AromTool: predicting aromatic stacking energy using an atomic neural network model.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Aromatic stacking interactions are crucial in protein-ligand binding.
• Accurate computational analysis of these interactions is vital for structure-based drug design.
• Existing methods may lack efficiency or precision for large-scale analysis.

Purpose of the Study:

• To develop a computational tool for analyzing aromatic stacking interactions.
• To integrate Behler-Parrinello neural networks (BPNN) for predictive modeling.
• To create an open-source Python package for benzene-containing aromatic stacking analysis.

Main Methods:

• Utilized Behler-Parrinello neural networks (BPNN) to build predictive models.
• Developed an open-source Python package named AromTool.
• Performed extensive testing and comparison with Density Functional Theory (DFT) calculations.

Main Results:

• AromTool demonstrates high precision in predicting aromatic stacking interactions.
• The tool offers excellent efficiency for high-throughput analysis.
• Achieved desirable accuracy comparable to DFT calculations.

Conclusions:

• AromTool provides a precise and efficient solution for aromatic stacking analysis.
• The package facilitates high-throughput screening in drug design.
• This tool advances computational approaches in understanding protein-ligand interactions.