_file)
if unit_mapping_file:
unit_mapping = utils.get_sample_mapping_dict(unit_mapping_file)
output_file = '.'.join(environment_file.split('.')[:-1]) + '.gexf'
utils.generate_gexf_network_file(
oligos,
samples_dict,
unit_percents,
output_file,
sample_mapping_dict=sample_mapping if sample_mapping_file else None,
unit_mapping_dict=unit_mapping if unit_mapping_file else None)
def main(
environment_file, sample_mapping_file=None, unit_mapping_file=None, min_abundance=0, min_sum_normalized_percent=1
):
samples_dict = utils.get_samples_dict_from_environment_file(environment_file)
oligos = utils.get_oligos_sorted_by_abundance(samples_dict, min_abundance=min_abundance)
unit_counts, unit_percents = utils.get_unit_counts_and_percents(oligos, samples_dict)
if sample_mapping_file:
sample_mapping = utils.get_sample_mapping_dict(sample_mapping_file)
if unit_mapping_file:
unit_mapping = utils.get_sample_mapping_dict(unit_mapping_file)
output_file = ".".join(environment_file.split(".")[:-1]) + ".gexf"
utils.generate_gexf_network_file(
oligos,
samples_dict,
unit_percents,
output_file,
sample_mapping_dict=sample_mapping if sample_mapping_file else None,
unit_mapping_dict=unit_mapping if unit_mapping_file else None,
)
from Oligotyping.utils.utils import get_samples_dict_from_environment_file
from Oligotyping.utils.utils import get_oligos_sorted_by_abundance
from Oligotyping.utils.utils import get_units_across_samples_dicts
from Oligotyping.utils.utils import get_unit_counts_and_percents
from Oligotyping.utils.cosine_similarity import get_oligotype_sets
from Oligotyping.utils.cosine_similarity import get_oligotype_sets_greedy
from Oligotyping.visualization.oligotype_distribution_stack_bar import oligotype_distribution_stack_bar
from Oligotyping.utils.utils import generate_ENVIRONMENT_file
input_file_path = sys.argv[1]
cosine_similarity_value = float(sys.argv[2])
sets_output_file_name = input_file_path + '-cos-%s-SETS' % cosine_similarity_value
environ_output_file_name = input_file_path + '-cos-%s-SETS-ENVIRON' % cosine_similarity_value
samples_dict = get_samples_dict_from_environment_file(input_file_path)
oligos = get_oligos_sorted_by_abundance(samples_dict)
unit_counts, unit_percents = get_unit_counts_and_percents(oligos, samples_dict)
samples = samples_dict.keys()
across_samples_sum_normalized, across_samples_max_normalized = get_units_across_samples_dicts(
oligos, samples_dict.keys(), unit_percents)
oligotype_sets = get_oligotype_sets_greedy(oligos,
across_samples_sum_normalized,
cosine_similarity_value,
sets_output_file_name)
print '%d sets from %d units' % (len(oligotype_sets), len(oligos))
samples_dict_with_agglomerated_oligos = {}
for sample in samples:
def oligotype_distribution_stack_bar(samples_dict, colors_dict, output_file = None, legend = False,\
colors_export = None, project_title = None, display = True, oligos = None):
samples = samples_dict.keys()
samples.sort()
if oligos == None:
oligos = get_oligos_sorted_by_abundance(samples_dict, oligos)
else:
oligos.reverse()
if colors_dict == None:
colors_dict = random_colors(copy.deepcopy(oligos))
samples_oligo_vectors = {}
for sample in samples:
vector = []
for oligo in oligos:
if samples_dict[sample].has_key(oligo):
vector.append(samples_dict[sample][oligo])
else:
vector.append(0)
samples_oligo_vectors[sample] = vector
samples_oligo_vectors_percent_normalized = {}
for sample in samples:
total_oligos_in_sample = sum(samples_oligo_vectors[sample])
vector = []
for oligo_abundance in samples_oligo_vectors[sample]:
vector.append(oligo_abundance * 100.0 / total_oligos_in_sample)
