Synthetic Accessibility Prediction for Small Organic Molecules
GASA (Graph Attention-based assessment of Synthetic Accessibility) is a deep neural network model for synthetic accessibility prediction based on graph attention mechanism. The server supports rapid prediction of synthetic accessibility for a large number of organic molecules, aiming to help pharmaceutical scientists to screen easy-to-synthesize molecules in drug design pipelines.

Cite GASA: Jiahui Yu, Jike Wang, Hong Zhao, Junbo Gao, Yu Kang, Dongsheng Cao, Zhe Wang, and Tingjun Hou. Organic Compound Synthetic Accessibility Prediction Based on the Graph Attention Mechanism.  J. Chem. Inf. Model. 2022, 62, 12, 2973-2986 DOI: 10.1021/acs.jcim.2c00038

Why we need synthetic accessibility prediction?
Discovering new drug candidates from the huge chemical space has always been a major challenge for medicinal chemists. As a mainstream technique in computer-aided drug design (CADD), virtual screening (VS) has been widely used to search for potential leads from large compound libraries. However, VS can only be used to search through existing molecules distributed in a limited chemical space. In contrast, de novo drug design can computationally generate novel structures from scratch with desirable therapeutic effects. Unfortunately, synthetic accessibility was rarely considered during the molecular generation process, and thus these computer-generated molecules were often hard or impossible to be synthesized. Therefore, synthetic accessibility assessment is clearly of critical importance. We proposed the GASA method for synthetic accessibility assessment of organic compounds and developed a free-access online service to help pharmaceutical scientist to quickly predict the synthetic feasibility of target molecules.
Location: Zijingang campus, Zhejiang university, 866 yuhangtang road, xihu district, hangzhou city, zhejiang province, China