User actions graph layout calculation

Question

So, I have a weighted directed graph. Each node represents a page in web site, each node represents user action on moving from one page to another, weight means the amount of users that have performed such action. I need to calculate a good informative layout for this graph, that will be informative for visual analysis, I want to see bottlenecks and popular user paths (there are start and end nodes). I want my solution to.be in python. What can I do?

Upd: I've tried Sugiyama, Kamada-Kawai and Fruchterman-Reingold algorythms, but they do not seem very informative, as my graph is strongly connected and they create a big unreadable mash (even in 1.5k edges with about 30 nodes)