Based on the hypergeometric distribution (all possible drug-disease pairs in CTD is 5239086, the number of positive pairs in CTD is 111481, the total number of drug-disease pairs considered 6260 and the number of positive pairs is 514), the result is extremely significant ( p -value is 6.19 × 10 −144 ) and shows that the top ranked predictions by iDrug are greatly enriched in the CTD dataset.
← all excerpts
iDrug: Integration of drug repositioning and drug-target prediction via cross-network embedding.
1
—
—