Prediction of Drug-Target Interaction on Jamu Formulas using Machine Learning Approaches

Jamu is an Indonesian herbal medicine that has many benefits. Prediction of drug-target interactions on Jamu formula using a graph-based approach was carried out, but the results were unsatisfactory with the area under the precision-recall curve (AUPR) of 0.70. This study develops a prediction model of drug-target interactions with machine learning approach using Support Vector Machine (SVM) and Random Forest (RF). The dataset used in this study as the same as the dataset in the previous research, obtained from Indonesian Jamu Herbs (IJAH) Analytics. The dataset represents interactions of compounds and proteins, including labels to indicate those of interactions. Principal Component Analysis (PCA) is used as feature reduction in the pre-processing stage. The prediction models using SVM and RF combined with PCA obtain the best AUPR results of 0.99. These results indicate that the machine learning approach has better performance than those of the graph-based approach in predicting drug-target interactions on Jamu formulas.