samples_oligo_vectors_percent_normalized[sample] = vector
# figure..
fig = plt.figure(figsize=(20, 10))
if legend:
plt.subplots_adjust(left=0.03, bottom=0.15, top=0.97, right=0.90)
else:
plt.subplots_adjust(left=0.03, bottom=0.15, top=0.97, right=0.99)
N = len(samples)
ind = np.arange(N)
width = 0.75
bars = []
colors_list = []
for i in range(0, len(oligos)):
values = [
samples_oligo_vectors_percent_normalized[sample][i]
for sample in samples
]
bottom = [
sum(samples_oligo_vectors_percent_normalized[sample][0:i])
for sample in samples
]
try:
color = HTMLColorToRGB(colors_dict[oligos[i]])
colors_list.append(colors_dict[oligos[i]])
except:
color = 'black'
colors_list.append('#000000')
p = plt.bar(ind, values, width, bottom=bottom, color=color)
bars.append(p)
if colors_export:
colors_list = reversed(colors_list)
colors_file = open(colors_export, 'w')
for oligo in oligos:
colors_file.write('%s\t%s\n' % (oligo, colors_dict[oligo]))
colors_file.close()
plt.ylabel('Oligotype Distribution', size='large')
plt.title('Stacked Bar Charts of Oligotype Distribution %s' \
% (('for "%s"' % project_title) if project_title else ''))
plt.xticks(ind + width / 2., samples, rotation=90, size='small')
plt.yticks([])
plt.ylim(ymax=100)
plt.xlim(xmin=-(width) / 2, xmax=len(samples))
if legend:
plt.legend([b[0] for b in bars][::-1],
oligos[::-1],
bbox_to_anchor=(1.01, 1),
loc=2,
borderaxespad=0.0,
shadow=True,
fancybox=True)
leg = plt.gca().get_legend()
ltext = leg.get_texts()
llines = leg.get_lines()
frame = leg.get_frame()
frame.set_facecolor('0.80')
plt.setp(ltext, fontsize='small', fontname='arial', family='monospace')
plt.setp(llines, linewidth=1.5)
if output_file:
plt.savefig(output_file)
if display:
try:
plt.show()
except:
pass
# -*- coding: utf-8 -*- # takes an environment file and a generates matching percent and count matrices. import sys from Oligotyping.utils.utils import get_samples_dict_from_environment_file from Oligotyping.utils.utils import get_oligos_sorted_by_abundance from Oligotyping.utils.utils import get_units_across_samples_dicts from Oligotyping.utils.utils import get_unit_counts_and_percents from Oligotyping.utils.utils import generate_MATRIX_files samples_dict = get_samples_dict_from_environment_file(sys.argv[1]) oligos = get_oligos_sorted_by_abundance(samples_dict) oligos.reverse() unit_counts, unit_percents = get_unit_counts_and_percents(oligos, samples_dict) samples = sorted(samples_dict.keys()) generate_MATRIX_files(oligos, samples, unit_counts, unit_percents, sys.argv[1] + '-MATRIX-COUNT', sys.argv[1] + '-MATRIX-PERCENT')
def oligotype_distribution_across_samples(samples_dict, colors_dict, output_file = None, legend = False, project_title = None, display = True, oligos = None):
samples = samples_dict.keys()
samples.sort()
if oligos == None:
oligos = get_oligos_sorted_by_abundance(samples_dict, oligos)
else:
oligos.reverse()
if colors_dict == None:
colors_dict = random_colors(copy.deepcopy(oligos))
oligo_percents = {}
max_normalized_across_samples_vectors = {}
sum_normalized_across_samples_vectors = {}
for oligo in oligos:
percents = []
for sample in samples:
if samples_dict[sample].has_key(oligo):
percents.append(samples_dict[sample][oligo] * 100.0 / sum(samples_dict[sample].values()))
else:
percents.append(0.0)
oligo_percents[oligo] = percents
for oligo in oligos:
max_normalized_across_samples_vectors[oligo] = [p * 100.0 / max(oligo_percents[oligo]) for p in oligo_percents[oligo]]
sum_normalized_across_samples_vectors[oligo] = [p * 100.0 / sum(oligo_percents[oligo]) for p in oligo_percents[oligo]]
