# ---------------------------------------------------------------------------- # Copyright (c) 2019-2023, QIIME 2 development team. # # Distributed under the terms of the Modified BSD License. # # The full license is in the file LICENSE, distributed with this software. # ---------------------------------------------------------------------------- from qiime2.plugin.testing import TestPluginBase from qiime2.plugins import rescript import qiime2 import pandas as pd import pandas.util.testing as pdt from rescript.dereplicate import _backfill_taxonomy from rescript._utilities import _return_stripped_taxon_rank_list import_data = qiime2.Artifact.import_data class TestDerep(TestPluginBase): package = 'rescript.tests' def setUp(self): super().setUp() self.dereplicate = rescript.actions.dereplicate self.seqs = import_data( 'FeatureData[Sequence]', self.get_data_path('derep-test.fasta')) self.fungi_seqs = import_data( 'FeatureData[Sequence]', self.get_data_path('derep-test-fungi.fasta')) self.taxa = import_data( 'FeatureData[Taxonomy]', self.get_data_path('derep-taxa.tsv')) self.fungi_taxa = import_data( 'FeatureData[Taxonomy]', self.get_data_path('derep-test-fungi.tsv')) self.seqsnumericids = import_data( 'FeatureData[Sequence]', self.get_data_path( 'derep-test-numericIDs.fasta')) self.taxanumericids = import_data( 'FeatureData[Taxonomy]', self.get_data_path( 'derep-taxa-numericIDs.tsv')) def test_dereplicate_uniq(self): seqs, taxa, = self.dereplicate( self.seqs, self.taxa, mode='uniq', rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__chondroitinus', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__brevis', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__damnosus', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei', 'C1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Pediococcus;s__acidilacti', 'A3': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B2': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # use derep_prefix=True;should still obtain same result if the prefix # seqs bear unique taxonomic labels, as seen in this test case seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='uniq', derep_prefix=True, rank_handles=['disable']) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_uniq_99_perc(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='uniq', perc_identity=0.99, rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__chondroitinus', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__brevis', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__damnosus', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei', 'C1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Pediococcus;s__acidilacti', 'A3': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B2': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # use derep_prefix=True;should still obtain same result if the prefix # seqs bear unique taxonomic labels, as seen in this test case seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='uniq', perc_identity=0.99, derep_prefix=True, rank_handles=['disable']) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_lca(self): seqs, taxa, = self.dereplicate( self.seqs, self.taxa, mode='lca', rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__damnosus', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei', 'C1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_super_lca_majority(self): seqs, taxa, = self.dereplicate( self.seqs, self.taxa, mode='super') exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__damnosus', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei', 'C1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Pediococcus;s__acidilacti'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_super_lca_majority_perc99(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='super', perc_identity=0.99) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__acidilacti', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # test that LCA taxonomy assignment works when derep_prefix=True # here derep_prefix + LCA leads to collapsed C-group seqs + LCA taxonomy def test_dereplicate_prefix_lca(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='lca', derep_prefix=True, rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # test that LCA taxonomy assignment works when derep_prefix=True # seqs F05, F06, F07 are identical. Thus having the longest shared prefix. # The representative taxonomy should appear in the output as F05, as that # has the shorter taxonomy. Thus the LCA is set to that. # Here derep_prefix + LCA taxonomy def test_dereplicate_prefix_lca_fungi(self): seqs, taxa, = self.dereplicate(self.fungi_seqs, self.fungi_taxa, mode='lca', derep_prefix=True, rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'F01': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota', 'F02': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Debaryomycetaceae;' 'g__Candida-Lodderomyces_clade;s__Candida_parapsilosis', 'F03': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Debaryomycetaceae;' 'g__Candida-Lodderomyces_clade;s__Candida_parapsilosis', 'F04': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Debaryomycetaceae;' 'g__Candida-Lodderomyces_clade;s__Candida_parapsilosis', 'F05': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Saccharomycetaceae;g__Lachancea', 'F08': 'd__Eukaryota', 'F09': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Saccharomycetaceae;' 'g__Saccharomyces;s__Saccharomyces_cerevisiae' }}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_lca_99_perc(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='lca', perc_identity=0.99, rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_majority(self): seqs, taxa, = self.dereplicate( self.seqs, self.taxa, mode='majority') exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__damnosus', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei', 'C1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Pediococcus;s__acidilacti'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # test that majority taxonomy assignment works when derep_prefix=True # all C-group seqs should be merged, and P. acidilacti is the majority def