Accommodating Statistical and Physics-Based Computational Protocols for Molecular Glue Model and Evaluation.

Area of Science:

• Biochemistry
• Computational Biology
• Drug Discovery

Background:

• Molecular glues (MGs) are promising for modulating protein-protein interactions (PPIs).
• High-resolution structural techniques like cryo-electron microscopy are resource-intensive, limiting rational MG design.
• Diverse mechanisms of MG action necessitate varied design strategies.

Purpose of the Study:

• To explore computational approaches for rational molecular glue design.
• To validate computational models against experimental data for specific MG systems.
• To elucidate dynamic mechanisms of molecular glue formation and guide future compound development.

Main Methods:

• Utilized AlphaFold-Multimer and molecular docking for three distinct molecular glue systems.
• Validated computational models using experimental crystal structures.
• Employed molecular dynamics simulations to capture atomic-resolution dynamics and water-mediated interactions.

Main Results:

• Established a foundation for developing more effective molecular glues through computational modeling.
• Elucidated dynamic mechanisms underlying molecular glue formation.
• Identified key factors influencing kinetic profiles and specificity, including peripheral molecular events.

Conclusions:

• Structure-based computational approaches are crucial for rationalizing molecular glue design.
• Molecular dynamics simulations provide valuable insights into dynamic interactions and can guide the design of MGs with tailored properties.
• Computational methods offer a more accessible alternative to resource-intensive techniques for MG development.