I am looking for a way to compare entire subnetworks within the context of encompassing PPI, say Y2H, and calculating how far (shortest path) each node of a disease subnetwork is against each node of another disease subnetwork, and averaging this number over the span of the entire subnetwork - to get a distance between two gwas disease gene sets

If you want to learn how to write your own code, Think Complexity has a section on working with graphs.