QSAR Classification of Beta-Secretase 1 Inhibitor Activity in Alzheimer's Disease Using Ensemble Machine Learning Algorithms

Abstract

This study focuses on the development of a machine learning ensemble approach for the classification of Beta-Secretase 1 (BACE1) inhibitors in Quantitative Structure-Activity Relationship (QSAR) analysis. BACE1 is an enzyme linked to the production of amyloid beta peptide, a significant component of Alzheimer's disease plaques. The discovery of effective BACE1 inhibitors is difficult, but QSAR modeling offers a cost-effective alternative by predicting the activity of compounds based on their chemical structures. This study evaluates the performance of four machine learning models (Random Forest, AdaBoost, Gradient Boosting, and Extra Trees) in predicting BACE1 inhibitor activity. Random Forest achieved the highest performance, with a training accuracy of 98.65% and a testing accuracy of 82.53%. In addition, it exhibited superior precision, recall, and F1-score. Random Forest's superior performance was a result of its ability to capture a wide variety of patterns and its randomized ensemble approach. Overall, this study demonstrates the efficacy of ensemble machine learning models, specifically Random Forest, in predicting the activity of BACE1 inhibitors. The findings contribute to ongoing efforts in Alzheimer's disease drug discovery research by providing a cost-effective and efficient strategy for screening and prioritizing potential BACE1 inhibitors.

Enhancing the Red Wine Quality Classification Using Ensemble Voting Classifiers

“…Each algorithm brings unique strengths to the classification task. Random forest is known for its robustness and accuracy [18,19], XGBoost for its efficiency and predictive capabilities [20,21], SVM for its versatility in handling both linear and nonlinear datasets [22,23], and AdaBoost for its proficiency in mitigating class imbalances by giving more weight to misclassified samples [24]. To create the ensemble, we used the hard voting method, combining predictions from the individual models [25].…”

Classifying Beta-Secretase 1 Inhibitor Activity for Alzheimer’s Drug Discovery with LightGBM

“…Current methodologies for classifying beta-secretase 1 inhibitors predominantly lean on traditional statistical models and conventional machine-learning techniques [19], [20]. While these approaches have yielded foundational insights, they frequently encounter limitations when confronted with high-dimensional data and the intricate nonlinear relationships inherent in molecular datasets.…”

Student Perspectives on the Role of Artificial Intelligence in Education: A Survey-Based Analysis

“…In recent years, AI has witnessed important breakthroughs, particularly in machine learning [3,4] and natural language processing [5,6]. The remarkable success of AI in addressing complex tasks like image recognition [7][8][9], language translation [10], and predictive modeling [11,12] has sparked the development of even more sophisticated applications across various sectors of life.…”