all ProteinSequence tests pass

isovar/protein_sequence.py

@@ -412,9 +412,15 @@ def subsequence(self, start_idx, end_idx):
         """
         old_length = len(self)
 
+        start_idx, end_idx = normalize_base0_range_indices(
+            start_idx=start_idx,
+            end_idx=end_idx,
+            sequence_length=old_length)
 
         amino_acids = self.amino_acids[start_idx:end_idx]
+
         new_length = len(amino_acids)
+
         # if we lose amino acids from the end of the sequence then it
         # can't end with a stop codon anymore
         ends_with_stop_codon = (
@@ -428,13 +434,16 @@ def subsequence(self, start_idx, end_idx):
         # then the start/end will be be clipped to 0. If the mutation is
         # to the right of the new subsequence then the start/end will both
         # be clipped to the subsequence length
-        mutation_start_idx = max(0, self.mutation_start_idx - start_idx)
-        mutation_end_idx = min(new_length, self.mutation_end_idx - start_idx)
+        mutation_start_idx = \
+            min(new_length, max(0, self.mutation_start_idx - start_idx))
+        mutation_end_idx = \
+            min(new_length, max(0, self.mutation_end_idx - start_idx))
+
         # number of mutant amino acids in the new subsequence
         num_mutant_aa = mutation_end_idx - mutation_start_idx
         # a deletion is considered a mutant sequence if the amino acids to
         # the left and right of it are both present
-        deletion = (num_mutant_aa == 0) and (0 < start_idx < new_length)
+        deletion = (num_mutant_aa == 0) and (0 < mutation_start_idx < new_length)
         contains_mutation = self.contains_mutation and (
             (num_mutant_aa > 0) or deletion
         )

test/test_protein_sequences.py

@@ -16,8 +16,8 @@
 from isovar.protein_sequence import ProteinSequence
 from isovar.protein_sequence_helpers import sort_protein_sequences
 
-def test_protein_sequence_num_mutant_amino_acids():
-    # testing that we got the correct number of amino acids (1)
+def test_protein_sequence_substitution():
+    # testing that we got the correct properties in the case
     # where "SIINFEKL" was mutated into "SIINFEQL"
     p = ProteinSequence(
         amino_acids="SIINFEQL",
@@ -27,13 +27,19 @@ def test_protein_sequence_num_mutant_amino_acids():
         ends_with_stop_codon=True,
         frameshift=False,
         translations=[])
+    eq_(p.amino_acids, "SIINFEQL")
+    eq_(len(p), 8)
+    eq_(p.contains_mutation, True)
+    eq_(p.contains_deletion, False)
+    eq_(p.frameshift, False)
+    eq_(p.ends_with_stop_codon, True)
     eq_(p.num_mutant_amino_acids, 1)
-    eq_(p.num_mutant_amino_acids, len(p.mutant_amino_acids))
+    eq_(p.mutant_amino_acids, "Q")
 
 
-def test_protein_sequence_num_mutant_amino_acids_deletion():
-    # testing that we got the correct number of amino acids (0)
-    # where "SIINFEKL" was mutated into "SIINFEL"
+def test_protein_sequence_deletion():
+    # testing that we got the correct properties in the case
+    # where "SIINFEKL" was mutated into "SIINFEL" by deletion of one amino acid
     p = ProteinSequence(
         amino_acids="SIINFEL",
         contains_mutation=True,
@@ -42,10 +48,17 @@ def test_protein_sequence_num_mutant_amino_acids_deletion():
         ends_with_stop_codon=True,
         frameshift=False,
         translations=[])
+    eq_(p.amino_acids, "SIINFEL")
+    eq_(len(p), 7)
     eq_(p.num_mutant_amino_acids, 0)
-    eq_(p.num_mutant_amino_acids, len(p.mutant_amino_acids))
+    eq_(p.mutant_amino_acids, "")
+    eq_(p.contains_mutation, True)
+    eq_(p.contains_deletion, True)
+    eq_(p.frameshift, False)
+    eq_(p.ends_with_stop_codon, True)
 
