A Pareto Algorithm for Efficient De Novo Design of Multi-functional Molecules.

Area of Science:

• Computational chemistry and cheminformatics
• Drug discovery and development
• Materials science and catalysis

Background:

• De novo molecular design aims to generate novel synthesizable molecules with desired properties.
• Optimizing multiple, often conflicting, properties simultaneously is a significant challenge in molecular design.
• Existing algorithms may struggle with generating molecules that meet stringent structural and physicochemical constraints.

Purpose of the Study:

• To introduce and describe a Pareto sorting algorithm integrated into the Synopsis de novo design program.
• To demonstrate the algorithm's efficacy in optimizing multiple properties for complex molecular design tasks.
• To address the issue of false positive hits in structure-based drug design.

Main Methods:

• Implementation of a Pareto sorting algorithm within the Synopsis de novo design framework.
• Application of the algorithm to two distinct design challenges: FGFR/VEGFR inhibitors and zeolite structure-directing agents.
• Incorporation of structural and physicochemical constraints and essential target interactions into the design process.

Main Results:

• The Pareto sorting algorithm enables simultaneous optimization of multiple molecular properties.
• Demonstrated successful application in designing dual and selective FGFR/VEGFR inhibitors.
• Achieved successful design of organic structure-determining agents (OSDAs) for zeolite synthesis.
• Significantly improved the generation of molecules with hard-to-meet constraints.

Conclusions:

• Pareto sorting enhances de novo design by enabling multi-objective optimization.
• The algorithm improves the generation of molecules with complex constraints, applicable to drug discovery and materials science.
• Integrating constraints and target interactions can mitigate false positives in de novo drug design.