Dense granular materials provide a rich setting to explore networks and topology. Of particular interest here are states near the jamming transition. The idea of jamming is simple: if a collection of grains is too loose, then it will not be mechanically stable. If it is very dense, there will be more than enough force-bearing contacts to make the system stable or jammed. The concept of states that are marginally stable is then a natural one. Until recently, it was thought that density alone controls the jamming transition. We have shown that this is not the case: both the density and stress anisotropy control jamming of frictional grains. This means that characterizing the networks of contacts of force-bearing grains is crucial to understanding jamming. In this talk, I will explore the use of computational homology and network methods to characterize force networks near jamming.