R is pretty common, but if you don't know R, you could try ngsplot. It's an R script you call from the command line with various flags and config files. It can create a lot of the average gene type plots.

I ran into the same problem a while ago -- I saw a lot of fancy ChIP-seq figures with no information in the methods to describe how they were made.

So I wrote metaseq (paper, github, docs). At the same time, others wrote ngsplot and deepTools to address the same lack of tools.

ngsplot and deepTools are great for use from the command line; metaseq is more of a toolbox. Example 1 walks you through making a custom ChIP-seq figure and gives a taste for what you can do with the package.

You need to know some basic programming skills.

Annotate your peaks first ( how far is the peak from the TSSs), then you count the peak numbers in each different categories(0-5kb, 5-10kb etc) and then plot a bar graph.

Easier way, try to use GREAT tools

A lot of graphs of this type can be created using CEAS (http://liulab.dfci.harvard.edu/CEAS/)

CEAS is also a part of Cistrome pipeline, which can be accessed from Galaxy web interface (http://cistrome.org/ap/) or from desktop application Unipro UGENE (http://ugene.unipro.ru)

ChIPseeker can creates similar figures: http://ygc.name/2014/04/30/visualization-methods-in-chipseeker/

I prefer Genomation.

High quality output and very flexible and easy to use! If you know some R.