Universal Activation Index for Class A GPCRs.

Area of Science:

• Biochemistry and Structural Biology
• Computational Chemistry
• Pharmacology

Background:

• G-protein-coupled receptors (GPCRs) are crucial drug targets.
• Understanding GPCR activation states is vital for drug development.
• Existing methods for classifying GPCR activation states have limitations.

Purpose of the Study:

• To develop and validate a predictive model for Class A GPCR activation states.
• To utilize molecular dynamics simulations and X-ray crystallography data for model training and testing.
• To provide a tool for classifying both experimental and computational GPCR structures.

Main Methods:

• Training an index using interhelix distances from microsecond molecular-dynamics simulations.
• Testing the model on 268 published X-ray structures.
• Developing both three-class and two-state models for GPCR activation.

Main Results:

• The three-class model achieved high accuracy: 63% for active, 81% for intermediate, and 89% for inactive states.
• The two-state model showed even higher accuracy: 94% for active and 99% for inactive states.
• Intermediate structures were classified with a 2:1 ratio of active to inactive.

Conclusions:

• The developed model accurately predicts GPCR activation states from structural data.
• The model's reliance on interhelix distances aligns with known GPCR activation mechanisms.
• The model is accessible as a Python script and web tool, facilitating broader application in GPCR research.