test_dereplicate_prefix_majority(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='majority', derep_prefix=True) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__acidilacti', 'B1a': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__vaginalis', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_majority_perc99(self): seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='majority', perc_identity=0.99) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__alvei', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__casei', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Pediococcus;s__acidilacti', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_dereplicate_majority_perc98_fungi(self): seqs, taxa, = self.dereplicate(self.fungi_seqs, self.fungi_taxa, mode='majority', perc_identity=0.98, rank_handles=['disable']) exp_taxa = pd.DataFrame({'Taxon': { 'F01': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Debaryomycetaceae;' 'g__Candida-Lodderomyces_clade;s__Candida_parapsilosis', 'F05': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Saccharomycetaceae;g__Lachancea;' 's__Lachancea_thermotolerans', 'F09': 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Saccharomycetaceae;' 'g__Saccharomyces;s__Saccharomyces_cerevisiae' }}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) # the above tests check actual derep functionality;this test just makes # sure that the same tests/modes above operate on numeric seq IDs, using # the same test data above (with numeric IDs). # See https://github.com/bokulich-lab/RESCRIPt/issues/49 def test_dereplicate_numericIDs(self): self.dereplicate(self.seqsnumericids, self.taxanumericids, mode='uniq') self.assertTrue(True) self.dereplicate(self.seqsnumericids, self.taxanumericids, mode='lca') self.assertTrue(True) self.dereplicate(self.seqsnumericids, self.taxanumericids, mode='majority') self.assertTrue(True) # Now test with backfilling. These parameters were chosen to set up a # variety of backfill levels. def test_dereplicate_lca_99_perc_backfill(self): # Note it is okay for inconsistent spacing around the ';' for # delimiting ranks. See `C1` for example. All that matters # is that `;` is delimiting the ranks. seqs, taxa, = self.dereplicate(self.seqs, self.taxa, mode='lca', perc_identity=0.99, rank_handles=['kingdom', 'phylum', 'class', 'order', 'family', 'genus', 'species']) exp_taxa = pd.DataFrame({'Taxon': { 'A1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Bacillales;' 'f__Paenibacillaceae;g__Paenibacillus;s__', 'B1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__Lactobacillus;s__', 'C1': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;' 'f__Lactobacillaceae;g__;s__', 'B1b': 'k__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales' ';f__Lactobacillaceae;g__Lactobacillus;s__pseudocasei'}}) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_derep_lca_90_perc_backfill_fungi(self): seqs, taxa, = self.dereplicate(self.fungi_seqs, self.fungi_taxa, mode='lca', perc_identity=0.90, rank_handles=[ 'domain', 'superkingdom', 'kingdom', 'subkingdom', 'phylum', 'subphylum', 'class', 'order', 'superfamily', 'genus', 'species']) exp_taxa = pd.DataFrame({'Taxon': { 'F01': 'd__Eukaryota;sk__;k__;ks__;p__;ps__;c__;o__;sf__;g__;s__'} }) pdt.assert_frame_equal(taxa.view(pd.DataFrame).sort_index(), exp_taxa.sort_index(), check_names=False) pdt.assert_index_equal(seqs.view(pd.Series).sort_index().index, exp_taxa.sort_index().index, check_names=False) def test_backfill_taxonomy(self): default_rank_handle = "d__; p__; c__; o__; f__; g__; s__" def _backfill_series(series, rank_handles=default_rank_handle): rank_handles = _return_stripped_taxon_rank_list(rank_handles) return series.apply(_backfill_taxonomy, args=([rank_handles, ';'])) taxa = self.taxa.view(pd.Series).sort_index() taxa = taxa.apply(lambda x: ';'.join( _return_stripped_taxon_rank_list(x))) exp_taxa = taxa.copy() # note: taxonomy is unchanged if rank handles are shorter than taxon backfilled_taxa = _backfill_series(taxa, "my;taxonomy;is;too;short") pdt.assert_series_equal(backfilled_taxa, exp_taxa, check_names=False) # manually backfill to match expected exp_taxa.loc['C1b'] += ';g__;s__' # backfill with defaults backfilled_taxa = _backfill_series(taxa) pdt.assert_series_equal(backfilled_taxa, exp_taxa, check_names=False) # trim back arbitrarily to backfill again trimmed_taxa = backfilled_taxa.apply( lambda x: ';'.join(x.split(';')[:3])) # manually backfill exp_taxa = trimmed_taxa.apply(lambda x: x + ';o__;f__;g__;s__') backfilled_taxa = _backfill_series(trimmed_taxa) pdt.assert_series_equal(backfilled_taxa, exp_taxa, check_names=False) # backfill to root # note: taxon labels can never be empty, so this test covers cases # where there is no classification beyond root/domain/kingdom backfilled_taxa = _backfill_series(trimmed_taxa.apply(lambda x: 'd__')) exp_taxa = trimmed_taxa.apply(lambda x: ';'.join( _return_stripped_taxon_rank_list( default_rank_handle))) pdt.assert_series_equal(backfilled_taxa, exp_taxa, check_names=False) # backfill custom labels custom_rank_handles = "p;e;a;n;u;t;s" exp_taxa = trimmed_taxa.apply(lambda x: x + ';n;u;t;s') backfilled_taxa = _backfill_series(trimmed_taxa, custom_rank_handles) pdt.assert_series_equal(backfilled_taxa, exp_taxa, check_names=False) def test_backfill_long_taxonomy(self): default_rank_handle = ('d__; sk__; k__; ks__; p__; ps__; c__; o__; ' 'sf__; g__; s__; ssb__; for__') rank_handles = _return_stripped_taxon_rank_list(default_rank_handle) taxa = ['d__Eukaryota; sk__Nucletmycea; k__Fungi; ks__Dikarya; ' 'p__Ascomycota', 'd__Eukaryota', 'd__Eukaryota; sk__Nucletmycea; k__Fungi; ks__Dikarya; ' 'p__Ascomycota; ps__Saccharomycotina; c__Saccharomycetes; ' 'o__Saccharomycetales; sf__Debaryomycetaceae; ' 'g__Candida-Lodderomyces_clade; s__Candida_parapsilosis'] backfilled_taxa = [_backfill_taxonomy(t, rank_handles, ';') for t in taxa] exp_taxa = ['d__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__;c__;o__;sf__;g__;s__;ssb__;for__', 'd__Eukaryota;sk__;k__;ks__;p__;ps__;c__;o__;sf__;g__;s__;' 'ssb__;for__', 'd__Eukaryota;sk__Nucletmycea;k__Fungi;ks__Dikarya;' 'p__Ascomycota;ps__Saccharomycotina;c__Saccharomycetes;' 'o__Saccharomycetales;sf__Debaryomycetaceae;' 'g__Candida-Lodderomyces_clade;s__Candida_parapsilosis;' 'ssb__;for__'] self.assertEqual(backfilled_taxa, exp_taxa)