Image source: © 2025 Krish Capital Pty.Ltd

Highlights

  • Study published in Science outlines expanded protein target space for CRBN-based degraders
  • Over 100 new target classes identified using Monte Rosa’s proprietary AI/ML QuEEN™ engine
  • Research demonstrates capacity to address previously inaccessible, disease-relevant protein structures

Monte Rosa Therapeutics, Inc. (Nasdaq: GLUE) announced the publication of new discoveries featured on the cover of Science. The article, “Mining the CRBN Target Space Redefines Rules for Molecular Glue-induced Neosubstrate Recognition,” presents how Monte Rosa’s AI and machine learning engine uncovered a broad range of human proteins potentially accessible to cereblon (CRBN)-based degradation.

Monte Rosa Therapeutics is a clinical-stage biotechnology company developing molecular glue degrader (MGD) medicines for diseases in oncology, autoimmune, and inflammatory areas. Its QuEEN™ (Quantitative and Engineered Elimination of Neosubstrates) discovery engine integrates AI-guided chemistry, structural biology, and proteomics to design MGDs with high selectivity. The company holds a global license agreement with Novartis for VAV1-directed MGDs and has a collaboration with Roche focused on cancer and neurological disease targets.

The publication outlines how the company’s proprietary algorithms identified new protein surfaces capable of recruiting cereblon for targeted protein degradation. This expanded the potential reach of MGDs across more than 100 target classes, many of which are currently considered inaccessible to small molecule binding.

Monte Rosa’s approach integrates internal datasets with geometric deep learning to characterize protein surfaces and define engagement rules between protein targets, small molecules, and E3 ligases. These findings are being applied to the company’s pipeline of MGDs and ongoing clinical programs in immunology, inflammation, and oncology.

The paper details the company’s methodology, including the use of geometric deep learning to encode protein surface patches and identify structures across the proteome that can mediate protein-protein interactions. According to the publication, these discoveries were made using Monte Rosa’s proprietary AI and ML approaches.