PRECOG: Predicting coupling probabilities of G-protein coupled receptors

PRECOG predicts coupling probabilities between Class A G-protein coupled receptors (GPCRs) and individual heterotrimeric G-proteins from receptor sequence inputs to inform receptor signaling specificity and rational mutational design.

Key Features:

• Individual G-Protein Prediction: Predicts coupling probabilities to individual heterotrimeric G-proteins rather than broad subfamily assignments.
• Sequence and Structural Feature Analysis: Utilizes sequence- and structure-based features derived from Class A GPCRs for prediction.
• Logistic Regression Model: Implements logistic regression algorithms trained on a curated dataset of Class A GPCRs.
• Input Flexibility: Accepts wild-type and mutated receptor sequences as inputs for prediction.
• Rational Mutation Design: Suggests mutations to rationally design artificial GPCRs with altered coupling properties.

Scientific Applications:

• Understanding Signal Transduction: Deciphers specific GPCR–G-protein couplings to inform analyses of cellular signaling pathways.
• Drug Discovery and Development: Aids design of targeted therapeutics by predicting and enabling modification of receptor coupling properties.
• Biological Research: Supports studies of GPCR evolution, function, and interaction dynamics.

Methodology:

PRECOG employs logistic regression trained on a curated dataset of Class A GPCRs using both sequence and structural features.