From c203a3be8bb168d9aa9fe0ba5e6e9f365b4bfa9b Mon Sep 17 00:00:00 2001
From: Kirill Bessonov <kbessonov@gmail.com>
Date: Thu, 7 Nov 2024 16:30:08 -0500
Subject: [PATCH] Updated bowtie2 setting to be compatible with long reads

---
 ectyper/genomeFunctions.py     | 4 +++-
 ectyper/predictionFunctions.py | 3 +--
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/ectyper/genomeFunctions.py b/ectyper/genomeFunctions.py
index fd57562..ebef144 100644
--- a/ectyper/genomeFunctions.py
+++ b/ectyper/genomeFunctions.py
@@ -232,8 +232,10 @@ def assemble_reads(reads, bowtie_base, combined_fasta, temp_dir, cores=1):
     bowtie_run = [
         'bowtie2',
         '--threads',f'{cores}',
+        '--local',
         '--score-min L,1,-0.5',
         '--np 5',
+        '--no-unal',
         '-x', bowtie_base,
         '-U', reads,
         '-S', sam_reads
@@ -248,7 +250,7 @@ def assemble_reads(reads, bowtie_base, combined_fasta, temp_dir, cores=1):
         'samtools', 'view',
         '--threads', f'{cores}',
         '-S',
-        '-F 4',
+        '-F 4', #filter out the unmapped reads
         '-q 1',
         '-b',
         '-o', bam_reads,
diff --git a/ectyper/predictionFunctions.py b/ectyper/predictionFunctions.py
index b01fc83..797d64a 100644
--- a/ectyper/predictionFunctions.py
+++ b/ectyper/predictionFunctions.py
@@ -534,7 +534,6 @@ def get_prediction(per_genome_df):
 
     #find pairwise distance between all alleles indentified by BLAST search and find alleles with IDENTICAL O-antigen score
     #Note that difference in score of 0 between O antigens is considered to be identical
-
     identicalscorestupleslist = [(orow, ocol, abs(i - j)) for ocol, i in scorestupleslist for orow, j in scorestupleslist
                             if i - j == 0 and (orow == selectedOantigen or ocol == selectedOantigen) and
                             (orow != ocol)]
@@ -548,7 +547,7 @@ def get_prediction(per_genome_df):
 
 
     if highsimilarity_oantigens:
-        mixedoantigen = [selectedOantigen] + highsimilarity_oantigens
+        mixedoantigen = sorted([selectedOantigen] + highsimilarity_oantigens, key = lambda x: re.search('O(\d+)',x).group(1), reverse=True)
         serotype['O']['serogroup'] = "/".join(mixedoantigen)
         LOG.info("Highly similar O-antigen candidates were found for {}".format(mixedoantigen))
     elif selectedOantigen != "":