def test_map_to_2D_dict(self):
"""Checks map_to_2D_dict is sane"""
# Creates a pseudo-opening function
test_map = StringIO(
'#SampleID\tBIRTH_YEAR\tDEATH_YEAR\tSEX\tPROFESSION\tHOME_STATE\n'
'00010\t1954\t2006\tmale\tMechanic\tKansas\n'
'00100\t1954\t1983\tfemale\tHunter\tKansas\n'
'00200\tNA\t2009\tfemale\tNurse\tMinnesota\n'
'00111\t1979\t2007\tmale\tHunter\tImpala\n'
'00112\t1983\t2006\tmale\tHunter\tImpala\n'
'00211\t1990\t2009\tmale\tStudent\tMinnesota\n')
# Sets up the known dictionary
known_dict = {'00010': {'#SampleID': '00010', 'BIRTH_YEAR': '1954',
'DEATH_YEAR': '2006', 'SEX': 'male',
'PROFESSION': 'Mechanic', 'HOME_STATE':
'Kansas'},
'00100': {'#SampleID': '00100', 'BIRTH_YEAR': '1954',
'DEATH_YEAR': '1983', 'SEX': 'female',
'PROFESSION': 'Hunter', 'HOME_STATE':
'Kansas'},
'00200': {'#SampleID': '00200', 'BIRTH_YEAR': 'NA',
'DEATH_YEAR': '2009', 'SEX': 'female',
'PROFESSION': 'Nurse',
'HOME_STATE': 'Minnesota'},
'00111': {'#SampleID': '00111', 'BIRTH_YEAR': '1979',
'DEATH_YEAR': '2007', 'SEX': 'male',
'PROFESSION': 'Hunter', 'HOME_STATE':
'Impala'},
'00112': {'#SampleID': '00112', 'BIRTH_YEAR': '1983',
'DEATH_YEAR': '2006', 'SEX': 'male',
'PROFESSION': 'Hunter', 'HOME_STATE':
'Impala'},
'00211': {'#SampleID': '00211', 'BIRTH_YEAR': '1990',
'DEATH_YEAR': '2009', 'SEX': 'male',
'PROFESSION': 'Student', 'HOME_STATE':
'Minnesota'}}
# Checks the test dictionary is loaded properly and equals the known
test_dict = map_to_2D_dict(test_map)
self.assertEqual(test_dict, known_dict)
else:
with biom_open(args.input) as fp:
tax_table = parse_biom_table(fp)
# Checks the output directory is sane.
if not args.output:
parser.error('An output directory must be supplied.')
elif not exists(args.output):
mkdir(args.output)
output_dir = args.output
if args.mapping and not isfile(args.mapping):
parser.error('The supplied mapping file does not exist in the path.')
elif args.mapping:
mapping = map_to_2D_dict(open(args.mapping, 'U'))
else:
mapping = args.mapping
# Parses the sample IDs as a list
if args.samples:
samples_to_analyze = []
for sample in args.samples.split(','):
samples_to_analyze.append(sample)
else:
samples_to_analyze = None
main(taxa_table=tax_table,
output_dir=output_dir,
mapping=mapping,
samples_to_analyze=samples_to_analyze)
def main(otu_table,
mapping_data,
cat_tables,
output_dir,
sample_type='fecal',
samples_to_plot=None,
legend=False,
xaxis=True,
debug=False):
"""Creates stacked bar plots for an otu table
INPUTS:
otu_table -- an open OTU table
mapping_data -- a tab delimited string containing the mapping data
passed from the mapping file.
categories -- a dictionary keying a mapping category to the
corresponding biom table
output_dir -- the location of the directory where output files should
be saved. If this directory does not exist, it will be
created.
samples_to_plot -- a list of sample ids to plot. If no value is passed,
then all samples in the biom table are analyzed.
debug -- ignore properly handling Michael Pollan's sample
OUTPUTS:
A pdf of stacked taxonomy will be generated for each sample and saved
in the output directory. These will follow the file name format
Figure_4_<SAMPLEID>.pdf
"""
# Sets constants for analyzing the data
LEVEL = 2
CATEGORY = 'taxonomy'
NUM_TAXA = 9
NUM_CATS_TO_PLOT = 7
# Sets up plotting constants
COLORMAP = array([[0.8353, 0.2421, 0.3098], [0.9569, 0.4275, 0.2627],
[0.9922, 0.6824, 0.3804], [0.9961, 0.8784, 0.5351],
[0.9020, 0.9608, 0.5961], [0.6706, 0.8667, 0.6431],
[0.4000, 0.7608, 0.6471], [0.1961, 0.5333, 0.7412],
[0.3333, 0.3333, 0.3333]])
FIG_DIMS = (4.44444, 3.33333)
AXIS_DIMS = array([[0.05, 0.05], [0.95, 0.95]])
