Fig. 1

DT2Vec pipeline. a1, a2 Drug-drug (DDS) and protein–protein (PPS) similarity graphs based on similarity used as input of embedding method to generate vectors. a3 Graph of DTIs. b Graph-embedding developed by node2vec to map nodes (in DDS and PPS) to vectors (in this figure drugs and target mapped to 2D-vector, x and y). c Known DTIs (positive and negative) were divided into 10% independent dataset (external testset) and 90% internal test and train (tenfold cross-validation). d Drug and protein vectors were concatenated and labeled as positive (1) or negative (0) and an XGBoost model was trained on the cross-validation datasets. The best model over the tenfold cross-validation on the internal testset was selected and applied on the external testset. The XGBoost model in c, d was repeated 5 times and the average performance of internal and external testsets was reported