diff --git a/README.md b/README.md
index 40b317e..1010d96 100644
--- a/README.md
+++ b/README.md
@@ -106,6 +106,15 @@ logoplotwithcrosslink(pfm, C; rna=true)
 ```
 which will automatically assume a uniform background of `[0.25, 0.25, 0.25, 0.25]`.
 
+Use the command
+```
+save_crosslinked_logoplot(pfm, background, C, "demo2.png")
+```
+or 
+```
+save_crosslinked_logoplot(pfm, C, "demo2.png") # uniform background
+```
+to save the plot.
 
 ### Multiplexed crosslinking tendencies
 
@@ -126,6 +135,14 @@ You'd get
 
 Here, different colors indicate different crosslinking signatures, and their height is proportional to the crosslinking tendency at each position in the PWM. 
 
+Similar to above, use
+```
+save_crosslinked_logoplot(pfm, background, C2, "demo3.png"; rna=true)
+```
+to save the plot.
+
+
+
 ## Plot your PWM with highlighted regions
 Sometimes you may have columns that you want to highlight, for example, when you have transcription factors binding sites embedded in a (long) transposable element (e.g. see figure 4 in this [paper](https://academic.oup.com/bioinformatics/article/39/6/btad378/7192989)). Then, what you can do is to provide a vector of `UnitRange{Int}` to highlight the regions of interest, e.g. 
 ```
@@ -147,6 +164,9 @@ logoplot_with_highlight_crosslink(pfm, background, C2, highlighted_regions2)
 
 ![highlight-pfm](demo/demo5.png)
 
+Use 
+
+
 <!-- 
 save_crosslinked_logoplot(pfm, C, "tmp.png"; highlighted_regions=highlighted_regions2) -->
 
diff --git a/demo/demo2.png b/demo/demo2.png
index 6fefe66..d57d268 100644
Binary files a/demo/demo2.png and b/demo/demo2.png differ
diff --git a/demo/demo3.png b/demo/demo3.png
index 03d7d4b..368cd65 100644
Binary files a/demo/demo3.png and b/demo/demo3.png differ