# Common taxa are designated before processing to remain constant.
COMMON_TAXA = [(u'k__Bacteria', u'p__Firmicutes'),
(u'k__Bacteria', u'p__Bacteroidetes'),
(u'k__Bacteria', u'p__Proteobacteria'),
(u'k__Bacteria', u'p__Actinobacteria'),
(u'k__Bacteria', u'p__Verrucomicrobia'),
(u'k__Bacteria', u'p__Tenericutes'),
(u'k__Bacteria', u'p__Cyanobacteria'),
(u'k__Bacteria', u'p__Fusobacteria')]
SKIPSET = set(('Sample', 'Average', 'MP'))
# Names categories being plotted
if sample_type == 'fecal':
michael_pollan = '10317.000007108'
cat_list = [
'You', 'Average', 'Similar Diet', ' Similar BMI', 'Same Gender',
'Similar Age', 'Michael Pollan'
]
order = [
'Sample', 'Average', 'DIET_TYPE', 'BMI_CAT', 'SEX', 'AGE_CAT', 'MP'
]
elif sample_type == 'skin':
michael_pollan = '10317.000007113'
cat_list = [
'You', 'Average', 'Similar Cosmetic Use', 'Same Dominant Hand',
'Same Gender', 'Same Age', 'Michael Pollan'
]
order = [
'Sample', 'Average', 'COSMETICS_FREQUENCY', 'DOMINANT_HAND', 'SEX',
'AGE_CAT', 'MP'
]
elif sample_type == 'oral':
michael_pollan = '10317.000007109'
cat_list = [
'You', 'Average', 'Similar Diet', 'Flossing Frequency',
'Same Gender', 'Same Age', 'Michael Pollan'
]
order = [
'Sample', 'Average', 'DIET_TYPE', 'FLOSSING_FREQUENCY', 'SEX',
'AGE_CAT', 'MP'
]
else:
raise ValueError('%s is not a supported sample type.' % sample_type)
# Gets the mapping file
map_dict = map_to_2D_dict(mapping_data)
# Gets the category file dictionary summarized with the common categories
# Generates the category file dictionary
categories = parse_category_files(raw_tables=cat_tables,
common_groups=COMMON_TAXA[:8],
level=LEVEL,
metadata=CATEGORY)
# Summarizes taxonomy for the category
(whole_sample_ids, whole_summary, new_common_taxa) = \
summarize_common_categories(biom_table=otu_table,
level=LEVEL,
common_categories=COMMON_TAXA[:8],
metadata_category=CATEGORY)
# Converts the final taxa to a cleaned up list
# Converts final taxa to a clean list
common_phyla = []
for taxon in new_common_taxa:
common_phyla.append(taxon[1].strip(' p__').strip('[').strip(']'))
new_common_taxa = common_phyla
# Checks that the crrect sample ids are plotted
if samples_to_plot is None:
sample_ids = whole_sample_ids
else:
sample_ids = samples_to_plot
if len(sample_ids) > 1:
# TODO: make the rest of the code reflect this...
raise ValueError("SCRIPT NO LONGER SUPPORTS MULTIPLE SAMPLES")
# Identifies Michael Pollan's pre-ABX sample
if debug:
mp_sample_pos = 2
else:
mp_sample_pos = whole_sample_ids.tolist().index(michael_pollan)
mp_sample_taxa = whole_summary[:, mp_sample_pos]
# Gets the table average
table_average = mean(whole_summary, 1)
# Generates a figure for each sample
for idx, sample_id in enumerate(whole_sample_ids):
if sample_id in sample_ids:
meta_data = map_dict[sample_id]
# Prealocates a numpy array to hold the data
tax_array = zeros((NUM_TAXA, NUM_CATS_TO_PLOT))
# Adds preset values to the array so the first column is the sample
# the second column is the average and the last column is Michael
# Pollan
tax_array[:, 0] = whole_summary[:, idx]
tax_array[:, 1] = table_average
tax_array[:, -1] = mp_sample_taxa
# Adds the categories to the table in the listed order
for idx, cat in enumerate(order):
# Skips over undesired categories
if cat in SKIPSET:
continue
# Gets the sample metadata
mapping_key = meta_data[cat]
# Pulls taxonomic summary and group descriptions
tax_summary = categories[cat]['Summary']
group_descriptions = categories[cat]['Groups'].tolist()
# Appends plotting tables
try:
mapping_col = group_descriptions.index(mapping_key)
except:
raise ValueError('The %s cannot be found in %s.' %
(mapping_key, cat))
tax_array[:, idx] = tax_summary[:, mapping_col]
# Sets up the file to save the data
filename = pjoin(output_dir, 'figure4.pdf')
# Plots the data
render_barchart(data_table=tax_array,
x_axis=False,
group_names=new_common_taxa,
legend=False,
sample_names=cat_list,
y_axis=False,
axis_dims=AXIS_DIMS,
fig_dims=FIG_DIMS,
file_out=filename,
show_edge=False,
colors=COLORMAP)
def main(otu_table, mapping_data, categories, output_dir, \
samples_to_plot = None, legend = False, xaxis = True):
"""Creates stacked bar plots for an otu table
INPUTS:
otu_table -- an open OTU table
mapping_data -- a tab delimited string containing the mapping data
passed from the mapping file.
