The AI-driven Drug Design (AIDD) platform: an interactive multi-parameter optimization system integrating molecular...

Area of Science:

• Computational chemistry
• Medicinal chemistry
• Drug discovery

Background:

• Computer-aided drug design (CADD) has advanced significantly, with in silico designed molecules reaching clinical trials.
• Early CADD focused on target affinity; however, pharmacokinetic and ADMET properties are crucial for successful drug development.
• Multi-parameter optimization is increasingly integrated into drug design platforms.

Purpose of the Study:

• Introduce the Artificial Intelligence-driven Drug Design (AIDD) platform.
• Demonstrate AIDD's capability to automate drug design and multi-objective optimization.
• Illustrate AIDD's application using Plasmodium falciparum dihydroorotate dehydrogenase inhibitors.

Main Methods:

• Integration of high-throughput physiologically-based pharmacokinetic simulations (GastroPlus) and ADMET predictions (ADMET Predictor).
• Utilization of an advanced evolutionary algorithm for multi-objective optimization, distinct from current generative models.
• Iterative generation of novel molecules based on activity, pharmacokinetic, and ADMET estimations.

Main Results:

• The AIDD platform successfully generates novel molecular sets with desired activity and lead-like properties.
• The workflow integrates diverse computational tools for comprehensive drug design.
• Demonstrated efficacy in generating potential drug candidates for specific targets, such as malaria.

Conclusions:

• AIDD offers an automated and efficient approach to multi-parameter drug design.
• The platform accelerates the identification of drug candidates with improved pharmacokinetic and safety profiles.
• AIDD represents a significant advancement in leveraging AI for rational drug discovery.