Graph theory - is it of any use in machine learning?

I am aware of probabilistic graphical models, but somehow I don't see alot of graph theory in that area of ML... don't know if there is any other useful application of graph theory? Actually, what I'm asking is this: is graph theory worth learning if one is interested in statistical/machine learning? Is analysis of complex networks intertwined with ML? Or am I going to waste my time? :/