At the homepage for Garib Murshudov’s lab, there is a page dedicated to EM fitting tools and scripts. Among these scripts are some for Coot which adds jiggle-fit, morph modeling and user-define restraints capabilities to Coot (you can read a bit about them here).
If you have not used Jiggle-fit yet, I can highly recommend it, especially for low resolution de-novo building.
Here is an example I did using jiggle-fit in Coot using my own map and a Coot generated ideal RNA double helix.
I just placed the helix randomly somewhere in the density,that I want to fit the helix into, like so:
I then ran jiggle fit on the whole molecule (I used 20000 loops) and about 30 sec later I get this:
Due to the relative low resolution of the map, it can be a bit difficult to see just how good the fit actually is, but this zoom-in may help a bit:
It is a very efficient tool, but of course it does not always work this well.