DeepADRA2A: predicting adrenergic α2a inhibitors using deep learning.

Area of Science:

• Pharmacology
• Computational Chemistry
• Artificial Intelligence in Drug Discovery

Background:

• Adrenergic α2a (ADRA2A) receptors regulate vital physiological functions, including blood pressure and heart rate.
• Dysregulation of ADRA2A is linked to hypertension and other cardiovascular conditions.
• Identifying ADRA2A inhibitors is crucial for treating related disorders.

Purpose of the Study:

• To develop accurate artificial intelligence (AI) models for predicting Adrenergic α2a (ADRA2A) receptor inhibitors.
• To provide a faster and more cost-effective alternative to conventional drug discovery methods.
• To expedite the identification of potential therapeutic agents targeting ADRA2A.

Main Methods:

• Employed four machine learning (ML) and deep learning (DL) algorithms.
• Utilized diverse molecular descriptors (1D, 2D, and molecular fingerprints) for model training.
• Evaluated model performance on training and test datasets.

Main Results:

• The deep learning (DL) based model exhibited superior predictive performance.
• Achieved high accuracy rates of 98.25% on the training dataset and 97.23% on the test dataset.
• Demonstrated the efficacy of DL for identifying ADRA2A inhibitors.

Conclusions:

• Deep learning models offer a powerful and efficient tool for predicting Adrenergic α2a (ADRA2A) inhibitors.
• AI-driven approaches can significantly streamline the drug discovery and development process.
• The developed model is publicly available to facilitate further research.