categories -- a dictionary keying a mapping category to the
corresponding sample IDs and taxonomy for a collapsed
biom table
output_dir -- the location of the directory where output files should be
saved. If this directory does not exist, it will be created.
samples_to_plot -- a list of sample ids to plot. If no value is passed,
then all samples in the biom table are analyzed.
OUTPUTS:
A pdf of stacked taxonomy will be generated for each sample and saved
in the output directory. These will follow the file name format
Figure_4_<SAMPLEID>.pdf
"""
# Sets constants
LEVEL = 2
FILEPREFIX = 'Figure_4_'
MICHAEL_POLLAN = '000007108.1075657'
NUM_TAXA = 9
NUM_CATS_TO_PLOT = 7
# Loads the mapping file
map_dict = map_to_2D_dict(mapping_data)
(common_taxa, whole_sample_ids, whole_summary) = \
summarize_human_taxa(otu_table, LEVEL)
# Converts final taxa to a clean list
common_phyla = []
for taxon in common_taxa:
common_phyla.append(taxon[1].strip(' p__').strip('[').strip(']'))
common_taxa = common_phyla
# Checks that the correct sample ids are plotted
if samples_to_plot == None:
sample_ids = whole_sample_ids
else:
sample_ids = samples_to_plot
# Identifies Michael Pollan's pre-ABX sample
mp_sample_pos = whole_sample_ids.index(MICHAEL_POLLAN)
mp_sample_taxa = whole_summary[:,mp_sample_pos]
# Loads the category dictionary
categories = load_category_files(category_fp, LEVEL)
# Generates a figure for each sample
for idx, sample_id in enumerate(whole_sample_ids):
if sample_id in sample_ids:
# Preallocates a numpy array for the plotting data
tax_array = zeros((NUM_TAXA, NUM_CATS_TO_PLOT))
meta_data = map_dict[sample_id]
cat_list = ['You', 'Average', 'Similar Diet', ' Similar BMI',
'Same Gender', 'Similar Age',
'Michael Pollan', '']
#cat_list.append('Your Fecal Sample')
#cat_list.append('Average Fecal Samples')
tax_array[:,0] = whole_summary[:,idx]
tax_array[:,1] = mean(whole_summary, 1)
cat_watch = 2
# Identifies the appropriate metadata categories
for cat in categories:
# Pulls metadata for the sample and category
mapping_key = meta_data[cat]
# Pulls taxonomic summary and group descriptions for the category
tax_summary = categories[cat]['Taxa Summary']
group_descriptions = categories[cat]['Groups']
# Amends plotting tables
try:
mapping_col = group_descriptions.index(mapping_key)
except:
raise ValueError, 'The %s cannot be found in %s.' \
% (mapping_key, cat)
tax_array[:,cat_watch] = tax_summary[:,mapping_col]
cat_watch = cat_watch + 1
tax_array[:,-1] = mp_sample_taxa
# Plots the data
filename = pjoin(output_dir, '%s%s.pdf' \
% (FILEPREFIX, sample_id))
plot_american_gut(tax_array, filename)
def test_map_to_2D_dict(self):
"""Checks map_to_2D_dict is sane"""
# Creates a pseudo-opening function
test_map = StringIO(
'#SampleID\tBIRTH_YEAR\tDEATH_YEAR\tSEX\tPROFESSION\tHOME_STATE\n'
'00010\t1954\t2006\tmale\tMechanic\tKansas\n'
'00100\t1954\t1983\tfemale\tHunter\tKansas\n'
'00200\tNA\t2009\tfemale\tNurse\tMinnesota\n'
'00111\t1979\t2007\tmale\tHunter\tImpala\n'
'00112\t1983\t2006\tmale\tHunter\tImpala\n'
'00211\t1990\t2009\tmale\tStudent\tMinnesota\n')
# Sets up the known dictionary
known_dict = {
'00010': {
'#SampleID': '00010',
'BIRTH_YEAR': '1954',
'DEATH_YEAR': '2006',
'SEX': 'male',
