Chapter 14 - Generative adversarial networks in protein and ligand structure generation: a case study

Abstract

Determining and designing the structure and function of the protein has deepened our understanding of biology at a cellular and molecular level. There are numerous proteins whose structures are not known. However, prediction of protein structure is possible using amino acid sequences, if available. However, creating new protein structures in a principled and methodical way is very challenging and time-consuming. Due to the advancement in deep learning and computational modeling, exceptional results in protein generation have been achieved. It is necessary to create de novo protein designs to fully utilize the application of protein structures in technological, scientific, and medical applications. In this chapter, we have discussed the fundamental concepts of generative adversarial networks (GANs) and their applications in protein structure and ligand generation. This chapter also presents a few case studies for generating protein structures using GANs and other generative models. Further, challenges and future research directions in the area have been discussed.