Utilizing the Mol2Vec Algorithm to Extract Features of HIV Drug Interactions
Felix Indra Kurniadi, Risma Yulistiani · 2024
HIV/AIDS is a global pandemic that targets the immune system and afflicts millions of individuals worldwide. This study centers on the classification of drug-target interactions specific to HIV. The study utilized a dataset obtained from MoleculeNet, and feature extraction was conducted using mol2vec on the Simplified Molecular Input Line Input System (SMILES) string format. Furthermore, we conducted a comparative analysis of the outcomes from this research by employing various machine learning techniques, including K-nearest Neighbor, Support Vector Machine, Decision Tree, and Random Forest. The studies yielded conclusive findings, demonstrating the effectiveness of the proposed strategy in extracting the features responsible for achieving an accuracy value of 0.96 across all four methods. Nevertheless, the issue of data imbalance significantly impacts the precision, recall, and F1-score values, resulting in unsatisfactory outcomes.