Deep generative models and Artificial Intelligence (AI) have recently made strides that have demonstrated their value in the medical field, particularly in the drug discovery and development process. The developer and user must decide which protocols to take into account, which elements to carefully examine, and how deep generative models may combine the necessary disciplines in order to properly deploy AI. This study provides an updated and user-friendly reference for the large computational drug discovery and development community by summarizing traditional and recently emerging AI methodologies. From various angles, we introduce deep generative models and discuss the theoretical underpinnings of describing chemical and biological structures as well as their practical applications. We go over the data issues and technical difficulties and highlight the multimodal deep generative models potential for speeding up drug discovery.
