Fix remaining problem for #470 by overriding HCV subtype elsewhere.

cfe-lab · Jun 19, 2019 · 45eb80a · 45eb80a
1 parent da6fa92
commit 45eb80a
Show file tree

Hide file tree

Showing 3 changed files with 45 additions and 10 deletions.
diff --git a/README.md b/README.md
@@ -9,8 +9,8 @@
 [travis]: https://travis-ci.org/cfe-lab/MiCall
 [Code Coverage]: https://codecov.io/github/cfe-lab/MiCall/coverage.svg?branch=master
 [codecov]: https://codecov.io/github/cfe-lab/MiCall?branch=master
-[DOI]: https://zenodo.org/badge/DOI/10.5281/zenodo.2644171.svg
-[zenodo]: https://doi.org/10.5281/zenodo.2644171
+[DOI]: https://zenodo.org/badge/DOI/10.5281/zenodo.1289989.svg
+[zenodo]: https://doi.org/10.5281/zenodo.1289989
 
 Maps all the reads from a sample against a set of reference sequences, then
 stitches all the reads into consensus sequences and coverage maps.

diff --git a/micall/resistance/resistance.py b/micall/resistance/resistance.py
@@ -128,7 +128,7 @@ def combine_aminos(amino_csv, midi_amino_csv, failures):
         in the midi_amino_csv file.
     """
     is_midi = True
-    midi_rows = {}  # {seed: [row]}
+    midi_rows = {}  # {genotype: [row]}
     midi_start = 231
     midi_end = 561
     is_midi_separate = midi_amino_csv.name != amino_csv.name
@@ -203,14 +203,16 @@ def combine_midi_rows(main_rows, midi_rows, seed):
     positions = sorted({pos
                         for pos in chain(main_row_map.keys(),
                                          midi_row_map.keys())})
+
+    for midi_row in midi_row_map.values():
+        midi_row['seed'] = seed  # Override MIDI seed with main seed.
     for pos in positions:
         main_row = main_row_map.get(pos)
         midi_row = midi_row_map.get(pos)
         if midi_row is None:
             if pos <= 336:
                 yield main_row
         elif main_row is None:
-            midi_row['seed'] = seed  # Override MIDI seed with main seed.
             yield midi_row
         elif (pos <= 336 and
               int(main_row['coverage']) > int(midi_row['coverage'])):

diff --git a/micall/tests/test_resistance.py b/micall/tests/test_resistance.py
@@ -323,9 +323,9 @@ def test_combine_aminos_ns5b_multiple_subtypes():
 seed,region,q-cutoff,query.nuc.pos,refseq.aa.pos,\
 A,C,D,E,F,G,H,I,K,L,M,N,P,Q,R,S,T,V,W,Y,*,X,partial,del,ins,clip,g2p_overlap,coverage
 HCV-4m,HCV4-ED43-NS5b,15,1,558,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4m,HCV4-ED43-NS5b,15,1,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4m,HCV4-ED43-NS5b,15,4,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4m,HCV4-ED43-NS5b,15,7,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,4,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,7,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,10,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
 """)
     midi_amino_csv.name = 'midi_amino.csv'
     expected_csv = """\
@@ -335,9 +335,42 @@ def test_combine_aminos_ns5b_multiple_subtypes():
 HCV-4c,HCV4-ED43-NS5b,15,4,2,0,0,0,0,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
 HCV-4c,HCV4-ED43-NS5b,15,7,228,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
 HCV-4c,HCV4-ED43-NS5b,15,1,558,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4c,HCV4-ED43-NS5b,15,1,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4c,HCV4-ED43-NS5b,15,4,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
-HCV-4c,HCV4-ED43-NS5b,15,7,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,4,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,7,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,10,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+"""
+
+    check_combination(amino_csv, midi_amino_csv, expected_csv)
+
+
+def test_combine_aminos_ns5b_multiple_subtypes_with_overlap():
+    """ Main sample's subtype overrides MIDI subtype when they disagree. """
+    amino_csv = StringIO("""\
+seed,region,q-cutoff,query.nuc.pos,refseq.aa.pos,\
+A,C,D,E,F,G,H,I,K,L,M,N,P,Q,R,S,T,V,W,Y,*,X,partial,del,ins,clip,g2p_overlap,coverage
+HCV-4c,HCV4-ED43-NS5b,15,1,1,0,0,0,0,0,0,0,0,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
+HCV-4c,HCV4-ED43-NS5b,15,4,2,0,0,0,0,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
+HCV-4c,HCV4-ED43-NS5b,15,7,228,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
+""")
+    amino_csv.name = 'main_amino.csv'
+    midi_amino_csv = StringIO("""\
+seed,region,q-cutoff,query.nuc.pos,refseq.aa.pos,\
+A,C,D,E,F,G,H,I,K,L,M,N,P,Q,R,S,T,V,W,Y,*,X,partial,del,ins,clip,g2p_overlap,coverage
+HCV-4m,HCV4-ED43-NS5b,15,1,228,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,4,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,7,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4m,HCV4-ED43-NS5b,15,10,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+""")
+    midi_amino_csv.name = 'midi_amino.csv'
+    expected_csv = """\
+seed,region,q-cutoff,query.nuc.pos,refseq.aa.pos,\
+A,C,D,E,F,G,H,I,K,L,M,N,P,Q,R,S,T,V,W,Y,*,X,partial,del,ins,clip,g2p_overlap,coverage
+HCV-4c,HCV4-ED43-NS5b,15,1,1,0,0,0,0,0,0,0,0,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
+HCV-4c,HCV4-ED43-NS5b,15,4,2,0,0,0,0,10000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,10000
+HCV-4c,HCV4-ED43-NS5b,15,1,228,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,4,559,0,0,0,0,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,7,560,0,0,0,0,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
+HCV-4c,HCV4-ED43-NS5b,15,10,561,20000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,20000
 """
 
     check_combination(amino_csv, midi_amino_csv, expected_csv)