With the advancement of natural product chemical biology, many innovative bioactive molecules
have been discovered from natural products. However, many of these natural products do not
conform to Lipinski’s “Rule-of-Five,” placing them in the Beyond Rule-of-Five chemical space.
Current computer science and deep learning technologies struggle to sufficiently analyze
compounds in this Beyond Rule-of-Five space, necessitating the development of computational
techniques capable of handling more diverse chemical structures. Here, I will introduce the
research achievements in computer science technology as a compass for exploring the vast
chemical space, focusing on the latest deep learning technologies. Through examples such as
molecular design using generative AI, target-binding peptide design using AlphaFold AI, and
precise natural product derivative design using free energy perturbation simulations, we will
discuss how computer science can contribute to the discovery and functional elucidation of natural
products in the Beyond Rule-of-Five chemical space in the field of natural product chemical biology.