Large-Scale FMO-MP2 Calculations of the Spike Protein Droplet Model

Area of Science:

• Computational chemistry
• Structural biology
• Biophysics

Background:

• The SARS-CoV-2 spike protein is a critical target for antiviral drug development.
• Accurate theoretical modeling of large proteins like the spike protein presents significant computational challenges.

Purpose of the Study:

• To benchmark the Fragment Molecular Orbital (FMO) method for modeling the SARS-CoV-2 spike protein.
• To assess the computational efficiency of FMO on a large-scale supercomputer.

Main Methods:

• Utilized the Fragment Molecular Orbital (FMO) method at the second-order Møller-Plesset perturbation (MP2) level.
• Employed 100 structure samples from molecular dynamics (MD) simulations for both closed (PDB ID: 6XLU) and open (PDB ID: 6XM0) spike protein forms.
• Performed calculations on the Fugaku supercomputer, processing approximately 20,000 fragments per structure.

Main Results:

• Achieved a benchmark calculation time of approximately 2 hours per structure using 8 racks of the Fugaku supercomputer.
• Demonstrated the feasibility of applying FMO to large protein systems, such as the SARS-CoV-2 spike protein.
• Successfully modeled both closed and open conformations of the spike protein.

Conclusions:

• The Fragment Molecular Orbital (FMO) method is a viable and efficient approach for theoretical studies of large viral proteins like the SARS-CoV-2 spike protein.
• Large-scale supercomputing resources like Fugaku are essential for enabling such complex molecular modeling.
• These findings support the use of FMO for future investigations into spike protein dynamics and interactions.