Development of Scoring-Assisted Generative Exploration (SAGE) and Its Application to Enzyme Inhibitor Design

In this study, an effective computational methodology named Scoring-Assisted Generative Exploration (SAGE) was devised by integrating the GEGL framework and multiple QSAR models. De novo molecular design, which involves exploring chemical space for drug-like molecules with desired properties, has been significantly advanced by deep learning techniques. Generative deep learning has revolutionized the field of de novo molecular design by enabling direct learning from input data without relying on human-made rules. This study introduces an innovative computational method, Scoring-Assisted Generative Exploration (SAGE), designed to enhance chemical diversity and optimize properties through virtual synthesis simulation, generation of bridged bicyclic rings, and application of multiple scoring models for drug-likeness. SAGE was tested on six protein targets and successfully generated high-scoring molecules within reasonable steps by optimizing for target specificity, synthetic accessibility, solubility, and metabolic stability. Additionally, SAGE identified a top-ranked molecule as a dual inhibitor of acetylcholinesterase and monoamine oxidase B, demonstrating its capability to optimize multiple properties simultaneously. These findings underscore the potential of SAGE and de novo design strategies in advancing drug discovery and development. With the ability to rapidly explore vast chemical spaces and generate novel molecules with desired properties, deep learning-based approaches like SAGE have the potential to revolutionize the field of drug discovery and development.