Fig. 1From: Prediction of compound-target interaction using several artificial intelligence algorithms and comparison with a consensus-based strategyPerformance target-centric models (TCM) on VDS2. TCM were trained with decision tree (DT), gaussian naive bayes (GM), k-nearest neighbors (KNN), random forest (RF) and support vector machine (SVM). The descriptors used were: a morgan's fingerprint (FGP), b molecular properties (DSC) and c the fusion of FGP and DSC (FUS)Back to article page