Federated SPARQL query performance evaluation for exploring disease model mouse: combining gene expression, orthology, and disease knowledge graphs

Table 6 The average runtime from 10 executions of the SPARQL query Examples 7, 8, 9, and 10 without using the subqueries in Query subpart 4

Query approach	Federated or Centralized	Anatomical parts where genes were expressed	The runtime of 0 subqueries (Query subpart 4)	The difference between the Federated and the Centralized
Example 7	Federated	prefrontal cortex	48 s	34 s (between Example 7 and 9), and 42 s (between Example 8 and 10)
Example 8	Federated	skin of body	58 s
Example 9	Centralized	prefrontal cortex	14 s
Example 10	Centralized	skin of body	16 s

Bold values represent the best values in each column. We executed the federated search from the BioResource MetaDB SPARQL endpoint to the official Bgee SPARQL endpoint. We performed the centralized search from the BioResource MetaDB SPARQL endpoint to the Bgee data stored in the BioResource MetaDB
The numbers of retrieved genes in Examples 7 and 9 were the same at 42,448 (Additional file 19, and 21). The numbers of retrieved genes in Examples 8 and 10 were the same and 45,724 (Additional file 20, and 22)

ISSN: 1472-6947