-def test_protein_subsequence_overlaps_mutation():
+
+def test_protein_subsequence_overlaps_substitution():
     # testing that we got the correct properties for case where "SIINFEKL" was
     # mutated into "SIINFEQL" and then sliced to keep just "FEQL"
     p = ProteinSequence(
@@ -60,10 +73,13 @@ def test_protein_subsequence_overlaps_mutation():
     eq_(p2.amino_acids, "FEQL")
     eq_(p2.contains_mutation, True)
     eq_(p2.contains_deletion, False)
+    eq_(p2.frameshift, False)
+    eq_(p2.ends_with_stop_codon, True)
     eq_(p2.num_mutant_amino_acids, 1)
     eq_(p2.mutant_amino_acids, "Q")
+    eq_(len(p2), 4)
 
-def test_protein_subsequence_does_not_overlap_mutation():
+def test_protein_subsequence_does_not_overlap_substitution():
     # testing that we got the correct properties for case where "SIINFEKL" was
     # mutated into "SIINFEQL" and then sliced to keep just "FEQL"
     p = ProteinSequence(
@@ -78,8 +94,57 @@ def test_protein_subsequence_does_not_overlap_mutation():
     eq_(p2.amino_acids, "SIIN")
     eq_(p2.contains_mutation, False)
     eq_(p2.contains_deletion, False)
+    eq_(p2.frameshift, False)
+    eq_(p2.ends_with_stop_codon, False)
     eq_(p2.num_mutant_amino_acids, 0)
     eq_(p2.mutant_amino_acids, "")
+    eq_(len(p2), 4)
+
+
+def test_protein_subsequence_overlaps_deletion():
+    # testing that we got correct properties for the case
+    # where "SIINFEKL" was mutated into "SIINFEL" (by a deletion of "K")
+    # and then we took the subsequence "FEL"
+    p = ProteinSequence(
+        amino_acids="SIINFEL",
+        contains_mutation=True,
+        mutation_start_idx=len("SIINFE"),
+        mutation_end_idx=len("SIINFE"),
+        ends_with_stop_codon=True,
+        frameshift=False,
+        translations=[])
+    p2 = p.subsequence(len("SIIN"), None)
+    eq_(len(p2), 3)
+    eq_(p2.amino_acids, "FEL")
+    eq_(p2.mutant_amino_acids, "")
+    eq_(p2.contains_deletion, True)
+    eq_(p2.contains_mutation, True)
+    eq_(p2.frameshift, False)
+    eq_(p2.ends_with_stop_codon, True)
+    eq_(p2.num_mutant_amino_acids, 0)
+
+
+def test_protein_subsequence_does_not_overlap_deletion():
+    # testing that we got correct properties for the case
+    # where "SIINFEKL" was mutated into "SIINFEL" (by a deletion of "K")
+    # and then we took the subsequence "SIINFE"
+    p = ProteinSequence(
+        amino_acids="SIINFEL",
+        contains_mutation=True,
+        mutation_start_idx=len("SIINFE"),
+        mutation_end_idx=len("SIINFE"),
+        ends_with_stop_codon=True,
+        frameshift=False,
+        translations=[])
+    p2 = p.subsequence(None, len("SIINFE"))
+    eq_(len(p2), 6)
+    eq_(p2.amino_acids, "SIINFE")
+    eq_(p2.contains_deletion, False)
+    eq_(p2.contains_mutation, False)
+    eq_(p2.frameshift, False)
+    eq_(p2.ends_with_stop_codon, False)
+    eq_(p2.mutant_amino_acids, "")
+    eq_(p2.num_mutant_amino_acids, 0)
 
 def test_sort_protein_sequences():
     protseq_most_reads = make_dummy_protein_sequence(