Multi-objective Genetic Algorithm for De Novo Drug Design (MoGADdrug)

Area of Science:

• Computational chemistry
• Medicinal chemistry
• Drug discovery

Background:

• A multi-objective genetic algorithm for De novo drug design (MoGADdrug) was developed.
• The algorithm designs novel drug-like molecules computationally, mimicking reference molecules.
• It utilizes fragments from approved drugs and fragment libraries, combined via specified rules.

Purpose of the Study:

• To present a novel multi-objective genetic algorithm (MoGADdrug) for de novo drug design.
• To enable the discovery of new drug-like molecules computationally.
• To optimize both structural similarity to reference molecules and drug-likeness.

Main Methods:

• A genetic algorithm encodes molecular fragments as variable-length genes.
• Genetic operators are applied across generations to evolve new molecular structures.
• A weighted sum approach simultaneously optimizes structural similarity and drug-likeness.

Main Results:

• The algorithm's performance was evaluated using five reference molecules: Lidocaine, Furano-pyrimidine derivative, Imatinib, Atorvastatin, and Glipizide.
• Newly designed molecules were analyzed using ZINC and PubChem databases.
• Docking investigations were performed on the designed molecules.

Conclusions:

• The MoGADdrug algorithm provides a computational approach for designing novel drug-like molecules.
• The method allows for optimization of key drug properties during the design process.
• Validation through database analysis and docking confirms the potential of the designed molecules.