normalize distance to [0, 1) with f(average_dist) = 0.5

cfe-lab · Sep 20, 2023 · a4d5c39 · a4d5c39
1 parent 7394784
commit a4d5c39
Show file tree

Hide file tree

Showing 7 changed files with 1,037 additions and 1,034 deletions.
diff --git a/intact/intact.py b/intact/intact.py
@@ -623,6 +623,9 @@ def get_indel_impact(alignment):
             orf_alignment = (exp_protein, "-" * len(exp_protein))
             best_match.distance = aligner.match_score - ((aligner.open_gap_score + len(exp_protein) * aligner.extend_gap_score) / max(1, len(exp_protein)))
 
+        average_distance = 0.62
+        best_match.distance = 1 - (average_distance / (best_match.distance + average_distance)) # normalise it to the [0, 1) interval, with f(average_distance) = 0.5.
+
         # Max deletion allowed in ORF exceeded
         if deletions > e.deletion_tolerence:
 

diff --git a/tests/expected-results-large-csv/orfs.csv b/tests/expected-results-large-csv/orfs.csv
diff --git a/tests/expected-results-large/orfs.json b/tests/expected-results-large/orfs.json
diff --git a/tests/expected-results-single-csv/orfs.csv b/tests/expected-results-single-csv/orfs.csv
diff --git a/tests/expected-results-single/orfs.json b/tests/expected-results-single/orfs.json
diff --git a/tests/expected-results-small-csv/orfs.csv b/tests/expected-results-small-csv/orfs.csv
diff --git a/tests/expected-results-small/orfs.json b/tests/expected-results-small/orfs.json