def test_is_latin(self):
self.assertTrue(is_latin('Escherichia coli'))
self.assertTrue(is_latin('Rickettsia felis'))
self.assertFalse(is_latin('Enterobacteriaceae'))
self.assertFalse(is_latin('Escherichia coli O157:H7'))
self.assertFalse(is_latin('Citrobacter sp. A293'))
self.assertFalse(is_latin('bacterium LF-3'))
self.assertFalse(is_latin('Firmicutes bacterium CAG:129'))
self.assertFalse(is_latin('unidentified enterobacterium'))
self.assertFalse(is_latin('Staphylococcus aureus subsp. aureus'))
def identify_taxonomy(self):
"""Identify taxonomy of genomes.
"""
print('Identifying taxonomy of genomes...')
n = self.df.shape[0]
def report_diff(msg):
nonlocal n
n_ = self.df.shape[0]
if n_ < n:
print(' ' + msg.format(n - n_))
n = n_
# remove non-capitalized organism names
if self.capital:
self.df = self.df[self.df['organism_name'].apply(is_capital)]
report_diff('Dropped {} genomes without captalized organism name.')
# block certain words in organism names
if self.block:
self.block = list_from_param(self.block)
self.df = self.df[~self.df['organism_name'].
apply(contain_words, args=(self.block, ))]
report_diff('Dropped {} genomes with one or more blocked words in '
'organism name.')
# remove original species information
self.df.drop(columns=['species_taxid'], inplace=True)
# drop genomes whose taxIds are not in taxdump
self.df.dropna(subset=['taxid'], inplace=True)
self.df['taxid'] = self.df['taxid'].astype(str)
self.df = self.df[self.df['taxid'].isin(self.taxdump)]
report_diff('Dropped {} genomes without valid taxId.')
# assign genomes to species (represented by taxId not name)
self.df['species'] = self.df['taxid'].apply(taxid_at_rank,
rank='species',
taxdump=self.taxdump)
# drop genomes without species taxId
self.df.dropna(subset=['species'], inplace=True)
report_diff('Dropped {} genomes without valid species taxId.')
# drop genomes without Latinate species name
if self.latin:
self.df = self.df[self.df['species'].apply(
lambda x: is_latin(self.taxdump[x]['name']))]
report_diff('Dropped {} genomes without Latinate species name.')
print('Done.')
# include/exclude taxIds
if self.taxids:
self.taxids = set(list_from_param(self.taxids))
print(f'{"Ex" if self.exclude else "In"}cluding '
f'{len(self.taxids)} custom TaxIDs...')
self.df = self.df[self.df['taxid'].apply(lambda x: is_ancestral(
x, self.taxids, self.taxdump) != self.exclude)]
report_diff('Dropped {} genomes.')