In this work, we aimed to present realistic evaluation settings to predict DDIs using knowledge graph embeddings. We have applied Logistic Regression, Naive Bayes and Random Forest on Drugbank knowledge graph with the 10-fold traditional cross validation using RDF2Vec, TransE and TransD. We also propose a simple disjoint cross-validation scheme to evaluate drug-drug interaction predictions for the scenarios where the drugs have no known DDIs.
We performed cross-validation using different setting:
Proposed disjoint CV:
To cite this work
Celebi, Remzi, Huseyin Uyar, Erkan Yasar, Ozgur Gumus, Oguz Dikenelli, and Michel Dumontier. "Evaluation of knowledge graph embedding approaches for drug-drug interaction prediction in realistic settings." BMC bioinformatics 20, no. 1 (2019): 1-14.