Given two semantic knowledge graphs that describe the pathophysiology (nodes are proteins, genes etc., edges are relationships like increase, decrease, associations) of two different diseases, find the best algorithms to compare them, to find the "shared mechanisms", to identify comorbidity.
I thought to use some semantic similatiry measures like Jaccard distance, or performing a network-based analysis using community detection, centrality measures.Could it be the right way of doing so?
If I want to use query language, like cypher? Do I check for common nodes, triples, sub-graphs?Should I costruct a new network/graph for shared entities?