Neural network-based QSAR and insecticide discovery: spinetoram.

Area of Science:

• Agricultural Chemistry
• Insecticide Development
• Computational Chemistry

Background:

• Spinosyns are fermentation-derived insecticides with a need for improved efficacy and spectrum.
• Modifying complex spinosyn molecules often resulted in decreased activity, posing a challenge for development.

Purpose of the Study:

• To identify novel synthetic directions for spinosyn chemistry to enhance insecticidal properties.
• To overcome limitations in traditional quantitative structure-activity relationship (QSAR) studies for spinosyns.

Main Methods:

• Exploration of quantitative structure-activity relationships (QSAR) for spinosyns.
• Application of artificial neural networks (ANN) to analyze spinosyn QSAR data.
• Synthesis and testing of ANN-guided analogs and other spinosyn derivatives.

Main Results:

• Artificial neural networks successfully identified new synthetic routes for improved spinosyn activity.
• ANN-based analogs, combined with structure-activity relationship data, led to the discovery of spinetoram (XDE-175).
• Spinetoram demonstrated enhanced efficacy and an expanded insecticidal spectrum compared to existing spinosyns.

Conclusions:

• Artificial neural networks are effective tools for guiding the discovery of improved agrochemicals.
• Spinetoram represents a significant advancement in insecticide technology, offering superior performance while retaining a positive environmental and toxicological profile.
• The study highlights the potential of computational approaches in optimizing complex natural product derivatives for agricultural applications.