Directed Chemical Evolution via Navigating Molecular Encoding Space

Abstract Deep-learning techniques have significantly advanced small-molecule drug discovery. However, a critical gap remains between representation learning and small molecule generations, limiting their effectiveness in developing new drugs. We introduce Ouroboros, a unified framework that integrates molecular representation learning with generative modeling, enabling efficient chemical space exploration using pre-trained molecular encodings. By reframing molecular generation as a process of encoding space compression and decompression, Ouroboros resolves the challenges associated with iterative molecular optimization and facilitates directed chemical evolution within the encoding space. Comprehensive experimental tests demonstrate that Ouroboros significantly outperforms conventional approaches across multiple drug discovery tasks, including ligand-based virtual screening, chemical property prediction, and multi-target inhibitor design and optimization. Unlike task-specific models in traditional approaches, Ouroboros leverages a unified framework to achieve superior performance across diverse applications. Ouroboros offers a novel and highly scalable protocol for rapid chemical space exploration, fostering a potential paradigm shift in AI-driven drug discovery. Competing Interest Statement H.L., M.L., C.C., C.L. and J.Z are affiliated with DeepMed Technology (Suzhou) Co., Ltd, but this affiliation did not influence the study design, data analysis, or interpretation of results. The authors declare no other competing interests. Data availability All molecular and benchmark datasets used in this study are collected from public resources and can be downloaded from https://zhanglab.comp.nus.edu.sg/Ouroboros/ and https://github.com/Wang-Lin-boop/Ouroboros.