Typically, the way I would approach this is by importing all of the data and then doing some simple visualizations. For example, you could map the fold change for each data type as a ring around the nodes in an interaction network. You could normalize the data across all channels and then do a correlation network, but I frankly think that may not be as interesting as looking for differential perturbation within the context of an interaction network. At the end of the day, the answer really depends on what you are hoping to acheive with your analysis -- what is your biological question?
-- scooter