'PROFESSION': 'Mechanic',
'HOME_STATE': 'Kansas'
},
'00100': {
'#SampleID': '00100',
'BIRTH_YEAR': '1954',
'DEATH_YEAR': '1983',
'SEX': 'female',
'PROFESSION': 'Hunter',
'HOME_STATE': 'Kansas'
},
'00200': {
'#SampleID': '00200',
'BIRTH_YEAR': 'NA',
'DEATH_YEAR': '2009',
'SEX': 'female',
'PROFESSION': 'Nurse',
'HOME_STATE': 'Minnesota'
},
'00111': {
'#SampleID': '00111',
'BIRTH_YEAR': '1979',
'DEATH_YEAR': '2007',
'SEX': 'male',
'PROFESSION': 'Hunter',
'HOME_STATE': 'Impala'
},
'00112': {
'#SampleID': '00112',
'BIRTH_YEAR': '1983',
'DEATH_YEAR': '2006',
'SEX': 'male',
'PROFESSION': 'Hunter',
'HOME_STATE': 'Impala'
},
'00211': {
'#SampleID': '00211',
'BIRTH_YEAR': '1990',
'DEATH_YEAR': '2009',
'SEX': 'male',
'PROFESSION': 'Student',
'HOME_STATE': 'Minnesota'
}
}
# Checks the test dictionary is loaded properly and equals the known
test_dict = map_to_2D_dict(test_map)
self.assertEqual(test_dict, known_dict)
elif not isfile(args.input):
parser.error("The supplied taxonomy file does not exist in the path.")
else:
with biom_open(args.input) as fp:
tax_table = parse_biom_table(fp)
# Checks the output directory is sane.
if not args.output:
parser.error("An output directory must be supplied.")
elif not exists(args.output):
mkdir(args.output)
output_dir = args.output
if args.mapping and not isfile(args.mapping):
parser.error("The supplied mapping file does not exist in the path.")
elif args.mapping:
mapping = map_to_2D_dict(open(args.mapping, "U"))
else:
mapping = args.mapping
# Parses the sample IDs as a list
if args.samples:
samples_to_analyze = []
for sample in args.samples.split(","):
samples_to_analyze.append(sample)
else:
samples_to_analyze = None
main(taxa_table=tax_table, output_dir=output_dir, mapping=mapping, samples_to_analyze=samples_to_analyze)
def main(otu_table, mapping_data, cat_tables, output_dir, sample_type='fecal',
samples_to_plot=None, legend=False, xaxis=True, debug=False):
"""Creates stacked bar plots for an otu table
INPUTS:
otu_table -- an open OTU table
mapping_data -- a tab delimited string containing the mapping data
passed from the mapping file.
categories -- a dictionary keying a mapping category to the
corresponding biom table
output_dir -- the location of the directory where output files should
be saved. If this directory does not exist, it will be
created.
samples_to_plot -- a list of sample ids to plot. If no value is passed,
then all samples in the biom table are analyzed.
debug -- ignore properly handling Michael Pollan's sample
OUTPUTS:
A pdf of stacked taxonomy will be generated for each sample and saved
in the output directory. These will follow the file name format
Figure_4_<SAMPLEID>.pdf
"""
# Sets constants for analyzing the data
LEVEL = 2
CATEGORY = 'taxonomy'
NUM_TAXA = 9
NUM_CATS_TO_PLOT = 7
# Sets up file name constants
FILEPREFIX = 'Figure_4_'
FILE_END = '.pdf'
# Sets up plotting constants
COLORMAP = array([[0.8353, 0.2421, 0.3098],
[0.9569, 0.4275, 0.2627],
[0.9922, 0.6824, 0.3804],
[0.9961, 0.8784, 0.5351],
[0.9020, 0.9608, 0.5961],
[0.6706, 0.8667, 0.6431],
[0.4000, 0.7608, 0.6471],
[0.1961, 0.5333, 0.7412],
[0.3333, 0.3333, 0.3333]])
FIG_DIMS = (4.44444, 3.33333)
AXIS_DIMS = array([[0.05, 0.05],
[0.95, 0.95]])
