In a groundbreaking study, researchers have developed a supercomputer-driven pipeline for in-silico drug discovery, leveraging enhanced sampling molecular dynamics (MD) and ensemble docking techniques. This innovative approach aims to accelerate the identification of potential drug candidates by utilizing the computational power of supercomputers to simulate and analyze the dynamic properties of protein binding sites.
The methodology hinges on ensemble docking, which integrates MD results to dock compound databases into multiple conformations of protein binding sites. This strategy effectively captures the dynamic nature of these sites, enhancing the accuracy of docking predictions. The study’s focus was on 24 systems derived from eight proteins in the SARS-CoV-2 proteome, the virus responsible for COVID-19.
The MD simulations employed temperature replica exchange enhanced sampling, a technique that utilizes massively parallel supercomputing to expedite the exploration of protein configurational space. By harnessing the Summit supercomputer at Oak Ridge National Laboratory, the researchers achieved an unprecedented rate of over one millisecond of enhanced sampling MD per day.
For the docking process, the team used AutoDock Vina to dock repurposing databases to ten distinct configurations of each of the 24 SARS-CoV-2 systems. Remarkably, they demonstrated that with AutoDock-GPU on Summit, they could perform exhaustive docking of one billion compounds in less than 24 hours, showcasing the incredible computational efficiency and capability of their pipeline.
Preliminary results from this study highlight the potential of this supercomputing-driven approach in identifying promising drug candidates against SARS-CoV-2. The researchers also outlined future enhancements to the pipeline, including the integration of quantum mechanical (QM) methods, machine learning (ML), and artificial intelligence (AI) for clustering MD trajectories and rescoring docking poses.
This study represents a significant advancement in the field of computational drug discovery, particularly in response to emergent viral threats like COVID-19. The combination of enhanced sampling MD and ensemble docking, powered by state-of-the-art supercomputers, offers a promising path forward in the rapid identification and development of effective therapeutics.
You can access the full paper here.