def test_read_open_ms(self):
exp = ca.read_ms(self.openms_csv, normalize=None, data_file_type='openms')
# test we get the MZ and RT correct
self.assertIn('MZ', exp.feature_metadata)
self.assertIn('RT', exp.feature_metadata)
self.assertEqual(exp.feature_metadata['MZ'].iloc[1], 118.0869)
self.assertEqual(exp.feature_metadata['RT'].iloc[1], 23.9214)
# test normalizing
exp = ca.read_ms(self.openms_csv, normalize=10000, data_file_type='openms')
assert_array_almost_equal(exp.data.sum(axis=1), np.ones(exp.shape[0]) * 10000)
# test load sparse
exp = ca.read_ms(self.openms_csv, sparse=True, normalize=None, data_file_type='openms')
self.assertEqual(exp.sparse, True)
def test_read_gnps_ms(self):
# load the gnps exported table with associated sample metadata and cluster info
exp = ca.read_ms(self.gnps_table, sample_metadata_file=self.gnps_map,
data_file_type='gnps-ms2', normalize=None)
# verify the load extracts required fields to metadata
self.assertEqual(exp.data[2, 3], 139692)
# # test normalizing
exp = ca.read_ms(self.gnps_table, sample_metadata_file=self.gnps_map,
data_file_type='gnps-ms2', normalize=10000)
assert_array_almost_equal(exp.data.sum(axis=1), np.ones(exp.shape[0]) * 10000)
# # test load sparse
exp = ca.read_ms(self.gnps_table, sample_metadata_file=self.gnps_map,
data_file_type='gnps-ms2', normalize=None, sparse=True)
self.assertEqual(exp.sparse, True)
def test_get_spurious_duplicates(self):
# load an mzmine2 metabolomics table, and associated gnps clusterinfo file
exp = ca.read_ms(self.mzmine2_csv,
sample_metadata_file=self.gnps_map,
data_file_type='mzmine2',
use_gnps_id_from_AllFiles=False,
normalize=None)
# get rid of the all 0s metabolite (to get rid of std=0 warning)
exp = exp.filter_sum_abundance(0.1)
res = exp.get_spurious_duplicates()
# no samples filtered away
self.assertEqual(res.shape[0], 6)
# default parameters don't identify and suspicious features
self.assertEqual(res.shape[1], 0)
res = exp.get_spurious_duplicates(mz_tolerance=100, rt_tolerance=0.5)
self.assertEqual(res.shape[1], 0)
res = exp.get_spurious_duplicates(rt_tolerance=1)
self.assertEqual(res.shape[1], 0)
res = exp.get_spurious_duplicates(mz_tolerance=100, rt_tolerance=1)
self.assertEqual(res.shape[1], 2)
res = exp.get_spurious_duplicates(mz_tolerance=100,
rt_tolerance=1,
corr_thresh=0.2)
self.assertEqual(res.shape[1], 4)
def test_read_open_ms_samples_rows(self):
exp = ca.read_ms(self.openms_samples_rows_csv, normalize=None, sample_in_row=True, data_file_type='openms')
# test we get the MZ and RT correct
self.assertIn('MZ', exp.feature_metadata)
self.assertIn('RT', exp.feature_metadata)
self.assertAlmostEqual(exp.feature_metadata['MZ'].iloc[1], 118.0869)
self.assertAlmostEqual(exp.feature_metadata['RT'].iloc[1], 23.9214)
def test_read_biom_ms(self):
# load a biom table with MZ/RT in featureID, and associated gnps clusterinfo file
exp = ca.read_ms(self.ms_biom_table, sample_metadata_file=self.gnps_map,
data_file_type='biom', use_gnps_id_from_AllFiles=False, normalize=None)
self.assertIn('MZ', exp.feature_metadata)
self.assertIn('RT', exp.feature_metadata)
self.assertEqual(exp.feature_metadata['MZ'].iloc[1], 899.53)
