def parse_biom(biom):
with open(biom) as fh:
bt = parse_table(fh)
df = bt.to_dataframe()
otu_df = [
["Samples", "OTUs", "OTU Total Count", "OTU Table Density"],
[
bt.length("sample"),
bt.length("observation"),
bt.sum(),
"{:f}".format(bt.get_table_density()),
],
]
otu_df = pd.DataFrame(otu_df)
# make the first row the column labels
otu_df.rename(columns=otu_df.iloc[0], inplace=True)
# drop the first row
otu_df = otu_df.reindex(otu_df.index.drop(0))
# build the table
summary_table = otu_df.to_html(
index=False,
bold_rows=False,
classes=["table", "table-bordered"],
table_id="otuTable",
)
# fix for rst
summary_table = summary_table.replace("\n", "\n" + 10 * " ")
return df, summary_table
def read_biom_file(self):
with open(self.args.biom_file) as f:
table = parse_table(f)
import pdb; pdb.set_trace()
pass
def mothur_counts_to_biom(mothur_f):
mothur_biom = parse_table(mothur_f)
mothur_biom.type = u'OTU table'
filter_biom = mothur_biom.filter(lambda val, id_, metadata: id_ in 'total',
invert=True)
return (filter_biom)
def get_biom_table_from_file(path):
with warnings.catch_warnings():
warnings.simplefilter("ignore")
from biom import parse_table
with open(path) as f:
table = parse_table(f)
return table
return None
def mothur_counts_to_biom(mothur_f):
mothur_biom = parse_table(mothur_f)
mothur_biom.type = u'OTU table'
filter_biom = mothur_biom.filter(
lambda val, id_, metadata: id_ in 'total', invert=True)
return(filter_biom)
def parse_table_to_biom(table_lines, table_format="tab-delimited",\
biom_format = 'otu table'):
"""Read the lines of an open trait table file, and output a .biom table object
The trait table must be either a biom file, or a picrust tab-delimited file
table_format -- must be either 'tab-delimited' or 'biom'
"""
return parse_table(table_lines)
def parse_table_to_biom(table_lines, table_format="tab-delimited",\
biom_format = 'otu table'):
"""Read the lines of an open trait table file, and output a .biom table object
The trait table must be either a biom file, or a picrust tab-delimited file
table_format -- must be either 'tab-delimited' or 'biom'
"""
return parse_table(table_lines)
def test_filter_contaminated_libraries(self):
"""Test filtering process for heavily contaminated libraries"""
test_biom = self.test_biom
exp_filtered_biom = parse_table(test_biom_file_filt)
contaminant_otus = set(["otu3", "contam1", "contam2", "contam4"])
obs_filtered_biom = filter_contaminated_libraries(test_biom, contaminant_otus, 0.8)
self.assertEqual(exp_filtered_biom, obs_filtered_biom)
def _read_biom(fp, transpose=True):
'''Read in a biom table file.
Parameters
----------
fp : str or file object
file path to the biom table
transpose : bool (True by default)
Transpose the table or not. The biom table has samples in
column while sklearn and other packages require samples in
row. So you should transpose the data table.
Returns
-------
sid : list of str
the sample ids
fid : list of str
the feature ids
data : numpy array (2d) of float
the table
feature_md : pandas.DataFrame
the feature metadata (if availble in table)
'''
logger.debug('loading biom table %s' % fp)
if hasattr(fp, 'read'):
table = biom.parse_table(fp)
else:
table = biom.load_table(fp)
sid = table.ids(axis='sample')
fid = table.ids(axis='observation')
logger.info('loaded %d samples, %d features' % (len(sid), len(fid)))
data = table.matrix_data
feature_md = _get_md_from_biom(table)
if transpose:
logger.debug('transposing table')
data = data.transpose()
return sid, fid, data, feature_md
#!/usr/bin/env python
import sys
from skbio import TreeNode
from biom import parse_table
tree_fp = sys.argv[1]
biom_fp = sys.argv[2]
tree = TreeNode.read(tree_fp)
biom = parse_table(open(biom_fp,'r'))
new_names = {}
for otu,sample in biom.nonzero():
if otu not in new_names:
new_names[otu] = '{0}_{1}'.format(otu,sample)
else:
new_names[otu] = '{0}_{1}'.format(new_names[otu],sample)
for tip in tree.tips():
otu = tip.name
tip.name = new_names[otu]
for node in tree.non_tips():
node.name = None
print tree
import numpy.core._methods
import numpy.lib.format
import numpy as np
from io import StringIO
# https://github.com/biocore/biom-format
# http://biom-format.org/documentation/quick_usage_examples.html#examples
from biom import parse_table, Table
dir = "C:/Projekti/PycharmProjects/UMCGMicrobiomeWeb/example_data"
path_tax = dir + '/example1_metaphlan.txt'
path_tax_marged = dir + 'example_metaphlan_merged.tsv'
with open(path_tax) as f:
table = parse_table(f)
print("version ", table.format_version)
print("id ", table.table_id)
print("dtype ", table.dtype)
print("type ", table.type)
print("version ", table.format_version)
print("generated_by ", table.generated_by)
print("table_id ", table.table_id)
print("ids ", table.ids()) # ['Metaphlan2_Analysis']
# print("ids ", table.to_json())
# with open('c:/tmp/metha.json', 'w') as file:
# file.write(table.to_json("janko"))
table = table.norm(axis='sample', inplace=False)
def setUp(self):
"""Init variables for the tests """
self.test_biom = parse_table(test_biom_file)
#!/usr/bin/env python
import sys
from skbio import TreeNode
from biom import parse_table
tree_fp = sys.argv[1]
biom_fp = sys.argv[2]
tree = TreeNode.read(tree_fp)
biom = parse_table(open(biom_fp, 'r'))
new_names = {}
for otu, sample in biom.nonzero():
if otu not in new_names:
new_names[otu] = '{0}_{1}'.format(otu, sample)
else:
new_names[otu] = '{0}_{1}'.format(new_names[otu], sample)
for tip in tree.tips():
otu = tip.name
tip.name = new_names[otu]
for node in tree.non_tips():
node.name = None
print tree
def load_biom(biom_filepath):
with open(biom_filepath) as f:
t = parse_table(f)
return t
#!/usr/bin/env python
import sys
from biom import parse_table
deblur_biom = parse_table(open(sys.argv[1], 'r'))
sample_counts = dict(zip(deblur_biom.ids(axis='sample'),[0] * len(deblur_biom.ids(axis='sample'))))
for seq,sample in deblur_biom.nonzero():
for i in range(int(deblur_biom.get_value_by_ids(seq,sample))):
sample_counts[sample] += 1
print ">{0}_{1}\n{2}\n".format(sample,sample_counts[sample],seq)
