Adaptive Space Search-based Molecular Evolution Optimization Algorithm.

Area of Science:

• Computational Chemistry
• Drug Discovery
• Bioinformatics

Background:

• Drug development heavily relies on lead compound optimization, a complex process due to vast chemical spaces.
• Current combinatorial optimization methods struggle with exploring diverse molecular structures and identifying novel drug candidates.

Purpose of the Study:

• To develop a novel algorithm for efficient and effective molecular optimization in drug discovery.
• To address limitations of existing methods in exploring chemical space and optimizing molecular properties.

Main Methods:

• Propose an adaptive space search-based molecular evolution optimization algorithm (ASSMOEA).
• Implement three modules: molecule-specific search space construction, evolutionary optimization, and adaptive space expansion.
• Utilize a fragment similarity tree and dynamic mutation for guided optimization, and an encoder-encoder structure for space expansion.

Main Results:

• ASSMOEA demonstrates superior performance in molecular optimization compared to existing methods.
• The algorithm significantly enhances the efficiency of the molecular optimization process.
• ASSMOEA shows a robust ability to explore and identify optimal solutions within complex chemical spaces.

Conclusions:

• ASSMOEA offers a more effective approach to lead compound optimization in drug discovery.
• The algorithm's adaptive search strategy improves the exploration of chemical diversity.
• This method has the potential to accelerate the identification of novel drug candidates.