self.assertEqual(exp.feature_metadata['RT'].iloc[0], 314)
def test_read_mzmine2_ms_with_idstr(self):
# load an mzmine2 metabolomics table with the sampleids inflated with additional info
exp = ca.read_ms(self.mzmine2_with_idstr_csv, sample_metadata_file=self.gnps_map,
use_gnps_id_from_AllFiles=False, cut_sample_id_sep='_', normalize=None)
self.assertEqual(exp.feature_metadata['MZ'].iloc[1], 200)
self.assertEqual(exp.feature_metadata['RT'].iloc[0], 1)
self.assertEqual(exp.sample_metadata['field2'][0], 'f')
self.assertEqual(exp.data[2, 1], 35900)
self.assertEqual(exp.data.shape, (6, 6))
def test_read_mzmine2_ms(self):
# load an mzmine2 metabolomics table, and associated gnps clusterinfo file
exp = ca.read_ms(self.mzmine2_csv, sample_metadata_file=self.gnps_map,
data_file_type='mzmine2', use_gnps_id_from_AllFiles=False, normalize=None)
self.assertIn('MZ', exp.feature_metadata)
self.assertIn('RT', exp.feature_metadata)
self.assertEqual(exp.feature_metadata['MZ'].iloc[1], 200)
self.assertEqual(exp.feature_metadata['RT'].iloc[0], 1)
self.assertEqual(exp.data[2, 1], 35900)
def test_merge_similar_features(self):
# load an mzmine2 metabolomics table, and associated gnps clusterinfo file
exp = ca.read_ms(self.mzmine2_csv,
sample_metadata_file=self.gnps_map,
data_file_type='mzmine2',
use_gnps_id_from_AllFiles=False,
normalize=None)
# no merging since features are far away
res = exp.merge_similar_features()
self.assertEqual(res.shape[1], 6)
# a little merging
res = exp.merge_similar_features(mz_tolerance=100, rt_tolerance=1)
self.assertEqual(res.shape[1], 3)
self.assertEqual(res.feature_metadata.at[85022, '_calour_merge_ids'],
'85022;93277')
# a lot of merging
res = exp.merge_similar_features(mz_tolerance=400, rt_tolerance=6)
self.assertEqual(res.shape[1], 2)
self.assertEqual(res.feature_metadata.at[121550, '_calour_merge_ids'],
'121550')
def test_filter_mz_rt(self):
# load an mzmine2 metabolomics table, and associated gnps clusterinfo file
exp = ca.read_ms(self.mzmine2_csv,
sample_metadata_file=self.gnps_map,
data_file_type='mzmine2',
use_gnps_id_from_AllFiles=False,
normalize=None)
# mz filtering
res = exp.filter_mz_rt(100)
self.assertEqual(len(res.feature_metadata), 1)
self.assertEqual(res.feature_metadata['MZ'].values, [100])
res = exp.filter_mz_rt([100, 201])
self.assertEqual(len(res.feature_metadata), 1)
self.assertEqual(res.feature_metadata['MZ'].values, [100])
res = exp.filter_mz_rt([100, 201], mz_tolerance=1)
self.assertEqual(len(res.feature_metadata), 2)
npt.assert_array_equal(res.feature_metadata['MZ'].values, [100, 200])
res = exp.filter_mz_rt([100, 201], negate=True)
self.assertEqual(len(res.feature_metadata), 5)
# rt filtering
res = exp.filter_mz_rt(rt=[1, 2.5])
self.assertEqual(len(res.feature_metadata), 1)
self.assertEqual(res.feature_metadata['RT'].values, [1])
res = exp.filter_mz_rt(rt=[1, 2.5], rt_tolerance=0.5)
self.assertEqual(len(res.feature_metadata), 3)
npt.assert_array_equal(res.feature_metadata['RT'].values, [1, 2, 3])
# complex - both mz and rt
res = exp.filter_mz_rt([101, 200, 400, 505], [1, 3, 4, 5],
mz_tolerance=2)
self.assertEqual(res.shape[1], 2)
