Background
Develop small-molecule inhibitors with enhanced chemical space for a popular oncology target.
Challenge
- Highly saturated intellectual property landscape with ~50 existing paftents.
- Challenging computation modeling of the pocket, as it is located at thhe junction of highly flexible protein domains.
- The surge of drug innovation is vigorous, and discovering novel compoundIs with broad chemical diversity is crucial.
- Currently, there is a lack of mature and stable Al models to explore the vast chemical space, and there are no convenient interactive tools available for practical application.
Key Results
- Generated 1 million compounds with XMolGen and narrowed odown to 30 novel candidates using XFEP
- Identified 10 active molecular scaffolds, unlocking new chemical spaces for hit and lead optimizations
