Merge pull request #4 from phac-nml/fix/empty-output

Bug Fix for When Mash Output is Empty
phac-nml · Nov 23, 2020 · 4f42693 · 4f42693
2 parents cd37a82 + e282174
commit 4f42693
Show file tree

Hide file tree

Showing 6 changed files with 70 additions and 33 deletions.
diff --git a/refseq_masher/__init__.py b/refseq_masher/__init__.py
@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 
-__version__ = '0.1.1'
+__version__ = '0.1.2'
 program_name = 'refseq_masher'
 program_summary = 'Mash MinHash search your sequences against a NCBI RefSeq genomes database'
 program_desc = program_summary + '''

diff --git a/refseq_masher/cli.py b/refseq_masher/cli.py
@@ -134,12 +134,21 @@ def contains(mash_bin, output, output_type, top_n_results, min_identity, max_pva
                                    max_pvalue=max_pvalue,
                                    min_identity=min_identity,
                                    parallelism=parallelism)
-        if top_n_results > 0:
-            df = df.head(top_n_results)
-        dfs.append(df)
-    logging.info('Ran Mash Screen on all input. Merging NCBI taxonomic information into results output.')
-    dfout = merge_ncbi_taxonomy_info(pd.concat(dfs))
-    logging.info('Merged taxonomic information into results output')
-    logging.info('Reordering output columns')
-    dfout = order_output_columns(dfout, MASH_SCREEN_ORDERED_COLUMNS)
-    write_dataframe(dfout, output_path=output, output_type=output_type)
+
+        if df is not None:
+            if top_n_results > 0:
+                df = df.head(top_n_results)
+            dfs.append(df)
+
+    logging.info('Ran Mash Screen on all input.')
+
+    if len(dfs) > 0:
+        logging.info('Merging NCBI taxonomic information into results output.')
+        dfout = merge_ncbi_taxonomy_info(pd.concat(dfs))
+        logging.info('Merged taxonomic information into results output')
+        logging.info('Reordering output columns')
+        dfout = order_output_columns(dfout, MASH_SCREEN_ORDERED_COLUMNS)
+        write_dataframe(dfout, output_path=output, output_type=output_type)
+
+    else:
+        logging.info('There were no matches found.')
diff --git a/refseq_masher/mash/parser.py b/refseq_masher/mash/parser.py
@@ -112,14 +112,20 @@ def mash_screen_output_to_dataframe(mash_out: str) -> pd.DataFrame:
         mash_out: Mash screen stdout
 
     Returns:
-        (pd.DataFrame): Mash screen output table ordered by `identity` and `median_multiplicity` columns in descending order
+        (pd.DataFrame): Mash screen output table ordered by `identity` and `median_multiplicity` columns in descending
+            order, or None if the Mash output is missing
     """
-    df = pd.read_table(StringIO(mash_out))
-    ncols = df.shape[1]
-    df.columns = MASH_SCREEN_COLUMNS[:ncols]
-    df.sort_values(by=['identity', 'median_multiplicity'], ascending=[False, False], inplace=True)
-    match_ids = df.match_id
-    refseq_matches = [parse_refseq_info(match_id=match_id) for match_id in match_ids]
-    dfmatch = pd.DataFrame(refseq_matches)
-    dfmerge = pd.merge(dfmatch, df, on='match_id')
-    return dfmerge
+
+    dfmerge = None
+
+    if len(mash_out) > 0:
+        df = pd.read_table(StringIO(mash_out))
+        ncols = df.shape[1]
+        df.columns = MASH_SCREEN_COLUMNS[:ncols]
+        df.sort_values(by=['identity', 'median_multiplicity'], ascending=[False, False], inplace=True)
+        match_ids = df.match_id
+        refseq_matches = [parse_refseq_info(match_id=match_id) for match_id in match_ids]
+        dfmatch = pd.DataFrame(refseq_matches)
+        dfmerge = pd.merge(dfmatch, df, on='match_id')
+
+    return dfmerge
diff --git a/refseq_masher/mash/screen.py b/refseq_masher/mash/screen.py
@@ -27,7 +27,7 @@ def vs_refseq(inputs: Union[str, List[str]],
         parallelism: Mash screen number of parallel threads to spawn
 
