Abstract:
New drug design is a huge project, and the failure rate is very high. On the basis of the existing active compounds, the design of "me-too" drugs can increase the success rate. The lead compound can provide the same or similar spatial structure as the original active compound, and a novel skeleton. However, the skeleton of the lead compounds needn't have good physicochemical properties. Designing "me-too" drugs based on lead compounds can significantly reduce the difficulty of new drug development and improve the development efficiency and success rate. How to accurately grasp the spatial structure of the lead compounds, designing a novel structure, a reasonable valence bond, and a skeleton with the same or similar spatial structure as the lead compound are the key issues for the design of "me-too" drugs. We design and write an automated new drug skeleton design software based on the extraction of structural long chains, fragment matching and splicing: ChemCloser. The software can provide researchers with all core frameworks with reasonable structure and the same or similar spatial structure as the lead compounds. Researchers may find frameworks with novel structures, good physical properties and stable chemical properties in these core frameworks. After splicing pharmacodynamic groups on these skeletons, it can produce the same or similar biological activity as the original active compound.