Design perspectives of an evolutionary process for multi-objective molecular optimization

Abstract

Simultaneous optimization of several physiochemical properties is an important task in the drug design process. Molecule optimization formulated as optimization problems usually provide several conflicting objectives. The number of molecular properties as well as the cost-intensive methods of molecule property prediction are stringent requirements to in silico aided drug design process. Numerical approximations of the physiochemical molecule properties are challenging and in vitro methods are still essential. The objective of an in silico aided drug design process is the evolution of a multi-objective evolutionary process with the potential of a selected number of improved molecule identification within a very low iteration number for a further efficient laboratory examinations. This paper presents design perspectives of a multi-objective evolutionary process that identifies a wide variety of genetic different but selected number of optimized molecules within a low number of generations. Its search behavior is compared to NSGA-II. Furthermore, limitations of the proposed algorithm are demonstrated with regard to its performance in many-objective molecular optimization and potential concepts of adaptations for this purpose are discussed.