     Returns:
-        (pd.DataFrame): Parsed Mash screen results dataframe
+        (pd.DataFrame): Parsed Mash screen results dataframe or None if the output of Mash was empty
     """
     cmd_list = [mash_bin, 'screen',
                 '-v', str(max_pvalue),
@@ -40,9 +40,14 @@ def vs_refseq(inputs: Union[str, List[str]],
         cmd_list.append(inputs)
     else:
         raise TypeError('Unexpected type "{}" for "inputs": {}'.format(type(inputs), inputs))
+
     logging.info('Running Mash Screen with NCBI RefSeq sketch database '
                  'against sample "%s" with inputs: %s', sample_name, inputs)
     exit_code, stdout, stderr = run_command(cmd_list, stderr=None)
+
     df = mash_screen_output_to_dataframe(stdout)
-    df['sample'] = sample_name
+
+    if df is not None:
+        df['sample'] = sample_name
+
     return df
diff --git a/tests/data/small.fastq b/tests/data/small.fastq
@@ -0,0 +1,4 @@
+@READ1
+AGCTACGTAGCTATATATTATAGCGTAGCTCGTGCGCTAGCTAGCTATATGCTGATGCTCTAGCTATATTATACTAGCTATAGCTAAAACGCTGCTCGTAA
++
+CCCCCCCCCCCCCCCCCCCCC;CCCCCCCCCCCCCC;CCCCCCCCCCCCCCCCC-CCCCCCCCCCCCCCCCCCCCCCCCCC;CCCCCCCCCCCCC-CCCCC
diff --git a/tests/test_contains.py b/tests/test_contains.py
@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-
+import unittest
 from io import StringIO
 
 from click.testing import CliRunner
@@ -13,16 +13,29 @@
 def runner():
     return CliRunner()
 
+    def test_fasta(runner):
+        fasta = 'tests/data/Se-Enteritidis.fasta'
+        result = runner.invoke(cli.contains, ['--top-n-results', '1',
+                                              '-p', '4',
+                                              fasta])
+
+        assert result.exit_code == 0, 'Exit code should be 0'
+
+        expected_top_tax_name = 'Salmonella enterica subsp. enterica serovar Enteritidis'
+        df = pd.read_table(StringIO(result.output))
+        assert df.top_taxonomy_name.str.contains(expected_top_tax_name).all(), \
+            'Top Mash RefSeq result should have "{}" in the top_taxonomy_name field'.format(expected_top_tax_name)
+
+
+class TestContains(unittest.TestCase):
 
-def test_fasta(runner):
-    fasta = 'tests/data/Se-Enteritidis.fasta'
-    result = runner.invoke(cli.contains, ['--top-n-results', '1',
-                                          '-p', '4',
-                                          fasta])
+    def test_small_fastq(self):
+        fastq = 'tests/data/small.fastq'
+        runner = CliRunner()
 
-    assert result.exit_code == 0, 'Exit code should be 0'
+        with self.assertLogs(level='INFO') as cm:
+            result = runner.invoke(cli.contains, ['-p', '4', fastq])
+            assert "There were no matches found." in " ".join(cm.output)
 
-    expected_top_tax_name = 'Salmonella enterica subsp. enterica serovar Enteritidis'
-    df = pd.read_table(StringIO(result.output))
-    assert df.top_taxonomy_name.str.contains(expected_top_tax_name).all(), \
-        'Top Mash RefSeq result should have "{}" in the top_taxonomy_name field'.format(expected_top_tax_name)
+        assert result.exit_code == 0, 'Exit code should be 0'
+        assert len(result.stdout) == 0, 'No output results on empty data'