AI-Aided Design of Novel Targeted Covalent Inhibitors against SARS-CoV-2

Area of Science:

• Drug discovery and development
• Artificial intelligence in medicine
• Virology

Background:

• Existing drug repurposing strategies for SARS-CoV-2 have proven ineffective.
• The 3C-like main protease (3CLpro or Mpro) is a crucial target for developing novel antiviral therapies against coronaviruses.

Purpose of the Study:

• To design novel chemical entities targeting the SARS-CoV-2 3CLpro.
• To leverage artificial intelligence and fragment-based drug design for identifying potential antiviral lead compounds.

Main Methods:

• Development of an advanced deep Q-learning network with fragment-based drug design (ADQN-FBDD).
• Utilized a structure-based optimization policy (SBOP) for refining lead compounds.
• Generated potential drug candidates based on the PDB ID: 6LU7 structure.

Main Results:

• The AI model successfully generated 47 lead compounds targeting SARS-CoV-2 3CLpro.
• Structure-based optimization yielded related derivatives of the initial lead compounds.
• All generated compounds are available in a molecular library for further research.

Conclusions:

• The novel AI-driven approach (ADQN-FBDD) shows promise for identifying effective antiviral agents against SARS-CoV-2.
• The generated lead compounds and derivatives represent valuable candidates for future drug development efforts.
• This study highlights the potential of AI in accelerating the discovery of new therapeutics for viral infections.