use most frequent element to determine scramble check direction

intact/intact.py

@@ -8,6 +8,7 @@
 import tempfile
 import csv
 from dataclasses import dataclass
+from collections import Counter
 from Bio import AlignIO, Seq, SeqIO, SeqRecord
 from scipy.stats import fisher_exact
 
@@ -159,6 +160,12 @@ def is_sorted(lst):
     return True
 
 
+def most_frequent_element(lst):
+    counter = Counter(lst)
+    most_common = counter.most_common(1)
+    return most_common[0][0] if most_common else None
+
+
 def remove_5_prime(blast_rows):
     # HIV 5' region can easily map to its 3' region because they are identical.
     # Such a maping would not constitute a scramble, so we ignore the 5' region for this check.
@@ -194,7 +201,7 @@ def is_scrambled(seqid, blast_rows):
         return None
 
     ignored_5_prime.sort(key=lambda x: x.qstart)
-    direction = ignored_5_prime[0].sstrand
+    direction = most_frequent_element(x.sstrand for x in ignored_5_prime)
     if direction == "plus" and is_sorted(x.sstart for x in ignored_5_prime):
         return None
     elif direction == "minus" and is_sorted(x.send for x in reversed(ignored_5_prime)):

tests/test_scramble.py

@@ -3,7 +3,19 @@
 import os
 
 import intact.intact as intact
-from intact.intact import is_scrambled, contains_internal_inversion, IntactnessError
+from intact.intact import is_scrambled, contains_internal_inversion, IntactnessError, most_frequent_element
+
+@pytest.mark.parametrize("lst, expected", [
+    ([1, 2, 3, 4, 2, 2, 3, 1, 4, 4, 4], 4),
+    ([1, 2, 3, 4, 2, 2, 3, 1], 2),
+    ([], None),
+    ([5], 5),
+    ([1, 2, 3, 4, 5], 1),
+    ([1, 2, 3, 4, 5, 6, 7, 8, 9, 10], 1),
+    ([1, 2, 3, 2, 4, 5, 4, 6, 6, 6], 6)
+])
+def test_most_frequent_element(lst, expected):
+    assert most_frequent_element(lst) == expected
 
 @pytest.mark.parametrize("lst, expected", [
     ([1, 2, 3, 4, 5], True),
@@ -114,7 +126,7 @@ def test_is_scrambled_mixed_direction():
     ]
     result = is_scrambled("id", blast_rows) or contains_internal_inversion("id", blast_rows)
     assert isinstance(result, IntactnessError)
-    assert result.error == intact.INTERNALINVERSION_ERROR
+    assert result.error == intact.SCRAMBLE_ERROR
 
 def test_is_scrambled_single_row_plus_strand():
     # Test case with a single row aligned in the "plus" strand