From b953615c258fa01279f3067005576f9c3732d0e2 Mon Sep 17 00:00:00 2001
From: Alex Bartlett <abartlett@nextrna.com>
Date: Tue, 3 May 2022 19:23:30 +0000
Subject: [PATCH] fix umi bug

---
 HISTORY.md               | 4 ++++
 seqcluster/libs/fastq.py | 2 +-
 setup.py                 | 2 +-
 3 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/HISTORY.md b/HISTORY.md
index c3f32e7..045f138 100644
--- a/HISTORY.md
+++ b/HISTORY.md
@@ -1,3 +1,7 @@
+- 1.2.9
+
+  * Fix bug where UMI is mistakenly detected in read names containing "ILLUMINA"
+
 - 1.2.8
 
   * Fix bug when detecting the end of the cluster due to big gaps in biopython function
diff --git a/seqcluster/libs/fastq.py b/seqcluster/libs/fastq.py
index 0e411bc..13b8b11 100644
--- a/seqcluster/libs/fastq.py
+++ b/seqcluster/libs/fastq.py
@@ -17,7 +17,7 @@ def collapse(in_file):
         line = handle.readline()
         while line:
             if line.startswith("@"):
-                if line.find("UMI") > -1:
+                if line.find("UMI_") > -1:
                     logger.info("Find UMI tags in read names, collapsing by UMI.")
                     return collapse_umi(in_file)
                 seq = handle.readline().strip()
diff --git a/setup.py b/setup.py
index fa368e7..00540ae 100644
--- a/setup.py
+++ b/setup.py
@@ -13,7 +13,7 @@ def readme():
 
 
 setup(name='seqcluster',
-      version='1.2.8',
+      version='1.2.9',
       description='Small RNA-seq pipeline',
       long_description=readme(),
       long_description_content_type="text/markdown",