diff --git a/denovo-assembly.qmd b/denovo-assembly.qmd
index 2cbf507d..8eaa666c 100644
--- a/denovo-assembly.qmd
+++ b/denovo-assembly.qmd
@@ -754,3 +754,5 @@ Be aware of the sequencing depth when you assemble your sample. This sample used
 The original sample had almost 200 million reads, however, I subsampled it to less than 5 million reads. You usually need a lot of sequencing data for _de novo_ assembly and definitely more data than for reference-alignment based profiling. However, it also heavily depends on the complexity of the sample. So the best advice is: **just give it a try**!
 
 :::
+
+## References