GPCR-A17 MAAP: mapping modulators, agonists, and antagonists to predict the next bioactive target

Area of Science:

• Pharmacology and Bioinformatics
• Computational Drug Discovery

Background:

• G Protein-Coupled Receptors (GPCRs) are crucial for cellular signaling and represent significant drug targets.
• The GPCR-A17 subfamily is implicated in various diseases, necessitating novel therapeutic strategies.

Purpose of the Study:

• To develop an advanced machine learning model for predicting the functional roles of agonists, antagonists, and modulators in GPCR-A17 interactions.
• To accelerate the drug discovery process for GPCR-A17-related diseases.

Main Methods:

• An ensemble machine learning model, GPCR-A17 Modulator, Agonist, Antagonist Predictor (MAAP), was developed.
• MAAP integrates XGBoost, Random Forest, and LightGBM algorithms.
• The model was trained on a dataset of over 3,000 ligands and 6,900 protein-ligand interactions from multiple databases.

Main Results:

• The MAAP model demonstrated strong predictive performance with high F1 scores and AUC values on both testing and independent validation datasets.
• A Ki-filtered subset of interactions further improved predictive accuracy.
• Achieved F1 scores of 0.9179 (testing) and 0.7151 (independent validation), with AUCs of 0.9766 and 0.8591, respectively.

Conclusions:

• GPCR-A17 MAAP effectively predicts ligand interactions, significantly aiding experimental validation.
• The model accelerates the identification of potential drug candidates for GPCR-A17 targets.
• The developed tool and data are publicly available on GitHub to support further research.