# Common taxa are designated before processing to remain constant.
COMMON_TAXA = [(u'k__Bacteria', u'p__Firmicutes'),
(u'k__Bacteria', u'p__Bacteroidetes'),
(u'k__Bacteria', u'p__Proteobacteria'),
(u'k__Bacteria', u'p__Actinobacteria'),
(u'k__Bacteria', u'p__Verrucomicrobia'),
(u'k__Bacteria', u'p__Tenericutes'),
(u'k__Bacteria', u'p__Cyanobacteria'),
(u'k__Bacteria', u'p__Fusobacteria')]
SKIPSET = set(('Sample', 'Average', 'MP'))
# Names categories being plotted
if sample_type == 'fecal':
michael_pollan = '000007108.1075657'
cat_list = ['You', 'Average', 'Similar Diet', ' Similar BMI',
'Same Gender', 'Similar Age', 'Michael Pollan']
order = ['Sample', 'Average', 'DIET_TYPE', 'BMI_CATEGORY', 'SEX',
'AGE_CATEGORY', 'MP']
elif sample_type == 'skin':
michael_pollan = '7113.1075702'
cat_list = ['You', 'Average', 'Similar Cosmetic Use',
'Same Dominant Hand', 'Same Gender', 'Same Age',
'Michael Pollan']
order = ['Sample', 'Average', 'COSMETICS_FREQUENCY',
'DOMINANT_HAND', 'SEX', 'AGE_CATEGORY', 'MP']
elif sample_type == 'oral':
michael_pollan = '7109.1075688'
cat_list = ['You', 'Average', 'Similar Diet', 'Flossing Frequency',
'Same Gender', 'Same Age', 'Michael Pollan']
order = ['Sample', 'Average', 'DIET_TYPE', 'FLOSSING_FREQUENCY',
'SEX', 'AGE_CATEGORY', 'MP']
else:
raise ValueError('%s is not a supported sample type.' % sample_type)
# Gets the mapping file
map_dict = map_to_2D_dict(mapping_data)
# Gets the category file dictionary summarized with the common categories
# Generates the category file dictionary
categories = parse_category_files(raw_tables=cat_tables,
common_groups=COMMON_TAXA[:8],
level=LEVEL,
metadata=CATEGORY)
# Summarizes taxonomy for the category
(whole_sample_ids, whole_summary, new_common_taxa) = \
summarize_common_categories(biom_table=otu_table,
level=LEVEL,
common_categories=COMMON_TAXA[:8],
metadata_category=CATEGORY)
# Converts the final taxa to a cleaned up list
# Converts final taxa to a clean list
common_phyla = []
for taxon in new_common_taxa:
common_phyla.append(taxon[1].strip(' p__').strip('[').strip(']'))
new_common_taxa = common_phyla
# Checks that the crrect sample ids are plotted
if samples_to_plot is None:
sample_ids = whole_sample_ids
else:
sample_ids = samples_to_plot
# Identifies Michael Pollan's pre-ABX sample
if debug:
mp_sample_pos = 2
else:
mp_sample_pos = whole_sample_ids.tolist().index(michael_pollan)
mp_sample_taxa = whole_summary[:, mp_sample_pos]
# Gets the table average
table_average = mean(whole_summary, 1)
# Generates a figure for each sample
for idx, sample_id in enumerate(whole_sample_ids):
if sample_id in sample_ids:
meta_data = map_dict[sample_id]
# Prealocates a numpy array to hold the data
tax_array = zeros((NUM_TAXA, NUM_CATS_TO_PLOT))
# Adds preset values to the array so the first column is the sample
# the second column is the average and the last column is Michael
# Pollan
tax_array[:, 0] = whole_summary[:, idx]
tax_array[:, 1] = table_average
tax_array[:, -1] = mp_sample_taxa
# Adds the categories to the table in the listed order
for idx, cat in enumerate(order):
# Skips over undesired categories
if cat in SKIPSET:
continue
# Gets the sample metadata
mapping_key = meta_data[cat]
# Pulls taxonomic summary and group descriptions
tax_summary = categories[cat]['Summary']
group_descriptions = categories[cat]['Groups'].tolist()
# Appends plotting tables
try:
mapping_col = group_descriptions.index(mapping_key)
except:
raise ValueError('The %s cannot be found in %s.'
% (mapping_key, cat))
tax_array[:, idx] = tax_summary[:, mapping_col]
# Sets up the file to save the data
filename = pjoin(output_dir, '%s%s%s'
% (FILEPREFIX, sample_id, FILE_END))
# Plots the data
render_barchart(data_table=tax_array,
x_axis=False,
group_names=new_common_taxa,
legend=False,
sample_names=cat_list,
y_axis=False,
axis_dims=AXIS_DIMS,
fig_dims=FIG_DIMS,
file_out=filename,
show_edge=False,
colors=COLORMAP)
