def convertBiomFileToStampProfile(self, file_name, output_name, metadata_name):
""" Function taken from PICRUSt by Morgan Langill.
https://github.com/mlangill/get_mgrast_data/blob/master/biom_to_stamp.py
"""
#allow file to be optionally gzipped (must use extension '.gz')
ext=splitext(file_name)[1]
if (ext == '.gz'):
table = parse_biom_table(gzip.open(file_name,'rb'))
else:
table = parse_biom_table(open(file_name,'U'))
metadata_name = metadata_name.split('(')[0].rstrip()
if metadata_name is None or metadata_name == '<observation ids>':
max_len_metadata = 0
elif table.observation_metadata and metadata_name in table.observation_metadata[0]:
#figure out the longest list within the given metadata
max_len_metadata = max(len(p[metadata_name]) for p in table.observation_metadata)
else:
QtGui.QMessageBox.information(self, 'Unrecognized metadata file', "'" + metadata_name + "' was not found in the BIOM table.", QtGui.QMessageBox.Ok)
return
#make the header line
header=[]
#make simple labels for each level in the metadata (e.g. 'Level_1', 'Level_2', etc.) "+1" for the observation id as well.
for i in range(max_len_metadata):
header.append('Level_'+ str(i+1))
header.append('Observation Ids')
#add the sample ids to the header line
header.extend(table.sample_ids)
fout = open(output_name, 'w')
fout.write("\t".join(header) + '\n')
#now process each observation (row in the table)
for obs_vals, obs_id, obs_metadata in table.iter(axis='observation'):
row=[]
if max_len_metadata > 0:
row = obs_metadata[metadata_name]
# add blanks if the metadata doesn't fill each level
if len(row) < max_len_metadata:
for i in range(max_len_metadata - len(row)):
row.append('unclassified')
#Add the observation id as the last "Level"
if isNumber(obs_id):
row.append('ID' + obs_id)
else:
row.append(obs_id)
#Add count data to the row
row.extend(map(str,obs_vals))
fout.write("\t".join(row) + '\n')
fout.close()
def determineColumns(self, data, profileTree):
firstDataRow = data[1].split('\t')
# first column entry that is numeric is assumed to be from first sample
firstSampleIndex = 0
for entry in firstDataRow:
if isNumber(entry):
break
firstSampleIndex += 1
# get hierarchical and sample names
headings = data[0].split('\t')
headings = map(string.strip, headings)
profileTree.hierarchyHeadings = headings[0:firstSampleIndex]
profileTree.sampleNames = headings[firstSampleIndex:]
def determineColumns(self, data, profileTree):
firstDataRow = data[1].split('\t')
# first column entry that is numeric is assumed to be from first sample
firstSampleIndex = 0
for entry in firstDataRow:
if isNumber(entry):
break
firstSampleIndex += 1
# get hierarchical and sample names
headings = data[0].split('\t')
headings = map(string.strip, headings)
profileTree.hierarchyHeadings = headings[0:firstSampleIndex]
profileTree.sampleNames = headings[firstSampleIndex:]
def sort(self, Ncol, order):
'''
Sort table by given column number.
'''
if len(self.arraydata) == 0:
return
self.emit(QtCore.SIGNAL("layoutAboutToBeChanged()"))
dataIsNumeric = isNumber(self.arraydata[0][Ncol])
if dataIsNumeric:
self.arraydata = SortTableNumericStrCol(self.arraydata, Ncol)
else:
self.arraydata = SortTableStrCol(self.arraydata, Ncol)
if order == QtCore.Qt.DescendingOrder:
self.arraydata.reverse()
self.emit(QtCore.SIGNAL("layoutChanged()"))
def sort(self, Ncol, order):
'''
Sort table by given column number.
'''
if len(self.arraydata) == 0:
return
self.emit(QtCore.SIGNAL("layoutAboutToBeChanged()"))
dataIsNumeric = isNumber(self.arraydata[0][Ncol])
if dataIsNumeric:
self.arraydata = SortTableNumericStrCol(self.arraydata, Ncol)
else:
self.arraydata = SortTableStrCol(self.arraydata, Ncol)
if order == QtCore.Qt.DescendingOrder:
self.arraydata.reverse()
self.emit(QtCore.SIGNAL("layoutChanged()"))
def convertBiomFileToStampProfile(self, file_name, output_name,
metadata_name):
""" Function taken from PICRUSt by Morgan Langill.
https://github.com/mlangill/get_mgrast_data/blob/master/biom_to_stamp.py
"""
#allow file to be optionally gzipped (must use extension '.gz')
ext = splitext(file_name)[1]
if (ext == '.gz'):
table = parse_biom_table(gzip.open(file_name, 'rb'))
else:
table = parse_biom_table(open(file_name, 'U'))
metadata_name = metadata_name.split('(')[0].rstrip()
if metadata_name is None or metadata_name == '<observation ids>':
max_len_metadata = 0
elif table.observation_metadata and metadata_name in table.observation_metadata[
0]:
#figure out the longest list within the given metadata
max_len_metadata = max(
len(p[metadata_name]) for p in table.observation_metadata)
else:
QtGui.QMessageBox.information(
self, 'Unrecognized metadata file',
"'" + metadata_name + "' was not found in the BIOM table.",
QtGui.QMessageBox.Ok)
return
#make the header line
header = []
#make simple labels for each level in the metadata (e.g. 'Level_1', 'Level_2', etc.) "+1" for the observation id as well.
for i in range(max_len_metadata):
header.append('Level_' + str(i + 1))
header.append('Observation Ids')
#add the sample ids to the header line
header.extend(table.sample_ids)
fout = open(output_name, 'w')
fout.write("\t".join(header) + '\n')
#now process each observation (row in the table)
for obs_vals, obs_id, obs_metadata in table.iter(axis='observation'):
row = []
if max_len_metadata > 0:
row = obs_metadata[metadata_name]
# add blanks if the metadata doesn't fill each level
if len(row) < max_len_metadata:
for i in range(max_len_metadata - len(row)):
row.append('unclassified')
#Add the observation id as the last "Level"
if isNumber(obs_id):
row.append('ID' + obs_id)
else:
row.append(obs_id)
#Add count data to the row
row.extend(map(str, obs_vals))
fout.write("\t".join(row) + '\n')
fout.close()
