diff --git a/Wolf-tutorial-with-the-OBITools3.md b/Wolf-tutorial-with-the-OBITools3.md
index 82b22a9..ed31ca4 100644
--- a/Wolf-tutorial-with-the-OBITools3.md
+++ b/Wolf-tutorial-with-the-OBITools3.md
@@ -53,11 +53,13 @@ And unzip it:
 
 	obi alignpairedend -R wolf/reads2 wolf/reads1 - > aligned_reads.fastq
 
+### 3. Remove sequence records with a low overlap alignment score:
+
 We can check the average alignment score (corresponding approximately to the similarity between the sequences in the overlap):
 
 	obi stats -a score_norm wolf/aligned_reads
 
-### 3. Remove sequence records with a low overlap alignment score:
+And keep only sequences with a high overlap alignment score: 
 
 	obi grep -p "sequence['score_norm'] > 0.8" wolf/aligned_reads wolf/good_sequences