def test_sum_taxids(self):
me = Analyze()
me.input_tax = {'S1': '1', 'S2': '3'}
def _hits_df(d):
return pd.Series(d, name='taxid').to_frame()
me.data = {
'S1': [{
'hits': _hits_df({
'a': '4',
'b': '6'
})
}, {
'hits': _hits_df({
'a': '4',
'c': '8'
})
}],
'S2': [{
'hits': _hits_df({
'b': '6',
'd': '1'
})
}]
}
obs = me.sum_taxids()
exp = {'1', '3', '4', '6', '8'}
self.assertSetEqual(obs, exp)
def test_calc_scores(self):
columns = ('id', 'taxid', 'score')
# helper for making hit table
def _hits_df(data):
return pd.DataFrame(data, columns=columns).set_index('id')
me = Analyze()
me.taxdump = taxdump_from_text(taxdump_proteo)
add_children(me.taxdump)
me.groups = {
'self': {'561', '562', '585056'},
'close': {'543', '91347', '1236'}
}
me.data = {
'S1': [{
'score': 100,
'hits': _hits_df((('P1', '561', 100), ('P2', '562', 95)))
}, {
'score': 90,
'hits': _hits_df((('P3', '561', 81), ('P4', '543', 72)))
}],
'S2': [{
'score':
96,
'hits':
_hits_df(
(('P5', '561', 90), ('P6', '543', 84), ('P7', '620', 66)))
}]
}
me.weighted = True
me.match_th = 0.9
me.calc_scores()
# helper for get scores
def _prot_scores(prot):
return [prot[x] for x in ('self', 'close', 'distal')]
s1_1 = me.data['S1'][0]
self.assertListEqual(s1_1['hits']['group'].tolist(), ['self', 'self'])
self.assertListEqual(_prot_scores(s1_1), [1.95, 0.0, 0.0])
self.assertEqual(s1_1['match'], '0')
s1_2 = me.data['S1'][1]
self.assertListEqual(s1_2['hits']['group'].tolist(), ['self', 'close'])
self.assertListEqual(_prot_scores(s1_2), [0.9, 0.8, 0.0])
self.assertEqual(s1_2['match'], '0')
s2_1 = me.data['S2'][0]
self.assertListEqual(s2_1['hits']['group'].tolist(),
['self', 'close', 'distal'])
self.assertListEqual(_prot_scores(s2_1), [0.9375, 0.875, 0.6875])
self.assertEqual(s2_1['match'], '620')
def test_make_score_table(self):
me = Analyze()
me.output = self.tmpdir
me.data = {
'S1': [{
'id': 'P1',
'length': 100,
'match': '0',
'self': 1.5,
'close': 0.75,
'distal': 0.0,
'hits': pd.DataFrame([0] * 3)
}, {
'id': 'P2',
'length': 120,
'match': '1224',
'self': 1.625,
'close': 0.225,
'distal': 0.375,
'hits': pd.DataFrame([0] * 5)
}],
'S2': [{
'id': 'P1',
'length': 225,
'match': '620',
'self': 2.35,
'close': 1.05,
'distal': 0.75,
'hits': pd.DataFrame([0] * 6)
}]
}
me.make_score_table()
obs = me.df.values.tolist()
exp = [['S1', 'P1', 100, 3, 1.5, 0.75, 0, '0'],
['S1', 'P2', 120, 5, 1.625, 0.225, 0.375, '1224'],
['S2', 'P1', 225, 6, 2.35, 1.05, 0.75, '620']]
self.assertListEqual(obs, exp)
fp = join(self.tmpdir, 'scores.tsv')
with open(fp, 'r') as f:
obs = [x.split('\t') for x in f.read().splitlines()[1:]]
exp = [[str(y) for y in x] for x in exp]
self.assertListEqual(obs, exp)
remove(fp)
def test_assign_taxonomy(self):
# input are two genomes with defined taxonomy
me = Analyze()
me.input_tax = 'S1:561,S2:620' # Escherichia and Shigella
me.data = {}
me.taxdump = taxdump_from_text(taxdump_proteo)
me.assign_taxonomy()
# test input taxonomy extraction
self.assertDictEqual(me.input_tax, {'S1': '561', 'S2': '620'})
# test taxonomy refinement
exp = {
'1', '131567', '2', '1224', '1236', '91347', '543', '561', '620'
}
self.assertSetEqual(set(me.taxdump.keys()), exp)
# test LCA discovery
self.assertEqual(me.lca, '543')
# helper for making hit table
def _hits_df(d):
return pd.Series(d, name='taxid', dtype=object).to_frame()
# input is one genome with defined taxonomy
me = Analyze()
me.data = {'S1': [{'hits': pd.DataFrame(columns=['taxid'])}]}
me.input_tax = '561' # Escherichia
me.taxdump = taxdump_from_text(taxdump_proteo)
me.assign_taxonomy()
self.assertDictEqual(me.input_tax, {'S1': '561'})
# input taxonomy not found in database
me.input_tax = '1234'
me.taxdump = taxdump_from_text(taxdump_proteo)
with self.assertRaises(ValueError) as ctx:
me.assign_taxonomy()
msg = 'TaxID 1234 is not present in taxonomy database.'
self.assertEqual(str(ctx.exception), msg)
# input are two genome whose taxonomies are to be inferred based on
# search results
me = Analyze()
me.input_tax = None
me.data = {
'S1': [{
'hits': _hits_df({
'P1': '561',
'P2': '562'
})
}, {
'hits': _hits_df({
'P3': '543',
'P4': '561'
})
}],
'S2': [{
'hits': _hits_df({
'P5': '562',
'P6': '585056'
})
}, {
'hits': _hits_df({
'P7': '561',
'P8': '1038927'
})
}, {
'hits': _hits_df({'P9': '2580236'})
}]
}
me.input_cov = 75
me.taxdump = taxdump_from_text(taxdump_proteo)
me.assign_taxonomy()
self.assertDictEqual(me.input_tax, {'S1': '543', 'S2': '561'})
self.assertEqual(me.lca, '543')
# cannot auto-infer taxonomy
me.data['S3'] = [{'hits': _hits_df({})}]
me.taxdump = taxdump_from_text(taxdump_proteo)
with self.assertRaises(ValueError) as ctx:
me.assign_taxonomy()
msg = 'Cannot auto-infer taxonomy for S3. Please specify manually.'
self.assertEqual(str(ctx.exception), msg)
# invalid input taxonomy string
me.input_tax = '561'
with self.assertRaises(ValueError) as ctx:
me.assign_taxonomy()
msg = 'Invalid input taxonomy format.'
self.assertEqual(str(ctx.exception), msg)
