Integrating protein similarity and drug similarity.

Drug-target interaction prediction accuracies using three methods and KL1LR are presented. For KL1LR, two cases of with and without protein similarity are presented. The first column and second row represent the protein similarity and drug similarity used in each method, respectively. Both ChEMBL and STITCH are used as data sources, as shown in the second column. The height of the signature kernel is 1. and are the drug and protein kernels, respectively. ECN denotes the EC numbers. For the choice of parameter values, see Table S7.