Machine learning combines atomistic simulations to predict SARS-CoV-2 Mpro inhibitors from natural compounds

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular biology

Background:

• The COVID-19 pandemic continues to cause global damage.
• SARS-CoV-2 main protease (Mpro) is a key therapeutic target for COVID-19 treatment.

Purpose of the Study:

• To computationally identify promising SARS-CoV-2 Mpro inhibitors from natural compounds.
• To validate a combined machine learning and atomistic simulation strategy for drug discovery.

Main Methods:

• Machine learning (ML) model trained to screen natural compounds.
• Atomistic simulations including molecular docking and molecular dynamics.
• Linear interaction energy (LIE) scheme for affinity calculation.

Main Results:

• Good agreement observed between ML predictions and atomistic simulations.
• Ten natural compounds identified as potential SARS-CoV-2 Mpro inhibitors.
• Seven compounds demonstrated high-nanomolar binding affinity.

Conclusions:

• The combined ML and atomistic simulation approach is reliable for rapid inhibitor prediction.
• This strategy can aid in the development of therapeutics against current and future SARS-CoV-2 variants.