consider ways to package edge and node data?

[diegospa99]

from sam on slack: "and b) regardless of dependencies, what are good options for intuitively organizing and indexing NICHES data, particularly given the fundamental non-barcode-alignmentbetween node-indexed data (cells, N2C, C2N, Neighborhood signaling, and nieghborhood composition assays) versus edge-indexed data (C2C)? this is probably an important starting point for whiteboarding and creative thinking..."

[RaredonLab]

I think that we may want to experiment with using dataframes, with the rows being either nodes or edges, and the columns being metadata AND/OR features. This will facilitate ggplot plotting and SQL language interface, naturally. But we need to research S3 and S4 objects, and research potential scaling issues. With how many rows / columns can such a structure allow fluid handling?