From 7ff07a2fb294ed1bf664030659cc4e1f96fbc08c Mon Sep 17 00:00:00 2001
From: Jeremy Arbesfeld <50678786+jarbesfeld@users.noreply.github.com>
Date: Thu, 17 Oct 2024 09:09:59 -0400
Subject: [PATCH] Update alignment code (#68)

---
 src/dcd_mapping/align.py | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/src/dcd_mapping/align.py b/src/dcd_mapping/align.py
index fc2d7fb..9925089 100644
--- a/src/dcd_mapping/align.py
+++ b/src/dcd_mapping/align.py
@@ -235,7 +235,8 @@ def _get_best_hsp(
     """Retrieve preferred HSP from BLAT Hit object.
 
     We select the hsp object with the lowest distance from the start of the
-    corresponding gene and the highest BLAT score
+    corresponding gene and the highest BLAT score. We omit the first sorting step
+    when a gene symbol is not associated with a score set
 
     :param hit: hit object from BLAT result
     :param urn: scoreset identifier for use in error messages
@@ -245,7 +246,11 @@ def _get_best_hsp(
     :raise AlignmentError: if hit object appears to be empty (should be impossible)
     """
     best_hsp = None
-    hsp_list = sorted(hit, key=lambda hsp: abs(hsp.hit_start - gene_location.start))
+    hsp_list = (
+        sorted(hit, key=lambda hsp: abs(hsp.hit_start - gene_location.start))
+        if gene_location
+        else hit
+    )
     hsp_list = sorted(
         hsp_list,
         key=lambda hsp: (hsp.query_end - hsp.query_start) / output.seq_len,