def write_hgt_list(self, sample):
"""Write a list of predicted HGTs to an output file.
Parameters
----------
sample : str
sample Id
Notes
-----
The output file has three columns: protein Id, silhouette score,
potential donor.
The donor will be the LCA of top hits as determined by `find_match`.
However, if `donor_rank` is specified, a donor below this rank will be
raise to this rank; a donor above this rank will be discarded.
"""
taxdump, name, rank = self.taxdump, self.donor_name, self.donor_rank
df_ = self.df[self.df['hgt'] & (self.df['sample'] == sample)]
print(f' {sample}: {df_.shape[0]}.')
with open(join(self.output, 'hgts', f'{sample}.txt'), 'w') as f:
for row in df_[['protein', 'silh', 'match']].itertuples():
# format donor taxon
match = row.match
if rank and match != '0':
match = taxid_at_rank(match, rank, self.taxdump) or '0'
if name:
match = taxdump[match]['name'] if match != '0' else 'N/A'
f.write(f'{row.protein}\t{row.silh:g}\t{match}\n')
def infer_self_group(self):
"""Infer self group automatically.
"""
# just use LCA
if not self.self_rank:
self.self_tax = [self.lca]
# try to raise LCA to given rank, but if LCA is already above that
# rank, just use LCA
else:
tid_ = taxid_at_rank(self.lca, self.self_rank, self.taxdump)
self.self_tax = [tid_ or self.lca]
def infer_self_group(self):
"""Infer self group automatically.
Notes
-----
Assign `self_tax` as top-level taxId(s) of the self group.
"""
# just use LCA
if not self.self_rank:
self.self_tax = [self.lca]
# try to raise LCA to given rank, but if LCA is already above that
# rank, just use LCA
else:
tid_ = taxid_at_rank(self.lca, self.self_rank, self.taxdump)
self.self_tax = [tid_ or self.lca]
def test_taxid_at_rank(self):
taxdump = taxdump_from_text(taxdump_archaea)
self.assertEqual(taxid_at_rank('1538547', 'genus', taxdump), '1655637')
self.assertEqual(taxid_at_rank('1538547', 'phylum', taxdump),
'1655434')