Chapter 13 - Polymer Design Case Study

This chapter discusses the performance of a genetic algorithm (GA)-based approach for large-scale molecular design with the help of a large polymer design case study. A bigger problem of polymer design is presented, where the objective of considering the bigger problem is two-fold: primarily, the investigation of the efficacy of the genetic design system for problems with much larger and more complex design spaces, and second to describe the extension of the original GA framework by incorporating higher-level chemical knowledge to enable better handling of constraints, such as chemical stability and molecular complexity. The chapter introduces the large-scale polymer design problem. Results for the standard as well as for the knowledge augmented genetic design framework are presented in the chapter. Some aspects, concerning parametric sensitivity and robustness of Gas, are discussed. It was found that, despite the tremendous increase in the search space size and the complex nonlinear group interactions, the genetic design was generally able to find the target molecules. The chapter concludes by stating that the problem independent, efficient nature of the versatile genetic approach, and the ease with which chemical, biological, design, or process knowledge and constraints can be incorporated make the genetic design framework very appealing for computer-aided molecular design (CAMD) and worthy of further investigation for large-scale molecular design problems.