Discovery of Potential Neonicotinoid Insecticides by an Artificial Intelligence Generative Model and Structure-Based...

Area of Science:

• Agricultural Chemistry
• Computational Chemistry
• Insecticide Development

Background:

• Discovering novel neonicotinoid insecticides is crucial for effective pest management.
• Existing insecticides face challenges like resistance and environmental impact.

Purpose of the Study:

• To leverage artificial intelligence (AI) and virtual screening for identifying new neonicotinoid insecticide leads.
• To validate the efficacy and binding modes of AI-generated compounds.

Main Methods:

• Constructed a deep generative model using a recurrent neural network and transfer learning.
• Applied hierarchical virtual screening and similarity searches to filter generated molecules.
• Conducted bioassays and molecular docking/dynamics simulations for lead compounds.

Main Results:

• The AI model successfully generated potential neonicotinoid compounds with accurate molecular grammar.
• Compounds A2 and A5 exhibited mortality rates of 52.5% and 50.3% against Aphis craccivora.
• Docking and molecular dynamics studies confirmed similar binding modes to known neonicotinoids.

Conclusions:

• AI-driven discovery is effective for identifying novel insecticide scaffolds.
• Compounds A2 and A5 represent promising leads for new neonicotinoid insecticides.
• The integrated AI and screening strategy accelerates pesticide discovery pipelines.