Documenting algorithms

[TomSmithCGAT]

The network-based algorithms are described in at least places (GR publication, UMI-tools docs & CGAT blog post). On reflection, the description of the 'adjacency' method is insufficiently precise in some/all places.

I suggest we include a specific readthedocs page to walk through the network methods using the example in the publication and refer users to this in the command line help