Modified Electrostatic Complementary Score Function and Its Application Boundary Exploration in Drug Design.

Area of Science:

• Computational chemistry
• Drug discovery
• Machine learning applications

Background:

• Machine learning (ML) models offer rapid and accurate predictions of molecular properties, rivaling traditional quantum chemistry methods.
• Electrostatic potential (ESP) prediction is a key ML application, crucial for understanding molecular interactions.
• Electrostatic complementarity (EC) quantifies ligand-target binding, vital for efficient drug design, yet EC score functions and their scope remain underexplored.

Purpose of the Study:

• To introduce a novel electrostatic complementarity (EC) score function.
• To validate the effectiveness of the proposed EC score function.
• To analyze the applicability domain and scope of the EC score.

Main Methods:

• Development of a modified EC score function based on existing models.
• Validation of the EC score function using Pearson's R correlation coefficient.
• Investigation of the applicability domain and defining indices for the EC score.

Main Results:

• The proposed EC score function demonstrates effectiveness in quantifying molecular complementarity.
• Validation confirms the reliability of the new EC score function.
• The study provides insights into the scope and applicability of EC scores in drug design.

Conclusions:

• The novel EC score function shows promise for improving drug design efficiency.
• Understanding the applicability domain is crucial for the effective use of EC scores.
• This research contributes to the advancement of computational drug discovery tools.