High-Performance Computing for Protein Structure Prediction Using ML and GPUs

EasyChair Preprint no. 13758

Abstract

The field of protein structure prediction has witnessed significant advancements due to the integration of high-performance computing (HPC) and machine learning (ML) techniques, especially with the use of graphics processing units (GPUs). This paper explores the transformative impact of HPC and ML on predicting protein structures, which is crucial for understanding biological functions and developing therapeutic interventions. The introduction of GPUs has revolutionized computational biology by offering parallel processing capabilities that significantly reduce the time required for complex calculations. Machine learning models, particularly deep learning algorithms, have demonstrated unprecedented accuracy in predicting protein folding and interactions by analyzing vast datasets of known protein structures. This synergy between ML and HPC facilitates the development of predictive models that are not only faster but also more accurate and scalable. The paper discusses key methodologies, including the use of neural networks and advanced optimization techniques, and highlights successful case studies where GPU-accelerated ML models have outperformed traditional approaches. Additionally, it addresses the challenges associated with data quality, computational costs, and the integration of ML models into existing HPC frameworks. The findings underscore the potential of combining ML and GPUs to accelerate biomedical research and drug discovery, paving the way for innovations in personalized medicine and biotechnology.

Keyphrases: Graphics Processing Units (GPUs), High Performance Computing (HPC), Machine Learning (ML)

@Booklet{EasyChair:13758,
  author = {Abill Robert},
  title = {High-Performance Computing for Protein Structure Prediction Using ML and GPUs},
  howpublished = {EasyChair Preprint no. 13758},

  year = {EasyChair, 2024}}