Download PDFOpen PDF in browserGPU-Accelerated Machine Learning Models for Drug Discovery in Computational BiologyEasyChair Preprint 1481514 pages•Date: September 12, 2024AbstractDrug discovery is a complex and time-consuming process that requires the identification of potential therapeutic compounds from vast chemical libraries. In recent years, the integration of machine learning (ML) with computational biology has revolutionized this field by enabling faster and more accurate predictions of drug-target interactions, molecular properties, and toxicity profiles. This paper explores the use of GPU-accelerated machine learning models in drug discovery, focusing on how the parallel processing capabilities of GPUs significantly reduce the computational burden associated with large-scale simulations and predictive modeling. Key advancements in deep learning architectures and generative models, such as convolutional neural networks (CNNs) and graph neural networks (GNNs), are discussed in the context of their application to molecular dynamics, protein-ligand binding, and virtual screening. Additionally, the study highlights the role of transfer learning and active learning strategies in enhancing model accuracy and adaptability in drug discovery workflows. Through case studies, this research demonstrates the potential of GPU-powered ML models to accelerate the identification of novel drug candidates, improve lead optimization, and ultimately shorten the drug development timeline. Keyphrases: GPU acceleration, drug discovery, machine learning
|