Proteomics Driven AI/ML Powers Rational Molecular Glue Degrader Design
1:30 pm - Wednesday 13th January 2027
- Introducing sphereOME: Leveraging deep proteomic and interactomic profiling of nearly 10,000 molecular glue degraders to map the accessible CRBN neosubstrate space across 11,000 proteins
- Harnessing sphereAI for rational degrader design: Learning the chemical features that drive selective degradation and predicting degradation potential for compound-target pairs, including targets without known degron motifs
- Closing the design loop with AI-driven optimization: Using reinforcement learning to generate novel degrader candidates that balance on-target efficacy with minimized off-target risk and improved selectivity profiles
