DrugECs: An Ensemble System with Feature Subspaces for Accurate Drug-Target Interaction Prediction

Figure 2

The flowchart of the ensemble system for the drug-target prediction. It illustrates the system for the 7 top principal components. The “PCA 1” denotes the features of protein targets created by the first top principal component, while the “” means the first feature group of drugs from 1 to 206 created by the “PaDEL” software. Each feature group consists of the roughly same number of features. Therefore each instance is composed of one “PCA” feature group and one “PaDEL” feature group. In total, there are 49 combinations for the case of the 7 top principal components.