Identifying Potential BACE1 Inhibitors from the ChEMBL Database Using Machine Learning and Atomistic Simulation...

Area of Science:

• Computational chemistry
• Drug discovery
• Neuroscience

Background:

• Alzheimer's disease (AD) is characterized by amyloid-β (Aβ) plaque formation.
• Inhibiting β-site amyloid precursor protein-cleaving enzyme 1 (BACE1) is a key therapeutic strategy to reduce Aβ production.
• Accelerating the discovery of effective BACE1 inhibitors is crucial for developing novel AD treatments.

Purpose of the Study:

• To develop and apply a computational framework combining machine learning (ML) and atomistic simulations for rapid identification of potential BACE1 inhibitors.
• To screen a large chemical library and identify promising drug candidates for Alzheimer's disease.

Main Methods:

• Machine learning models were trained on experimental binding affinity data for BACE1 ligands.
• High-accuracy ML models were used to screen over two million compounds from the CHEMBL33 library.
• Top-hit compounds were further analyzed using molecular docking and Fast Pulling of Ligand (FPL) simulations.

Main Results:

• ML models demonstrated high predictive accuracy for ligand binding affinity.
• The screening identified a shortlist of potent BACE1 inhibitor candidates.
• FPL simulations provided detailed insights into the binding stability and interaction mechanisms of BACE1-ligand complexes.

Conclusions:

• The integrated computational approach effectively accelerates the discovery of novel BACE1 inhibitors.
• The identified compounds and binding insights can guide the rational design of next-generation Alzheimer's therapeutics.
• This strategy offers a powerful platform for developing targeted treatments for neurodegenerative diseases.