# figure..
fig = plt.figure(figsize=(20, 10))
if legend:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.80)
else:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.99)
plt.rcParams.update({'axes.linewidth' : 0.1})
plt.rc('grid', color='0.70', linestyle='-', linewidth=0.1)
plt.grid(True)
plt.subplot(2, 1, 1)
plt.grid(True)
N = len(samples)
ind = np.arange(N)
width = 0.75
lines = []
for i in range(0, len(oligos)):
oligo = oligos[i]
try:
color = HTMLColorToRGB(colors_dict[oligos[i]])
except:
color = 'black'
if len(oligos) < 50:
plt.plot(max_normalized_across_samples_vectors[oligo], color=color, linewidth = 3, alpha = 0.3, zorder = i)
plt.plot(max_normalized_across_samples_vectors[oligo], color=color, linewidth = 5, alpha = 0.2, zorder = i)
p = plt.plot(max_normalized_across_samples_vectors[oligo], color=color, linewidth = 1, alpha = 0.9, zorder = i)
lines.append(p)
plt.ylabel('MAX Normalized', size='large')
plt.title('Normalized Oligotype Distributions Across Samples %s' \
% (('for "%s"' % project_title) if project_title else ''))
plt.xticks(ind, ['' for d in samples], rotation=90, size='small')
plt.yticks([])
plt.ylim(ymax = 100)
plt.xlim(xmin = -(width) / 2, xmax = len(samples) - 0.5)
if legend:
plt.legend([b[0] for b in lines][::-1], oligos[::-1], bbox_to_anchor=(1.01, 1), loc=2, borderaxespad=0.0, shadow=True, fancybox=True)
leg = plt.gca().get_legend()
ltext = leg.get_texts()
llines = leg.get_lines()
frame = leg.get_frame()
frame.set_facecolor('0.80')
plt.setp(ltext, fontsize='small', fontname='arial', family='monospace')
plt.setp(llines, linewidth=1.5)
plt.subplot(2, 1, 2)
if legend:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.80)
else:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.99)
plt.rcParams.update({'axes.linewidth' : 0.1})
plt.rc('grid', color='0.70', linestyle='-', linewidth=0.1)
plt.grid(True)
for i in range(0, len(oligos)):
oligo = oligos[i]
try:
color = HTMLColorToRGB(colors_dict[oligos[i]])
except:
color = 'black'
if len(oligos) < 50:
plt.plot(sum_normalized_across_samples_vectors[oligo], color=color, linewidth = 3, alpha = 0.3, zorder = i)
plt.plot(sum_normalized_across_samples_vectors[oligo], color=color, linewidth = 5, alpha = 0.2, zorder = i)
p = plt.plot(sum_normalized_across_samples_vectors[oligo], color=color, linewidth = 1, alpha = 0.9, zorder = i)
plt.ylabel('SUM Normalized', size='large')
plt.xticks(ind, samples, rotation=90, size='small')
plt.yticks([])
plt.ylim(ymax = 100)
plt.xlim(xmin = -(width) / 2, xmax = len(samples) - 0.5)
if output_file:
plt.savefig(output_file)
if display:
try:
plt.show()
except:
pass
def oligotype_distribution_stack_bar(samples_dict, colors_dict, output_file = None, legend = False,\
colors_export = None, project_title = None, display = True, oligos = None):
samples = samples_dict.keys()
samples.sort()
if oligos == None:
oligos = get_oligos_sorted_by_abundance(samples_dict, oligos)
else:
oligos.reverse()
if colors_dict == None:
colors_dict = random_colors(copy.deepcopy(oligos))
samples_oligo_vectors = {}
for sample in samples:
vector = []
for oligo in oligos:
if samples_dict[sample].has_key(oligo):
vector.append(samples_dict[sample][oligo])
else:
vector.append(0)
samples_oligo_vectors[sample] = vector
samples_oligo_vectors_percent_normalized = {}
for sample in samples:
total_oligos_in_sample = sum(samples_oligo_vectors[sample])
vector = []
for oligo_abundance in samples_oligo_vectors[sample]:
vector.append(oligo_abundance * 100.0 / total_oligos_in_sample)
samples_oligo_vectors_percent_normalized[sample] = vector
# figure..
fig = plt.figure(figsize=(20, 10))
if legend:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.90)
else:
plt.subplots_adjust(left=0.03, bottom = 0.15, top = 0.97, right = 0.99)
N = len(samples)
ind = np.arange(N)
width = 0.75
bars = []
colors_list = []
for i in range(0, len(oligos)):
values = [samples_oligo_vectors_percent_normalized[sample][i] for sample in samples]
bottom = [sum(samples_oligo_vectors_percent_normalized[sample][0:i]) for sample in samples]
try:
color = HTMLColorToRGB(colors_dict[oligos[i]])
colors_list.append(colors_dict[oligos[i]])
except:
color = 'black'
colors_list.append('#000000')
p = plt.bar(ind, values, width, bottom=bottom, color=color)
bars.append(p)
if colors_export:
colors_list = reversed(colors_list)
colors_file = open(colors_export, 'w')
for oligo in oligos:
colors_file.write('%s\t%s\n' % (oligo, colors_dict[oligo]))
colors_file.close()
plt.ylabel('Oligotype Distribution', size='large')
plt.title('Stacked Bar Charts of Oligotype Distribution %s' \
% (('for "%s"' % project_title) if project_title else ''))
plt.xticks(ind+width/2., samples, rotation=90, size='small')
plt.yticks([])
plt.ylim(ymax = 100)
plt.xlim(xmin = -(width) / 2, xmax = len(samples))
if legend:
plt.legend([b[0] for b in bars][::-1], oligos[::-1], bbox_to_anchor=(1.01, 1), loc=2, borderaxespad=0.0, shadow=True, fancybox=True)
leg = plt.gca().get_legend()
ltext = leg.get_texts()
llines = leg.get_lines()
frame = leg.get_frame()
frame.set_facecolor('0.80')
plt.setp(ltext, fontsize='small', fontname='arial', family='monospace')
plt.setp(llines, linewidth=1.5)
if output_file:
plt.savefig(output_file)
if display:
try:
plt.show()
except:
pass
