TCB Members, NAMD, and VMD Part of Team Winning the ACM Gordon Bell Special Prize for COVID-19 Research

Members of the Theoretical and Computational Biophysics Group were part of a multi-institutional interdisciplinary team awarded at Supercomputing 2020 with the internationally recognized ACM Gordon Bell Special Prize for COVID-19 Research, for 2020.

Announcement of the team's winning paper at Supercomputing 2020.
The winning team developed a new AI-driven workflow that incorporated state-of-the-art software tools for structure preparation, molecular dynamics simulation, accelerated weighted ensemble sampling, simulation analysis, and visualization, using the combined computing resources, performance analysis tools, and computing staff of several of the largest supercomputer centers in the United States. The new workflow, software tools, and computing power were used to provide scientists with unprecedented views into the dynamics of the SARS-CoV-2 spike protein, its interaction with the human receptor ACE2, and the full virion, leading to several new scientific discoveries, and providing ready-to-use tools for ongoing and future multi-scale modeling efforts. Team members from U. Illinois provided key methodological and scalability advances in the NAMD molecular dynamics simulation software, and technological improvements to VMD, a key molecular modeling tool used to prepare, visualize, and analyze the SARS-CoV-2 spike and virion simulations. The NAMD and VMD software advances provided by U. Illinois team members enabled the science campaign to more efficiently utilize state-of-the-art supercomputers at national computing centers, including Summit, the most powerful supercomputer in the United States, operated by Oak Ridge National Laboratory. These achievements required the unique skills of a diverse multi-institutional team to overcome both technical challenges and an extremely compressed research timeline. Every team member played a vital role in achieving the final outcome.

AI-Driven Multiscale Simulations Illuminate Mechanisms of SARS-CoV-2 Spike Dynamics

Abstract: We develop a generalizable AI-driven workflow that leverages heterogeneous HPC resources to explore the time-dependent dynamics of molecular systems. We use this workflow to investigate the mechanisms of infectivity of the SARS-CoV-2 spike protein, the main viral infection machinery. Our workflow enables more efficient investigation of spike dynamics in a variety of complex environments, including within a complete SARS-CoV-2 viral envelope simulation, which contains 305 million atoms and shows strong scaling on ORNL Summit using NAMD. We present several novel scientific discoveries, including the elucidation of the spike's full glycan shield, the role of spike glycans in modulating the infectivity of the virus, and the characterization of the flexible interactions between the spike and the human ACE2 receptor. We also demonstrate how AI can accelerate conformational sampling across different systems and pave the way for the future application of such methods to additional studies in SARS-CoV-2 and other molecular systems.

Team presentation at Supercomputing 2020:

SARS-CoV-2 spike protein and full virion model. Amaro Lab, UCSD.

NAMD scaling on Summit and Frontera for 8.5M-atom
spike-ACE2 complex (upper lines) and 305M-atom
virion (lower line). Thin reference lines show ideal
linear scaling.

Weighted ensemble methods, as implemented in WESTPA,
provided tremendous enhancement of molecular dynamics
sampling of rare events, yielding effective performance
levels orders of magnitude faster than can be achieved
through brute force parallel computation alone.

t-SNE plot of latent space resulting from DeepDriveMD
pipeline, showing weighted ensemble training data
(thin traces) and testing data large traces) from
subsequent simulations of the 8.5M-atom spike system,
colored by RMSD. Outliers from these simulations
are represented with large spheres.

Map of research team members, their institutions,
and supercomputer centers used for the research.

Research Team: